Phase II Archaeological Testing and Evaluation of Six Precontact Sites in Hoke and Richmond Counties, Fort Bragg and Camp Mackall, NC.

john cable

The Directorate of Public Works, under the administration of the National Park Service, sponsored the current project to conduct archaeological testing and evaluation-related cultural resource management tasks at Fort Bragg and Camp Mackall, North Carolina (Figure 1). Palmetto Research Institute performed the Phase II testing and evaluation of six precontact archaeological sites under a Contract (C5890020435) with the National Park Service that was structured to extend up to a period of five years and to be administered in separate delivery orders The six precontact sites selected for evaluation in the DO5 package were situated in two areas (Figure 2). Three of the sites (31HK2502, 31HK2510 and 31HK2521) were located along a restricted stretch of Rock¬fish Creek, between its confluences with Piney Bottom Creek and Gum Branch, in Hoke County on the western side of the Fort Bragg Military Reservation. The remaining three sites (31RH478, 31RH480 and 31RH491) were situated within the boundaries of Camp Mackall in Richmond County. Site 31RH491 occupied a portion of the Drowning Creek terrace between Big Muddy and Aberdeen creeks. Sites 31RH478 and 31RH480 were situated in the headwaters of an upland tributary drainage of Big Muddy Creek that is now impounded by a beaver dam just above Moss Gill Lake. Sites 31HK2510 and 31RH491 were demonstrated to possess extensive in tact Middle and Early Holocene occupation surfaces containing a wide range of cultural phases and settlement types. Test unit excavations revealed that the cultural components contained within this horizon, which is situated between 20 and 50 to 60 cm bs, are affiliated with a full range of phases associated with the Middle and Early Archaic periods. Some locations within the matrix of the surface are highly reoccupied and deposits from separate occupations are mixed, while other locations exhibit nearly pure residues from single occupations. Even in highly mixed situations, however, test unit and sample block data reveal that material from individual occupations can be segregated through detailed lithic raw material analysis and the structural and organizational characteristics of the separate occupations can be made accessible for study. The importance of the Quewhiffle site (31HK2521) site is contained in the extensive Woodland period occupation. Only rarely are sites with such dense Woodland occupation encountered in the uplands of the North Carolina Sandhills. The fact that the site is situated on a ridge terrace of a major Sandhills stream, Rockfish Creek, should not be overlooked. However, the reasons for the dense Woodland occupation at this site in par¬ticular must go beyond this simple fact, as there are many sites along the creek without comparable Woodland occupation. Directly below the site is an extensive wetland created by a decreased stream gradient within a widened-out basin. Swampy and continually wet soil patches such as this would have provided an opportune microenvironment to cultivate certain of the Eastern Agricultural Complex (Ford 1985:347-349) seed crops such as sunflower (Helianthus annus), sumpweed (Iva annua var. macrocarpa), goosefoot (Chenopodium bushi¬anum), maygrass (Phalaris caroliniana), knotweed (Polygynum erectum L.), little barley (Hordeum pusillum Nutt.) and giant ragweed (Ambrosia trifida). The status of indigenous cultivation has not yet been elucidated or confirmed for the Sandhills, but sites such as Quewhiffle may supply this evidence. Numerous Late Woodland sites in the Coastal Plain and Pied¬mont have produced examples of indigenous seed crops (Scary and Scary 1997) and it is likely that these subsis¬tence patterns were common to cultures throughout the Southeast Atlantic Slope by at least the Late Woodland time horizon. If the Quewhiffle site owes it’s dense Woodland occupation to indigenous seed crop cultivation, then we should not expect it to be unique. Agriculturalists and cultivators are brought to this adaptation by population packing (Binford 1983:210-213; Netting 1968, 1990, 1993) and we can expect the Sandhills streams to be dotted with Quewhiffle-like sites at locations of expanded floodplain development if Late Woodland populations in the region were cultivating crops. Site 31HK2502, located immediately south of the Quewhiffle site, may represent another node in this Late Woodland settlement system along Rockfish Creek. A wider view of one of these linear stream settlement patterns is supplied by dense Woodland period sites along McPherson Creek and its tributaries (Cable and Cantley 2005a) on the east side of Fort Bragg.

FINAL REPORT Phase II Archaeological Testing and Evaluation of Six Precontact Sites in Hoke and Richmond Counties, Fort Bragg and Camp Mackall, NC (C5890020435-D5095060228) Palmetto Research Institute 1419 Salem Church Rd Irmo, SC 29063 September 5, 2011 Form Approved OMB No. 0704-0188 REPORT DOCUMENTATION PAGE Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 222024302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS. 1. REPORT DATE (DD-MM-YYYY) 08-18-2011 2. REPORT TYPE Final 3. DATES COVERED (From - To) July 2007 - August 2011 5a. CONTRACT NUMBER C5890020435-D5095060228 4. TITLE AND SUBTITLE Phase II Archaeological Testing and Evaluation of Six Precontact Sites in Hoke and Richmond Counties, Fort Bragg and Camp Mackall, NC. 5b. GRANT NUMBER 5c. PROGRAM ELEMENT NUMBER 6. AUTHOR(S) 5d. PROJECT NUMBER John S. Cable 5e. TASK NUMBER 5f. WORK UNIT NUMBER 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT NUMBER AND ADDRESS(ES) Palmetto Research Institute 1419 Salem Church Rd Irmo, SC 29063 9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S) Department of the Army Directorate of Public Works 2715 Reilly Road, Stop A Fort Bragg, North Carolina 28310 National Park Service Southeast Archaeological Center 2035 East Paul Dirac Drive Johnson Building, Suite 120 Tallahassee, Florida 32310 11. SPONSOR/MONITOR’S REPORT NUMBER(S) 12. DISTRIBUTION / AVAILABILITY STATEMENT Approved for Public Release 13. SUPPLEMENTARY NOTES 14. ABSTRACT The Public Works Business Center, under the administration of the National Park Service, sponsored the current project to conduct archaeological testing and evaluation-related cultural resource management tasks at Fort Bragg, North Carolina. Palmetto Research Institute performed the Phase II testing and evaluation of six precontact archaeological sites under a Contract (C5890020435) with the National Park Service that was structured to extend up to a period of five years and to be administered in separate delivery orders. This project (D5095060228) is the fifth of a series of delivery orders (DO5) to be conducted under the agreement. Phase II investigations were undertaken in compliance with the National Historic Preservation Act (Public Law 89-665, as amended by Public Law 96-515) Guidelines for Federal Agency Responsibilities, under Sections 106 and 110 of the National Historic Preservation Act, Army Regulation AR 200-1, and 36 CFR 800 (Protection of Historic and Cultural Properties). All six precontact sites represent seasonal re-occupations by a wide range of cultural groups extending over a period of at least 10,000 years. Most of these sites contained a high degree of horizontal and vertical integrity and consequently decisions concerning eligibility were based on redundancy, component representation, and the relative comprehensiveness of the data collected during Phase II investigations. Experimental field methodologies were deployed on the sites to confront the problems of low artifact density and small individual occupation sizes. Shovel testing and test unit excavations yielded data on a wide range of occupation types that will contribute to a fuller understanding of prehistoric settlement patterns and mobility strategies in the Sandhills region of North Carolina and to a broader understanding of the social and economic evolution of hunter-gatherer and tribal groups in North America. Three of the six sites were recommended eligible for inclusion on the National Register of Historic Places under criterion “d," for their scientific value. The recommended eligible sites include 31HK2510, 31HK2521 and 31RH491. Sites 31HK2502, 31RH478 and 31RH480 were recommended not eligible for inclusion on the National Register due to considerations of preservation of critical components and research redundancy. 15. SUBJECT TERMS 16. SECURITY CLASSIFICATION OF: a. REPORT U b. ABSTRACT U 17. LIMITATION OF ABSTRACT c. THIS PAGE U UU 18. NUMBER OF PAGES 638 + Front Matter & Appendices 19a. NAME OF RESPONSIBLE PERSON John S. Cable 19b. TELEPHONE NUMBER (include area code) (803) 732-6170 Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std. Z39.18 FINAL REPORT Phase II Archaeological Testing and Evaluation of Six Precontact Sites in Hoke and Richmond Counties, Fort Bragg and Camp Mackall, NC (C5890020435-D5095060228) Author: John S. Cable John S. Cable Principal Investigator Submitted to National Park Service Southeast Archeological Center 2035 East Paul Dirac Drive Johnson Building, Suite 120 Tallahassee, Florida 32310 and Department of the Army Directorate of Public Works Attn: IMSE-BRG-PWE 2715 Reilly Road, Stop A Fort Bragg, North Carolina 28310 Palmetto Research Institute 1419 Salem Church Rd Irmo, SC 29063 September 5, 2011 Table of Contents Table of Contents....................................................................................................................................................... i List of Figures........................................................................................................................................................... v List of Tables........................................................................................................................................................... xi Abstract.................................................................................................................................................................. xvi Acknowledgements..............................................................................................................................................xviii VOLUME 1 Chapter 1. Introduction................................................................................................................................. 1 Chapter 2. Environment................................................................................................................................ 5 INTRODUCTION............................................................................................................................................. 5 REGIONAL PHYSIOGRAPHY....................................................................................................................... 5 GEOLOGY AND SOILS ................................................................................................................................. 5 REGIONAL CLIMATE.................................................................................................................................... 8 BIOGEOGRAPHY............................................................................................................................................ 8 PALEOENVIRONMENT............................................................................................................................... 12 Paleoclimate............................................................................................................................................. 12 Paleovegetation........................................................................................................................................ 13 Paleofauna................................................................................................................................................ 17 Paleoshorelines......................................................................................................................................... 18 Chapter 3. Culture History......................................................................................................................... 21 PALEOINDIAN PERIOD .............................................................................................................................. 21 ARCHAIC PERIOD........................................................................................................................................ 23 WOODLAND PERIOD.................................................................................................................................. 32 MISSISSIPPIAN PERIOD.............................................................................................................................. 36 PROTOHISTORIC AND EARLY HISTORIC NATIVE AMERICANS....................................................... 37 Chapter 4. Research Design....................................................................................................................... 41 APPROACH TO RESEARCH........................................................................................................................ 41 OCCUPATION CLUSTERS........................................................................................................................... 42. Quadrat Sampling (Sample Blocks)......................................................................................................... 50 Occupation Cluster Characteristics.......................................................................................................... 52 Population Estimates................................................................................................................................ 55 REGIONAL SETTLEMENT MODELS......................................................................................................... 56 REGIONAL INTEGRATION......................................................................................................................... 57 Chronology Building................................................................................................................................ 57 Mobility and Subsistence-Settlement Systems........................................................................................ 59 ANALYSIS SYSTEMS................................................................................................................................... 62 Precontact Ceramics................................................................................................................................. 62 Lithic Artifacts.......................................................................................................................................... 65 FIELD METHODS......................................................................................................................................... 70 Previous Investigations............................................................................................................................ 70 Site Map................................................................................................................................................... 70 Site Boundary Definition.......................................................................................................................... 71 Shovel Tests.............................................................................................................................................. 71 Test Units.................................................................................................................................................. 71 i Table of Contents (Continued) Chapter 4. Research Design (Continued) FIELD METHODS (Continued)..................................................................................................................... 70 Photography............................................................................................................................................. 72 GPS Data.................................................................................................................................................. 72 NRHP EVALUATIONS.................................................................................................................................. 72 CURATION..................................................................................................................................................... 72 NATIVE AMERICAN GRAVES PROTECTION AND REPATRIATION ACT (NAGPRA)....................... 73 Chapter 5. 31HK2502................................................................................................................................... 75 PREVIOUS RESEARCH................................................................................................................................ 75 FIELDWORK OVERVIEW............................................................................................................................ 78 ARTIFACT INVENTORY.............................................................................................................................. 83 Lithic Raw Material Types....................................................................................................................... 83 Lithic Artifacts.......................................................................................................................................... 91 Precontact Ceramics............................................................................................................................... 109 OCCUPATION PATTERNS.......................................................................................................................... 119 SAMPLE BLOCK INVESTIGATIONS....................................................................................................... 129 TEST UNITS................................................................................................................................................. 140 Test Unit 1 (SB 14)................................................................................................................................. 141 Test Unit 2 (SB 3)................................................................................................................................... 143 Test Unit 3 (SB 1)................................................................................................................................... 147 EVALUATION.............................................................................................................................................. 147 Chapter 6. 31HK2510................................................................................................................................. 153 PREVIOUS RESEARCH.............................................................................................................................. 153 FIELDWORK OVERVIEW.......................................................................................................................... 156 ARTIFACT INVENTORY............................................................................................................................ 162 Lithic Raw Material Types..................................................................................................................... 162 Lithic Artifacts........................................................................................................................................ 170 Precontact Ceramics............................................................................................................................... 190 Historic Artifacts.................................................................................................................................... 192 OCCUPATION PATTERNS.......................................................................................................................... 193 SAMPLE BLOCK INVESTIGATIONS....................................................................................................... 209 TEST UNITS................................................................................................................................................. 219 Test Unit 1 and 4 (SB 4)......................................................................................................................... 220 Test Unit 2 (SB 6)................................................................................................................................... 223 Test Unit 3 (SB 9)................................................................................................................................... 226 EVALUATION.............................................................................................................................................. 230 Chapter 7. 31HK2521................................................................................................................................. 235 PREVIOUS RESEARCH.............................................................................................................................. 235 FIELDWORK OVERVIEW.......................................................................................................................... 238 ARTIFACT INVENTORY............................................................................................................................ 244 Lithic Raw Material Types..................................................................................................................... 246 Lithic Artifacts........................................................................................................................................ 251 Precontact Ceramics............................................................................................................................... 271 OCCUPATION PATTERNS.......................................................................................................................... 305 SAMPLE BLOCK INVESTIGATIONS....................................................................................................... 319 ii Table of Contents (Continued) Chapter 7. 31HK2521 (Continued) TEST UNITS................................................................................................................................................. 329 Block 1 (TUs 4 & 9)/SB8....................................................................................................................... 331 Block 2 (TUs 1, 6, 10 and 11)/SB3........................................................................................................ 343 Block 3 (TUs 2, 5, 7 and 8)/SB 16 ........................................................................................................ 348 TU3 (SB 14)........................................................................................................................................... 353 EVALUATION.............................................................................................................................................. 361 VOLUME 2 Chapter 8. 31RH478.................................................................................................................................... 363 PREVIOUS RESEARCH.............................................................................................................................. 363 FIELDWORK OVERVIEW.......................................................................................................................... 366 ARTIFACT INVENTORY............................................................................................................................ 370 Lithic Raw Material Types..................................................................................................................... 375 Lithic Artifacts........................................................................................................................................ 377 Historic Artifacts.................................................................................................................................... 381 OCCUPATION PATTERNS.......................................................................................................................... 381 SAMPLE BLOCK INVESTIGATIONS....................................................................................................... 384 Sample Block 1 (SB1)............................................................................................................................ 384 Sample Block 2 (SB2)............................................................................................................................ 387 TEST UNITS................................................................................................................................................. 390 EVALUATION.............................................................................................................................................. 398 Chapter 9. 31RH480.................................................................................................................................... 399 PREVIOUS RESEARCH.............................................................................................................................. 399 FIELDWORK OVERVIEW.......................................................................................................................... 402 ARTIFACT INVENTORY............................................................................................................................ 410 Lithic Raw Material Types..................................................................................................................... 410 Lithic Artifacts........................................................................................................................................ 414 Precontact Ceramics............................................................................................................................... 423 Historic Artifacts.................................................................................................................................... 431 OCCUPATION PATTERNS.......................................................................................................................... 431 SAMPLE BLOCK INVESTIGATIONS....................................................................................................... 444 TEST UNITS................................................................................................................................................. 450 Test Unit 1 (SB1).................................................................................................................................... 450 Test Unit 2 (SB2).................................................................................................................................... 453 Test Units 3 and 4 (SB4)........................................................................................................................ 456 EVALUATION.............................................................................................................................................. 462 Chapter 10. 31RH491.................................................................................................................................. 463 PREVIOUS RESEARCH.............................................................................................................................. 463 FIELDWORK OVERVIEW.......................................................................................................................... 468 ARTIFACT INVENTORY............................................................................................................................ 471 Lithic Raw Material Types..................................................................................................................... 471 Lithic Artifacts........................................................................................................................................ 483 Precontact Ceramics............................................................................................................................... 508 iii Table of Contents (Continued) Chapter 10. 31RH491 (Continued) Historic Artifacts.................................................................................................................................... 518 OCCUPATION PATTERNS.......................................................................................................................... 518 SAMPLE BLOCK INVESTIGATIONS....................................................................................................... 531 TEST UNITS................................................................................................................................................. 537 Test Unit 1 (SB10).................................................................................................................................. 541 Test Unit 2 (SB1).................................................................................................................................... 544 Test Unit 3 (SB6).................................................................................................................................... 546 Test Unit 4 (SB4).................................................................................................................................... 548 Test Unit 5 (SB14).................................................................................................................................. 552 EVALUATION.............................................................................................................................................. 554 Chapter 11. Synthesis.................................................................................................................................. 557 VERTICAL COHERENCE OF DEPOSITS................................................................................................. 557 OCCUPATION TYPES AND SETTLEMENT PATTERNS........................................................................ 563 Occupation Types................................................................................................................................... 563 Microenvironmental Patterns................................................................................................................. 573 Settlement and Mobility Overview........................................................................................................ 577 CULTURE HISTORIC SEQUENCING....................................................................................................... 580 FINAL COMMENTS........................................................................................................................................... 594 Chapter 12. Recommendations............................................................................................................... 595 References Cited............................................................................................................................................ 599 Appendix A: Provenience File................................................................................................................. 639 Appendix B: Artifact Inventory.............................................................................................................. 883 Appendix C: Core Data............................................................................................................................ 1009 Appendix D: Biface Data......................................................................................................................... 1115 Appendix E: Projectile Point Data...................................................................................................... 1119 Appendix F: Scraper and Retouched Flake Data......................................................................... 1125 Appendix G: Utilized Flake Data......................................................................................................... 1131 Appendix H: Ground and Pecked Stone and Cobble Tool Data............................................. 1135 Appendix I: Precontact Ceramic Data............................................................................................... 1139 Appendix J: GPS Data............................................................................................................................. 1177 Appendix K: Lithic Thin-section Report, Keith Seramur, P. G., PC.................................... 1181 iv List of Figures VOLUME 1 Chapter 1. Introduction 1. Location Map, Fort Bragg and Camp Mackall, NC...........................................................................................2 2. Location of Sites in DO5 Package (Projection: UTM NAD83, Scale: 1:300,000)............................................3 Chapter 2. Environment 3. Physiographic Regions Surrounding Fort Bragg and Camp Mackall, NC.........................................................6 4. Location of Pollen Sites Discussed in the Text Relative to Modern Vegetation Associations .......................14 Chapter 3. Culture History 5. Generalized Cultural Chronology for the Paleoindian and Archaic Periods of North Carolina (Data Drawn from Anderson et al. 1990, Sassaman et al. 1990, and Sassaman et al. 2002)...........................23 6. Ceramic Sequence for Sandhills Region of North Carolina (data derived from Herbert 2003).....................32 Chapter 4. Research Design 7. 8. 9. 10. 11. 12. 13. Historic Artifact Distributions and Components at Salt Pond Plantation, Francis Marion NF, SC.................44 Precontact Occupation Clusters at Salt Pond Plantation, Francis Marion NF, SC ..........................................45 Two-household !Kung Residential Camp of Tanagaba (from Yellen 1977)....................................................46 Simulated Reoccupations of Tanagaba-like Settlements Over Shovel Test Grids of Varying Intervals..........47 Shovel Test Pattern, 38SU136/137, Poinsett Range, South Carolina...............................................................48 Vertical Density Distributions of Precontact Sherds, 38SU136/137................................................................49 Vertical Density Distributions of Precontact Lithics, 38SU136/137...............................................................50 Chapter 5. 31HK2502 14. Location of Site 31HK2502 (McCain, NC 7.5’ USGS Quadrangle) Scale: 1:24,000......................................76 15. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31HK2502 (Scale: 1 inch = 200m, Projection: UTM, NAD83).........................................................................................77 16. Site 31HK2502, Looking East into Rockfish Creek Floodplain from Site Datum..........................................78 17. Site 31HK2502, Looking North into Rockfish Creek Floodplain from N530/E480........................................79 18. Site 31HK2502, Base Map...............................................................................................................................80 19. Site 31HK2502, Phase I Survey Sketch Map, TRC (from Grunden and Ruggiero 2006)...............................81 20. Site 31HK2502, Stage I Shovel Test Sample...................................................................................................82 21. Site 31HK2502, Stage I and II Shovel Test Sample ........................................................................................84 22. Site 31HK2502, Stage III Shovel Test Sample Showing Locations of Sample Blocks...................................85 23. Site 31HK2502, Locations of Test Units .........................................................................................................86 24. Cores and Bifaces, Site 31HK2502 .................................................................................................................94 25. Projectile Points, Site 31HK2502.....................................................................................................................98 26. Unifaces, Site 31HK2502...............................................................................................................................102 27. Miscellaneous Stone Tool Classes, Site 31HK2502.......................................................................................108 28. Anvil Stone (a1190), Site 31HK2502.............................................................................................................110 29. Hanover I and II Series Ceramics, Site 31HK2502........................................................................................114 30. Ceramic Rim Profiles, Site 31HK2502..........................................................................................................115 v List of Figures (Continued) Chapter 5. 31HK2502 (Continued) 31. 32. 33. 34. 35. 36. 37. 38. Hanover III, Cape Fear III and Yadkin III Series Ceramics, Site 31HK2502................................................118 Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31HK2502.........121 Density Distribution of Quartz Debitage, Stage I and II Shovel Test Sample, Site 31HK2502....................122 Density Distribution of Precontact Sherds, Stage I and II Shovel Test Sample, Site 31HK2502..................125 Profile Drawing, TU1, North Wall, 31HK2502..............................................................................................142 Profile Drawing, TU2, North Wall, 31HK2502..............................................................................................145 Profile Drawing, TU3, North Wall, 31HK2502..............................................................................................148 Hypothesized Type II Occupation at 31HK1214 (from Cable and Cantley 2005b)......................................150 Chapter 6. 31HK2510 39. Stage III Sample Frame Surrounding Test Unit 5, Site 31CD1065................................................................154 40. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31HK2510 (Scale: 1 inch = 200m, Projection: UTM, NAD83).......................................................................................155 41. Site 31HK2510, Looking South Down the Rockfish Creek Channel from Near N420/E540.......................156 42. Site 31HK2510, Looking South Down Dirt Access Road from N440/E520.................................................157 43. Site 31HK2510, Base Map.............................................................................................................................158. 44. Site 31HK2510, Looking Southeast Across Dirt Access Road from N440/E520 Toward Rockfish Creek Floodplain............................................................................................................................................159 45. Site 31HK2510, Phase I Survey Sketch Map, TRC (from Grunden and Ruggiero 2006).............................160 46. Site 31HK2510, Stage I Shovel Test Sample.................................................................................................161 47. Site 31HK2510, Stage I and II Shovel Test Sample.......................................................................................163 48. Site 31HK2510, Stage III Shovel Test Sample Showing Locations of Sample Blocks.................................164 49. Site 31HK2510, Locations of Test Units........................................................................................................165 50. Cores, Site 31HK2510....................................................................................................................................173 51. Bifaces, Site 31HK2510.................................................................................................................................176. 52. Projectile Points, Site 31HK2510...................................................................................................................178 53. Unifaces, Site 31HK2510...............................................................................................................................183 54. Miscellaneous Stone Tools, Site 31HK2510..................................................................................................188 55. Hammer/Anvil Stone (a288), Site 31HK2510...............................................................................................189 56. Precontact Ceramics, Site 31HK2510............................................................................................................194 57. Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31HK2510 ........196 58. Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31HK2510.........197 59. Locations of Stone Tools of Probable Early Archaic Association, 31HK2510..............................................201 60. Comparison of Soil Depth and Early Archaic Stone Tool Distributions, 31HK2510....................................202 61. Woodland Period Artifact Distributions, 31HK2510.....................................................................................204 62. Profile Drawing, TU1 and TU4, South Wall, 31HK2510...............................................................................221 63. Early Archaic Living Floor, Sample Block 4, 31HK2510.............................................................................223 64. Profile Drawing, TU2, North Wall, 31HK2510..............................................................................................224 65. Pee Dee Projectile Points, TU2, 31HK2510...................................................................................................226 66. Profile Drawing, TU3, East Wall, 31HK2510................................................................................................227 67. Hypothesized Type II Occupation at 31HK1214 (from Cable and Cantley 2005b)......................................231 vi List of Figures (Continued) Chapter 7. 31HK2521 68. Location of Site 31HK2521 (McCain, NC 7.5’ USGS Quadrangle) Scale: 1:24,000....................................236 69. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31HK2521 (Scale: 1 inch = 200m, Projection: UTM, NAD83).......................................................................................237 70. View of Site 31HK2521, Looking Northeast from a Point Southwest of the Site Boundary........................238 71. Site 31HK2521, Base Map.............................................................................................................................239 72. View of Site 31HK2521, Looking Northeast onto the Apex of Ridge from N480/E460...............................240 73. View of Site 31HK2521, Looking North Toward Rockfish Creek Floodplain from the Apex of the Ridge at N520/E500...................................................................................................................241 74. Site 31HK2521, Phase I Survey Sketch Map, TRC (from Grunden and Ruggiero 2006).............................242 75. Site 31HK2521, Stage I Shovel Test Sample ................................................................................................243 76. Site 31HK2521, Stage I and II Shovel Test Sample.......................................................................................244 77. Site 31HK2521, Stage III Shovel Tests and Sample Block Locations...........................................................245 78. Site 31HK2521, Test Unit Locations..............................................................................................................246 79. Cores, Site 31HK2521....................................................................................................................................255 80. Bifaces, Site 31HK2521.................................................................................................................................258 81. Projectile Points, Site 31HK2521 ..................................................................................................................261 82. Unifaces, Site 31HK2521 ..............................................................................................................................266 83. Grinding and Hammer Stones, Site 31HK2521.............................................................................................272 84. Hammer/Anvil Stones, Site 31HK2521.........................................................................................................273 85. Miscellaneous Stone Tools, Site 31HK2521..................................................................................................274 86. New River Series Ceramics, Site 31HK2521.................................................................................................278 87. Rim Profiles, Site 31HK2521.........................................................................................................................279 88. Hanover I Series Ceramics, Site 31HK2521..................................................................................................282 89. Hanover II Check Stamped, Site 31HK2521.................................................................................................284 90. Hanover II Cord Marked, Fabric Impressed and Incised, Site 31HK2521....................................................285 91. Hanover III Fabric Impressed, Site 31HK2521..............................................................................................288 92. Hanover IIId Fabric Impressed (p1978), Vessel 154, Site 31HK2521...........................................................289 93. Hanover IIIa Fabric Impressed Interior Finish Treatments, Site 31HK2521.................................................290 94. Hanover III Cord Marked, Site 31HK2521....................................................................................................291 95. Hanover III Cord Marked and Paddle Edge Cord Marked Interior Finish Treatments, Site 31HK2521.......292 96. Hanover III Check Stamped and Simple Stamped, Site 31HK2521..............................................................293 97. Yadkin III Fabric Impressed, Site 31HK2521................................................................................................298 98. Yadkin III Cord Marked, Site 31HK2521......................................................................................................299 99. Yadkin III Cord-Wrapped Paddle Edge Stamped, Simple Stamped, Check Stamped and Plain, Site 31HK2521...............................................................................................................................................300 100. Cape Fear Series Ceramics, Site 31HK2521..................................................................................................304 101. Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31HK2521.........309 102. Density Distribution of Quartz Debitage, Stage I and II Shovel Test Sample, Site 31HK2521....................310 103. Density Distribution of New River Series Ceramics, Stage I and II Shovel Test Sample, Site 31HK2521 ..............................................................................................................................................313 104. Density Distribution of Hanover I Series Ceramics, Stage I and II Shovel Test Sample, Site 31HK2521 ..............................................................................................................................................314 105. Density Distribution of Hanover II Series Ceramics, Stage I and II Shovel Test Sample, Site 31HK2521...............................................................................................................................................315 106. Density Distribution of Hanover III and Yadkin III Series Ceramics, Stage I and II Shovel Test Sample, Site 31HK2521.............................................................................................................316 vii List of Figures (Continued) Chapter 7. 31HK2521 (Continued) 107. Density Distribution of Cape Fear III Series Ceramics, Stage I and II Shovel Test Sample, Site 31HK2521.............................................................................................................317 108. Profile Drawing, TU4 and TU9, East Wall, 31HK2521.................................................................................338 109. Profile Drawing, TU1, TU6, TU10 and TU11, East Wall, 31HK2521...........................................................345 110. Plan View of SB16 and the Block 3 Test Units, 31HK2521..........................................................................351 111. Profile Drawing, TU2 and TU7, Block 2, North Wall, 31HK2521................................................................352 112. Locations of Intrusive Artifacts in Lower Levels, SB16, 31HK2521...........................................................353 113. Distribution of Quartz Debitage and Tools, SB16 and TU Block 3, 31HK2521...........................................354 114. Spatial Relationships of Sherd Aggregations and the Quartz Debitage Element, SB16, 31HK2521............356 115. Profile Drawing, TU3, North Wall, 31HK2521..............................................................................................359 VOLUME 2 Chapter 8. 31RH478 116. Location of Site 31RH478 (Pine Bluff, NC 7.5’ USGS Quadrangle) Scale: 1:24,000..................................364 117. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31RH478 (Scale: 1 inch = 200m, Projection: UTM, NAD83).......................................................................................365 118. View of Site 31RH478, Looking North to SB2 from N480/E500................................................................366 119. Site 31RH478, Base Map...............................................................................................................................367 120. View of Site 31RH478, Looking Southwest from Dirt Road at N520/E540.................................................368 121. Site 31RH478, Phase I Survey Sketch Map, TRC (from Ruggiero 2003).....................................................369 122. Site 31RH478, Stage I Shovel Test Sample...................................................................................................371 123. Site 31RH478, Stage I and II Shovel Test Sample.........................................................................................372 124. Site 31RH478, Stage III Shovel Test Sample and Locations of Sample Blocks............................................373 125. Site 31RH478, Test Unit Locations................................................................................................................374 126. Lithic Artifacts, Site 31RH478.......................................................................................................................380 127. Density Distribution of MMR Debitage, SB1, Site 31RH478.......................................................................385 128. Density Distribution of UER Debitage, SB1, Site 31RH478.........................................................................386 129. Density Distribution of UER Debitage, SB2, Site 31RH478.........................................................................388 130. Locations of Test Units in SB2, Site 31RH478..............................................................................................391 131. Profile Drawing, TU1 and TU4, West Wall, Site 31RH478...........................................................................393 132. Profile Drawing, TU2, North Wall, Site 31RH478.........................................................................................394 133. Profile Drawing, TU3, South Wall, Site 31RH478.........................................................................................395 134. Hypothetical Model of Multi-Household Camp, SB2, Site 31RH478...........................................................397 Chapter 9. 31RH480 135. Location of Site 31RH480 (Pine Bluff, NC 7.5’ USGS Quadrangle) Scale: 1:24,000..................................400 136. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31RH480 (Scale: 1 inch = 200m, Projection: UTM, NAD83).......................................................................................401 137. View of Site 31RH480, Looking Southeast from the Central Trail at N510/E490 Toward the Beaver Pond.................................................................................................................................402 138. Site 31RH480, Base Map...............................................................................................................................403 139. Site 31RH480, Phase I Survey Sketch Map, TRC (from Ruggiero 2003).....................................................404 viii List of Figures (Continued) Chapter 9. 31RH480 (Continued) 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. Site 31RH480, Stage I Shovel Test Sample...................................................................................................406 Site 31RH480, Stage I and II Shovel Test Samples.......................................................................................407 Site 31RH480, Stage III Shovel Test Sample and Locations of Sample Blocks............................................408 Site 31RH480, Test Unit Locations................................................................................................................409 Chipped Stone Tools, Site 31RH480..............................................................................................................418 Miscellaneous Stone Tools and Minerals, Site 31RH480..............................................................................423 New River II and Hanover II Series Ceramics, Site 31RH480......................................................................427 Rim Profiles, Site 31RH480...........................................................................................................................428 Hanover III and Yadkin III Series Ceramics, Site 31RH480..........................................................................429 Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31RH480............433 Density Distribution of Quartz Debitage, Stage I and II Shovel Test Sample, Site 31RH480.......................434 Density Distribution of New River Series Sherds, All Shovel Tests, Site 31RH480.....................................436 Density Distribution of Hanover II Series Sherds, All Shovel Tests, Site 31RH480.....................................437 Density Distribution of Hanover III, Yadkin III and Cape Fear III Series Sherds, All Shovel Tests, Site 31RH480.................................................................................................................................................438 Profile Drawing, TU1, South Wall, 31RH480................................................................................................452 Profile Drawing, TU2, West Wall, 31RH480.................................................................................................454 Profile Drawing, TU3 and TU4, South Wall, 31RH480.................................................................................457 Density Distribution of Vessel 1 Sherds, SB4, 31RH480 ..............................................................................459 Hypothesized Palmer II Type II Element, SB4, 31RH480.............................................................................460 Distribution of Biotite Crystal Fragments, SB4, 31RH480............................................................................461 Chapter 10. 31RH491 160. Location of Site 31RH491 (Pine Bluff, NC 7.5’ USGS Quadrangle) Scale: 1:24,000.................................464 161. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31RH491 (Scale: 1 inch = 200m, Projection: UTM, NAD83).......................................................................................465 162. View of Site Drowning Creek Looking North from the Bridge on General Ridgeway Road........................466 163. Site 31RH491, Base Map...............................................................................................................................467 164. View of Site 31RH491 Looking West from N510/E510 to the Wildlife Plot................................................468 165. View of Site 31RH491 Looking South from N460/E540 along Grade Cut Paralleling the Gravel Road.....469 166. View of Site 31RH491 Looking Southeast from N370/E570 toward Gun Implacement..............................470 167. View of Site 31RH491 Looking South from N520/E530 Down the Gravel Road........................................471 168. Site 31RH491, Phase I Survey Sketch Map, TRC (from Grunden and Ruggiero 2006)...............................472 169. Site 31RH491, Stage I Shovel Test Sample...................................................................................................473 170. Site 31RH491, Stage I and II Shovel Test Samples.......................................................................................474 171. Site 31RH491, Stage III Shovel Tests and Sample Block Locations.............................................................475 172. Site 31RH491, Test Unit Locations................................................................................................................476 173. Cores and Type I Bifaces, Site 31RH491.......................................................................................................485 174. Type II and III Bifaces and Biface Tools, Site 31RH491...............................................................................488 175. Projectile Points, Site 31RH491.....................................................................................................................491 176. End and Side Scrapers, Site 31RH491...........................................................................................................498 177. Pointed Scrapers and Other Retouched Tools, Site 31RH491........................................................................506 178. Miscellaneous Stone Tools, Site 31RH491....................................................................................................509 179. New River and Hanover I and II Series Ceramics, Site 31RH491.................................................................514 180. Rim Profiles, Site 31RH491...........................................................................................................................516 ix List of Figures (Continued) Chapter 10. 31RH491 (cont.) 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. Hanover III, Yadkin III and Cape Fear III Series Ceramics, Site 31RH491..................................................516 Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31RH491............521 Density Distribution of Quartz Debitage, Stage I and II Shovel Test Sample, Site 31RH491.......................522 Density Distribution of New River Series Sherds, All Shovel Tests, Site 31RH491.....................................524 Density Distribution of Hanover I and II Series Sherds, All Shovel Tests, Site 31RH491............................525 Density Distribution of Hanover III, Yadkin III and Cape Fear Series Sherds, All Shovel Tests, Site 31RH491.................................................................................................................................................526 Profile Drawing, TU1, West Wall, 31RH491.................................................................................................542 Profile Drawing, TU2, North Wall, 31RH491................................................................................................545 Profile Drawing, TU3, North Wall, 31RH491................................................................................................547 Profile Drawing, TU4, West Wall, 31RH491.................................................................................................550 Profile Drawing, TU5, East Wall, 31RH491..................................................................................................553 Chapter 11. Synthesis 192. 193. 194. 195. 196. 197. 198. 199. Two-household !Kung Residential Camp of Tanagaba (from Yellen 1977)..................................................564 Panch Pera Camp in 1961, Four Birhor Bands (Adapted from Binford 1983:141).......................................566 Hypothesized Type II Occupation at 31HK1214 (from Cable and Cantley 2005b)......................................568 Patterns of Residential Camp Movement and the Distribution of Logistical Camps and Loci (Adapted from Binford 1982:10, Figure 2)............................................................................................570 Walukaritji, Australian Aborigine Unmarried Men’s Camp (from Hayden 1979).........................................572 Frequency Histogram of Type I Residential Elements by Microenvironmental Zone and Archaeological Site........................................................................................................................................576 Frequency Histogram of Ceramic Clusters by Microenvironmental Zone and Archaeological Site ............577 Regional Map of Sites and Locations Mentioned in the Text........................................................................583 x List of Tables VOLUME 1 Chapter 4. Research Design 1. Summary of Cubic Meter Allocations by Site, Delivery Order 5...................................................................70 Chapter 5. 31HK2502 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. Total Chipped Stone Inventory, 31HK2502.....................................................................................................88 Distribution of Debitage Type by Raw Material Type, 31HK2502..................................................................92 Percent Cortex on Debitage by Raw Material Type, 31HK2502.....................................................................92 Summary Data for Cores, 31HK2502..............................................................................................................93 Summary Data for Bifaces, 31HK2502............................................................................................................95 Summary Data for Projectile Points, 31HK2502.............................................................................................97 Summary Data for Scrapers and Retouched Flakes, 31HK2502...................................................................101 Summary Data for Utilized Flakes, 31HK2502.............................................................................................107 Hanover I Ceramic Inventory, 31HK2502.....................................................................................................113 Hanover II Ceramic Inventory, 31HK2502....................................................................................................116 Hanover III Ceramic Inventory, 31HK2502...................................................................................................117 Cape Fear III and Yadkin III Ceramic Inventory, 31HK2502........................................................................117 Debitage Inventory for Stage I and II Shovel Tests, 31HK2502....................................................................123 Diagnostic Stone Tool Inventory, All Proveniences, 31HK2502...................................................................123 Precontact Ceramic Inventory, All Proveniences, 31HK2502.......................................................................124 Recognized Elements for the Stage I and II Shovel Test Sample, 31HK2502...............................................127 Vertical Distribution of Recognized Elements, 31HK2502...........................................................................128 Vertical Distribution of Debitage, Sample Blocks 1—5, 31HK2502.............................................................130 Vertical Distribution of Debitage, Sample Blocks 6—10, 31HK2502...........................................................131 Vertical Distribution of Debitage, Sample Blocks 11—15, 31HK2502.........................................................132 Vertical Distribution of Debitage, Sample Blocks 16—19, 31HK2502.........................................................133 Vertical Distribution of Stone Tools by Sample Block, 31HK2510....................................................... 134-135 Vertical Distribution of Precontact Ceramics by Sample Block, 31HK2502.................................................136 Shovel Test Outcome Model for the Purpose of Identifying Element Types.................................................137 Inferred Element Types by Sample Blocks, 31HK2502......................................................................... 138-139 Summary Data for Test Units, 31HK2502......................................................................................................141 Stone Tools and Debitage by Level for TU 1 and SB 14, 31HK2502............................................................143 Stone Tools and Debitage by Level for TU 2 and SB 3, 31HK2502..............................................................146 Precontact Ceramics from TU 2, 31HK2502.................................................................................................146 Stone Tools and Debitage by Level for TU 3 and SB 1, 31HK2502..............................................................149 Chapter 6. 31HK2510 32. 33. 34. 35. 36. 37. 38. 39. Total Chipped Stone Inventory, 31HK2510...................................................................................................166 Distribution of Debitage Type by Raw Material Type, 31HK2510................................................................171 Percent Cortex on Debitage and Core Flakes by Raw Material Type, 31HK2510........................................171 Summary Data for Cores, 31HK2510............................................................................................................172 Summary Data for Bifaces, 31HK2510..........................................................................................................175 Summary Data for Projectile Points, 31HK2510...........................................................................................177 Summary Data for Scrapers and Retouched Flakes, 31HK2510...................................................................182 Summary Data for Utilized Flakes, 31HK2510.............................................................................................187 xi List of Tables (Continued) Chapter 6. 31HK2510 (Continued) 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. Precontact Ceramic Inventory, 31HK2510.....................................................................................................193 Historic Artifact Inventory, 31HK2510..........................................................................................................194 Debitage Inventory by Level for Stage I and II Shovel Tests, 31HK2510.....................................................198 Diagnostic Archaic Stone Tool Inventory, All Proveniences, 31HK2510......................................................199 Vertical Distribution of Debitage in Mid-slope Stage I and II Shovel Tests, 31HK2510..............................199 Vertical Distribution of Debitage for Stage I and II Shovel Tests Minus the Mid-slope Sample, 31HK2510......................................................................................................................................................200 Comparison of Vertical Distribution of Debitage Between Early Archaic Localities, 31HK2510................203 Vertical Distribution of Woodland Ceramics and Projectile Points, All Proveniences, 31HK2510...............203 Recognized Elements for the Stage I and II Shovel Test Sample, 31HK2510...............................................207 Inferred Culture-Historic Associations of Recognized Lithic Elements, 31HK2510....................................208 Inferred Culture-Historic Associations of Recognized Type I Elements, 31HK2510....................................209 Vertical Distribution of Debitage, Sample Blocks 1—6, 31HK2510.............................................................210 Vertical Distribution of Debitage, Sample Blocks 7—12, 31HK2510...........................................................211 Vertical Distribution of Debitage, Sample Blocks 13—18, 31HK2510.........................................................212 Vertical Distribution of Stone Tools by Sample Block, 31HK2510....................................................... 213-214 Shovel Test Outcome Model for the Purpose of Identifying Element Types.................................................215 Inferred Element Types by Sample Blocks, 31HK2510......................................................................... 217-218 Summary Data for Test Units, 31HK2510......................................................................................................219 Stone Tools and Debitage by Level for TUs 1 and 4 and SB 4, 31HK2510.................................................222 Stone Tools and Debitage by Level for TU 2 and SB 6, 31HK2510..............................................................225 Stone Tools and Debitage by Level for TU 3 and SB 9, 31HK2510..............................................................228 Debitage Characteristics of Type I Elements Composed of Type I RT Material, 31HK2510........................229 Chapter 7. 31HK2521 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. Lithic Raw Material Frequencies by Chipped Stone Tool Class, 31HK2521................................................248 Distribution of Debitage Type by Raw Material Type, 31HK2521................................................................252 Percent Cortex on Debitage and Percent Core Flakes by Raw Material Type, 31HK2521...........................253 Summary Data for Cores, 31HK2521............................................................................................................254 Summary Data for Bifaces, 31HK2521..........................................................................................................257 Summary Data for Projectile Points, 31HK2521...........................................................................................259 Summary Data for Scrapers and Retouched Flakes, 31HK2521...................................................................265 Summary Data for Utilized Flakes, 31HK2521.............................................................................................270 New River Ceramic Inventory, 31HK2521....................................................................................................277 Hanover I Ceramic Inventory, 31HK2521.....................................................................................................281 Hanover II Ceramic Inventory, 31HK2521....................................................................................................283 Percentage of Surface Treatment Types by Hanover II Paste Subvariants, 31HK2521.................................286 Hanover III Ceramic Inventory, 31HK2521...................................................................................................287 Percentage of Surface Treatment Types by Hanover III Paste Subvariants, 31HK2521...............................294 Yadkin III Ceramic Inventory, 31HK2521.....................................................................................................297 Percentage of Surface Treatment Types by Yadkin III Paste Subvariants, 31HK2521..................................301 Cape Fear III Ceramic Inventory, 31HK2521................................................................................................303 Contingency Table of Sand Density and Grain Size, Cape Fear III Sample, 31HK2521..............................305 Percentage of Exterior Surface Treatment Types by Ceramic Series Variant, 31HK2521.............................307 Debitage Inventory by Level for Stage I and II Shovel Tests, 31HK2521.....................................................311 xii List of Tables (Continued) Chapter 7. 31HK2521 (Continued) 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. Vertical Distribution of Ceramic Sub-series, All Proveniences, 31HK2521..................................................311 Diagnostic Stone Tool Inventory, All Proveniences, 31HK2521...................................................................312 Recognized Elements for the Stage I and II Shovel Test Sample, 31HK2521...............................................318 Ceramic Vessel Representation by Ceramic Sub-series for Stage I and II Shovel Tests, 31HK2521............320 Inferred Culture-Historic Associations of Recognized Lithic Elements, 31HK2521....................................321 Inferred Culture-Historic Associations of Recognized Type I Elements, 31HK2521....................................322 Vertical Distribution of Debitage, Sample Blocks 1—5, 31HK2521.............................................................323 Vertical Distribution of Debitage, Sample Blocks 6—9, 31HK2521.............................................................324 Vertical Distribution of Debitage, Sample Blocks 10—14, 31HK2521.........................................................325 Vertical Distribution of Debitage, Sample Blocks 15—18, 31HK2521.........................................................326 Vertical Distribution of Stone Tools by Sample Block, 31HK2521....................................................... 327-328 Vertical Distribution of Ceramic Types in Sample Blocks, 31HK2521................................................. 329-333 Shovel Test Outcome Model for the Purpose of Identifying Element Types.................................................333 Inferred Element Types by Sample Blocks, 31HK2521......................................................................... 334-336 Summary Data for Test Units, 31HK2521.....................................................................................................337 Stone Tools and Debitage by Level for SB 8 and Block 1, 31HK2521.........................................................340 Vertical Distribution of Ceramics in SB 8 and Block 1, 31HK2521..............................................................341 Partial Completeness Percentages for Vessels in SB 8 and Block 1, 31HK2521...........................................342 Stone Tools and Debitage by Level for Block 2, 31HK2521.........................................................................346 Vertical Distribution of Ceramics Block 2 Test Units, 31HK2521................................................................347 Partial Completeness Percentages for Vessels in SB 3 and Block 2, 31HK2521...........................................349 Stone Tools and Debitage by Level for SB 16 and Block 3, 31HK2521.......................................................355 Vertical Distribution of Ceramics in Test Units of Block 3, 31HK2521........................................................357 Partial Completeness Percentages for Vessels in SB 16 and Block 3, 31HK2521.........................................358 Stone Tools and Debitage by Level for SB 14 and TU 3, 31HK2521............................................................360 Vertical Distribution of Ceramics in SB 14 and TU 3, 31HK2521................................................................360 Partial Completeness Percentages for Vessels in SB 14 and TU 3, 31HK2521.............................................361 VOLUME 2 Chapter 8. 31RH478 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. Lithic Raw Material Frequencies by Chipped Stone Tool Class, 31RH478..................................................376 Weight and Frequency Distributions of Debitage Type by Raw Material Type, 31RH478...........................378 Percent Cortex on Debitage and Percent Core Flakes by Raw Material Type, 31RH478.............................379 Vertical Distribution of Chipped Stone Artifacts, Stage I and II Shovel Tests, 31RH478.............................382 Vertical Distribution of Element Types, 31RH478.........................................................................................384 Vertical Distribution of Chipped Stone Artifacts in SB 1, 31RH478.............................................................385 Shovel Test Outcome Model for the Purpose of Identifying Element Types.................................................386 Inferred Element Types in SB 1, 31RH478....................................................................................................387 Vertical Distribution of Chipped Stone Artifacts in SB 2, 31RH478.............................................................388 Vertical Distribution of Debitage for the UER Elements, SB 2, 31RH478....................................................389 Inferred Element Types in SB 2, 31RH478....................................................................................................389 Debitage Reduction Profiles for Elements in SB 2, 31RH478.......................................................................390 Summary Data for Test Units, 31RH478........................................................................................................392 xiii List of Tables (Continued) Chapter 8. 31RH478 (Continued) 122. Vertical Distribution of Chipped Stone Artifacts in Test Units, 31RH478.....................................................396 123. Debitage Reduction Profiles for Elements in SB 2 and Associated Test Units, 31RH478.............................397 Chapter 9. 31RH480 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. Lithic Raw Material Frequencies by Chipped Stone Tool Class, 31RH480..................................................411 Distribution of Debitage Type by Raw Material Type, 31RH480..................................................................415 Percent Cortex on Debitage and Percent Core Flakes by Raw Material Type, 31RH480.............................416 Summary Data for Cores, 31RH480..............................................................................................................417 Summary Data for Bifaces, 31RH480............................................................................................................419 Summary Data for Scrapers and Retouched Flakes, 31RH480......................................................................420 Summary Data for Utilized Flakes, 31RH480...............................................................................................422 Precontact Ceramic Sherd and Vessel Representation, 31RH480..................................................................425 Vertical Distribution of Chipped Stone Artifacts, Stage I and II Shovel Tests, 31RH480.............................435 Vertical Distribution of Precontact Ceramics, All Proveniences, 31RH480..................................................435 Recognized Elements for the Stage I and II Shovel Test Sample, 31RH480.................................................441 Representation of Vessels by Sub-Series, Stage I and II Shovel Tests, 31RH480.........................................442 Inferred Culture-Historic Associations of Recognized Lithic Elements, 31RH480.......................................443 Vertical Distribution of Debitage, Sample Blocks 1—4, 31RH480...............................................................445 Vertical Distribution of Debitage, Sample Blocks 5—8, 31RH480...............................................................446 Vertical Distribution of Ceramic Types in Sample Blocks, 31RH480...........................................................447 Shovel Test Outcome Model for the Purpose of Identifying Element Types.................................................447 Inferred Element Types by Sample Blocks, 31RH480........................................................................... 448-449 Summary Data for Test Units, 31RH480........................................................................................................450 Stone Tools and Debitage by Level for TU 1, 31RH480...............................................................................453 Debitage Reduction Profiles by Raw Material Type, TU 1 and SB 1, 31RH480...........................................453 Stone Tools and Debitage by Level for TU 2, 31RH480...............................................................................455 Debitage Reduction Profiles by Raw Material Type, TU 2 and SB 2, 31RH480...........................................455 Vertical Artifact Distributions, TU 3 and TU 4, 31RH480.............................................................................458 Debitage Reduction Profiles for USR (1) , TUs 3 and 4 and SB 4, 31RH480...............................................461 Chapter 10. 31RH491 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. Lithic Raw Material Frequencies by Chipped Stone Tool Class, 31RH491..................................................478 Distribution of Debitage Type by Raw Material Type, 31RH491..................................................................482 Percent Cortex on Debitage and Percent Core Flakes by Raw Material Type, 31RH491.............................482 Summary Data for Cores, 31RH491..............................................................................................................484 Summary Data for Bifaces, 31RH491............................................................................................................486 Summary Data for Projectile Points, 31RH491.............................................................................................490 Summary Data for End Scrapers, 31RH491...................................................................................................497 Summary Data for Side Scrapers, 31RH491..................................................................................................502 Vertical Distribution of Scrapers and Retouched Flakes, 31RH491..............................................................503 Summary Data for Other Scrapers and Retouched Flakes, 31RH491............................................................505 Summary Data for Utilized Flakes, 31RH491...............................................................................................508 Precontact Ceramic Sherd and Vessel Representation, 31RH491..................................................................510 Sherd Counts and Weights for Ceramic Types, 31RH491..............................................................................511 xiv List of Tables (Continued) Chapter 10. 31RH491 (Continued) 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. Historic Artifact Inventory, 31RH491............................................................................................................519 Vertical Distribution of Debitage by Raw Material Type, Stage I and II Shovel Tests, 31RH491................523 Vertical Distribution of Precontact Ceramics, All Shovel Tests, 31RH491...................................................523 Recognized Elements for the Stage I and II Shovel Test Sample, 31RH491.................................................528 Representation of Ceramic Vessels by Sub-series, Stage I and II Shovel Tests, 31RH491...........................530 Inferred Culture-Historic Associations of Recognized Lithic Elements, 31RH491.......................................531 Inferred Culture-Historic Associations of Recognized Type I Elements, 31RH491......................................531 Vertical Distribution of Debitage by Sample Block, 31RH491............................................................. 533-535 Vertical Distribution of Ceramic Types in Sample Blocks, 31RH491...........................................................536 Shovel Test Outcome Model for the Purpose of Identifying Element Types.................................................536 Inferred Element Types by Sample Blocks, 31RH491........................................................................... 538-540 Summary Data for Test Units, 31RH491........................................................................................................541 Stone Tools and Debitage by Level for TU 1, 31RH491...............................................................................543 Technological Profiles for Debitage, TU 1 and SB 10, 31RH491..................................................................544 Ceramic and Lithic Artifact Inventory by Level, TU 2, 31RH491.................................................................546 Lithic Artifact Inventory by Level, TU 3, 31RH491......................................................................................549 Technological Profiles for Debitage, TU 3 and SB 6, 31RH491....................................................................549 Lithic Artifact Inventory by Level, TU 4, 31RH491......................................................................................551 Technological Profiles for Debitage, TU 3 and SB 6, 31RH491....................................................................552 Lithic Artifact Inventory by Level, TU 6, 31RH491......................................................................................554 Technological Profiles for Debitage, TU 6 and SB 14, 31RH491..................................................................555 Chapter 11. Synthesis 183. Comparison of the Vertical Distributions of Lithic and Ceramic Artifacts by Delivery Order......................559 184. Vertical Distribution of Diagnostic Stone Tools from Stage I and II Shovel Tests, All Delivery Orders.....560 185. Percentage Data for the Vertical Distributions of Diagnostic Stone Tools by Cultural Period for Stage I and II Shovel Tests, All Delivery Orders.....................................................................................560 186. Vertical Distribution of Scrapers from Shovel Tests, All Delivery Orders....................................................561 187. Vertical Distribution of Early Archaic Projectile Points from Shovel Tests, All Delivery Orders.................561 188. Comparison of Occupation Characteristics by Site, All Delivery Orders......................................................574 189. Distribution of Ceramic Clusters Identified in Stage I Shovel Tests by Site, All Delivery Orders................575 190. Comparison of Projected to Expected Sherd Weight Based on Vessel Populations for Sites in the Delivery Order 5 Package......................................................................................................580 191. Percentage of Exterior Surface Treatment Types by Ceramic Series Variant, 31HK2521.............................581 192. Comparison of Exterior Surface Treatment Types by Middle Woodland Paste Variants at Sites 38SU136/137 and 38SU141, Poinsett Range, SC (Data from Cable 2002)..........................................586 193. Hanover III and Yadkin III Exterior Surface Treatment Percentages, 31HK2521.........................................586 194. Comparison of McLean Mound and Quewhiffle Site Arenite-tempered Surface Treatments.......................588 xv Abstract The Directorate of Public Works, under the administration of the National Park Service, sponsored the current project to conduct archaeological testing and evaluation-related cultural resource management tasks at Fort Bragg and Camp Mackall, North Carolina (Figure 1). Palmetto Research Institute performed the Phase II testing and evaluation of six precontact archaeological sites under a Contract (C5890020435) with the National Park Service that was structured to extend up to a period of five years and to be administered in separate delivery orders. This project (D5095060228) is the fifth of a series of delivery orders (DO5) to be conducted under the agreement. Phase II investigations were undertaken in compliance with the National Historic Preservation Act (Public Law 89-665, as amended by Public Law 96-515) Guidelines for Federal Agency Responsibilities, under Sections 106 and 110 of the National Historic Preservation Act, Army Regulation AR 200-4, and 36 CFR 800 (Protection of Historic and Cultural Properties). Fieldwork extended from July 17 to November 28, 2007. John Cable served as Principal Investigator and Field Director for the project and he was assisted in the field by Ray Talley. Crew that worked on the project at various times consisted of George Price, Graham Elmore, Geoffry Lipscomb, Neal Sexton and Kevin Enlow. Fieldwork consumed 306 person days. The six precontact sites selected for evaluation in the DO5 package were situated in two areas (Figure 2). Three of the sites (31HK2502, 31HK2510 and 31HK2521) were located along a restricted stretch of Rockfish Creek, between its confluences with Piney Bottom Creek and Gum Branch, in Hoke County on the western side of the Fort Bragg Military Reservation. The remaining three sites (31RH478, 31RH480 and 31RH491) were situated within the boundaries of Camp Mackall in Richmond County. Site 31RH491 occupied a portion of the Drowning Creek terrace between Big Muddy and Aberdeen creeks. Sites 31RH478 and 31RH480 were situated in the headwaters of an upland tributary drainage of Big Muddy Creek that is now impounded by a beaver dam just above Moss Gill Lake. Phase II investigations resulted in recommending three of these sites for inclusion on the NRHP under criterion “d,” based on their scientific importance in accordance with 36CFR60.4. These are 31HK2510, 31HK2521 and 31RH491. The remaining three sites were recommended not eligible. This group included sites 31HK2502, 31RH478 and 31RH480. Although each of these latter sites were demonstrated to contain in tact cultural deposits, the information potential of their archaeological records was judged redundant to that of the recommended eligible sites, where this information was more accessible due to lower degrees of deposit mixing and/or the presence of a wider range of components. Sites 31HK2510 and 31RH491 were demonstrated to possess extensive in tact Middle and Early Holocene occupation surfaces containing a wide range of cultural phases and settlement types. Test unit excavations revealed that the cultural components contained within this horizon, which is situated between 20 and 50 to 60 cm bs, are affiliated with a full range of phases associated with the Middle and Early Archaic periods. Some locations within the matrix of the surface are highly reoccupied and deposits from separate occupations are mixed, while other locations exhibit nearly pure residues from single occupations. Even in highly mixed situations, however, test unit and sample block data reveal that material from individual occupations can be segregated through detailed lithic raw material analysis and the structural and organizational characteristics of the separate occupations can be made accessible for study. The importance of the Quewhiffle site (31HK2521) site is contained in the extensive Woodland period occupation. Only rarely are sites with such dense Woodland occupation encountered in the uplands of the North xvi Carolina Sandhills. The fact that the site is situated on a ridge terrace of a major Sandhills stream, Rockfish Creek, should not be overlooked. However, the reasons for the dense Woodland occupation at this site in particular must go beyond this simple fact, as there are many sites along the creek without comparable Woodland occupation. Directly below the site is an extensive wetland created by a decreased stream gradient within a widened-out basin. Swampy and continually wet soil patches such as this would have provided an opportune microenvironment to cultivate certain of the Eastern Agricultural Complex (Ford 1985:347-349) seed crops such as sunflower (Helianthus annus), sumpweed (Iva annua var. macrocarpa), goosefoot (Chenopodium bushianum), maygrass (Phalaris caroliniana), knotweed (Polygynum erectum L.), little barley (Hordeum pusillum Nutt.) and giant ragweed (Ambrosia trifida). The status of indigenous cultivation has not yet been elucidated or confirmed for the Sandhills, but sites such as Quewhiffle may supply this evidence. Numerous Late Woodland sites in the Coastal Plain and Piedmont have produced examples of indigenous seed crops (Scary and Scary 1997) and it is likely that these subsistence patterns were common to cultures throughout the Southeast Atlantic Slope by at least the Late Woodland time horizon. If the Quewhiffle site owes it’s dense Woodland occupation to indigenous seed crop cultivation, then we should not expect it to be unique. Agriculturalists and cultivators are brought to this adaptation by population packing (Binford 1983:210-213; Netting 1968, 1990, 1993) and we can expect the Sandhills streams to be dotted with Quewhiffle-like sites at locations of expanded floodplain development if Late Woodland populations in the region were cultivating crops. Site 31HK2502, located immediately south of the Quewhiffle site, may represent another node in this Late Woodland settlement system along Rockfish Creek. A wider view of one of these linear stream settlement patterns is supplied by dense Woodland period sites along McPherson Creek and its tributaries (Cable and Cantley 2005a) on the east side of Fort Bragg. Management Summary Table, DO 5 SUMMARY TABLE: DO5 SITES GPS: DIFFERENTIAL CORRECTIONS SITE NAD27 COMPONENT IDENTIFICATIONS NAD83 RECOMMENDATION NORTHING EASTING NORTHING EASTING 31HK2502 3883961 655482 3884144 655499 Taylor; Palmer; Morrow Mountain; Guilford; Early, Middle and Late Woodland Not Eligible 31HK2510 3883400 655596 3883614 655613 Dalton; Palmer; LeCroy; Morrow Mountain; Early, Middle and Late Woodland Eligible 31HK2521 3884323 655114 3884533 655135 Possible Paleoindian; Possible Hardaway-Dalton; Palmer; Morrow Mountain; Early, Middle and Late Woodland Eligible 31RH478 3877686 637990 3877893 638009 Archaic Not Eligible 31RH480 3878011 637751 3878226 637763 Palmer; Early, Middle and Late Woodland Not Eligible 31RH491 3878606 639654 3878835 639664 Possible Hardaway-Dalton; Taylor; Palmer; Kirk/Stanly Stemmed; Morrow Mountain, Guilford; Savannah River; Early, Middle and Late Woodland Eligible xvii Acknowledgements In an ongoing project of this scale, the efforts of many individuals are key to its successful completion. Three different National Park Service COTRs (Contract Officer’s Technical Representatives) have given their time and counsel since 2002. These include Dr. David G. Anderson, now an anthropology professor at the University of Tennessee, Emily Yates, who provided vital support along the way, and Dr. Michael Russo, who has steered the final stages of the project. The author is indebted to each of these individuals for their important contributions. Equally important have been the contributions of the personnel of the Cultural Resources Program. at Fort Bragg. Dr. Linda Carnes-McNaughton, Jeff Irwin, Dr. Joe Herbert and Charles Heath have provided cogent critiques and suggestions for the design of fieldwork and interpretations of the data. Joe Herbert’s advice and commentray were of especially helpful throughout the course of the project. The efforts and assistance of Stacy Culpepper and Christopher Moore, formerly of the Fort Bragg Cultural Resource Program, are also greatly appreciated. Finally, the author would like to extend his warm gratitude and thanks to his old friend, Ray Talley, who provided crucial assistance and supervision. His exceptional commitment to the quality of the fieldwork and his toil and sweat were crucial contributions to the ultimate success of this long-term project. xviii Chapter 1. Introduction The Directorate of Public Works, under the administration of the National Park Service, sponsored the current project to conduct archaeological testing and evaluation-related cultural resource management tasks at Fort Bragg and Camp Mackall, North Carolina (Figure 1). Palmetto Research Institute performed the Phase II testing and evaluation of six precontact archaeological sites under a Contract (C5890020435) with the National Park Service that was structured to extend up to a period of five years and to be administered in separate delivery orders. This project (D5095060228) is the fifth of a series of delivery orders (DO5) to be conducted under the agreement. Phase II investigations were undertaken in compliance with the National Historic Preservation Act (Public Law 89-665, as amended by Public Law 96-515) Guidelines for Federal Agency Responsibilities, under Sections 106 and 110 of the National Historic Preservation Act, Army Regulation AR 200-1, and 36 CFR 800 (Protection of Historic and Cultural Properties). Fieldwork extended from July 17 to November 28, 2007. John Cable served as Principal Investigator and Field Director for the project and he was assisted in the field by Ray Talley. Crew that worked on the project at various times consisted of George Price, Graham Elmore, Geoffry Lipscomb, Neal Sexton and Kevin Enlow. Fieldwork consumed 306 person days. The six precontact sites selected for evaluation in the DO5 package were situated in two areas (Figure 2). Three of the sites (31HK2502, 31HK2510 and 31HK2521) were located along a restricted stretch of Rockfish Creek, between its confluences with Piney Bottom Creek and Gum Branch, in Hoke County on the western side of the Fort Bragg Military Reservation. The remaining three sites (31RH478, 31RH480 and 31RH491) were situated within the boundaries of Camp Mackall in Richmond County. Site 31RH491 occupied a portion of the Drowning Creek terrace between Big Muddy and Aberdeen creeks. Sites 31RH478 and 31RH480 were situated in the headwaters of an upland tributary drainage of Big Muddy Creek that is now impounded by a beaver dam to form a northern extension of Moss Gill Lake. The six archaeological sites selected for evaluation in the DO5 package are situated in upland environments at the heads or along the courses of tributary creeks that ultimately drain into Cape Fear River. These locales were the focus of extensive reoccupation by mobile and sometimes seasonally sedentary hunter-gatherer and horticultural forager groups, spanning the full reach of the modern Holocene geological epoch, which began about 12,000 years ago. Although individual campsites were small, normally covering less than 25 m2, the resulting phenomena that we refer to as archaeological sites in these cases often achieve immense size due to the repeated accumulation of generally unrelated occupations. Larger scale patterns of adaptation and regional settlement organization are to be found in these palimpsest accumulations of human history, but archaeologists are challenged to recover this information in a systematic and coherent manner due to the extreme dispersion of individual elements of each culture-historic period. The report is divided into 12 chapters including this introduction. Chapter 2 presents an environmental overview of the Sandhills region surrounding Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 1 2 GREENE 83˚ WILKES MCDUFFIE LAURENS CALDWELL SALUDA NEWBERRY UNION CHEROKEE 82˚ EDGEFIELD AIKEN ALEXANDER YORK GASTON LEXINGTON FAIRFIELD CHESTER LINCOLN CATAWBA CLEVELAND BURKE e SOUTH CAROLINA GREENWOOD COLUMBIA LINCOLN MCCORMICK ABBEVILLE ANDERSON SPARTANBURG POLK g RUTHERFORD MCDOWELL i d Greenville GREENVILLE HENDERSON e R s AVERY n PATRICK FLOYD IREDELL 81˚ CALHOUN RICHLAND SUMTER Columbia KERSHAW LANCASTER UNION RANDOLPH ANSON 80˚ CHATHAM LEE ORANGE MARLBORO WILLIAMSBURG FLORENCE MARION SCOTLAND GEORGETOWN DILLON JOHNSTON Raleigh WAKE FRANKLIN HORRY pe Ca ar R Fe SAMPSON BRUNSWICK BLADEN COLUMBUS CUMBERLAND Fayetteville HARNETT 79˚ ROBESON HOKE GRANVILLE VANCE HALIFAX GREENE 78˚ CAPE FEAR . ONSLOW Ne u PITT rR JONES Ta EDGECOMBE LENOIR NEW HANOVER NORTHAMPTON R oano 77˚ se lico Rive r PAMLICO Pam 77˚ DARE 100 KM 76˚ O c e a n 100 Mi CAPE HATTERAS Parallel scale at 35˚N 0˚E CAPE LOOKOUT Pamlico Sound HYDE TYRRELL Albemarle Sound CURRITUCK VIRGINIA BEACH Virginia Beach A t l a n t i c 0 0 PASQUOTANK CAMDEN WASHINGTON BEAUFORT CARTERET erCRAVEN Riv MARTIN BERTIE 76˚ Norfolk CHESAPEAKE CHOWAN PERQUIMANS GATES SUFFOLK Chesapeake ISLE OF WIGHT Portsmouth HERTFORD SOUTHAMPTON SUSSEX GREENSVILLE DINWIDDIE PENDER DUPLIN WAYNE WILSON NASH BRUNSWICK 78˚ WARREN LUNENBURG MECKLENBURG CHARLOTTE Durham DURHAM PERSON HALIFAX 79˚ FORT BRAGG MOORE RICHMOND DARLINGTON CHESTERFIELD CLARENDON ALAMANCE NORTH CAROLINA CAMP MACKALL LEE Charlotte CASWELL PITTSYLVANIA Greensboro GUILFORD ROCKINGHAM HENRY MONTGOMERY DAVIDSON STANLY ROWAN CABARRUS DAVIE FORSYTH STOKES Winston-Salem YADKIN SURRY 80˚ FRANKLIN VIRGINIA CARROLL PULASKI MECKLENBURG WILKES ALLEGHANY GRAYSON WYTHE 81˚ ee R. Figure 1. Location Map, Fort Bragg and Camp Mackall, NC. MORGAN OGLETHORPE ELBERT HART l u YANCEY a t MITCHELL BUNCOMBE B n UNICOI i CARTER ASHE WATAUGA JOHNSON WASHINGTON SULLIVAN er Riv SMYTH Pee D 84˚ WALTON OCONEE MADISON FRANKLIN STEPHENS OCONEE PICKENS TRANSYLVANIA . ROCKDALE BANKS JACKSON BARROW HALL HABERSHAM MADISON Asheville HAYWOOD n JACKSON RABUN a u WASHINGTON ton ls Ho 82˚ RUSSELL Brood R. Atlanta DE KALB FORSYTH DAWSON WHITE TOWNS CLAY SWAIN MACON c i M o hR DOUGLAS CHEROKEE PICKENS a GRAHAM l h SEVIER COCKE GREENE anna Sav COBB LUMPKIN UNION CHEROKEE p a BLOUNT GEORGIA GILMER FANNIN A p MONROE LOUDON Knoxville JEFFERSON HAMBLEN HAWKINS SCOTT WISE Great 34˚ BARTOW GORDON KNOX er iv hR HANCOCK LEE nc Cli 83˚ HARLAN . POLK MCMINN ROANE ANDERSON GRAINGER CLAIBORNE BELL LESLIE eR MURRAY RHEA CAMPBELL WHITLEY KNOX TENNESSEE UNION SCOTT MCCREARY MORGAN 84˚ ok an Ro . R ke 35˚ 36˚ WAYNE LAUREL KENTUCKY PULASKI 34˚ 35˚ 36˚ Chapter 1. Introduction Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC Chapter 1. Introduction 3890000 3900000 680000 670000 ek 650000 CAMP MACKALL am r D e Be a v Cr Mud d y B ig 31RH478 640000 alf C m 660000 Roc k r ow ni 31RH491 Q 31RH480 R iver L it tle 31HK2502 31HK2510 Gum Br Fie l d Br D 3880000 Low 31HK2521 e A Pi ney Bo C tt o 650000 k ee Cr Cr u 660000 FORT BRAGG ng e wh iffl r Br Cr Cr berd ee n Hor se 640000 e Cr 3870000 670000 680000 690000 3880000 690000 3870000 er h fis Cr 3890000 3900000 Figure 2. Location of Sites in DO5 Package (Projection: UTM NAD83, Scale: 1:300,000). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 3 Chapter 1. Introduction Fort Bragg. Chapter 3 presents an overview of the precontact culture history of the region. The research design and general field and analytical methodologies are described in Chapter 4. This is followed by descriptions of the results of the field investigations conducted at each site (Chapters 5 through 10). Chapter 11 presents a synthetic interpretation of the results, while the final chapter, Chapter 12, reviews the evaluations of each site and presents suggested data recovery plans for those sites evaluated eligible for inclusion on the National Register of Historic Places. Bibliographic references are presented at the back of the volume. Finally, data appendices are provided in an accompanying volume and on a data CD attached to each report. 4 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC Chapter 2. Environment INTRODUCTION Ethnohistoric accounts of early historic Indian groups indicate that the inter-riverine upland zones of the Coastal Plain region were occupied only on a seasonal or temporary basis and primarily served as locations for gathering nuts and plant material as well as hunting game (Jones 1978; Waddell 1980; Cable and Cantley 1998; Cantley and Cable 2002). Archeological investigations within this zone tend to suggest that this pattern of land use typified the earlier precontact period of occupation as well. Of course, it is not unlikely that the specific settlement and subsistence strategies of the groups that utilized this zone varied over time as shifts in overall adaptational patterns occurred, and it is through the definition and analysis of these subtle variations in land use that the archeology of this zone can contribute to a broader understanding of the character of precontact human settlement in the Sandhills and inner Coastal Plain of North Carolina. REGIONAL PHYSIOGRAPHY Fort Bragg is located in the Sandhills physiographic province that forms a narrow band of xeric, sandy uplands stretching from Virginia south/southwest to Texas. In North Carolina, the Sandhills region ranges from approximately 5–40 miles in width, and is bounded by the Piedmont/Fall Line to the west and the Coastal Plain to the east (Figure 3). The region is characterized by gently undulating topography with well-defined, broad ridges with gentle slopes in upland locations and more broken terrain along streams (Colby 1958:27; Hudson 1984:2). It slopes toward the south and east with elevations ranging from a low of 270 feet AMSL to its highest point of 527 feet AMSL near McCain, in Hoke County (Hudson 1984:2). Fort Bragg sits on a watershed divide with its northern half feeding into the Lower Little River drainage, and its southern half drained by tributaries of Rockfish Creek. Both the Lower Little River and Rockfish Creek flow eastward until they join the Cape Fear River, which eventually empties into the Atlantic Ocean near Wilmington, North Carolina. The Sandhills region and its coarse sandy uplands acts as a natural aquifer creating abundant underground sources of water (Colby 1958:27). Percolation of rainfall is often interrupted by underlying argyllic or clayey horizons that funnel the water laterally, creating numerous springs or seepages on side slopes that provide the origin of many first order streams in the project area (Benson 1998:6; Clement et al. 1997). Previous studies (Braley and Schuldenrein 1993; Clement et al. 1997; Loftfield 1979) have documented the importance of springs or seepages to precontact settlement in the region. GEOLOGY AND SOILS Located east of the Fall Line, the bedrock in the Sandhills is composed of volcanic slate that is encountered at depths of 60 to 122 meters below Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 5 34˚ 35˚ 76˚ 79˚ ee R. 80˚ s ill L. Murray 81˚ Columbia n Sa 82˚ L. Greenwood Clark Hill L. . n L. Keowee 83˚ hR 84˚ L. Sidney Lanier 34˚ 35˚ Carters L. Atlanta Blue Ridge L. Allatoona L. GEORGIA Hiwassee L. L. Ocoee Watts Bar L. A p p a Nottely L. Chatuge L. aFontana L. l Brenau L. L. Burton a i h c Fort Loudoun L. 36˚ Secession L. Hartwell Res. L. Jocassee Asheville M o Douglas L. Knoxville Cherokee L. TENNESSEE L. Marion dH Wateree L. SOUTH CAROLINA Greenville L. William C Bowen e l u B n u er Riv nch Cli anna Sav Figure 3. Physiographic Regions Surrounding Fort Bragg and Camp Mackall, NC. Pee D Great Camp Mackall Fall Line L. Wylie L. Tillery L. Lure L. James t Ho Norris L. 78˚ L. Waccamaw . kR ac R Bl ar Fe pe Ca r ve Ri Fort Bragg L. Hickory a Boone L. i R n i d s g e South Holston L. r ive nR lsto r Rive ell Pow KENTUCKY 77˚ 0 CAPE FEAR Coastal Plain Fayetteville uth So L. Norman Charlotte Piedmont High Rock L. Badin L. Greensboro Winston-Salem High Point Philpott L. VIRGINIA NORTH CAROLINA Raleigh Durham Roxboro L. 79˚ 81˚ 82˚ 83˚ L. Cumberland 84˚ A t l a n t i c 100 KM CAPE LOOKOUT 0 Ne . rR Ta R oano Roanoke Rapids L. John H. Kerr Res. 78˚ Meherrin River u se eR ok an Ro . R ke Brood R. 6 O c e a n 100 Miles CAPE HATTERAS r . Riv e way R. Notto 80˚ Pamlico Sound L. Mattamuskeet Pam lico Rive r Pungo L. Phelps L. New L. Albemarle Sound Virginia Beach L. Drummond Chesapeake 77˚ Portsmouth Norfolk 76˚ 36˚ Chapter 2. Environment ground surface (Hudson 1984:2). Overlying this bedrock are Cretaceous period (135–65 million years ago or mya) cross-bedded sand containing clay balls and iron cemented concretions or ferricretes, sandstone, and mudstone (Bartlett 1967; USDA 1984:2; North Carolina Geological Survey 1985) attributed to the Cape Fear and upper Middendorf formations. The Cape Fear formation deposit is a non-marine delta formation, which is often exposed along entrenched streams in the project area (Sohl and Owens 1991:191–192, 220). The relationship between the Cape Fear and underlying bedrock can be seen near Erwin, North Carolina where the Cape Fear River has cut into the underlying crystalline, metamorphic basement, and contains numerous small boulders and gravel. Further downstream, these gravel decrease in size and abundance and the formation consists of interbedded clays and sands (Sohl and Owens 1991:192). In contrast to the Cape Fear formation sands, Middendorf sands and gravels are more frequently exposed along valley slopes and eroded ridges (Barlett 1967). Dating to the later Tertiary period (65-2 mya), the Middendorf age material is thought to have been transported by streams and rivers originating in the Piedmont and subsequently deposited in the project area. The upper layer of the Middendorf deposit (also referred to as the Tuscaloosa formation by some researchers) contains small quartz gravels, which could have been utilized by precontact groups. Idol and Pullins (2001:21) noted the presence of pebble-sized gravels along the side slopes of ridges, but failed to find definitive evidence for their use as a raw material for the production of stone tools. Observations made by this author of firebreaks cut along the side and toe slopes of some ridges noted the presence of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC Chapter 2. Environment pebbles and gravels or cobbles measuring up to 10 cm in diameter. This observation in addition to the recovery of a relative large number of cortical flakes and cobble fragments on sites above these firebreaks suggest that material from this geologic formation was exploited when possible. West of Fort Bragg is the Piedmont Physiographic Province that consists of a series of northeast/ southwest-trending lithotectonic belts. The belts are defined by the age and metamorphic grades of rock found in each belt and are named, from east to west, the Carolina Slate Belt, Charlotte Belt, Kings Mountain Belt, and the Inner Piedmont Belt (Overstreet and Bell 1965). This study is concerned primarily with the Carolina Slate Belt. Generally, each lithotectonic belt contains rocks of higher-grade metamorphism near the belt’s center and rocks of lesser metamorphic grades along their perimeters. In some cases, geologic discontinuities (faults) mark the boundary between two belts. Scientific dating of the belts by K-Ar, Rb-Sr, and 40Ar/39 methods indicate that the regional metamorphic events that created the belts occurred at different times and in some belts, at more than one time between 510-265 million years ago (mya) (Bulter 1991:127, 129). Diabase dikes that originating from later periods of volcanic activity, (195-205 mya) also are distributed throughout the belts. The Carolina Slate Belt is dated to the Taconic orogeny (500-450 mya) and consists of sedimentary and volcanic rocks that were subjected to low-grade regional metamorphism and intrusive igneous dikes. Sedimentary rocks found in this belt occur in the Fall Line district and vary from strongly bedded and finely laminated to massively bedded deposits of graywacke, siltstone, sandy siltstone, and fine sandstone. The predominate rock type found in this belt is finely laminated slate (metamorphosed siltstone) which also has been misidentified as volcanic slate, shale, mudstone, argillite, and siltstone. Quartzite (metamorphosed sandstone), conglomerate (sedimentary), and limestone (sedimentary) occur less frequently in this belt. In addition to the sedimentary and metamorphic rocks, both flow and pyroclastic types of volcanic rocks occur in the Carolina Slate Belt. These rocks are interbedded with the metamorphic and sedimentary rock formations and include such types as felsic tuff, welded vitric tuff, breccia, flow banded rhyolite, porphyritic rhyolite, plain rhyolite, and basalt (Novick 1978:427–431). Although these latter rock types comprise much of the precontact lithic assemblages found in the Sandhills region and at Fort Bragg, most known outcrops and quarry sites occur in the south-central Piedmont province, a distance of approximately 60 km (40 miles). At present, the closest known outcrop of metavolcanic bedrock is located 15 km west at the Moore/Hoke County line (Braley 1990; North Carolina Department of Conservation and Development 1958), in the vicinity of Camp Mackall. Another possible source of this material could exist along the ancient terraces and floodplain of the Cape Fear River. The river could have cut through older cobble-laden delta deposits and/or moved cobbles to levee or point bar locations accessible to precontact groups. At present, systematic studies of the Cape Fear valley and its potential as a raw material source area, has gone unrealized. Other rock types found in the Carolina Slate Belt that have archeological importance include intrusive igneous dikes, and other rock types associated with dike formation or fault alteration that include rocks of siliceous and mafic origin. Intrusive igneous dikes are common along the eastern boundary of the belt near the Fall Line. These dikes contain olivine diabases of mafic origin (Ragland 1991:174) and vary in width from 0.9-30 m (3-100 ft) (Sundelius 1970:363). South of Fort Bragg in Chesterfield County, South Carolina, abundant quartz veins occur along the Pageland fault zone where the zone crosscuts argillites and metavolcanic rocks of the South Carolina slate belt (Luce and Bell 1981:9). Possibly of archaeological importance, geological studies conducted along the Pageland Fault zone indicate that siliceous rocks and/or quartz veins occur where the fault trace crosses areas consisting of argillites and other metavolcanic derived bedrock. Conversely, in areas were the fault zone crosses non-metavolcanic bedrock, siliceous rocks and quartz rarely occur as large veins. If this observation is true for other areas within the slate belt then future studies of dikes or fault zones may prove valuable for predicting the possible locations of high-grade, siliceous raw materials that were highly sought after for stone tool manufacture (Abbott et al. 2001, Cantley 2000). Blaney, Gilead, and Lakeland soil associations that are found on broad areas of nearly level and Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 7 Chapter 2. Environment gently sloping topography characterize the project area. These soils are differentiated by minor differences in color hues for their respective profiles that distinguish between the individual soil types comprising the associations. All of the sites included in the present project are located on areas mapped as Blaney loamy sand, Candor sand, and Vaucluse loamy sand. The Blaney loamy sand (BaB) profile provided in the soil survey is summarized as follows. The surface horizon is typically a plowed or disturbed A stratum 0 to 10-cm below surface (cmbs) thick consisting of dark grayish brown (10YR 4/2) loamy sand with weak fine friable granular structure. It is then underlain by an E horizon, or an older and depleted A horizon, which contains 10 to 63-cmbs of light yellowish brown (2.5Y 6/4) loamy sand, single grained. Substrata horizons vary between 63 and 85-cmbs, labeled as the Bt1 horizon – brownish yellow (10YR 6/6) sandy clay loam, and the Bt2 horizon that is characterized as reddish-yellow (7.5YR 6/6) sandy clam loam, which occurs between 85 and 155-cmbs. These substrata horizons feature weak, medium-to-coarse subangular blocky structures that are firm yet brittle in places (Hudson 1984). Candor sand (CaB) is similar to the Blaney soil mapping series and contains an Ap horizon, which is described as a dark grayish brown (10YR 4/2) sand seen from 0 to 23-cmbs (Hudson 1984). An older E horizon, characterized as a yellowishbrown (10YR 5/4) sand ranges from 23 to 50-cmbs. Underlying this zone, a Bt horizon, which is noted as yellowish-brown (10YR 5/6) loamy sand, is evident from 50 to 75-cmbs. Substrata zones are depicted as an E’1 horizon characterized as brownish yellow (10YR 6/6) single grained sand occurring between 75 and 83-cmbs and an E’2 horizon [also noted as a brownish yellow (10YR 6/6) sand] occurring from 83 to 150-cmbs. Other horizons include a B’t horizon identified as a strong brown (7.5YR 5/6) sandy clay loam occurring between 150 and 200-cmbs, and an underlying BC horizon when present. The undulating horizons have developed from residuum processes where stratums are buried from the ridge crest downward to the side slopes. Vaucluse loamy sand (VaB) also belongs to the Blaney-Gilead-Lakeland mapping unit (Hudson 1984). This series consists of well drained soils and 8 is characterized as an A horizon from 0 to 10-cmbs thick consisting of dark grayish brown (10YR 4/2) loamy sand with weak fine friable granular structure. This is underlain by an E horizon 10 to 23-cmbs of a yellowish brown (10YR 5/4) loamy sand stratum that contains some A horizon characteristics intermixed such as fine quartz grains and a friable granular structure. Various Bt horizons are represented in this unit. The Bt1 horizon consists of a yellowish red (5YR 5/6) sandy clay loam occurring between 23 and 63-cmbs. The Bt2 zone that is described as a yellowish red (5YR 5/8) sandy clay loam with distinct red (2.5YR 4/8) and yellowish brown (10YR 5/8) mottled sand inclusions occurring between 63 and 93-cmbs. Finally, the Bt3 horizon or a red (2.5YR 4/8) sandy clay loam occurs between 93 and 125-cmbs. Like Candor sand, the Vaucluse series also derives its underlying horizons from loamy sediments formed on upland slopes. REGIONAL CLIMATE Relatively short, cold winters and hot, humid summers typify the modern climate of the study area. Average daily temperature for the winter in Fayetteville is 44 degrees Fahrenheit with an average minimum of 31 degrees. Winter precipitation originates from continental fronts out of the north and west. Summers are dominated by warm, moist, tropical air masses, and precipitation during this season is generally produced by convection storms. The average daily temperature for the summer in Fayetteville is 78 degrees with an average maximum of 89 degrees. Precipitation is distributed nearly even throughout the year with 60 percent of the rainfall between the months of April and September. Average annual precipitation is equal to 43 inches, of which, approximately 3 inches falls as snow (Hudson 1984:2). Spring usually represents the driest season, but rare drought conditions can occur in the fall. BIOGEOGRAPHY Within the project vicinity, three different ecosystems are known to exist. The most prominent of these in the immediate project vicinity is the Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC Chapter 2. Environment upland forest communities generally assignable to oak-pine (Braun 1950), longleaf pine-wiregrass, and loblolly-shortleaf pine associations. These upland communities are concentrated on the tops and side slopes of ridge systems. Secondly, swamp or wetland communities form within the project vicinity at poorly drained locations such as river bottoms and in the uplands near seeps or springs. Freshwater stream environments constitute the third ecosystem, which is confined to river and tributary channels. From the perspective of precontact subsistence, the interriverine uplands of the Coastal Plain have been characterized as a perpetual “food-poor” pine barren, dominated by long-leaf pine forest with very low species diversity (Larson 1970, 1980, Milanich 1971). Reconstructing pre-European forest distributions, however, is a difficult task due to the great successional impacts of historic (and also precontact) landuse and, consequently, much controversy exists concerning the composition and distribution of “pristine” climax vegetation in the Southeastern United States (Delcourt and Delcourt 1977, 1987; Quarterman and Keever 1962; Shelford 1963). Quarterman and Keever (1962) have argued that the current closed canopy loblolly-shortleaf pine dominated forests of the Coastal Plain are the product of modern forestry management practices and other types of historic landuse, and that these forests are replaced by a Southern mixed hardwood climax when allowed to mature. Nevertheless, given the abundance of sub-climax soil conditions (eg. saturation), it is probable that natural forests would have resembled a mosaic of mixed hardwood and pine associations prior to the nineteenth century (Brooks and Canouts 1984:10-13, Widmer 1976:9). Bartram’s description of the interior Coastal Plain along the Savannah River in the late eighteenth century conforms well to this reconstruction (Harper 1958:19-20). Sub-climax conditions also appear to have been fostered by forest fires, which tend to interrupt normal successional processes. An important and once abundant community that is maintained principally by fire is the longleaf pine-wiregrass association (Platt et al. 1988:491, Russo et al. 1993). It is estimated that this association has been reduced by as much as 98 percent since presettlement times because of modern forestry practices (Bennett and Pittman 1991, Croker 1987, Noss 1989). In its pristine state this community consists of homogeneous and scattered stands of mature longleaf pine intermixed with occasional oaks and dense clumps of young pines. The understory is composed of low-lying forbs, shrubs, and grasses. Wiregrass is identified as a dominant in this community because of its consistent association with longleaf pine in old-growth tracts, a factor brought about by its own dependence on fires for reproduction. Forestry studies contradicting earlier assumptions indicate that longleaf pine-wiregrass associations are actually characterized by high species diversity rather than ecological homogeneity (Frost et al. 1986). The role of Native Americans in perpetuating and fostering longleaf pine forests and savannahs through controlled burning has also been appreciated for some time (Platt et al. 1988, Robbins and Myers 1989). Ethnohistoric accounts indicate that a popular form of surround hunting employed by Southeastern aboriginal groups involved the use of fire lines of several miles in extent set in the dried detritus of the forest floor (see accounts by Bartram, Calderon, DuPratz, Lawson, and Smith in Swanton 1946:319320). Such practices would have regularly removed the young seedlings of climax species, preventing them from maturing at a normal rate. In combination with other land modification involving the clearing of forest for settlements and agricultural fields, aboriginal land-use practices not only perpetuated sub-climax forests, but also created pine parklands or savannas. Widmer’s (1976) model for reconstructing the pre-settlement (pre-European) vegetation of the interior uplands is useful when examining the Coastal Plain’s forest structure. He identified three “pristine” subsystems, including the southern mixed hardwood forest, the longleaf pine forest and pine savannas. The latter two represent subclimax communities owing their existence to both natural and cultural causes, while the former constituted the mature climax vegetation of pre-settlement times. Upland communities were primarily restricted to the ancient terraces and ridges where soils were drier. Pine-savannas, however, were a specialized community associated with aboriginal swidden or field-rotation agriculture and were primarily confined to well-drained bottomland and stream terraces. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 9 Chapter 2. Environment In the Inner Coastal Plain region, the mixed hardwood subsystem is most likely composed of two basic community types in the vicinity of the project area today: 1) mesic slope hardwoods, and 2) upland mesic hardwoods. These two communities appear to approximate the normal range of variability associated with the mixed hardwood subsystem on the Coastal Plain of North Carolina. The structure and composition of the mesic slope hardwood communities correspond closely with Braun’s (1950) mixed mesophytic forest type. In the lowlands of the Southeast such communities typically occur on dissected riverbluffs, ravines and high bottomland where edaphic conditions are moist but well drained. This community has also been referred to as “beech ravine” (Kohlsaat 1974), “ravine slope” (Hartshorn 1972), or “bluff and slope forest” (Wharton 1978) in more locally based studies. Dominants in the North Carolina mesic slope hardwoods communities consist of beech, bull bay, laurel oak, red maple, black gum, tulip tree, sweet gum, and loblolly pine. The upland mesic hardwoods community corresponds to Braun’s (1950) “oak-hickory forest” type and represents the climax vegetation of the Coastal Plain according to Quarterman and Keever (1962). Dominants of this community, which tend to occupy the majority of the area on ridge tops, consist of beech, laurel oak, bull bay, white oak, sweet gum, mockernut hickory, water oak, southern red oak, pignut hickory, and black gum. The long leaf pine subsystem occurs in xeric, well-drained, sandy locations, seasonally flooded landforms and in mesic situations where fire has interrupted but not inhibited successional processes. (Bennett and Pittman 1991, Platt et al. 1988, Sandifer et al. 1980:439, Noss 1989:211). Fire-maintained stands of long leaf pine may contain only a two-tiered structure including a canopy of predominantly longleaf pine and a limited herbaceous layer composed of such commonly abundant species as wiregrass, ported nut rush, camphorweed, beggar ticks, panic grass, broom-straw, bracken fern, aster, goat’s rue, and thoroughwort. In the successional phase of development, however, these forests are generally three-tiered, containing in addition a tall shrub layer. Other dominants common to both areas include immature pines and hardwoods, wiregrass, bitter gallberry, running oak, stagger bush, blueberry, and huckleberry. The successional type eventually develops into mixed pine and pine-mixed hardwood communities. In these communities long leaf pine is often replaced by slash, loblolly, and short-leaf pine species. These successional types were not as common in precontact times, but the intensity of land modification was probably sufficient to perpetuate these associations in one form or another in restricted patches. Unfortunately, very little is known about the pine-savanna subsystem. Lawson (Lefler 1967:34) provides a description of one large patch of savanna adjacent to a Congeree settlement in 1701: ...about Noon, we pass’d by several fair Savanna’s, very rich and dry; seeing great Copses of many Acres that bore nothing but Bushes, about the bigness of Box-trees; which (in the Season) afford great Quantities of small Black-berrys....Hard by the Savanna’s we found the Town....The Town consists not of above a dozen Houses, they having other stragling Plantations up and down the Country, and are seated upon a small Branch of Santee River. Their Place hath curious dry Marshes, and Savanna’s adjoining to it, and would prove an exceeding thriving Range for Cattle, and Hogs.... Lawson’s use of the term plantations conveys the impression that much of the river valley margin of each of the tribes he described was punctuated with these clearings, or savannas, and that some patches were planted while the majority were unattended. The Santee Indian plantations, for instance, were described by Lawson as “lying scattering here and there, for a great many Miles” (Lefler 1967:24-25). The presence of bushes and briers on the Congeree savannas, moreover, suggests that the abandoned fields may have been maintained within a fallow rotation, as the early successional stage evidenced by this scrub vegetation would have been replaced by immature pines and hardwoods within 5-20 years after abandonment of the field (Odum 1971:261). Undoubtedly, other successional stages of pine forest were also present along these river bottoms and terraces, reflecting yet earlier concentrations of aboriginal farming communities. Odum’s (1960) study of “old field” succession is probably a useful analog with which to model these bottomland savannas. In the initial stage Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 10 Chapter 2. Environment of succession the open field is colonized by forbes and grasses over a period of two years. By the third year, sedges and shrubs begin to dominate and over a period of three to 20 years shrubs and immature trees replace the grasses and forbes. Young pine forests are established after about 25 years, and between about 75 and 100 years the mature pine forest is replaced by hardwoods under optimal climax conditions. Fauna of the inter-riverine uplands reflects a typical terrestrial forest assemblage. Because of greater mobility, however, the distribution of the member species can rarely be limited to a specific forest type, habitat, or even a particular ecosystem. The pine-mixed hardwood and mixed hardwood communities contain the greatest abundance and diversity of terrestrial faunal species of the upland ecosystem communities. This has been detailed most for avian species (see Johnston and Odum 1956), but it also holds true for all other classes as well. At the base of the faunal food chain is a class of animals including nematodes, arthropods, and myriapods, that spend all or portions of their lives within the soil matrix of the forest (Kevan 1968). Some of the more prevalent species of soil fauna in the region are nematodes, mites, springtails, and earthworms. A diverse assemblage of insects resides in these forests. Some of the more common species include mosquitos, flies, midges, wasps, bees, sawflies, grasshoppers, butterflies, moths, termites, dragonflies, mantids, crickets, cockroaches, katydids, cicadas, trips, aphids, and pine beetles (Sandifer et al. 1980:453-455). Amphibians and reptiles generally occupy moist habitats within the uplands such as leaf-litter, burrows, and temporary pools, and feed on soil fauna and insects. Numerous salamanders, hylid frogs or treefrogs, and toads dominate the amphibious fauna, while a wide array of lizards and snakes comprise the majority of the reptile species. Turtles are rare in the upland ecosystem, and are generally represented by only the eastern box turtle in the project area region of North Carolina. The most common lizards include the green anole, ground skink, six-lined racerunner, and the eastern five-lined skink. A group of small snakes occupy the leaflitter habitat. The eastern scarlet snake, mole king snake, brown snake, northern redbelly snake, southeastern crown snake, eastern coral snake, pine woods snake, and the scarlet king snake tend to occur in this habitat in pine domi- nated communities. A number of larger snakes are less specific to habitat and include the southern black racer, corn snake, yellow rat snake, eastern hognose snake, southern hognose snake, eastern kingsnake, eastern coachwhip, and the eastern garter snake. Vipers tend to inhabit hardwood communities and the more common species of viper in the North Carolina Coastal Plain include the southern copperhead, cottonmouth water moccasin, pygmy rattle snake, and canebrake rattle snake. Avian species tend to occupy very specialized niches in the forest and as such their habitat and forest associations tend to be better defined than species of the other faunal groups. Pine forests exhibit the lowest bird densities and species diversity. Only thirteen dominant species are listed for this forest type by Sandifer et al.(1980:465) including one large predator, the screech owl, and a series of primarily insectivorous birds including the red-bellied woodpecker, eastern wood pewee, southern crested flycatcher, the Carolina chickadee, the brown-headed nuthatch, the eastern bluebird, two warblers, summer tanager, and Bachman’s sparrow. The ground-feeding bobwhite and the common crow complete the list of dominants. Vultures, several species of hawk, numerous additional insectivores, and turkey comprise minor components of the avian assemblage. Thirty-two avian species are considered dominant in upland pine-mixed hardwood and mixed hardwood communities (Sandifer et al. 1980:469-470). The overall structure of this list, however, is very similar to the one produced for the pine communities. The screech owl remains the single large predator and insectivore species are the most abundant. Three species of woodpecker (ie. pileated, red-bellied, and downy), the blue jay, the mourning dove, the Carolina chickadee, the Carolina wren, the common crow, the hermit thrush, the tufted titmouse, the robin, the catbird, the blue-gray gnatcatcher, the cardinal, and various species of vireos, warblers, and sparrows comprise the list of dominants. Numerous additional moderately important and minor species are also listed including various hawks, vultures, owls, insectivores, and the turkey. Dominant mammalian herbivores of the Upland forests of the Coastal Plain consist of whitetailed deer, squirrels, the eastern wood rat, and the cotton mouse. The oposum and raccoon comprise the Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 11 Chapter 2. Environment dominant omnivores, while major carnivores include the gray and red fox, the striped skunk, the shorttailed shrew, the long-tailed weasel, the bobcat, and the black bear (Sandifer et al. 1980:472-478). Presettlement assemblages also included cougar, gray wolf, and possibly minor numbers of elk and bison (Penny 1950). Mammalian species generally do not occupy overly specialized niches and they can range across very large areas. Deer, however, tend to aggregate in hardwood patches where browse and nut mast is more plentiful. Very few species would have occupied the pine-savanna patches on a permanent basis, but such communities would have provided an important “edge”-type feeding source for mammalian herbivores and omnivores, and also predatory avian and reptilian species hunting for rodents and lagomorphs (Odum 1960). The primary mammalian dominants of old field communities in the region today consist of the eastern cottontail, cotton rat, eastern mole, least shrew, and the striped skunk (Sandifer et al. 1980:472473). The marsh rabbit also extends its range into such locations when feeding pressures increase in the swamps. The white-tailed deer, raccoon, and oposum are nocturnal visitors to such patches to feed, and are generally accompanied by most of the major mammalian predators of the upland forest. PALEOENVIRONMENT The eastern United States has undergone rather dramatic environmental change since the beginning of human occupation in the New World, which can only be confidently extended back to about 12,000 B.P. (see Dincauze 1984; Haynes et al. 1984; Kelly and Todd 1988; Lynch 1990; Meltzer 1989). The Meadowcroft Rockshelter investigations in Pennsylvania as well as the excavations at Cactus Hill in Virginia may increase this time depth to as much as 14,000 to 16,000 B.P. (Adovasio et al. 1990), but claims of much earlier occupation during the Pleistocene glacial period are today discounted by many researchers (Butzer 1988). The changes that occurred during this time frame, of course, were global in scale, and were associated with the termination of the last Pleistocene glaciation, generally referred to as the Wisconsin Glaciation in North American Quaternary stratigraphy. In this section we will examine the available evidence for paleoclimatic, paleovegeta- tional, paleofaunal, and paleoshoreline change during the terminal Pleistocene and Holocene. Paleoclimate The Wisconsin glaciation began about 115,000 B.P. and was characterized by a process of fluctuating expansion that reached its maximum extent at about 18,000 B.P. By the end of the Wisconsin, the Laurentide ice sheet of eastern North America had migrated as far south as southern Indiana and Ohio, which exerted a powerful influence on atmospheric circulation, depressing and weakening the force of summer monsoons (Gates 1976; Kutzbach et al. 1998; Webb and Kutzbach 1998). Climatic conditions were much colder and drier than today and plant species were depressed considerably south of their present ranges (Whitehead 1973). Recent palynological analysis of three wetland locations on Fort Bragg and Camp MacKall indicate that that prairielike vegetation characterized the region in the last interstadial, prior to about 20,000 B.P. (Goman and Leigh 2003). A series of environmental changes beginning around 14,000 B.P. are now well documented in the paleoenvironmental record, and provide evidence for a major climatic warming trend which eventually ushered in the Holocene, or modern, period at about 10,000 B.P. (Hare 1976). The major continental ice masses began to retreat, ocean fronts shifted poleward, the area of sea ice contracted, sea level rose, and certain middle latitude lakes became desiccated (Kutzbach 1983). The reasons for these dramatic changes are not yet well understood, but simulation models of global atmospheric circulation suggest that cyclical, long-period variation in the earth’s orbital parameters may explain the alternating reoccurrence of ice ages and interglacial periods (Andrews 1973; Gates 1976; Hare 1976; Hays et al. 1976; Kutzbach 1983; Kutzbach et al. 1998; Otto-Bliesner et al. 1982). By 10,000 B.P., global ice volumes reached minimum levels, the North American ice sheets were gone, most plant species had reached the poleward limits of their migrations, and modern circulation patterns were established (Kutzbach 1983; Wendland 1978). The date of 10,000 B.P. is rather widely accepted as the beginning of the Holocene because of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 12 Chapter 2. Environment these factors (Davis 1976; 1983; Watts 1983; Wright 1978). The period between 14,000 B.P. and 10,000 B.P. is viewed as transitional between the Late Wisconsin full-glacial and the Holocene and is commonly referred to as the Late-Glacial period of the Late Wisconsin (Watts 1980a, 1983). It is inferred from sedimentation patterns and vegetational associations that the period was cooler, but also wetter, than today, while the early Holocene marks a period of warming and drying conditions (Davis 1983, Watts 1983). A rather steep gradient of warming temperatures is hypothesized for the early and middle Holocene, with maximum summer radiation peaking between 7,000 B.P. and 5,000 B.P. when temperatures averaged 2° C to 3° C higher than today. This climatic optimum corresponds to the hypsithermal or altithermal episode, which was continental in scale and possibly timetransgressive by latitude (Wright 1976). After the climatic optimum, temperatures appear to have gradually cooled, although they remained above modern levels until the Little Ice Age, dated between A. D. 1450 and A. D. 1850 (Davis 1983). Paleovegetation These climatic changes had a profound effect on the biogeographic structure and composition of plant and animal communities throughout the world. In the southeastern United States the trajectory of change moved from a full-glacial vegetation of pinespruce parkland, through a wide range of mesic deciduous forest assemblages during the late-glacial period, and an oak-dominated deciduous forest peaking during the hypsithermal (Davis 1983, Watts 1983). These changes were time-transgressive, with analogous shifts in assemblage composition occurring earlier at more southerly latitudes and at lower elevations. Detailed mapping of dated pollen spectra from paleoenvironmental sites throughout the eastern United States has recently demonstrated that the poleward migration of plant species during the late-glacial-early Holocene warming trend was accomplished on an individual basis and migration rates differed according to seed dispersal patterns, tolerance ranges of individual species, and the locations of refugia (Davis and Botkin 1985; Delcourt and Delcourt 1987; Webb 1987, 1988). Consequently, the modern forest types, which represent “true” Holocene climax associations, did not appear until the trajectory of climatic and environmental change had stabilized in the middle Holocene. Prior to this period the palynological record is characterized by a bewildering array of ephemeral species associations, varying along latitude and elevation gradients, which have no modern analogs and are best referred to as vegetational assemblages rather than forest types or formations. The modern vegetational patterns of the Southeast, in fact, did not emerge until the hypsithermal episode, when sea level began to stabilize and high water tables allowed the expansion of swamps and the establishment of the pine-dominated Southern Coniferous Forest along the Atlantic Coastal Plain (Watts 1980a; Webb 1987, 1988). Principal fossil pollen sites (Figure 4) used to develop current paleovegetational reconstructions on the Atlantic Coastal Plain include: (1) Lake Annie (Watts 1975a), Mud Lake (Watts 1969), and Sheelar Lake (Watts and Stuiver 1980) in Florida, (2) Lake Louise and Pennington (Watts 1971) on the southern Georgia Coastal Plain, (3) Goshen Springs (Delcourt 1979) on the Coastal Plain of southern Alabama, (4) Quicksand, Bob Black, and Green ponds (Watts 1970) and Pigeon Marsh (Watts 1975b) in the Ridge and Valley Province of northwestern Georgia, (5) Anderson Pond (Delcourt 1979) on the Cumberland Plateau in Tennessee, (6) White Pond (Watts 1980b) on the Fall-Line in central South Carolina, (7) Singletary Lake, Jerome Bay and Rockyhock Bay (Whitehead 1973) on the North Carolina Coastal Plain, (8) the Dismal Swamp (Whitehead and Oaks 1979) in coastal Virginia, and (10) Hack Pond in the Shenandoah Valley, Virginia (Craig 1970). Added to this list is the recent palynological study of three wetalnd sites at Fort Bragg and Camp MacKall (Goman and Leigh 2003). The developmental history of the region can be divided into five major periods: 1) the Late Wisconsin (22,000-13,500 B.P.), 2) the Late-Glacial transition (13,500-10,000 B.P.), 3) the early Holocene (10,000-8,000 B.P.), 4) the Hypsithermal (8,0006,000 B.P.), and 5) the late Holocene (6,000 B.P.Present). The vegetational patterns of each of these periods will be briefly reviewed below with specific reference to the Coastal Plain of the Carolinas. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 13 Chapter 2. Environment 1 2 3 7 Fort Bragg 8 15 4 5 6 9 nti la At 14 10 n ea c cO Pollen Sites 11 12 Modern Forest Associations 1. Hack Pond 2. Dismal Swamp 3. Rockyhock Bay 4. Jerome Bay 5. Singletary Lake 6. White Pond 7. Anderson Pond 8. Pigeon Marsh 9. Bob Black Pond 10. Lake Louise 11. Sheelar Lake 12. Mud Lake 13. Lake Annie 14. Goshen Springs 15. Bogwater Complex Oak-Hickory Forest Mixed Mesophytic Forest Appalachian Oak Forest Northern Hardwood Forest Spruce-Pine Forest 13 North 0 100 Miles Oak-Hickory-Pine Forest Southern Mixed Forest Pocosin Southern Floodplain Forest Palmetto Prairie Everglades Mangrove Forest Figure 4. Location of Pollen Sites Discussed in the Text Relative to Modern Vegetation Associations. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 14 Chapter 2. Environment Late Wisconsin (22,000-13,500 B.P.) Late Wisconsin vegetation in the southeastern United States has been inferred to represent a pine parkland with minor components of spruce, fir, and broad-leaved hardwoods (Watts 1980a:392-393; 1983:302-304). Pine comprises 60 to 80 percent of the pollen spectra from this period, and it has been argued that most of this pollen belongs to Pinus banksiana or the jack-pine. Three spruce species appear to have been widespread across the Atlantic Slope during this period, Picea glauca, P. rubens, and P. mariana. In the Piedmont and foot-hills of the Ridge and Valley Province, low frequencies of broad-leaved hardwoods including oak (Quercus sp.), ironwood (Ostrya/Carpinus), and hickory (Carya sp.) were present, while these arboreals were relatively rare on the Coastal Plain. The herbaceous dominants are principally associated with prairies today, and included wormwood (Artemisia), ragweed (Ambrosia), other composites (Tubuliflorae), grasses, and sedges. Although the Late Wisconsin vegetation of the Southeast seems to have been relatively homogeneous in its composition, important altitudinal and latitudinal clines have been identified (see Watts 1980a:393, Whitehead 1973). Significant levels of spruce are documented as far south as the Tunica Hills of Louisiana (Delcourt and Delcourt 1977), but very little, if any, spruce is present in pollen spectra and macroplant samples from Florida. At Sheelar Lake in northern Florida, broad-leaved trees and herbs comprised 30 percent of the Late Wisconsin assemblage, while spruce was virtually absent (Watts and Stuiver 1980). North Carolina and Virginia, although still dominated by jack pine, were situated on the southern latitudinal transition to the spruce woodlands of the Northeast, and spectra from these states consistently show higher percentages of spruce than those from more southerly states. In addition, they contain a suite of “northern” herbaceous species including clubmoss (Lycopodium sp.), burnet (Sanguisorba canadensis), and curlygrass fern (Schizaea pusilla). The southern limit of this transitional zone appears to have been located between Singletary Lake on the North Carolina Coastal Plain and White Pond, near Columbia, S.C. (Watts 1980a:393). The South Coastal Plain region of North Carolina would have been situated near this boundary during the Late Wisconsin and we would expect to find a transitional type vegetational assemblage. This period represents a major hiatus of sedimentation around Fort Bragg, but evidence from the PAW2 core at the Peatland Atlantic White Cedar Forest site suggests that Late Glacial vegetation here was pine-oak woodland with riparian species such as alder locally important (Goman and Leigh 2003:52). This forest type is consistent with a climate cooler and mositer than today’s. Late-Glacial Transition (13,500-10,000 B.P.) Between 13,500 and 11,000 B.P. time-transgressive, but abrupt changes occur in the pollen spectra of the Southeast (Watts 1983:305-306). These changes are reflected in increased abundances of mesic deciduous species such as hickory, beech, birch, and hemlock. The principal mechanism for these changes was the post-Wisconsin warming trend, which allowed the northward migration and range expansion of broad-leaved deciduous trees from restricted refugia in various Southeastern localities. Florida sites evidence the earliest transition. At Sheelar Lake in northern Florida the Late-Glacial transition is dated as early as 14,600 to 14,000 B.P with the resumption of sedimentation (Watts and Stuiver 1980). At this time the pollen spectrum of the site was still dominated by a Late-Glacial, droughtadapted woodland/prairie consisting of oak, hickory, hackberry (Celtis), cedar (Juniperus or Chamaecyparis), and herbs (especially Ambrosia). However by 13,500 B.P., this dry woodland had been replaced by a more mesic forest assemblage that included significant proportions of beech as well as oak and hickory. This evidently represented only a brief florescence that was soon replaced by an oak-pine forest between 13,500 and 11,200 B.P. Vegetational transition at White Pond in the South Carolina midlands occurred somewhat later, at around 12,800 B.P. (Watts 1980a). Here, there was a seemingly abrupt replacement of jack-pine and spruce forest with oak, hickory, beech, ironwood, and elm (Ulmus). Again, this mesic assemblage was short-lived and by 9,500 B.P. an oak-pine forest had been established across North Carolina. The transition in North Carolina began around 11,000 B.P. as evidenced by a sharp rise in oak pollen at Rockyhock Bay and Singletary Lake (Whitehead 1973). Signifi- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 15 Chapter 2. Environment cant proportions of hemlock, beech, birch, and ironwood also occur at this time, while the appearance of hickory is not well documented until around 9,000 B.P. The mesic forest was only weakly developed in North Carolina as compared to more southerly localities, and was also differentiated from the latter by “northern” species such as hemlock and birch (Watts 1980a:399). Early Holocene (10,000-8,000 B.P.) Between 9,000 and 10,000 B.P., the mesic forests of the Southeastern Coastal Plain were replaced by more xeric-adapted forests of oak, hickory, and pine. Broad-scale vegetational changes throughout the eastern United States during this period indicate that climatic conditions shifted significantly at the beginning of the Holocene, which marks the establishment of the modern climatic regime (Davis 1983:176-177). It was also during this time period that the dramatic rise in post-Pleistocene sea level began to stabilize, and by about 9,000 B.P., sea level was only several meters lower than it is today Current reconstructions suggest that both a drop in precipitation and increased temperatures ushered-in the Holocene and ultimately provided the impetus for the continued altitudinal and latitudinal migration of the mesic-adapted species northward and upward. Latitudinal and altitudinal vegetation gradients in the southeastern United States during this time consisted of three major bands. These included an oak savanna in peninsular Florida, an oak-hickory-southern pine forest along the Gulf Coastal Plain and the southern Atlantic Coastal Plain terminating in central South Carolina, and a mixed hardwoods association extending across North Carolina and most of the Cumberland Plateau (Delcourt and Delcourt 1981:147-149). The higher elevations of the Appalachians still supported a refugium for the Late Wisconsin jack pine-spruce forest. Mesic hardwood species were predominantly distributed northward in a band of mixed conifer-northern hardwood associations stretching from the Mid-Atlantic states into the Mid-continent of Indiana and Illinois. The Hypsithermal (8,000-6,000 B.P.) The early Holocene temperature gradient increased rather dramatically until it peaked sometime between 9,000 and 5,000 B.P. By 9,000 B.P. temperatures are estimated to have been approximately equal to today, and sometime during the interval 8,000 to 6,000 B.P., it is hypothesized that temperatures were significantly higher (Davis 1983:176). This interval has been variously referred to as the altithermal, the mid-Holocene temperature maximum, and the hypsithermal. The climatic conditions of this period are not well understood at present and there is some evidence to suggest that it was latitudinally time-transgressive. It was during this time that the oak-dominated deciduous forest of the eastern United States reached its maximum distribution, and hickory experienced a florescence (Webb 1988:402). The relatively flat latitudinal vegetation banding of the earlier periods is replaced by a decidedly northward curve along the Atlantic Slope (Delcourt and Delcourt 1981:151). By the end of this period modern vegetation distributions were established throughout the Southeast. A decidedly new forest type, the southern pine forest, replaced the oak-hickory-southern pine forests along the Gulf Coastal Plain and the Atlantic Coastal Plain as far north as southern Virginia (Delcourt and Delcourt 1981; Watts 1979, 1983; Webb 1988). Associated with this new forest type was the expansion of swamp species such as cypress (Taxodium), sweet-gum (Liquidamber), and tupelo or black-gum (Nyssa), whose distribution had previously been confined primarily to the Mississippi basin and delta. Swamping and the establishment of the southern pine forest in the Coastal Plain appear to have been brought about by the processes of sea level stabilization and accompanying stream gradient flattening. By 6,000 to 5,000 B.P. the formation of the modern swamps along the Coastal Plain was essentially completed (Brooks et al. 1989, Watts 1980a). The issues surrounding the true climax vegetation of the Southeastern Coastal Plain have been reviewed earlier in this chapter (see also Quarterman and Keever 1962). Regardless of the controversy, it is a fact that pollen profiles throughout this region document a significant increase in southern pine during the hypsithermal, and it would appear that changes in water table conditions resulted in a significant representation of “post-climax” conditions which would Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 16 Chapter 2. Environment have inhibited a homogeneous development of the oak-hickory-southern pine association. Delcourt and Delcourt (1981:150) contend that the oak-hickorysouthern pine association existed as a true climax vegetation type only along the Piedmont of the Atlantic Slope and the Ozark Highlands. Late Holocene (6,000 B.P.- Present) Very little broad-scale change has occurred in vegetational and ecological distributions in the Southeast since about 6,000 B.P. Delcourt and Delcourt’s (1981) paleovegetation maps of 5,000 B.P. and 200 B.P., in fact, are virtually identical. The most significant changes to occur during this time interval were the continued development of swampy freshwater wetlands on the Coastal Plain with the resumption of somewhat moister climatic conditions and the development and expansion of estuarine and salt marsh habitats along the coast. The initial development of estuaries is dated by archeological inference to around 4200 B.P. along the South Carolina coast (Brooks et al. 1989:93-94). It is evident that the most dramatic impact of these trends was experienced along the sea island region of Georgia and South Carolina and the Outer Banks region of North Carolina where the complex geological structure of the submergent coastline provided conditions most favorable to the development of estuarine and marshland ecosystems. The same effects were not so strongly felt along the South Coastal Region of North Carolina, where the more homogeneous strand structure predominated. Nevertheless, significant precontact and ethnohistoric utilization of estuarine resources has been documented along the North Carolina coast as well (see Phelps 1983). Locally, vegetation assemblages appear to have achieved a “modern” composition. Oak-pine forests predominated and Pinus palustris became the dominant upper story species (Goman and Leigh 2003:54). Sedimentation resumed across the region at about 6,000 B.P. Paleofauna The dramatic changes that occurred during the Late-Glacial transition were accompanied by major animal extinctions, the most obvious of which involved the large mammal component of the Late Pleistocene Rancholabrean faunal assemblage (Voorhies 1974:85). At least thirty-three different genera of large mammals from this assemblage did not survive the transition, including the equids, camelids, and the Proboscidea (Grayson 1987, Martin 1984:361-363, Meltzer and Mead 1985). The issue of whether these extinctions were totally independent of human predation has never been sufficiently resolved (see Martin and Klein 1984), but the coincidence of widespread human colonization of the New World at this time does suggest at least a contributory role for Paleoindian populations. It is at times easy to ignore the fact that Late Pleistocene faunal assemblages also contained most of the smaller, less-glamorous species that inhabit this region today. Semken (1983:192) describes the Late Pleistocene fauna of the United States as consisting of a combination of components including extinct megavertebrates, extant mega- and microvertebrates of temperate regions, and now disjunct large and small northern species. Modern faunal assemblages, as a consequence, are often considered to represent an “impoverished residuum” of the diverse and dense Late Pleistocene Rancholabrean assemblage (Martin 1967; Martin and Webb 1974; Semken 1974, 1983). Webb (1981) distinguishes three late Pleistocene faunal zones in the Southeast (see also Goodyear et al. 1989:22, 1) a subtropical zone covering Florida and southern Georgia, 2) a temperate zone extending northward into the central South Carolina area, and 3) a boreal zone extending from central South Carolina through North Carolina and up into the Mid-Atlantic region. The locations of these zones conform well to the Late Wisconsin and Late-Glacial transition vegetation bands previously discussed. The subtropical faunal zone correlates with the scrub oak dune prairie of peninsular Florida, the temperate zone corresponds to the oak-hickory-southern pine forests/savanna of the Gulf Coast and the southern Atlantic Slope, and the boreal zone extends from the central South Carolina ecotone of mixed conifer and northern mesic hardwoods northward into the jack pine-spruce parkland of North Carolina and the Mid-Atlantic. The subtropical assemblage consisted of species adapted to a warm, moist, and relatively equable climate such as the giant tortoise (Geoche- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 17 Chapter 2. Environment lone crassiscutta), deer (Oedicoleus virginianus), glyptodonts, giant ground sloth, the American mastodon (Mammut americanum), tapir, peccary, giant beaver, capybaras, alligator, and turtles. Species of the narrow temperate zone consisted of the Columbian and woolly mammoths (Mammuthus columbi and M. primigenius), bison (Bison), horse (Equus), camelids (Camelops and Hemiauchenia), the American mastadon, and deer. The boreal zone fauna was represented by the woolly mammoth, caribou (Rangifer tarandus), horse, and bison. The ranges of the Rancholabrean megafauna, of course, cross-cut the zones of demarcation, and Webb (1981) has argued that the temperate zone may have exhibited the greatest species diversity owing to its transitional or ecotonal position. Again, the Strand zone would have occupied a transitional position between the northern boreal assemblage and the temperate zone assemblage. There is definitive evidence to determine that the earliest known human populations in North America, identified as Paleoindian, did exploit the Rancholabrean megafauna. This evidence is primarily derived from the western United States, but several examples have been documented from the eastern United States as well, including the Kimmswick mastadon in Missouri (Graham et al. 1981), the butchered giant tortoise carcass from Little Salt Spring, Florida (Clausen et al. 1979), the Bison antiquus skull with a fragment of a chert projectile point embedded in it from the Wacissa River Valley, north Florida (Webb et al. 1984), and the worked proboscidean bone from sink holes in north Florida (Dunbar et al. 1974.). A review of radiocarbon dated assemblages, however, indicates that human exploitation of megafauna in North America was probably limited to a period of less than 1,000 years (Meltzer and Mead 1985), and that the major peak of extinctions occurred between 11,500 and 11,000 B.P. Haynes et al. (1984) cogently argue that only the very earliest Paleoindian groups (ie. Clovis), in fact, exploited these large animals prior to their extinction. Current evidence indicates that the post-Pleistocene mass extinctions were virtually complete as early as 10,800 to 10,000 B.P., and that only modern faunal assemblages can be found after this time range (Goodyear et al. 1989:25). Paleoshorelines A final factor of importance in evaluating Late Glacial and Holocene paleoenvironments is sea level fluctuations. It has been known for some time that sea levels have changed time and again over the history of the earth and it has been determined that these changes are related to a number of complexly interrelated factors. The five basic determining factors include: (1) tectonic changes or movements in the earth’s crust, (2) glacial isostacy or rebound, (3) hydro-isostacy or rebound, (4) geoidal changes, and (5) glacio-eustatic movements due to alternating processes of glaciation and deglaciaition (Bowen 1978:158). During the last glaciation sea levels on the eastern North American coastline varied between 90 and as much as 130 meters below present sea levels, which resulted in exposing most, if not all, of the continental shelf to terrestrial life forms. The Late Glacial and Holocene sea level rise is attributable primarily to processes associated with deglaciation. We know from studies in the southeast (see Colquhoun and Brooks 1986) that sea level did not attain dynamic equilibrium until about 4,000 years ago. Prior to this sea levels were generally greater than 2 or 3 meters lower than today. At 8,000 years ago sea levels are estimated to have been at approximately 8 meters below present levels and a rather steep rise in sea level is documented from the beginning of the Late Glacial until more stable conditions were established at around 6,000 years ago. This, of course, resulted in a rapid inundation of the continental shelf and the destruction of established estuarine fisheries and shellfish habitats, which were not established again until around 4,000 B.P. During the Early Holocene and Late Glacial periods lake sediments indicate that the Coastal Plain was more xeric than today and was essentially devoid of major swamp wetlands (Brooks et. al. 1989:91-92) due to lowered sea levels. Evidence exists, however, for increased precipitation during the Holocene in the form of organically rich clays dating to this period in Dismal and Okefenokee swamps. Conditions became drier in the Hypsithermal and the general xeric nature of the Coastal Plain did not change significantly until about 5,000 years ago. At this time extensive peat deposits are detected throughout the Coastal Plain and it is inferred that this was the initiation of major swamp forming conditions. This is Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 18 Chapter 2. Environment associated with a rise in sea level to within several meters of present day levels and it is argued the two processes were linked through a concomitant rise in freshwater hydrologic levels. Clearly, then, prior to sea level stabilization at about 5,000 to 4,000 years ago, the Coastal Plain was a relatively xeric environment generally devoid of the extensive swamp and bay formations we see throughout the area today. Adaptation was geared to the hunting and collecting of terrestrial fauna and flora tethered to creek beds and river valleys. At about 4,200 B.P., however, we begin to see evidence for intensive shell collecting and estuarine exploitation in the newly formed estuaries and major interior settlement around swamp wetlands. This period corresponded with the emergence of the ceramic Late Archaic cultures of the major river valleys and coast. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 19 Chapter 2. Environment Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 20 Chapter 3. Culture History Lands within what is now Fort Bragg are situated at the interface between the well-defined cultural sequences of the North Carolina Piedmont and Coastal Plain Provinces. The following discussion draws extensively from the work of others (Coe 1964; Claggett and Cable 1982; Phelps 1983; Ward 1983; Oliver 1981, 1983; Anderson and Hanson 1988; Smith 1986; Daniel 1998). Archaeologists have divided the prehistory of North Carolina’s Coastal Plain region into three general stages (Paleoindian, Archaic, and Woodland), based for the most part on inferred economic adaptations and ceramic traditions (see Figures 5 and 6). A fourth possible stage of development, the pre-Clovis, predates the Paleoindian and is a highly contested unit of cultural division within North and South America. As of the present, no pre-Clovis sites have been identified in North Carolina. PALEOINDIAN PERIOD (10,500-8,000 B.C.) The earliest evidence for human occupation in North Carolina began during the Paleoindian period near the end of the late Pleistocene. The Paleoindian period is generally dated from ca. 10,000 and 8,000 B.C. throughout North America. While various chronologies and artifact sequences have been proposed over the years (Haynes et al. 1984, Gardner and Verrey 1979, Oliver 1981), the on most often cited chronology for the southeastern United States today, is proposed by Anderson et al. (1996:7). These authors subdivide the 2,500-year span into three subperiods. The earliest subperiod is represented by the Clovis Point that has been more securely dated from sites in the southwestern and plains regions to ca. 9,250 and 8,950 B.C. The middle subperiod (ca. 8.950-8,550 B.C.) is identified by assemblages containing Cumberland, Simpson, and Suwannee fluted and unfluted projectile points. In contrast, the late subperiod (ca. 8,550-7,550 B.C.) is characterized by fluted Beaver Lake and Quad types and fluted and unfluted Dalton and Hardaway point styles. Oliver’s (1981) proposed revision of the North Carolina Piedmont sequence extends the temporal range of the Paleoindian period back to 12,000 B.C. and includes the Palmer Cornernotched type into the terminal phase of the PaleoIndian period. This latter projectile point type is most commonly recognized as Early Archaic (see Goodyear et al. 1979) and this perspective is adopted in this overview. Traditional interpretations of Paleoindian Period subsistence practices have relied on a view of Paleoindians as hunters of late Pleistocene megafauna. Research beginning in the mid to late 1980s indicates that reliance on megafauna may have been the norm in the western part of North America, while plants and small game comprised a larger portion of the southeastern Native American’s diet (Sassaman et al. 1990:8). Although, the Paleoindian period corresponds with the final stages of the late Pleistocene megafauna extinction, when 35 to 40 known species of large mammals became extinct (Martin 1984, Pielou 1991), only a few examples of the direct exploitation of megafauna in the southeastern United States have been documented. Most of these come from wetland and underwater sites in Florida and include a butchered giant tortoise carcass recovered from Little Salt Spring (Clausen et al. 1979), a bison Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 21 Chapter 3. Culture History antiquus skull with an embedded projectile point fragment in the Wacissa River valley (Webb et al. 1984), and a worked proboscidean bone from a sink hole in northern Florida (Dunbar et al. 1990). More “modern” species such as caribou have been recovered at Holcombe Beach, Michigan (Cleland 1965) and Dutchess Cave Quarry, New York (Funk 1977), and white-tailed deer and wapiti have been positively identified at Meadowcroft Rockshelter, Pennsylvania (Adovasio et al. 1977). The importance of meat in the Paleoindian diet, however, can sometimes be overemphasized. Ethnobotanical remains from Meadowcroft Rockshelter and Shawnee-Minisink in Pennsylvania (McNett et al. 1977) indicate that secondary resources including fish, birds, hawthorne, and nuts were also incorporated into various eastern woodland Paleoindian subsistence systems. Paleoindian occupation in the Southeastern United States is one characterized by high mobility, high range (territorial) mobility, low population density, and a focal hunting economy (Anderson and Joseph 1988, Gardner 1979, Goodyear 1979, Goodyear et al. 1989, Meltzer 1988, Smith 1986, Steponaitis 1986, Williams and Stoltman 1965). However, some researchers are beginning to question these traditional views and are advocating new theories. One such theory is that Paleoindians were less mobile, and selected choice areas for initial settlement. Only after this initial area was colonized, did Paleoindian groups expand into other regions (Sassaman et al. 1990:8). Another theory stipulates that early Holocene mobility patterns should have shifted from logistically based settlement systems to more residentially mobile systems as temperatures warmed and the homogeneity of resource distributions increased (Cable 1982a). Contrary to the traditional view (Caldwell 1958) of a gradual shift toward more sedentary systems through time, more recent studies (Anderson and Hanson 1988; Anderson and Schuldenrein 1983, 1985) argue that Paleoindian and initial Early Archaic populations may have maintained more stable residences than those of the later Early Holocene and Middle Holocene. In spite of increasing research into Paleoindian sites, the Southeast in general and North Carolina specifically, has few sites with diagnostic Paleoindian artifacts and even fewer sites offering more than surface materials. In northern Virginia, Gardner (1974) has proposed a Paleo-Indian settlement model based on his excavations and surveys in and around the Flint Run site. The model states that a dependence on highly siliceous lithic resources to maintain technological “readiness” was necessary given the highly mobile nature of Paleoindian groups. Perkinson’s (1971, 1973) North Carolina fluted point study suggests that Paleo-Indian site densities may have been higher in the Piedmont than in the Coastal Plain. In fact, his numbers indicate that Paleo-Indian occupation in the Coastal Plain was very limited, as only 15 percent (13 of 83) of the points came from Coastal Plain counties. Daniel’s (1998) more recent geographic study of 189 North Carolina Paleoindian projectile points (Clovis, Quad, Redstone, and Simpson types) yielded the same results with a large percentage (90%) of the points clustering near lithic source areas in the western mountains and Piedmont regions of the state. The remaining 10 percent (n=19) occurred within the Coastal Plain where lithic resources are believed to be scarce. One of the Paleoindian points in Daniel’s study was recovered at site 31HK118 on Fort Bragg. Two other fluted points from this site are known to exist, but are not curated on the installation (Griffin et al. 2001). Another fluted point was found at Site 31CD145 on an upland watershed divide six miles west of Stedman, North Carolina (Robinson 1986). In addition, a late Paleoindian Hardaway Point was recovered at site 31SP103, a large multicomponent site, located along the South River (Hackbarth and Fournier-Hackbarth 1981). To the south of Fort Bragg in South Carolina, a different picture of Paleoindian settlement patterns emerges. A study of the South Carolina Paleopoint Database reveals a relative high density of Paleoindian points in the northern counties of South Carolina within the Fall Line and Piedmont counties including Kershall (n=20), York (n=12), Lexington (n=11), Lancaster (n=8), and Chesterfield (n=7). All the counties, except York County, are located along the Fall Line in an area much like the environment of Fort Bragg. Because of the apparent heavy concentration of Paleoindian points in this zone, Goodyear et al. (1989:44) have speculated that this zone evidenced a disproportionately high rate of reoccupation or was the location for prolonged, seasonal base camp occupations. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 22 Chapter 3. Culture History ARCHAIC PERIOD (8,000-1,000 B.C.) PERIOD 3,000 Otarre Stemmed LATE ARCHAIC 4,000 5,000 Guilford/Brier Creek Lanceolate MIDDLE ARCHAIC 7,000 Morrow Mountain I/II Stemmed Stanly/Kirk Stemmed 8,000 9,000 Savannah River Stemmed MALA/Allendale Stemmed EARLY ARCHAIC Uncalibrated Radiocarbon Years B.P. 6,000 Diagnostic Projectile Point Types Bifurcate Variants Palmer/Kirk Corner-Notched Taylor/Hardaway Side-Notched 10,000 11,000 PALEOINDIAN Dalton Cumberland/Simpson/ Quad Clovis 12,000 PRECLOVIS 13,000 Figure 5. Generalized Cultural Chronology for the Paleoindian and Archaic Periods of North Carolina (Data Drawn from Anderson et al. 1990, Sassaman et al. 1990, and Sassaman et al. 2002). The Archaic Period has been defined as a shift towards increasingly new dietary patterns reflecting a variety of birds, fish, mammals, and reptiles in the site assemblages. The Archaic sequence has been traditionally divided into three subperiods: the Early Archaic (8000-6,000 B.C.), the Middle Archaic (6,000-3,000 B.C.) and the Late Archaic (3,000-1,000 B.C.). In general, the Archaic is viewed as a lengthy period of adjustment to changing environments brought about by the Holocene warming trend and rising sea level. Caldwell’s (1958) model of wide-niche or “broad spectrum” hunter-gatherer adaptations continues to succinctly define the period for most archeologists. However, the differences between the cultures at either end of the sequence are immense and indicate that major cultural and adaptive changes occurred during the Archaic period. The adaptation to a warmer, post-Pleistocene environment is reflected in the tools associated with the Early Archaic period. These lithic assemblages exhibit a number of similarities with those of the Paleoindian period. Projectile points (ie, Palmer, Big Sandy, and Kirk Corner-Notched styles) remain stylistically formalized and show evidence of economizing resharpening strategies, hafted end scrapers continue to be well represented and there is an emphasis on the curation and use of cryptocrystaline raw material such as chert and high grade metavolcanics. Cleland (1976) has suggested that these attributes indicate a continued focus on the hunting and processing of big game animals. In support of this Goodyear et al. (1979:104) note that plant processing tools such as grinding stones are extremely rare in Early Archaic deposits. Chapman (1977:95, 116) reports the presence of eight grinding slabs in Kirk Corner-Notched deposits at Ice House Bottom in Tennessee, but was unable to demonstrate the reliance on, or even presence of, “weed seeds” in the flotation samples from these levels. Acorn and hickory nutshells, however, were abundant. Faunal remains from Early Archaic associations in the Southeast indicate a wide spread emphasis on white-tailed deer, but a variety of smaller game including gray squirrel, raccoon, turkey and box turtle have also been identified (Goodyear et al. 1979:105). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 23 Chapter 3. Culture History Subsistence data then, suggest that hunting large game (i.e., white-tailed deer, elk, and bison and antelope on the western margin of the eastern woodlands) was indeed a major element of Early Archaic economies as was true for the Paleoindian period. But it is also true that there was also significant energy devoted to nut and seed gathering (i.e., oak, hickory, black walnut, hackberry, persimmon, copperleaf, pigweed, goosefoot, maygrass, knotweed, purslane, grape, etc.) and the trapping of smaller terrestrial game and aquatic resources (i.e. mussels, fish, turtle, ducks, geese, quail, turkey, beaver, squirrel, skunk, bobcat, opossum, porcupine, raccoon, otter, etc.). In fact, a review of subsistence data from major Dalton and Early Archaic contexts in the Southeast leads Bruce Smith (1986:10) to observe that Early Archaic subsistence systems were diverse, “providing little support for the existence of a focal economy,” and that the available faunal-floral assemblages resemble the “broad spectrum” composition of those of later assemblages in important ways. He further notes that the subsistence resources commonly associated with the Early Archaic period indicate significant exploitation of both upland, closed canopy, climax forest and edge areas such as river valleys where stages of early succession were fostered by unstable geomorphological conditions and possibly precontact land use practices. The widespread occurrence of Early Archaic sites throughout the Southeast, and in both riverine and non-riverine settings (Sassaman 1996, Daniel 1992, O’Steen 1992, Goodyear et al. 1979:105, Ward 1983), suggests population increases from the Paleoindian period and perhaps a greater emphasis on foraging strategies. The few excavations located along or near the Atlantic Slope that have recovered preserved Early Archaic living surfaces (Anderson and Schuldenrein 1985, Broyles 1971, Chapman 1975, Claggett and Cable 1982, Coe 1964) do not exhibit evidence of long term habitation. The principal features of these floors consist of rock clusters, hearths, small pits, raw material caches, and, very occasionally, grinding slabs. Evidence of shelters (i.e., postholes) has not been positively identified and it is speculated that they were temporary huts as opposed to permanent domiciles. This kind of pattern is consistent with a residentially mobile settlement system, in which various site types may occur. While individual researchers sometime disagree on terminology and the methods used to identify various site types, all generally agree that Early Archaic assemblages can be divided into base camps, foraging camps, and special-purpose sites (Anderson and Hanson 1988, Anderson and Joseph 1988, O’Steen et al. 1986, Cable 1996, Daniel 1998, Kimball 1996). The scale of Early Archaic settlement systems has been difficult to define, as well. Goodyear (1983) has suggested a Fall Line centered settlement system for the Early Archaic period similar to that proposed for the Paleoindian period described earlier. Anderson and Hanson (1988) have elaborated on this general scheme by proposing a seasonal round for Early Archaic systems in which the Piedmont was occupied during the summer and early fall, the Coastal Plain was visited in the spring, and the Fall-Line was inhabited during the fall and winter. Occupation of the Fall Line is characterized by the establishment and/ or reoccupation of fall aggregation sites and winter base camps, while the Piedmont and Coastal Plain were exploited by dispersed foraging units. It is further proposed that the territory of each Early Archaic band was distributed along a major drainage and that the South Atlantic Slope contained eight such bands distributed from northern Florida to Pamlico Sound, North Carolina (Anderson and Hanson 1988). Daniel (1994, 1998), on the other hand, presents a persuasive argument for a cross-watershed Early Archaic settlement model. He suggests groups were tethered to high-grade lithic sources and they used the available resources of several drainages in the coarse of their seasonal or yearly rounds. The mid-Holocene warming trend, the Hypsithermal, has been recognized as the principal cause of subsistence and settlement changes during the Middle Archaic. Stemmed points replaced earlier notched points, with the Kirk Stemmed/Serrated (6,000-5,800 B.C.), Stanly (6,000-5,000 B.C.), Morrow Mountain (5,500-4,000 B.C.) and Guilford (4,000-3,000 B.C.) being the most common diagnostics of this period. Other technological changes noted in Middle Archaic assemblages include the discontinued use of end scrapers (Cable 1982b, Kimball and Chapman 1977), raw material proportions tend to reflect local availability (House and Ballanger 1976), and cryptocrystaline materials are de-emphasized as distance to source increases (Goodyear et al. 1979: 111). Also, the use of storage facilities and human Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 24 Chapter 3. Culture History interments increased during this period (Griffin 1974, Wetmore 1986, and Chapman 1977). One commonly referenced trend is the notion that ground stone tools increase dramatically during the Middle Archaic. The large ground stone tool assemblage from the Early Archaic deposits at Rose Island (Chapman 1975:153170), however, has led Bruce Smith (1986:18-21) to cogently argue that there does not appear to be a measurable difference in the subperiods on the basis of present evidence. Consequently, he argues that there is no compelling reason to suggest that a technological revolution took place during the Middle Archaic in the Southeast. Research in South Carolina has greatly influenced archaeological perceptions of Middle Archaic technology and mobility/settlement strategies employed along the Southeast Atlantic Slope (Ward 1983, Poplin et al. 1993, Blanton and Sassaman 1989, Anderson 1996, Kowalewski 1995). The consensus is that Middle Archaic technology could be characterized by localized raw material procurement and use, generalized toolkits with little concern for extended curation of tools, and a lessened concern with the quality of raw materials used to manufacture tools (Blanton 1983). Furthermore, the “Adaptive Flexibility” model, as it came to be known, characterized Middle Archaic occupations as containing highly redundant assemblages whereby all stages of lithic tool reduction regularly co-occurs on sites and quartz was selected over less abundant, higher-quality stone as a “trade-off between curation and expediency” (Sassaman 1983:84). As more Middle Archaic components are being intensively examined, evidence is accumulating for the presence of more complex and varied mid-Holocene assemblages. Intensive testing and data recovery projects have successfully identified Middle Archaic period residential occupations along the South Carolina Fall Line and Coastal Plain. The Middle Archaic components along the Fall Line have yielded preserved surface and basin-shaped hearths, and a diverse tool assemblage including Morrow Mountain bifaces, scrapers, gravers, spokeshaves, and groundstone tools presumably used for numerous maintenance and production related activities (Wetmore 1986, Radisch personal communication 2000, O’Steen 1994). In contrast to these highly diverse assemblages, less diverse Middle Archaic assem- blages in the uplands of the Coastal Plain have been identified and found to exhibit internal structural patterns consistent with short-term residential camps. These camps conform well to Yellen’s (1977) models of hunter-gatherer camp structure that consist of huts, close-by exterior hearths identified by tool and faunal bone concentrations, and nuclear area artifact scatters (Cable et al. 1996; Cable and Cantley 1998, and Cantley and Cable 2002). Contrary to the prevailing view of Middle Archaic groups adopting an expedient technology, these assemblages contain significant percentages of reused and curated tools. Work at Richburg Quarry (38CS217), a quarry located in the “quartz rich” landscape of the Piedmont, illustrates how Middle Archaic groups sought high-quality knappable quartz outcrops for the production of efficient, transportable, and functionally diverse tool forms (i.e., bifacial cores and preforms) without substantially affecting their carrying costs (Cantley 2000). This curation strategy would be important for groups who practice high residential mobility, but are constrained in terms of how far they can move, and therefore may not always be located near a high-quality source of raw material. Other studies linking hunter-gatherer mobility/settlement models to the regional lithic resource base and raw material procurement patterns have attempted to document seasonal movements and group territories. Sassaman and Anderson (1994:106107, 124,127-128) documented the widespread occurrence of Morrow Mountain period groups over most of the Piedmont with the greatest number of occupations occurring in the inner Piedmont between the Broad and Savannah rivers. Tippett (1992:136), on the other hand, proposed a settlement model tethered to individual drainages whereby various Morrow Mountain groups inhabited the Yadkin-Pee Dee drainage, Broad-Congaree drainage, and the Savannah River drainage. Later during the later Middle Archaic period, an analysis of Guilford points represented in the state-wide collector survey suggests a shift in territory (Charles 1981:53). This data indicates that the later Guilford period is characterized by a reduction in group territory to an area between the Broad and Yadkin-Pee Dee rivers in the northern Piedmont and Fall Zone of South Carolina. Archaeological surveys conducted within the Sumter National Forest provide additional support for the Broad River being a territorial boundary with few Guilford assem- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 25 Chapter 3. Culture History blages occurring south and west of the Broad River (Benson 1995; Bates, personal communication 1998). The Yadkin-Pee Dee drainage poses another type of boundary where an abrupt change occurs in Guilford tool assemblages. West of the Yadkin River, Guilford assemblages are manufactured predominately from quartz, while east of the Yadkin River, assemblages are made almost exclusively of metavolcanic materials (Coe 1964, Abbott personal communication 1998). Cantley (2000) has suggested that the YadkinPee Dee river drainage may have served as a possible social boundary (see Wobst 1977) that divided contemporary groups and established rights and privileges over the use of specific raw material resources within each group territory. Fort Bragg located north and east of the Yadkin-Pee Dee drainage should therefore reflect a prevalence of metavolcanic manufactured tools. Similar to the South Carolina Fall Line/Sandhills, numerous Middle Archaic sites have been discovered at Fort Bragg (Benson 1995). Most if not all of these sites are characterized as technologically simple, low-density occupations with little evidence of differentiation in site function (Griffin et al. 2001). It should be noted that few of these sites have undergone intensive investigation to determine their composition and structure. Climatic and environmental pressures to adjust settlement systems in the direction of greater residential mobility in the Middle Holocene may have been offset at some point by range reduction due to tighter population packing (Anderson and Joseph 1988:130-131). One factor indicating range reduction pressures is the shift toward a heavy reliance on local lithic materials during the Middle Archaic (Blanton and Sassaman 1989, Sassaman 1983, 1988). Greater residential mobility may very well have typified Early Archaic and early Middle Archaic settlement systems regardless of gradual range reduction processes (Sassaman 1988), but other factors toward the latter half of the Middle Archaic period may have hastened a shift back toward logistical strategies, albeit within a much reduced range. One such factor affecting the Coastal Plain and coastline was the formation of swamps and estuaries as sea level began to stabilize (Brooks et al. 1989). Moreover, the Middle Holocene climate appears to have been drier, but also more variable, suggesting to Blanton and Sassaman (1989) that at least the Coastal Plain environment was changing toward a greater degree of patchiness and therefore would have presented Middle Holocene foragers with the opportunity to exploit an environment with increasing spatial resource segregation. Consequently, pressures toward the reversion to logistically oriented settlement systems may have been manifest earlier in the Coastal Plain than in the Piedmont. The mid-Holocene sub-epoch (8,000 to 3,000 BP) spanned a period of adjustment in which the cooler (slightly warmer summers and colder winters) postglacial climate warmed and environments began to stabilize and take on the structure that we observe in modern times (Schuldenrein 1996:3-9). Sea level rise slowed substantially and estuarine productivity exploded along the margins of the Coastal Plains. In the Southeast interior aquatic biomes emerged and flourished as stream channels became increasingly entrenched. Stream gradients also diminished due to sea level rise, ultimately resulting in overbank deposition of bed loads and meandering stream channels. In addition, sink hole basins that were generally dry during the early Holocene filled as water tables rose with sea level. Pollen cores further indicate that the oak dominated forests of the early Holocene were replaced by pine and swamp formations during the mid-Holocene along the Atlantic and Gulf coasts (Watts et al. 1996:29). The replacement was timetransgressive, beginning at about 8,500 BP in the Carolinas and emerging much later in more southerly latitudes. In Florida pine forests were not established until around 4,500 BP. Environmental and climatic changes were neither synchronous nor equivalent across this large geographic region, but the most significant change for human adaptation was the significantly increased productivity of aquatic biomes in estuaries, swamps and interior river valleys. It is generally believed that at the start of the mid-Holocene, precontact cultures in the Eastern Woodlands represented simple huntergatherers subsisting on terrestrial plants and animals and conducting a largely transhumant lifeway in small social groups, with many sub-seasonal residential moves (Anderson 1996; Cable 1982; Caldwell 1958; Sassaman 1983). Later in the period, however, evidence of resource intensification, increased group size, greater sedentism and perhaps even greater social complexity emerges. These trends are manifest most notably in aquatic biomes that supported high densities of fresh- and salt-water molluscan fauna. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 26 Chapter 3. Culture History Throughout the Mid-South and lower Midwest prime freshwater mussel beds anchored around shoals formed in the newly entrenched river channels and by 6,500 to 5,000 BP some Archaic groups were intensively exploiting these resources. On the Green River in central Kentucky massive shell midden deposits of as much as 2 to 3 m in depth were built up along stream banks by Archaic groups (Crothers and Bernbeck 2004; Marquardt and Watson 1983; Moore 1916; Webb 1946, 1950). Other food remains in the deposits included a wide range of terrestrial game, nutshell and fruit seeds. Structural remains have not been positively identified in the middens, but clusters of fire-cracked rock and large hearths for food preparation are common, as are burials. At the Indian Knoll site more than 1,100 burials were excavated. Similar freshwater shell-bearing archaeological deposits have been identified in the Upper and Middle Ohio river valleys, the Tennessee River system in Tennessee and Alabama, the Savannah and Ogeechee rivers in South Carolina and Georgia, the hilltop middens on the Wabash River in West Virginia and the St. Johns River in Florida. Together with the Green River Archaic, Claassen (1996:236-238) refers to these foci as the Shell Mound Archaic (SMA). Her review of available radiocarbon dates from the various areas indicates a range of ages between 7,180 and 2,400 BP. The Tennessee system begins about the same time as the Green River Archaic and continues as late as 2,400 BP. The middens on the Ohio and St. John’s rivers were apparently founded between 6,000 and 5,000 BP. Beginning slightly later, between about 5,000 and 4,5000 BP, were Stallings Island and Rabbit Mount on the Savannah River. The Wabash River middens began much later, at about 3,400 BP. The spread of intensive freshwater shellfish exploitation from its heartland in Kentucky and Tennessee at 7,000 to 6,500 BP to other interior river systems over the next several millennia may evidence gradually expanding population pressure on traditional forager subsistence strategies. The impressive shell middens of the Green River Archaic were originally interpreted as the refuse from permanent villages (Funkhouser and Webb 1932), but later evaluations of these and other large middens in the SMA suggest that they more likely represent seasonal aggregation camps (Hensley 1991; Hofman 1985). In contrast to permanent villages, aggregation camps are defined as locations that are temporarily occupied by otherwise dispersed family groups that come together during periods of seasonal resource abundance to arrange marriages, transact gift exchanges, conduct rituals and bury their dead. Evidence of sturdily constructed houses, which provides strong evidence for a degree of settlement permanence although not always permanent villages, is not abundant in the mid-Holocene archaeological record. Structural features in deposits of this age, however, are not likely to be well preserved. Sassaman and Ledbetter (1996) review the meager structural data of interest to this discussion. The earliest evidence they recognize for mid-Holocene structures comes from the Koster site in the lower Illinois River Valley. In Horizon 8C, dated to 7,300 to 6,850 BP, Brown and Vierra (1983:184) report an occupation surface containing a series of deep postholes that appear to represent structures. One posthole pattern appears to represent the outline of a rectangular structure measuring 4.5 x 5.0 m. The depth and size of the postholes suggest that the partially exposed structures evidence significant labor investment in permanent shelters. Similar structures were identified in the higher 6A and 6B Horizons at Koster, dating to the period between 5,800 and 4,900 BP. Although the mid-Holocene occupants of Koster were not intensively exploiting shellfish, Brown and Vierra (1983) argue that the degree of settlement permanence evidenced by the Koster structures was facilitated by the development of productive slack-water environments along the Illinois River. In general, these lowland sites are not regarded as permanent villages, but as seasonally revisited base camps positioned adjacent to productive shoals, oxbow lakes and other backwater features (B. Smith 1986:25). Sassaman and Ledbetter (1996:78) note that several SMA sites have yielded evidence of structures as well. Sykes-White Springs and Benton phase clay floors dating between about 6,300 and 5,000 BP have been excavated at sites in the Tombigbee River basin (Ensor and Studer 1983; Rafferty et al. 1980). The Perry and Mulberry Creek (Webb and DeJarnette 1942:61, 238), Eva (Lewis and Lewis 1961:15), Indian Knoll (Webb 1946:129) and Robinson (Morse 1967) sites have yielded small numbers of similar clay floors with postholes, hearths and Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 27 Chapter 3. Culture History pits. Ten rectangular prepared clay platform structures with postholes and hearths associated with the Late Archaic period have also been excavated at the Riverton site on the Wabash River (Winters 1969:98). These structures measure on average about 4.5 x 3.0 m, a size that inferentially would be appropriate to house a nuclear family. The best evidence of structural remains for Savannah River SMA sites comes from the Late Archaic Mims Point site near Stallings Island (Sassaman 1993a). At least three structures have been identified there, defined by partial arcs of postholes, in two cases, hearths, storage pits and refuse filled pits. Although Mims Point has been severely damaged by pot hunting, a semi-circular house pattern was inferred to surround a communal space of 200 m2. Although the mid-Holocene age architectural data are sparse, they clearly demonstrate that more stable seasonal habitation sites were being established along some interior river systems in the Southeast and Midwest as early as 7,000 BP, where thriving aquatic biomes were forming around shoals, oxbow lakes and other backwater features as a consequence of stream channel stabilization and entrenchment. Freshwater shellfish exploitation was an important feature of this development, but it was not a required condition. Another subsistence intensification strategy that emerges during the mid-Holocene was horticultural production and incipient domestication of native plants. Domestication of plants and a gradually increasing reliance on them occurred in the very same river valleys that produced the rich aquatic resources so important to the SMA regional systems. Spring flooding of these major river courses created open habitats in aggrading floodplains that promoted colonizing weedy plants to manifest characteristics that were pre-adapted for genetic manipulation and domestication (B. Smith 1992a: 3). Smith (2007) recognizes four major native crops that were domesticated in the mid-Holocene in the Eastern Woodlands. These include marshelder (Iva annua), goosefoot (Chenopodium berlandieri), squash (Cucurbita pepo) and sunflower (Helianthus annuus). Current evidence reveals that marshelder, squash and sunflower were domesticated between 5,500 and 4,800 BP, while goosefoot seeds exhibit a domesticated morphology by 3,800 BP. Importantly, Smith (1992b: 282-283) points out that the process of domestication began much earlier than this general time horizon. Anthro- pogenic habitat modification and encouragement of these crops may have encompassed several millennia, perhaps beginning as early as 7,000 BP. Bruce Smith envisions a general sequence of domestication beginning with the introduction of the wild progenitors of future cultigens from floodplain stands into the shell and dirt middens of human settlements, the environment he refers to as domestilocalities, followed by a gradual recognition of the value of these species and encouragement of their propagation. Finally, a process of selection would have created quasi-cultigens that were further encouraged through the establishment and expansion of prepared gardens. In spite of regular and systematic flotation analysis from Piedmont and Coastal Plain mid-Holocene contexts, no evidence of domesticated crops has yet been produced (Gremillion 1996:111). Taking stock of these features (e.g., subsistence intensification, plant domestication, thick occupation middens, cemeteries and variably sturdy architecture) has moved many to suggest that the Shell Mound Archaic represents a manifestation of hunter-gatherer societies undergoing a transformation to a more sedentary lifeway and to a more complexly organized society (Brown 1985; Claassen 1996; B. Smith 1986; Winters 1974). Claassen (1996) has argued that many of the shell middens represent ceremonial burial mounds that were intentionally constructed or later adapted to this function. Moreover, she views the establishment of formal burial areas within constructed mounds as indication of ranked or hierarchical social organization. Crothers and Bernbeck (2004:405-406) offer an alternative interpretation of the Green River Archaic middens. First, burial placement appears to have been ad hoc rather than planned, suggesting the absence of the kind of protocol that would be expected in a purposeful construction. Younger burials commonly intrude older ones in a haphazard manner and burial location (e.g., within or outside of middens) varies by site. Moreover, midden structure varies between heaped deposit and horizontal strata. Second, although the placement of burials within the middens might indicate that feasting was an important feature of burial ceremonialism, this activity does not appear to have been segregated from mundane refuse discard behavior. Third, there are no clear grave good associations between subgroups that would indicate rank or wealth differences. Crothers and Bernbeck (2004:414-415) argue that Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 28 Chapter 3. Culture History these characteristics more likely support the hypothesis that the Green River Archaic groups were organized as foraging groups that were basically egalitarian in organization. Non-ranked societies, however, operate at various levels of social organization based on resource structure and regional population levels. Clearly, the SMA evidences a level of society above that of simple, low population density foragers in that the riverine midden sites appear to reflect higher population levels and a degree of subsistence intensification. Johnson and Earle (1987:316-317) identify a level of cultural development that they refer to as the Local Group that may provide an analogy. These societies are basically organized at the nuclear and extended family level, but natural and demographic forces bring them together into aggregations for socializing and feasting. The aggregates cannot be maintained permanently due to social tensions or seasonal resource distributions, though, and they fission into scattered and isolated family forager groups. Social relationships are fluid and the aggregating group composition can vary from year to year. Crothers and Bernbeck (2004) argue that the unplanned cemeteries of the SMA midden sites are consistent with this kind of societal flux. In situations where regional population growth continues, however, social and resource pressures lead to greater competitive frictions and violent attacks between families. As internecine warfare intensifies alliances tend to be formed between local groups and Big Man Collectivities emerge. A “big man” is a transitory leader who unifies a local group or alliance of local groups for the organization of defense, war expeditions and sometimes also group economic endeavors. Local groups take on greater importance and ceremonial activities generally involve aggregations of local groups participating in defense alliances. Exchange networks are reinforced, social transactions take place and “big men” are afforded the opportunity to display their power and viability to lead the group or the larger intergroup collectivity. Because the power of “big men” is transitory and alliances between groups are fragile, however, institutionalized authority and the hierarchical stability that comes with the next level of cultural development, the Chiefdom, does not emerge. If population growth continued after the founding of each of the SMA midden sites, it could be predicted that some river valleys may have supported Big Man Collectivities. Evidence of this might come in the form of the identification of significant levels of deliberate trauma in burial populations, evidence of trophy dismemberment, mass burials with evidence of violent deaths or the discovery of rare high status burials in an otherwise undifferentiated population. Evidence of deliberate trauma does occur in SMA burial populations, but the evidence of intensified warfare is relatively equivocal (M. Smith 1996:142143). Over the past several decades evidence has come to light supporting the conclusion that midHolocene estuarine environments along the Gulf and Atlantic coasts provided an aquatic biome as rich as that supplied by the interior river systems of the Midcontinent. Earlier, it was assumed that rapid sea level transgression and drier conditions conspired to inhibit extensive estuarine development. Offshore sampling has consistently identified estuarine deposits dating as far back as 8,000 to 10,000 BP at numerous locations from Virginia to Texas (Blanton 1996, Russo 1996a:179). Although much of the archaeological record is submerged, mounting evidence of intensive mid-Holocene occupation along the coastal fringe comparable to the SMA settlements of the Midcontinent has been documented. Due to its location on a relatively unique topographically prominent Pleistocene parabolic dune, the Horr’s Island site represents one of the best examples of this former estuarine settlement system (Russo 1996a:194-195). The village consists of a 5 m deep oyster midden containing evidence of yearround collection of slatwater shellfish and small fish species. An arcuate shell ridge encircles the village midden, the latter of which extends over a distance of about 300 m. At least three and possibly four purposefully constructed ceremonial shell and sand mounds ranging from 1.5 to 6 m in height are situated around the perimeter of the midden deposit. Corrected radiocarbon dates indicate construction dates of 5,200 to 4,600 BP for the mounds. Additional shell midden of shallower depth abuts the east and west ends of the core of the site for another 200 to 300 m. The village is characterized by a maximal accumulation of organically stained living surfaces spanning about 4 m of deposit height. Limited excavations identified the posthole patterns of numerous struc- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 29 Chapter 3. Culture History tures averaging about 3 m in diameter with hearths and pits. Russo contends that converging lines of evidence suggest that the Horr’s Island was a permanently occupied village. More than 60 ceremonial mound groups in Arkansas, Louisiana, Mississippi and Florida are suspected to date to the mid-Holocene (Russo 1996b), which suggests that other Gulf Coast localities may contain submerged sites comparable to Horr’s Island. The evidence at present is only spotty. The greatest concentration of presumptive mid-Holocene mounds occurs in Louisiana. One of the most spectacular of these mound sites is Watson Break in northeast Louisiana, which is comparable in age and structure to the Horr’s Island site. It consists of a circular ridge connecting at least 10 mounds measuring more than 300 m in diameter (Russo 1996a:279-280). The tallest mound stands over 7 m in height. Radiocarbon dates from the ridge and one of the mounds range in age from 5,850 to 4,150 BP. The fact that Watson Break is not situated in an estuarine environment, but in the floodplain of the Ouachita River indicates that freshwater aquatic biomes of interior Gulf Coast streams were also sufficient to support intensive occupation and ceremonial mound building. Available radiocarbon dates associated with the confirmed mounds range in age from around 5,700 BP to as late as about 2,500 BP. Russo (1996b:260) states that the mounds are variably massive “tumuli that may contain human burials, ceremonial objects, tombs, earthen platforms, and structures, reflecting construction episodes and shapes indicating that mound construction was intentional and purposeful for activities unrelated to or beyond the simple disposal of refuse and other mundane activities” (Russo 1996a:260). No evidence of domesticated crops has yet been recovered from these mound sites and it is argued that the builders subsisted primarily on a mixture of aquatic (i.e. freshwater and saltwater shellfish, fish and turtles) and terrestrial faunal resources. More ethnobotanical sampling from these sites will be required to prove that cultivation of incipient domesticates was not an important element of Gulf Coast mid-Holocene mound builders. Monumental architecture with a series of building stages like the ones Russo (1996b:271-273) describes for the Horr’s Island mounds is strong evidence that some late midHolocene groups on the Gulf Coastal Plain may have developed hierarchically ranked societies of Chiefdom level organization. The later Poverty Point site on the Macon Ridge in northeastern Louisiana, which is dated between about 3,300 and 3,000 BP represents the ultimate culmination of this system of mound builders (Gibson 1996). Large-scale exploitation of estuarine biomes by cultures on the South Carolina and Georgia coasts apparently does not occur until about 4,200 BP. Marine shell middens and larger features known as shell rings associated with the pottery producing Stallings Island and Thom’s Creek cultures are extensively distributed along the Sea Islands region of the coast, but does not extend in significant densities north of the Santee River in central South Carolina (Sassaman 1993b). Russo (1996a:186-187) notes, however, that there are numerous aceramic estuarine shell deposits below these occupations that could potentially contain earlier occupations. He further notes that several radiocarbon samples from coastal archaeological sites evidence possible earlier mid-Holocene occupation. An answer to this question awaits more systematic offshore testing. Sassaman (1993b) suggests that intensive occupation on the coast of South Carolina and Georgia began with movements down the river systems of the Atlantic Slope from populations assigned to the SMA (i.e. the Stallings Island Culture on the Savannah River). This migration was supposedly stimulated by the initial explosion of estuarine habitats as sea level rise lessened and stabilized (see Brooks et al. 1989). The character of the coastal occupation is speculated to represent year-round settlement (DePratter 1979), but there has been only minimal archaeological investigation at hypothesized habitation sites. The significance of shell rings has been the subject of great debate. Some have argued that they grew naturally from shellfish refuse accumulations and served only as bases for habitation (Trinkley 1980). Others have argued that they were ceremonial features (Waring 1968; Michie 1979). Cable’s (Cable et al. 1993) archival re-analysis of the Spanish Mount data, a shell ring or shell mound located on Edisto Island, demonstrates rather conclusively that this feature was not built haphazardly from serial dumping episodes but rather was built-up laterally and in layers as would be characteristic of an intentionally constructed monument. The argument that the shell rings Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 30 Chapter 3. Culture History represent intentionally constructed monuments gains more credence now that the large array of mid-Holocene earthen and shell ceremonial mounds have been identified and confirmed along the Gulf Coast (see Russo 1996b; Sassaman and Anderson 1994:157). Most models of mid-Holocene settlement patterns stipulate that populations dispersed from their riverine and coastal aggregation sites into the hinterlands during certain seasons of the year (Sassaman and Ledbetter 1996:86-87; B. Smith 1986). Contemporary sites in the upland zones tend to be small and clearly indicate that fissioning groups were small, consisting of elemental family units of one or a few nuclear families or extended families. Seasonally dispersed groups, then, probably lived in social units very similar in structure to their forager ancestors who distributed themselves in a uniform manner across the landscape to minimize resource procurement costs (Johnson and Earl 1987:314-315). Because of the small and mundane nature of such sites, though, information about them has been slow to develop and consequently data on group size, duration of stay, group composition, subsistence strategies and reoccupation patterns are sparse. Sassaman and Ledbetter (1996:87) observe that the necessary continuity of occupation was probably not repeated enough at these upland sites to justify prevalent construction of substantial architecture, but they cite several Late Archaic upland sites that have actually yielded such evidence. At the Millbranch Site in Warren County, Georgia a large midden filled pithouse was excavated (Ledbetter 1995). The foundation pit measured 4 x 5 m and contained a hearth and a series of substantial perimeter postholes. Radiocarbon assays date the house to around 3,850 BP. Several other upland Late Archaic pihouses have been identified in Tennessee at the Cherry (Lewis and Kneberg 1947) and Spring Creek (Peterson 1973) sites and at the F. L. Brinkley Midden site (Otinger et al. 1982) in northeast Mississippi. Thus, by about 3,500 to 4,000 BP, more permanent architecture was being constructed on upland sites across a large area of the Southeast. This would suggest that some upland sites were either being occupied year-round as isolated residences or hamlets or that family groups were returning to the same residential sites year after year due to the emergence of usufruct rights on particular upland tracts brought about by population packing and territorial contraction. Conditioned reoccupation of the same location, then, would have justified the increased labor investment necessary to construct permanent houses. Fort Bragg is situated on the margin of the extensive record of mid-Holocene intensification and increased group sizes. Intensive exploitation of freshwater shellfish resources has not been documented on the Atlantic Slope in the river systems north of the Savannah River. Although intensive harvesting of saltwater shellfish is evident along the North and South Carolina coasts, almost all of these middens derive from the later Woodland and Mississippian periods north of the Santee River. Moreover, there has not yet been any evidence produced to suggest that these cultures were encouraging or cultivating quasi-domesticated crops. The mid-Holocene archaeological record of this region appears to reflect simpler forager systems characterized by high residential mobility, dispersed populations and subsistence strategies based entirely on wild resources (Blanton and Sassaman 1989; Cable 1982, 2009; Cable and Cantley 1998, 2005a; Cable et al. 1998; Cantley and Cable 2002; Sassaman 1983; Sassaman and Anderson 1994). Dense concentrations of Middle and Late Archaic debris have been documented in a number of Riverine settings in the Carolinas, however, including the Doerschuk site on the Yadkin River in the Uwharrie Mountains of North Carolina (Coe 1964), the Gaston site on the Roanoke River in Halifax County, North Carolina (Coe 1964), the Haw River in Chatham County, North Carolina (Claggett and Cable 1982), the Nipper Creek site on a tributary of Broad River in Piedmont South Carolina (Wetmore and Goodyear 1986) and the Manning site on the Congaree River terrace near Columbia, South Carolina (Anderson 1979). Perhaps the best evidence of intensified, semi-sedentary habitation along a major stream is provided by the Gaston site Late Archaic component. It included a stained midden, stone-lined hearths, high artifact density, steatite vessel fragments and full-grooved axes. Coe (1964: 119) observed that these traits suggested a “larger group occupying the site over a longer continuous period than had been true of the earlier periods.” Structural remains have not been identified on any of the mid-Holocene deposits at these sites, however, and the special role Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 31 of riverine sites in the region has yet to be defined. Available evidence suggests that riverine localities may have been utilized more intensively than upland ones, but whether this factor also indicates that larger social formations such as multi-family aggregations met in the river valleys is not known. PERIOD Ceramic Series 500 PROTOHISTORIC Undefined Hanover II Series 1000 LATE WOODLAND Chapter 3. Culture History Perhaps the most significant development distinguishing the early portion of the Woodland period from the Late Archaic was the full-blown emergence of what Ford (1985:347-349) refers to as the Eastern Agricultural Complex throughout many regions of the southeastern United States. This complex was primarily composed of indigenous species of seed-producing commensal weeds including sunflower (Helianthus annus) sumpweed (Iva annua var. macrocarpa), goosefoot (Chenopodium bushianum), maygrass (Phalaris caroliniana), knotweed (Polygynum erectum L.), little barley (Hordeum pusillum Nutt.), and giant ragweed (Ambrosia trifida). The former three exhibit signs of domestication (i.e., enlarged achenes, decreased seed coat thickness, brittle rachis, etc.) by the terminal phases of the Late Archaic, while the others appear to have been intentionally transported and cultivated in Late Archaic and Woodland contexts (Cowan 1984). The bottle gourd and squash, as discussed previously, represented very early Mexican domesticate introductions and along with this seed complex comprised the basis of the Early Woodland gardening subsystem. Maize was a relatively late entrant into the eastern woodlands, with an initial date of appearance of only about A.D. 300 (Yarnell and Black 1985). In spite of the rather Hanover I Series MIDDLE WOODLAND 1500 2000 EARLY WOODLAND 2500 Cape Fear, Yadkin, Deptford Series New River Series 3000 3500 LATE ARCHAIC The Woodland period in North Carolina spans the time interval from 1,000 B.C. to A.D. 1500 and is divided into “Early” (1,000-300 B.C.), “Middle” (300 B.C.-A.D. 800), and “Late” (A.D. 800-1500) subperiods. In most regions of the Southeast, the Late Archaic-Woodland transition is seen as continuity with the emergent patterns of intensification gradually building in magnitude (Steponaitis 1986:378-379). These patterns consisted of an increased emphasis on gardening and exploitation of seeds, greater adjustments toward sedentism, and elaboration of mortuary ritual and political control. Uncalibrated Radiocarbon Years B.P. WOODLAND PERIOD (1,000 B.C.- A.D. 1500) Thom’s Creek Series Stallings Series Figure 6. Ceramic Sequence for Sandhills Region of North Carolina (data derived from Herbert 2003). substantial evidence for agriculture, isotopic analyses of Early and Middle Woodland skeletal populations do not indicate a dependence on cultigens (Bender et al. 1981, Boutton et al. 1984, Van der Merwe and Vogel 1978). Early cultivation in the eastern woodlands may very well have not represented an economically more important food source than a number of wild food subsystems such as nuts, aquatic resources, and deer (Steponaitis 1986:379). Evidence gathered in the southeastern United States for sturdy, if not permanently occupied, houses is abundant for the Early Woodland Period. Along the Gulf and Atlantic coasts, the massive shell middens of the Late Archaic subperiod are replaced by Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 32 Chapter 3. Culture History more diffuse scatters of shell that are interpreted as the refuse from individual households (Milanich and Fairbanks 1980). Settlements appear to be small, ranging in size from about 5 to 10 households, and cover less than a hectare in area. Similarly small Early Woodland settlements with ample remains of houses have been investigated in the interior Southeast and in the mountains and Piedmont of the Atlantic Slope (Faulkner and McCollough 1978, Keel 1976, Kline et al. 1982, McNutt and Weaver 1983). Generally these settlements are viewed as seasonal in nature, but annually re-occupied. Some late Middle Woodland settlements in the interior Southeast have been hypothesized to represent permanent villages due to the apparent pairing of cold- and warm-season structures, sturdier house construction, larger and more dense midden formation, and the absence of “cachetype” storage facilities (Steponaitis 1986:381). To date, evidence of Woodland period villages or large, single component sites possibly representing occupation(s) by a single or related groups is lacking at Fort Bragg (Clement et al. 1997). Instead, the emerging picture of the region’s occupational history can be characterized as one of seasonal ephemeral campsites occupied by single and/or multiple households and short-duration special purpose sites. This pattern of land-use does not appear to change significantly throughout the various Woodland periods. The Sandhills region of North Carolina has yielded little data useful for constructing a local artifact sequence or chronology. Instead, archaeologists have relied on diagnostic artifacts, projectile points and ceramics described for the Piedmont and Coastal Plain provinces, to reconstruct the precontact landscape of the Sandhills. A notable exception to this general condition is represented by the thermoluminescence dating of Woodland period ceramics conducted by Herbert (2003) upon which the regional sequence presented in Figure 6 is based. The application of the TL technique to the dating of ceramics is controversial, but the reults of the program to date are promising. One could argue with the interpretations of the datasets, but a roughly coherent chronological sequence of dates have been shown to be associated with the predicted trajectory of local ceramic development. In the Piedmont and Blue Ridge provinces, it has been argued that the Gypsy-Stemmed and Swan- nanoa-Stemmed projectile points evolved out of the earlier, Late Archaic Savannah River point type tradition (Oliver 1985). Gypsy points were recovered in the Early Woodland deposits at Doerschuk and Gaston sites (Oliver 1985), while Swannanoa points have been identified in similarly dated deposits at the Warren Wilson site (Keel 1976, Oliver 1985). The Early Woodland deposits at Doerschuk contained not only Gypsy-Stemmed points, but also Badin Triangular points and Badin ceramics. The Early Woodland ceramic assemblage for the eastern Piedmont and Coastal Plain regions is not well understood. Coe (1964:27-30) defined Badin ceramics as exemplifying cord-marked, fabricimpressed, or plain surface treatments with a hard, compact paste with a very fine river sand temper. Unfortunately, the description of Badin ceramics is very similar to the Middle Woodland Vincent series pottery of Virginia’s Roanoke Basin and the New River and Deep Creek ceramics of the Coastal Plain. Since no absolute dates exist to place the Badin ceramic assemblage within a secure temporal framework, questions persist as to its placement in the local cultural sequence. East of the Piedmont on the Coastal Plain, Early Woodland period assemblages include Thom’s Creek, Hamp’s Landing, New River, and Deep Creek ceramics. Thom’s Creek pottery in North Carolina is fairly restricted to the lower southeast corner of the state in Brunswick and New Hanover counties; however, a few specimens have been recovered as far north and west as the Fort Bragg vicinity (Griffin et al. 2001). Hamp’s Landing series is differentiated from other ceramics by its limestone or marl-tempered paste (Hargrove 1993, Hargrove and Eastmean 1997, Hebert 1997, Hebert and Mathis 1996). The series has been recovered in features with stratigraphic contexts that are intermediate between Early Woodland Thom’s Creek and Middle Woodland Hanover pottery horizons along the lower Cape Fear drainage and along the coastal margin as far north as Carteret County. Radiocarbon dates ranging between ca. 2000 to 500 B.C have been obtained from features containing these ceramics. It is generally recognized today that the Oak Island series originally defined as a shell-tempered series by South (1960), is actually limestone-tempered and as a consequence this series nomenclature has been retired in favor of inclusion within the rubric of the Hamp’s Landing series (Herbert 2010, personal communication; Mathis 1999). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 33 Chapter 3. Culture History To date, no Hamp’s Landing ceramics have been recovered from the Fort Bragg project vicinity. these ceramics occurred late in the Yadkin sequence. Once again, no radiocarbon dates are available. The remaining two Early Woodland pottery types identified within the North Carolina Coastal Plain are the New River and Deep Creek ceramics. These ceramic types occur south and north of the Neuse River, respectively. These two Coastal Plain types share the same surface treatments and are believed to be contemporaneous with Piedmont Badin ceramics. Three TL samples taken from New River sherds suggest a date range from about 1,200 to 400 B.C. for this ceramic series (Herbert 2003). Throughout the Southeast and Midwest the later part of the Early Woodland and the Middle Woodland periods mark the beginnings of a very distinctive mortuary complex(es) characterized by the incorporation of burial mound features. These features are commonly regarded as evidence for the emergence of segmentary lineages, systems of ranked social status, and “big-man” leadership roles (Brose and Greber 1979, Smith 1986:45-50, Steponaitis 1986:382-383). Typically, such systems are unstable and particularistic in the details of integration and the regional diversity of mortuary ritual evinced in these burial mounds is generally regarded as a reflection of these characteristics. During the Middle Woodland period (300 B.C.-A.D. 800) differences between the northern and southern regions of the North Carolina Coastal Plain continue to be expressed in ceramic typology. The North Carolina northern Coastal Plain region is the Mount Pleasant series. This series will not be described since it is located some distance from the project area. Along the southern coastal region, Phelps (1983) incorporates both the Hanover (grogtempered) and Cape Fear (sand-tempered) series into the Cape Fear phase. More recently, the Hamp’s Landing series has been added to the southern coastal sequence. All of these ceramic series exhibit cord marked, fabric or net-impressed surface treatments. Unfortunately, few absolute dates exist to aid in the placement of these ceramics into a temporal sequence. Radiocarbon dates from South and North Carolina suggests a date range from 200 B.C. to A.D. 500 for the Hanover series, while only one available date for Cape Fear ceramics was calibrated to A.D. 1028 (Eastman 1994). Yadkin series ceramics mark the advent of the Middle Woodland period in the Piedmont region. This series was defined by Coe (1964) and represents an evolutionary trend from the earlier Badin ceramic series. The major difference between these two pottery types is that the Yadkin series incorporates the use of angular fragments of crushed quartz into the paste. Otherwise the surface treatments remain the same with mostly fabric-impressed, cord-marked, and some check-stamped examples occurring at the Doerschuk site (Davis 1987). At Town Creek, Coe (1995) added simple stamping as a new surface treatment to the Yadkin series. The presence of simple stamping in the Town Creek assemblage suggested to Coe that In North Carolina these mounds are lowlying, ranging between about 0.6 and 1.2 meters in height, circular to oval in shape, and vary between about 6 and 18 meters in diameter. J. A. Holmes, a geologist with the Department of the Interior, was the first to investigate and report on excavations into several of these mounds in 1883 (see MacCord 1966). His major area of concentration was in Duplin County, southeast of the Falls Lake region. Charles Peabody (1910) also excavated one of these features in Cumberland County in the early twentieth century. A number of other mounds including the McFayden Mound in Brunswick County (South 1966), the McLean Mound in Cumberland County (MacCord 1966) and the Red Springs (Keel 1970) and Buie (Wetmore 1978) mounds in Robeson County have been the subject of more recent investigations. Although not recognized as such, the Holiday “ossuary” site near Galivants Ferry on the Little Pee Dee River in Horry County, South Carolina may also represent the remnants of a sand burial mound (Rathbun 1989). Burials are of three types in these North Carolina mounds, all of which commonly occur in a single mound. These include cremations, bundle burials of varying degrees of completeness and flexed inhumations. Mound size tends to correlate with burial population. Stewart (1966:69) estimates that the Mclean Mound, which was about 18 m in diameter, contained about 500 individuals, while Holmes reported only 8 skeletons from a mound of about 7 m in diameter in Duplin County of which he excavated Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 34 Chapter 3. Culture History one-half. Other than the fact that these cemeteries are mounded, they exhibit some burial patterns not unlike those of Iroquoise and Algonkian ossuaries in the Middle Atlantic states according to Stewart (1966). The burial populations from both the sand mounds and the ossuaries reveal an under-representation of children, especially of infants. Moreover, both contain smaller numbers of adult males than females. Cremation is also a shared trait, although no ossuary has yet produced as many as the 32 individuals identified at the Mclean Mound. Stewart’s cranial measurements also suggest that the McLean Mound population is more closely aligned with the Middle Atlantic physical type than that of southerly groups. Herbert (personal communication, 2005), however, has pointed out that there are very real differences in burial practices between mounds and ossuaries. At the McClean Mound he argues that single individual interment was the mode and bodies were accumulated over a period of several hundred years. Colington phase (Algonkian) ossuaries, by contrast, appear to contain multiple interments associated with single events. Moreover, Cashie phase (Iroquoian) ossuaries tend to contain clusters of multiple individuals interred in numerous events. Very little information is available to discuss internal structural patterning in these mounds as the quality of reporting and the degree of preservation are both very poor. Holmes (1883) mentioned in several different places in his notes that he did not detect any submound disturbance, but he did observe a ringed depression around one mound that might resemble the borrow pit formations excavated around the central tomb features of the coastal Georgia mounds (see Thomas and Larsen 1979: 35). Peabody observed what he called a “sod line” under one burial mound, which he attributed to a recent forest fire. Alternatively, this may represent a premound clear burn patch that initialized mound construction, as discussed above for mounds on the Georgia coast. No mention has ever been made of central tombs or differentiated shell and/or sand lenses in the North Carolina mounds, but this may be a factor of the intense pot-hunting damage that has accrued over the years and the lack of modern excavation studies. On the basis of a very broad correlation of the burial mound trait in the eastern woodlands, Phelps (1983:35) has assigned the sand burial mounds in North Carolina to the Middle Woodland period. The single radiocarbon date of A.D. 970 from the McLean Mound (MacCord 1966:17) suggests either a very late Middle Woodland or early Late Woodland time range for this particular mound. The MacFayden Mound near Wilmington, NC contains predominantly Cape Fear series ceramics (South 1966) in its fill, which would correlate, in general, with the Wilmington or St. Catherines phases on the Georgia Coast (DePratter 1979). Experience from excavations in the Georgia Sea Island region (Thomas and Larsen 1979), however, would indicate that dating the use of the burial mounds from sherd inclusions can be very problematic. Since these mounds appear to be built-up from nearby fill, which often contains earlier midden, the best that can be said of the MacFayden Mound is that it cannot date earlier than the Cape Fear series, but can be younger. A recent overview of the cultural associations of sand burial mounds in North Carolina by Irwin et al. 1999), strongly suggests that they are affiliated with Late Woodland cultures. Across the eastern woodlands, the Late Woodland period (A.D. 800-1500) has often been characterized as a time of cultural decline, principally because of the apparent simplification of the burial complexes (Griffin 1952). This view is highly biased, in many respects, toward the events surrounding the collapse of the Hopewell Interaction sphere in the Midwest where dramatic declines in the diversity and “exotic” character of grave offerings occurred (Brose and Greber 1979). Over many other areas of the Southeast, however, the differences are less extreme, and, if anything, reflect a developmental continuum. The burial mound sequence of the Georgia coast exemplifies such a trajectory (Cable et al. 1991, Caldwell and McCann 1941, Larsen and Thomas 1982, Thomas and Larsen 1979). It is nevertheless generally held that this subperiod witnessed a decline in “big-man” authority systems, primarily as a response to population expansion, filling, and dispersal (Smith 1986:5253). Settlements apparently remained small and subsistence systems changed little, with the possible exception of an increased emphasis on maize agriculture. Ethnobotanical analyses of Late Woodland sites from the North Carolina Piedmont indicate that maize agriculture was well established by this time Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 35 Chapter 3. Culture History interval. At the Donnaha site, a large village in the upper Yadkin River Valley, corn was ubiquitous in the flotation samples and comprised over 40 percent of the recovered macrobotanical remains (Mikell 1987:11). Other cultigens in the samples included the common bean (Phaseolus vulgaris), another Mexican import, squash, and sumpweed. A wide array of wild plant remains were also recovered from the site including hickory nuts, acorns, grape, plum maypops, hackberry, walnut, hazelnut, and butternut, but the ubiquity and relative abundance of cultigens at the site demonstrates a heavy reliance on agricultural food production (Mikell 1987). This emphasis on agriculture is evidently seen at smaller, seasonal camps or hamlets of this time interval. At site 31Am278 (dated to around 800 rcy B.P.) a small Piedmont upland site in Alamance County, North Carolina, maize remains comprised 6 percent of the macroplant sample from an apparent storage pit (Cantley and Raymer 1990:68-75). The overall contribution of maize is considerably less than that of the Donnaha floral assemblage and this led the authors to suggest that the corn might have been transported to the site from a base camp. Goosefoot is also present in the samples from the pit and may represent yet another cultigen in the sample. Regardless of the function of the site (i.e., wild food procurement camp, farmstead, etc.), the presence of maize and other cultigens on small as well as large sites attests to the overall importance of agriculture during the Late Woodland subperiod in North Carolina. In spite of Ward’s (1983:72-73) reservations about the importance of agriculture in Woodland subsistence systems, recent ethnobotanical analyses indicate that it would be counterproductive to de-emphasize the role of maize agriculture during the Late Woodland. Late Woodland material culture in the Fort Bragg project area is expected to be influenced most by the Siouan-speaking inhabitants of the southern Coastal Plain as well as groups occupying the eastern Piedmont region. South (1976) characterized the ceramics during this period as belonging to the Oak Island phase. Oak Island ceramics are shelltempered and exhibit cord marked, net-impressed, plain, and fabric-impressed surface treatments. Phelps (1983:48) suggests that the increase in fabricimpressing and the presence of simple stamping may align the Oak Island phase with the Colington phase centered along northern Coastal region. The Uwhar- rie phase characterizes the Late Woodland Piedmont ceramics. This pottery is differentiated from the earlier Yadkin pottery by larger vessel sizes, scraped interior surfaces, and a higher proportion of crushed quartz in the paste that often times protrudes though both vessel walls (Coe 1952, 1995). Exterior surface treatments typically associated with this pottery type include net-impressed, plain, fabric-impressed, and cord-marked decorative types. MISSISSIPPIAN PERIOD (A.D. 1200 -1500) The hallmark of South Appalachian ceramic industries is, in the earlier phases at least, complicated stamped surface treatments (Ferguson 1971). This is contrasted by a continuance of net impressed, cord marked, fabric impressed, and simple stamped treatments in the more northerly Late Woodland ceramic assemblages (Davis 1987:211). Throughout the Southeast, the presence of this complex is regarded as a manifestation of chiefdom level organization and highly ranked or stratified social units (Steponaitis 1986:391, Smith 1986:55-56). Some of these societies have also been speculated to represent macropolities and to have been organized along the lines of paramount chiefdoms, with a series of internally centralized, quasi-autonomous polities loosely integrated into a hegemony around a paramount center. The most northerly expression of the South Appalachian Mississippian culture is the Pee Dee manifestations located approximately 40 miles west of Fort Bragg in the middle Yadkin River Valley (Coe 1952). Coe (1964:124) saw the South Appalachian manifestations of North Carolina as intrusive into the area and dated this event as occurring around A.D. 1500. The radiocarbon dates from the Town Creek Mound, however, indicate that the Pee Dee culture was present in North Carolina in the 13th and 14th centuries A.D. (Dickens 1970:79). Moreover, the ceramic time-line of the Pee Dee pottery correlates well with the trajectory of change in other South Appalachian regions that occurred between about A.D. 1150 and A.D. 1400 (DePratter and Judge 1986, Hally and Rudolph 1986). These lines of evidence strongly suggest that the Pee Dee culture appeared in North Carolina around A.D. 1200. Whether it represents an intrusion into this area is not known. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 36 Chapter 3. Culture History Pee Dee ceramics are generally stamped with a carved paddle or smoothed and burnished. While a small percentage of the vessels exhibit simple stamping or check stamping, the majority of stamped ceramics exhibit distinctive curvilinear and rectilinear designs. Decorations usually found on burnished vessels include applied rosettes, pellets, or punctations around the shoulder. Textile and corncobimpressions are present, but rarely occur in the Pee Dee ceramic assemblage. To date, there is only sparse evidence of Pee Dee occupations on Fort Bragg. Another potential Mississippian mound center is Buie Mound in Robeson County. This site contains a very anomalous sherd assemblage in its fill that Wetmore (1978:65) ascribes to the Pee Dee and Lamar periods. The predominant ceramic type here is a sand-tempered burnished plain ware. Inspection of the pictures of the associated decorated types in the report indicates the presence of features (ie. sloppy incised complicated stamped designs, below-the-rim reed punctations, and shoulder and rim ticks) that would correlate with a number of Lamar assemblages in other regions including the Mulberry phase (A.D. 1450-1550) in the Wateree Valley, South Carolina (DePratter and Judge 1986), the Dyar phase (A.D. 1450-1600) in northern Georgia, and terminal occupation ceramics from Town Creek (Reid 1967). If this association is valid, and the ubiquitous distribution of plain burnished and incised sherds in the fill of the mound supports this interpretation, then the Buie Mound represents the latest known evidence of the persistence of the burial mound phenomenon in the Southeast. PROTOHISTORIC AND EARLY HISTORIC NATIVE AMERICANS (A.D. 1500-1650) In recent years it has become apparent that the transition from late precontact to historic adaptations was accompanied by some rather dramatic changes in the settlement pattern, health status, social organization, and even regional location of Southeastern Indian groups (Smith 1987). Even before Europeans had established a strong presence in the Southeast, the effects of disease and demographic disruption were felt throughout the region. In North Carolina, Merrell (1987:19-21) identifies four stages of contact. The initial stage extended from about A.D. 1525 to 1625 and consisted of primarily indirect contacts with the Spanish and English involving the exchange of material goods and disease epidemics. The next stage was a relatively short interval, lasting about 50 years. During this time numerous clashes with Piedmont groups were undertaken by the English and European settlement extended into the southern Piedmont. The next brief stage, ending in the early eighteenth century, signaled an end to the semblance of a preserved Native American cultural system and the beginning of the historic period. This stage saw major conflicts throughout the Carolinas and numerous displacements and extinctions of native groups. As of yet, a confirmed Protohistoric occupation has not been recognized in the Sandhills area of Fort Bragg. It is quite possible that the region was vacant during this period and there are many examples of expansive regional abandonments and extreme depopulation events occurring around AD 1450 in the greater Southeast. The best known of these is the “Vacant Quarter” proposed by Stephen Williams (1983, 1990) for an area surrounding the Ohio-Mississippi River confluence that spread across the American Bottom in Illinois and southward to the Missouri Bootheel and eastward up the lower Ohio River Valley into Indiana. Williams based his hypothesis on the absence of protohistoric horizon markers in this region and subsequent investigations in various subregions of the Vacant Quarter (Cobb and Butler 2002; Price and Price 1990) support this hypothesis. Similar regional “abandonments”, however, appear to have occurred at other, more proximate locations to the North Carolina Sandhills. For instance, Anderson (1994) demonstrates that major sections of the Savannah River Valley below the Cherokee towns and all of the major mound sites were abandoned prior to AD 1500. Extensive survey in the Francis Marion National Forest on the central coast of South Carolina has resulted in the identification of another regional abandonment of Mississippian populations between AD1450 and 1500 (Cable 2002:474-485). Explanations for these vacated territories range from disease pandemics to environmental change/drought (Anderson 1994; Williams 1990), but it is clear that from a broader perspective these events can more properly be viewed as population rearrangements, as Williams (1990:176-179) describes it, as other regions flour- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 37 Chapter 3. Culture History ished. Clearly, the Appalachian Summit and foothills and the northern Piedmont and Coast of North Carolina supported large protohistoric horizon and contact populations (Davis 2002; Smith 2002; Ward and Davis 1999), suggesting that environmental change might pose a more likely explanation for immediately precontact population shifts. At present, however, there is simply not adequate evidence to determine if the Sandhills of North Carolina experienced a regional abandonment in the late precontact time frame. The lack of archaeological evidence for such populations may just as easily be explained by the underdeveloped state of the ceramic sequence. The current sequence for the region suggests that after about AD 1000 the grog-tempered Hanover series became singularly predominant in local assemblages and, based on dated contexts from surrounding areas, the pattern may have extended into the fifteenth century (Herbert 2003). The later series variant, Hanover II, is distinguished from the earlier variant, Hanover I, by temper configuration (Herbert 2003:159-160). The Hanover I series (AD 400 to 1000) is characterized by the presence of a small amount of finely crushed grog added to a primarily sand-tempered paste, while the Hanover II series (after AD 1000) consists of a paste dominated by grog with sparse medium background sand. Hanover I is characterized by fabric impressed, cord marked and check stamped surface treatments, while Hanover II ceramics are overwhelmingly represented by fabric impressed surfaces (Herbert 2003:191). In this attribute, Hanover II correlates well with the Colington/ Townsend Fabric Impressed type that dominates the early Late Woodland assemblages of the Algonquian settlements on the northern and central North Carolina coast. The precise temporal span of Hanover II series has not been established, but radiocarbon data tabulated by Herbert (2003:195-199) indicates the likelihood that it extends at least into the early AD1400s and two TL-dates on specifically Hanover II Fabric Impressed yielded means of AD 1481 and 1503 respectively. However, some sherds classified as Hanover I series produced equally late TL- and radiocarbon dates, indicating that the typological system may not be sufficient for serriating grog-tempered ceramics along a temporal trajectory at present. Complicating matters is the apparent presence of a yet to be recognized and defined Late Wood- land sand-tempered series. At the McLean and Buie Mounds, sand-tempered plain and fabric impressed wares are found in association with Hanover II ceramics (Herbert 2003:193). This sand-tempered ware has been uncritically assigned to the Cape Fear series, but as Herbert notes, the surface treatments would be unusual for the Cape Fear series, which is dominated by cord marking. Herbert further observes that the disparity in grog-temper proportions between the two mounds suggests that the ceramic assemblages from the mounds may represent a sequence, with the Buie Mound occupying a later chronological position. The McLean Mound contains 86 percent grog-tempered sherds, while the Buie mound is dominated by sandtempered wares and contains only 12 percent grogtempered pottery. Moreover, the Buie Mound assemblage is dominated by sand-tempered burnished plain (79 percent) and various examples in the collection exhibit below-the-rim reed punctations and shoulder and rim ticks, all characteristics of Mississippian (Reid 1967) and Late Woodland assemblages of the Piedmont (Ward and Davis 1999:108), especially the Dan River series. Wetmore (1978:44-45) also reports that about 6 percent of the burnished plain sherds from the Buie Mound exhibit crude and narrow incised decorations including motifs with series of three and four parallel lines and appended or incased semicircles and, also, curved line motifs. These incised motif characteristics are consistent with patterns that emerge in Lamar assemblages throughout the Southeast after about AD1450 (Hally 1994:147), parallel lines narrow, increase in number and are incorporated into more complex designs with curvilinear elements. The Buie Mound assemblage, then, appears to correlate most closely with the Caraway phase. As Ward and Davis (1999:137) note, a formal description of Caraway ceramics has not been published, but they describe it as the “southern Piedmont’s version of the widespread Lamar style.” They indicate that Caraway assemblages are dominated by smoothed and burnished surfaces with minor contributions of simple stamping, complicated stamped, brushed, cob impressed and net impressed surface finishes. Drawing on their excavations at Lower Saratown, the Wall site and the Hairston site, they further observe that the smoothed and burnished types probably date later than the stamped and impressed types. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 38 Chapter 3. Culture History Clearly, the character of the Buie Mound assemblage suggests close ties to the Piedmont Caraway phase or to a post-Leak phase (see Boudreaux 2005) Mississippian Lamar tradition to the south. An issue that should be resolved, then, is whether it constitutes a cultural continuity with a terminal Hanover II phase as represented by the McLean Mound assemblage, or whether it signifies a late precontact/ protohistoric intrusion from the south or the northwest Piedmont. If the latter hypothesis holds, then an assemblage dominated by a sand-tempered ceramic series would be predicted to replace the McLean assemblage containing fabric impressed and possibly simple stamped surface treatments similar to the sequence trajectory seen in the Cashie (Byrd 1997; Phelps 1983:43-46) and Colington/Townsend (Gardner 1990) series to the north and east. The Paraham Mound site (31RB33), located approximately 45 to 50 miles southeast of Fort Bragg, may contain a ceramic assemblage derived from the Buie Mound assemblage (see Heath and Irwin 2007: 32-36, Wetmore 1969). Wetmore (1969) indicates that although the site name implies the presence of a sand burial mound, the feature so identified was demonstrated to represent a purely natural topographic rise. The Native American ceramics found at the site, however, consisted primarily of sand-tempered plain and burnished plain exterior surface treatments and incised and punctated decorations, which would suggest a clear connection to Lamar or Caraway assemblages. The site also produced a small collection of seventeenth/eighteenth century historic artifacts that may be associated with the Native American occupation. These included dark green bottle glass, blue and white glass trade beads and possibly small metal bells. Based on these potential associations, Heath and Irwin (2007: 33) suggest that the primary Native American ceramic assemblage at the site may be affiliated with either the historic Cape Fear or Waccamaw tribes of the early eighteenth century. It is generally assumed that the indigenous population of North Carolina Sandhills belonged to the eastern Siouan languistic family. However, the sixteenth century configuration of native tribes occupying the region is not established by historic documents. Speculation persists that Francisco Girebillo of the first Ayllon expedition in 1521 may have landed at the mouth of the Pee Dee River in South Carolina and from there travelled inland along the Pee Dee River valley and/or up the Waccamaw River valley, where he made contact with the Waccamaw Indians (Milling 1969:203; Ross 1999:139). According to Swanton (1946:203), the Waccamaw tribe was first identified by Francisco de Chicora, a native guide accompanying Girebillo, as members of the province of Guacaya. In 1670, the Waccamaw inhabited the Waccamaw River and formed close alliances with the Pee Dee and Winyaw tribes. Very little is directly known about the linguistic affiliations of these tribes, but Mooney (1894:76-78) included them within his roster of eastern Siouan speakers. The Winyaw occupied locations on the western side of the Pee Dee River near its mouth, while the Pee Dee tribe was settled on the middle course of the Pee Dee River on its eastern bank, downstream from the Sara near Cheraw, SC. The latter, later to be known as the Cheraw, had migrated extensively across the Piedmont during the sixteenth and seventeenth centuries due to wars with the Iroquois and English, settling for a short time on the east bank of the Pee Dee River opposite Cheraw, SC in Marlboro County by the end of the Yamassee War in 1715 (Mooney 1894:59-60). A native group that is said to have been closely aligned with the larger Waccamaw province was the Cape Fear Indians, who occupied stretches of the Cape Fear River in Brunswick County, NC. Swanton (1946:103) noted that a native name for this group had not been preserved. Only the names of a single village, Necoes, and its chief, Wat Coosa, have survived in the historic record. Three separate attempts by Europeans to settle near the Cape Fear Indians between 1661 and 1665 failed for various reasons. After the Yamassee War, the Cape Fear Indians were removed to South Carolina where they occupied portions of St. Stephens and St. Johns parishes along with remnants of the Pee Dee tribe under the leadership of an individual named “King Johnny.” The modern Lumbee tribe, which occupies portions of Robeson, Scotland and Hoke counties just to the south and east of Fort Bragg, is generally viewed as an amalgamation of displaced contact period tribal remnants. Swanton (1933:5) observes that most of the remnants probably spoke Siouan dialects, including the Cheraw, Keyauwee, Eno, Shakori, Waccamaw and Cape Fear. Minor elements of Algonquian and Iroquoian (Cherokee) speakers might also contribute to the mixture, as well as European. The existence of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 39 Chapter 3. Culture History the Lumbee tribe in the Robeson County region can probably be traced back to at least the 1730s to 1750s and they were recognized legally as a tribal entity by the state of North Carolina as early as 1885 (Ross 1999:115-116). Whether any of the antecedents of these historic entities occupied the Sandhills during the late precontact period is not known. It is likely, however, that the late precontact cultures of the Sandhills were closely aligned with the tribes of the southern tidewater coast. The apparent retirement of the White Oak/Oak Island series provides a basis for merging the coastal zone and Sandhills of the Cape Fear River Valley into a single coherent ceramic tradition during the Late Woodland period. White Oak and Oak Island ceramics were formerly considered a shelltempered correlate of the Late Woodland Townsend/ Colington series, but now they are thought to be representative of an Early to Middle Woodland marltempered series (see Herbert 2003:192-193; Herbert and Mathis 1999; Mathis 1999). The Hanover series, in particular, bridges the archaeological records of these zones (Herbert 2003:231-238) and supplies the corpus of a derivative Late Woodland stratum. The hypothesized McLean Mound assemblage would suggest a transitional phase dominated by Hanover II Fabric Impressed with minority sand-tempered fabric impressed (McLean) followed by a yet unrecognized sand-tempered phase dominated by fabric impressed and later simple stamped surface treatments. River running through Fort Bragg has yet to produce such a candidate. Recognizing short-term Protohistoric occupations in the region will require detailed ceramic analyses to consistently distinguish sandand grog-tempered ceramic series. That potential terminal Late Woodland/Protohistoric occupation exists within the confines of Fort Bragg is demonstrated by the presence of a full range of Late Woodland period projectile point types (Cable 2008:179; Cable and Cantley 2005a:94-97). Excavation of discrete shortterm occupation deposits will provide an opportunity to associate clusters of projectile points types with ceramic vessel concentrations in the future, which will undoubtedly shed light on the cultural chronology of late precontact and protohistoric groups in the region. In some cases, potential terminal Late Woodland and Protohistoric ceramics in the region have most probably been subsumed under the rubric of the Cape Fear series. Complicating these matters on Fort Bragg is that most of the Woodland occupations represent short-term or seasonal settlements by small groups at locations that were reoccupied over long periods. The resultant debris fields would consist of the accumulations formed by a wide range of Woodland phases rather than relatively pure assemblages formed by long-term, sedentary village occupations. Protohistoric and historic contact villages in the Sandhills region, would most likely be located on major stream courses as they are in other regions (Ward and Davis 1999). Potential village sites, then, should be expected to occur along the Cape Fear, Lumber and Little rivers, although relatively extensive archaeological survey and testing along the stretch of Little Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 40 Chapter 4. Research Design Our primary approach to conducting Phase II investigations at Fort Bragg is to build a database that allows the evaluation of the occupation histories of individual sites within the framework of regional settlement analysis. The research design and methodology employed in the conduct of fieldwork and analysis on this project is presented in this chapter. APPROACH TO RESEARCH In general, regional research designs incorporate procedures for three levels of archaeological investigation (Redman 1973:64). These include surveys (reconnaissance and intensive), testing projects, and full-blown excavations. The current project may seem somewhat poorly adapted to this scheme since the primary data collection will consist exclusively of test excavations. If the work is to be integrated into a larger management program, however, it is necessary to develop linkages to surveys on the one hand and excavations on the other. This is particularly crucial when the overwhelming majority of the site evaluations will be made in relation to criterion d of the National Register of Historic Places (36CFR60.4), which is concerned with the scientific value of properties. Because of the complexities and broad scope of most scientific inquiries, probability sampling has come to represent a necessary step in deriving an adequate view of the phenomena under study. Sampling at this scale requires a rigorously controlled design to insure that the targeted variability is captured and described. In the case of regional archaeological approaches, this variability is manifest in two major dimensions, environmental zones and site types. Without an accurate picture of these dimensional associations, it is not possible to adequately assess the role(s) of a particular site within the larger regional system. Nor is it possible to systematically determine if a particular site contains information that is redundant in relation to the existing corpus of eligible sites. The survey data from the Fort, because of their broad geographic range and large sample size, are the basis for identifying the critical dimensional relationships between site type and environmental zone. A site typology, however, must be built and developed in accordance with the kinds of theoretical goals that are selected to guide research. As stated above, we propose to describe and elucidate the regional organization of each of the cultural systems registered in the archaeological record of the Fort and to address issues concerning diachronic change. The themes of historic context identified and developed specifically in Appendix XI of the ICRMP (2007) constitute the subject matter for the investigation. Since these organizational themes are principally concerned with adaptations to the environment by cultural systems, the site typology should be constructed to reflect function and mobility strategies. In addition, because sites are merely the locations of behaviors that may have changed over time, the typology must also be organized at the sub-site level to reflect occupation component patterns. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 41 Chapter 4. Research Design Once this information is synthesized into a regional model of diachronic land use, sites to be tested can be successfully evaluated within the context of a regional database. Models of component function and mobility can be investigated and refined through greater excavation coverage on sites selected for testing and this, in turn, can provide clear directions for data recovery operations on eligible sites. Over the course of the contract, field techniques and analytical procedures will be adjusted as feedback is integrated into the historical and cultural contexts of the Fort. OCCUPATION CLUSTERS Archaeologists have traditionally viewed the site as the primary unit of analysis in settlement pattern studies. This is unfortunate because, in reflection, the site is not the unit that we need to isolate when undertaking settlement reconstruction. In the most general way, we want to observe the relationship between discrete occupations, defined as the uninterrupted use of a place by participants in a cultural system (see Binford 1982:5), and the features of the landscape. Sites, on the other hand, are places in the landscape that have served as the stage for any number of occupations, usually by a long sequence of cultural systems. Consequently, sites consist of agglomerations of occupations that can serve to obscure and confound our understanding of how individual cultural systems operated, if they are allowed to serve as the primary analytical unit. This is not to say that archaeologists naively equate occupations with sites, but our measures of occupation variability tend to be made at the site level and consequently they are unspecific and at times misleading. As Dunnell (1971:150-153) has observed, there are difficult methodological problems posed when attempting to isolate occupations within sites, particularly when we attempt to segregate deposits into assemblages. Moreover, whole-site approaches have been devised that provide some basis for observing long-term land-use patterns (see Binford 1982). However, these approaches lean heavily on synchronicity concepts such as site function (i.e. field camps, residences, special use areas, etc.) and component composition in their construction. In monotonous situations where function shifts little with reoccupation, site level characterization can be useful, but in dynamic, multi-component situations, cultural and chronological relationships will be obscured. The “occupation cluster” concept is used to organize component characterization in the Fort Bragg database. The concept has been used successfully in South Carolina (see Cable et al. 1994: 69-74 and 153-170; Cliff and Cable 1999: 413-433) and its application provides a basis for systematically unpacking sites to isolate component structures with survey and testing data. The occupation cluster represents a discrete concentration of diagnostic artifacts of a particular culture historic phase or period. A single component may be comprised of any number of occupation clusters and these clusters can represent functional, temporal, and/or organizational variability within the component. Moreover, a single occupation cluster may contain temporally offset or overlapping occupations of a single component. Thus, an occupational cluster is an empirically defined spatial unit and does not necessarily represent a discrete, synchronous deposit. It is doubtful that the temporal relationships of occupation clusters comprising a single component can be defined in a satisfactory manner with survey or testing data. However, a question such as this might achieve extreme relevance in organizing a data recovery program. For instance, a series of small clusters might represent re-occupations by single-family or specially comprised task groups or alternatively they might be ultra-contemporaneous members of a multifamily residence. The extant methodology for defining occupation clusters has involved the generation of functionally and/or chronologically sensitive diagnostic artifact density maps from shovel test grids. The accuracy and specificity of the component models produced from this procedure vary with the size of the sampling interval and the types of occupation clusters present. Large, dense clusters are defined with relative ease and with a great degree of confidence. An example of a large component comprised of a series of linked, functionally differentiated occupation clusters is provided by the Salt Pond Plantation historic component (Cable et al. 1994:86-94). Salt Pond Plantation was a small, eighteenth century satellite plantation owned and operated by Nathaniel Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 42 Chapter 4. Research Design Johnson, whose principal residence was Silk Hope Plantation on the Cooper River north of Charleston, SC. Historic records indicate that the plantation was permanently inhabited by an overseer and a small number of African slaves. In addition, two wood frame houses were known to exist on the property, one used by the overseer and the other used by the Johnson family and guests when visiting the plantation. A 20-m interval shovel test grid was placed over the presumptive location of the plantation and a peculiar H-shaped pattern of late eighteenth century artifacts measuring about 200 x 275 m was identified by density contour mapping (Figure 7). Subsequent mapping of various functional artifact categories revealed discrete occupation clusters within the component corresponding to two structures, an outdoor kitchen area and a slave row. The residential zone of the plantation also contained extensive precontact occupation (Figure 8). Most of the ceramic material could be associated with the Mississippian period. Nine Mississippian ceramic clusters were defined, four of which were large and appeared to form a circle around a central zone of low artifact density. The physical characteristics of this four-cluster aggregation suggested a compact village, perhaps with an encircling palisade wall and central plaza. Shorter interval testing at Salt Pond would undoubtedly have provided a more accurate definition of the occupation clusters associated with each of the components, and probably would have resulted in the identification of additional small clusters across the site. At every stage of archaeological investigation, we are confronted with problems of sampling and projecting population estimates from the partial data we recover. Statistical methods based on Monte Carlo re-sampling simulations exist which provide a basis for making such estimates from sampling results (Blank, Seiter, and Bruce 2001; Simon 1997; for an archaeological application see Cable and Donaldson 1988). The ability to generate population estimates of small occupation clusters from shovel test samples is a crucial step in implementing the methodology developed here. This is especially true for locations such as Fort Bragg, where the majority of clusters are small and the sampling interval is not close enough to guarantee full identification. Our experience in working on heavily reoccupied precontact sites comprised primarily of small occupation clusters in sandy substrates is that a fairly close correspondence between testing results and excavated occupation clusters is achieved at a shovel test grid interval of 5 m (see Cable and Cantley 1998; Cantley and Cable 2002; Cable et al. 1998). However, individual hunter-gatherer occupations along the Atlantic Slope generally range from 2 to 5 m in diameter, and, under these conditions, 5-m interval shovel testing cannot provide a very effective basis for developing site functional models. Much closer intervals would be required to adequately sample individual occupations. As a heuristic, consider the dimensional characteristics of !Kung bushmen camps as described by Yellen (1977). Figure 9 displays a redrafted plan of a two-household bushman residential camp known as Tanagaba. Two families slept and conducted most of their camp activities within a 5 m diameter area, but not all of this area would be archaeologically visible. The primary artifact deposits associated with the camp occurred between the hut entrances and the adjacent hearths, encompassing debris scatters measuring less than 2.5 m in diameter each. Even if the occupations at Fort Bragg are not functionally equivalent to this bushman campsite, they are structurally and dimensionally very similar. Now consider how a palimpsest scatter of such occupations across a ridge top would relate to shovel test grids of different sizes (Figure 10). At a 10-m interval, none of the positive tests would link a common occupation. Contouring at this extremely coarse-grained level combines unrelated information into a single algorithmic distribution and creates misleading and incomplete density maps. The problem is compounded by artifact recovery success. Just because you are excavating in the area of a former camp does not mean you will recover anything from it. On the other hand, you might have been fortunate to come down on the center of the debris scatter near a hearth and thusly conclude that this might be a good place to excavate a test unit. Had the grid been shifted a meter in any direction, however, the return may have been singularly unimpressive and your attentions might have shifted elsewhere. Only two of 30 hearth areas in the hypothetical site would have been intercepted or nearly intercepted by 10 m interval shovel testing. Testing at 5-m intervals would Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 43 Chapter 4. Research Design Figure 7. Historic Artifact Distributions and Components at Salt Pond Plantation, Francis Marion NF, SC. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 44 Chapter 4. Research Design Figure 8. Precontact Occupation Clusters at Salt Pond Plantation, Francis Marion NF, SC. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 45 Chapter 4. Research Design Tanagaba: !Kung Residential Campsite 0 * 2 Hearth ** 3 4 5 Meters * Hut N 1 * ** * Hearth Hut * * * ** Branch Water Root Bone Wood Shavings Spiny Melon Tsin Bean Pods Figure 9. Two-household !Kung Residential Camp of Tanagaba (from Yellen 1977). increase the coherence of the distributional sample significantly, as an increased number of adjacent tests would potentially yield artifacts from the same occupation. However, if most of the occupations consisted of single-family units with an average diameter of 3 m, none of the tests would have produced related materials, even at this relatively high sampling intensity. Finally, at an intensity of 2.5 m, an accurate map of debris concentrations could be achieved, and there would be enough negative tests between occupations or reoccupation clusters to identify the rough boundaries of the phenomena of interest. Intervals below 2.5 m intervals, however, would be required to sufficiently sample individual occupations to generate structural data. Another methodological problem that is encountered at Fort Bragg is the recognition and identification of Archaic occupation clusters. Projectile points are rarely recovered from shovel tests and this is the primary diagnostic artifact used to identify Archaic phases. However, there is ample evidence that lithic material found at depths below 30 or 40 cm in sandy matrices along the Atlantic Slope is pre- dominantly affiliated with Archaic and Paleoindian phases. Michie (1990) observed, over a large sample of South Carolina Coastal Plain sites, that Archaic and Paleoindian components consistently derived from depths of 30 to 70 cm below ground surface. Moreover, the phases were positioned in a vertically coherent pattern relative to one another. Late Archaic material was consistently found between 28 and 35 cm, Middle Archaic assemblages occurred between 35 and 55 cm, Early Archaic components were positioned between 50 and 60 cm, and Paleoindian materials were located below 60 cm. Today we know that these depth ranges vary from one depositional environment to the next (see Brooks et al. 1998), but the relative vertical positions of the components are always expressed in a coherent sequence. Woodland and Mississippian materials, by contrast, are concentrated in the upper 30 cm of deposit. A graphic demonstration of the Woodland (ceramic)/Archaic (lithic) vertical dichotomy is provided by the density distributions of ceramics and lithics by level from shovel test sampling at Site 38SU136/137 on Poinsett Range in Sumter County, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 46 Chapter 4. Research Design 1.25 m Interval Hu t Hea rth ** Hea rth Hu t 550 * * * ** * * Hut * ** Hearth ** * * * * Hearth Hut * * Hut 2.5 m Interval * * t * * * t t rth Hea * * * t Hu ** 530 * * * h Hu Hearth * ** * art He * ** * * h art He Hut Hearth * Hearth * He Hu t Hut ** * ** Hu h art He Hut * ** * * * h art He ** art t Hu H * ** h art He t Hu ut * ** * * h * * * * h art He * * * * * ** ** * * * h art He t Hu Hu t * ** Hu * Hut 540 ** * ** Hearth * * ** Hearth Hut Hearth * * * * ** h art He * ** h art He * * t ** * rth Hea Hu * t Hu * Hut ** Heart * h Hut h * Heart ** * * * * * * * * * ** h art He * * ** t Hearth Hut Hut ** * Hearth * * ** * * * Hut Hu Hearth * ** * * ** h art He Hut * t Hu Hearth * 520 510 500 500 510 520 530 540 550 Figure 10. Simulated Reoccupations of Tanagaba-like Settlements (Different Colors Represent Separate Occupations) Over Shovel Test Grids of Varying Intervals. South Carolina (Cantley and Cable 2002). Here, shovel tests were excavated in two levels, 0-30 cm (Level 1) and below 30 cm (Level 2) on a grid of 10 m with a sub-sample of 5-m interval frames (Figure 11). Ceramics were clearly concentrated in Level 1 (Figure 12), while a majority of the lithic material was positioned in Level 2 (Figure 13). Although Woodland occupations also contained lithic artifacts, Archaic occupations, as confirmed by block excavations, contained higher lithic densities. Clearly, level excavation of shovel tests provides a basis to identify pre-ceramic occupation clusters in sites with sandy matrices. This was also demonstrated during the previous DO1 testing project conducted by Palmetto Research at Fort Bragg, where a tri-partite arbitrary level scheme was deployed (see Cable and Cantley 2005a): (1) 0-30 cm bs, (2) 30-55 cm bs, and (3) > 55 cm bs. This provided a basis for not only the differentiation Woodland and Archaic, but also for differentiating later and earlier Archaic deposits. An additional Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 47 Chapter 4. Research Design 620 600 580 560 540 520 38SU136/137 500 480 460 440 420 828 10-meter Interval Shovel Tests 400 Big Bay 380 Systematic, Judgemental Shovel Tests Topographic Contours: Interval = 0.1 meters d 360 rt Di 340 a Ro 320 0 300 40 meters North 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 11. Shovel Test Pattern, 38SU136/137, Poinsett Range, South Carolina. level was inserted into the scheme during the DO4 fieldwork to achieve an even finer resolution. The four levels were structured in this manner: (1) 0–20 cm bs, (2) 20–30 cm bs, (3) 30–50 cm bs, and (4) > 50 cm bs. An additional principal can provide guidance in distinguishing clusters made up of lithic debris. Our experience in the South Carolina Coastal Plain (see Cantley and Cable 2002:266-267) indicates that occupation floors tend to be composed of a limited range of lithic raw material types. Most commonly, a single raw material type comprises an entire concentration. This holds true for both debitage and tools. Low lithic raw material diversity would seem to be a consistent feature of short duration campsites, particularly when raw material sources are not available in the immediate vicinity. This would appear to be the case at Fort Bragg as well (see Benson 1999:2534; Daniel and Butler 1996). In these situations, raw material cores are imported to a campsite and reduced to manufacture and replenish tool kits. In situations where occupation duration does not exceed several days there is little opportunity to reduce a variety of cores from different sources. The consistent occurrence of spatially segregate debitage concentrations of single lithic raw material types provides an important organizing principle for identifying occupation clusters. Benson’s (2000:622) observations concerning precontact sites tested in the Overhills Tract indicate that raw material segregation commonly occurs in the Fort Bragg area, as would be predicted given short occupation spans and extra-local raw material sources. This principal should not be uncritically applied, however, as there are a number of situations that could bring different raw material types together in the output of a single occupation. For instance, long curated tools and cores made of different raw materials might find their way to a campsite where they were reduced or used and deposited along with the dominant by-products of cores from the last source visited. Another example would involve the rendezvoused aggregation of multiple households traveling from different locations and carrying lithic Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 48 Chapter 4. Research Design 580 560 540 520 500 480 460 Big Bay 440 420 Level 1 400 380 Key: Sherd Density Contours (per stp) ≥ 0.5 < 1 sherd ≥ 5< 6 sherds ≥ 1 < 2 sherds ≥ 6 < 7 sherds ≥ 2 < 3 sherds ≥ 7 < 8 sherds ≥ 3 < 4 sherds ≥ 8 < 9 sherds ≥ 4 < 5 sherds ≥ 10 sherds D irt R oa d 360 340 320 300 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 0 40 meters North Topographic Contour Intervals = 0.10 m 500 520 540 560 580 600 620 640 660 580 560 540 520 500 480 Big Bay 460 440 420 Level 2 400 380 d Key: Sherd Density Contours (per stp) ≥ 0.5 < 1 sherd ≥ 5< 6 sherds ≥ 1 < 2 sherds ≥ 6 < 7 sherds ≥ 2 < 3 sherds ≥ 7 < 8 sherds ≥ 3 < 4 sherds ≥ 8 < 9 sherds ≥ 4 < 5 sherds ≥ 10 sherds D irt R oa 360 340 320 300 200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 0 40 meters North Topographic Contour Intervals = 0.10 m 500 520 540 560 580 600 620 640 660 Figure 12. Vertical Density Distributions of Precontact Sherds, 38SU136/137. raw material from different sources. Individual concentrations in this instance, however, might retain raw material homogeneity. Acknowledging these limitations, however, lithic raw material segregation can be effective in tentatively distinguishing overlapping lithic concentrations in both Archaic and Woodland contexts. Moreover, in Woodland contexts lithic and ceramic concentrations are spatially discrete and represent functionally distinct occupation clusters (see Cable et al. 1998). In combination with level excavation of shovel tests, the raw material segregation principal provides a basis for generating identifications of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 49 Chapter 4. Research Design 580 560 540 520 500 480 Big Bay 460 440 420 Level 1 400 380 d Key: Lithic Density Contours (per stp) ≥ 0.5 < 1 lithic ≥ 5< 6 lithics ≥ 1 < 2 lithics ≥ 6 < 7 lithics ≥ 2 < 3 lithics ≥ 7 < 8 lithics ≥ 3 < 4 lithics ≥ 8 < 9 lithics ≥ 4 < 5 lithics ≥ 10 lithics D irt R oa 360 340 320 300 200 220 240 260 280 300 320 340 360 380 400 420 440 0 40 meters North Topographic Contour Intervals = 0.10 m 460 480 500 520 540 560 580 600 620 640 660 580 560 540 520 500 480 Big Bay 460 440 420 Level 2 400 380 d Key: Lithic Density Contours (per stp) ≥ 0.5 < 1 lithic ≥ 5< 6 lithics ≥ 1 < 2 lithics ≥ 6 < 7 lithics ≥ 2 < 3 lithics ≥ 7 < 8 lithics ≥ 3 < 4 lithics ≥ 8 < 9 lithics ≥ 4 < 5 lithics ≥ 10 lithics D irt R oa 360 340 320 300 200 220 240 260 280 300 320 340 360 380 400 420 440 0 40 meters North Topographic Contour Intervals = 0.10 m 460 480 500 520 540 560 580 600 620 640 660 Figure 13. Vertical Density Distributions of Precontact Lithics, 38SU136/137. lithic clusters with culture-chronological implications. For instance, a cluster composed of aphyric rhyolite recovered from Level 2 would suggest a Late or Middle Archaic affiliation. Over a large sample of clusters, then, it might be possible to begin to construct chronological raw material utilization patterns based on debitage outputs to complement and contrast patterns derived from diagnostic projectile points. Quadrat Sampling (Sample Blocks) The traditional approach to Phase II testing is to establish a systematic shovel test sample at an intermediate sized of interval of 10 or 15 m to define site boundaries and then to select locations to excavate 1 x 1 m test nits to amplify the artifact density patterns discovered from shovel testing. As we con- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 50 Chapter 4. Research Design tinued to investigate the distributions of occupation clusters during the DO1 Phase II testing project, we became increasingly aware of the inadequacies of this program. Since most clusters were less than 5 m in diameter, sampling at intermediate shovel test intervals was not effective in deriving probability samples of these phenomena. Consequently, it would not be effective in developing occupation models simply because the phenomena could not be sufficiently defined. A method we feel is better suited to Phase II investigations on low-density, palimpsest precontact sites is quadrat sampling. Smaller areas selected because of specific shovel test discovery phenomena (i.e. a lithic tool, a debitage concentration, a partial vessel concentration, a cluster of fire-cracked rock, etc.) can be examined at a much more intensive sampling intensity to yield information nearly tantamount to small block excavations. An array of 1.25 m interval shovel tests can potentially recover a set of 6 to 9 contiguous shovel tests from single occupations on the size order of 2.5 to 5.0 m in diameter. These samples have tremendous interpretive and explanatory power because they actually constitute a formidable probability sample of individual occupations at a sampling fraction of 10.33 percent. At this sampling intensity, it is possible to generate statistical data on the size and content of individual occupations and occasionally recover information on their culture-historic associations. Because we know that tools and large pieces of lithic raw material tend to be clustered at a point on the periphery of debitage concentrations (assumed hearthing areas), it is also possible in many cases to use the 1.25 m interval data to “mine” or “krig” for additional tools and diagnostics. Two methods are available for this third stage of investigation, microinterval shovel testing and test unit excavations. The next increment of shovel test interval is 0.625 m and this yields an area sample fraction of 33.7 percent, an exceedingly powerful probability sample if applied to the entire target occupation and an effective tool, as well, when “mining” for tools and diagnostics. The quadrat sample approach is particularly effective at the Phase II level of investigation because it allows for a detailed evaluation of the horizontal integrity of deposits without the need to impact a large area of the site. It also provides a basis for building a regional settlement data base with valid and accurate depictions of individual occupation types, many of which can be associated with a particular culture historic phase or period. This second benefit often times goes unappreciated at the Phase II level of investigation. One reason for this is that Phase II projects have not been traditionally geared to recovering specific and detailed data of this type. Another is that there does not seem to be an appreciation for how research results play a role in evaluating sites. The standard approach is to demonstrate that the deposits of a site are well preserved. Once integrity is demonstrated a particular characteristic of the archaeological record of the site might be drafted into service to justify an eligible recommendation. This might be either the identification of a particular component of interest or simply the presence of high artifact density. This is a vapid strategy and cannot lead to a sophisticated understanding of the archaeological record of a region. It is also a very poor fit at Fort Bragg. From our two seasons of work here we now know that almost every site retains a sufficient degree of integrity to inform on a host of important research issues and we will show this with the results reported herein. It is also clear that the identification of components at a particular site is a rather serendipitous process at the level of Phase II coverage. You can demonstrate the presence of a particular component, but you are not on equally firm ground in suggesting that a component is not present based on the light coverage these sites receive. Moreover, artifact density is a function of settlement type variability and probably raw material source proximity. Focusing solely on high-density deposits will bias the site sample and actually impede synthesis. Regional reconstruction requires that you have a sufficient sample of the full range of behaviors that transpired to produce a settlement strategy, not just the most archaeologically visible ones. If the sample is not effectively constructed at the Phase II level, it is difficult to see how it would ever be. This problem must be confronted directly through the development of strategies that will allow the exercise of proper judgment in selecting a coherent and efficient regional sample of sites. Two strategies are available, redundancy and sufficient recovery. Both strategies require a fair amount of documentation and description on a site-by-site basis. The former strategy relies on a synthetic picture of “site type” representation within the various microenvi- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 51 Chapter 4. Research Design ronments and geographic localities of the Fort. Sites would be evaluated based on their goodness of fit for complementing a regional data set. At the outset, this strategy will be difficult to implement because we do not yet have an environmental model to draw on, nor is there a clear idea of the environmental representation of the current sample of eligible sites. The second strategy requires the gathering of sufficient information at the Phase II level to argue that the important values of certain sites have been recovered. This strategy is most effectively deployed on smaller sites where a relatively large proportion of the available occupations can be identified and described. The mode of description we have proposed is close interval, systematic shovel test grids spaced at 1.25 m and 0.625 m intervals. The former interval yields ideally a 10.33 percent area sample of an occupation or occupation element, while the latter produces 33.7 percent samples. The actual sample fraction, of course, will vary based on the size of the cluster and the fit between it and the shovel test pattern. Both sample fractions generate rigorous data with which to develop regional models of settlement. Occupation Cluster Characteristics Various characteristics of occupation clusters can inform on the organization of regional systems. Our ability to accurately measure these characteristics, however, depends on the number of sample points (i.e. shovel tests) that define a cluster. Obviously, an occupation cluster represented by a single shovel test is subject to a high degree of stochastic variation. Larger clusters represented by a number of sample points, or smaller clusters more intensively sampled, on the other hand, will yield more reliable measurements, simply because they contain larger sample sizes. In this regard, survey data will less reliably characterize the quantitative properties of occupation clusters than data derived from testing projects as a rule. The two characteristics that have exhibited the most utility for settlement pattern studies are Cluster Size and Cluster Artifact Density (Cable et al. 1994; Cliff and Cable 1999). Another characteristic that may have selective value in comparing the internal structure of larger cluster types is Cluster Spatial Point Pattern. Although procedures for measuring these variables were provided in the initial research design, on-the-ground applications of more intensive shovel test sampling strategies required some modifications that will be discussed below. In addition, close-interval shovel testing allowed the identification of core reduction strategies from debitage profiles and this variable has taken on increased importance in differentiating occupation cluster types. Cluster Size Our original plan was to use the Delaunay triangulation algorithm to define cluster size. This method is not as sophisticated as grid model techniques, but it more faithfully honors the actual sample point values and also approximates more closely hand drawn contours so that computer mapping is not essential to achieve data comparability. Moreover, the method could be used to approximate cluster sizes from small numbers of positive shovel tests. During fieldwork, however, it became clear that we would have a difficult time identifying clusters at relatively large shovel test intervals due to the paucity of diagnostics. When shovel test intervals were lowered to sample restricted areas we achieved a degree of success in defining individual clusters by raw material distributions, the identification of tool clusters, and occasionally the recovery diagnostic artifacts. This dictated that we focus on a smaller number of clusters than originally estimated and sample each more intensively. Consequently, we shifted to a grid-based contouring algorithm to estimate cluster size, which is more accurate when larger sample sizes are available. Clusters were defined at shovel test intervals of 2.5 m and 1.25 m, resulting in samples of as many as 6 to 16 shovel tests within a single cluster. Artifact Density Artifact density for each occupation cluster was calculated as a weighted mean of artifact frequencies from shovel tests. Density was standardized by unit area. A weighted mean is preferable to the common arithmetic mean because it helps control for sampling error related to non-homogeneous data distributions. The weighted mean measurement applied here is described by Long and Rippetaeu (1974:208) and is calculated by constructing a fre- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 52 Chapter 4. Research Design quency histogram of diagnostic artifact frequencies by shovel test. Factoring is applied to progressively give greater weight to the more frequent shovel test results. For example, an occupation cluster comprised of eight shovel test results, two of which produced no artifacts, three of which yielded one artifact each, two of which contained two artifacts each, and one of which produced five artifacts associated with the specific cluster would yield a weighted mean of 1.0 artifacts per shovel test. Each result is assigned a weighting factor from 1 to 4 based on its frequency of occurrence ranking. In the example above a weight factor of 2.5 was assigned to results 0 and 2 because they have equal frequencies (n=2) of occurrence. The average of ranks 2 and 3, the positions they occupy, is 2.5. The weighted mean would then be calculated using this general formula: (r1 x w1) + (r2 x w2) + ... (rn x wn) –––––––––––––––––––––––––––– w1 + w2 + ... wn Where r equals the individual shovel test result and w equals the weighting factor associated with each shovel test result. Each separate result is multiplied by its assigned weight factor and the sum of these weighted results is divided by the sum of the weights for each shovel test as opposed to the actual number of shovel tests (n), which would represent the denominator in the standard arithmetic mean. The non-weighted arithmetic mean for these data would equal 1.5, but the weighted mean is lower (1.0) and compensates for the potential bias posed by the one unusually high yield shovel test. It is also desirable to standardize mean density so that area densities can be easily calculated and compared. A standardized area of 1 m2 is an easily compared value. The mean calculated above indicates a density of 1.00 artifacts per 30-cm shovel test, an area of 0.09 m2. This yields a mean density of 11.11 artifacts per m2 for the occupation cluster. It is important to precisely excavate shovel tests to ensure comparable data on artifact density. During the field phase a consistent unit size was maintained by the use of a wooden template. pation clusters at sample fractions (percent area of an occupation cluster) of 4 to 36 percent. With this information as well as cluster size, we were able to extend the density measures for each occupation to estimates of total quantities of artifacts. Most of the clusters sampled corresponded to debitage concentrations ranging from primary core reduction loci to secondary core reduction and tool maintenance stations. Spatial Point Pattern Spatial point pattern is a measure that informs on the distributional characteristics of an occupation cluster and is adapted from the quadrat count models of ecology (see Pielou 1969) and geography (Haggett et al. 1977:414-417). These models evaluate the spacing relationships of scatters of points by overlaying quadrat grids. Each grid quadrat serves as a sampling unit and the number of points contained within each quadrat is tabulated. The frequency distribution of point counts per quadrat is then compared to a theoretical distribution generated from any number of statistical processes. The simplest of these is the Poisson process. Importantly, the Poisson process provides measures of not only the amount, but also the kind of order present in an observed spatial point pattern. Since the mean and variance in a Poisson distribution are equal, the variance/mean ratio, Di, in a purely random point pattern is 1.0. Moreover, the ratio is expected to exceed 1.0 as clustering increases, while systematic patterns are indicated by ratios of less than 1.0. A formal test, the d statistic, for significant departure from the random pattern of an observed ratio has been developed by Bartko et al. (1968), which approximates a Chi-square test. This measure is most appropriate for evaluating point patterns in large, heterogeneous clusters such as Mississippian residential sectors, multifamily hamlets, or slave rows. The precontact occupations so far isolated at Fort Bragg, however, are simple, homogeneous phenomena that are not well suited to this type of measure. At present, then, spatial point pattern statistics have not been applied to the recovered occupation clusters. The extremely close-interval shovel test patterns applied to restricted areas of the sites provided an opportunity to statistically sample individual occuPhase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 53 Chapter 4. Research Design Debitage Profiles Debitage profiles of individual concentrations supply important information concerning the functions of occupation clusters. Those clusters defined by close-interval shovel testing generally produced sufficient data to characterize the kind of core reduction strategies that transpired during an occupation. Reduction strategies were identified through comparisons of debitage characteristics and debitage density calculations. Five strategies were recognized: (1) Low-density debitage scatters containing only or mostly small thinning flakes (BTFs) with little or no cortex on dorsal surfaces indicated only tool maintenance activities. (2) Low-density debitage scatters containing greater than 15% large flake sizes and high proportions of flakes of bifacial retouch (FBRs) and other early stage debitage represent limited tool manufacture from prepared biface cores. These assemblages generally contained little or no debitage with dorsal cortex. Reduction is inferred to have involved only one core. (3) Low-density debitage scatters with 30% to 60% large flake sizes that were dominated by early stage debitage classes (i.e. chunks, core flakes, core rejuvination flakes) with relatively high incidences (> 3–4%) of cortex represent the limited reduction of an unmodified core or natural cobble. Reduction is inferred to have involved only one core. (4) High-density debitage concentrations containing greater than 15% large flake sizes and high proportions of FBRs represent tool manufacture from prepared biface cores. These assemblages generally contained little or no debitage with dorsal cortex. Reduction was inferred to have involved only more than one core. (5) High-density debitage scatters containing 30% to 60% large flake sizes that were dominated by, or had relatively high proportions of, early stage debitage classes (i.e. chunks, core flakes, core rejuvination flakes) with relatively high incidences (4–50%) of cortex represent the reduction of multiple unmodified cores or natural cobbles. Debitage classes and flake size categories are defined in Chapter 6. Evidence of tool manufacture was considered an indication of residential occupation and some kinds of specialized logistical camps, while its absence was inferred to be an indication of very short-term logical camps. Cluster Types Short-interval shovel testing in a judgmental manner in DO1 and in a systematic strategy in DO2 has provided a basis for defining a number of occupation cluster types. This terminology is used in this report to interpret the functions of individual occupation clusters. Four types are now recognized and some inter-type variability may be present: (1) Type I occupations were composed of debitage concentrations of single raw material types which were produced from either unmodified core or biface core reduction. Tool clusters commonly consisted of tools and manufacturing rejects made of the same raw material comprising the associated debitage concentrations, indicating tool manufacture for use during the stay as well, probably, for tooling up upon abandonment. Some examples of discordant raw material suites between tool clusters and debitage concentrations were also noted. The discarded tools in these situations are regarded as tool-kit residuum brought to the site and deposited after use and retooling. This site type is inferred to represent a forager household or small multi-household residence occurring either in isolation or within a larger aggregation of households. (2) Type II occupations exhibit a similar spatial organization to that of Type I occupations, but exhibit much higher tool-to-debitage ratios and may commonly be comprised of diverse raw materials. Lithic reduction on these sites consisted of tool maintenance and late stage shaping of tools from flake blanks or reduced biface cores. These occupations are inferred to represent single or small, multiple households of high technology foragers and may occur in isolation or as members of a larger aggregation of households as well. (3) Type III occupations consist of clusters of extremely low-density debitage scatters and small Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 54 Chapter 4. Research Design numbers of tools discarded in a disorderly fashion. Debitage profiles suggest that only tool maintenance activities produced the debitage scatters as opposed to tool manufacturing. This type is believed to represent logistical camps or extraction loci produced by special task groups. (4) Type IV occupations exhibit debitage profiles indicative of core reduction to produce tools. Tools are scarce and combined with the fact that the debitage scatters are low-density in nature suggests short-term occupation. Within this type, however, may be a number of functional types that are difficult to distinguish because of the low frequency of artifacts. Included functions might be short-term residences and logistical camps. Also subsumed under this category are sherd clusters associated with the Woodland period. These clusters tend to represent isolated deposits with very little associated lithic material. Debitage scatters and stone tools have not been found in association with these features, although several may have associated fire-cracked rock concentrations. It is speculated that these sherd clusters represent plant food processing stations occupied by especially composed female task groups. Occupation may not have involved overnight stays, but the camps may have been reused serially over the course of a season and over a period of years by the same or related social units. These types can occur as isolated occupations or they may represent elements of multi-household camps. Unfortunately, our testing and data recovery strategies are geared only to define and expose these elements. Larger exposures and more efficient data recovery techniques are badly needed to allow us to see larger community patterns. Population Estimates During the earlier DO2 investigation (Cable and Cantley 2005b), a method was developed to generate population estimates of occupations or occupation elements from close-interval shovel test grids of 2.5 m or less in recovery units referred to as “sample blocks.” These blocks ranged from 5 to 15 m on a side and in the larger exposures succeeded in identifying potential community patterns. Attempts were made to distribute these sample blocks across the sites to approximate proportional representations of occupation types. Sample areas ranged between 4.3% and 14.9% of the smaller sites and 0.7% to 2.1% of the larger ones. After corrections for occupation intensity, it was estimated that 5,174 occupation elements were contained within the 8-site sample. That may seem like a large number of occupations, but if this figure were spread out over a 12,000-year occupation span, it would indicate an average occupation intensity of 0.054 occupation elements per year per site. Proportional data on occupation types and cultural phase associations were also generated from these data, providing an extremely detailed look at the character of occupation over time at each site. Different methods of calculation and investigation were required for the DO3, DO4 and DO5 investigations, as these sites were sampled at much lower intensities. Shovel test intervals ptimarily consisted of only 2.5-, 5- and 10 m-intervals. Closeinterval shovel tests of 1m- and 1.25 m-intervals were used only sparingly to precisely locate test units. Procedures to estimate occupation elements from the 10 m-interval Stage I sample and from the 2.5 m- and 5 m-interval sample block grids were developed to compensate from this shift in strategy. The weakness of the Stage I sample is that exposures of individual occupations, which range from about 1 to 5 m in diameter, are incomplete and consequently prone to identification errors. The strength is that it achieves an unbiased, even coverage of the site. The advantage of sample block exposures is that individual occupations are better defined spatially and vertically in “micro-view” and a more accurate assessment of occupation history and occupation type proportions can be generated. The weakness of this form of sample is that coverage is biased because it focuses on sections of sites that may be uncharacteristically dense. The interplay of the two data sets, however, moves us toward an accurate picture of occupation and functional land use patterns. The Stage I estimate is generated by counting elements exposed in each shovel test. At this sampling interval we are assured that the presence of an element will not be repeated in adjacent shovel tests based on the size of elements, which we know ranges between about 1 and 5 or 6 m in diameter from our intensive investigation in the DO2 investigation. Elements are identified by lithic raw material groups and Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 55 Chapter 4. Research Design by precontact sherd sub-series/vessels. This runs the risk of overestimating occupations because more than one raw material group may comprise a single occupation. The term “element” recognizes this fact, focusing on depositional events and episodes rather than entire occupations. Elements can represent segments of contemporaneous multi-family occupations, separate occupations of various function, or contemporary members of multiple raw material reduction episodes. Elements have horizontal dimensions and, as a result, sample units that are smaller than those dimensions benefit as sample space is expanded beyond the actual area sampled (see Cable and Donaldson 1988; Rice 1987; Rice and Plog 1983). In the case of the typical element found in the Fort Bragg area, which has a mean diameter of about 3 m, a shovel test covering .09 m2 samples only .09% of a 10 m square, but it samples 9.0% of the mean feature or element space within this area. A 10 m-interval shovel test pattern actually samples about 9% of the available feature or element space. Theoretically, the features or elements contianed within this space will be intersected by a shovel test excavated in its center. Those elements that are not intersected, but occupy a portion of the target feature space, would be members of an adjacent feature space of the same size. The sample frame, then, consists of contiguous cells of feature spaces each measuring 3 m in diameter. The population estimate, then, requires only a simple arithmetic calculation of counting the occupation elements identified in the shovel tests and dividing this figure by 9%, the tehretical size of the sampled area, or the effective sample area. The results appear to be of the same magnitude as the estimates based on the sample block approach developed for the DO2 investigation, which suggests that this method can be effectively applied not only at the Phase II level, but also at the Phase I survey level to generate a regional database. The Stage I population estimates for the six sites in the DO3 project totaled 3,477 elements. Shorter interval shovel tests will yield higher feature space sampling fractions. For instance, 5-m intervals will produce a 36% effective sample area, while a 2.5 m-interval will theoretically yield a 100% sample, given an average element diameter of 3 m. The blocks were primarily sampled at a 2.5 m intensity, but in some cases either 5 m-intervals or combinations of 2.5 and 5 m-intervals were deployed. After adjustments to integrate the various sample coverages and to address boundary coverage issues as discussed individually for each site, the combined population estimate using the sample block method was 4,070 elements in the DO3 package. The Stage I estimate is probably more accurate given its unbiased coverage, but in four of the cases, the projections were extremely close. In the other cases sample block coverage was probably biased for higher occupation intensity locations. In addition, an extra large block was excavated at a location that yielded an unusually large number of ceramic elements many of which may be associated with debitage scatters, resulting in double counting of occupations. One problem with the method is that it assumes that all elements within the feature space will be recognized in the associated shovel test. Since elements are identified from the recovery of artifacts, it is likely that in some instances an element will go unrecognized, given the small area covered by a shovel test. This will result in an under-estimation of the element populations of sites. The next phase in improving the accuracy and precision of the method would be to empirically test the macro-interval sample results against large nested blocks of microinterval shovel tests, most preferably at an interval of 1.25 m. This would provide an understanding of the relationship between the method and real element distributions on sites, which would, in turn, suggest ways to correct and modify the approach to achieve greater accuracy. The method, at present, represents a heuristic device to compare and contrast occupation data between sites. It will move from a relative measure to a specific measure when it receives greater scrutiny from empirical trial runs. REGIONAL SETTLEMENT MODELS Once the database is constructed, it will be possible to develop classification systems with which to create environmental zone and occupation cluster typologies. The environmental zone typology will attempt to devise a breakdown that will not only distinguish microenvironments, but that will also apply the crosscutting dimension of geographic location. In Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 56 Chapter 4. Research Design this way, equivalent microenvironments can be differentiated to achieve even coverage of the Fort. This provides a basis for monitoring historical processes of geography that may have little relation to the precise micro-environmental context of a location. Migration and population center proxemics are examples of processes that can introduce settlement variability into a landscape that would be independent of microenvironment (see Haggett et al. 1977). Once occupation cluster and environmental zone typologies are developed from the database, regional settlement models for each phase can be generated. This will proceed in two ways. First, associational relationships between environmental zone and occupation cluster types are to be examined. This will be accomplished through the application of Chi-square and non-parametric statistical measures of association. The other approach is to examine occupation cluster distributions on the landscape. A study of component distributions is useful as a basis for identifying settlement shifts through time. The approach can also be used to identify spacing relationships within single component distributions that may reveal the structure of land-use systems. An update on the development of the database is presented in the synthetic chapter (Chapter 16) in the DO2 report (Cable and Cantley 2005b). One hundred fourteen occupations/occupation elements have been identifiedso far and these have contributed to the formulation of the ccupation type construction discussed above. The metric and qualitative data for each occupation are reported in Appendix M of the DO2 report. Because micro-interval shovel tests were disallowed in the DO4 investigation, no new occupations were added to the database, since the required sample density could not be achieved. REGIONAL INTEGRATION Interpretation of the regional models will be accomplished within the framework of established cultural and historic contexts. Because of the synthetic nature of the approach proposed here, it is likely that new insights can result in yearly modifications and redirections of such contexts. Appendix XI of the Fort Bragg ICRMP presents a detailed discussion of the current cultural and historic contexts for the Fort. A distilled overview of this document follows with suggestions of how a regional approach can improve our ability to address research questions and to evaluate the potential of specific properties to contribute to answering these questions. Two major themes are identified in the discussion of precontact cultures represented on the Fort. One theme concerns the building of a culture chronology, while the other focuses on settlement and mobility systems. Both themes are presented with the problem of scale. The arbitrary boundaries of Fort Bragg do not correspond, most likely, to the territorial ranges of the past cultural groups that occupied the region. The Fort is situated on a watershed divide centered on a large ridge spine running in an east-west direction in the Sandhills physiographic province (ICRMP 2001, Appendix XI:2). Most drainages are small, ranging between ranks of 1 and 2. A portion of the Little River floodplain is situated on the northeastern boundary of the Fort, but most of the area is well-drained Sandhill ridge fingers jutting out from the divide. Thus, most cultural systems using the area had larger territorial ranges than the geographic area of the Fort and incorporated a great deal more micro-environmental variability in their adaptations. Any approach devised to explain cultural systems must control for this limitation and develop methods to compensate for it. Fort Bragg CRM personnel have responded by extending their range of study to include extra-local studies of lithic and clay sources, sand burial mound investigations (Irwin et al. 1999), and ceramic sequencing (Herbert 1999). Chronology Building Intensive surveys conducted at Fort Bragg have succeeded in recovering a wide range of diagnostic artifacts spanning the Paleoindian, Archaic, and Woodland periods. Building a local chronological sequence out of this corpus of information is underway, and attention to specific data needs during the site relocation and testing project will greatly advance this endeavor. Three areas of investigation are of particular relevance in this context. These are: 1) correlation of unidentified projectile point styles with sequences from adjoining regions, 2) detailed analysis of ceramics, and 3) identification of rela- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 57 Chapter 4. Research Design tive and absolute dating opportunities. The ceramic TL-datinfg project conducted by Herbert (2003) is a prime example of how a regional chronology can eventually be constructed. Many of the projectile points identified on the Fort correspond closely to the material Coe (1964) described from the Hardaway and Doerschuk sites, but other points have yet to be correlated with projectile point sequences from neighboring regions. Daniel’s (1998) reanalysis of the Hardaway Site indicates that a larger range of morphological variation was present there than was originally described. Moreover, researchers across the Atlantic Slope are beginning to appreciate that many unclassified types in their regions correspond closely to defined styles in other regions. The Early Archaic Big Sandy point, a large side-notched type originally identified in Alabama and Tennessee (Cambron and Hulse 1960; Kneberg 1956), is apparently quite common at Fort Bragg (see Benson 1998). It is also prevalent in the north Coastal Plain of South Carolina (Cable et al. 1998; Cable and Cantley 1998). Other types that can be correlated with point styles from more distant regions are now finding their way into regional sequences along the Atlantic Slope. Cantley (1998; 2002) has convincingly demonstrated the existence of a wide variety of Woodland triangular and stemmed points at Poinsett Range in Sumter County, South Carolina that closely correlate with the Tennessee sequence. In addition, Sassaman et al. (1990) have recognized a number of stemmed points (MALA) that associate with the terminal Middle Archaic period along the Savannah River that bear a remarkable resemblance to the late Middle Archaic Benton, Sykes, and White Springs styles in Tennessee and Alabama (see DeJarnette et al. 1962:70; Lewis and Lewis 1961:40-41). Goodyear (pers com. 2002) has found a large concentration of MALA points on the Lower Savannah River as well, and Cantley (2002) reports Sykes points in stratigraphic contexts consistent with terminal Middle Archaic at Poinsett Range in central South Carolina. These examples illustrate that greater stylistic similarity in projectile point sequences prevails across the Southeast than is commonly recognized and that attention to reported type descriptions can be useful in classifying unidentified point styles in local sequences. A quick scan of illustrations of the projectile points collected during sur- veys at Fort Bragg indicates that there are a number of unidentified points that would probably correlate with existing type descriptions from adjoining regions. Attention to this detail would significantly increase the time sensitivity of the local sequence. Herbert (2003) has built a ceramic sequence for the Cape Fear drainage system by combining a detailed, multi-attribute analysis centered around paste characteristics and surface treatment and an innovative thermoluminescent dating program that generates absolute dates for individual sherds. Preliminary dating and typological evidence suggest that various paste modes or series are chronologically ordered. The limestone-tempered Hamp’s Landing series may date as early as 1500 or 2000 B.C., while the coasre sand-tempered New River series is somewhat later and has a chronological position clearly within the Early Woodland time range. The sherdtempered Hanover and sand-tempered Cape Fear series date to later temporal positions and are regarded as Middle and Late Woodland affiliates. Preliminary information on a crushed rock series suggests that it may correlate with the Piedmont-based Yadkin series and may follow New River in the sequence (Herbert 2002, personal communication). The next step is to build from these basic observations a culture-chronological phase sequence for the Cape Fear drainage system. This will require data from controlled excavations in which ceramic assemblages can be defined by recovery from sealed features or from single component deposits. In addition, phase breakdown may be facilitated by microseriational studies of series groupings in which the relationship between sub-series paste variation and surface treatment are monitored (see Cable 1998). During the site relocation and testing project numerous opportunities to investigate and recover samples from discrete deposits will be presented and these situations can yield data sets that would provide an opportunity to conduct microseriational analysis. These same contexts may offer an opportunity to recover additional TL dates or radiocarbon assays that would bolster the time sensitivity of the microseriation. Palmetto Research has devised a multi-attribute ceramic analysis of appropriate detail for conducting this sort of investigation. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 58 Chapter 4. Research Design As is stated in the statement of work (N5890020160), there may be situations in which the opportunity for absolute dating will become manifest. In these cases, consultation with the NPS COTR and Fort Bragg CRM personnel will determine the feasibility of submitting samples for special analysis based on context quality and other concerns. At present radiocarbon and ceramic thermoluminescence dating are the primary techniques used on the Fort. Clearly, if we are to operationalize a component/occupation cluster approach to regional analysis, improving the specificity of the phase sequence will provide great benefit. Mobility and Subsistence Settlement Systems Modeling settlement systems within the confines of Fort Bragg is difficult. Two factors contribute heavily to this. First, is the aforementioned restricted scale of the Fort, which probably does not encompass the full territorial range of any of the cultural systems that occupied the area. Therefore, we have, as a unit of study, a small window exposing only a portion of the systems we want to study. Second, the survey data, which must serve as the foundation for any regional study, is generated at the site level but needs to be transformed into component level units to facilitate clear models of settlement type distribution and interaction. The first limitation is unavoidable and must be compensated for by extra-local studies that serve to define the larger system in ways that make the materials on the Fort more properly understood within a regional framework (e. g. Irwin et al. 1999). Reworking the survey data can mitigate the latter. Models to explain Paleoindian and Archaic period settlement/mobility systems in the Southeast are broad-based generalizations that receive much of their structure from empirical studies of one or a small number of sites (Anderson and Sassaman 1996). There is general agreement that territorial ranges were large, but contracted with time, and that the character of lithic assemblages reflects a shift from curated to expedient organization. Very little is actually known about Paleoindian site types and distributions anywhere in the Southeast. It is generally accepted that site types consisted of small residences and logistical camps, territories were large, and mobility was extremely high (see Cable 1982a; Anderson and Hanson 1988). The combination of high residential mobility and curated organization has been described by Kelly and Todd (1988) as a “High Technology Forager” adaptation. Early Archaic systems, by contrast, appear to have had larger residences that were probably occupied for longer durations (see Anderson and Hanson 1988; Cable 1996; Daniel 1996, 1998; Kimball 1996), as well as small residences. These larger residences may have served as seasonal aggregation sites that were, perhaps, the scenes of bulk deer harvesting activities. During the Early Holocene, then, there appears to have been a shift from systems of extremely large geographic range and high residential mobility to systems with mixed mobility strategies. These mixed systems appear to have been characterized by an alternating yearly cycle of high residential mobility and high logistical mobility strategies. It is assumed that the Early Archaic period was characterized by smaller territorial ranges and greater population densities than were extant in the Paleoindian period. Precisely how large these territories might have been and in what manner they were distributed across geographic space has become a matter of great importance in understanding regional dynamics. One of the first attempts to address such phenomena was the “band-macro-band” model proposed by Anderson and Hanson (1988). They hypothesized that Early Archaic groups were organized into a two-tiered social system. The minimal unit was a band numbering 50 to 150 people, while the maximum unit of the system was a macro-band of some 500 to 1500 people, a minimal mating network according to Wobst (1974). Eight macro-bands were speculated to have occupied the drainages of the Atlantic Slope from the Ocmulgee River in Georgia to the Neuse River in North Carolina. Each macro-band was said to have maintained individual river basins as a territory and to have seasonally exploited the physiographic variability that cross cut each basin. A primary argument mustered in support of socially closed drainages was that extra-local raw material use was greatest along rather than between drainages. The lower frequency of extra-local material between drainages was explained by indirect acquisition (i.e. exchange). Daniel (1996) has raised questions concerning this model based primarily on the problem of equifinality in distinguishing direct from indirect Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 59 Chapter 4. Research Design acquisition, since similar proportional patterns can be generated from other processes as well as exchange. Daniel argues that the patterns identified by Anderson and Hanson (1988) are just as easily explained as a function of the availability and proximity of lithic sources and that a model of cross-drainage territorial ranges is just as easily supported by the distributions they site. In response to this critique, Sassaman (1996) has examined drop-off curves for lithic raw material from transects in South Carolina and concludes that the primary route of isotropic raw material (i.e. Coastal Plain chert and rhyolite) was along rather than across drainages. Moreover, he concluded that cross drainage movement was characterized by plateaus in the drop-off curves, suggesting that cross drainage movement may have been accomplished by a different mechanism such as exchange. A problem with this interpretation is that the precise locations of the chert and rhyolite sources are not known. There are recorded chert outcrops in the closer Edisto basin which might have been used by these groups and the large quantity of rhyolite at Poinsett Range suggests that closer sources of this raw material may also be extant. If so, then the lateral movement of the groups would be much smaller and the territorial range could be seen as limited to the Wateree-Santee drainage. Lithic Sourcing, unfortunately, was not a part of the original study. Clearly, lithic source studies and unambiguous identification of raw material types is an extremely important facet of any regional study of territorial range. Palmetto Research will work closely with Fort Bragg CRM personnel to insure accurate identification of lithic material recovered during the project. Raw material availability is important for evaluating the far-reaching scope of the scale of adaptation in hunter-gatherer groups. However, until we develop better controls on lithic sources we may never bridge the equifinality problem. The recent work at Poinsett Range in South Carolina exemplifies this problem (Cantley and Cable 2002). Here, a large number of small residential sites of the Early Archaic period have been defined through shovel testing and block excavations. The occupation clusters consist of what appear to be both multi-household and single household camps occupied for short durations. Middle Achaic systems along the Southeast Atlantic Slope are commonly characterized as foraging economies with high residential mobility (Blanton and Sassaman 1989; Cable 1982a; Sassaman and Anderson 1996). Population density is assumed to have increased from the levels of the Early Archaic period, and, by implication, territorial ranges were shrinking. Most comparisons of population density have been made strictly on the basis of site frequencies, which is not an ideal measure because it does not control for site function. Breaking sites down into their component units as suggested in this proposal may provide a basis for more effectively evaluating regional population levels over time. In the Midsouth and Midwest there is ample evidence for the beginnings of a settlement hierarchy and at least semisedentary villages along major streams during this period (Brown 1985), but these features have not yet been documented on the Atlantic Slope. One possible exception is the terminal Middle Archaic component at the Big Pine Tree site on the lower Savannah River, which contains a dense assemblage of tools and debitage (Goodyear 2002, pers. com.). Interestingly, the camps show an equal representation of Coastal Plain chert and rhyolite, which have mutually exclusive distributions by individual campsites. Poinsett Range is approximately equidistant between the Allendale chert dominated Lower Savannah River and the rhyolite dominated Pee Dee drainage. Based on these facts, one could conclude that the groups occupying the large Carolina Bay at Poinsett Range on a seasonal basis were actually moving between the Pee Dee and Savannah River during a seasonal round; stopping at the bay twice a year. Coastal Plain chert dominated residences, then, would be interpreted as camps established while moving away from the Savannah River and rhyolite dominated residences would represent movement originating from the Pee Dee area. This reconstruction would support Daniel’s (1996) critique that territories may have been formed across drainages. Very little is actually known about Late Archaic settlement systems in North Carolina. From surrounding regions there is ample evidence of social organizational complexity (Smith 1986). Storage technology (i.e. baskets, storage pits, and steatite and ceramic vessels) proliferated during this period and sedentary villages with sturdily-built habitation structures became common. Subsistence intensification is Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 60 Chapter 4. Research Design also well documented in estuarine and riverine settings throughout the Mid-south and southern coastal zones. In South Carolina, shell rings appear, which may have been monumental ceremonial features around which regional Late Archaic populations were tethered (Cable et al. 1993, 1994; Michie 1980). Perhaps the best evidence of intensified, semi-sedentary habitation along a major stream in eastern North Carolina is provided by the Gaston Site Savannah River component, which included a stained midden, stonelined hearths, high artifact density, steatite vessel fragments, and full-grooved axes. Coe (1964: 119) observed that these traits suggested a “larger group occupying the site over a longer continuous period than had been true of the earlier periods.” In South Carolina, the interior creeks of the Coastal Plain were intensively inhabited by what appear to be seasonal Late Archaic residences (see Cable et al. 1994). These residences were spaced at regular intervals in a manner sufficient to effectively divide the uplands into equal foraging zones for exploitation by similarly constituted social groups, perhaps extended families. Fort Bragg may contain similar interior Late Archaic sites. The ICRMP (2001: XI-34) indicates that Woodland period occupation on the Fort is characterized by small, ephemeral camps in the uplands and small habitation sites on stream margins. Their distribution patterns are not distinguishable from Archaic patterns, with the possible exception of a tendency to be located to water features (see Clement et al. 1997). Large village sites have not been documented, although floodplains of major streams are not well represented within the boundaries of the Fort. The small ephemeral sites in the uplands are regarded as generally reflecting small nuclear or extended family camps rather than logistical camps and are generally represented by “pot busts” or vessel aggregates. This same pattern has been identified over large areas of the South Carolina Coastal Plain and similar functions have been inferred (Cable and Cantley 1998; Cable et al. 1998; Cantley and Cable 2002). Block excavations have supported this inference. These camps consist of one to a few sherd aggregates each representing a single vessel, a discrete debitage concentration, and a cluster of tools situated along the edge of the debitage concentration. The tool cluster is inferred to represent the general vicinity of a hearth, and this is supported many times by a coter- minous calcined bone concentration. This basic unit is inferred to consist of a hut containing a ceramic vessel(s), an exterior lithic reduction concentration and a hearth area positioned directly outside of the entrance to the hearth. These basic units can occur alone or in a small cluster, the latter being inferred to represent a multi-family camp. Secondary refuse deposits have yet to be isolated around these camps. It is likely that larger and more permanent Woodland habitation sites will be found in abundance along the Cape Fear and Little Rivers, but these areas are not systematically investigated at present. The ICRMP (2001:XI-36) mentioned two village candidates from riverine settings. One of these apparently contains large pit features, which is a good indicator of permanent habitation. In addition, the Cape Fear drainage contains numerous sand burial mounds dating to the Woodland period. Irwin et al. (1999) suggest that these were uninhabited or vacant centers that served as the hub for ritual ceremony and exchange in a similar manner to that specified by Clay (1998) for the Adena mound complex. In this light, the mounds would have served as an integrative mechanism for a geographically dispersed population. It is also suggested that the presence of extralocal sumptuary goods in these mounds indicates that the Woodland populations of the Sandhills were participating in an extra-local exchange network. One potential motivation that is offered for this participation was to form extra-local alliances in a situation where neighboring groups were exerting pressure on land and resources. Tracking the development and demise of this system through the Woodland period provides an important research focus for Woodland investigations on the Fort. Very little evidence documents Pee Dee culture occupation on the Fort (ICRMP 2001:XI-31). The same is true for Protohistoric and early historic aboriginal cultures (ICRMP 2001:XI-33). The near absence of Pee Dee material may indicate that the area of the Fort represented a cultural boundary in the late precontact period, separating indigenous Late Woodland groups from what may have been an intrusive Mississippian group at Town Creek (the Pee Dee culture). On the other hand, it may represent an occupation hiatus during this time frame. The reasons for the dearth of materials related to Protohistoric and early historic groups may be different. It is clear from Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 61 Chapter 4. Research Design historic records that aboriginal groups inhabited the Cape Fear River Valley in the seventeenth century (Hilton 1967:72-79). Moreover, based on the similarity of ceramics from known Protohistoric and historic villages in the northeast Piedmont (see Dickens et al. 1987; Ward and Davis 1993) to Woodland types, it is entirely possible that ceramics of this time period have simply gone unrecognized. ANALYSIS SYSTEMS The artifact analysis program developed for the project was directed toward the description and interpretation of two major classes of material, precontact ceramics and lithics. The analysis system used for each class is presented below. Precontact Ceramics The precontact ceramic program was developed in conjunction with Herbert’s (2003) variables and typological nomenclature for the southern Coastal Plain of North Carolina. Twenty-five variables were recorded during the analysis or in subsequent interpretive variable recombination. Vessel form variables included: (1) ceramic form, (2) rim type, (3) lip form, (4) rim orientation, (5) rim decoration, and (6) wall thickness. Variables of surface finish consisted of: (1) exterior treatment, (2) weave type, (3) stamping pattern, (4) cordage diameter, and (5) interior treatment. Paste variables included: (1) quartz sand density, (2) dominant quartz sand grain size, (3) clay temper type, (4) dominant clay temper size range, (5) clay temper density, (6) paste texture, (7) other temper type and (8) paste hardness. Three culture historic variables were recorded including (1) ceramic series, (2) ceramic series variant, and (3) series variant type. Two additional variables were measured, sherd size class and weight. This was done to control comparisons of type representation and to monitor formation processes contributing to the vertical and horizontal distributions of ceramic items. Finally, maximum vessel identifications were made during the analysis for assemblages recovered from intensively sampled areas. Maximum vessel counts are used later in the interpretive chapters to evaluate reoccupation and component structure. Below, each variable is described in terms of its measurement and the recognized variable states. Ceramic Form Ceramic Form refers to the part of a vessel or other ceramic object represented by an individual ceramic item. The collection consisted exclusively of container sherds (ie. body, rim and basal sherds). Small, eroded and spalled sherds that had little or no diagnostic value and were classified as residual sherds. Surface Treatment Variables Five variables were differentiated to describe aspects of surface finish on sherds. These are exterior treatment, warp/weft type, stamping pattern, cordage diameter, and interior treatment. Each of these is described below. Exterior Surface Treatment. This variable refers to the type of decoration or finishing technique exhibited on the exterior of individual sherds. In some situations the treatment was obscured by erosion and in these cases a series of observationally qualified categories were recorded (i.e. indeterminate, indeterminate decorated, indeterminate stamped, indeterminate textile, etc.). Surface treatment information was not recorded for residual sherds, although occasionally this attribute could be observed on members of this sub-sample. Predominant exterior surface treatment types consisted of fabric impressed, cord marked and check stamped. Weave Type. This variable describes variation in the types of weave manifest on fabricimpressed sherds. Following Herbert (2003), four main weave types were identified in the collection. Fabric with pliable, cordage warps and intertwined wefts of variable coarseness is described as flexible weave. The other three types of weave contain rigid, non-cordage warps and they are distinguished by the relative coarseness of the weft. Coarse wefts exceed 2 mm in diameter, medium wefts range between 1 and 2 mm in diameter, and fine wefts consist of cord- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 62 Chapter 4. Research Design age with diameters of less than 1 mm in diameter. Stamping Pattern. This variable describes the application patterns of cord marked, net impressed and simple stamped sherds. Cord marking occurred in only two basic patterns. Parallel-overstamped consisted of parallel cord stamping placed at roughly parallel and partially overlapping orientations. Included within this category are also examples of parallel stamped cord impressions without visible evidence of overstamping. In all instances, cross-stamping was oblique rather than precisely perpendicular and the application was somewhat sloppily executed. Cordage Diameter: When possible, average cord impression diameters were measured with a digital caliper on individual cord marked sherds. Diameters ranged between 1 and 2.5 mm. Interior Treatment. This variable describes the techniques used to finish the interior surface of vessels. Numerous techniques and conditions have been recognized in the ceramic collections recovered from the various delivery order packages. All of the interiors were at least smoothed, but within this wide category a wide range of completeness occurred. These categories consisted of regularized, smoothed finishes, floated and regularly smoothed, smoothed irregular, smoothed irregular/scraped, scraped, smoothed scraped, smoothed irregular wiped, and smoothed wiped. The first two categories most frequently represent bowl interiors, while the remaining types are more commonly associated with openmouthed jars. A small proportion of the observable interiors generally cannot be classified (i.e. eroded or indeterminate), due to erosion, spalling, or some other obstructing condition. Paste Variables Eight variables describing aspects of paste composition were recorded in this study. These are quartz sand density, dominant quartz sand grain size, clay temper type, dominant clay temper particle size, clay temper density, paste texture, and other temper type. Each variable is described below. Quartz Sand Density. All sherds in the collection contained some amount of quartz sand in the paste. This variable attempts to estimate macroscopi- cally the relative abundance of quartz sand in individual sherd paste bodies. Three ordinal categories were developed: sparse (S), moderate (M), and abundant (A). Sparse density describes sherd cores containing less than 5 percent quartz sand. Moderate density identifies sherd cores with approximately 5 to 15 percent quartz sand and abundant density describes those cores with > 15 percent quartz sand. Dominant Quartz Sand Grain Size. All sherds contained a range of sand grain sizes, but this variable was designed to characterize the dominant grain size(s). The Wentworth (1933) Scale was used to describe individual grain size. Grain size was ordinated in the following manner: (1) fine sand (1/16-1/4 mm in diameter), (2) medium sand (1/4-1/2 mm in diameter), (3) coarse sand (1/2-1 mm in diameter), (4) very coarse (1-2 mm in diameter), (5) granule (2-4 mm in diameter), and (6) pebble (> 4 mm). Pastes were evaluated by combinations of these size classes in an attempt to characterize dominant grain size representation. In addition, secondary dominance was recorded when relevant by separating dominant and secondary grain ranges with a forward slash in the analysis. Seven grain size combinations are commonly identified in the Fort Bragg ceramic collections associated with the delivery order packages: (1) fine, (2) fine to medium, (3) medium (4) medium to coarse, (5) medium to very coarse, (6) medium to granule, and (7) coarse to granule. Clay Temper Type. All sherds containing clay temper were assigned to the Hanover series. Within the broad category of clay temper, however, two types are recognized in the region. These are crushed clay particles without visible signs of clastic inclusions and crushed grog containing clastic inclusions. All of the material in the collection was classified as grog temper, most of which appeared to be crushed sherd. In many instances sherd surface features were observed on individual particles, but timeconstraints prevented the application of a systematic procedure to record these occurrences in an unbiased and controlled manner. Dominant Clay Temper Particle Size. A great deal of variation in clay temper size was noted in the Hanover collection. This variable was designed to group the observed variation into ordinal categories similar to the dominant size range procedure devel- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 63 Chapter 4. Research Design oped for quartz sand. Here, three size ranges of clay/ grog temper have been identified: small (< 1mm in diameter), medium (1 to 2.5 mm in diameter), and large (>2.5 mm in diameter). From this classification, four combinations were recognized: small (S), small to medium (SM), medium (M), and medium to large (ML). Clay Temper Density. The density of grog temper in sherd cores was ranked by the recognition of three variable states. Abundant (A) density refers to cores that are composed of greater than 25 percent clay temper. Moderate (M) density consisted of ranges between 10 and 25 percent of the sherd core. Finally, low (L) density was identified in situations where less than 10 percent of the sherd core was composed of sherd/grog temper. Paste Texture. Two broad categories of paste texture were present in the collection, gritty and fine. Gritty paste is friable and “gritty” to the feel due to a large amount of fine to medium sand in the paste. Fine paste texture was associated with harder ceramics containing very little fine to medium grained background sand in the paste. Larger quartz sand particles, however, were present in fine pastes, often in large amounts. Gritty pastes were most common in New River series material, but were present throughout the sequence in variable proportions. Other Temper Type. In several instances, other clastics and/or temper inclusions have been identified in sherd collections from the delivery order packages. Five different types have so far been recognized: (1) crushed sandstone, (2) crushed sandstone and feldspar sand, (3) feldspar sand, (4) crushed quartz, (5) granule sized subangular quartz sand, (6) crushed arenite/quartzite and (7) crushed and ground indurated granite. Crushed rock is diagnostic of the Yadkin series on Fort Bragg (see Herbert 2003:158). Paste Hardness. It became clear during the analysis that there existed a great range of variability in the hardness characteristics of individual sherds in the collection. A hardness scale consisting of three attribute states was developed to describe this variation. These consisted of: (1) friable, (2) compact and (3) hard. The attribute state of “friable” corresponded to sherds that were easily crushed on their edges with only minimal application of pressure with the thumb. A “compact” state was identified in instances where moderate pressure with the thumb was required to break the edge of the sherd. Finally the “hard” state was reserved for those sherds whose edges were not easily broken even with extreme thumb pressure. Generally, pottery achieves greater hardness when one or more of three conditions prevail (Shepard 1954:114). These are: (1) the pottery is made from a low-fusing, dense-firing clay, (2) the pottery is fired at a high temperature and/or (3) the pottery is fired in an atmosphere promoting liquification in the sintering process. It is common to observe in the local ceramic sequences of the Southeast, a progression from friable ceramic series to much harder, more durable pottery in the later stages of cultural development. Clearly, by the Mississippian period potters had developed technological strategies to produce extremely hard, sometimes vitrified pottery and it is quite possible that this advancement was meant to adapt to a more sedentary, agricultural life way in which durability and longevity of pots would have been advantageous. Similar adaptive technology may have been incorporated into sedentary, agricultural societies peripheral to the Southeast Appalachian Mississippian culture as well, and for the same reasons. As of yet, a sophisticated and quick method for measuring hardness in large samples of “primitive” pottery does not exist due to its porosity and heterogeneity of paste composition. The “thumb” test, although somewhat imprecise, provides a relatively objective basis for identifying relative hardness in large collections of precontact pottery. Herbert (personal communication, August, 2011) has recently pointed out to the author that mechanical devices are available that are capable of making precise measurements of vessel wall strength (see Grimshaw 1971; P. Rice 1987:357-363; Simon and Coghlan 1989). These methods could not be applied to every sherd in a large ceramic collection, but they might very well provide a basis for confirming and calibrating the “thumb” technique in characterizing what Prudence Rice (1987:360) describes as “compressive strength.” She (P. Rice 1987:354-357) reserves the term “hardness” for specifically measuring the surface strength of ceramics. It is hoped that including this additional variable in the current study will aid in distinguishing between “earlier” varieties of sand- and grog-tem- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 64 Chapter 4. Research Design pered ceramics and “later” ones. It may also point the way toward more detailed studies of numerous measures of ceramic strength in precontact ceramic series, which could inform on variation in firing technology and other factors of ceramic production containing functional and chronological significance. Sherd Wall Thickness Sherd wall thickness was measured in hundredths of mm using digital sliding calipers. Only those specimens that had both interior and exterior surfaces preserved were measured. Rim thickness was consistently taken at a point 10 cm below the lip. Sherd Size Sherd size was measured by placing sherds on a graduated and nested set of squares or rectangles encompassing specific area ranges. These area ranges were transformed into size classes in the following manner: (1) Class 1: < 1 cm2, (2) Class 2: 1 cm2 to < 4 cm2, (3) Class 3: 4 cm2 to < 9 cm2, (4) Class 4: 9 cm2 to < 16 cm2, (5) Class 5: 16 cm2 to < 25 cm2, (6) Class 6: 25 cm2 to < 36 cm2, (7) Class 7: 36 cm2 to < 49 cm2, (8) Class 8: 49 cm2 to < 64 cm2 (9) Class 9: 64 cm2 to < 81 cm2, (10) Class 10: 81 cm2 to < 100 cm2, (11) Class 11: 100 cm2 to < 121 cm2, and (12) Class 12: 121 cm2 to < 144 cm2. Weight The weight of individual sherds was measured with a digital scale. Readings were recorded in hundredths of grams. Rim Variables Four variables were measured to characterize aspects of rim morphology. These were: (1) Rim Type, (2) Lip Form, (3) Lip/Neck Orientation, and (4) Rim Decoration. Rim Type describes the construction of the rim. All of the rims (n=28) in the DO4 collection were plain. Lip Form monitored the shape of rim lips. Four shapes are generally recognized: (1) flat, (2) sub-rounded, (3) rounded, and (4) truncated/rounded. Neck Orientation describes the angle at which a neck diverges from the body of a vessel. Most of necks in the collection were straight, which would indicate a dominance of open-mouthed jars. Other neck orientations generally found in the delivery order packages include slightly incurved, re-curved and flared. These forms would suggest, just as did the interior finish data, a minority component of incurved and re-curved bowls and constricted neck jars. Rim Decoration recorded the presence of specially applied decorative elements to lips or to the area just below the rim on the interior surface. Fabric impressions and cord marking are the most common rim decoration types in the delivery order packages. Vessel Identification Sherds that were recognized to be members of individual vessels were identified during the course of the analysis and were recorded in a column on the analysis sheet under the heading “VID,” which stands for “vessel identification number.” Identified vessels ranged from single sherds to larger sherd aggregates that would qualify as partial or reconstructible vessels. Culture Historic Identification Three variables were recorded for culture historic identification purposes. “Series” corresponds to the classic series defined for the region by Herbert (2003). These include Thom’s Creek, New River, Yadkin, Cape Fear, Hanover I, and Hanover II. “Series Variant” refers to variations in the traditional series categories based on variation in paste. Finally, “Series Variant Type” refers to combinations of surface treatment types and series variants. Lithic Artifacts The precontact lithic program was organized around techno-functional, core reduction, and culture-historic themes. The structure of the analysis was modeled on the Haw River attribute system (see Claggett and Cable 1982) and the system developed by Daniel (1998) for the Hardaway Site reanalysis. Projectile point and other formal tool types were iden- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 65 Chapter 4. Research Design tified with reference to pertinent type descriptions from adjoining regions (see Anderson et al. 1982, 1990; Cantley 1998, 2002; Chapman 1985; Claggett and Cable 1982; Coe 1964, 1995; Daniel 1996, 1998; Goodyear et al. 1989; Keel 1976; Oliver 1985; Tippit and Daniel 1987; Ward and Davis 1993). Lithic Raw Material Classes Lithic raw materials utilized by precontact groups in the Sandhills of North Carolina consist primarily of Slate Belt metavolcanics, vein and crystalline quartz and, rarely, chert, slate, metasiltstone, unidentified silicates and orthoquartzite. Most of the quartz appears to have derived from stream gravels based on the high incidence of water-worn cobbles in the archaeological collections in the region. Orthoquartzite is typically found in Coastal Plain formations and is abundant in the Black Mingo and Black Creek formations of central South Carolina along the Santee and Black River terraces (see Anderson et al. 1982:). The origins of the rare silicates and cherts are unclear. The bulk of the metavolcanic material is seen as coming from distant bedrock outcrop sources in the Carolina Slate Belt (see Daniel and Butler 1996; Benson 2002). Before this thesis is accepted without critical evaluation, however, there are factors concerning the assemblages recovered from archaeological sites in the region that must be considered. Numerous of the metavolcanic debitage assemblages from sites tested in this delivery order do not appear to exhibit evidence of the necessary raw material conservation strategies that would be expected of material being transported over distances as far away (50 or more miles) as the Uwharrie Mountains or other Slate Belt quarries. Raw material being transported over these distances would be expected to be brought in as portable, prepared cores and flake blanks and the debitage assemblages would primarily consist of late stage reduction flakes with very little cortex. Instead, many of the debitage concentrations are characterized by early and late stage reduction debitage and an unexpectedly high incidence of cortex, indicating the reduction of unmodified cores. Moreover, some of these concentrations contain numerous preform rejects, another line of evidence suggesting nearby metavolcanic sources. Finally, unmodified metavolcanic cores with water-worn cortical rinds have been recovered, suggesting that a significant proportion of the metavolcanic material used by precontact groups in the region was probably extracted from local stream gravels. Moore and Irwin (2003) discuss one such gravel source (Cape Fear Group) discovered on a Cape Fear River terrace, east of Fayetteville. Gravels of this sort occur extensively across major stream terrace systems at old point bar and channel formations and we should expect numerous gravel sources throughout the Cape Fear River terraces and also in the major tributary systems of Little River and Rock Fish Creek. These features can occur up to several miles from current major river channels. During precontact times, such sources would have been accessible at locations where drainage channels of all sizes cut across these old point and channel bar formations and on some eroded ridges. The original source of these gravels, of course, would have been hard rock outcrops in the Carolina Slate Belt and we would expect the metavolcanic composition of the gravels to reflect the full range of the drainage catchment zone of particular streams. Consequently, such gravel beds should reflect a wider range of variability of rock types than would be present at any single outcrop source. Of interest in this regard is the Neodynium isotopic composition of the Cape Fear gravel group reported by Moore and Irwin (2003: Figure 4). Four samples were analyzed from this group, two of which plotted clearly outside of all of the Slate Belt quarry locations that were sampled in the study while the remaining two were positioned significantly closer to the Uwharrie 1 Slate Belt group than they were to the two outlier samples from the Cape Fear gravel group. A larger sample of gravels from this source would probably show an even greater range of variability reflecting the alluvial dynamic of the drainage system. A concerted effort should be made to not only describe the Piedmont rock outcrop quarries, but also the composition of gravel stringers associated with local stream terraces surrounding Fort Bragg to build a complete lithic procurement model for the region. Physical variability within each main raw material category was partitioned into categories with two purposes in mind. First, the assemblages were characterized in terms of traditional hard rock geological terminology to achieve comparability with previous projects. Second, an attempt was made to Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 66 Chapter 4. Research Design further partition some of the hard rock classes into “core groupings” to distinguish between similar raw material subclasses that may have derived from different cores or post-depositional weathering processes. This was done to aid spatial analysis in closeinterval shovel test grids where overlapping or complex multi-household occupations were suspected. Core grouping analysis was undertaken principally in the metavolcanic category where subtle differences in patination color, core color, phenocryst size and density, and matrix texture could be successfully monitored and distinguished. Debitage The debitage classification, as stated above, is adapted from formal (Bradley 1973; Frison and Bradley 1980; House and Wogaman 1978; Newcomer 1971), attribute (Moore 2002; Shott 1994), and mass (Ahler 1989; Moore 2002) analysis approaches. Debitage classes were devised to identify both reduction stages and reduction/production systems (ie. biface core reduction, directional core reduction, and flake blank production). This classification was supplemented with the recording of four attributes: (1) size class, (2) percent cortex, (3) condition and (4) weight. Each of these attributes and the utilized debitage classes area discussed below. Size Class, Condition and Weight. Each piece of debitage was subjected to size class and weight measurements to provide a basis for undertaking mass analysis. Debitage size class distributions are often used to characterize the relative stage of reduction represented across a series of sites or assemblages (Ahler 1989; More 2002; Sassaman 1993; Shott 1994). In the current study, we were specifically interested in generating size class distributions to derive a picture of the kind and stage of reduction associated with individual occupations. Mass analysis is commonly applied to only whole pieces of debitage, but this practice depresses sample sizes and, consequently, can skew the results. Control was established by recording the condition (whole or broken) of each piece of debitage, allowing the analysis to include either exclusively whole pieces or the entire sample. Size class was measured by placing an individual item on a graduated and nested set of squares encompassing specific area ranges. These area ranges were transformed into size classes in the following manner: (1) Class 1: < 0.25 cm2, (2) Class 2: 0.25 cm2 to 1 cm2, (3) Class 3: 1 cm2 to 4 cm2, (4) Class 4: 4 cm2to 9 cm2 (5) Class 5: 9 cm2 to 16 cm2, (6) Class 6: 16 cm2 to 25 cm2, (7) Class 7: 25 cm2 to 36 cm2, (8) Class 8: 36 cm2 to 49 cm2, (9) Class 9: 49 cm2 to 64 cm2, and (10) Class 10: 64 cm2 to 81 cm2. Since most of the debitage recovered was small, the size class analysis was supplemented by wieth measurements in hundredths of grams to better differentiate mass comparisons of the smaller flake ranges. Percent Cortex. Cortex, either from waterworn stream gravel or from weathered talus debris, was occasionally observed on debitage. It is generally true that assemblages containing higher proportions of debitage with cortex indicate earlier stage reduction activities. The relative percentage of cortex on a dorsal or exposed face of a piece of debitage was recorded as an ordinal variable. A value of zero was assigned to debitage lacking cortex, while values of 1, 2, 3, and 4 were assigned to pieces respectively exhibiting less than 25 percent, 25 to 50 percent, 50 to 75 percent, and 75 to 100 percent cortex. Formal Debitage Classes. Most of the debitage collected during the project represent by-products of biface reduction strategies. The two main strategies consist of flake blank production and subsequent bifacial shaping of bifacial tools and projectile points, and biface core thinning to make bifacial tools. The terminology used for this debitage analysis is principally an adaptation of the “FBR model” developed by House and Ballenger (1976:89-92) and House (1977). (1) Flake of Bifacial Retouch (FBR). This class represents relatively large flakes removed primarily during the early stage of biface reduction. They are relatively thin, exhibit feathered edges, and generally contain numerous flat flake scares on their dorsal surfaces. Platforms are thick and multifaceted, containing a portion of the former bifacial edge of the core. In Ahler’s (1989) terminology, FBRs represent non-marginal percussion flakes in which the point of impact is directed behind the edge of the striking platform. Striking platform angles are acute. Generally, it is assumed that FBRs have been removed by a hard hammer. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 67 Chapter 4. Research Design (2) Biface Thinning Flake (BTF). These are small flakes with thin striking platforms consistent with marginal percussion strategies (see Ahler 1989). BTFs are assumed to have been produced during late stage core reduction and in final tool shaping and edge trimming. Both hard and soft percussion were used to remove these flakes and although most represent the byproducts of biface reduction, some were no doubt removed during directional core reduction and in the manufacture of unifacial tools. (3) Core Flake. Core flakes exhibit clear evidence of being removed from directional cores. Striking platforms exhibit none or few facets and platform angles achieve nearly 90°. Dorsal scars are generally linear, representing the prior removal of elongate flakes from core faces. Platforms are thick and represent non-marginal percussion flaking with a hard hammer. This class includes only flakes removed in preparing a directional core for flake blank removal. Flake blanks serving as cores for the production of tools are discussed within the framework of core classes. As familiarity with the lithic assemblages at Fort Bragg as progressed, it has become increasingly evident that the reduction of block cores to produce flake blanks for unifacial and bifacial tool manufacture constitutes the primary debitage signature. Traditional biface core reduction was of only limited occurence as is demonstrated by the predominance of core flakes in the collections and the generally small sizes of biface reduction flake types (i.e. FBRs and BTFs). (4) Core Rejuvination Flake. Flakes fitting this class are generally wedge-shaped chunks of material exhibiting platform faceting on one side. It is assumed that these flakes represent the repairing of the striking platform edge on a directional core to restore conditions appropriate for flake blank removal. (5) Chunk. These items are relatively large, angular pieces of debitage that lack striking platforms and other characteristics of typical conchoidal flakes (House and Ballenger 1977:89). Chunks are differentiated from core fragments by a lack of striking platforms and flake scars from detached flakes. (6) Flake Fragments. This class consists of debitage containing some characteristics of conchoidal fracture (i.e. flake scars on the dorsal surface, bulbs of percussion, feathered margins, etc.), but lack a striking platform that would allow a more precise identification (i.e. core flake, FBR, BTF, core rejuvination flake). Cores Four major categories of core were recognized in the lithic assemblage: (1) Unmodified Cores, (2) Directional Cores, (3) Core Fragments, and (4) Flake Blanks. Biface cores, which are also used in flake blank production, will be discussed in the succeeding section where they are integrated into a continuous biface reduction stage classification. Unmodified Cores. Unmodified masses of lithic raw material were classified as unmodified cores because they would have required human transport to be deposited in site locations. Directional Cores. This core type consists of relatively large masses of material that exhibit flake scars running in one or more directions from flat striking platforms rather than from biface edges. The platform angle approximates 90°. In certain cases it was possible to discern the number of directions inwhich flakes were removed. Uni-directional cores exhibit a single platform and a single flake removal direction. Bi-directional cores exhibit two platforms and multi-directional cores include cores exhibiting more than tow directions of flake removal. Core Fragments. This category consists of angular pieces of raw material containing portions of one or more flake scars suggestive of directional core reduction. They lack evidence of a striking platform, however, and as a result they cannot be further classified. Some of these specimens may represent fragments from bifaces or biface cores. Flake Blanks. Flake blanks are flakes removed from a core for the purpose of further modification (Bradley 1973:6) or use as a tool. Both directional (core flakes and blade flakes) and biface core (Levallois flakes, FBRs) flakes are represented in this class. Core flakes have been defined previously. Blade flakes are also core flakes, but they are elongated and approach the 1:2 width-to-length ratio characteristic of blades. Levallois flakes are Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 68 Chapter 4. Research Design large flakes removed longitudinally from the face of a biface core. The dorsal surface contains the facial ridge of the former biface core and gives the appearance of a turtle’s back. The use of the term “Levallois flake” may be somewhat misleading because the term was defined for a special kind of biface thinning flake removed from handaxes in European Acheulian assemblages. However, the technique employed is identical to that observed on some flakes from the study collection. Moreover, Levallois flakes are a common feature of most bifacial reduction industries throughout the world (see Newcomer 1971:90). Bifaces Bifaces were partitioned into two major categories, cores/performs and tools. The biface core/ perform group was further divided into three classes describing a continuous reduction trajectory from early to late stage production following the classification outlined by Daniel (2002). Biface Cores/Preforms. These items were derived from both large mass packages (i.e. unmodified and/or prepared cores) and flake blanks. Determination that a biface was derived from a flake blank was made when flake characteristics were observed on the item, such as remnant striking platforms, flake curvature, or bulbs of percussion from flake detachment. Each biface category as adapted from Daniel (2002:51-54) is described below. (1) Type I Bifaces. These bifaces are irregularly shaped flake blanks and core masses with large conchoidal scars forming at least one section of a sinuous edge along the core edge. No evidence of secondary thinning scars is observable and a relatively high thickness to width ratio obtains. This category includes both biface cores, which were constructed to produce flake blanks, and early stage preform rejects. (2) Type II Bifaces. These bifaces, in the main, represent intermediate stage preform rejects. They are roughly flaked and exhibit only large, conchoidal flake scars. No edge trimming is evident, but an ovate to lanceolate outline and the perimeter of the edge margin is entirely retouched. Facial retouch, however, may not be complete and features of original flake blanks may be observable. (3) Type III Bifaces. Type III Bifaces represent late stage preform rejects. The edges exhibit secondary thinning and shaping scars and facial retouch is globally distributed. Outline shapes are well defined and reflect ovate and lanceolate forms. Sometimes features of diagnostic projectile point haft elements are present, allowing correlation with specific culture-historic types Most of these specimens exhibit step and hinge fractures that terminated production. Stage III Bifaces tend to exhibit the lowest thicknessto-width ratios. Biface Tools. These items represent bifacially shaped items that were used as tools instead of rejects form the manufacturing process. Projectile Points Projectile points consist of all bifacially flaked tools with retouched haft elements or fragments with fine bifacial retouch characteristic of finished projectile points. Generally, projectile points represent the hafted tip elements of spear shafts or arrows, although they also functioned as hafted knives for butchering and other activities. Their importance to archaeological analysis is enhanced due to the distinctive stylistic chnages expressed in the forms through time. Unifaces Unifaces consist of tools that are retouched along the margins of a flake in a single direction (unifacial edge shaping). Retouch was applied to shape a working edge or bit and, sometimes, a butt element for hafting. Within this tool class are a number of specific uniface forms that are discussed when appropriate in the site discussions. Retouched Flake Tools This category consists of thin flakes that were modified by retouch to create a range of tool types that are not appropriately subsumed under the rubric of uniface types. Generally such tools exhibit serated or bifacial edges produced from retouch. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 69 Chapter 4. Research Design Utilized Flakes Previous Investigations This category consists of unmodified flakes exhibiting wear patterns on one or more edges, indicating their use as expedient tools. Generally, use wear is manifested in the form of edge attrition and bifacial nibbling scars indicative of cutting functions. Secondarily, some flakes exhibit unifacial nibbling and edge polish suggesting light to heavy scraping functions. Cobble Tools This category includes itmes fashioned from river cobbles that exhibit faceting or battering wear indicative of use as hammer stones or grinding stones. Rarely, pecked stone artifacts such as banner stones (i.e. atl-atl weights) are identified in Fort Bragg assemblages. FIELD METHODS The original delivery order called for 139 m3 of shovel test volume and 14 m3 of test unit volume spread across the six sites in variable amounts (Table 1). Site-by-site volume allocations were adjusted as site boundaries expanded and strategies shifted. The total volume excavated of 193.2 m3, however, exceeded the original allocation by a sustantial amount. This surplus volume was excavated without additional cost to the government. Sites to be tested under this contract had an associated record of previous investigation. All pertinent reports, maps, site forms, and collections were reviewed prior to fieldwork. None of the reports suggested testing plans for any of the sites in the package. Efforts were made to relocate and map all previously excavated shovel tests on the sites and when successful were plotted on the field maps. Artifact density maps of the Phase I surveys were not generated because the shovel test patterns were irregular and not amenable detailed identifications of artifact concentrations. Since the Phase I grids were irregular and difficult to relocate, we established new grids at locations approximating the previous site datum according to GPS coordinates. Site Map A map of each site was generated using an electronic transit and meter tapes. A standard grid was established on-site and was used to locate and map all shovel tests and test units. The grid coordinate system was calibrated in metric units and originated at the site datum. The site datum was assigned the arbitrary coordinates “N500 E500.” This system was used to assign provenience data to all recovery contexts. The grid was constructed on a 10-meter interval. The site data established during earlier surveys were relocated and used as the data to orient the Table 1. Summary of Cubic Meter Allocations by Site, Delivery Order 5. SUMMARY OF EXCAVATED VOLUME, DO5 SITE 31HK2502 31HK2510 31HK2521 31RH491 31RH480 31RH478 GRAND TOTAL SHOVEL TESTS Original Stage I Allocation 17.0 9.0 33.0 28.1 21.0 13.5 55.0 42.6 12.0 15.0 1.0 10.8 139.0 119.0 Stage II/III Total 10.2 10.9 9.7 12.0 6.1 7.4 56.3 19.2 39.0 23.2 54.6 21.1 18.2 175.3 TEST UNITS Original Actual Allocation 2.0 1.9 3.0 2.1 2.0 4.7 3.0 3.0 2.0 2.8 2.0 3.4 14.0 17.9 GRAND TOTAL 21.1 41.1 27.9 57.6 23.9 21.6 193.2 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 70 Chapter 4. Research Design Phase II investigation. The geographic location of the datum was recorded using a Trimble GeoExplorer 3 GPS unit, which has 1-5 meter accuracy with base station DGPS correction. Locations of all shovel tests, test units, and cultural and natural features were mapped and illustrated on the site base maps. Elevation data was recorded for all old and new excavation units mapped on the site. The topographic features of the sites were mapped using a point density of 100 per hectare and a contour interval of 30-cm resolution was produced for each site. Contour mapping extended beyond the established site boundaries a sufficient distance to accurately depict the topographic situation of the site. All previously excavated units found in the field were mapped. Earlier maps of each site are illustrated in the report indicating the locations of previous nonrelocated test units relative to the updated site maps. Isolated surface artifact finds were piece-plotted and their locations indicated on the site map. When dense overgrowth is encountered, lines of site were cut along grid lines and along other axes critical for the accurate mapping of the site. were taken to a depth of at least 30 cm and if substrate was not reached, they were excavated to depths of at least 75 cm. Any shovel tests excavated during this project extended to depths equal to those excavated during previous investigations at individual sites. Fill from shovel tests was screened through 1/4-inch mesh hardware cloth to recover artifacts. In concordance with the research design, shovel tests were excavated in three arbitrary levels when possible. Level 1 extended to 30 cm below ground surface, Level 2 ranged between 30 and 55 cm, and Level 3 included deposits below 55 cm. A field log was maintained for all shovel tests and included the following information: (1) grid coordinates, (2) size, (3) maximum depth, (4) Number of Levels, (5) soil zones, (6) soil zone depth, and (7) artifact contents by level. Soil zones were described in standard, soil science terminology and coloration was recorded using Munsell color charts. Sufficient numbers of profiles were drawn of shovel test exposures to clearly delineate natural and cultural strata. The shovel test log data is provided in an Excel spreadsheet as appendix in this report. All shovel tests were backfilled prior to the end of fieldwork at each site. Site Boundary Definition Test Units Site boundaries were determined through the deployment of 10-meter shovel test grids oriented along cardinal directions. Shovel tests excavation continued along each grid point until two negative shovel tests were encountered along each cardinal direction. Surface artifact finds were also considered to represent positive tests and were closed out in the same manner as positive shovel tests when they occurred on the periphery of a site. The precise site boundary was determined by connecting the midpoints between the last positive and the first negative test in all directions. Test units measured 1 x 1 m or larger. Test units were not be placed in areas of low artifact density or outside of the established site boundaries unless this placement was approved by the NPS COTR or Fort Bragg CRM personnel. Excavation proceeded in arbitrary 10 cm levels with the exception of plow zones. Plow zones were removed as single levels. In situations where it was not obvious whether the upper A-horizon represents a plow zone or some other kind of deposit, excavation proceeded in arbitrary 10 cm levels. In some situations, 5 cm arbitrary levels were excavated to better define potential vertical separation of cultural deposits. All test units were excavated to compact B-Horizons and at least one unit at each site was excavated into the subsoil to evaluate its potential to yield pre-Clovis cultural deposits. Dispersion of individual test units was the preferred placement strategy unless an unusual situation arose that required a larger contiguous exposure, such as the definition of a feature or deep deposits. Shovel Tests Shovel tests were square and measured 30 x 30 cm. Square wooden templates were deployed to maintain a uniform shovel test size. Each test was excavated to subsoil, which was defined as B-Horizons of Late Pleistocene age or older with little likelihood of containing cultural deposits. All shovel tests Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 71 Chapter 4. Research Design The placement of test units was guided and justified with reference to artifact density maps generated from shovel test data. Deposits were screened through 1/4-inch hardware cloth to recover artifacts. Scaled profile drawings of all sidewalls were made and all soil zones and strata were described using standard soil science terminology. Deposit coloration was described using a Munsell color chart. Scaled plan and profile drawings were made of each feature, as well as staged drawings depicting the order of fill removal. Feature data included horizontal dimensions, depth, orientation, and associations. Test Unit logs were maintained in the field. Photography Both Black and White and Color 35-mm photographs were taken of each subject. Photographs documented the site area, field activities, excavation units, and other findings of the investigation conducted at each site. At least two photos of the site area were taken. All photos of units and features contain scales, north arrows, and a restaurant-style menu board recording site number, provenience, date, and subject. A photo log was maintained in the field and included the following information: 1) roll #, 2) negative # for print film, 3) slide # for slide film, 4) Fort Bragg Accession #, 5) name of contractor, 6) delivery order number, 7) direction of view, 8) subject matter, and 9) date. This information is placed on each slide or print. GPS Data A Trimble GeoExplorer 3 unit was used to take GPS readings in the field. The unit has an accuracy rating of 1 to 5 m with post-processing software differential correction. Differential correction was performed using GPS Pathfinder Office software. GPS data were collected in WGS84, NAD27, and NAD83 formats, and are supplied in Appendix A. NRHP EVALUATIONS The main goal of the evaluative process to be implemented within the framework of this proj- ect was to arrive at definitive statements concerning the eligibility of the targeted properties. In the past, the common approach, as in many areas, has been to make such decisions within a virtual vacuum because a regional database has not been built. A database can provide a powerful tool for determining eligibility for four reasons. First, it will quantify what we know about the cultural systems in the region. Second, it will supply a basis for assessing redundancy by allowing for the stratification of the study area into meaningful environmental units in which the representative completeness of site or component types can be monitored. Third, it will serve as a basis for linking historic contexts, which are component-based in their construction, with the archaeological record. Fourth, it provides managers with a more sophisticated model for identifying which data values at a potentially eligible site are actually important within the perspective of the entire region. That is, a site may contain multiple components, but through regional quantification it may be determined that the information from only one or a small number of the components is important due to redundancy. This would allow for much more specific and efficient data recovery programs that focus on a manageable fraction of the occupation history of a particular site. There is an understandable concern on the part of researchers to recover information on all occupations when data recovery projects are implemented and this is largely the consequence of site-based evaluations. Unfortunately, this is an extremely wasteful approach and when we must consider what to do with scarce funding, it behooves us to seek other approaches that will achieve a more optimal fit with our goals and objectives. It is the opinion of Palmetto Research that we will not be successful in implementing these new directions until we begin to realize the research potential of survey and testing data, which are the foundations of regional databases. NRHP evaluations in this report are made within the framework of a developing regional database. Significance assessments are linked to the cultural and historical contexts discussed in the ICRMP (2001) and in this document through a consideration of the role each identified component played in its respective regional system. The regional database quantifies the relative importance of each component for reconstructing regional systems and data redundancy is analyzed. Detailed data recovery plans are developed for each site considered eligible. These Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 72 Chapter 4. Research Design plans focus on the relevant components and present efficient research programs that specify the objectives of excavation and appropriate field and analytical strategies to achieve these goals. Quantitative estimates of the volume and location of excavation units required are presented and the method of analysis is detailed. CURATION All artifacts, field and analysis notes, photographs, slides and negatives, electronic media, collected, generated, and/or produced as a consequence of this project is to be prepared for curation at Fort Bragg. Curation procedures were performed in accordance with the Archaeological Curation Standards and Guidelines, Office of State Archaeology, Guidelines for the Disposition of Archaeological and Human Remains, 36 CFR 79 (Curation of Federally Owned and Administered Archaeological Collections, and the Fort Bragg Curation Guidellines). ratory records, artifact inventories) were submitted to Fort Bragg for permanent curation. NATIVE AMERICAN GRAVES PROTECTION AND REPATRIATION ACT (NAGPRA) No Native American human remains were recovered during the project. If they had been during the course of this project, the Fort Bragg Cultural Resource Manager/Archaeologist would have notified immediately. In addition, work would cease in any unit producing the human remains until proper consultation could be arranged. Native American groups with potential interest in Native American Graves Protection and Repatriation Act-related cultural resources identified at Fort Bragg during the course of testing carried out under this contract will be notified of project results by Fort Bragg in accordance with Fort Bragg NAGPRA Standing Operating Procedures. Human remains will also be treated in accordance with existing federal guidelines All materials collected during the project, including artifacts, floral and faunal remains, and soil samples were cleaned, stabilized and treated as appropriate. All collection materials were clearly labeled using a permanent medium. Provenience data was organized by a bag list. Artifacts were separated into two classes and boxed separately by site. Class 1 artifacts include diagnostic or extraordinary items that are likely to be examined regularly by researchers at Fort Bragg or that are suitable for exhibition (e. g. projectile points, diagnostic sherds, brass buttons, etc.). Class 2 artifacts consist of items without diagnostic value or that are unlikely to be examined regularly by researchers at Fort Bragg (e.g., lithic debitage, residual ceramic sherds, nail fragments, soil samples). Associated records include site forms, original field notes, prepared maps or drawings, photographic materials, oral histories, artifact inventories, laboratory reports, computerized data on CD, diskette, or tape, NRHP nomination forms, reports, bibliography of all resources consulted including public and archival records, and administrative records (36 CFR 79.4(a)(2)). All original paper records (e.g., field notes, site maps, topographic quad maps, laboPhase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 73 Chapter 4. Research Design Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 74 Chapter 5. 31HK2502 Site 31HK2502 is located in Hoke County in the western portion of Fort Bragg. The site can be reached by traveling approximately 0.45 miles eastnortheast along Firebreak Road 33 from its juncture with Moore County Road to the west bank of Rockfish Creek. At this point, the site can be accessed by travelling by foot approximately 55 m northwest from the road. Topographically, the site occupies what appears to be an alluvial levee feature formed immediately adjacent to the channel of Rockfish Creek situated just bellow an upland terrace ridge (Figures 14 and 15). During wet periods, the levee frequently is surrounded by floodwaters impounded by the higher terrace slope to the west. The SCS soil survey for Hoke County, NC (Hudson 1981:81) classifies the landform as Blaney loamy sand (2 to 8 percent slopes), a well drained soil that typically forms on side slopes and narrow upland ridges. A small, unnamed tributary disembogues into the creek on its eastern bank directly opposite the site. Substantial tributary confluences occur at 0.6 km (Calf Branch) and 1.2 km (Piney Bottom Creek) north of the site and at 1.6 km (Gum Branch) south of it. Although the site has been subject to tree thinning, the surface is relatively free of modern disturbances from road construction (Figures 16 and 17). Elevations immediately on-site ranged between about 70 m amsl at the edge of the creek channel to about 71.7 m amsl along the apex of the levee (Figure 18). Average slope from the creek channel to the apex of the levee is estimated at 4.3 percent. The northwestern end of the site has been eroded by channeling cut during flood stage. Maximum site dimensions were documented through 10-m interval shovel testing at 130 m northwest-southeast x 40 m northeast-southwest, covering an area of approximately 0.72 acres. Vegetation across the site consisted of a thinned stand of mature long-leaf pines and bottomland hardwoods, with an understory of turkey oak and wiregrass. PREVIOUS RESEARCH Site 31HK2502 was originally recorded by TRC Garrow Associates, Inc in the spring of 2003 during a Phase I survey of portions of the DD3 Training Area (Grunden and Ruggiero 2006:402-403). The site boundaries established at that time covered the entire levee feature and correspond closely to those later established by PRI (Figure 19). Site definition proceeded through the deployment of 15 m-interval shovel tests, of which 24 were excavated in total. Eleven tests running along the spine of the levee produced artifacts, while negative tests were recorded around the slope of the landform leading down to the creek channel and to an inchoate drainage channel behind and to the west of the levee. Grunden and Ruggiero (2006:402) described three soil zones exposed in shovel tests and in a 50 cm x 50 cm test unit. The upper zone was composed of a dark gray loamy sand and extended to a depth of 20 cm bs. The intermediate zone consisted of a light Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 75 Chapter 5. 31HK2502 656000 657000 3885000 3885000 3886000 655000 3886000 654000 3884000 3883000 3882000 3882000 3883000 3884000 31HK2502 654000 655000 656000 657000 Figure 14. Location of Site 31HK2502 (McCain, NC 7.5’ USGS Quadrangle) Scale: 1:24,000. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 76 655000 655200 655400 655800 655600 656000 eak 31 3884800 3884800 Chapter 5. 31HK2502 3884600 3884600 Firebr Ro F 32 3884400 h C r ee k fis ck 3884400 ak re ireb 3884200 3884200 31HK2502 eak 33 or eC ou nt yR 3884000 oa d 3883400 3883400 3883600 Mo 3883800 eak 34 Firebr 3883600 3883800 3884000 Firebr 655000 655200 655400 655600 655800 656000 Figure 15. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31HK2502 (Scale: 1 inch = 200m, Projection: UTM, NAD83). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 77 Chapter 5. 31HK2502 Figure 16. Site 31HK2502, Looking East into Rockfish Creek Floodplain from Site Datum. yellowish brown sand and terminated at 80 cm bs at the contact with an olive yellow sandy clay substrate. The recovered artifact assemblage (n=38) was composed of quartz, rhyolite and metavolcanic debitage, a retouched flake of rhyolite and a metavolcanic core/ cobble fragment. Most of the artifacts were recovered between 20 and 60 cm bs, although several may have originated at 65 to 70 cm bs. The surveyors observed that the landform was virtually intact and had sustained few modern disturbances other than prescribed bur activities. They concluded that the deposits at the site were most likely intact and that the chances are good that cultural features are preserved in the matrix. The site was recommended to be potentially eligible for inclusion on the National Register of Historic Places due to the ability of the site to inform on site structure and precontact adaptation in the Sandhills. FIELDWORK OVERVIEW Fieldwork activities were implemented in four stages, each stage building on the information generated from the previous stage. Stage I investigation consisted of a program of 10 m-interval shovel tests that were excavated to establish firm site boundaries. Site boundaries were determined by the documentation of two consecutive negative shovel tests in all directions on a 10-m grid, or in some cases when wetland was encountered. Once artifact distributions from the Stage I program were mapped and evaluated, the Stage II investigation involved the excavation of closer-interval shovel tests at 5 m-intervals in most of the artifact bearing areas of the site. Stage III investigations involved the deployment of closeinterval shovel tests of 1.25 m-intervals around the perimeter of targeted shovel tests yielding deposits Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 78 Chapter 5. 31HK2502 Figure 17. Site 31HK2502, Looking North into Rockfish Creek Floodplain from N530/E480. deemed of theoretical importance for evaluating the site. These close-interval shovel test locations were identified as sample blocks and generally consisted of eight shovel tests excavated around a targeted Stage I or Stage II shovel test. The sample blocks were also instrumental in guiding the precise placement of 1 x 1 m test units, which constituted the final stage of the investigation. Test units were placed in locations judged to provide optimal opportunities to reconstruct the occupation history of the site and to assess the integrity and scientific importance of the deposits. the Phase I system. Relative topographic elevations at each 10 m grid intersection were recorded with the total station and marked with a pin flag. The elevation values (in meters) illustrated on the base map (Figure 18) are approximated real elevations based on the correlation of GPS points with a geo-referenced copy of the USGS, McCain, NC 7.5-minute topographic quadrangle map. Cultural and landform features were also mapped and additional elevations were taken when appropriate to produce the topographic map. Stage I investigations on site 31HK2502 were initiated by re-locating the datum established by Grunden and Ruggiero (2006) and laying out a 10-m interval grid with a total station. The datum, which was placed at the location of survey shovel test T65J3, was easily relocated and was assigned the Phase II grid coordinates of N500/E500 in agreement with Shovel tests measured 30-x-30-cm and were square in shape. Excavation proceeded in 10 cm arbitrary levels measured from the surface. Fortyfour (44) Stage I shovel tests were excavated at 10-m intervals to define the limits of the site (Figure 20). Twenty (20) of the tests were positive. Maximum site boundaries were established at 130 m x 40 m, cover- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 79 Chapter 5. 31HK2502 620 610 70.5 71.0 600 Floodplain 590 Floodplain 580 71 .5 570 .0 72 560 550 . 540 31HK2502 530 71.5 Rock fi 70 .5 ek Cre sh 520 71 .0 71 .5 510 Floodplain 500 72 .0 71.5 490 72.5 480 73.0 470 460 N 0 450 440 400 410 40 m 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 18. Site 31HK2502, Base Map. ing an area of about 2,900 m2, or approximately 0.72 acres. Mean shovel test depth for the Stage I sample was calculated at 60.45 cm bs (SD=12.38) with a mode of 60 cm bs and a range of 20 to 80 cm bs. Excavation was terminated in the upper 5 to 10 cm of the substrate, an olive yellow (2.5Y 6/6) to brownish yellow (7.5 YR 6/6) fine sandy loam. The SCS soil survey for Hoke County (Hudson 1984:31) maps the area of the site as Blaney loamy sand, which is a well-drained upland soil situated on side slopes and Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 80 - 515 - 500 - 485 - 470 - 455 - 440 - 425 Chapter 5. 31HK2502 - 575 TR 64 TR 65 - 560 Ro 62-J2 ckf ish Cre - 545 ek 64-J1 - 530 South Wall Centimeters 0 10 64-J2 2.5Y 4/1 Dark gray Loamy sand - 515 31HK2502 - 500 65-J3 20 30 - 485 2.5Y 6/4 Light yellowish brown Sand - 470 2.5 6/8 Olive Yellow Sandy clay S GPS datum/Positive ST Negative transect ST Positive ST Positive transect ST Negative ST Site boundary Surface positive (if any) Contour (approximate) 50 cm x 50 cm test unit Wetland (if any) N Scale in Meters 0 15 30 45 Shovel test pits shown are not to scale Figure 19. Site 31HK2502, Phase I Survey Sketch Map, TRC (from Grunden and Ruggiero 2006). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 81 Chapter 5. 31HK2502 620 610 70.5 71.0 600 Floodplain 590 Floodplain 580 71 .5 570 .0 72 560 550 . 540 31HK2502 530 71.5 Rock fi 70 .5 ek Cre sh 520 71 .0 71 .5 510 Floodplain 500 72 .0 71.5 490 72.5 480 73.0 470 Positive ST Negative ST 460 0 450 440 N 400 410 40 m 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 20. Site 31HK2502, Stage I Shovel Test Sample. narrow ridges. The upper portion of the on-site soil profile corresponds fairly well to the typical Blaney pedon. The A-horizon is weakly formed, extending to a depth of no more than about 10 to 15 cm bs. The upper 5 cm of the A-horizon evidences organic contributions from prescribed burn episodes. The “undis- turbed” A-horizon is composed of grayish brown (10YR 5/2 to 2.5Y 5/2) sandy loam. The light coloration of the A-horizon and its irregular contact with the underlying E-horizon sequence strongly suggests that its formation is natural and has not been impacted by plowing. The upper E-horizon (E1) layer is light Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 82 Chapter 5. 31HK2502 yellowish brown (2.5Y 6/3) to pale brown (10YR 6/3) fine sand, while the lower (E2) layer is lighter, ranging from very pale yellow (10YR 7/3) to pale yellow (2.5Y 7/3), and it is composed of fine sand. The layers transition at about 30 to 35 cm bs, when both are present. The substrate, however, is not typical of Blaney loamy sand, which is composed of a sequence of dense and compact, light brown sandy clay loam Bt-horizons. Instead, the on-site substrate is composed of only weakly cemented to loose fine sandy loam. In combination with the topographic setting of the site, adjacent to Rockfish Creek, and the distinctive “island-like” appearance of the landform, the composition of the substrate suggests that the landform may represent an alluvial feature (levee?) that has remained stable for a period long enough to sustain weak soil formation. The Stage I sample provided a basis for defining a Stage II sample frame within which to deploy a 5 m-interval shovel test grid. This guaranteed that the artifact bearing locations of the site would receive systematic sampling at a relatively intensive sample unit density, which would yield a detailed picture of the occupation history of the site. Ninety-eight (98) Stage II shovel tests were excavated at 5-m intervals within the sample frame, 55 of which yielded artifacts (Figure 21). Mean shovel test depth for the Stage II sample was 64.59 cm bs (SD=13.64) with a mode of 60 cm bs and a range of 10 to 100 cm bs. The slightly deeper average of shovel tests in the Stage II sample is a function of the fewer number of tests excavated on the slopes of the levee where substrate was encountered at much shallower depths. The Stage III shovel tests sample was distributed between 19 locations anchored by Stage I or II shovel tests containing data relevant to evaluating the occupational history and integrity of the site matrix (Figure 22). One hundred fifty-four (154) Stage III shovel tests were excavated at generally 1.25-m intervals around targeted Stage I and II shovel tests. One hundred thirty-four (134) of these tests yielded artifacts. Mean shovel test depth for the Stage III sample was 65.84 cm bs (SD=11.59) with a mode of 60 cm bs and a range of 10 to 90 cm bs. Once the results of the Stage III shovel test sample were analyzed, specific locations yielding shovel test outcomes of interest to evaluating the site were selected for test unit excavation. Stage III shovel testing around targeted shovel test outcomes provided the basis for precise placement of test units. Three 1 m-x-1-m test units were excavated in total and their locations are illustrated in Figure 23. The units were terminated at 60 and 70 cm bd. ARTIFACT INVENTORY Artifacts collected during the investigation overwhelmingly consisted of chipped stone debitage and tools. The overall precontact inventory consisted of 1,574 items including 984 pieces of lithic debitage, 65 chipped stone tools, 18 cores, 1 tabular slate fragment, 2 indeterminate slate fragments, a flake adz fragment, a possible grinding stone, an anvil stone, a grooved abrader, a possible hearth deflector/rest, 3 steatite bowl fragments, 95 precontact ceramics, and 401 rocks and fire-cracked rocks. Appendix B contains the artifact database for the project, while appendices C through I provide additional data on lithic chipped stone and ground stone tools and precontact ceramics. Descriptions of the various artifact and lithic raw material types are presented below. Lithic Raw Material Types Lithic raw material identification proceeded on two different levels. At the macro-scale, individual specimens were typed in accordance with major hard rock geological terminology to achieve comparability with previous projects at Fort Bragg. At the micro-level, however, an attempt was made to further partition some of the hard rock types into “core groupings” to distinguish between specimens derived from single hard rock types that may have originated from different cores. This was done to facilitate spatial analysis in close-interval shovel test grids where overlapping of distinct deposits or complex multi-household occupations were suspected. Core grouping analysis was undertaken principally in metavolcanic types where subtle differences in patina coloration, core color, phenocryst size and density, and matrix texture was easily discerned and monitored. Descriptions of the various rock types and the associated core grouping sub-types recognized in the collection are presented below. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 83 Chapter 5. 31HK2502 620 610 70.5 71.0 600 Floodplain 590 Floodplain 580 71 .5 570 .0 72 560 550 . 540 31HK2502 530 71.5 Rock fi 70 .5 ek Cre sh 520 71 .0 71 .5 510 Floodplain 500 72 .0 71.5 490 72.5 480 73.0 470 Positive ST Negative ST 460 0 450 440 N 400 410 40 m 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 21. Site 31HK2502, Stage I and II Shovel Test Sample. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 84 Chapter 5. 31HK2502 620 610 70.5 71.0 600 Floodplain 590 Floodplain 580 71 .5 570 .0 72 6 560 9 13 550 7 12 5 19 31HK2502 8 540 530 71.5 11 17 520 4 1 72 .0 71.5 490 .0 510 71 .5 500 .5 10 71 14 2 73.0 470 Positive ST Negative ST 460 N 0 450 440 Floodplain 3 16 72.5 480 Rock fi 70 18 ek Cre sh 15 . 400 410 40 m 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 22. Site 31HK2502, Stage III Shovel Test Sample Showing Locations of Sample Blocks. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 85 Chapter 5. 31HK2502 620 610 70.5 71.0 600 Floodplain 590 Floodplain 580 71 .5 570 .0 72 560 550 . 540 31HK2502 530 71.5 Rock fi 70 .5 ek Cre sh 520 .5 71 71 .0 TU3 510 Floodplain TU2 500 72 TU1 .0 71.5 490 72.5 480 73.0 470 460 N 0 450 440 400 410 40 m 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 23. Site 31HK2502, Locations of Test Units. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 86 Chapter 5. 31HK2502 Metavolcanic Types Metavolcanic types, as a group, dominate the chipped stone collection from the site (Table 2). They represent 81.4 percent (n=869) of the total chipped stone artifact inventory (i.e. debitage and chipped stone cores and tools). The best reflection of actual representation, however, is supplied by the combined Stage I and II shovel test samples, which is unbiased in terms of spatial coverage. In this sample, metavolcanic material composes 80.4 percent of the chipped stone. In this case, the total inventory is nearly identical to the unbiased sample. Descriptions of the five identified metavolcanic types and associated core group subtypes follow. The typological system varies somewhat from earlier reports in response to the recent stone quarry sourcing study for the Carolina Slate Belt conducted by Steponaitis et al. (2006). (1) Mill Mountain Rhyolite (MMR). MMR was not clearly differentiated in the sourcing study (Steponaitis et al. 2006), but it is a readily recognizable type in the assemblages from Fort Bragg. Mill Mountain is located between the Uwharries Southern and Uwharries Eastern source areas on the opposite bank of the Yadkin River from Table Top Mountain. It was not included in either of these source areas, however, because of the presence of distinctive clear, glassy quartz phenocrysts. Daniel and Butler (1996:18) describe this type as a medium gray, aphanitic, rhyolite porphyry with sparse, glassy phenocrysts of quartz generally less than 1 mm in diameter. Very fine-grained, disseminated grains of pyrite are also present. Thin sections indicate a microcrystalline matrix composed primarily of feldspar and quartz, with some biotite, chlorite, and disseminated pyrite. The specimens assignable to this type from the 31HK2502 collection are characterized by a dark reddish gray to moderate brown coloration when not patinated. Four subtypes were assigned to the type, based primarily on the completeness of the observation field and patination. Subtype R8 corresponds to specimens that exhibit sparse glassy quartz phenocrysts only. Subtypes R8Hhf, R8Hhm and R8Hhc are fine-, medium- and coarse-grained specimens respectively that exhibit sparse glassy quartz, white subhedral quartz phenocrysts and, rarely, feldpsar phenocrysts. Although these phenocryst types were not included in the type description, the subtypes containing them were included as a variant of Mill Mountain Rhyolite due to the co-occurrence with glassy quartz phenocrysts. Mill Mountain Rhyolite makes up 5.4 percent of the combined Stage I and II chipped stone sample. (2) Type I Rhyolite Tuff (Type I RT). An exact correlate of this type was not discernible in the sourcing study (Steponaitis et al. 2006). Benson (1999:30) describes it as fine-grained dark green to gray material that looks and feels very similar to Piedmont chert. He noted that plagioclase phenocrysts were sporadically present. A single subtype was identified in the 31HK2502 collection, Subtype 2A, which is the weathered state of Subtype R2. Subtype R2 is only rarely identified in assemblages on Fort Bragg. It appears very similar to Subtype R3t, but its coloration can be described as a dusky blue green. The texture is microcrystalline and the edges are semi-translucent. Subtype R2A specimens exhibit a pale green patina. Darker green splotches and swirls on the faces of these specimens represent more resistant matrix. The matrix is aphantitic and microcrystalline in texture and is highly isotropic in fracture characteristics. The origin of this material is not well understood and it is possible that the type represents a very fine-grained metamudstone similar to some of those described for the Chatham Pittsboro source (see Moore and Irwin 2006:27-28; Stoddard 2006:57). In support of this is the occurrence of what appear to be bedding laminations in some of the patinated material. Moreover, the color of the patinated specimens is similar to that reported for the Chatham Pittsboro source. Type I Rhyolite Tuff makes up 3.3 percent of the combined Stage II and II chipped stone sample. (3) Uwharries Eastern Rhyolite (UER). Stoddard (2006:52) describes the specimens from this sourcing area as light to dark gray metadacite porphyry or crystal lithic tuff. All samples collected contained plagioclase feldspar and white quartz phenocrysts, ranging from less than 2 percent to 7 percent phenocryst density. Common metamorphic minerals within the matrices include biotite, stilpnomelane and calcite. Similar dacites and rhyolites are typical of the Uhwarries Southeastern source (Stoddard 2006:55-56) as well, but many of these specimens contain silica levels beyond the range typical of igneous rocks. Macroscopic criteria for distinguishing between lithic material from these two sources does Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 87 Chapter 5. 31HK2502 Table 2. Total Chipped Stone Inventory, 31HK2502. TOTAL CHIPPED STONE INVENTORY: 31HK2502 Utilized Flake Side Scraper Pointed Scraper End Scraper Retouched Flake Projectile Point Core Biface Debitage RAW MATERIAL TYPE Flake Adz CHIPPED STONE CLASSES GRAND TOTAL Metasedimentary Metasedimentary, gray 2 Metasedimentary, green 3 1 3 1 4 Metavolcanic Indeterminate metavolcanic 2 Mill Mountain Rhyolite 50 Type I Rhyolite Tuff 71 2 2 2 1 1 1 Uwharries Eastern Rhyolite 295 3 4 Uwharries Southern Rhyolite 257 5 2 1 2 2 2 Uwharries Western Rhyolite 126 1 1 4 1 1 1 174 1 9 1 2 56 4 78 1 3 1 7 8 284 5 1 140 1 309 Quartz Crystal Quartz White Quartz 2 1 186 Other Material Brown Chert 1 1 Orthoquartzite 1 1 Quartzite 1 1 GRAND TOTAL 984 13 18 not at present exist. Consequently, specimens from both source areas may be present in the analyzed lithic assemblages at Fort Bragg. Typical specimens from 31HK2502 are characterized by dark gray (nonpatinated) and grayish green (patinated) matrices with splotches of silicate and variable percentages (2 to 4 percent) of white quartz and plagioclase feldspar phenocrysts. Some specimens exhibit purplish to brown background matrices. Quartz phenocrysts predominate over feldspar in the examples from the site. Type I Rhyolite, which has been distinguished separately in previous reports (Cable and Cantley 2005b, 2006, 2010), is also subsumed under this type. Five subtypes were identified in the collection: R8Ihc, R8Ihf, R8Ihm, R8hf, and R8hff. The first three subtypes contained white quartz and plagioclase feldspar phenocrysts in variable percent- 1 11 3 3 1 20 13 1067 ages, while the latter two subtypes exhibited the same background matrix, but phenocrysts could not be observed. The absence of phenocrysts in these latter two subtypes is assumed to result from natural variability in the distribution of low-levels of phenocrysts within the type and problems with narrow observation fields in some specimens. Subtypes R8Ihf (fine), R8Ihm (medium) and R8Ihc (coarse) distinguish texture variability within the larger type. UER makes up 22.1 percent of the combined Stage I and II chipped stone sample. (4) Uwharries Southern Rhyolite (USR). This type includes sources found on Morrow Mountain and surrounding areas, most notably Table Top Mountain. Daniel and Butler (1996:10-13) describe this type as a dark gray, aphanitic, aphyric rhyolite, that commonly exhibits flow-banding. Weather- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 88 Chapter 5. 31HK2502 ing tends to bring out the flow lines visually, which generally alternate in a pattern of light and dark gray. Although the matrix is usually homogeneous, some specimens exhibit small spherulites of less than 1 mm in diameter (see Stoddard 2006:52). Thin sections indicate a microcrystalline intergrowth of feldspar and quartz, with minor biotite and chlorite elements. The individual minerals are difficult to distinguish and strings of dark minerals can occur along fracture planes. R3mp. All of this material has been patinated to a grayish green or greenish gray color. Non-patinated specimens may have been assigned to one of the other Uwharrie source areas in the analysis, but the distinctive glomerocyst structure suggests that this was not a major problem. Due to patination, most of the specimens in the collection were characterized by fine to medium texture. UWR makes up 23.8 percent of the combined Stage I and II chipped stone sample from 31HK2502. The specimens identified as USR in the site assemblage are composed of two larger subtype groupings, the 3m and the R3/R3t groups. The 3m Group corresponds to material that typically occurs on Morrow Mountain and includes a series of subtypes that describe variation in flow banding, texture and patination. Subtype R3m is non-patinated and it exhibits a dark gray fine-grained matrix devoid of phenocrysts. R3mc represents a coarser textured version of R3m. R3mf contains flow bands but otherwise is identical to R3m. Subtypes R3mD, R3mf, R3ms and R3msf are patinated and range in color from grayish green to greenish gray to moderate yellow or tan. R3mD and R3mDf appear to represent patinated specimens of R3m and R3mf respectively. R3ms and R3msf exhibit blotches and linear arrays of silicate containing biotite and perhaps chlorite elements. They too, appear to be patinated forms of R3m and R3mf. The R3/R3t group is composed of an extremely fine grained or microcrystalline dark gray matrix that is semi-translucent when viewed on attenuated edges. Blotches of feldspar are commonly seen on exposed surfaces, but they do not form phenocrysts. R3 and R3t are non-patinated, while R3p and R3pat are fully patinated examples. Both are tan in color, but R3p also exhibits finely disseminated biotite specks. Finally, R3tf constitutes a flow-banded version of R3t. USR makes up 25.8 percent of the Stage I chipped stone sample. (6) Indeterminate Metavolcanic. A small amount of the metavolcanic material identified in the analysis could not be further classified to type due to excessive hydration. These were simply classified as “indeterminate metavolcanic.” Only two pieces of debitage were placed in this category and neither of these was recovered from the combined Stage I and II sample. (5) Uwharries Western Rhyolite (UWR). This source area includes specimens from Wolf Den Mountain and Falls Dam (Stoddard 2006:52). This material is described as gray to black microcrystalline felsic volcanic rock that contains rounded plagioclase phenocrysts and glomerocrysts. Green biotite and pale green amphibole are dispersed locally in the matrix and spherulites are sometimes present. A single subtype was identified in the collection, Quartz Types Pure quartz forms as veins in igneous and metamorphic rock formations characterized by slow crystallization. This generally produces anhedral, milky structures (Spock 1962). In some cases, however, crystallization occurs earlier in the sequence, producing clear crystals with euhedral structure. Both vein and euhedral crystal quartz were utilized by precontact groups in the project area, but most of it appears to have come from river gravel deposits. A great deal of variation was evident in the vein quartz category ranging from mixed clear and cloudy exposures to opaque milky white. Since these characteristics appeared to be repeatedly present on single cores when debitage concentrations were analyzed, all of this variation was subsumed under the single white quartz subtype Q5. The category “crystal quartz” (CQ) was retained for instances where glasslike examples and crystal facets strongly suggested a euhedral crystal origin. White quartz represented 18.3 percent of the combined Stage I and II chipped stone sample, while only one piece of crystal quartz (0.4 percent) was recovered in the sample. In addition, 8 pieces of fire-cracked rock derived from quartz cobble fragments were recovered from the entire Phase II collection at the site. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 89 Chapter 5. 31HK2502 Minority Types Several other rock types were recognized in the 31HK2502 collection. Descriptions of each are presented below. (1) Arenite. This material has been previously identified as ferruginous sandstone, but thinsectioning of a piece of this material indicates that a more accurate characterization is arenite, or hematite cemented quartz sandstone (Appendix K). Three pieces of arenite were classified as fire-cracked rock. It occurs in a variety of shapes including flattened and irregular nodules. The exterior of the nodules consist of fine-grained ruddy brown precipitate while the centers consist of dense, reddish brown to purple, densley cemented sandstone. The material was commonly used for rock hearths by the precontact inhabitants of Fort Bragg. Two hundred and sixty-three pieces of arenite were collected during the investigation, totaling nearly 5.35 kg. In addition to serving as a hearth rock constituent, the material also appears to have been used for grinding activities. One item exhibits apparent facets on a flat face, suggested that it was used as a grinding stone (271265a66). Another large, thin piece of arenite (271265a812) may have been used as a hearth deflector or hearth rest rock. One other piece of arenite appears to have been removed from a larger block of material through percussive blows. (2) Brown Chert. A single, small piece of debitage (271265m679) was made of a medium to grayish brown, opaque chert with high luster. Its lustrous appearance may have been the result of heattreatment. Numerous chert sources have been identified in the Durham-Wadesboro Basin in the eastern Piedmont (Lautzenheizer 1996), but most of the identified types are dark gray in coloration, although a porous brown type has also been recognized. The dark coloration of the specimen, however, would generally suggest a Piedmont source. (3) Indurated Granite. A single specimen of fine-grained granite referred to here as “indurated granite” was recovered at 31HK2502. It was fire-cracked and had been broken into three pieces together weighing 293 g. The sample was thin-sectioned and identified as a fine-grained granite composed of subhedral plagioclase feldspar, subhedral quartz magnetite, sub- hedral muscovite and subhedral zircon (Appendix K). Some quartz occurred “as intergrowths within alkali feldspar crystals (this texture is referred to as myrmeckitic texture and is often observed in rocks that have been altered by fluids during metamorphism).” Outcrops of this material are unknown in the region, but Seramur (personal communication, August, 2011) notes that its distinctive structure would make determining its precise source probable. This material was extensively used as temper in portions of the Hanover III and Yadkin III series precontact ceramics recovered from the DO5 package sites. The outer surface of the material was rounded, suggesting that it was transported from its source by errosional and/or alluvial agents. It is possible that this material occurs as a minor element in stream gravels, perhaps deriving from Triassic Basin sediments. On the other hand, it may have been intentionally imported by potters to use as tempering material for pottery manufacture. If the latter is true, it would provide another line of evidence to suggest that pottery manufacture was undertaken locally. Site 31RH491 on Drowning Creek, to be discussed later, also produced small rock fragments of this material in precontact contexts. The specimen (m127) from 31HK2502 was found at a depth of 20 to 30 cm bs in a shovel test, in what also appears to be a precontact context. (4) Metasedimentary. Two types of fine to medium grained, homogeneous metasedimentary material were recovered during the investigation. The types are distinguished solely by color. One type was greenish gray (n=4), while the other was light gray (n=4). Although the material occurs only infrequently, it was apparently used to manufacture a wide range of tool types at the site. A Morrow Mountain II Stemmed base (271265a344), a Type III Side Scraper (271265a1166) and a fragment of a grooved abrader (271265a325) were all made from this material. The remaining items consisted of debitage. In spite of its low incidence, two of the flakes were recovered in the combined Stage I and II sample, possibly indicating a 0.8 percent representation at the site. The coarseness of the grain of both types suggests that they can be further classified as metasiltstone or fine metasandstone. Numerous source areas in the Durham-Wadesboro Basin contain similar rock types, including the Chatham Pittsboro, Person County and Chatham Siler City (Stoddard 2006:57-63). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 90 Chapter 5. 31HK2502 (5) Orthoquartzite. A single piece of debitage was made from a light tan, highly micaceous gritty material resembling orthoquartzite from the Santee River Valley of South Carolina (Anderson 1982:120124). That material derives from the Thanetian Black Mingo Formation and it is typically described as chert and chalcedony cemented quartz arenite or chert and chalcedony cemented sandstone. Large cobbles and chunks of lag gravel are found in terrace exposures along the Lower Santee and Black River valleys. (6) Petrified Wood. A single piece of petrified wood was recovered from the investigation. It was unmodified, but it was probably included in a rock cluster hearth along with associated sandstone conglomerate and quartzite. Petrified wood occurs in the sedimentary formations of the upper coastal plain in the general vicinity of Fort Bragg. (7) Sandstone Conglomerate. Also abundantly present in the deposit were chunks of sandstone conglomerate containing a wide range of sand and pebble sizes. Thin-sectioning of a piece of this material determined that individual grains of sand were cemented with clays (wacke) and exhibited udulatory extinction indicating that it originated from metamorphic rock (Appendix K). Occasionally interfaces with arenite could be observed on the chunks, indicating that the two materials were derived from the same formations (e.g. Middendorf). The matrix of the conglomerate consistently exhibited a grayish orange to dusky yellow coloration. Chunks of arenite were sometimes viewed in the matrix as well. This material appears to have been used to make rock hearths. One hundred twenty-eight pieces of sandstone conglomerate were collected, summing to about 3.8 kg. (8) Slate. Three pieces of a dark to medium grayish green, laminated material identified as slate were recovered during the investigation. Two of the items (271265a212) consisted of indeterminate fragments, while the other (271265a195) was thin, tabular fragment possibly representing a fragment of a pendant. Evidence of human modification, however, could not be discerned on the item. (9) Steatite. Three sherds (271265a902) from a single steatite bowl were recovered in a shovel test during the investigation. The talc-schist mate- rial was medium light gray in color. Steatite vessel fragments are sporadically found on Fort Bragg (see McNutt and Gray 2007; Millis et al. 2010; Ruggiero 2003). The material is commonly found at locations in the Piedmont, although the nearest known quarries are in South Carolina and south central Virginia (see Millis et al. 2010:460-461; Truncer 2004). (10) Grano-diorite. An anvil stone fashioned from a flat river cobble composed of a light colored igneous rock was recovered from a shovel test during the investigation. The igneous material is tentatively identified as grano-diorite. Grano-diorite is gradational between granite and diorite and is formed when plagioclase replaces orthoclase in sufficient proportions during magma cooling (Spock 1962:58). Lithic Artifacts Basic definitions of the various lithic artifact classes were presented in Chapter 4. Here, specific descriptions of the 31HK2502 lithic collection are presented. Debitage The debitage classification is adapted from formal (Bradley 1973; Frison and Bradley 1980; House and Wogaman 1978; Newcomer 1971), attribute (Moore 2002; Shott 1994), and mass (Ahler 1989; Moore 2002) analysis approaches. Debitage classes were devised to identify both reduction stages and reduction/production systems (ie. biface core reduction, directional core reduction, and flake blank production). This classification was supplemented with the recording of three attributes: (1) size class, (2) percent cortex, and (3) condition. Table 3 presents the complete inventory (all stages) of debitage type by raw material type from 31HK2502. Biface reduction types (FBRs and BTFs) comprise the majority of identifiable debitage, but directional core reduction as represented by core flakes also registers a significant presence in the collection. Core Flakes comprise 19 to 34 percent of the identifiable debitage types (i.e. core flakes, FBRs and BTFs) within the metavolcanic and quartz samples. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 91 Chapter 5. 31HK2502 The distribution of cortex on dorsal faces and platforms of debitage provides a basis for further evaluating the nature of reduction strategies associated with the various raw material types at 31HK2502 (Table 4). Dibble et al. (2005:550) indicate that raw nodule reduction from bifaces should yield between about 6 and 21 percent flakes with cortex for any single episode. It is likely that similar proportions characterize directional core reduction from raw nod- ules. Cortex percentages across the main raw material types, with one exception, ranges between 3.0 and 6.6 percent, suggesting that predominant on-site reduction activities started with partially prepared cores. The exception, Type I Rhyolite, is characterized by a cortex incidence of 26.7 percent, strongly arguing that cores of this raw material type consisted of unmodified nodules. Table 3. Distribution of Debitage Type by Raw Material Type, 31HK2502. DISTRIBUTION OF DEBITAGE TYPE BY RAW MATERIAL TYPE: 31HK2502 DEBITAGE TYPES RAW MATERIAL TYPE BTF Blade Flake Chunk Core Flake Cort. Chunk Flake Frag. FBR GRAND TOTAL Metasedimentary Metasedimentary, gray 2 Metasedimentary, green 1 1 8 2 1 3 17 16 50 Metavolcanic Indet. Metavolcanic 2 Mill Mountain Rhyolite 9 Type I Rhyolite Tuff 13 7 33 16 71 Uwharries Eastern Rhyolite 29 40 145 81 295 Uwharries Southern Rhyolite 41 37 125 53 257 Uwharries Western Rhyolite 25 19 51 31 126 12 122 13 174 1 1 1 2 2 Quartz Crystal Quartz 1 White Quartz 10 17 1 Other Raw Materials Brown Chert Orthoquartzite 1 Quartzite 1 1 GRAND TOTAL 127 1 18 1 124 2 500 212 984 Table 4. Percent Cortex on Debitage by Raw Material Type, 31HK2502. PERCENT CORTEX ON DEBITAGE BY RAW MATERIAL TYPE: 31HK2502 RAW MATERIAL TYPE TOTAL DEBITAGE PERCENT W/CORTEX Mill Mountain Rhyolite 50 4.0 Type I Rhyolite Tuff 71 26.7 Uwharries Eastern Rhyolite 295 4.4 Uwharries Southern Rhyolite 257 6.6 Uwharries Western Rhyolite 126 4.8 White Quartz 174 3.0 TOTAL 973 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 92 Chapter 5. 31HK2502 Cores Metric and attribute data on cores are provided in Appendix C. Three classes of cores were identified in the 31HK2502 collection: (1) Directional Cores, (2) Flake Blanks and (3) Core Fragments. Each class is discussed below (Table 5). (1) Directional Cores. Directional cores consist of relatively large masses of material that exhibit flake scars running in one or more directions from flat striking platforms rather than from biface edges. The platform angle approximates 90°. Two broken directional cores were identified in the collection, both of which exhibited flake scars removed in a single direction (unidirectional) from platforms devoid of flake scars (Figure 24:A-B). Specimen 271265a265 was made of white quartz, while Specimen 271265a455 was composed of Uhwarries Western Rhyolite. (2) Flake Blanks. Flake blanks were removed from a core for the purpose of further modification or use as a tool (Bradley 1973:6). However, they do not exhibit obvious macroscopic evidence of use or modification. Six flakes fit the physical requirements of this definition. All of the specimens were derived from directional cores and were classified as core flakes or blade flakes. All of the recognized flake blanks in the collection were composed of rhyolite. (3) Core Fragments. Angular pieces of raw material containing portions of one or more flake scars suggestive of directional core reduction, but lacking striking platforms were classified as core fragments. Ten such fragments were recognized in the collection, all but two of which consisted of white quartz. The exceptions were composed of Uhwarries Eastern and Mill Mountain Rhyolite, respectively. Bifaces Thirteen biface fragments were recovered during the investigation (Table 6). Bifaces are derived from both large mass packages (i.e. unmodified and/ or prepared cores) and flake blanks. Determination that a biface was derived from a flake blank was made when flake characteristics were observed on the item, such as remnant striking platforms, flake curvature, or bulbs of percussion from flake detachment. The Table 5. Summary Data for Cores, 31HK2502. CORES:31HK2502 SPECIMEN # BAG CORE TYPE PORTION FLAKE TYPE LITHIC RAW MATERIAL TYPE MAXIMUM THICKNESS (MM) MAXIMUM LENGTH (MM) MAXIMUM WIDTH (MM) WEIGHT (GM) 5.52 271265a36 16 Core Fragment Fragment Indeterminate White Quartz 11.03 21.63 20.55 271265a194 105 Core Fragment Fragment Indeterminate White Quartz 13.14 16.63 15.93 4.57 271265a239 134 Core Fragment Fragment Indeterminate White Quartz 19.48 33.98 19.70 14.13 271265a333 208 Core Fragment Fragment Indeterminate White Quartz 24.17 20.74 35.63 12.99 271265a811 434 Core Fragment Fragment Indeterminate White Quartz 17.11 38.75 26.50 15.28 271265a876 451 Core Fragment Fragment Indeterminate White Quartz 27.61 21.44 34.29 31.32 271265a961 496 Core Fragment Fragment Indeterminate White Quartz 9.88 25.90 17.55 5.62 271265a1013 524 Core Fragment Fragment Indeterminate Uwharries Eastern Rhyolite 9.94 6.87 19.81 1.75 271265a1119 553 Core Fragment Fragment Indeterminate White Quartz 17.97 12.06 32.03 5.34 271265a1135 558 Core Fragment Fragment Indeterminate Mill Mountain Rhyolite 13.22 12.53 19.74 4.06 271265a265 148 Unidirectional Core Fragment Fragment Indeterminate White Quartz 24.71 30.32+ 48.61 34.81 271265a455 266 Unidirectional Core Fragment Fragment Indeterminate Uwharries Western Rhyolite 28.81 34.33+ 54.02 47.96 271265a201 109 Flake Blank Whole Core Flake Mill Mountain Rhyolite 8.83 49.63 37.21 15.39 271265a565 326 Flake Blank Whole Core Flake Uwharries Eastern Rhyolite 4.77 42.48 29.98 6.15 271265a814 435 Flake Blank Whole Core Flake Uwharries Eastern Rhyolite 4.35 32.93 30.89 4.73 271265a977 507 Flake Blank Whole Core Flake Uwharries Eastern Rhyolite 4.32 24.90 36.79 2.84 271265a983 511 Flake Blank Whole Core Flake Uwharries Eastern Rhyolite 3.96 21.70 27.70 2.83 271265a1023 529 Flake Blank Whole Blade Flake Uwharries Southern Rhyolite 3.12 35.17 20.65 2.11 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 93 Chapter 5. 31HK2502 Figure 24. Cores and Bifaces, Site 31HK2502. A-B: Unidirectional Cores (a265 and a455 respectively), C-D: Type I Bifaces (a14 and a1001), E-F: Type III Bifaces (a717 and a268 respectively; a717 represents a Palmer II-III Preform). bifaces in the collection, however, do not have sufficient mass to suggest that they were derived from large biface cores. All appear to have been produced from flake blanks. Each biface category as adapted from Daniel (2002:51-54) is described below. (1) Type I Bifaces. Daniel (2002:51) describes Type I Bifaces as irregularly shaped flake blanks and core masses with large conchoidal scars forming at least one section of a sinuous edge along the core edge. No evidence of secondary thinning scars is observable and a relatively high thickness to width ratio obtains. Two Type I Bifaces were recognized in the collection. One specimen (271265a14) was only minimally flaked and characteristics of the original flake blank (core flake), including the striking platform, were observable (Figure 24:c). The other specimen (271265a1001) was a lateral section from a crystal quartz biface (Figure 24:d). (2) Type II Bifaces. These bifaces, in the main, represent intermediate stage preform rejects. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 94 271265a84 271265a530 271265a984 271265a1035 271265a1069 271265a912 271265a87 271265a268 271265a652 309 511 532 545 472 45 150 362 271265a1001 43 520 271265a14 271265a717 399 6 271265a140 SPECIMEN # 73 BAG Type III Biface Fragment Type III Biface Fragment Type III Biface Fragment Type III Biface Fragment Type II Biface Fragment Type II Biface Fragment Type II Biface Fragment Type II Biface Fragment Type II Biface Fragment Type I Biface Fragment Type I Biface Preform Biface Fragment TYPE Lateral Section Upper Blade Upper Blade Base Tip End Lateral Section End End Lateral Section Whole Whole End PORTION Core Flake BLANK TYPE Ovate Ovate Flat Base Ovate Ovoid/Ovate Ovoid Ovate SHAPE Palmer ASSOCIATION Bifaces: 31HK2502 Uwharries Eastern Rhyolite Mill Mountain Rhyolite Uwharries Southern Rhyolite Uwharries Southern Rhyolite Uwharries Eastern Rhyolite White Quartz Uwharries Southern Rhyolite Uwharries Southern Rhyolite Mill Mountain Rhyolite Crystal Quartz Uwharries Southern Rhyolite Uwharries Eastern Rhyolite Uwharries Western Rhyolite LITHIC RAW MATERIAL TYPE Table 6. Summary Data for Bifaces, 31HK2502. 4.47 7.78 7.36 9.84 7.32 8.39 7.36 6.46 8.10 7.53 4.31 18.23 7.12 Maximum Thickness (mm) 17.68 21.56 27.14 14.27 16.95 15.96 15.16 30.67 29.10 39.70 34.78 42.10 11.79 Maximum Length (mm) 9.48 24.97 26.43 27.03 17.24 27.97 14.72 13.65 12.57 15.99 24.73 33.02 15.60 Maximum Width (mm) 0.51 3.69 4.86 3.54 1.57 4.73 1.71 2.72 2.64 5.61 5.94 12.40 1.06 Weight (gm) Chapter 5. 31HK2502 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 95 Chapter 5. 31HK2502 They are roughly flaked and exhibit only large, conchoidal flake scars. No edge trimming is evident, but an ovate to lanceolate outline is generally recognizable and the perimeter of the edge margin is entirely formed. Facial retouch, however, may not be complete and features of original flake blanks are sometimes evident. Five Type II Biface fragments were identified in the collection (Table 6). (3) Type III Bifaces. Type III Bifaces represent late stage preform rejects. The edges exhibit secondary thinning and shaping scars and facial retouch is globally distributed. Outline shapes are well defined and reflect ovate and lanceolate forms. Most of these specimens exhibit step and hinge fractures that terminated production. Type III Bifaces tend to exhibit the lowest ratios of thickness-to-width. Sometimes features of diagnostic projectile point haft elements are present, allowing correlation with specific culture-historic types. In such cases, specimens are differentiated and identified as “Preforms.” One preform representing a blank for a Palmer Corner Notched projectile point was recognized in the collection (Figure 24:e). Small corner notches were removed from an ovate blade with a straight base, approximating the outline of a Palmer II/III Corner Notched outline. The preform was discarded when the upper part of the blade was accidentally removed during thinning. The preform was recovered from a Stage III shovel test (N566/E439.2) at a depth of 40 to 50 cm bs, a vertical position consistent with the inference that the item represents a Palmer II/III blank. Four other Type III Bifaces were identified in the collection (Table 6, Figure 24:f). Projectile Points Projectile points are bifacially flaked tools with retouched haft elements. In addition, projectile point fragments consist of specimens with the fine bifacial retouch characteristic of finished projectile points, but not necessarily exhibiting evidence of a haft element. The collection from 31HK2502 consisted of 6 diagnostic projectile points and 5 fragments (Table 7). Metric and descriptive data on the projectile points can be found in Appendix E. (1) Taylor Side Notched. A single, whole specimen (a500) recovered from the 31HK2502 investigation is identified as a Taylor Side Notched point (Figure 25:A). Although this type is more frequently found in South Carolina, its presence in North Carolina is gaining increasing recognition. Michie (1966:123-124; 1996:253-254) defined the type as an outcome of his excavation at the Taylor Site in central South Carolina. Morphological characteristics of the type include basal grinding, squared ears, deep side notches and alternate blade beveling. Taylor points have not been recovered in stratigraphic contexts, nor are there associated radiocarbon dates available. However, it is generally accepted that the morphological characteristics of the type place it in the very early range of the Early Archaic period, in a time horizon correlated with Bolen (Bullen 1975) and Big Sandy (DeJarnette et al. 1962:48; Lewis and Kneberg 1961:38) points. Michie suggests a chronological position of around 10,000 BP for the Taylor Side Notched point. The morphological characteristics of the point type place it between the Hardaway Side Notched and Small Dalton cluster on the early end and Palmer Corner Notched on the late end. Daniel (1998:53) recognized one specimen from the Hardaway site that resembles the type. The blade of the specimen was not beveled, as is common for the South Carolina sample. The specimen from 31HK2502 exhibited squared ears, small U-shaped side notches, a deeply concave, ground base and a non-serrated blade margin without alternate beveling. In these characteristics, it is almost identical to the Taylor Side Notched point illustrated by Daniel (1998:54, Figure 4.3:A). Two of the three Taylor Side Notched points recovered from the Taylor Site did not have serrations and alternate beveling did not occur on at least one specimen. The blade face has been carefully pressure flaked in a “chevron” pattern and none of the flake scars extends across the medial axis. The specimen was recovered at a depth of 40 to 50 cm bs in a Stage III shovel test excavated at grid point (N508.75/E487.5), a vertical position consistent with an Early Archaic age. (2) Palmer II Corner Notched. A single specimen (a502) was typed as a Palmer II Corner Notched point (Figure 25:B). Coe (1964:67) described the Palmer Corner Notched type as a “small cornernotched blade with a straight, ground base and pronounced serrations.” Blade edges were generally straight, but were secondarily both incurved and excurvate. Bases were commonly straight and typi- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 96 BAG 360 25 48 333 232 235 11 214 287 324 286 SPECIMEN No. 271265a646 271265a54 271265a92 271265a586 271265a386 271265a395 271265a25 271265a344 271265a502 271265a560 271265a500 Whole Blade Palmer II/III Corner Notched Taylor Side Notched Whole Base Base Uwharries Southern Rhyolite Uwharries Eastern Rhyolite Uwharries Western Rhyolite Metasedimentary, gray Uwharries Western Rhyolite Mill Mountain Rhyolite Uwharries Western Rhyolite Whole White Quartz Upper Blade Midsection Uwharries Eastern Rhyolite Uwharries Southern Rhyolite Uwharries Western Rhyolite LITHIC RAW MATERIAL TYPE Tip Tip Tip Midsection Lower Blade PORTION Palmer II Corner Notched Morrow Mt II Stemmed Morrow Mt II Stemmed Guilford Lanceolate, St. Base Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment TYPE WEIGHT (gm) 6.99 7.08 7.32 2.72 4.81 24.95 9.02 1.11 0.53 0.53 2.88 MAXIMUM THICKNESS (mm) 7.5 5.32 7.71 6.86 6.43 12.49 (7.61) (5.38) (3.44) (4.14) (6.57) 43.78 (36.44) 51.36 (21.94) (20.43) 86.17 (41.45) (17.81) (20.11) (16.47) (20.02) MAXIMUM LENGTH (mm) BASAL WIDTH (mm) 19.26 22.35 14.82 PROJECTILE POINTS:31HK2502 TANG WIDTH (mm) 17.77 22.52 17.46 15.09 24.01 36.6 24.84 23.63 SHOULDER WIDTH (mm) Table 7. Summary Data for Projectile Points, 31HK2502. TANG LENGTH (mm) 10.24 11.42 10.39 BLADE LENGTH (mm) 33.54 36.46 39.94 BASAL GRINDING Present Present BLADE SERRATION Present EDGE ANGLE 46° 64° 69° 58° 60° 53° 39° 56° 50° 51° 67° Deeply Concave Slightly Convex Rounded Rounded BASE SHAPE Chapter 5. 31HK2502 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 97 Chapter 5. 31HK2502 Figure 25. Projectile Points, Site 31HK2502. A: Taylor Side Notched (a500)), B: Palmer II Corner Notched (a502), C: Palmer II/III Corner Notched (a560), D-E: Morrow Mountain II Stemmed (a25 and a344 respectively), F: Guilford Lanceolate, Straight Base (a395). cally ground. Serrations were at times deep, averaging about 3 mm in width and 5 to 7 mm in length. The width of the shoulder barbs usually exceeded that of the base. Metric dimensions ranged between 28 and 60 mm in length, 15 and 25 mm in width, and 5 to 12 mm in thickness. At the Hardaway Site, Palmer Corner-Notched points were primarily associated with Level III, sandwiched between Hardaway SideNotched and Hardaway-Dalton styles in Level 4 and Kirk Serrated and Kirk Corner-Notched in Level II. As an outcome of the Haw River excavations, Cable (1982) further divided the Palmer type into three variants, Palmer I, II and III. The Palmer I style is smaller, typically exhibits ground bases and was found in a slightly lower vertical position at Haw River than the larger variants described as Palmer II and Palmer III. The latter is the largest of the three variants and is characterized by a low incidence of basal grinding Palmer II is stylistically intermediate and it may overlap the temporal distributions of the Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 98 Chapter 5. 31HK2502 other variants. Others suggest that all of this variation is largely a consequence of life-stage variation rather than temporal differences (Daniel 1998; Sassaman 2002). The time range of this style is generally accepted at 9500–8900 BP (Sassaman et al. 2002:10). The specimen from 31HK2502 was made of Uhwarries Southern Rhyolite. The base was heavily ground and it exhibited a slightly convex shape. The blade took on a “Christmas tree” appearance from episodes of resharpening and the margins were highly serrated but not beveled. The tang was wide and relatively long and by virtue of its intermediate size and ground base, it was typed as a Palmer II Corner Notched point. The point was recovered from a depth of 50 to 60 cm bs in the same Stage III shovel test (N508.75/E487.5) yielding the Taylor Side Notched point. (3) Palmer II/III Corner Notched. The blade of a corner-notched projectile point (Figure 25:C) was typed as a Palmer II/III Corner Notched point by virtue of its large size. The tang width was extremely wide, which is a characteristic of the larger cornernotched points in the Palmer continuum. In the absence of a base element, it cannot be determined if the point represents a Palmer II (ground base) or Palmer III (non-ground base) style. The flaking pattern of the blade is relatively random and resharpening retouch is confined to not more than 7 mm from the edges. The blade margins are straight and the blade face approaches the shape of an isosceles triangle. It was fashioned from porphyritic Uwharries Eastern Rhyolite. The point fragment was recovered from a Stage III shovel test at grid coordinates N503.75/493.75, from a depth of only 10 to 20 cm bs. (4) Morrow Mountain II Stemmed. Two specimens of this style were recovered from 31HK2502 (Figure 25:D-E). The Morrow Mountain II Stemmed type was originally defined by Coe (1964:37-43) from his excavations at the Doershuk Site and it is similar in morphology to the Morrow Mountain I Stemmed point. Both styles were characterized by contracting stems, but two differences were cited. The blade is proportionately longer and narrower and a noticeable angle occurs at the shoulder/stem juncture in the Morrow Mountain II Stemmed type. There is some suggestive evidence that Morrow Mountain II occurred at a slightly higher stratigraphic/vertical position than Type I in the Doershuk Site excavations, which led Coe (1964:43) to tentatively posit that the former was “transitional and survived to a later date.” This type is similar in length to the Morrow Mountain I Stemmed type (30–80 mm), but is noticeably narrower at the shoulder (18–30 mm). The specimens from 31HK2502 represent fragmentary bases. Specimen a25 (Figure 25:D) was fashioned from Uwharries Western Rhyolite and it was found at a depth of 20 to 30 cm bs in a Stage I shovel test excavated at N500/E501. The other specimen (a344) was made of a metasedimentary siltstone and it was recovered at a depth of 30 to 40 cm bs in a Stage III shovel test placed at N501.25/E499. The proximity of the points suggests that they may have been deposited in a single, Middle Archaic occupation episode. (4) Guilford Lanceolate, Round Base Variant. A single Guilford Lanceolate, Round Base point (a395) was recovered from 31HK2502 (Figure 25:F). Coe (1964:43) described the Guilford Lanceolate type as a “long, slender, but thick blade with straight, rounded, or concave base.” These are large points that ranged between 50 and 120 mm in Coe’s sample, while widths ranged between 20 and 35 mm. Guilford Lanceolate points were found sandwiched between the earlier Morrow Mountain Stemmed and the later Savannah River Stemmed types in Zone VI at the Doerschuk Site (see Coe 1964:34). An important factor of variation noted in Coe’s (1964:43) broadly defined Guilford Lanceolate type is the morphology of the basal contour. Concave, straight, and rounded bases are noted to exist in the Doershuk Site assemblage and it is clear from the overall physical resemblance and close stratigraphic associations that these variants are relatively contemporaneous. However, this particular variation may evidence a very fine-grained chronological sequence and, consequently, it is advisable to draw typological distinctions between the various basal contours (Cable and Cantley 2005). Adoption of this terminology will facilitate communication and preserve potentially important chronological variation in the Guilford Complex. The specimen from 31HK2502 is made from coarse-grained, porphyritic Mill Mountain Rhyolite. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 99 Chapter 5. 31HK2502 It was crudely shaped and the base was only haphazardly shaped to form a rounded contour. The width of the blade begins to contract at what appears to be the haft juncture, about 26 mm above the base. The edges of the blade are crudely shaped with large conchoidal flake scars originating at 90 degrees from the margins. It was found in a Stage III shovel test at a depth of 30 to 40 cm bs. The shovel test (N500/ E498.75) is part of Sample Block 3, which also produced the Morrow Mountain II Stemmed points discussed above. Unifaces Unifaces consist of tools that are retouched along flake margins in a single direction (unifacial edge shaping). Retouch was applied to shape a working edge or bit and, sometimes, a butt element for hafting. Limited bifacial retouch or resharpening is sometimes present along haft elements or working edges. Metric and attribute data for unifaces are presented in Appendix F. Twenty-eight items classified as unifaces were recognized in the collection (Table 8). These include 3 end scrapers, a pointed scraper, 20 side scrapers, a denticulate, a graver, a spoke shave and what might be the butt end of a hafted flake adz. The specimens are described below under the rubric of the typological system developed by Coe (1964) and elaborated by Daniel (1998). (1) End Scrapers. Three end scraper specimens were identified in the 31HK2502 collection. Daniel (1998:66) defines end scrapers as unifaces with “steep, regular unifacial retouch that forms a relatively narrow convex bit or working edge.” He further notes that the bit edge is commonly positioned in a transverse orientation to the long axis of the tool. Four main types and a series of subtypes have been recognized; including both hafted and unhafted forms. (a) Type II End Scrapers. A single Type IIb End Scraper (a143) was identified during the investigation. Coe (1964:76) defined the Type II End Scraper from his excavations at the Hardaway Site. The type consisted of flakes of variable shapes and sizes that were retouched only along the narrowest margins to form scraping edges. He recognized two varieties that Daniel (1998:72) later formalized. Type IIa End Scrapers consisted of large, thick irreg- ular flakes that were casually shaped at one end to produce a scraper edge. Cutting edges were always irregular in marginal outline and sharp, suggesting to Coe that they were used to work hard material such as wood or bone. They were primarily confined to Level II at the Hardaway Site, which was primarily associated with Kirk Stemmed and Serrated and Stanly Stemmed points. Type IIb End Scrapers were made on thin, narrow, prismatic flakes. The scraping edge was formed along the end opposite the striking platform. This variety was not numerous, but it had the same stratigraphic association as the Type IIa variety. Daniel (1998:72-73) expanded the type definitions to also include specimens with lateral margin retouch. The specimen from 31HK2502 was made of a fine-grained Uwharries Eastern Rhyolite (Figure 26:A). The blank on which the tool was manufactured was a core flake with a thick platform that was subsequently subjected to steep marginal retouch to form a bit edge. The use of the platform to make a bit edge was not recognized in the Hardaway collection, but the specimen is included here as an example of a Type IIb End Scraper because all of its other morphological characteristics conform to the type description. A similar use of the flake platform to create a bit on a Type IIb End Scraper has been documented at 31HT844 within the confines of Fort Bragg (Cable and Cantley 2006:311). Minimal retouch can be observed on the butt end of the tool and along the right lateral margin; probably to accommodate hafting. The former edge is naturally concave in outline and its juncture with the bit forms a graver spur. The bit is slightly convex in contour and the edge exhibits extensive dulling from use. Minimal bifacial resharpening scars were observed on the ventral surface of flake along the bit edge. The specimen was recovered from a Stage II shovel test (N535/E470) at a depth of 20 to 30 cm bd. (b) Type III End Scrapers. A Type III End Scraper (a206) made of Type I Rhyolite Tuff was recovered in a Stage II shovel test (N505/E505) at a depth of 40 to 50 cm bs (Figure 26:B). Coe (1964:76) defined this type as a rough and larger version of the finely made Type I End Scraper, or “tear-drop” end scraper. Daniel (1998:75) notes that Type III End Scrapers at Hardaway exhibit the widest median bit value (42.4 mm) of all end scrapers, suggesting to him that this tool needed more lateral and dorsal retouch Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 100 18 149 302 461 495 536 550 563 567 271265a43 271265a266 271265a520 271265a888 271265a960 271265a1042 271265a1101 271265a1154 271265a1171 271265a56 271265a1174 568 384 27 271265a809 520 433 271265a801 271265a694 428 271265a1166 271265a1004 71 566 271265a132 306 64 271265a118 271265a527 373 271265a673 563 445 271265a854 251 320 271265a552 271265a1155 175 271265a316 271265a423 114 271265a206 30 75 271265a143 271265a63 BAG SPECIMEN NO. Flake Adz Spoke Shave Graver/Burin Denticulate Type III Side Scraper Type III Side Scraper Type III Side Scraper Type III Side Scraper Type II c Side Scraper Type II b Side Scraper Type II b Side Scraper Type II b Side Scraper Type II a Side Scraper Type II a Side Scraper Type II a Side Scraper Type II a Side Scraper Type II a Side Scraper Type II a Side Scraper Type II a Side Scraper Type II a Side Scraper Type II a Side Scraper Type I Side Scraper Side Scraper Fragment Side Scraper Fragment Type I b Pointed Scraper Type IV End Scraper Type III End Scraper Type II b End Scraper UNIFACE/RETOUCHED FLAKE TYPE PORTION Butt End Fragment Whole Whole Fragment Whole Whole Core Flake Indeterminate Indeterminate Core Flake Blade Flake Core Flake Blade Flake Blade Flake Indeterminate Lateral Section Whole Indeterminate Indeterminate Core Flake Indeterminate Indeterminate Indeterminate Core Flake Indeterminate Indeterminate Core Flake FBR Core Flake Core Flake Indeterminate Core Flake Trapezoidal Lenticular Triangular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Trapezoidal Triangular Lenticular Trapezoidal Lenticular Trapezoidal Indeterminate Trapezoidal Stage II Biface Fragment Trapezoidal CROSS-SECTION Indeterminate Core Flake BLANK TYPE Fragment Fragment Whole Fragment Fragment Fragment Fragment Fragment Fragment Whole Whole Fragment Whole Fragment Fragment Fragment Fragment Bit Whole Uwharries Eastern Rhyolite Uwharries Southern Rhyolite Uwharries Southern Rhyolite Uwharries Western Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite Metasedimentary, green Uwharries Eastern Rhyolite Uwharries Western Rhyolite Uwharries Southern Rhyolite Uwharries Western Rhyolite Type I Rhyolite Tuff Uwharries Southern Rhyolite Uwharries Eastern Rhyolite Type I Rhyolite Tuff Uwharries Western Rhyolite Uwharries Southern Rhyolite Uwharries Southern Rhyolite White Quartz Uwharries Western Rhyolite Uwharries Southern Rhyolite Mill Mountain Rhyolite Uwharries Western Rhyolite Uwharries Southern Rhyolite Uwharries Western Rhyolite Uwharries Southern Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite LITHIC RAW MATERIAL TYPE RETOUCH LENGTH (mm) 9.72 11.6 35.86 (5.17) 20.84 25.33 100.67 (18.60) (13.51) (27.83) 18.84 (4.03) 34.71 3.14 (22.38) (48.83) (13.16) 17.02 33.85 (18.70) 21.56 (10.29) (32.56) (74.27) (22.84) (42.53) 62.19 MINIMUM EDGE ANGLE 62° 36° 55° 46° 48° 66° 48° 46° 57° 54° 44° 58° 56° 57° 57° 53° 62° 60° 56° 58° 77° 46° 54° 34° 66° 48° 68° 52° 72° 56° 49° 57° 55° 75° 62° 63° 58° 66° 88° 56° 59° 63° 69° 58° 74° MAXIMUM EDGE ANGLE SCRAPERS AND RETOUCHED FLAKES: 31HK2502 NUMBER OF RETOUCHED EDGES 1 1 2 1 1 1 2 1 1 2 1 1 2 1 1 2 1 1 1 1 1 1 1 2 2 3 3 MAXIMUM THICKNESS (mm) 15.34 3.72 3.89 11.14 2.45 2.45 4.83 5.02 3.75 2.51 2.06 2.22 2.9 4.51 4.54 3.97 2.33 3.37 12.19 3.37 7.16 11.41 6.02 6.57 7.16 14.52 11.19 9.17 (53.55) 20.48 64.64 33.52 45.44 60.42 21.36 (38.06) (32.13) 36.25 36.78 (31.40) 50.68 28.16 (58.86) (57.26) 34.25 MAXIMUM LENGTH (mm) Table 8. Summary Data for Scrapers and Retouched Flakes, 31HK2502. MAXIMUM WIDTH (mm) 41.23 21.68 21.76 34.23 13.15 20.58 26.52 16.82 17.77 (16.75) 45.04 18.24 (28.25) 26.86 20.75 (29.76) (26.69) 34.45 21.06 WEIGHT (gm) 30.60 1.57 1.46 23.87 0.58 1.16 4.89 8.89 1.41 0.39 0.84 0.81 1.49 2.64 1.68 1.43 2.21 1.26 14.12 2.59 5.03 13.32 1.23 3.38 10.31 26.6 6.07 4.84 Opposing Lateral Edges Ground Opposite Side Exhibits Bifacial Retouch Forming A Point. Probably Hafted Opposite Side Utilized (32°) Opposite Edge Utilized (27°) Alternating Retouch Side Retouch On Opposite Faces Of Flake Opposite Edge Utilized (48°) Retouch Discontinuous One Edge Contains Spoke Shave Concavity Alternate Face Retouch Graver Spur, Bit Made On Flake Platform, Two Non-Bit Edges Minimally Retouched NOTES Chapter 5. 31HK2502 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 101 Chapter 5. 31HK2502 Figure 26. Unifaces, Site 31HK2502. A: Type IIb End Scraper (a143), B: Type III End Scraper (a206), C: Type IV End Scraper (a316), D: Type I Side Scraper (a118), E-G: Type IIa Side Scrapers (a43, a266, a520 respectively) H: Type IIb Side Scraper (a63), I: Type IIc Side Scraper (a527), J-K: Type III Side Scraper (a132 and a1166 respectively), L: Type Ib Pointed Scraper (a552). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 102 Chapter 5. 31HK2502 to form a functional haft element. The bit width of Specimen 206a achieved a maximum dimension of 34.58 mm, which is within the range for the type. The haft element, or butt end, of the specimen was broken off in a hinge fracture termination, suggesting that it was discarded upon breakage. The bit end was steeply pressure flaked forming a convex working surface with edge angles ranging between 48° and 58°. Although edge rounding was a common occurrence on most end scraper types at the Hardaway Site, Daniel (1998:79-83) noted that none of the Type III end scrapers exhibited this wear pattern. The specimen from 31HK2502 exhibited a sharp bit edge, as well. Daniel suggests that edge dulling and rounding on end scrapers may evidence the scraping of, in particular, dry hides. Other functions proffered for scrapers not exhibiting edge rounding, were scraping of resistant material. However, since some examples within each of the other end scraper types also failed to exhibit edge dulling, it is likely that much of this variation is explained by the state of resharpening at the time of discard. Coe (1964:76) observed that all of the edges on Type III End Scrapers were irregular and sharp. Scraping of hard surfaces would not maintain a sharp edge. (c) Type IV End Scrapers. A single Type IV End Scraper made of non-patinated, aphyric Uwharries Southern Rhyolite was recovered from investigations at 31HK2502 (Figure 26:C). Daniel (1998:7576) extracted this type from the Hardaway scraper collection to segregate a group of informal end scrapers that he believed were hand-held during use. Type IV End Scrapers are predominantly unshaped with the exception of the bit end and they exhibit large bit widths. He notes, however, that approximately one half of the specimens displayed restricted retouch on edge margins near the bit. Daniel further observes that blank selection for Type IV End Scrapers was quite variable, suggesting that few design restrictions were placed on their manufacture. Moreover, bit placement was, most often, lateral to the striking platform of the blank. The overall morphological characteristics suggest to Daniel that the tool type was manufactured on an ad-hoc basis to be used in an immediate and specific task. In other words, this type was considered to represent an expedient end scraper form. The specimen from 31HK2502 was manufactured from a large Type I Biface reject that still contained the striking platform of the original flake blank. In this case, the blank consisted of a large core flake removed from a block core. The bit edge, as is common for the type, was placed lateral to the striking platform. A portion of the bit appears to have been broken off in a step fracture, probably during use. The remaining bit width is 35.08 mm. The outline of the working edge is variably plano-convex to plano-concave due to the remnant flaking patterns of the broken biface blank. The specimen was recovered in a Stage II shovel test (N545/E465) at a depth of 40 to 50 cm bs; a depth suggestive of Early Archaic association. (2) Side Scrapers. Twenty side scraper specimens were identified in the 31HK2502 collection. Daniel (1998:6683-84) defines side scrapers as unifaces containing “one or more unifacially retouched working edges that parallel the long axis of the flake blank.” Four side scraper types and numerous subtypes were identified by Daniel in the Hardaway scraper collection. Side scrapers from 31HK2502 can be placed in Types I, II and III. (a) Type I Side Scrapers. One Type I Side Scraper (a118) was identified in the 31HK2502 collection (Figure 26:D). Coe (1964:77) defined this scraper type as being made from large, wedgeshaped flakes produced from directional cores. One of the long sides of the flake blank was subjected to steep percussion retouch, while one or both ends of the blank also received retouch shaping. Daniel (1998:84) further adds that the bit edges were generally convex to straight in shape and cross-sections appeared wide and flat (lenticular). Type I Side Scrapers were equally distributed between Levels II, III and IV at the Hardaway Site, in association with Kirk and Palmer Corner Notched and Hardaway Side Notched points and Hardaway Blades. The specimen from 31HK2502 was made from a large core flake of finegrained, porphyritic Mill Mountain Rhyolite. One long side of the blank was steeply retouched along a straight segment measuring about 21.5 mm, while the opposite, unmodified and acute-angled edge exhibited nibble scars and dulling along its margin suggestive of cutting functions. Retouch along the scraping edge originated from the dorsal surface of the core flake, which equally as straight in outline as the ventral surface of the flake at this location. The scraper bit edge was lightly dulled from use. The angle along the working edge ranged between 77° and 88°, suggesting the scraping of resistant substances such as Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 103 Chapter 5. 31HK2502 wood or bone. The specimen was recovered from a Stage II shovel test (N550/E445) between 20 and 30 cm bs. (b) Type II Side Scrapers. Building on Coe’s (1964:79) original description for the Type II Side Scraper, Daniel (1998:84-93) separated individual specimens further into two types based on the size of the flake blanks used to manufacture the tools. Type IIa refers to the classic description of a Type II Side Scraper. That is, scrapers made on large, irregular flakes whose bit edges were retouched without concern for regularizing the marginal outline. Daniel’s Type IIb was reserved for very small flakes, usually of less than 30 mm in length. To these two subtypes, the current study adds an additional one. This subtype is referred to as “Type IIc Side Scrapers.” This subtype is similar morphologically to the other two subtypes, but it is distinguished by “alternating retouch” on the bit edge. Alternating retouch occurs on an edge of a lithic flake in such a way that it alternates between dorsal and ventral sides from one end to the other of the edge (see Bordes 1961; Crabtree 1972). This particular retouch pattern is created by rotating the face of the tool at approximately the midpoint of the working edge during its initial formation, resulting in balanced, continuous lines of retouch shifting from a point of origin on one face of the flake to the other. This technique is widely distributed in Gravettian and other Upper Paleolithic assemblages in the Old World and in Paleoindian assemblages in northeastern North America. The purpose of the pattern is unresolved at present. Coe (1964:79) observed that Type II Side Scrapers were strongly associated with the Palmer and Hardaway levels (III and IV) at the Hardaway Site. Nine Type IIa Side Scrapers were recognized in the 31HK2502 collection (Figure 26:E-G). They were generally manufactured on core flakes from block cores, but one specimen (a266) was fashioned from a large FBR. Angles on retouched edge margins ranged between 53° and 75°, although most were dominated by angles between 53° and 63° (Table 8). Two of the scrapers exhibited two retouched edges, while the remainder displayed only one. Individual specimens were predominantly recovered from vertical positions between 20 and 40 cm bs, although two specimens were found at depths of 40 to 50 cm bs. Three Type IIb Side Scrapers were recovered (Figure 26:H). These specimens were fashioned from thin flakes, although only one (a63) was complete enough to determine blank type. This was a small core flake removed from a block core. One fragmentary specimen (a423) exhibited evidence of retouch on opposite elongated sides. Specimen a63 was made of Type I Rhyolite Tuff and it was found at a depth of 40 to 50 cm bs in a Stage I shovel test (N550/E450). Specimen a423 was made of Uwharries Western Rhyolite and it was recovered between 40 and 50 cm bs in a Stage III shovel test (N541.25/468.75). Finally, Specimen a1155 was fashioned from Uwharries Southern Rhyolite and it was excavated from TU1 at a depth of 20 to 30 cm bd. A single Type IIc Side Scraper was identified in the collection (Figure 26:I). The subtype, as discussed above, is distinguished from the other two subtypes by the presence of alternating marginal retouch. Although only one specimen of this subtype was found at 31HK2502, several other specimens were recognized at other sites in this testing package. Specimen a527 was made of Uwharries Western Rhyolite and it consisted of a lateral section portion (Table 8). It was found at a depth of 50 to 60 cm bs in a Stage III shovel test (N511.25/E488.75) excavated in Sample Block 1. The low depth at which the scraper fragment was recovered is consistent with an Early Archaic or Paleoindian association. The steep marginal retouch was carefully applied, resulting in the formation of a working edge angle of 46° to 49°. (c) Type III Side Scrapers. Four Type III Side Scrapers were identified in the collection from 31HK2502 (Figure 26:J-K). Type III Side Scrapers were made on thin and narrow flakes (Coe 1964:79). Retouch occurred along the edges parallel to the long axis of the flakes and generally involved reshaping of only the working edges. No accommodations for hafting were present. This scraper type was found predominantly in Levels II and III at Hardaway, levels associated with Palmer through Morrow Mountain occupation (Coe 1964:73). Very few of these forms were found at Doerschuk, however, which suggests that their primary association may be with Early Archaic phases. Daniel (1998:93) indicates that the Type III Side Scrapers from the Hardaway collection were made on bladelike flakes, which, in this report, are referred to as blade flakes. One specimen (a132) Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 104 Chapter 5. 31HK2502 displayed two marginally retouched sides, while two (a801 and a1166) of the three remaining specimens exhibited acute, non-retouched utilized edges opposite the retouched edges. Specimen a1166 was manufactured from a green metasedimentary (metasiltstone) material, Specimen 801 was made of aphyric Uwharries Southern Rhyolite and Specimens a132 and a809 were composed of fine-grained, porphyritic Uwharries Eastern Rhyolite. Specimens a801 and a809 were recovered from shovel tests in Sample Block 15 at depths of 40 to 50 cm bs. Specimen a132 (Figure 26:J) was found at a depth of 10 to 20 cm bs in a Stage II shovel test (N555/E439), while Specimen a1166 (Figure 26:K) was recovered in a shovel test in Sample Block 14 at a depth of 30 to 40 cm bs. (d) Indeterminate Side Scrapers. Diagnostic characteristics of two of the side scrapers (a673 and a854) were insufficiently displayed to allow detailed typological assignments. Specimen a673 was made of Uwharries Western Rhyolite and it was recovered from a Stage III shovel test in Sample Block 10 at a depth of 30 to 40 cm bs. Specimen a854 was composed of aphyric Uwharries Southern Rhyolite. It was found at a depth of 10 to 20 cm bs in a Stage III shovel test excavated as part of Sample Block 8. (3) Pointed Scrapers. One specimen (a552) meeting the typological description of a Pointed Scraper was identified in the collection (Figure 26:L). This scraper type was infrequently found in the Hardaway Site assemblage. Coe (1964:79) described it as a tool manufactured from a thick flake in which steep unifacial retouch was applied to two sides of the dorsal surface to form a point at one end. The vertical distribution of the pointed scrapers at Hardaway was indistinguishable from Type I end scrapers and suggests, again, an Early Archaic association. Daniel (1998:102) recognizes two subtypes within the Hardaway collection. Type Ia corresponded to the original description provided by Coe, while Type Ib specimens were made on thinner flake blanks. The specimen from 31HK2502 meets the definitional criteria for the Type Ib Pointed Scraper. It was made on a broad, flat flake of indeterminate derivation owing to the absence of a striking platform on the fragment. A concavity on one side of the point exhibited unifacial retouch and dulling suggesting that the scraper was a composite tool incorporating a spoke shave feature as well. The specimen was composed of fine- grained, patinated Uwharries Western Rhyolite and the retouched surfaces exhibited lenticular cross-sections. It was recovered from a Stage III shovel test excavated in Sample Block 4 at a depth of 50 to 60 cm bs, indicating a probable Early Archaic affiliation. (4) Other Unifacial Tools. In addition to the more conventional scraper types, four other unifacial tools were recognized in the 31HK2502 collection. These consist of a denticulate, a graver/burin, a spoke shave and a possible flake adz. Each of these is described below. (a) Denticulates. A single denticulate (a56) was identified in the 31HK2502 collection (Figure 27:A). Denticulates represent flakes that exhibit a series of regularly spaced notches/projections along a working edge creating a saw-like appearance (Daniel 1998:103). The morphology of the working edge suggests that the tool functioned as a knife or saw for processing materials of soft to moderate hardness. The specimen from 31HK2502 was made of finegrained Uwharries Western Rhyolite (Table 8). It was manufactured from a large core flake that was subsequently shaped on two margins. The denticulated margin consisted of a series of three, broad, evenly spaced conchoidal flake scars forming four “teeth.” The opposite margin received a minimal amount of bifacial retouch along an acute working edge. Indentations at the termination of the retouched edges may indicate that the tool was hafted in this location. The indentation on the denticulated margin was steeply retouched in the same fashion as the teeth. The specimen was recovered from a Stage II shovel test (N 564/E438) at a depth of 30 to 40 cm bs. (b) Gravers/Burins. A single flake tool (a1004) fitting the description of a graver/burin was recognized in the collection (Figure 27:B). Gravers represent flakes with short unifacially flaked projections (Daniel 1998:104; Goodyear 1974:55). It is generally held that gravers served as engravers or points for slotting bone and wood. The projection on the specimen from 31HK2502 is unifacially shaped on one side, but the other side is step fractured, resembling the scar of a burin spall. Hence, the classification as a “graver/burin.” This step fracture may have only been fortuitously used to assist in creating the graver spur, however. The projection extends approximately 5.5 mm from the body of the flake. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 105 Chapter 5. 31HK2502 Evidence that the item was hafted, such as marginal edge damage or dulling, was not observed. The specimen was manufactured from a fine-grained, flowbanded Uwharries Southern Rhyolite. It was found at a depth of 40 to 50 cm bs in a Stage III shovel test excavated in Sample Block 7. Gravers are typically associated with Early Archaic assemblages and the depth at which this specimen was found is consistent with this affiliation. (c) Spoke Shaves. One specimen (a694) in the collection corresponds to the definition of a spoke shave (Figure 27:C). Typically, unifacial retouch is applied to a natural concavity on a flake margin to produce a strong working edge from which to scrape or shave narrow convex surfaces such as spear or arrow shafts. Goodyear (1974:50-53) identified both hafted and unhafted spoke shaves at the Brand Site, and Daniel (1998:104-105) recognized two hafted specimens in the Hardaway collection. Goodyear believes that most of the spoke shaves he recovered at the Brand Site were associated with the Dalton occupation. The specimen from 31HK2502 appears to represent a hand-held spoke shave, but the original flake blank has been snapped in its approximate center, so the actual issue of hafting cannot be resolved. The notch spans a straight-line width of approximately 11.5 mm and it was formed with marginal retouch extending onto the dorsal surface of the flake blank. Very short projections at the outer corners of the concavity appear to represent graver spurs, which is a common characteristic (Goodyear 1974:51-52) of hafted spoke shaves at the Brand Site. This phenomenon was also noted on the two hafted spoke shaves Daniel recognized in the Hardaway collection. The 31HK2502 specimen was manufactured from a fine-grained patinated Uwharries Southern Rhyolite, which exhibited a small amount of the original cortex of the block core. The tool was recovered from a Stage III shovel test in Sample Block 10 at a depth of 60 to 70 cm bs. This vertical position is consistent with an Early Archaic association. (d) Flake Adz. A large flake fragment (a1174) of coarse-grained porphyritic Uwharries Eastern Rhyolite exhibits characteristics that are consistent with a flake adz (Figure 27:D). The exterior margins of the broken flake are ground, creating a regularized oval outline and which could have aided in hafting it. No evidence of retouch is seen along the perimeter of the flake that might have facilitated hafting the flake or forming a sharpened adz bit. However, only approximately half of the item is represented and it is possible that it constitutes the butt end of the implement. Ethnographic studies of Australian Pitjantjatjara groups have documented the use of expedient adz flakes of similar morphology (Hayden 1979). In these instances, the bit edges gradually acquire a “retouched” edge with subsequent sharpening of dulled edges. At the time of initial hafting most of these tools were unretouched. Use wear consisted primarily of striations and polish/abrasion across the faces and edges of the flakes (Hayden 1979:56-57). The dorsal surface of the specimen from 31HK2502, which was mostly covered in cortex, appears to have experienced abrasive wear similar to what might be expected from adzing movements like those performed by the Aborigine. The item was recovered from TU1 at a depth of 30 to 40 cm bd. Utilized Flakes Thirteen un-retouched flakes exhibited wear patterns on one or more edges, indicating their use as expedient tools (Table 9). Metric and attribute data for utilized flakes are presented in Appendix G. All of these specimens exhibited edge attrition and bifacial nibbling scars indicative of cutting functions. Utilized edge angles fell between 21°and 44°. Most of the specimens appear to have been derived from block cores (i.e. core flakes, blade flakes and bladelet flakes), but two represented large FBRs. One specimen was fragmentary and a derivation could not be discerned. Aphyric raw material types (Uwharries Southern Rhyolite and Type I Rhyolite Tuff) dominate the collection. Use wear would be easier to detect on these microcrystalline rock types, suggesting that the porphyries might be under-represented because of observational difficulties. Miscellaneous Stone Tools Several non-chipped stone, lithic artifacts were recovered from the investigation at 31HK2502. These consist of a tabular slate fragment, a grooved abrader fragment, a possible grinding stone, a possible hearth deflector, three steatite bowl fragments and an anvil stone. Metric and attribute data for these Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 106 Chapter 5. 31HK2502 Table 9. Summary Data for Utilized Flakes, 31HK2502. WEIGHT (gm) EDGE ANGLE Fragment Cutting Blade Flake 1 Uwharries Southern Rhyolite 4.63 (37.11) 13.33 2.30 39° 271265a72 37 Whole Cutting Bladelet Flake 2 Type I Rhyolite Tuff 2.48 17.06 5.77 0.19 37° 271265a109 59 Fragment Cutting Core Flake 1 Type I Rhyolite Tuff 1.69 15.40 16.76 0.50 21° 271265a176 92 Whole Cutting Blade Flake 1 Uwharries Southern Rhyolite 3.84 28.93 17.87 1.58 25° 271265a334 209 Fragment Cutting FBR 1 Type I Rhyolite Tuff 2.28 (32.28) 26.95 2.44 25° 271265a420 249 Fragment Cutting Core Flake 2 Uwharries Western Rhyolite 3.72 (14.66) (17.89) 0.89 29° 271265a540 314 Whole Cutting Core Flake 2 Type I Rhyolite Tuff 7.30 45.98 30.80 8.14 21° 271265a692 383 Whole Cutting Blade Flake 2 Uwharries Southern Rhyolite 3.01 31.64 11.04 1.00 30° 271265a703 390 Whole Cutting FBR 1 Uwharries Southern Rhyolite 3.28 30.54 16.30 1.63 30° 271265a734 403 Whole Cutting Core Flake 2 Uwharries Southern Rhyolite 5.92 50.62 30.16 6.11 26° 271265a784 422 Fragment Scraping Flake Frag. 1 Uwharries Southern Rhyolite 4.94 (19.69) (7.47) 0.66 44° 271265a922 476 Fragment Cutting Blade Flake 1 Uwharries Southern Rhyolite 2.43 (23.10) (13.65) 0.79 31° 271265a1159 564 Whole Cutting FBR 1 Uwharries Southern Rhyolite 1.78 11.73 21.78 0.44 25° CONDITION BLANK TYPE LITHIC RAW MATERIAL TYPE MAXIMUM WIDTH (mm) MAXIMUM LENGTH (mm) 18 BAG NO. UTILIZED EDGES 271265a44 SPECIMEN # UTILIZED FLAKE FUNCTION MAXIMUM THICKNESS (mm) UTILIZED FLAKES: 31HK2502 artifacts are presented in Appendix H. (1) Tabular Slate Fragment. A single piece of (a195) tabular slate was recovered that might represent pendant fragment (Figure 27:E). Evidence of cutting marks or grinding striations was not observed. The fragment measured 3.13 mm in thickness, 20.49 mm in length and 19.32 mm in width. It was found at a depth of 30 to 40 cm in a Stage II shovel test (N485/ E505). (2) Grooved Abrader Fragment. A flat fragment of laminar metasedimentary material (a325) displayed a portion of a shallow, ground groove on one of its faces (Figure 27:F). The rounded groove measured 5.15 mm in width and was less than 1.5 mm in depth. Goodyear (1974:69-71) reports similar grooved abraders from the Dalton Deposits at the Brand Site. These abraders were made of locally available sandstone and ferruginous sandstone and could be segregated into two groups. One group consisted of relatively narrow grooves ranging between 3 mm and 8 mm in width, while the other was composed of specimens of 10 mm to 15 mm in width. The smaller group coincides with the dimensions of the specimen from 31HK2502. Goodyear suggested that the size and shape of this narrower group argued that they were used to manufacture thin cylindrical objects such as bone points, pins and perforating tools. Interestingly, Goodyear argued that the wider group of abraders was likely used in the manufacture of Dalton Adzes based on the contents of the Hawkins Cache (see Morse 1971:18). The specimen from 31HK2502 was recovered at a depth of 20 to 30 cm bs in a Stage III shovel test belonging to Sampling Block 2. It measures 8.32 mm in thickness, 41.08 mm in length and 9.41 mm in width. (3) Possible Grinding Stone. A moderate sized slab-like fragment (a66) of arenite displays an unusually smooth face on one of its faces (Figure 27:G). Smoothed surfaces occur naturally on pieces of arenite in the region, but this particular face exhibits patches of what appear to be grinding facets on Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 107 Chapter 5. 31HK2502 Figure 27. Miscellaneous Stone Tool Classes, Site 31HK2502. A: Denticulate (a56), B: Graver/Burin (a1004), C: Spoke Shave (a694), D: Flake Adz (a1174), E: Tabular Slate Fragment (a195), F: Grooved Abrader (a325), G: Grinding Stone (a66), H: Steatite Bowl Fragments (a902). its higher contours. It measures 35.63 mm in thickness, 63.12 mm in length and 56.63 mm in width and weighs 198.50 gm. It was found in a Stage I shovel test (N520/E480) at a depth of 30 to 40 cm bs. (4) Possible Hearth Deflector. Another flat piece of unmodified arenite may have served as a deflector stone in a hearth (a812). It measures 16.42 mm in thickness, 151.58 mm in length and 69.38 mm Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 108 Chapter 5. 31HK2502 in width and weighs 252.08 gm. It was found in a Stage III shovel test belonging to Sample Block 8 at a depth of of 40 to 50 cm bs. (5) Steatite Bowl Fragments. Three fragments (a902) from a single steatite bowl were recovered from a single Stage III shovel test in Sample Block 11 at a depth of 10 to 20 cm bs (Figure 27:H). The talc-schist material was medium light gray in color and porous. Steatite vessel fragments are sporadically found on Fort Bragg (see McNutt and Gray 2007; Millis et al. 2010; Ruggiero 2003). The material is commonly found at locations in the Piedmont, although the nearest known quarries are in South Carolina and south central Virginia (see Millis et al. 2010:460-461; Truncer 2004). The interiors of the sherds were smoothly finished, although the regular contours were broken by numerous open pockets. The exteriors, by contrast, were finished with regular, shallow grooves from a serrated smoothing tool. Wall thickness ranged from 9.28 mm to 11.02 mm. (6) Anvil Stone. A large anvil stone (a1190) was recovered from a Stage II shovel test in Sample Block 14 at a depth of 30 to 40 cm bs (Figure 28). The anvil was made from a flat cobble of grano-diorite measuring 25.72 mm in thickness, 198.46 mm in length and 120.51 mm in width. It weighed 1179.34 gm. Broad, shallow pitted depressions occurred in the center of the faces of both sides of the anvil. The largest depression measured 100.96 mm in diameter, while the smaller one achieved a maximum diameter of 74.71 mm. The depressions extended to maximum depths of only 2 to 3 mm. The cobble had been fractured along the long axis in the past, but this event preceded its use as an anvil since the face of the fracture was patinated, while the pits in the depressions were rough and fresh looking. The nature of the pitting further suggests that one of the primary uses of the anvil was to reduce hard rhyolite and quartz block cores. Precontact Ceramics Although diagnostic Woodland period projectile points were not recovered during the investigation at 31HK2502, fieldwork produced a relatively large collection of precontact ceramics. The collection consisted of 95 sherds, including representatives from four recognized ceramic series: (1) New River, (2) Hanover, (3) Cape Fear and (4) Yadkin. Fortynine individual vessels were recognized in this collection. Metric and attribute data on precontact ceramics are contained in Appendix I. Monitoring of temper types within series has proceeded in an ad hoc fashion in earlier investigations (see Cable and Cantley 2005a, b, 2006, Cable 2010), but the enormity of the inventory of temper constituents and various compositional combinations seen in ceramic collections makes it imperative that in the future a universal system be developed to accommodate cross-project comparability. Formerly, temper classes were assigned letter designations as they were historically recognized within each series, creating a confusing list of different letter designations for the same temper type within each series. Here, temper constituent designations are standardized into a single classificatory system. The letters “a” and “b” are reserved for the basic paste representatives diagnostic of each series as described previously. The remaining letter designations note the addition of other temper constituents to these basic paste categories as described below: (c) Crushed Arenite (d) Granule to Pebble Sized Angular/Subangular Quartz Sand and/or Crushed Quartz (e) Crushed and Ground Indurated Granite (f) Medium to Granule Sized Angular/Subangular Feldspar Sand More than one of these temper constituents was sometimes present in a single sherd. In these cases, the constituent designations would include combinations of letter designations (e.g. “c/e” or “e/f”, etc.). During the course of the analysis it also became apparent that some of the pottery exhibited extremely hard pastes, much harder, in fact, than the typical paste previously defined for each series. A hardness scale consisting of three attribute states was developed to describe this variation. These consisted of: (1) friable, (2) compact and (3) hard. The attribute state of “friable” corresponded to sherds that were easily crushed on their edges with only minimal application of pressure with the thumb. A “compact” state was identified in instances where moderate pres- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 109 Chapter 5. 31HK2502 Figure 28. Anvil Stone (a1190), Site 31HK2502. sure with the thumb was required to break the edge of the sherd. Finally the “hard” state was reserved for those sherds whose edges were not easily broken even with extreme thumb pressure. Generally, pottery achieves greater hardness when one or more of three conditions prevail (Shepard 1954:114). These are: (1) the pottery is made from low-fusing, dense-firing clay, (2) the pottery is fired at a high temperature and/or (3) the pottery is fired in an atmosphere promoting vitrification. It is common to observe in the local ceramic sequences of the Southeast, a progression from friable ceramics to much harder, more durable pottery in the later stages of cultural development. Clearly, Mississippian potters had developed technological strategies Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 110 Chapter 5. 31HK2502 to produce extremely hard, sometimes vitrified pottery and it is quite possible that this advancement was meant to adapt to a more sedentary, agricultural life way in which durability and longevity of pots would have been advantageous. Similar adaptive technologies may have been incorporated into sedentary, agricultural societies peripheral to the Southeast Appalachian Mississippian culture as well, and for the same reasons. Currently, a sophisticated method for measuring hardness in “primitive” pottery does not exist due to its porosity and heterogeneity of paste composition. The “thumb” test, although somewhat imprecise, provides a relatively objective and empirical basis for identifying relative hardness in precontact pottery assemblages. For an extended discussion of hardness in precontact ceramics see Chapter 4. This method of analysis resulted in segregating extremely hard or vitrified pottery that would have traditionally been placed in the Cape Fear (sandtempered), Hanover (sand- and grog-tempered) and Yadkin (crushed rock- and granule sand-tempered) series. As a means of preserving data to test the “hardness trajectory” hypothesis, these hard examples were separated into new series variants distinguished by the Roman numeral suffix “III.” Thus, extremely hard variants of the series are respectively referred to here as Hanover III, Cape Fear III and Yadkin III. If the efficacy of the hypothesis is later confirmed, then it would be appropriate to assign new series names to the Cape Fear III and Yadkin III variants since they are generally viewed as Middle Woodland series. The ceramics from each series are described below. New River Series New River series ceramics are associated with the Early Woodland period in the region. The series was originally defined by Loftfield (1976) who characterized it as a coarse sand-tempered pottery of compact composition and gritty consistency characterized by cord marked, fabric impressed, thong simple stamped and plain exterior surface treatments. Correlated series of the Early Woodland period include Lenoir (Crawford 1966) and Deep Creek (Phelps 1983). The New River series in the Fort Bragg area is poorly understood. Herbert (2003:156) indicates that both net-impressed and parallel/overstamped cord marked types are recognized, while Cable and Cantley (2005a, b) have also recognized simple stamped, fabric impressed and plain surface treatments. The predominant classificatory criterion used to identify the series is an abundance of quartz sand temper as Loftfield (1976) prescribed. Herbert (2003:156) proposes that an arbitrary cut-off at about 15% sand temper density should be used to distinguish Cape Fear (< 15%) and New River (>15%) in the Fort Bragg area. Cable and Cantley (2005a, b) have recognized two major paste variants in Fort Bragg assemblages. These are variants Ia and IIa. Variant Ia consists of sherds dominated by abundant medium to very coarse subangular to rounded quartz sand grains in a gritty, friable paste matrix. Variant IIa contains of a fine, compact, relatively harder paste matrix with abundant to moderately dense medium to very coarse quartz sand inclusions. Variant IIa may correlate with the traditionally defined Cape Fear series. Feldspar sand is sometimes found in variant Ia paste, and it is designated New River If. Only one New River series sherd was recognized in the collection from 31HK2502, a New River Ia sherd with an indeterminate and eroded exterior surface. The sherd was thick, measuring 11.59 mm. It was found at a depth of 20 to 30 cm bs in a Stage III shovel test belonging to Sample Block 19. The paste was friable and gritty to the feel. Subangular quartz sand was abundantly concentrated in the paste matrix and ranged between medium and coarse in grain size. Hanover Series South (1960:16-17; 1976: 28) defined the Hanover series from surface collections made in 1960 on sites located between Wilmington, NC and Myrtle Beach, SC. Pottery affiliated with this series is tempered with aplastic clay lumps, most of which consist of grog fragments from crushed sherds. A correlate on the north coastal plain of North Carolina is referred to as the Carteret series (Loftfield 1979:154157). Cord marked and fabric impressed exterior surface treatments were included in the original series description, but other grog-tempered types have been recognized subsequently. Herbert (2003:75) has identified minority proportions of check stamped, plain (smoothed), net impressed, simple stamped, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 111 Chapter 5. 31HK2502 punctate, and random straw bundle punctate surface treatments. The latter two types are traditionally subsumed under the Refuge series on the central South Carolina coast where grog/clay temper occurs as a major temper variant (see Cable 2002: 195-201). A great deal of variability in paste composition has been noted in Hanover assemblages, primarily related to differences in quartz sand contributions, clay versus grog distinctions, and the density and size of clay/grog particles (see Cable et al. 1998; Herbert 2003; 74-75). Herbert (2003:191-192) recognizes two broad variants that he believes are sequential. The earliest variant, Hanover I, is primarily sandtempered with minor amounts of finely crushed grog. Surface treatments attributed to Hanover I pottery consist of cord marking, fabric impressing and check stamping. He estimates an age range of AD 400-800 for Hanover I pottery. The later variant, Hanover II, is inferred to correlate with a Late Woodland temporal range (AD 800 through 1500). Hanover II pottery is characterized by pastes with only small amounts of sand and abundant grog particles. The dominant surface treatment of later variant is fabric impressing. Cable and Cantley (2005a, b) have followed Herbert (2003) in describing Hanover series ceramics in the Fort Bragg area. However, in an attempt to more fully describe the temper and paste variation, these basic groupings have been expanded to include sub-variants. As research has progressed, the number of sub-variants has increased to accommodate an ever-expanding range of temper constituents and paste qualities. There are three major paste variants found in Hanover pottery at Fort Bragg based on relative hardness. Hanover I is characterized by a friable paste, Hanover II exhibits a compact paste and Hanover III contains a hard to vitrified paste. Within each of these relative hardness groupings are two primary temper constituent designations. Designation “a” refers to pastes with sparse to moderate amounts of fine to medium quartz sand and abundant medium to large grog particles. Designation “b” includes pastes with moderate to abundant amounts of medium to coarse quartz sand and sparse to moderate densities of medium to large grog particles. Other temper constituents are sometimes added to these basic paste variants. These include crushed arenite (“c”), gran- ule to pebble sized angular to subangular quartz sand and, rarely, crushed quartz (“d”), crushed and ground indurated granite (“e”) and medium to granule sized subangular feldspar sand (“f”). Crushed arenite and indurated granite almost exclusively occur in Hanover III paste. Indurated granite temper was originally identified as sandstone conglomerate because of similar texture characteristics, but thin-sectioning of rock samples indicated that most of the temper in this category corresponds to a fine grained, indurated granite (see Appendix K). Some of the temper in this category may actually represent sandstone conglomerate, but this was not differentiated in the ceramic analysis due to the late submittal of samples. Eight of the sherds from the collection correspond to the Hanover I series variant (Table 10). Seven vessels are represented in the collection. Hanover Ib paste dominates and includes both check stamped and fabric impressed surface treatments (Figure 29:A-B). Two sherds of Hanover Id paste from the same vessel exhibited cross-stamped cord marking (Figure 29:C). One of the cord marked sherds was a rim that exhibited a tapered rounded lip and a flared orifice (Figure 30:A) suggestive of a flare-rimmed open mouth jar. The cordage impressions were not adequately displayed to determine cordage diameter. One of the fabric-impressed sherds (p222) displayed a medium weave pattern with a rigid warp. Surface treatment data from the sample are consistent with an early Middle Woodland association. Mean vessel width for the Hanover I sample was 7.81 mm (SD=1.35 mm). The Hanover II sample from the site was the largest, consisting of 63 sherds (Table 11). Fabric impressing occurred on 92.5 percent (n=37) of the sherds with recognizable surface treatments (n=40). Minority surface treatments included cord marking (n=2) and plain (n=1). Twenty-six separate vessels were represented in the Hanover II sample. Included in the sample were three paste sub-variants. Sub-variant IIa dominated, constituting 79.4 percent (n=50) of the sample, while Sub-variant IIb represented 15.9 percent (n=10) and Sub-variant IId contributed 4.8 percent (n=3). Rigid warps represented 75 percent (9 of 12) of the fabric impressed weave patterns in the Hanover IIa collection (Figure 29:D-F), while 25 percent was made up of flexible weave patterns (Figure 29:G). The other sub-variants had smaller recognizable surface treatment sample sizes. A rigid warp pat- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 112 Chapter 5. 31HK2502 Table 10. Hanover I Ceramic Inventory, 31HK2502. SIZE WALL THICKNESS (mm) VESSEL NO. 8.21 22 Gritty Body 3 7.70 15 Gritty Body 3 9.60 11 323 Gritty Body 2 9.64 27 271265p459 267 Eroded Body 3 8.47 22 Hanover Ib indet. dec. 271265p352 216 Smoothed Irregular Body 2 7.29 17 Hanover Id Cord Marked 271265p6 4 Cross Smoothed Irregular Body 3 6.39 23 Hanover Id Cord Marked 271265p9 5 Cross Eroded Rim 3 5.55 23 271265p458 267 Hanover Ib Check St. 271265p23 10 Hanover Ib Fabric Imp. 271265p222 122 Hanover Ib Fabric Imp. 271265p558 Hanover Ib indet. dec. Rigid/Med. tern was observed on one Hanover IId sherd (Figure 29:H), while two flexible weave patterns were present on Hanover IIb sherds. Stamping patterns on the two cord marked sherds were evenly divided between cross-stamped (Figure 29:I) and parallel (Figure 29:J). The latter exhibited wide cord impressions measuring around 3.4 mm in diameter. Two rim sherds (p1065 and p1066) with indeterminate surface treatment were present in the Hanover IIa sample. They can supply only minimal vessel form information due to their small size. The former rim exhibited a tapered rounded lip and a slightly flaring orientation, suggesting that it derived from a small, flare-rimmed, open mouth jar (Figure 30:B). The latter had a flat lip and a direct or straight orientation. Its small size, in combination with its morphological features, indicates a possible deep bowl form. A sherd of Hanover IIb Cord Marked (p64) also belonged to the rim of a vessel (Figure 29:J; Figure 30:C). This rim exhibited thickening to the outside and had a recurved profile suggestive of an open mouth jar with a slightly constricted neck. Mean vessel wall thickness for the Hanover II sample was 7.41 mm (SD=1.93 mm). The dominance of INTERIOR FINISH STAMP PATTERN 4 Hanover Ib Check St. WARP/WEFT Body BAG NO. Eroded CERAMIC TYPE SPECIMEN NO. SHERD TYPE HANOVER I CERAMICS: 31HK2502 fabric impressing in the Hanover II sample strongly suggests a late Middle Woodland or Late Woodland association. The Hanover III sample consists of 13 sherds (Table 12). All thirteen sherds represent different vessels. Of the eight sherds with recognizable surface treatments, five display fabric impressing (Figure 31:A-B) and three have cord marked surfaces (Figure 31:C-D). The reduced dominance of fabric impressing in the Hanover III sample could supply evidence of a temporal difference between Hanover II and III. Average cord impression width on the 31HK2502 cord marked sample ranges between 1.79 and 2.19 mm. Both flexible and rigid warps were observed on the fabric-impressed sherds. The finer Hanover IIIa paste dominates, comprising 61.5 percent of the sample. Hanover IIIb (23.1 percent) and Hanover IIId (15.4 percent) occur in much lower proportions. Interior finishing of vessel walls in the sample is undistinguished, with the exception of one sherd (p588) that exhibits broad scraping marks made with a serrated tool (Figure 31:E), which is a common attribute of the early Late Woodland Uwharrie series pottery of the central Piedmont (Ward and Davis 1999:101). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 113 Chapter 5. 31HK2502 Figure 29. Hanover I and II Series Ceramics, Site 31HK2502. A: Hanover Ib Check Stamped (p458), B: Hanover Ib Fabric Impressed (p222), Hanover Id Cord Marked (p9), D-G: Hanover IIa Fabric Impressed (p612, p548, p988 and p967 respectively), H: Hanover IId Fabric Impressed (p634), I: Hanover IIa Cord Marked (p556), J: Hanover IIb Cord Marked (p64). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 114 Chapter 5. 31HK2502 A B C D Figure 30. Ceramic Rim Profiles, Site 31HK2502. Vessel walls are relatively thin in the sample, yielding a mean of 6.48 mm (SD=1.08 mm). Rim sherds were not recovered for the Hanover III sample. Cape Fear Series Stanley South (1976:18-20) first described the Cape Fear series, or ware-group as he conceived it, from surface collections made in 1960 on sites located between Wilmington, NC and Myrtle Beach, SC. He applied the term to virtually all of the sandtempered cord marked, fabric impressed, and net impressed pottery within the collection. Many have suggested that the series was too broadly applied by South and other investigations in North Carolina indicate that such material should be separated into a sequence of series based on sand inclusion coarseness (see Anderson and Logan 1981:107-108; Trinkley 1981:11). Coastal North Carolina sequences contain at least two sand grain size modes that are used to distinguish series (Phelps 1983). The Deep Creek series consists of pastes with abundant, coarse sand and it is chronologically correlated with Deptford. Other series (i.e. Mount Pleasant and Cape Fear) contain finer grained pastes and they are roughly dated to a post-Deptford or late Middle to Late Woodland context (i.e. AD 500-900 or 1500 to 1100 BP). Herbert (2003:156) distinguishes the sandtempered New River and Cape Fear series by quartz sand density in the southern Sandhills region. Cape Fear series material is characterized by low to moderate densities of quartz sand amounting to less than 15% of the paste body. In addition, he places all perpendicular stamped cord marked exterior surfaces in the Cape Fear series, while all net-impressed is assigned arbitrarily to New River. Cable and Cantley (2005a, b) distinguished two series variants for Cape Fear in the Fort Bragg area based on quartz sand density and paste characteristics. Cape Fear I was characterized by a friable, gritty paste and a moderate density of medium to coarse quartz sand temper. Cape Fear II exhibited a finer, more compact paste with sparse medium to coarse quartz sand temper. Only two Cape Fear series sherds were recovered during the current investigation (Table 13). They derived from different vessels. Both exhibited a hard, almost vitrified paste with sparse fine to medium quartz sand temper, and, on this basis, they have been recognized as members of the newly defined Cape Fear III series. In addition, both displayed cord marked surface treatments, but only one exhibited a clear stamping pattern. It was parallel stamped (Figure 31:F). Cord impressions on the face of the sherd averaged about 2.0 mm in diameter. The other sherd was a small rim with a truncated rounded lip and straight orientation. The thick wall of the sherd (8.38 mm) suggested that it formerly belonged to a large, straight-sided open mouth jar (Figure 30:D). Cord impressions continued to the top of the lip. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 115 Chapter 5. 31HK2502 Table 11. Hanover II Ceramic Inventory, 31HK2502. Hanover IIa Cord Marked Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa Fabric Imp. Hanover IIa indet. Hanover IIa indet. Hanover IIa indet. Hanover IIa indet. Hanover IIa indet. Hanover IIa indet. Hanover IIa indet. Hanover IIa indet. Hanover IIa indet. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa indet. dec. Hanover IIa Plain Hanover IIb Cord Marked Hanover IIb Fabric Imp. Hanover IIb Fabric Imp. Hanover IIb Fabric Imp. Hanover IIb indet. dec. Hanover IIb indet. dec. Hanover IIb indet. dec. Hanover IIb indet. dec. Hanover IIb indet. dec. Hanover IIb indet. dec. Hanover IId Fabric Imp. Hanover IId Fabric Imp. Hanover IId indet. Cross Flexible/Fine Flexible/Fine Flexible/Med. Rigid/Coarse Rigid/Coarse Rigid/Fine Rigid/Fine Rigid/Fine Rigid/Medium Rigid/Medium Rigid/Medium Rigid/Medium Parallel Flexible/Fine Flexible/Fine Rigid/Medium Eroded Eroded Smoothed Irregular Smoothed Smoothed Irregular Smoothed Irregular Smoothed Irregular Eroded Gritty Smoothed Irregular Eroded Smoothed Smoothed Smoothed Eroded Eroded Eroded Eroded Eroded Eroded Smoothed Smoothed Irregular Smoothed Irregular Smoothed Irregular Smoothed Irregular Smoothed Irregular Eroded Smoothed Smoothed Smoothed Smoothed Irregular Eroded Smoothed Smoothed Smoothed Irregular Smoothed Irregular Indeterminate Eroded Eroded Eroded Eroded Eroded Eroded Smoothed Smoothed Smoothed Smoothed Irregular Smoothed Irregular Smoothed Irregular Smoothed Eroded Smoothed Smoothed Smoothed Irregular Eroded Smoothed Smoothed Smoothed Smoothed Smoothed Irregular Smoothed Irregular Eroded Eroded Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Body Rim Rim Body Body Body Residual Residual Residual Residual Residual Residual Body Body Body Body Body Body Body Body Body Body Body Body Body Rim Body Body Body Body Body Body Body Body Body Body Body Residual 4 3 3 3 3 4 3 2 3 2 4 1 2 2 2 2 2 2 2 1 1 2 2 3 2 3 1 1 3 2 2 1 2 1 2 1 1 4 2 2 2 1 3 2 2 3 3 2 2 3 3 2 1 3 3 4 3 2 3 2 3 2 1 7.45 7.29 9.34 6.86 3.25 12.19 11.00 8.44 8.23 9.01 9.76 7.53 6.78 2.01 5.80 7.25 7.76 6.50 7.40 9.01 8.88 8.48 6.83 7.60 7.37 5.97 9.14 8.74 6.85 5.49 5.90 9.80 7.62 1.18 7.30 6.17 7.77 7.92 7.81 9.11 7.79 7.12 6.74 6.77 7.77 6.68 6.48 VESSEL NO. WALL THICKNESS (mm) SIZE SHERD TYPE INTERIOR FINISH 322 456 414 500 15 347 348 338 317 339 515 545 547 547 15 218 314 314 416 541 551 414 414 466 475 476 545 545 416 542 414 544 545 545 545 545 401 89 89 91 416 416 464 89 91 89 133 544 544 508 31 68 68 14 49 223 224 224 225 204 355 547 135 STAMP PATTERN 271265p556 271265p883 271265p763 271265p967 271265p35 271265p612 271265p613 271265p595 271265p548 271265p596 271265p988 271265p1067 271265p1079 271265p1079 271265p35 271265p357 271265p542 271265p542 271265p768 271265p1050 271265p1103 271265p763 271265p763 271265p900 271265p920 271265p921 271265p1066 271265p1065 271265p767 271265p1052 271265p764 271265p1057 271265p1068 271265p1068 271265p1068 271265p1068 271265p727 271265p169 271265p169 271265p173 271265p766 271265p766 271265p896 271265p169 271265p173 271265p169 271265p238 271265p1055 271265p1056 271265p978 271265p64 271265p124 271265p124 271265p34 271265p93 271265p366 271265p367 271265p367 271265p369 271265p327 271265p634 271265p1080 271265p241 WARP/WEFT BAG NO. CERAMIC TYPE SPECIMEN NO. HANOVER II CERAMICS: 31HK2502 26 38 35 40 2 30 30 29 25 29 43 18 47 47 2 18 25 25 36 18 47 35 35 39 39 39 18 18 35 18 35 18 18 18 18 18 7 7 7 35 35 39 7 7 7 12 44 18 41 3 6 6 1 4 19 19 19 19 16 31 48 13 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 116 Chapter 5. 31HK2502 Table 12. Hanover III Ceramic Inventory, 31HK2502. Hanover IIIa Cord Marked 271265p745 407 Hanover IIIa Fabric Imp. 271265p1078 547 Hanover IIIa Fabric Imp. 271265p525 Hanover IIIa Fabric Imp. 271265p1074 SIZE WALL THICKNESS (mm) VESSEL No. Body 3 8.31 49 INTERIOR FINISH 550 SHERD TYPE Hanover IIIa Cord Marked STAMP PATTERN BAG NO. 271265p1095 CERAMIC TYPE WARP/WEFT SPECIMEN NO. HANOVER III CERAMICS: 31HK2502 Parallel Parallel Smoothed Smoothed Irregular Body 3 6.90 34 Smoothed Regular Body 3 8.90 47 304 Smoothed Body 2 7.01 24 546 Smoothed Irregular Body 2 8.25 46 Rigid/Medium Hanover IIIa indet. 271265p538 313 Eroded Body 2 6.94 23 Hanover IIIa indet. dec. 271265p635 356 Smoothed Irregular Body 3 8.14 32 Hanover IIIa indet. dec. 271265p588 334 Serrated Scraped Body 3 10.11 28 Hanover IIIb Fabric Imp. 271265p189 101 Flexible/Coarse Smoothed Irregular Body 4 8.74 8 Hanover IIIb Fabric Imp. 271265p190 102 Flexible/Medium Hanover IIIb indet. dec. 271265p199 108 Hanover IIId Cord Marked 271265p97 52 Hanover IIId indet. dec. 271265p981 510 Parallel Smoothed Body 2 6.78 9 Gritty Body 3 6.76 10 Smoothed Irregular Body 3 7.90 5 Smoothed Body 2 6.48 42 Table 13. Cape Fear III and Yadkin III Ceramic Inventory, 31HK2502. Smoothed SHERD TYPE INTERIOR FINISH WARP/WEFT STAMP PATTERN Parallel Body VESSEL NO. 406 WALL THICKNESS (mm) 271265p742 SIZE Cape Fear III Cord Marked BAG NO. CERAMIC TYPE SPECIMEN NO. CAPE FEAR III AND YADKIN III CERAMICS: 31HK2502 4 8.45 33 Cape Fear III Cord Marked 271265p726 401 Smoothed Rim 2 8.38 33 Yadkin IIIe Cord Marked 271265p1073 546 Smoothed Irregular Body 5 8.10 45 Yadkin IIIc Cord Marked 271265p880 454 Smoothed Regular Body 3 7.67 37 Yadkin IIIe/f indet. 271265p271 152 Eroded Body 4 9.11 14 Yadkin IIIe/f indet. 271265p406 241 Eroded Body 4 7.33 20 Yadkin IIIe/f Plain 271265p418 248 Gritty Body 3 8.74 20 Cross Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 117 Chapter 5. 31HK2502 Figure 31. Hanover III, Cape Fear III and Yadkin III Series Ceramics, Site 31HK2502. A: Hanover IIIa Fabric Impressed (p1078), B: Hanover IIIb Fabric Impressed (p189), C: Hanover III Cord Marked (a1095), D: Hanover IIIb Cord Marked (p97), E: Hanover IIIa indeterminate decorated, Interior Serrated Scraped (p588), F: Cape Fear Cord Marked (p742), G: Yadkin IIIc Cord Marked (p880), H: Yadkin IIIe Cord Marked (p1073), I: Yadkin IIIe/f Plain (p418), J: Yadkin IIIe/f indeterminate (p406). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 118 Chapter 5. 31HK2502 Although other data will be brought to bear on this subject in other chapters, the possibility that the Cape Fear series has been too broadly defined to include all sparse to moderate sand-tempered postNew River pottery is very real. The difference in Cape Fear cord marked to fabric-impressed ratios between the Haag collection (Herbert 2003:69-70) and Stanly South’s (1960) survey collection should not be dismissed in this regard. South calculated a cord marked to fabric-impressed ratio of 1.6:1.0, which suggested to him that Cape Fear succeeded Hanover. Herbert, by contrast, generated a reverse ratio of 0.39:1.0 and concluded that Cape Fear preceded Hanover. A good deal of this disparity can probably be attributed to definitional issues. Herbert assigned only perpendicular stamped (cross stamped) cord marked specimens to Cape Fear, while sand-tempered parallel stamped cord marked and net-impressed were included exclusively with the New River series. However, the mere recognition of these differences does not resolve the potential for unrecognized post-Hanover sandtempered series to exist in the regional record. The McLean (MacCord 1966) and Buie Mound (Wetmore 1978) ceramic assemblages, as discussed in Chapter 3 (see also the discussion in Herbert 2003:193), constitute two excellent candidates for Late Woodland and protohistoric sand-tempered series near Fort Bragg. The “hardness trajectory” hypothesis may provide a way out of this typologically confusing situation. Yadkin Series Coe (1964:30-32) defined Yadkin series pottery from his excavations at the Doershuck Site. Abundant, angular crushed quartz was the diagnostic trait of the series, which was characterized by three main surface treatments, fabric impressed, cord marked, and linear check stamped. Subsequent investigations at Town Creek (Coe 1995) also associated simple stamped exteriors with the series. In the southern Sandhills, the definition of Yadkin has been expanded to include all types of crushed rock temper (Herbert 2003:58–59). Currently, examples of crushed granite, quartzite, feldspar, and quartz have been identified in assemblages in the area. The five sherds that correspond to the temper sub-variants of the Yadkin series recognized in the 31HK2502 collection exclusively exhibited extremely hard to vitrified pastes. Therefore, they have been assigned to the Yadkin III series variant (Table 13). Yadkin IIIe (crushed indurated granite), IIIc (crushed arenite) and IIIe/f (crushed indurated granite and feldspar sand) sub-variants were present. Consistent with the Hanover III and Cape Fear III series, cord marking continues to be an important or dominant surface treatment type (Figure 31:G-H). The cord marked sherd with a recognizable stamp pattern is cross or perpendicular stamped. In addition, one sherd displayed a smoothed or plain exterior surface (Figure 31:I). The mean wall thickness for the sample is 8.19 mm (SD = 0.74 mm). Even though paste hardness was not monitored in the Yadkin series previously and the Yadkin sample from 31HK2502 is exclusively composed of Yadkin III sherds, softer paste variants were recognized at 31HK2521 in the DO5 site package. This would suggest that the ceramic sequence of the region is complexly interlinked and issues of overlapping series contemporaneity are no doubt extant. OCCUPATION PATTERNS Diagnostic markers of culture-chronological association were relatively numerous in the 31HK2502 artifact assemblage. The dominance of lithic artifacts and Archaic tool types indicates that the greatest proportion of the occupation debris was deposited during the Archaic period. Recovered projectile points document Early Archaic (i.e., one Taylor Side-Notched, one Palmer II Corner Notched and one Palmer II/III Corner Notched blade fragment) and Middle Archaic (i.e., two Morrow Mountain II Stemmed points and one Guilford Lanceolate, Rounded Base variant). A Palmer II/III Corner Notched preform blank adds further to the Early Archaic representation. In addition, a wide range of scraper types and specialized unifacial tools evidence especially Early Archaic occupation, although some of the forms have also been recognized in Middle Archaic deposits. Finally, a broad range of ceramic series document Early Woodland (i.e., New River Series), Middle Woodland (i.e., Hanover I Series), early, Late Woodland (i.e., Hanover II Series) and possibly later Late Woodland (i.e., Hanover III, Cape Fear III and Yadkin III series) occupations. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 119 Chapter 5. 31HK2502 The site is characterized by relatively high artifact density, suggesting frequent short-term reoccupation patterns. Artifact density in positive Stage I and II shovel tests (n=69) equaled a mean of 3.92 artifacts (SD = 2.58), indicating that two-thirds of the positive tests yielded densities between one and seven artifacts. The modal positive shovel test outcome across the sample was three artifacts. Mean artifact density for the entire site area (including negative Stage I and II shovel tests) was 1.91 artifacts (SD = 2.66). Metavolcanic debitage recovered during Stage I and II shovel testing was widely distributed across the levee (Figure 32), while quartz debitage presented clusters of limited extent (Figure 33). Vertical distributions by raw material type indicate that debitage is generally concentrated between Levels 4 and 5, at depths of 30 and 50 cm bs (Table 14). All of the raw material types conform to this pattern with the exception of Uwharries Southern Rhyolite, which displays a bimodal distribution, peaking in Levels 2 (10-20 cm bs) and 4 (30 to 40 cm bs). From what we have learned about the vertical distributions of cultural-chronological occupations on sites in the Sandhills (see Cable 2010; Cable and Cantley 2005a, b, 2006), these profiles indicate that the great majority of the occupations represented at the site are associated with the Archaic period. This is consistent with the diagnostic projectile points recovered during the investigation, which represent Early and Middle Archaic types, and the scraper assemblage, which is predominantly associated with Early Archaic occupations. Confirmation for this inference is supplied by viewing the vertical distributions of these tool types in Table 15. Tool occurrences (from all proveniences) peak in Levels 4 and 5. Two of the Palmer Corner Notched points are out of position (too high), most likely due to bioturbation dislocation or localized deflation. In general, however, the coherence of diagnostic types conforms to the vertical coherence model of Sandhills cultural deposits. Precontact ceramics, as would be stipulated by the vertical coherence model, are positioned higher than the Archaic deposits (Table 16). Their frequencies peak in Level 2 (10-20 cm bs) and 88 percent of ceramic collection occurs between Levels 1 and 3 (0 to 30 cm bs). Sample sizes are too small to reliably comment on inter-series vertical positions. All of the series appear to peak in Levels 2 or 3. The Stage I and II shovel test sample indicates that ceramic concentrations are clustered and limited in extent, suggestive of short-term occupation (Figure 34). A more detailed appreciation for precontact land-use patterns evidenced at the site can be derived from a consideration of occupation type distributions and associations. Currently, four occupation types based on chipped stone criteria have been recognized on sites in the Fort Bragg area (see Cable and Cantley 2005c, 2006): (1) Type I occupations consist of a debitage concentration and an associated tool cluster situated at a confined location along its periphery. It is inferred that the tool cluster was formed around a hearth and in several instances calcined bone fragments have been recovered from the deposit delimited by a tool cluster. It is also believed that ephemeral shelters were situated adjacent to these tool clusters and opposite the concentrations. Debitage concentrations appear to represent lithic reduction loci established away from a shelter or sleeping area where tools were manufactured during the stay at the camp. These concentrations are generally composed of single lithic raw material types, and the associated tool clusters commonly consist of manufacturing rejects and discarded broken or worn-out tools made of the same raw material that comprises the associated debitage concentration. Although many Type I residences occur in isolation and appear to be the by-products from single nuclear or small extended family occupations, others appear to represent multi-family occupations that extend beyond the ability of a single shovel test to identify. (2) Type II residences are similar in spatial organization to Type I residences, but they exhibit much higher tool-to-debitage ratios and may commonly be comprised of diverse lithic raw materials. Lithic reduction on these sites consisted of tool maintenance and late stage shaping of tools from flake blanks or reduced biface cores. Consequently, debitage density is much lower than that of Type I debitage concentrations and the scatters consist primarily of late stage reduction debitage. Type II occupations are inferred to be the output from single or small, multiple household aggregations and are thought to represent the residue from High Technology Forager Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 120 Chapter 5. 31HK2502 620 610 70.5 71.0 600 Floodplain 590 Floodplain 580 71 .5 570 .0 72 560 31HK2502 550 . 540 530 71.5 Rock fi 70 .5 ek Cre sh 520 71 .0 71 .5 510 Floodplain 500 72 .0 71.5 490 72.5 480 73.0 470 Positive ST Negative ST 460 0 450 440 N 400 410 40 m 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 32. Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31HK2502 (Contours = 5 pieces of debitage). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 121 Chapter 5. 31HK2502 620 610 70.5 71.0 600 Floodplain 590 Floodplain 580 71 .5 570 .0 72 31HK2502 560 550 . 540 530 71.5 Rock fi 70 .5 ek Cre sh 520 71 .0 71 .5 510 Floodplain 500 72 .0 71.5 490 72.5 480 73.0 470 Positive ST Negative ST 460 0 450 440 N 400 410 40 m 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 33. Density Distribution of Quartz Debitage, Stage I and II Shovel Test Sample, Site 31HK2502 (Contours = 5 pieces of debitage). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 122 Chapter 5. 31HK2502 Table 14. Debitage Inventory for Stage I and II Shovel Tests, 31HK2502. DEBITAGE, STAGE I AND II SHOVEL TESTS:31HK2502 LEVELS RAW MATERIAL TYPE 1 2 3 4 1 1 1 1 2 3 1 Metasedimentary, gray Mill Mountain Rhyolite Type I Rhyolite Tuff 6 Uwharries Eastern Rhyolite 3 4 Uwharries Southern Rhyolite 2 8 1 9 1 4 7 13 15 7 13 4 17 11 8 3 14 15 17 1 Crystal Quartz 7 2 3 Uwharries Western Rhyolite 2 51 55 1 1 52 1 White Quartz GRAND TOTAL 5 GRAND TOTAL 6 1 7 6 11 13 3 28 34 63 58 21 40 1 3 214 Table 15. Diagnostic Stone Tool Inventory, All Proveniences, 31HK2502. DIAGNOSTIC STONE TOOLS, ALL PROVENIENCES:31HK2502 LEVELS TOOL TYPE 2 3 4 1 1 1 1 2 Guilford Lanceolate, Round Base Morrow Mt II Stemmed 5 Palmer II Corner Notched Palmer II/III Corner Notched 1 1 1 1 1 1 1 2 Type I b Pointed Scraper 1 Type I Side Scraper 1 Type II a Side Scraper 3 Type II b End Scraper 1 Type II b Side Scraper 1 2 9 2 3 1 1 Type III End Scraper 1 1 Type IV End Scraper 1 1 1 2 4 1 Denticulate 1 1 Graver/Burin 1 1 1 Spoke Shave 1 1 4 Type II c Side Scraper GRAND TOTAL 1 1 Taylor Side Notched Type III Side Scraper 7 1 Palmer II/III Corner Notched Preform Side Scraper Fragment 6 3 GRAND TOTAL 7 9 11 3 1 1 1 34 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 123 Chapter 5. 31HK2502 Table 16. Precontact Ceramic Inventory, All Proveniences, 31HK2502. PRECONTACT CERAMICS, ALL PROVENIENCES:31HK2502 LEVELS CERAMIC SERIES 1 2 Cape Fear III 3 4 1 1 5 2 3 2 Hanover II 12 30 14 2 Hanover III 2 5 4 2 1 2 8 3 4 16 43 (HTF) residences (Spiess 1984; Todd 1983:231-233). The HTF model proposes a specialized forager adaptation combining high logistical and high residential mobility into a single settlement system characterized by highly curated technologies. Type II residences have thus far only been assigned to the Early Archaic period at Fort Bragg. The lithic raw material diversity characteristic of Type II residences suggests that these occupations represent the aggregation of social units that may have been seasonally dispersed, transporting a wide range of lithic raw materials from different sources. (3) Type III occupations are inferred to represent logistical camps and very short-term residences characterized by low-density debitage scatters and sporadic tool discard. (4) Type IV occupations represent extraction or processing loci that are not believed to have involved, necessarily, overnight stays. They are difficult, however, to distinguish from Type III occupations, even in situations of greater excavation exposure. It is difficult to distinguish effectively between these various occupation types at shovel test intervals of greater than 1.25 m, but a less precise classification can be used at greater shovel test 1 1 22 63 13 1 Yadkin III GRAND TOTAL 6 2 Hanover I New River I GRAND TOTAL 5 5 3 3 92 intervals to aid in population projections. High-density debitage concentrations tend to indicate Type I residences, while low-density scatters tend to signal short-term residences, special purpose camps and locations (i.e. Type II, III and IV occupations). Highdensity debitage concentrations are temporarily stipulated to correspond to shovel test outcomes of five or more pieces of debitage of a specific raw material subtype. Eventually, it will be necessary to confirm this arbitrary threshold by examining shovel test counts from a large sample of known high-density debitage concentrations. Deposits containing less than five pieces of debitage are viewed as more likely to represent low-density debitage scatters associated with the other three site types. Since shovel tests can intersect the peripheral zones of high-density debitage concentrations where debitage densities are much lower, however, it is likely that Type I residential occupations will be under-represented in macrointerval shovel test data. Because it is difficult to determine whether a single occupation is represented in more than one shovel test, in the cases of multi-family residences or associated special activity zones on the periphery of a camp, the term “element” is preferred over the term “occupation” to refer to the occupational data in shovel tests. At a sampling interval of 5-m or 10-m intervals, we are assured that the presence of an ele- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 124 Chapter 5. 31HK2502 620 610 70.5 71.0 600 Floodplain 590 Floodplain 580 71 .5 570 .0 72 31HK2502 560 550 . 540 530 71.5 Rock fi 70 .5 ek Cre sh 520 71 .0 71 .5 510 Floodplain 500 72 .0 71.5 490 72.5 480 73.0 470 Positive ST Negative ST 460 0 450 440 N 400 410 40 m 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 34. Density Distribution of Precontact Sherds, Stage I and II Shovel Test Sample, Site 31HK2502 (Contours = 1 sherd). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 125 Chapter 5. 31HK2502 ment will not be repeated in adjacent shovel tests. Elements are identified by lithic raw material subtypes and/or by precontact sherd scatters associated with single vessels. This methodology runs the risk of overestimating occupations because more than one raw material subtype may comprise a single occupation. The term “element” recognizes this fact, focusing on depositional events and episodes rather than entire occupations. Elements can represent segments of contemporaneous multi-family occupations, separate occupations of various function, or contemporary members of multiple raw material reduction episodes. The current methodology for identifying elements in shovel tests consists of running cross-tabs of lithic raw material subtypes or subtype groups by shovel test number in a spreadsheet. Once the crosstab is generated, the sum of each lithic raw material subtype in a shovel test is identified as an element. In some instances, the vertical separation between individual items of the same lithic subtype can be so great as to suggest the presence of distinct elements. Due to the vagaries of bioturbation and possible recovery errors in which higher positioned material can be included in lower levels through sidewall displacement, however, the lower item(s) is assigned to the higher positioned item(s) to define a single occupation element Elements have horizontal dimensions and, as a result, sample units that are smaller than those dimensions have the benefit of increased sample efficiency (see Cable and Donaldson 1988; Rice 1987; Rice and Plog 1983). The typical element found at Fort Bragg has a diameter of about 3 m. Thus, a shovel test of .09 m2 samples only .09 percent of the area of a 10 m-square, but it samples 9.0 percent of the theoretical mean element space of 9.0 m2. Consequently, a 10 m-interval shovel test pattern actually samples about 9.0 percent of the available element space at a site, while a 5 m-interval shovel test pattern samples about 36.0 percent of the available element space. A mean estimate of the population of elements at shovel test intervals of 10 m and 5 m, then, requires a simple arithmetic calculation of counting the occupation elements identified in shovel tests and dividing this figure respectively by 9.0 percent and 36.0 percent. Although Stage I shovel test samples have been relied upon in the past to produce population projections because of their even and complete coverage, a concerted effort was made in the DO5 package to extend 5 m-interval shovel tests across the entire site area so that advantage could be taken of greater sample density. Following the procedures outlined above, the Stage I and II shovel test sample was successful in identifying six Type I residence elements, and 146 indeterminate (Type II, III and IV) elements (Table 17). The Uhwarries Southern Rhyolite sample was segregated into two raw material groups (i.e., the R3m and R3t groups as discussed in the raw material description section of the chapter). It can further be estimated that the site contains a mean of about 422 elements, of which approximately 17 represent Type I residential elements. Stone tools were counted as elements of a particular lithic raw material subtype when found in isolation from debitage in a shovel test. The proportional representation of Type I elements of 4.0 percent at 31HK2502 indicates that the site was not a preferred location from longer-term, or more intensive residential occupation. In general, the landform on which the site rests was used for special purpose camps and extraction loci for groups that rarely stayed longer than several days. Expanded over nearly the full range of human occupation (about 10,000 years), the population estimate indicates that precontact groups, on average, visited the site once every 24 years. This occupation, however, was probably not evenly distributed through time. The site was no doubt used intensively by specific culturehistoric groups, probably for strings of continuous years punctuated by abandonments in response to long-term land-use patterns dependent upon resource depletion cycles. The site appears to have been most intensively utilized during the Early and Middle Archaic periods, with a secondary emphasis during the Middle and Late Woodland periods. The fact that the site rests on a levee of limited extent immediately adjacent to a major Sandhills stream served to restrict choices for camp relocations at this spot, resulting in higher than normal occupation superimposition. Mean element occurrence in individual Stage I and II shovel tests was 2.24 elements (SD = 1.11). Seventy-two percent of the shovel tests in this sample contained two or more elements, and 35 percent of these contained three or Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 126 Chapter 5. 31HK2502 Table 17. Recognized Elements for the Stage I and II Shovel Test Sample, 31HK2502. RECOGNIZED ELEMENTS FROM THE STAGE I AND II SHOVEL TEST SAMPLE:31HK2502 RAW MATERIAL TYPES TYPE I TYPE II, III, IV GRAND TOTAL PERCENTAGE OF ELEMENTS Quartz Crystal 0 1 1 0.7 Metasedimentary, Gray 0 1 1 0.7 Mill Mountain Rhyolite 0 10 10 6.6 Type I Rhyolite Tuff 0 7 7 4.6 Uwharries Eastern Rhyolite 1 28 29 19.1 Uwharries Southern Rhyolite (1) 0 21 21 13.8 Uwharries Southern Rhyolite (2) 1 26 27 17.8 Uwharries Western Rhyolite 2 28 30 19.7 White Quartz 2 24 26 17.1 GRAND TOTAL 6 146 152 100.0 more elements. Ten percent of the shovel tests actually yielded four to six elements. Although there is not a one-to-one correspondence between elements and individual occupations, the data clearly point to a high level of deposit superimposition. The impact of superimposition, however, may be slightly mitigated by the potential for an unusual concentration of Early Archaic Type II residences, which are characterized by high lithic raw material variability. This issue will be further developed in the next section, which examines the results of the Stage III Sample Block investigations. Not included in the population estimate was the Woodland period material. Following the procedure for population estimates outlined above for precontact ceramic vessels, it is projected that a mean of 39 Woodland period sherd scatters/occupation elements are present at the site. Since Woodland period occupations generally contain paired lithic debitage scatters and sherd clusters, it is likely that some of the elements identified in the lithic subtype tally are Woodland in association. Thus to avoid the issue of double counting, ceramics were not included in the element population estimate presented above. It is roughly estimated from this calculation, however, that about 24 percent of the occupation elements represented at 31HK2502 are affiliated with Woodland period occupations. The Stage I and II shovel test sample succeeded in identifying three Hanover I, seven Hanover II, three Hanover III and one Yadkin III elements. It is inferred from these data that about 21 percent of the Woodland occupation of the site is affiliated with the early Middle Woodland (Hanover I), approximately 50 percent occurred during the Middle to Late Woodland transition (Hanover II) and about 29 percent was associated with the later Late Woodland period (Hanover III and Yadkin III). Approximately 84 percent of the lithic debitage in the Stage I and II samples was recovered between 20 and 60 cm bs (Table 14), which is generally consistent with sites in the Sandhills of Fort Bragg that contain predominantly Archaic period occupations (Cable and Cantley 2005a, b, 2006). The presence of a relatively large sample of precontact ceramics, however, suggests that a good proportion of the debitage situated in Level 3 at 20 to 30 cm bs is affiliated with Woodland occupations. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 127 Chapter 5. 31HK2502 Numerous studies have documented vertical patterning in the relative depth of deposits in the unconsolidated sandy deposits of the Coastal Plain (Cable and Cantley 2005 b, 2006; Michie 1990). The actual depth patterns vary according to the character of the deposit. Those deposits with higher B-horizons tend to have compressed sequences, while those in deep C-horizon and E-Horizon sediments have more discreet and thicker sequencing. Michie’s (1990) C-horizon model developed from the Waccamaw Neck area on the central coast of South Carolina begins with Mississippian and Woodland materials in the upper 30 cm of sediment. This zone is underlain by ceramic Late Archaic occupations situated at 28 to 35 cm bs, Middle Archaic horizons at 35 to 55 cm bs, and Early Archaic corner-notched and side-notched components at 55 to 62 cm bs. This is generally a good model for Fort Bragg. However, the deposits at Fort Bragg appear to be somewhat more compressed, judging by the vertical positions of samples of diagnostic artifacts. At Fort Bragg three vertical groupings have been developed for deep Sandhills sites: (1) Late Archaic/Middle Archaic between 15 and 40 cm bs, (2) Middle Archaic/Early Archaic between 35 and 50 cm bs, and (3) Early Archaic below 50 cm bs. The divisions are broadly constructed and realistically anticipate a certain degree of vertical mixing and overlap between the various Archaic periods. This vertical model can be of great utility in reconstructing the occupation history of sites like 31HK2502 where the predominant artifact class is non-diagnostic lithic chipped stone debitage. The vertical positions of the various elements recognized in the Stage I and II shovel test sample provide a rough picture of the representation of occupation periods at the site (Table 18). Most elements were situated in a single level, but some, especially the higher density elements, spanned several levels. In the latter cases, the central tendency of the vertical distribution of a particular element was used to estimate its position. Vertical data from the shovel test sample was not precisely congruent with the model ranges discussed above. However, the following adaptation was developed to correlate culture-historic association with level data: (1) 0-20 cm bs, Woodland, (2) 20-40 cm bs, Woodland/Late Archaic/Middle Archaic, (3) 40-50 cm bs, Middle Archaic/Early Archaic and (4) 50 to 70 cm bs, Early Archaic. Approximately 48.7 percent of the occupation elements are positioned between 20 and 40 cm bs, indicating that Middle Archaic phases are an important co-dominant. Some of these elements are no doubt affiliated with Woodland occupations. Based on the overall representation of 24 percent, it is estimated that approximately 20 percent of the elements found between 20 and 40 cm bs are affiliated with the Woodland period. Only Table 18. Vertical Distribution of Recognized Elements, 31HK2502. 74 48.7 Middle/Early Archaic 2 2 4 3 9 10 6 36 23.7 Early Archaic/Paleoindian 1 2 5 5 4 3 2 22 14.4 10 7 29 21 27 30 26 152 100.0 GRAND TOTAL 1 1 PERCENTAGE 13.2 13 GRAND TOTAL 20 16 2 WHITE QUARTZ 5 12 1 UWHARRIES WESTERN RHYOLITE UWHARRIES SOUTHERN RHYOLITE (1) 1 8 Woodland/Late Archaic/Middle Archaic UWHARRIES SOUTHERN RHYOLITE (2) UWHARRIES EASTERN RHYOLITE 2 15 1 Woodland MILL MOUNTAIN RHYOLITE 5 3 METASEDIMENTARY, GRAY 5 5 INFERRED ASSOCIATION QUARTZ CRYSTAL TYPE I RHYOLITE TUFF VERTICAL DISTRIBUTION OF RECOGNIZED ELEMENTS:31HK2502 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 128 Chapter 5. 31HK2502 meager evidence (i.e., three steatite bowl fragments from a single, Stage III shovel test) of a Late Archaic presence was produced during the investigation, suggesting that the Late Archaic contribution to the occupation history of the site was minimal. About 14.4 percent of the occupation is unambiguously Early Archaic, but much of the occupation positioned between 40 and 50 cm bs is most likely also affiliated with Early Archaic phases. Thus, it is estimated that Early Archaic occupation is represented by approximately 30 to 35 percent of the elements. The upper 20 cm of deposit contains about 13.2 percent of the recognized elements and it is inferred that most of these are affiliated with Woodland occupation. SAMPLE BLOCK INVESTIGATIONS Nineteen shovel test outcomes from the Stage I and II sample were targeted for close-interval shovel testing to further elaborate the character of the occupations extant at the site. In general, a series of eight additional shovel tests were excavated at 1.25 m-intervals surrounding the target shovel test. These are referred to as sample blocks and their locations are illustrated in Figure 22. Sample Block (SB) locations were evenly distributed across the site to achieve a representative picture of the settlement record. Tables 19 through 22 tabulate the vertical distributions of debitage by raw material type for each of the sample blocks. Nearly all of the sample blocks yielded debitage profiles concentrated between 20 and 50 cm bs, in what would be primarily considered the Archaic zone of the site. Projectile point and scraper distributions (Table 23) indicate that Early Archaic occupation is most likely predominant in SB1, SB4, SB6, SB7, SB12, SB15, SB16 and SB19. Middle Archaic occupation is confirmed in only one of the sample blocks, SB3, where two Morrow Mountain II Stemmed and a Guilford Lanceolate, Round Base variant were recovered between 20 and 40 cm bs. It is likely, however, that Middle Archaic occupation is dominant in a number of the sample blocks lacking diagnostic tools. Two sample blocks (SB13 and SB18) yielded debitage concentrations in the upper 30 cm of deposit (see Tables 21 and 22). Ceramic distributions indicate that much of the debitage concentration in SB18 is most likely associated with Woodland occupation (see Table 24). Most of the debitage is concentrated below the ceramic deposits across the sample blocks. An exception is SB5 where the light debitage scatter is vertically congruent with Hanover and Yadkin series ceramics. The sample block investigation succeeded in demonstrating that ceramics are scattered across the entire span of the levee. The wide range of series, however, suggests that Woodland occupation was sporadic and short-term. The Stage I and II sample yielded evidence of 26 separate ceramic pots. From this it can be projected that Woodland deposit at the site contains a mean of 163 individual vessels, given a pot scatter element size of 4 m2. Evidence of whole pot concentrations was not found during the investigation, suggesting that a great deal of displacement and fragmentation occurred during the Woodland occupation. It is likely that this displacement has two sources, post-depositional disturbance from subsequent Woodland occupations and large sherd recycling for other uses once the original pot was broken. Since 49 individual vessels were identified just in the site collection sample, however, it is likely that this mean estimate under-represents the actual number of vessels or vessel fragments that were deposited on the site. Higher incidences of fragmentation and displacement will confound attempts to apply the “element space” calculation to estimate vessel populations because the average element size is invalidated. An idea of the magnitude of the sherd population at the site can be supplied by calculating a straight area mean. The sample of 142 Stage I and II shovel tests sums to an area of 12.78 m2. The site area is approximately 2,914 m2, which yields an area sample of 0.44 percent. Twenty-six sherds, weighing 100.59 gm, were recovered from the sample, resulting in a population mean estimate of sherds of 5,909 and a total sherd weight of 22.86 kg. If whole pots were entirely represented in the assemblage, it is doubtful that this total would be sufficient to cover the weight of only 10 to 20 typical open-mouth jars from the region (see Herbert and Irwin 2003:2). Given an average weight of 3 pounds (1,377 gm) for a typical pot, if there were 163 whole pots present on the site, a mean population weight of 223 kg would be required. This suggests that the intensity of pot recycling is great on Sandhills sites and that large sherds Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 129 Chapter 5. 31HK2502 Table 19. Vertical Distribution of Debitage, Sample Blocks 1—5, 31HK2502. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 1-5:31HK2502 SB1 RAW MATERIAL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL LEVEL 1 2 3 4 5 3 1 1 2 1 1 0 0 2 6 11 1 2 3 4 SB2 RAW MATERIAL 1 2 5 8 1 2 0 5 6 GRAND TOTAL 5 1 3 1 5 7 10 32 GRAND TOTAL 1 1 1 2 8 13 2 2 LEVEL 1 2 3 4 5 5 7 1 2 10 7 27 2 1 1 2 2 8 1 1 1 2 1 2 3 4 1 13 6 7 GRAND TOTAL 14 3 7 16 12 52 LEVEL 1 2 3 4 5 GRAND TOTAL 2 2 2 4 5 3 18 6 2 2 2 1 1 1 1 3 3 4 8 SB5 Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 2 2 1 Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL RAW MATERIAL 0 1 SB4 RAW MATERIAL 5 4 2 13 1 SB3 Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 8 LEVEL Quartzite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Western Rhyolite White Quartz GRAND TOTAL RAW MATERIAL 7 1 1 6 6 1 1 4 1 1 5 6 LEVEL 1 2 3 1 1 1 1 4 4 1 1 2 1 1 1 1 7 GRAND TOTAL 1 2 1 3 1 8 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 130 Chapter 5. 31HK2502 Table 20. Vertical Distribution of Debitage, Sample Blocks 6—10, 31HK2502. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 6-10:31HK2502 SB6 LEVEL RAW MATERIAL Orthoquartzite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 2 3 1 2 5 1 2 4 2 1 8 SB7 1 1 1 4 1 1 5 1 5 3 1 1 6 2 6 1 1 10 1 2 15 1 2 3 4 5 6 1 3 4 5 2 19 2 21 4 1 1 4 2 1 14 2 2 25 1 1 5 1 7 8 7 8 GRAND TOTAL 1 4 21 1 7 34 5 1 6 33 1 7 1 1 5 6 7 5 6 7 GRAND TOTAL 4 53 5 7 5 12 86 LEVEL 1 2 Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite GRAND TOTAL 3 1 1 2 SB10 Brown Chert Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 6 1 2 1 1 1 3 9 3 SB9 RAW MATERIAL 5 GRAND TOTAL 1 10 1 1 3 6 22 LEVEL RAW MATERIAL RAW MATERIAL 6 2 SB8 Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Western Rhyolite White Quartz GRAND TOTAL 5 LEVEL RAW MATERIAL Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 4 4 1 1 1 3 8 GRAND TOTAL 1 2 1 1 5 LEVEL 1 2 3 4 1 5 4 1 1 6 1 1 1 6 1 2 8 1 1 1 9 2 GRAND TOTAL 1 1 15 2 2 3 2 26 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 131 Chapter 5. 31HK2502 Table 21. Vertical Distribution of Debitage, Sample Blocks 11—15, 31HK2502. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 11-15:31HK2502 SB11 LEVEL RAW MATERIAL 1 Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 2 3 1 1 3 4 3 1 3 4 14 4 1 1 1 11 1 4 19 SB12 1 Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 4 6 3 4 5 1 1 5 1 5 2 6 4 10 1 4 1 20 1 2 3 4 5 2 3 2 8 2 1 16 4 1 7 3 3 1 1 3 14 4 1 SB14 6 7 8 2 7 GRAND TOTAL 15 12 1 4 3 35 GRAND TOTAL 9 2 20 6 1 38 6 LEVEL RAW MATERIAL 1 1 1 2 3 4 1 5 1 2 4 1 3 13 10 1 6 4 5 26 2 3 4 SB15 Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 6 1 1 9 LEVEL RAW MATERIAL RAW MATERIAL 1 2 SB13 Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL GRAND TOTAL 2 5 7 25 2 10 51 6 LEVEL RAW MATERIAL Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 5 5 6 1 7 2 2 3 1 14 1 5 6 7 1 4 1 1 GRAND TOTAL 1 26 1 13 8 12 61 7 2 5 LEVEL 1 1 1 3 6 2 1 1 2 15 8 1 7 7 8 GRAND TOTAL 3 18 4 2 1 3 31 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 132 Chapter 5. 31HK2502 Table 22. Vertical Distribution of Debitage, Sample Blocks 16—19, 31HK2502. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 16-19:31HK2502 SB16 LEVEL RAW MATERIAL Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz Grand Total 1 1 2 3 4 5 6 3 8 16 3 5 1 1 26 8 2 11 2 5 26 2 3 7 5 6 1 1 1 4 2 1 10 SB17 1 indet. metavolcanic Metasedimentary, green Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Western Rhyolite White Quartz Grand Total 2 3 4 1 1 1 1 2 4 2 1 2 GRAND TOTAL 38 3 22 4 11 78 GRAND TOTAL 1 1 1 8 1 8 4 24 1 2 4 7 3 1 1 2 9 1 2 5 6 1 LEVEL RAW MATERIAL 1 2/3 3 4 11 12 1 20 5 11 3 23 1 27 1 15 8 4 4 3 1 20 2 3 SB19 Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz Grand Total 2 1 1 1 SB18 RAW MATERIAL 8 LEVEL RAW MATERIAL Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz Grand Total 7 GRAND TOTAL 2 52 25 4 4 6 93 1 2 1 1 3 8 LEVEL 1 4 5 6 1 1 1 2 1 1 4 8 4 2 1 8 1 5 4 1 11 2 2 7 8 GRAND TOTAL 1 1 2 13 6 7 30 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 133 Chapter 5. 31HK2502 Table 23. Vertical Distribution of Stone Tools by Sample Block, 31HK2502. VERTICAL DISTRIBUTION OF STONE TOOLS IN SAMPLE BLOCKS:31HK2502 STONE TOOL TYPE LEVELS 1 2 3 4 5 6 7 8 GRAND TOTAL SB1 Palmer II Corner Notched 1 1 Taylor Side Notched 1 1 Type II a Side Scraper 1 1 Type II c Side Scraper 1 1 SB2 Core Fragment 1 Type I Biface 1 1 1 Utilized Flake 1 1 SB3 Guilford Lanceolate, Round Base Morrow Mt II Stemmed 1 Projectile Point, Upper Blade 1 1 1 2 1 1 SB4 Core Fragment 1 1 Type I b Pointed Scraper 1 Type II Biface Fragment 1 Utilized Flake 1 1 1 1 SB6 Denticulate 1 Preform (Palmer II/III) 1 1 Utilized Flake 1 1 1 SB7 Flake Blank 2 2 Graver/Burin 1 1 Type I Biface Fragment 1 1 Type II b Side Scraper 1 1 Type II Biface Fragment 1 1 SB8 Core Fragment 1 Flake Blank Side Scraper Fragment 1 2 1 1 1 1 Slab/Rock 1 Utilized Flake 1 SB9 Type II Biface Fragment 1 1 3 1 1 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 134 Chapter 5. 31HK2502 Table 23. Vertical Distribution of Stone Tools by Sample Block, 31HK2502 (Continued). VERTICAL DISTRIBUTION OF STONE TOOLS IN SAMPLE BLOCKS:31HK2502 STONE TOOL TYPE LEVELS 1 2 3 4 5 6 7 8 GRAND TOTAL SB10 Side Scraper Fragment 1 1 Spoke Shave 1 Type III Biface Fragment 1 1 1 Utilized Flake 1 1 SB11 Soapstone Vessel Sherd 3 Type II a Side Scraper 3 1 Type III Biface Fragment 1 1 Utilized Flake 1 1 1 SB12 Core Fragment 1 1 Type II a Side Scraper 1 1 SB13 Biface Fragment 1 Core Fragment 1 Flake Blank 1 1 1 1 Type II Biface Fragment 1 1 Type III Side Scraper 1 1 SB14 Anvil Stone 1 1 Unidirectional Core Fragment 1 1 SB15 Type III Side Scraper 2 Utilized Flake 2 1 1 SB16 Flake Blank 1 Palmer II/III Corner Notched 1 1 1 Projectile Point, Tip 1 1 SB17 Projectile Point, Lower Blade 1 Type III Biface Fragment 1 1 1 SB18 Utilized Flake 1 1 SB19 Type II b Side Scraper 1 Utilized Flake GRAND TOTAL 1 1 3 8 12 1 21 15 3 1 1 61 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 135 Chapter 5. 31HK2502 Table 24. Vertical Distribution of Precontact Ceramics by Sample Block, 31HK2502. VERTICAL DISTRIBUTION OF PRECONTACT CERAMICS IN SAMPLE BLOCKS: 31HK2502 LEVEL CERAMIC TYPE SB1 Hanover Id Cord Marked Hanover IIIa Fabric Impressed SB2 Hanover IIb indeterminate decorated SB3 Hanover Ib Check Stamped Hanover Ib indeterminate decorated Hanover IIa Fabric Impressed Hanover IIb indeterminate decorated Residual Sherd SB4 Hanover IIa Cord Marked Hanover IIa Fabric Impressed Hanover IIb Fabric Impressed Hanover IIIa indeterminate SB5 Hanover IIa Fabric Impressed Yadkin IIIc Cord Marked SB7 Hanover IIa Fabric Impressed Hanover IIa Plain Hanover IIb Cord Marked Hanover IIIf indeterminate decorated SB11 Hanover IIa Fabric Impressed Hanover IIa indeterminate decorated SB12 Hanover IIb Fabric Impressed SB14 Hanover Ib Check Stamped Hanover Ib indeterminate decorated SB16 Hanover Ib Fabric Impressed Hanover IIa Fabric Impressed Hanover IId Fabric Impressed Hanover IId indeterminate Hanover IIIa indeterminate decorated SB17 Hanover IIIa indeterminate decorated SB18 Cape Fear III Cord Marked Hanover IIa Fabric Impressed Hanover IIa Indeterminate Hanover IIa indeterminate decorated Hanover IIIa Cord Marked SB19 New River Ia indeterminate decorated Yadkin IIIe indeterminate Yadkin IIIe Plain GRAND TOTAL 1 2 3 4 1 5 6 7 8 1 2 1 1 1 1 1 1 1 1 4 1 1 1 2 1 1 1 1 2 1 1 5 1 1 1 1 1 2 1 1 1 2 2 1 1 1 1 1 1 1 1 1 1 1 1 3 1 2 2 1 1 1 1 2 1 4 1 1 1 1 1 1 1 1 1 2 1 3 2 1 1 1 2 4 3 2 1 1 1 1 1 1 1 1 13 GRAND TOTAL 24 13 5 1 3 0 0 59 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 136 Chapter 5. 31HK2502 from broken pots were probably curated and stored in caches or transported upon leaving camps. All of the sample blocks contain a wide variety of lithic raw material types. Much of this heterogeneity can be attributed to occupation superimposition, but some of it may result from site functional variability. Type II residences characteristically contain heterogeneous lithic raw material profiles and the high density of scraper forms in the stone tool inventory suggests that many of the occupations at the site may represent Early Archaic Type II residences. The sample block frames generally cover an area of only about 2.5 m square, which is smaller than the average element size. Moreover, it would be uncommon for the sample block to expose nearly whole elements. However, a method was devised during the DO4 investigation (Cable 2010) to identify and distinguish Type I and III elements that are mostly contained within the extent of a sample block. Three measures related to shovel test outcomes were developed to describe this variability (Table 25). These are: (1) number of positive shovel tests within which the element occurs, (2) number of shovel tests with greater than or equal to 5 pieces of debitage and (3) mean debitage frequency per positive test. Since Type I residential elements are characterized by high density debitage concentrations they are expected to yield high outcome values for each of these measures. Type III residences, by contrast, are defined by the presence of low-density debitage scatters of about the same size as Type I high density scatters. Consequently, they should be expected to measure high or moderate on the number of positive shovel tests within which they are represented, but low on the number of shovel tests yielding five or more pieces of debitage of a specific raw material type and low on the mean frequency of debitage per positive shovel test. Given the structure of the sample blocks, there are generally nine outcomes (shovel tests) to measure. Some elements present measures that are intermediate between the two extremes and these can only be classified as Type I/III. Type II residential elements would be expected to reflect similar patterns to Type III residences, but they should contain multiple raw material types and a greater density of tools. Clearly, the identification of Type II residences in sample blocks is subjective and it would require additional confirmation from expanded sample blocks or test unit excavation to firmly identify such elements. Table 26 summarizes the result of applying the shovel test model to the elements defined by lithic raw material type within each sample block. Generally, only elements that were represented by at least three shovel tests were analyzed, as those with lower representation are regarded as insufficiently complete or ineffectively sampled in the sample block frame. Two exceptions were made, however, because of high debitage count yields in limited numbers of shovel tests (i.e. a UER element in SB10 and a UWR element in SB17). As established by the Stage I and II element study, Type III elements predominate in the sample blocks. Seventy-four percent (n = 40) of the elements were classified as Type III, short-term residences. These elements are characterized by vari- Table 25. Shovel Test Outcome Model for the Purpose of Identifying Element Types. INFERRED ELEMENT TYPE NO. POSITIVE STPS NO. STPS WITH ≥ 5 DEBITAGE MEAN DEBITAGE/ POSITIVE TEST TYPE I HIGH HIGH HIGH TYPE I OR TYPE III HIGH MODERATE MODERATE TYPE III MODERATE LOW LOW Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 137 Chapter 5. 31HK2502 Table 26. Inferred Element Types by Sample Blocks, 31HK2502. MEAN DEBITAGE/ + SHOVEL TEST 3 0 1.67 III LA/MA Uwharries Eastern Rhyolite 3 0 1.00 III EA Type II Uwharries Southern Rhyolite (2) 3 0 1.67 III EA Type II Palmer II CN Uwharries Western Rhyolite 5 0 1.40 III EA Type II Taylor SN, Type IIc Side Scraper White Quartz 3 1 3.33 I/III MA/EA Type II Type IIa Side Scraper 5 0 1.60 III LA/MA TYPE II CANDIDATE NO. STs WITH ≥ 5 DEBITAGE Mill Mountain Rhyolite ELEMENTS ASSOCIATION NO. OF STs REPRESENTED INFERRED ELEMENT TYPE INFERRED ELEMENT TYPES IDENTIFIED IN SAMPLE BLOCKS: 31HK2502 STONE TOOLS SB1 SB2 White Quartz Core Fragment Type I Rhyolite Tuff Utilized Flake Uwharries Southern Rhyolite (1) Type I Biface Meta-sedimentary Grooved Abrader SB3 Uwharries Eastern Rhyolite 6 0 2.33 III MA Uwharries Southern Rhyolite (2) 3 0 2.33 III MA Uwharries Western Rhyolite 6 1 2.67 III MA White Quartz 5 0 2.40 III MA Morrow Mountain II Stemmed, PPK Upper Blade Meta-sedimentary Morrow Mountain II Stemmed Mill Mountain Rhyolite Guilford Lanceolate SB4 Uwharries Southern Rhyolite (2) 3 0 1.33 III LA/MA Type II Biface frag. Uwharries Western Rhyolite 5 0 1.00 III EA Type Ib Pointed Scraper White Quartz 3 0 1.00 III LA/MA Core frag. Type I Rhyolite Tuff Utilized Flake SB5 Uwharries Southern Rhyolite (2) 3 0 1.00 III LA/MA Uwharries Eastern Rhyolite 4 1 2.50 III EA Type II Preform (Palmer II/III) Uwharries Western Rhyolite 3 0 1.00 III EA Type II Denticulate White Quartz 3 0 2.00 III EA Type II SB6 Uwharries Southern Rhyolite (1) Utilized Flake SB7 Uwharries Eastern Rhyolite 3 0 1.33 III LA/MA Uwharries Southern Rhyolite (2) 6 1 3.50 I/III EA White Quartz 3 0 2.33 III W/LA Type I Rhyolite Tuff Crystal Quartz Flake Blank Type II Flake Blank, Graver/Burin, Type II Biface frag. Type II Type IIb Side Scraper Type I Biface frag. SB8 Type I Rhyolite Tuff 3 3 17.67 I EA Utilized Flake Uwharries Eastern Rhyolite 3 0 1.67 III LA/MA Flake Blank Uwharries Southern Rhyolite (1) 4 0 1.75 III EA Side Scraper frag. White Quartz 6 1 2.00 III MA/EA Core frag. (2) Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 138 Chapter 5. 31HK2502 Table 26. Inferred Element Types by Sample Blocks, 31HK2502 (Continued). MEAN DEBITAGE/ + SHOVEL TEST INFERRED ELEMENT TYPE ASSOCIATION 2 1 7.50 I/III LA/MA TYPE II CANDIDATE NO. STs WITH ≥ 5 DEBITAGE ELEMENTS NO. OF STs REPRESENTED INFERRED ELEMENT TYPES IDENTIFIED IN SAMPLE BLOCKS: 31HK2502 STONE TOOLS SB10 Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Type II Spoke Shave, Type III Biface frag., Utilized Flake Uwharries Western Rhyolite Type II Side Scraper frag. SB11 Uwharries Eastern Rhyolite 5 0 1.00 III LA Uwharries Southern Rhyolite (1) 4 0 1.75 III MA/EA Uwharries Southern Rhyolite (2) 8 2 3.13 I White Quartz 5 0 2.00 III Type Ib Side Scraper, Type III Biface frag., Utilized Flake EA LA/MA Steatite Bowl frag. (3) SB12 Mill Mountain Rhyolite 2 2 7.50 I MA/EA Uwharries Eastern Rhyolite 5 1 2.40 III MA/EA White Quartz 3 0 1.00 III MA/EA Core frag. Uwharries Southern Rhyolite (1) Type IIa Side Scraper SB13 Uwharries Eastern Rhyolite 4 0 2.25 III Uwharries Southern Rhyolite (2) 4 3 5.00 I EA W/LA Core frag., Type III Side Scraper Flake Blank Uwharries Western Rhyolite 3 0 2.00 III LA/MA Biface frag. White Quartz Type II Biface frag. SB14 Uwharries Eastern Rhyolite 6 2 4.33 I MA/EA Uwharries Southern Rhyolite (2) 6 1 2.17 III MA/EA Type II Type II Uwharries Western Rhyolite 4 0 2.00 III MA/EA Type II White Quartz 5 0 2.40 III MA/EA Type II Grano-Diorite Unidirectional Core frag. Anvil Stone SB15 Uwharries Eastern Rhyolite 7 1 2.57 III EA Type II Type III Side Scraper Uwharries Southern Rhyolite (1) 3 0 1.33 III EA Type II Type III Side Scraper, Utilized Flake Uwharries Eastern Rhyolite 8 4 4.75 I EA Uwharries Southern Rhyolite (2) 6 1 3.67 I/III EA Type II Uwharries Western Rhyolite 3 0 1.33 III EA Type II White Quartz 4 1 2.75 III EA Type II Uwharries Eastern Rhyolite 4 0 2.00 III MA/EA Uwharries Western Rhyolite 2 1 4.00 I/III LA/MA White Quartz 3 0 1.33 III LA/MA Uwharries Eastern Rhyolite 9 3 5.78 I W/LA Uwharries Southern Rhyolite (1) 3 2 8.33 I White Quartz 5 0 1.20 III SB16 Palmer II/III Corner Notched, Flake Blank PPK Tip SB17 PPK Blade frag., Type III Biface frag. SB18 W/LA Utilized Flake MA/EA SB19 Uwharries Southern Rhyolite (2) 6 1 2.17 III MA/EA Type II Uwharries Western Rhyolite 5 0 1.20 III EA Type II Type IIb Side Scraper, Utilized Flake Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 139 Chapter 5. 31HK2502 ably moderate to high shovel test representation (3 to 6 shovel tests), low frequencies of shovel tests containing greater than 5 pieces of debitage (0 to 1 shovel test) and debitage frequency means in positive shovel tests of less than 3.0. Type I elements were more numerous than expected, numbering eight of the 54 elements (15 percent) recognized for this study. These elements commonly had high shovel test representation (4 to 9 shovel tests), greater numbers of high debitage density shovel test outcomes (2 to 4 shovel tests) and high debitage mean densities, which ranged between 3.13 and 17.67 pieces of debitage. Ten percent (n = 6) of the elements exhibited intermediate characteristics and could not be differentiated. Cultural-chronological associations were inferred from depth profiles and from diagnostic stone tools that were matched to the raw material composition of the debitage concentrations. Sometimes diagnostic stone tool associations supplanted inferences based on depth patterns. In particular, some elements that occupied high vertical positions were assigned to the Early Archaic period based on diagnostic stone tool associations. This adjustment was found to be necessary in SB 6 and SB13. Nineteen Early Archaic, 13 Middle Archaic/Early Archaic, 4 Middle Archaic, 14 Late Archaic/Middle Archaic and 4 Woodland/ Late Archaic elements were identified in the study. Candidates for Type II elements were recognized in SB1, SB6, SB7, SB10, SB14, SB15, SB16 and SB 19 (see Table 26). In these instances, groups of Type III elements could hypothetically belong to single Type II residences. Since such elements have only been positively associated with the Early Archaic period, only elements of this inferred affiliation were considered for this procedure. Type II residences are characterized by low-density scatters of debitage composed of multiple raw material types and relatively high densities of stone tools of heterogeneous raw material composition. It is suggested that these elements represent multiple family units that came together for short durations to conduct activities such as communal hunts that resulted in abnormally high tool discard rates related to processing animals. Each of the identified sample blocks meets these criteria. High numbers of tools are found with multiple “Type III” debitage scatters of a depth or association to be consistent with an Early Archaic affiliation. SB16 contains what appears to be a Type II candidate and a Type I element containing a highdensity UER debitage concentration and an associated Palmer II/III Corner Notched blade fragment. The Type I element appears to represent a separate occupation event, as its vertical debitage profile is higher than the other elements (see Table 22). SB1 contains at least two distinct Early Archaic occupations (i.e. Taylor and Palmer II phases), but there are other Type III elements present that could belong to either occupation. Type II elements are rare and not well understood in the region. The possibility that these elements are present in unusually high frequencies at 31HK2502 is further evaluated in the test unit investigation to follow. TEST UNITS At the time of fieldwork, the character of the Early Archaic occupation was not fully appreciated, as many of the scrapers were not recognized until the analysis phase. Nevertheless, deposits were targeted that provided an opportunity to evaluate the possibility that Type II elements exist at 31HK2502. The number of test units would have been increased, had the potentially high Type II element frequency been realized at the time of fieldwork. Three 1-x-1-m test units were excavated at 31HK2502 (Table 27). Each unit was quartered and excavated in arbitrary 10 cm levels and each was taken down to approximately 5 cm below contact between the E2-horizon and the underlying substrate. The precise location of each test unit was determined after close-interval shovel testing was deployed around targeted shovel tests. The test units were placed in SB1 (TU3), SB3 (TU2) and SB14 (TU1), all on the southeast end of the site (Figure 23). SB3 contained at least two Middle Archaic occupations represented by Guilford Lanceolate and Morrow Mountain II Stemmed projectile points. SB1 contained portions of at least two Early Archaic occupations recognized by the presence of a Taylor Side Notched point and a Palmer II Corner Notched point. Each was associated with a Type III debitage scatter of matching raw material type and two other Type III elements of relative contemporaneity were identified, suggesting that at least one and possibly two Type II residences Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 140 Chapter 5. 31HK2502 Table 27. Summary Data for Test Units, 31HK2502. SUMMARY OF TEST UNITS: 31HK2502 TEST UNIT SIZE NORTH EAST NO. LEVELS MAX. DEPTH (cm bd) VOLUME 3 (m ) ARTIFACT DEPTH (cm bd) TU1 1x1m 495.5 500.5 6 60 0.60 50 TU2 1x1m 502.0 501.0 6 60 0.60 60 TU3 1x1m 509.0 488.0 7 70 0.70 70 were present. Finally, SB14 also contained a number of Type III elements of probable Early Archaic age that are also thought, in hindsight, to be viable Type II candidates. The results for each test unit are presented below. Test Unit 1 (SB 14) TU1 was placed at grid coordinates N495.5/ E500.5 (NE corner) in the southeast quadrant of SB14. Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of the unit (Figure 35). Four soil layers were exposed in the wall profiles, not including the upper humic zone of the A-horizon which was charcoal laden from prescribed burning activities. The A-horizon was positioned in the upper 10 to 12 cm of of the solum. It was composed of grayish brown (10YR 5/2) sandy loam. Its contact with the underlying E-horizon was irregular and faint, suggesting that it represents a natural soil horizon free of major disturbances such as plowing. The fine sand E-horizon transitioned from light yellowish brown (10YR 6/4) to pale brown (10YR 7/3) at about 28 to 30 cm bd. The upper zone is identified as the E1-horizon, while the lower zone is referred to as an E2-horizon. The brownish yellow (7.5YR 6/6) fine sandy loam substrate displays a sharp transition from the E2-horizon, suggesting that the latter lies unconformably above it. The loose substrate is identified as a B-horizon, but it is not typical of the Bt-horizon that characterizes the Blaney sand pedon. The origin of the substrate may be alluvial. SB14 yielded substantial evidence of three Type III elements and one Type I element within a vertical zone suggestive of Middle Archaic or Early Archaic occupation (Table 26). Diagnostic tools were not recovered, but a unidirectional core fragment of UWR and a large grano-diorite anvil stone were found between 30 and 40 cm bs. TU 1 was placed in a location that was designed to intersect the edges of the identified elements to increase the chances of finding stone tools, which are commonly found on the periphery of debitage concentrations and scatters. This strategy succeeded in the recovery of a relatively large inventory of stone tools, the majority of which consisted of side scrapers (Table 28). In general, side scrapers are indicative of Early Archaic assemblages, particularly when they occur in concentrations. The four side scrapers recovered from TU1 were made from different raw material types. One (a1154) matched the UER Type I element and another (a1155) matched the USR(2) Type III element. A possible adz flake (a1174) also matched the UER Type I debitage concentration. The remaining two scrapers were made from USR(1) and a green, fine-grained metasedimentary material (a1171 and a1166 respectively), both raw materials of which occur infrequently in the area of SB14. Additional USR(1) artifacts consisted of two pieces of debitage and a utilized flake (a1159) recovered from TU1. No other metasedimentary material was recovered. In addition, two core fragments were recovered, an indeterminate fragment of MMR (a1135) and a directional core made of UER (a455). The flake scars on the remaining face of the directional core (Figure 24:B) correspond to the size parameters of the flake tools in the inventory, indicat- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 141 Chapter 5. 31HK2502 cm bd 0 H Lev 1 A 10 Lev 2 E1 20 Lev 3 30 Tree Roots E2 Lev 4 40 Lev 5 50 Lev 6 B N495.5/E499.5 60 TU1 N495.5/E500.5 0 1m Soil Strata H: Dark Gray (10YR 4/1) Prescribed Burn Charcoal and Loamy Sand A: Grayish Brown (10YR 5/2) Sandy Loam E1: Light Yellowish Brown (10YR 6/4) Fine Sand E2: Very Pale Brown (10YR 7/3) Fine Sand B: Brownish Yellow (7.5YR 6/6) Fine Sandy Loam Figure 35. Profile Drawing, TU1, North Wall, 31HK2502. ing that most of the reduction activity in the deposit involved flake blank production from block cores for the purposes of manufacturing expedient flake tools such as scrapers and utilized flakes. Vertical distributions of artifacts register slightly higher in the test unit level data than they do in the shovel tests in SB14. Artifacts are centered at 20 to 30 cm bd in the test unit and are predominantly found between 10 and 40 cm bd. The same debitage profiles are centered between 30 and 40 cm bs in shovel tests and concentrate mainly between 20 and 50 cm bs (Table 26). The same structure of elements is maintained in both data sets and it is likely that the differences are explained in terms of the difference in recovery technique. Although the elements are characterized by a relatively high vertical position, the high incidence of side scrapers strongly argues for Early Archaic associations. Of course, each of these elements may represent separate occupations, but there is equally strong evidence that most of them belong to a Type II residence or residences. Type II residences are similar in spatial organization to Type I residences, but they exhibit much higher tool-to-debitage ratios and may commonly be comprised of diverse lithic raw materials. It is speculated that raw material heterogeneity is a function of groups travelling from different locales to aggregate at selected places on the landscape at specific times. It is likely in these situations that individual groups will be provisioned with a different assemblage of raw material types, even if the groups were foraging within the same general region. Lithic Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 142 Chapter 5. 31HK2502 Table 28. Stone Tools and Debitage by Level for TU 1 and SB 14, 31HK2502. STONE TOOLS AND DEBITAGE BY LEVEL FOR TU1 AND SB14: 31HK2502 TU1 RAW MATERIAL TYPE Type I Rhyolite Tuff Mill Mountain Rhyolite Metasedimentary, Green Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 2 4 1 2 7 4 1 1 3 7 18 1 2 1 1 5 1 2 4 1 3 13 10 1 6 4 5 26 6 1 7 2 2 3 1 14 GRAND TOTAL 2 0 0 12 1 5 3 14 37 2 1 1 6 4 5 10 2 DEBITAGE LEVELS 3 4 5 SB14 RAW MATERIAL TYPE Grano-Diorite Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL DEBITAGE LEVELS 3 4 5 2 1 2 5 GRAND TOTAL 2 STONE TOOLS LEVELS 3 4 Core Fr. IIa SS UF IIb SS 2 III SS Flake Adz IIa SS STONE TOOLS LEVELS 3 4 Anvil Stone 1 26 1 13 8 12 61 Uni Core Fr. Core Fr.: Core Fragment, IIa SS: Type IIa Side Scraper, IIb SS: Type IIb Side Scraper, III SS: Type III Side Scraper, Uni Core Fr.: Unidirectional Core fragment, UF: Utilized Flake reduction on these sites consisted of tool maintenance and late stage shaping of tools from flake blanks or reduced biface cores. Consequently, debitage density is much lower than that of Type I debitage concentrations and the scatters consist primarily of late stage reduction debitage. Type II occupations are inferred to be the output from single or small, multiple household aggregations and are thought to represent the residue from High Technology Forager (HTF) residences (Spiess 1984; Todd 1983:231-233). The HTF model proposes a specialized forager adaptation combining high logistical and high residential mobility into a single settlement system characterized by highly curated technologies. The element that stands out from this potential grouping is the UER debitage concentration, which displays shovel test outcome characteristics that would set it apart as representative of a Type I residence. The UER debitage, however, is not effectively different from the reduction profiles of any of the other raw material types in the sample. They are characterized by small core flakes and BTFs and only small to moderately sized FBRs, which is not indicative of major tool production episodes, as one would expect for a Type I residence. The total weight of all debitage recovered from SB14 and TU1 is only 56.45 gm (about 2 ounces). Nearly as much raw material is tied up in MMR core fragment (4.06 gm) and the UER directional core fragment (47.96 gm). Moreover, the ratio of chipped stone tools to debitage is only 1:5.7 gm by weight and 1:16.5 by count. If the flake adz is added to the weight calculation, the ratio reduces to 1:1.39 gm. These characteristics are well within the expectations for Type II residences and, although the exposure is not sufficient to demonstrate this contention, there is a strong possibility that all of the elements, including the UER element, are members of at least one Type II residence. Because of the heterogeneous nature of the raw material composition of such elements, the degree to which occupational deposits are mixed cannot be objectively evaluated without seeing larger distributional patterns in the area of SB14. Test Unit 2 (SB 3) TU2 was placed at grid coordinates N502.0/ E501.0 (NE corner) in the northwest quadrant of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 143 Chapter 5. 31HK2502 SB3. Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of the unit (Figure 36). Four soil layers were exposed in the wall profiles, not including the upper humic zone of the A-horizon, which was charcoal laden from prescribed burning activities. The exposed profile is identical to that described above for TU1, which is only about 6 m distant. The lower A-horizon transitions into the E1-horizon at 10 cm bd. The E1-/E2-horizon contact occurs at about 25 to 30 cm bd and the brownish yellow (7.5YR 6/6) fine sandy loam substrate begins at about 55 cm bd and displays a sharp, flat contact with the overlying E2-horizon. SB3 yielded evidence of four Type III elements (Tables 26 and 29) and produced three Middle Archaic projectile points, two Morrow II Stemmed points (Figure 25:D-E) and a Guilford Lanceolate, Round Base point (Figure 25:F). One of the Morrow Mountain II Stemmed points (a25) was made of UWR and matched the raw material composition of a UWR debitage scatter displaying the characteristics of a Type III element. In addition, a UWR upper blade fragment from another projectile point (a386) was recovered at a distance of 1.25 m from Specimen a25, suggesting the presence of a tool cluster inferred to represent a hearth area associated with the Type III element. The UWR element was centered between 30 and 40 cm bs, a depth consistent with a Middle Archaic deposit. The blade fragment was also recovered at this depth, while the point was found between 20 and 30 cm bs. The other Morrow Mountain II Stemmed point (a344) was manufactured from a greenish gray, fine-grained metasedimentary material that was unmatched by any of the debitage recovered. It was found in a Stage III shovel test located 2 m west of the UWR tool cluster and the points’ relationship with this Type III element could not be determined on the basis of the shovel test sample. It too was recovered at a depth of 30 to 40 cm bs, at a depth consistent with a Middle Archaic deposit. The Guilford Lanceolate (a395), a later Middle Archaic style, was made of MMR, a raw material type that was unrepresented by debitage in the sample block. It was found in the southwest quadrant of the sample block at a depth of 30 to 40 cm bs. Other Type III elements identified in the sample block were composed of UER, USR(2) and white quartz. All of these elements were centered at 30 to 40 cm bs as well. TU2 was placed in the northwest quadrant of SB3 to further investigate the nature of the Middle Archaic surface, which was centered at a depth of 30 to 40 cm bs according to the shovel tests data. As was true of the TU1/SB14 excavation, however, the vertical positions of the various elements in TU2 were shifted one level higher, centering at a depth of 20 to 30 cm bd (Table 29). Additional Middle Archaic diagnostics were not recovered, but the presence of MMR and metasedimentary debitage that could be respectively matched to the Guilford Lanceolate and Morrow Mountain II Stemmed points was documented. The unit produced a UER Type II Biface fragment (a1069), a white quartz core fragment (a1119) and a Type I Rhyolite side scraper fragment (a1101). The biface matches the UER debitage scatter, which was substantially expanded in representation by the test unit excavation, and the core fragment matches the white quartz Type III element. The side scraper is probably Early Archaic in association. Three pieces of matching Type I Rhyolite debitage were also recovered in the test unit, suggesting the presence of a Type III Early Archaic element. The apparent compressed nature of the deposits inhibits using vertical debitage profiles to distinguish separate occupations. However, the wide range of lithic raw material types and the presence of matching stone tools increase the likelihood that these elements can be analytically separated. The identification of Middle Archaic tool cluster also demonstrates that the apparent deposit mixing has not significantly impacted the integrity of the spatial organization of the occupations. The results of the of the investigation suggest that the area of SB3 contains numerous Type III short-term camps primarily associated with the Middle Archaic period. The structure of the debitage scatters associated with the Middle Archaic Type III elements are similar to that described above for the Early Archaic deposits. Flake size is generally small and the total weight of the debitage inventory is relatively low (65.62 gm or 2.31 ounces). Moreover, the sum of stone tool weights is high (41.50 gm minus the weight of the Type IIa Side Scraper which is probably affiliated with an Early Archaic occupation) relative to debitage weight. The chief difference between the individual Type II and III elements is that tool diversity is much higher in the former. Typically, Middle Archaic occupations contain only cores, bifaces and projectile points, while Early Archaic occupations Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 144 Chapter 5. 31HK2502 0 H Lev 1 cm bd A 10 Lev 2 E1 20 Lev 3 30 Tree Roots Lev 4 40 E2 Lev 5 50 Lev 6 B N502.0/E500.0 60 TU2 0 N502.0/E501.0 1m Soil Strata H: Dark Gray (10YR 4/1) Prescribed Burn Charcoal and Loamy Sand A: Grayish Brown (10YR 5/2) Sandy Loam E1: Light Yellowish Brown (10YR 6/4) Fine Sand E2: Very Pale Brown (10YR 7/3) Fine Sand B: Brownish Yellow (7.5YR 6/6) Fine Sandy Loam Figure 36. Profile Drawing, TU2, North Wall, 31HK2502. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 145 Chapter 5. 31HK2502 Table 29. Stone Tools and Debitage by Level for TU 2 and SB 3, 31HK2502. STONE TOOLS AND DEBITAGE BY LEVEL FOR TU2 AND SB3: 31HK2502 TU2 RAW MATERIAL TYPE Metasedimentary, green Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 2 9 6 1 2 2 4 20 SB3 RAW MATERIAL TYPE Metasedimentary, Gray Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 2 DEBITAGE LEVELS 3 4 2 1 1 2 14 4 4 1 1 5 3 1 8 1 33 15 DEBITAGE LEVELS 3 4 5 1 1 1 2 1 2 3 4 1 13 7 1 2 10 7 27 5 6 GRAND TOTAL 2 2 11 3 25 5 7 4 15 72 5 GRAND TOTAL 1 1 6 2 4 IIa SS II Bif Core Fr. 2 0 0 14 3 7 16 12 52 2 1 1 2 2 8 STONE TOOLS LEVELS 3 STONE TOOLS LEVELS 3 4 MMt II St Guilford MMt II St PPK Fr. Core Fr.: Core fragment, Guilford: Guilford Lanceolate, Round Base variant, MMt II St: Morrow Mountain II Stemmed, PPK Fr.: Projectile Point, Upper Blade fragment, II Bif: Type II Biface fragment, IIa SS: Type IIa Side Scraper contain these items and a wide variety of scrapers and other unifacial tools. Investigations in the area of SB3 also yielded a relatively large sample of precontact ceramics. Shovel testing produced eight sherds representing Hanover I and II series types (Table 24). Excavation of the test unit produced an additional 20 sherds representing Hanover II and III and Yadkin III types (Table 30). In total, 10 individual vessels are represented in the combined inventory. The presence of four series and 10 vessels in this small space, with an Table 30. Precontact Ceramics from TU 2, 31HK2502. PRECONTACT CERAMICS FROM TU2: 31HK2502 LEVELS 1 2 3 GRAND TOTAL Hanover IIa Fabric Impressed 1 5 1 7 Hanover IIa Indeterminate 1 5 6 Hanover IIa Indeterminate decorated 2 2 Hanover IId Fabric Impressed 1 1 CERAMIC TYPE Hanover IIIa Cord Marked 1 Hanover IIIa Fabric Impressed 2 2 Yadkin IIIb Cord Marked 1 1 GRAND TOTAL 1 2 16 2 20 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 146 Chapter 5. 31HK2502 average of only 2.8 sherds per vessel, points to a high degree of disorganization in the ceramic deposits, as discussed earlier. The ceramic inventory from TU2 was concentrated between 10 and 20 cm bs, while the Archaic occupation was centered only about 10 cm deeper, at 20 to 30 cm bs. The MMR and debitage scatter is centered between 10 and 20 cm bs and could provide a candidate for a Woodland period debitage scatter. However, the presence of a MMR Guilford point in the sample block suggests that this scatter is affiliated with a Guilford occupation. Test Unit 3 (SB1) TU3 was placed at grid coordinates N509.0/ E488.0 (NE corner) in the southwest quadrant of SB1. Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of the unit (Figure 37). The unit was terminated at 70 cm bd. Four soil layers were exposed in the wall profiles, not including the upper humic zone of the A-horizon, which was charcoal laden from prescribed burning activities. The exposed profile is identical to that described above for TUs 1 and 2. The lower A-horizon transitions into the E1-horizon at 10 to 12 cm bd. The E1-/E2-horizon contact occured at 30 cm bd and the brownish yellow (7.5YR 6/6) fine sandy loam substrate begins at about 65 cm bd and displays a sharp, flat contact with the overlying E2-horizon. SB1 yielded evidence of a series of Type III elements, two of which were associated with diagnostic Early Archaic projectile points (Tables 26 and 31). A UWR Taylor Side Notched point (Figure 25:A, a500) matched a low-density UWR debitage scatter and a UWR Type II Side Scraper (Figure 26:I, a527). This apparent Type III element was centered between 40 and 50 cm bs. A matched Type III element of USR(2) consisted of a Palmer II Corner Notched point (Figure 25:B, a502) and a low density debitage scatter also centered between 40 and 50 cm bs. The different projectile point styles establish that a sequence of Early Archaic occupation was extant in the area of the block. In addition to these, two other probable Early Archaic Type III elements were identified through shovel testing based on depth and tool associations. These consist of a white quartz debitage scatter and a matching Type IIa Side Scraper (Figure 26:G, a520), and low-density UER debitage scatter. An MMR Type III element appeared to be positioned somewhat higher in the sediment and was assigned a Late Archaic/Middle Archaic age. Just as was true for the TU1/SB14 investigation, the low-density Early Archaic occupations in SB1 may represent separate Type III short-term residences or they may be members of Type II camps. TU3 was placed immediately adjacent to the Stage I shovel test at N510/E490 that produced both projectile points to further evaluate the structure of the identified elements. This location yielded very little additional information. Only five pieces of debitage and a Type IIa Side Scraper were recovered (Table 31). The scraper (a1042) was made of UWR and matched the Taylor Side Notched element. All of the elements recovered are characterized by small core flakes and BTFs and only small to moderately sized FBRs, which is quite analogous to the inventory recovered from the TU1/SB14 sample. The total weight of all debitage recovered from SB1 and TU3 is only 22.75 gm (about 0.8 ounces). More raw material mass (31.27 gm) is actually taken up by the chipped stone tools. The ratio of chipped stone tools to debitage is only 1:0.73 gm by weight and 1:7.4 by count. These characteristics are well within the expectations for Type II residences and, although the exposure is not sufficient to demonstrate this contention, there is a strong possibility that all of the elements are members of two superimposed Type II residences. Broader exposures wherein the larger spatial patterns of elements can be traced and reconstructed is the only means by which individual Type III elements can be effectively differentiated from Type II residences. Type II residences would be expected to exhibit a coherent, semicircular pattern of elements oriented around a communal are with multiple tool clusters, whereas Type III elements should be randomly distributed to one another and commonly contain only single tool clusters. EVALUATION Site 31HK2502 is situated on a low, “islandlike” rise immediately adjacent to Rockfish Creek in Hoke County. The landform has the topographic characteristics of an alluvial levee, an interpretation Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 147 Chapter 5. 31HK2502 cm bd 0 H A Lev 1 10 A’ E2’ Lev 2 E1 20 Lev 3 30 Tree Roots Lev 4 40 E2 Lev 5 50 Lev 6 60 Lev 7 B 70 N509.0/E487.0 TU3 0 N509.0/E488.0 1m Soil Strata H: Dark Gray (10YR 4/1) Prescribed Burn Charcoal and Loamy Sand A: Grayish Brown (10YR 5/2) Sandy Loam A’: Displaced A-Horizon E1: Light Yellowish Brown (10YR 6/4) Fine Sand E2: Very Pale Brown (10YR 7/3) Fine Sand E2’: Displaced E2-Horizon B: Brownish Yellow (7.5YR 6/6) Fine Sandy Loam Figure 37. Profile Drawing, TU3, North Wall, 31HK2502. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 148 Chapter 5. 31HK2502 Table 31. Stone Tools and Debitage by Level for TU 3 and SB 1, 31HK2502. STONE TOOLS AND DEBITAGE BY LEVEL FOR TU3 AND SB1: 31HK2502 TU3 RAW MATERIAL TYPE indet. metavolcanic Mill Mountain Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 2 3 1 1 6 7 1 7 GRAND TOTAL 2 2 5 1 3 1 5 7 10 32 1 1 1 0 2 0 3 3 2 6 11 1 2 1 DEBITAGE LEVELS 4 5 1 1 1 2 1 1 5 1 4 2 6 13 6 0 GRAND TOTAL 1 2 1 0 1 5 1 SB1 RAW MATERIAL TYPE Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL DEBITAGE LEVELS 4 5 4 STONE TOOLS LEVELS 5 6 STONE TOOLS LEVELS 5 6 IIa SS 4 Tay SN IIa SS P II CN IIc SS P II CN: Palmer II Corner Notched, Tay SN: Taylor Side Notched IIa SS: Type IIa Side Scraper, IIc SS: Type IIc Side Scraper, that is supported by the texture and composition of the substrate, which is a non-consolidated fine sandy loam rather than a compact clay loam as is common on upland hilltops in the region. This basal stratum may in fact represent a remnant of the original T1 terrace of the creek not yet removed by lateral incising of the stream channel. Regardless of the origin of the landform, it provided precontact hunter-gatherer groups with an ideal location from which to exploit a segment of the Rockfish Creek floodplain. The Archaic occupation of the site is primarily associated with the Middle and Early Archaic periods. The dominant settlement theme during this time was short-term residential and special purpose camps occupied for periods of only several weeks at the most. Long-term land-use probably involved serial reoccupations by family groups for several years punctuated by extended periods of abandonment as resource depletion reached a point where the foraging radius could no longer produce sufficient yields. Such a scenario would create an archaeological record that might be characterized by the reoccupation of the same camp locations on the landform by lineally related and/or demographically cycling family units for a period of years until the settlement continuity was broken by longer periods of abandonment. Test unit excavation demonstrated that the vertical separation between Archaic deposits is compressed and not as clear as seen on some uplands sites in the region. However, it also showed that individual occupational events could be analytically recognized and segregated. Debitage and tools could be easily matched by raw material type and the elements of individual camps such as tool clusters have been preserved. The Archaic occupations consist primarily of Type I and III elements. These elements are characterized by very short-term occupation and they are difficult to distinguish from limited exposures because they share many compositional qualities. The are both characterized by low debitage densities, limited debitage weight due to reduction strategies focused on tool maintenance and flake blank production to fashion expedient unifacial tools and unretouched flake tools. Type II residences, however, are characterized by multiple elements of heterogeneous raw material composition, high stone tool fre- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 149 Chapter 5. 31HK2502 Flake Tool (R3B) (Fig. 24:J) Flake Blank (Q4) Flake Blank (Q4) (Fig. 14:H) Type III Side Scraper (Fig. 22:H) Flake Tool (R3A) (Fig. 23:F) 592.5 Flake Tool (Q4) Type II Biface Frag (R3D) Type II Biface (R3D) Flake Tool (R3F) (Fig. 24:I) Flake Tool (R3F) Palmer I CornerNotched(R3E) (Fig. 17:C) TU5 Serrated PPK Tip (Q4) Type II Biface Frag. (Q4) Flake Tool (Q4) (Fig. 24:F) Flake Tool (R3B) Type II End Scraper (R3A) (Fig. 21:H-I) Flake Tool (R2A) (Fig. 24:G) 590.0 Palmer II Corner-Notched (R3E) (Fig. 17:D) Morrow Mt. II Stemmed (R7) (Fig. 18F) 587.5 540.0 Flake Tool (R2) 537.5 0 542.5 N 2.5 meters Sample Block 6: 31HK1214 Quartz (Q4) Debitage (cont=1, begins @ 2, max=6) R2 Group Debitage (cont=1, begins @ 2, max=7) R3 Group Debitage (cont=1, brgins @ 2, max=7) R7 Group Debitage (cont=1, max=2) Figure 38. Hypothesized Type II Occupation at 31HK1214 (from Cable and Cantley 2005b). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 150 Chapter 5. 31HK2502 quencies and high tool diversity. Only a few Type II residences have been identified in the region because of the necessity to expose large areas to confirm their existence. The best example (Figure 38) of one of these occupation types is supplied by an extensive closeinterval shovel test block and associated test unit excavation at 31HK1214 (see Cable and Cantley 2005b:204-210). Occupation HK1214.SB6.R2/ R3.1 was comprised of low density scatters of Wolf Den Mountain Rhyolite (UWR), Type I Rhyolitic Tuff and Quartz and a wide array of chipped stone tools. The tool inventory consisted of two Palmer Corner-Notched points, five flake tools, two Type II biface fragments, and a type II end scraper of UWR and USR(1); two flake blanks, two flake tools, one serrated projectile point tip, and a Type II biface of quartz; and a Type III side scraper and two flake tools of Type I Rhyolitic Tuff. Tool clusters were identified at either end of the debitage scatter, which suggests the presence of at least two social units, either nuclear family-like units or specially comprised task groups. Generally, it is assumed that tools will be discarded at relatively low rates, especially curated ones, which would imply that the Type II residences involved lengthy rather than short duration stays. Cable (1996), however, has argued that high discard rates for curated tools can result from bulk processing activities such as the butchering and skinning of ungulates from mass kills. This would require aggregations of deer populations in the Southeast, which tend to occur in winter months when snow accumulations drive deer to gather in areas of high canopy cover where browse material is left exposed and optimal bedding areas exist (Gates and Harman 1980). Certainly, the Early Holocene environment of the Carolinas was far from a boreal forest, but the characteristic that may link these disparate cultural systems is an economic focus on highly mobile and dispersed ungulate populations, in this case deer. Opportunities for such aggregations probably waned with time and this may explain why Type II occupations have not yet been recognized for deposits of later age. Woodland periods. These occupations appear to have also involved short-term seasonal occupation of the landform to exploit floodplain resources. Precontact ceramics were distributed across the entire landform, but a high degree of disorganization and displacement was demonstrated from their analysis. No potential whole pot deposits were identified and 49 individual vessels were recognized in the total inventory of 95 sherds. Although such patterns can be produced by extensive historic disturbances such as plowing, it is likely that in this case the explanation relates to actual behaviors characterizing the Woodland occupants of the site. The A-horizon at the site was demonstrated to be virtually undamaged by historic disturbances and the primary concentration of ceramics on the site is actually situated below the A-horizon at depths of 10 to 20 cm bs. Most of the ceramic disorganization, then, must be the consequence of processes such as reoccupation and recycling of broken pots and large sherds, even to the point of removing partial vessels and large sherds upon site abandonment. Site 31HK2502 is recommended not eligible for inclusion on the National Register of Historic Places under criterion d, “that have yielded, or may be likely to yield, information important in prehistory or history’ in accordance with 36CFR60.4. Even though important archaeological deposits are present at the site, most specifically Type II and III Early Archaic elements and the relatively intense Woodland ceramic deposit, the degree of reoccupation intensity is high and substantial deposit mixing was documented. Analytical methods would provide a basis to mitigate some of the deposit mixing that could not be resolved by recovery methods, but a nearly identical occupation history is contained at nearby 31HK2510, which supports a much lower incidence of superimposition of separate occupations. That site is recommended eligible in this report and the information contained at 31HK2502 is considered redundant to that preserved in much better condition at 31HK2510. It is estimated that about a quarter of the occupation that occurred at the site was associated with the Woodland period. A long history of Woodland occupation was documented, principally ranging between the early Middle Woodland and Late Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 151 Chapter 5. 31HK2502 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 152 Chapter 6. 31HK2510 Site 31HK2510 is located in Hoke County in the western portion of Fort Bragg (Figure 39). The site can be reached by traveling approximately 0.3 miles east-northeast along Firebreak Road 33 from its juncture with Moore County Road to a two-track road on the right. Take the road south for approximately 0.1 miles. The site is primarily situated west of the road at this point going south. Topographically, 31HK2510 occupies a stretch of the upland terrace ridge on the west bank of Rockfish Creek (Figures 40 and 41). The SCS soil survey for Hoke County, NC (Hudson 1981:81) classifies the landform as Blaney loamy sand (2 to 8 percent slopes), a well drained soil that typically forms on side slopes and narrow upland ridges. The confluence of Rockfish Creek and Gum Branch is situated approximately 0.8 km (0.5 miles) south of the site. Although the site has been subject to tree thinning, the surface is relatively free of modern disturbances. However, a number of foxholes from military training exercises have been excavated along the west side of the two-track road (Figure 42). Elevations within the site boundaries ranged between about 71 m amsl at the edge of the creek channel to about 74.0 m amsl on the ridge top (Figure 43). Average slope from the creek channel to the top of the ridge is estimated at 3.9 percent. A small tongue of higher ground juts out into the narrow floodplain in the center of the site, allowing sufficient drainage for precontact occupation to extend below the elevation of the road to the east (Figure 44). A small springhead alluded to by Grunden and Ruggiero (2006:404) from the Phase I report originates at the north end of the site just below the road and drains into the creek (Figure 43). As documented by 10 m-interval shovel testing, maximum site dimensions were established at 180 m north-south x 100 m east-west, covering an area of approximately 2.77 acres (11,200 m2). Vegetation across the site consisted of a thinned stand of mature long-leaf pines and bottomland hardwoods, with an understory of turkey oak and wiregrass. PREVIOUS RESEARCH Site 31HK2510 was originally recorded by TRC Garrow Associates, Inc in the spring of 2003 during a Phase I survey of portions of the DD3 Training Area (Grunden and Ruggiero 2006:404-406). The site boundaries established at that time correspond closely to those later established by PRI (Figure 45). Site definition proceeded through the deployment of 15 m-interval shovel tests. Seventy-seven shovel tests were excavated in total, 17 of which were positive. Grunden and Ruggiero (2006:404) described three soil zones exposed in shovel tests and in a 50 cm x 50 cm test unit. The upper zone was composed of dark gray loamy sand and extended to a depth of 10 cm bs. The intermediate zone consisted of a light yellowish brown sand that terminated between 75 and 80 cm bs at the contact with strong brown clay substrate. The recovered artifact assemblage (n=45) consisted of 36 rhyolite flakes, a rhyolite core fragment, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 153 Chapter 6. 31HK2510 656000 657000 656000 657000 3885000 3884000 3884000 3885000 3886000 655000 3886000 654000 3883000 3882000 3882000 3883000 31HK2510 654000 655000 Figure 39. Location of Site 31HK2510 (McCain, NC 7.5’ USGS Quadrangle), Scale: 1:24,000. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 154 Chapter 6. 31HK2510 655400 655600 655800 656000 3884000 eak 32 Firebr eak 33 Firebr h fis ck Ro re oo M 3883600 ty un 3883400 Co eak 34 Firebr 3883400 3883600 31HK2510 3883800 655200 3883800 3884000 655000 3883200 3883000 3882800 3882600 3882600 3882800 3883000 3883200 Creek ad Ro 655000 655200 655400 655600 655800 656000 Figure 40. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31HK2510 (Scale: 1 inch = 200m, Projection: UTM, NAD83). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 155 Chapter 6. 31HK2510 Figure 41. Site 31HK2510, Looking South Down the Rockfish Creek Channel from Near N420/E540. 5 quartz flakes, 1 quartzite flake, a cord marked sandtempered sherd and an indeterminate sand-tempered sherd. Most of the artifacts were recovered between 20 and 60 cm bs, although several may have originated at 70 to 75 cm bs. The surveyors characterized the site as a low-density artifact scatter dominated by lithic debitage. Based on the vertical depth of the artifacts and a tendency for horizontal clustering to occur, it was suggested that the cultural deposits were intact. Consequently, it was concluded that the site was potentially eligible for inclusion on the National Register of Historic Places due to its ability to inform on site structure and precontact adaptation in the Sandhills. FIELDWORK OVERVIEW Phase II fieldwork was implemented in four stages, each stage building on the information generated from the previous stage. Stage I investigation consisted of a program of 10 m-interval shovel tests that were excavated to establish firm site boundaries. Site boundaries were determined by the documentation of two consecutive negative shovel tests in all directions on a 10-m grid, or in some cases when wetland was encountered. Once artifact distributions from the Stage I program were mapped and evaluated, the Stage II investigation involved the excavation of closer-interval shovel tests at 5 m-intervals in Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 156 Chapter 6. 31HK2510 Figure 42. Site 31HK2510, Looking South Down Dirt Access Road from N440/E520. most of the artifact bearing areas of the site. Stage III investigations involved the deployment of closeinterval shovel tests of 1.25 m-intervals around the perimeter of targeted shovel tests yielding deposits deemed of theoretical importance for evaluating the site. These close-interval shovel test locations were identified as sample blocks and generally consisted of eight shovel tests excavated around a targeted Stage I or Stage II shovel test. The sample blocks were also instrumental in guiding the precise placement of 1 x 1 m test units, which constituted the final stage of the investigation. Test units were placed in locations judged to provide optimal opportunities to reconstruct the occupation history of the site and to assess the integrity and scientific importance of the deposits. Stage I investigations were initiated by relocating the datum established by Grunden and Ruggiero (2006) and laying out a 10-m interval grid with a total station. The datum, which was placed at the location of survey shovel test T14-2, was easily found and assigned grid coordinates of N500/E500 in agreement with the Phase I system. Relative topographic elevations at each 10 m grid intersection were recorded with the total station and marked with a pin flag. The elevation values (in meters) illustrated on the base map (Figure 43) are approximated real elevations based on the correlation of GPS points with a geo-referenced copy of the USGS, McCain NC 7.5minute topographic quadrangle map. Cultural and landform features were also mapped and additional Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 157 Chapter 6. 31HK2510 530 520 Two-Track Road oc kf ish R Spring Head 510 Cr eek 73.0 500 .0 74 490 .5 73 480 470 460 450 31HK2510 440 430 73.0 420 Flood Plain 71 .5 410 400 0 72 .5 74. 390 72.0 380 370 74.0 360 350 N 340 0 40 m 330 320 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 43. Site 31HK2510, Base Map. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 158 Chapter 6. 31HK2510 Figure 44. Site 31HK2510, Looking Southeast Across Dirt Access Road from N440/E520 Toward Rockfish Creek Floodplain. elevations were taken when appropriate to produce the topographic map. Shovel tests measured 30-x-30-cm and were square in shape. Excavation proceeded in 10 cm arbitrary levels measured from the surface. One hundred sixty-five Stage I shovel tests were excavated at 10 m-intervals to define the limits of the site (Figure 46). Sixty of the tests were positive. Maximum site boundaries were established at 180 m x 100 m, covering an area of about 11,200 m2, or approximately 2.77 acres. Mean shovel test depth for the Stage I sample was 69.21 cm bs (SD=18.41) with a mode of 70 cm bs and a range of 20 to 100 cm bs. Excavation was terminated in the upper 5 to 10 cm of the substrate. The SCS soil survey for Hoke County (Hudson 1984:31) maps the area of the site as Blaney loamy sand, which is a well-drained upland soil situated on side slopes and narrow ridges. The on-site soil profile corresponds well to the typical Blaney pedon. The A-horizon extends to a depth of about 15 cm bs and appears to represent an old plow zone. The upper 3 to 5 cm of the A-horizon evidences organic contributions from prescribed burn episodes, while below this surface layer, the A-horizon is composed of grayish brown (10YR 5/2 to 2.5Y 5/2) sand and loamy sand. The upper E-horizon (E1) layer, when present, is composed of brownish yellow (10YR 6/6) to yellow (10YR 7/6) sand. The lower E2-horizon sand is lighter, ranging from very pale yellow (10YR 7/3) to pale yellow (2.5Y 7/3). The layers transition at about 30 to 35 cm bs, when both are present. The Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 159 - 545 - 530 - 515 - 500 - 485 - 470 - 455 - 440 - 425 Chapter 6. 31HK2510 - 530 - 515 S 13-5 - 500 14-2 - 485 Centimeters 0 13-7 2.5Y 4/2 Dark grayish brown Sand 10 Rockfish Creek South Wall - 470 15-2 two-track 14-3 13-6 15-3 14-4 - 455 - 440 - 425 20 - 410 30 13-8 2.5Y 6/4 Light yellowish brown Sand 15-4 14-5 S - 395 - 380 7.5YR 5/8 Strong brown clay S TR 13 TR 15 - 365 TR 14 GPS datum/Positive ST Negative transect ST Positive ST Positive transect ST Negative ST Site boundary Surface positive (if any) Contour (approximate) 50 cm x 50 cm test unit Wetland (if any) N Scale in Meters 0 15 30 45 Shovel test pits shown are not to scale Figure 45. Site 31HK2510, Phase I Survey Sketch Map, TRC (from Grunden and Ruggiero 2006). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 160 Chapter 6. 31HK2510 Positive ST 530 520 Negative ST Two-Track Road oc kf ish R Spring Head 510 Cr eek 73.0 500 .0 74 490 .5 73 480 470 460 450 31HK2510 440 430 73.0 420 Flood Plain 71 .5 410 400 0 72 .5 74. 390 72.0 380 370 74.0 360 350 N 340 0 40 m 330 320 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 46. Site 31HK2510, Stage I Shovel Test Sample. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 161 Chapter 6. 31HK2510 Bt-horizon, which typically begins between 50 and 100 cm bs across the site, consists of brownish yellow (10YR 6/8) to olive yellow (2.5YR 6/6) sandy clay loam of dense and compact structure. The Stage I sample provided a basis for defining a Stage II sample frame within which to deploy a 5 m-interval shovel test grid. This guaranteed that the artifact bearing locations of the site would receive systematic sampling at a relatively intensive sample density, so that a detailed picture of the occupation history of the site could be generated. Two hundred thirty-three Stage II shovel tests were excavated at 5 m-intervals within the sample frame (Figure 47). One hundred thirty-two were positive. Mean shovel test depth for the Stage II sample was 70.52 cm bs (SD=20.23) with a mode of 70 cm bs and a range of 10 to 110 cm bs. These statistics are nearly identical to those of the Stage I sample. The Stage III shovel test sample was distributed in 17 locations anchored by Stage I or II shovel tests containing data relevant to evaluating the occupational history and integrity of the site matrix (Figure 48). One hundred forty-eight Stage III shovel tests were excavated at typically 1.25 m-intervals around targeted Stage I and II shovel tests. One hundred twenty-two of these tests yielded artifacts. Mean shovel test depth for the Stage III sample was 73.72 cm bs (SD=18.35) with a mode of 70 cm bs and a range of 10 to 110 cm bs. Once the results of the Stage III shovel test sample were analyzed, specific locations yielding shovel test outcomes of interest to further evaluating the site were selected for test unit excavation. Stage III shovel testing around targeted shovel test outcomes provided the basis for precise placement of test units. Four 1 m-x-1-m test units were excavated in total and their locations are illustrated in Figure 49. The units were terminated between 40 and 70 cm bd. ARTIFACT INVENTORY Artifacts collected during the investigation overwhelmingly consisted of chipped stone debitage and tools. The precontact inventory consisted of 1,763 items including 1,224 pieces of lithic debitage, 90 chipped stone tools, 18 cores, a tabular metasedimentary fragment, two possible grinding stone fragments, a hammer stone spall, a hammer stone/anvil, 18 precontact ceramics, a large piece of biotite, and 407 rocks and fire-cracked rocks. A small number of historic artifacts were also recovered (n=11), which are primarily associated with military training activities. Appendix B contains the artifact database for the project, while appendices C through I provide additional data on lithic chipped stone and ground stone tools and precontact ceramics. Descriptions of the various artifact and lithic raw material types are presented below. Lithic Raw Material Types Lithic raw material identification proceeded on two different levels. At the macro-scale, individual specimens were typed in accordance with major hard rock geological terminology to achieve comparability with previous projects at Fort Bragg. At the micro-level, however, an attempt was made to further partition some of the hard rock types into “core groupings” to distinguish between specimens derived from single hard rock types that may have originated from different cores. This was done to facilitate spatial analysis in close-interval shovel test grids where overlapping of distinct deposits or complex multi-household occupations were suspected. Core grouping analysis was undertaken principally in metavolcanic types where subtle differences in patina coloration, core color, phenocryst size and density, and matrix texture was easily discerned and monitored. Descriptions of the various rock types and the associated core grouping sub-types recognized in the collection are presented below. Metavolcanic Types Metavolcanic types, as a group, dominate the chipped stone collection from the site (Table 32). They represent 79.4 percent (n=1,058) of the total chipped stone artifact inventory (i.e. debitage and chipped stone cores and tools). The best reflection of actual representation, however, is supplied by the combined Stage I and II shovel test samples, which is unbiased in terms of spatial coverage. In this sample, metavolcanic material composes 91.1 percent of the Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 162 Chapter 6. 31HK2510 530 520 Positive ST Negative ST Two-Track Road oc kf ish R Spring Head 510 Cr eek 73.0 500 .0 74 490 .5 73 480 470 460 450 31HK2510 440 430 73.0 420 Flood Plain 71 .5 410 400 0 72 .5 74. 390 72.0 380 370 74.0 360 350 N 340 0 40 m 330 320 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 47. Site 31HK2510, Stage I and II Shovel Test Sample. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 163 Chapter 6. 31HK2510 530 520 Two-Track Road oc kf ish R Springhead 510 500 490 eek 73.0 13 14 Cr .0 74 .5 73 480 1 470 31HK2510 460 17 450 5 440 12 8 430 7 11 3 73.0 420 18 15 Flood Plain 10 71 .5 410 6 400 72 .5 9 0 74. 390 72.0 2 380 16 370 4 74.0 360 350 N 340 0 40 m 330 320 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 48. Site 31HK2510, Stage III Shovel Test Sample Showing Locations of Sample Blocks. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 164 Chapter 6. 31HK2510 530 520 Two-Track Road oc kf ish R Spring Head 510 Cr eek 73.0 500 .0 74 490 .5 73 480 470 31HK2510 460 450 440 430 73.0 420 Flood Plain 410 71 .5 TU2 400 0 72 .5 74. 390 TU3 72.0 380 TU1 TU4 370 74.0 360 350 N 340 0 40 m 330 320 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 49. Site 31HK2510, Locations of Test Units. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 165 Chapter 6. 31HK2510 Table 32. Total Chipped Stone Inventory, 31HK2510. TOTAL CHIPPED STONE INVENTORY: 31HK2510 UTILIZED FLAKES SIDE SCRAPERS RETOUCHED FLAKE PROJECTILE POINTS CORES BIFACES DEBITAGE RAW MATERIAL TYPE END SCRAPERS CHIPPED STONE TOOL CLASSES GRAND TOTAL Metavolcanic indeterminate metavolcanic Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite 7 7 76 420 2 65 1 1 78 1 1 14 438 3 4 2 77 2 4 5 7 133 7 4 265 3 Uwharries Southern Rhyolite (1) 112 2 1 Uwharries Southern Rhyolite (2) 236 6 2 10 54 1 1 2 Uwharries Western Rhyolite 1 1 60 1 1 2 268 Quartz Crystal Quartz White Quartz 249 3 11 1 2 Rose Quartz 1 1 Fine Grained Quartzite 1 1 Metasedimentary, gray 1 1 Slate 2 2 Other Material GRAND TOTAL 1224 14 18 4 23 1 20 28 1332 chipped stone. Descriptions of the five identified metavolcanic types and associated core group subtypes follow. The typological system varies somewhat from earlier reports in response to the recent stone quarry sourcing study for the Carolina Slate Belt conducted by Steponaitis et al. (2006). (1) Mill Mountain Rhyolite (MMR). MMR was not clearly differentiated in the sourcing study (Steponaitis et al. 2006), but it is a readily recognizable type in the assemblages from Fort Bragg. Mill Mountain is located between the Uwharries Southern and Uwharries Eastern source areas on the opposite bank of the Yadkin River from Table Top Mountain. It was not included in either of these source areas, however, because of the presence of distinctive clear, glassy quartz phenocrysts. Daniel and Butler (1996:18) describe this type as a medium gray, aphanitic, rhyolite porphyry with sparse, glassy phenocrysts of quartz generally less than 1 mm in diameter. Very fine-grained, disseminated grains of pyrite are also present. Thin sections indicate a microcrystalline matrix composed primarily of feldspar and quartz, with some biotite, chlorite, and disseminated pyrite. The specimens assignable to this type from the 31HK2510 collection are characterized by a dark reddish gray to moderate brown coloration when not Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 166 Chapter 6. 31HK2510 patinated. Five subtypes were assigned to the type, based primarily on the completeness of the observation field and patination. Subtype R8 corresponds to specimens that exhibit sparse glassy quartz phenocrysts only. Subtype R8c is a coarse grained variant of R8. Subtypes R8Hhf, R8Hhm and R8Hhc are fine-, medium- and coarse-grained specimens respectively that exhibit sparse glassy quartz, white subhedral quartz phenocrysts and, rarely, feldpsar phenocrysts. Although these phenocryst types were not included in the type description, the subtypes containing them were included as a variant of Mill Mountain Rhyolite due to the co-occurrence with glassy quartz phenocrysts. Mill Mountain Rhyolite makes up 6.7 percent of the combined Stage I and II chipped stone sample. (2) Type I Rhyolite Tuff (Type I RT). An exact correlate of this type was not discernible in the sourcing study (Steponaitis et al. 2006). Benson (1999:30) describes it as fine-grained dark green to gray material that looks and feels very similar to Piedmont chert. He noted that plagioclase phenocrysts were sporadically present. Five subtypes were identified in the 31HK2510 collection. Subtype R2 is only rarely identified in assemblages on Fort Bragg. It appears very similar to Subtype R3t, but its coloration can be described as a dusky blue green. The texture is microcrystalline and the edges are semi-translucent. Subtype R2A specimens exhibit a pale green patina, but broken edges reveal that the underlying material corresponds to R2. Darker green splotches and swirls on the faces of these specimens represent more resistant matrix. The matrix is aphantitic and microcrystalline in texture and is highly isotropic in fracture characteristics. Also recognized were subtypes R2Am, R2Ac and R2Af. Respectively, these subtypes correspond to medium-grained, coarse-grained and flow banded or laminated specimens. The origin of this material is not well understood and it is possible that the type represents a very fine-grained metamudstone similar to some of those described for the Chatham Pittsboro source (see Moore and Irwin 2006:27-28; Stoddard 2006:57). In support of this is the occurrence of what appear to be bedding laminations in some of the patinated material. Moreover, the color of the patinated specimens is similar to that reported for the Chatham Pittsboro source. Type I Rhyolite Tuff makes up 37.7 percent of the combined Stage II and II chipped stone sample, an uncommonly high occurrence for this raw material type. Generally, Type I Rhyolite Tuff com- prises less than 5 percent of chipped stone assemblages in precontact sites on Fort Bragg. (3) Uwharries Eastern Rhyolite (UER). Stoddard (2006:52) describes the specimens from this sourcing area as light to dark gray metadacite porphyry or crystal lithic tuff. All samples collected contained plagioclase feldspar and white quartz phenocrysts, ranging from less than 2 percent to 7 percent phenocryst density. Common metamorphic minerals within the matrices include biotite, stilpnomelane and calcite. Similar dacites and rhyolites are typical of the Uhwarries Southeastern source (Stoddard 2006:55-56) as well, but many of these specimens contain silica levels beyond the range typical of igneous rocks. Macroscopic criteria for distinguishing between lithic material from these two sources does not at present exist. Consequently, specimens from both source areas may be present in the analyzed lithic assemblages at Fort Bragg. Typical specimens from 31HK2510 are characterized by dark gray (nonpatinated) and grayish green (patinated) matrices with splotches of silicate and variable percentages (2 to 4 percent) of white quartz and plagioclase feldspar phenocrysts. Some specimens exhibit purplish to brown background matrices. Quartz phenocrysts predominate over feldspar in the examples from the site. Type I Rhyolite, which has been distinguished separately in previous reports (Cable and Cantley 2005b, 2006, 2010), is also subsumed under this type. Three subtypes were identified in the collection: R8Ihf, R8Ihm, R8Ihc, all of which contained white quartz and plagioclase feldspar phenocrysts in variable percentages. They are distinguished from each other by differences in grain. Respectively, the subtypes correspond to fine-, medium- and coarsegrained specimens. UER makes up 6.6 percent of the combined Stage I and II chipped stone sample. (4) Uwharries Southern Rhyolite (USR). This type includes sources found on Morrow Mountain and surrounding areas, most notably Table Top Mountain. Daniel and Butler (1996:10-13) describe this type as a dark gray, aphanitic, aphyric rhyolite, that commonly exhibits flow-banding. Weathering tends to bring out the flow lines visually, which generally alternate in a pattern of light and dark gray. Although the matrix is usually homogeneous, some specimens exhibit small spherulites of less than 1 mm Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 167 Chapter 6. 31HK2510 in diameter (see Stoddard 2006:52). Thin sections indicate a microcrystalline intergrowth of feldspar and quartz, with minor biotite and chlorite elements. The individual minerals are difficult to distinguish and strings of dark minerals can occur along fracture planes. The specimens identified as USR in the 31HK2510 assemblage are composed of two larger groupings, the 3m and the R3/R3t groups. The 3m Group corresponds to material that typically occurs on Morrow Mountain and includes a series of subtypes that describe variation in flow banding, texture and patination. Subtype R3m is non-patinated and it exhibits a dark gray fine-grained matrix devoid of phenocrysts, while R3mf corresponds to flow banded specimens of R3m material. Subtypes R3mD, R3mDf, R3mDs and R3mDfL are patinated and range in color from grayish green to greenish gray to moderate yellow or tan. R3mD and R3mDf represent patinated specimens of R3m and R3mf respectively. R3mDs and R3mDfL exhibit blotches and linear arrays of silicate containing biotite and perhaps chlorite elements, respectively. They too, appear to be patinated forms of R3m and R3mf. The R3/R3t group is composed of an extremely fine grained or microcrystalline dark gray matrix that is semi-translucent when viewed on attenuated edges. Blotches of feldspar are commonly seen on exposed surfaces, but they do not form phenocrysts. R3 and R3t are non-patinated, while R3p, R3pat and R3patf are fully patinated subtypes. All three are tan in color, but R3p also exhibits finely disseminated biotite specks and R3patf exhibits flow banding. Finally, R3tf constitutes a flow-banded version of R3t. USR makes up 34.3 percent of the Stage I and II chipped stone sample. The R3/R3t group (Group 1) makes up 11.5 percent of the Stage I and II sample, while the R3m group (Group 2) comprises 22.8 percent. (5) Uwharries Western Rhyolite (UWR). This source area includes specimens from Wolf Den Mountain and Falls Dam (Stoddard 2006:52). This material is described as gray to black microcrystalline felsic volcanic rock that contains rounded plagioclase phenocrysts and glomerocrysts. Green biotite and pale green amphibole are dispersed locally in the matrix and spherulites are sometimes present. Three subtypes were identified in the collection. The main subtype, R3mp, has been patinated to a grayish green or greenish gray color. Non-patinated specimens may have been assigned to one of the other Uwharrie source areas in the analysis, but the distinctive glomerocyst structure suggests that this was not a major problem. Due to patination, most of the specimens in the collection were characterized by fine to medium texture. Minor subtypes include R3mpc (coarse-grained specimens) and R3mps (specimens with splotches of greenish silicate embayments). UWR makes up 5.2 percent of the combined Stage I and II chipped stone sample from 31HK2510. (6) Indeterminate Metavolcanic. A small amount of the metavolcanic material identified in the analysis could not be further classified to type due to excessive hydration. These were simply classified as “indeterminate metavolcanic.” Only seven pieces of debitage were placed in this category in the Stage I and II sample. Quartz Types Pure quartz forms as veins in igneous and metamorphic rock formations characterized by slow crystallization. This generally produces anhedral, milky structures (Spock 1962). In some cases, however, crystallization occurs earlier in the sequence, producing clear crystals with euhedral structure. Both vein and euhedral crystal quartz were utilized by precontact groups in the project area, but most of it appears to have come from river gravel deposits. A great deal of variation was evident in the vein quartz category ranging from mixed clear and cloudy exposures to opaque milky white. Since these characteristics appeared to be repeatedly present on single cores when debitage concentrations were analyzed, all of this variation was subsumed under the single white quartz subtype Q5. The category “crystal quartz” (CQ) was retained for instances where glass-like examples and crystal facets strongly suggested a euhedral crystal origin. “Rose quartz” was reserved for quartz specimens with a high degree of rubification, a commonly observed attribute in quartz cobbles. White quartz represented 8.4 percent of the combined Stage I and II chipped stone sample, while only one piece each of crystal quartz (0.1 percent) and rose quartz (0.1 percent) were members of the sample. In addition, 23 pieces and one piece, respectively, of fire-cracked rock derived from white quartz Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 168 Chapter 6. 31HK2510 and rose quartz cobble fragments were collected in the entire Phase II collection at the site. Minority Types Several other rock types were recognized in the 31HK2510 collection. Descriptions of each are presented below. (1) Arenite. This material has been previously identified as ferruginous sandstone, but thinsectioning of a piece of this material indicates that a more accurate characterization is arenite, or hematite cemented quartz sandstone (Appendix K). Three pieces of arenite were classified as fire-cracked rock. It occurs in a variety of shapes including flattened and irregular nodules. The exterior of the nodules consist of fine-grained ruddy brown precipitate while the centers consist of dense, reddish brown to purple, densley cemented sandstone. The material was commonly used for rock hearths by the precontact inhabitants of Fort Bragg. Three hundred thirty pieces of arenite were collected during the investigation, totaling 9.02 kg. In addition to serving as a hearth rock constituent, the material also appears to have been used for grinding activities. At 31HK2510, a single arenite fragment (a849) exhibiting what appears to be a flat surface with grinding facets was recovered. Although such items could have been used to grind vegetal material, they more likely represent abraders for stone tool manufacturing. (2) Fine Grained Quartzite. A very fine grained and dense quartzite was sparingly recognized in the collection. The material ranged from dark yellowish orange to grayish brown in color and in all instances was derived from river cobbles. Three pieces were recognized, in the forms of a piece of debitage, a hammer stone spall (a1058) and a complete hammer stone/anvil stone (a288). (3) Metasedimentary. A single piece of fine to medium grained, homogeneous metasedimentary debitage of light gray color was recovered during the investigation. The types are distinguished solely by color. Although the material occurs only infrequently, it was apparently used to manufacture a wide range of tool types in the region. At nearby 31HK2502, a Morrow Mountain II Stemmed base (271265a344), a Type III Side Scraper (271265a1166) and a fragment of a grooved abrader (271265a325) were all made from this material. The material represents only 0.1 percent of the chipped stone inventory in the Stage I and II shovel test sample. A large slab-like river cobble fragment (a580) of greenish metasedimentary material was recovered in a Stage III shovel test at 31HK2510. The relative coarseness of the grain suggests that the material can be further classified as metasiltstone or fine metasandstone. Numerous source areas in the Durham-Wadesboro Basin contain similar rock types, including the Chatham Pittsboro, Person County and Chatham Siler City (Stoddard 2006:57-63). (4) Sandstone Conglomerate. Also abundantly present in the deposit were chunks of sandstone conglomerate containing a wide range of sand and pebble sizes. Thin-sectioning of a piece of this material determined that individual grains of sand were cemented with clays (wacke) and exhibited udulatory extinction indicating that it originated from metamorphic rock (Appendix K). Occasionally interfaces with arenite could be observed on the chunks, indicating that the two materials were derived from the same formation (e.g. Middendorf). The matrix of the conglomerate consistently exhibited a grayish orange to dusky yellow coloration. Chunks of arenite were sometimes viewed in the matrix as well. This material appears to have been used to make rock hearths. Forty-seven pieces of sandstone conglomerate were collected, summing to about 1.17 kg. Only one of these specimens exhibited clear evidence of firecracking, but it is likely that all of this material was brought to the site to be used as hearth rock. A single fragment containing a flat grinding facet was also recognized (a341). (5) Slate. Two pieces of a dark to medium grayish green, laminated material identified as slate were recovered during the investigation. Both consisted of angular fragments classified as debitage. It is likely that they were by-products from the attempted shaping of a blocky slate core. (6) Float. Seven pieces of small, water worn nodules of lithic material were recovered that are worth mentioning because they were either brought to the site by precontact groups or were transported and deposited during flood events. One piece was Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 169 Chapter 6. 31HK2510 made of rose quartz, but the other six were composed of metavolcanic material similar in composition and structure to the raw material types described for the chipped stone inventory. Represented in the sample are single pieces of Type I Rhyolite, Uwharries Eastern Rhyolite and Uwharries Southern Rhyolite and three pieces of indeterminate metavolcanic material. If they were brought to the site by natural forces or as unintentional inclusions within rock hearth bundles, it would suggest that Uwharries metavolcanic material could be present in stream terrace formations in the area or in exposures of Triassic Basin sediments crosscut by streams. Rogers (2006) discusses the distribution of Carolina Slate Belt elements and the Triassic Basin in locations near Fort Bragg. Basic definitions of the various lithic artifact classes were presented in Chapter 4. Here, specific descriptions of the 31HK2510 lithic collection are presented. The distribution of cortex on dorsal faces and platforms of debitage provides a basis for further evaluating the nature of reduction strategies associated with the various raw material types at 31HK2510 (Table 34). Dibble et al. (2005:550) indicate that raw nodule reduction from bifaces should yield between about 6 and 21 percent flakes with cortex for any single episode. It is likely that similar proportions characterize directional core reduction from raw nodules. Three of the raw material types exhibit percentages (7.4 to 9.3 percent) within the range of raw nodule reduction. These are Type I Rhyolite Tuff, USR (1) and UWR. All of these have correspondingly high core flake percentages, indicating that the predominant strategies in these cases involved raw nodule and prepared block core reduction to produce flake blanks. White quartz is characterized by extremely high core flake percentages, but low cortical percentages, suggesting a dominance of prepared block core reduction. The remaining metavolcanic types exhibit similar patterns to white quartz, arguing that they were also characterized by prepared block core reduction. Debitage Cores The debitage classification is adapted from formal (Bradley 1973; Frison and Bradley 1980; House and Wogaman 1978; Newcomer 1971), attribute (Moore 2002; Shott 1994), and mass (Ahler 1989; Moore 2002) analysis approaches. Debitage classes were devised to identify both reduction stages and reduction/production systems (ie. biface core reduction, directional core reduction, and flake blank production). This classification was supplemented with the recording of three attributes: (1) size class, (2) percent cortex, and (3) condition. Metric and attribute data on cores are provided in Appendix C. Three classes of cores were identified in the 31HK2510 collection: (1) Directional Cores, (2) Flake Blanks and (3) Core Fragments. Each class is discussed below (Table 35). Lithic Artifacts Table 33 presents the complete inventory (all field stages) of debitage type by raw material type from 31HK2510. Biface reduction types (FBRs and BTFs) comprise the majority of identifiable debitage, but directional core reduction as represented by core flakes also registers a significant presence in the collection. Core Flakes comprise 13 to 49 percent of the identifiable debitage types (i.e. core flakes, FBRs and BTFs) within the metavolcanic and white quartz samples. (1) Directional Cores. Directional, or block, cores consist of relatively large masses of material that exhibit flake scars running in one or more directions from flat striking platforms rather than from biface edges. The platform angle approximates 90°. Directional cores can be further divided into those with scars emanating from only one platform (unidirectional) and cores with multiple platforms (multidirectional). Generally, multidirectional cores represent the final stage of core use as new platforms are formed to extend their lives. In addition, directional cores are sometimes split in two in the later life stages to create thick core flake blanks, either by intention or because of a flake production error. These are called “split cores.” Six directional cores were recognized in the collection, all of which were made of quartz. These consisted of three unidirectional cores (Figure Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 170 Chapter 6. 31HK2510 Table 33. Distribution of Debitage Type by Raw Material Type, 31HK2510. DISTRIBUTION OF DEBITAGE TYPE BY RAW MATERIAL TYPE: 31HK2510 CORTICAL CHUNK FLAKE FRAGMENT FBR CORE REJUVINATION FLAKE CHUNK BTF RAW MATERIAL TYPE CORE FLAKE DEBITAGE TYPES 1 5 1 7 GRAND TOTAL METAVOLCANIC indeterminate metavolcanic Metasedimentary, gray 1 1 Mill Mountain Rhyolite 13 6 32 25 76 Type I Rhyolite Tuff 96 55 217 52 420 Uwharries Eastern Rhyolite 10 3 35 17 65 Uwharries Southern Rhyolite (1) 27 11 54 20 112 Uwharries Southern Rhyolite (2) 27 40 111 58 236 2 11 24 17 54 175 9 249 Uwharries Western Rhyolite QUARTZ White Quartz 28 2 34 1 Rose Quartz 1 1 Fine Grained Quartzite 1 1 Slate 2 2 OTHER TYPES GRAND TOTAL 203 2 161 1 1 657 199 1224 Table 34. Percent Cortex on Debitage and Core Flakes by Raw Material Type, 31HK2510. PERCENT CORTEX AND CORE FLAKES BY RAW MATERIAL: 31HK2510 RAW MATERIAL TYPE Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite TOTAL DEBITAGE PERCENT W/CORTEX PERCENT CORE FLAKES 76 3.9 13.6 420 7.4 27.1 65 3.1 10.0 Uwharries Southern Rhyolite (1) 112 8.9 19.0 Uwharries Southern Rhyolite (2) 236 3.0 32.0 54 9.3 36.7 249 1.6 48.6 Uwharries Western Rhyolite White Quartz TOTAL 1212 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 171 Chapter 6. 31HK2510 Table 35. Summary Data for Cores, 31HK2510. CORES: 31HK2510 SPECIMEN # BAG CORE TYPE PORTION FLAKE TYPE LITHIC RAW MATERIAL TYPE MAXIMUM THICKNESS (MM) MAXIMUM LENGTH (MM) MAXIMUM WIDTH (MM) WEIGHT (GM) 271266a130 91 Multidirectional Core Whole White Quartz 31.37 45.85 30.06 59.25 271266a1219 688 Split Core Whole White Quartz 50.33 50.31 27.63 76.90 271266a1220 688 Split Core Fragment White Quartz 34.56 43.48 25.89 49.10 271266a131 91 Unidirectional Core Whole White Quartz 22.01 49.93 37.86 43.95 271266a132 91 Unidirectional Core Fragment White Quartz 30.34 50.17 24.69 34.16 271266a250 168 Unidirectional Core Whole White Quartz 32.72 49.50 42.98 79.97 271266a126 89 Flake Blank Fragment FBR Uwharries Southern Rhyolite (2) 3.92 25.77 28.06 2.56 271266a493 303 Flake Blank Whole FRB Uwharries Southern Rhyolite (1) 3.39 38.67 34.66 3.98 271266a679 413 Flake Blank Whole Core Flake 271266a45 29 Core Fragment Fragment Uwharries Southern Rhyolite (2) 10.69 39.43 39.55 14.91 Uwharries Western Rhyolite 15.67 21.43 4.59 4.81 271266a116 81 Core Fragment Fragment White Quartz 33.26 17.64 16.49 25.45 271266a998 582 Core Fragment Fragment White Quartz 25.75 44.17 18.14 10.83 271266a1019 591 Core Fragment Fragment White Quartz 19.90 36.93 21.76 18.11 271266a1092 639 Core Fragment Fragment White Quartz 40.85 24.55 13.39 17.46 271266a1123 648 Core Fragment Fragment Uwharries Eastern Rhyolite 25.72 39.98 32.40 42.59 271266a1123 648 Core Fragment Fragment Uwharries Eastern Rhyolite 39.99 30.48 18.01 19.84 271266a1123 648 Core Fragment Fragment Uwharries Eastern Rhyolite 29.42 25.02 13.99 10.29 271266a1214 687 Core Fragment Fragment White Quartz 19.84 26.00 14.42 10.42 50:A-B), one multidirectional core (Figure 50:C) and two split cores (Figure 50:D-E). All of these cores appear to be “spent,” as it would be difficult to remove additional flakes that would be serviceable as tool blanks. Core thickness, the distance from the top of the platform to the base of the core, would define the maximum available length of flake blanks. On these cores, thickness generally ranged from 22 to 35 mm. A thickness of about 50 mm was measured on one of the split cores (Figure 50:D, Specimen a1219). One of the spent unidirectional cores (Figure 50:A, Specimen a132) had a platform composed entirely of water worn cortex, but none of the other specimens exhibited any patches of cortex. Most of these cores were recovered from a small area surrounding a Stage I shovel test excavated at N370/E480. This shovel test produced the multidirectional core (Figure 50:C, Specimen a130), the two unidirectional cores (Figure 50:A-B, Specimens a131 and a132 respectively) and the bit end of a large Type III End Scraper (Specimen a129). These items were tightly packed between 30 and 40 cm bs in the shovel test, suggesting that they might belong to a cache. Short interval shovel testing around this test in SB4 failed to yield additional matching cores or tools, but a test unit (TU1) excavated immediately to the east of it produced the two white quartz split cores (Figure 50:D-E, Specimens a1219 and a1220 respectively) and an additional quartz core fragment. These items were found at 30 to 40 cm bd, suggesting that they might have belonged to the cache as well, or to a more scattered tool cluster associated with a residential occupation. (2) Flake Blanks. Flake blanks were removed from a core for the purpose of further modification or use as a tool (Bradley 1973:6). However, they do not exhibit obvious macroscopic evidence of use or modification. Three flakes fit the physical requirements of this definition. Two of these were derived from large FBRs, while the other was a core flake. The lengths Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 172 Chapter 6. 31HK2510 Figure 50. Cores, Site 31HK2510. A-B: Unidirectional Cores (a132 and a250 respectively), C: Multidirectional Core (a130), D-E: Split Core (a1219 and a1220 respectively), F: Core Fragment (a1123). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 173 Chapter 6. 31HK2510 of the flakes ranged between 25 and 40 mm. The recognized flake blanks in the collection were composed of rhyolite. (3) Core Fragments. Angular pieces of raw material containing portions of one or more flake scars suggestive of directional core reduction, but lacking striking platforms were classified as core fragments. Nine such fragments were recognized in the collection. Included were three fragments from a single, larger block core fragment of UER (Figure 50:F, Specimen a123), a UWR fragment and five white quartz fragments. All of these appear to represent fragments from directional or block core fragments. Bifaces Fourteen biface fragments were recovered during the investigation (Table 36). Four specimens were indeterminate fragments. Bifaces are derived from both large mass packages (i.e. unmodified and/ or prepared cores) and flake blanks. Determination that a biface was derived from a flake blank was made when flake characteristics were observed on the item, such as remnant striking platforms, flake curvature, or bulbs of percussion from flake detachment. The bifaces in the collection, however, do not have sufficient mass to suggest that they were derived from large biface cores, with the exception of a Dalton Adz, which will be discussed below. The remaining specimens appear to have been produced from flake blanks. Each biface category as adapted from Daniel (2002:51-54) is described below. (1) Type I Bifaces. Daniel (2002:51) describes Type I Bifaces as irregularly shaped flake blanks and core masses with large conchoidal scars forming at least one section of a sinuous edge along the core edge. No evidence of secondary thinning scars is observable and a relatively high thickness to width ratio obtains. One Type I Biface (Figure 51:A, Specimen a137) was recognized in the collection. It consisted of the midsection of a relatively narrow biface with only large flake scars exhibited. (2) Type II Bifaces. These bifaces, in the main, represent intermediate stage preform rejects. They are roughly flaked and exhibit only large, con- choidal flake scars. No edge trimming is evident, but an ovate to lanceolate outline is generally recognizable and the perimeter of the edge margin is entirely formed. Facial retouch, however, may not be complete and features of original flake blanks are sometimes evident. Only Two Type II Biface fragments were identified in the collection (Figure 51:B). (3) Type III Bifaces. Type III Bifaces represent late stage preform rejects. The edges exhibit secondary thinning and shaping scars and facial retouch is globally distributed. Outline shapes are well defined and reflect ovate and lanceolate forms. Most of these specimens exhibit step and hinge fractures that terminated production. Type III Bifaces tend to exhibit the lowest ratios of thickness-to-width. Sometimes features of diagnostic projectile point haft elements are present, allowing correlation with specific culture-historic types. In such cases, specimens are differentiated and identified as “Preforms.” Six Type III Biface fragments were identified the collection (Figure 51:C-F). Two of these (Figure 51:E-F, Specimens a390 and a779 respectively) were made on small flakes of USR (1) that exhibited only careful marginal retouch extending not more than 2 mm from the edge. The interior faces do not show extensive bifacial retouch scars. The outlines are not regular, but both specimens appear to represent small triangular preforms. The shortened marginal retouch is consistent with this interpretation. (4) Dalton Adz. A bifacial adz (Figure 51:G, Specimen a302) corresponding to the dimensional and morphological characteristics of a Dalton Adz was recovered from a shovel test in SB6 at a depth of 20 to 30 cm bs. A less well shaped adz identified as a Dalton Adz was recovered during Phase II investigations at 31CD919 (Cable and Cantley 2005a:8889). Dalton adzes are common elements of Dalton phase occupations in northeast Arkansas (Goodyear 1974; Morse and Goodyear 1973). However, similar forms have been recovered from other Late Paleoindian/Early Archaic assemblages on the Atlantic Slope (see Daniel 2002:55). The specimen from 31HK2510 measures 74.98 mm in length, 38.61 mm in width, and 18.29 mm in thickness. Interestingly, none of the Brand Site adzes, of which there were 72, was whole. (Goodyear 1974:41-42). All of these fragments were repurposed as knives, wedges, end scrapers or flake cores. One incidence of a ground adz Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 174 Chapter 6. 31HK2510 Table 36. Summary Data for Bifaces, 31HK2510. BIFACES: 31HK2510 SPECIMEN # BAG TYPE PORTION SHAPE ASSOCIATION LITHIC RAW MATERIAL TYPE Maximum Thickness (mm) Maximum Length (mm) Maximum Width (mm) Weight (gm) 271266a137 96 Type I Biface Fragment Midsection Uwharries Southern Rhyolite (2) 9.10 15.41 25.61 4.10 271266a106 74 Type II Biface Fragment End Ovate White Quartz 7.06 14.03 24.37 1.95 271266a761 456 Type II Biface Fragment Upper Blade Ovate Type I Rhyolite Tuff 271266a75 53 Type III Biface Fragment Upper Blade Lanceolate Uwharries Western Rhyolite 271266a319 205 Type III Biface Fragment Tip Ovate 271266a667 409 Type III Biface Fragment Lateral Section 271266a720 437 Type III Biface Fragment Lateral Section 271266a390 235 Type III Biface/Preform Fragment Base Square 6.40 30.15 23.69 3.97 10.09 30.89 17.77 6.50 White Quartz 7.13 18.93 14.92 1.67 Ovate Uwharries Southern Rhyolite (2) 4.30 33.60 18.31 2.63 Oval Uwharries Southern Rhyolite (2) 8.38 21.95 15.18 2.33 Triangular? Uwharries Southern Rhyolite (1) 3.67 13.64 12.97 0.76 Triangular? Uwharries Southern Rhyolite (1) 2.80 12.98 18.08 0.78 Dalton Type I Rhyolite Tuff 18.29 74.98 38.61 63.57 271266a779 471 Type III Biface/Preform Fragment Base Square/ Concave 271266a302 196 Dalton Adz Whole Ovoid 271266a754 453 Biface Fragment Lateral Section Uwharries Southern Rhyolite (2) 3.81 7.64 13.44 0.43 271266a758 453 Biface Fragment Lateral Section Uwharries Southern Rhyolite (1) 6.97 20.48 11.24 0.95 271266a956 563 Biface Fragment Midsection 271266a1228 692 Biface Fragment Lateral Section Oval Uwharries Southern Rhyolite (2) 5.65 7.54 17.95 0.74 White Quartz 9.20 21.81 33.70 5.67 spall being reused as an end scraper was documented at 31HK1214 during Phase II investigations (Cable and Cantley 2005b:86). The bit end of the specimen from 31HK2510 has been ground into a straight edge. Striations on both faces of the bit run parallel to the bit edge and are probably remnants from sharpening the edge with an abrader. Polish dulling over the striations is likely a consequence of use. Both lateral edges and the butt end have been ground, probably to accommodate a haft. It is generally held that chipped stone adzes of this type were used for heavy-duty wood working actions such as chopping and planing (see Goodyear 1974:41; Semenov 1964:126). Projectile Points Projectile points are bifacially flaked tools with retouched haft elements. In addition, projectile point fragments consist of specimens with the fine bifacial retouch characteristic of finished projectile points, but not necessarily exhibiting evidence of a haft element. The collection from 31HK2510 consisted of nine diagnostic projectile points and 14 frag- ments (Table 37). Metric and descriptive data on the projectile points can be found in Appendix E. (1) Small Dalton. A deeply concave, ground base fragment (Figure 52:A, a160) from a Small Dalton point was recovered at a depth of 20 to 30 cm bs in a Stage II shovel test in SB5. A similar basal fragment was recovered from 31HT947 in a Phase II investigation (Cable 2010:395-397). The form of the upper blade was not observable on this fragment and it was classified as either a Small Dalton or Hardaway Side Notched point. Both of these Late Paleoindian/ Early Archaic point types exhibit deeply recurved bases with outside notches giving the ears a “horned appearance” (see Coe 1964:67). Daniel (1998:52-55) recognized both types in the Hardaway collection. Referring to the seven recognized Dalton points as Small Dalton Points, he notes that they range from classic Dalton forms as described by Morse and Morse (1983) to forms with a somewhat more developed notch that were casually identified as “Hardapalmers” during the 1970 excavations at the Hardaway site because they exhibited characteristics transitional between Hardaway and Palmer points. Daniel notes continuity in the morphological expressions of these early lanceolate and widely side-notched types Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 175 Chapter 6. 31HK2510 Figure 51. Bifaces, Site 31HK2510. A: Type I Biface (a137), B: Type II Biface (a761), C-F: Type III Bifaces (a667, a75, a390, a779 respectively), G: Dalton Adz (a302). spanning the typological variability contained in the Hardaway-Dalton, Small Dalton and Hardaway Side Notched types. Thus, the typological distinctiveness of these types is yet to be fully clarified. The basal width of the specimen from 31HK2510 is 21.94 mm, which falls in the center of the ranges for the set of Small Dalton points that Daniel (1998:55, Figure 4.4) illustrates. The position taken here is that these transitional forms should be assigned to the typological category “Small Dalton.” Dalton points in the Southeast are dated to around 10,500 to 9,900 BP (Goodyear 1982:389-392). (2) Palmer I Corner Notched. Coe (1964:67) described the Palmer Corner Notched type as a “small corner-notched blade with a straight, ground base and pronounced serrations.” Blade edges were generally straight, but were secondarily both incurved and excurvate. Bases were commonly straight and typically ground. Serrations were at times deep, averaging about 3 mm in width and 5 to 7 mm in length. The width of the shoulder barbs usually exceeded that of the base. Metric dimensions ranged between 28 and 60 mm in length, 15 and 25 mm in width, and 5 to 12 mm in thickness. At the Hardaway Site, Palmer Corner-Notched points were primarily associated with Level III, sandwiched between Hardaway SideNotched and Hardaway-Dalton styles in Level 4 and Kirk Serrated and Kirk Corner-Notched in Level II. As an outcome of the Haw River excavations, Cable (1982) further divided the Palmer type Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 176 442 271266a730 275 550 194 109 582 77 49 51 92 240 331 353 412 466 535 536 564 586 645 695 271266a454 271266a926 271266a300 271266a160 271266a997 271266a110 271266a69 271266a73 271266a133 271266a399 271266a529 271266a559 271266a675 271266a773 271266a902 271266a903 271266a960 271266a1005 271266a1101 271266a1233 8 670 271266a1176 271266a13 BAG SPECIMEN No. Yadkin Large Triangular, Var. B Yadkin Large Triangular, Var. E Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Projectile Point Fragment Small Dalton Point Pee Dee Serrated LeCroy Bifurcated Stem Morrow Mountain I Stemmed Palmer I Corner Notched Palmer I Corner Notched Palmer I Corner Notched TYPE White Quartz Uwharries Southern Rhyolite (2) Tip Midsection Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Uwharries Eastern Rhyolite Tip Tip Tip/Ear Tip/Ear Tip Midsection Tip Base Blade Midsection Lateral Section Mill Mountain Rhyolite Uwharries Western Rhyolite Uwharries Southern Rhyolite (1) Tip Midsection Upper Blade White Quartz Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Western Rhyolite Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Type I Rhyolite Tuff LITHIC RAW MATERIAL TYPE Tip Tip Base Base Base Whole Whole Whole Upper Blade Missing Base/Should er Lateral Sec/Base PORTION WEIGHT (gm) 0.22 0.39 0.95 0.98 0.72 0.64 0.48 4.84 0.94 3.16 2.48 0.34 0.12 0.14 2.95 3.36 2.77 0.63 7.17 3.68 2.61 6.02 4.37 MAXIMUM THICKNESS (mm) 3.29 3.49 6.39 4.66 4.98 5.36 4.16 6.79 7.16 7.41 2.55 2.90 3.13 1.67 5.13 7.14 6.11 3.50 7.48 6.02 6.40 6.65 5.83 MAXIMUM LENGTH (mm) (8.46) (15.07) (11.72) (14.96) (11.85) (10.76) (9.79) (29.84) (9.91) (25.01) (24.76) (12.72) 7.61 (12.34) (22.35) (25.96) (15.70) 16.57 44.53 32.13 (26.58) (33.60) (29.25) 19.00 (25.18) 23.71 21.94 10.22 18.30 19.88 16.03 BASAL WIDTH (mm) PROJECTILE POINTS: 31HK2510 TANG WIDTH (mm) 11.02 15.52 16.92 13.13 15.06 SHOULDER WIDTH (mm) 13.31 22.65 21.13 24.91 8.44 11.13 7.47 7.21 21.53 10.36 TANG LENGTH (mm) Table 37. Summary Data for Projectile Points, 31HK2510. BLADE LENGTH (mm) 33.40 24.66 BASAL GRINDING Present Present Present Present EDGE ANGLE 46° 58° 49° 55° 82° 56° 59° 53° 48° 49° 21° 40° 69° 60° 53° 58° 60° 64° 59° Present BLADE SERRATION One Margin 23° 39° 73° 57° Deeply Concave Rounded Rounded Straight Straight Straight BASE SHAPE Chapter 6. 31HK2510 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 177 Chapter 6. 31HK2510 Figure 52. Projectile Points, Site 31HK2510. A: Small Dalton (a160), B-D: Palmer I Corner Notched (a926, a454, a13 respectively), E: LeCroy Bifurcated Stem (a1176), F: Morrow Mountain I Stemmed (a730), G: Yadkin Large Triangular, Variant B (a997), H: Yadkin Large Triangular, Variant E (a110), I: Pee Dee Serrated (a300). into three variants, Palmer I, II and III. The Palmer I style is smaller, typically exhibits ground bases and was found in a slightly lower vertical position at Haw River than the larger variants described as Palmer II and Palmer III. The latter is the largest of the three variants and is characterized by a low incidence of basal grinding. It would correspond closely to the ideal type description for the Kirk Corner Notched type (Coe 1964:69-70). Palmer II is stylistically intermediate and it may overlap the temporal distributions of the other variants. It is larger than Palmer I specimens, but is characterized by a high incidence of basal grinding. Others suggest that all of this variation is largely a consequence of life-stage variation rather than temporal differences (Daniel 1998; Sassaman 2002). Although this may be the case for Palmer II and III, this explanation does not take into account the much narrower tang widths of Palmer I specimens. The time range of this style is generally accepted at 9500–8900 BP (Sassaman et al. 2002:10). Three points from 31HK2510 were identified as Palmer I Corner Notched (Table 37). Specimen a926 (Figure 52:B) was found between 30 and 40 cm Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 178 Chapter 6. 31HK2510 bs in a Stage III shovel test excavated in SB17. It was manufactured from USR (2) and the base was straight and heavily ground. The blade outline was slightly convex and both margins were serrated. Specimen a454 (Figure 52:C) exhibited a relatively wide tang width relative to blade length, which is a characteristic of Palmer II and III styles, but its small overall size places it in the Palmer I type. It was recovered from a Stage II shovel test excavated in SB9 at a depth of 30 to 40 cm bs. It too was fashioned from USR (2). The base was straight and heavily ground, but the central part of the basal margin had been indented by flaking from either use or intentional thinning. The nonserrated blade margins exhibited a slightly concave outline due to resharpening. One shoulder had been broken off, but had been repaired through marginal retouch applied to one face. Specimen a13 (Figure 52:D) represented a laterally snapped fragment consisting of a partial base, one shoulder and a portion of one blade margin. It was found at a depth of 30 to 40 cm bs in a Stage I shovel test that became part of SB1. The specimen was also made of USR (2) and exhibited a non-serrated, convex blade outline on the observable margin. The base appears to have been straight and exhibited heavy basal grinding. (3) LeCroy Bifurcated Stem. A single LeCroy Bifurcated Stem point (a1176) was recognized in the 31HK2510 collection (Figure 52:E). This point type has been described by Lewis and Kneberg (1955), Kneberg (1956:27-28), and Broyles (1966:27). The stem is deeply notched, but in contrast to St. Albans Side-Notched, the basal ears are pointed and directed nearly straight downward from the shoulder. Blade edges are typically straight and serrated, sometimes with sharp barbs. Basal grinding is absent. The LeCroy Bifurcated Stem type is widely distributed across the Mid-Continent (see Justice 1987:92), but is probably more numerous along the Southeast Atlantic Slope than commonly believed. Chapman (1985:Table 7-1) dates the type to 6500 to 5800 radiocarbon years BC, which indicates that it post-dates Palmer/Kirk Corner Notched and St. Albans Side Notched and pre-dates Kirk Stemmed and Serrated. Specimen a1176 was recovered from TU3 in SB9 at a depth of 40 to 50 cm bd. It was made of nonpatinated Type I Rhyolite Tuff (Subtype R2), which is a commonly used raw material type to make Bifurcate Tradition point styles in the region. The upper blade was missing, removed from the body of the point with a snap fracture. The blade margins appear to have been straight in outline and heavily resharpened. One shoulder had been broken and efforts had been made to repair it by marginal retouch applied to one face. (4) Morrow Mountain I Stemmed. A single Morrow Mountain I Stemmed point (a730) was identified in the collection from 31HK2510 (Figure 52:F). The Morrow Mountain I Stemmed type was originally defined by Coe (1964:37-43) from his excavations at the Doershuk Site and it is similar in morphology to the Morrow Mountain II Stemmed point. Both styles were characterized by contracting stems, but two differences were cited. The blade is proportionately longer and narrower and a noticeable angle occurs at the shoulder/stem juncture in the Morrow Mountain II Stemmed type. The shoulder of the Morrow Mountain I style, by contrast, curved into the stem without a break in the marginal contour. There is some evidence to suggest that Morrow Mountain II occurred at a slightly higher stratigraphic/vertical position than Type I in the Doershuk Site excavations, which led Coe (1964:43) to tentatively posit that the former was “transitional and survived to a later date.” This type is similar in length to the Morrow Mountain II Stemmed type (30–80 mm), but is noticeably broader at the shoulder (18–30 mm). The specimen from 31HK2510 consists of a basal fragment in which the upper blade was snapped off at the approximate mid point of the blade. It was found at a depth of 30 to 40 cm bs in a Stage III shovel test excavated in SB7. It was made of USR (2) material. One side of the blade dipped toward the central spine of the point at a noticeably different contour than the opposite side, indicating that the blade had been repaired prior to the upper blade being snapped off. The very tip of the base was missing, having been detached from the stem with a hinge fracture. Basal ear fragments from Morrow Mountain Stemmed points, owing to the contracting outline of the stem, take on a similar appearance to blade tips and are often times confused. Two possible basal ear fragments from Morrow Mountain Stemmed points were identified in the collection (see Table 37). (5) Yadkin Large Triangular. Coe (1964:4549) defined the Yadkin Large Triangular type from excavations conducted at the Doerschuk Site. The Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 179 Chapter 6. 31HK2510 type was described as “a large, symmetrical, and well made triangular point.” Bases were usually concave and blade margins ranged from excurvate to straight to incurvate. Coe (1964:45) observed that all specimens were finished by careful pressure flaking. As is true of the Badin type, a great deal of morphological variation is contained within the Yadkin Large Triangular type description. Although this variation may simply reflect life cycle stages within a single constellation of typological traits, it is just as possible that it represents a chronological sequence of design style changes. If the stratigraphic and spatial dimensions of this variation are not monitored through typological recognition, this hypothesis will remain inadequately addressed. For this reason, Cable (2010:416) has proposed that various shape configurations of the Yadkin type be segregated into several variants, as was also done for the Badin type. Four variants are identified in the plate of Yadkin points in the Doerschuk Site report (Coe 1964: Figure 42): (1) Variant A: triangular blade w/excurvate blade margins and concave base, (2) Variant B: triangular blade with straight blade margins and concave base, (3) Variant C: triangular blade with incurvate blade margins and concave base, and (4) Variant D: Lanceolate blade with excurvate blade margins and concave bases with side notched ears. Two specimens were typed as Yadkin Large Triangular points from the 31HK2510 collection. Specimen a997 corresponds to Variant B (Figure 52:G). It exhibits a straight outline on one blade margin, while the other effects a concave outline as a consequence of partly unsuccessful attempt to thin the bulb of percussion and platform of the original flake blank. The upper blade was broken off in an irregular step fracture. It was crudely shaped with large percussion flakes. The specimen was found at a depth of 10 to 20 cm bs in a Stage III shovel test excavated in SB18. The other specimen (Figure 52:H, a110) was asymmetrical due to breakage repair, but the unrepaired side exhibited extremely straight blade and base margins. The damaged side had been shortened by a large percussion flake applied to one side of the face, removing the original blade/base juncture. The upper blade was broken off in a hinge fracture. The faces of the point exhibited large percussion flake scars, but the margins were shaped with care- ful marginal retouch, some of which appeared to have been produced pressure flaking. Its dimensions were similar to the Variant B specimen, indicating that it belongs in the range of variation associated with larger, Middle Woodland triangular points. Consequently, it was assigned to another Yadkin Large Triangular variant category, Variant E. It resembles the straight-based examples of the Roanoke triangular point described by South (2005:62). South noted that some of the “smaller and more equilateral” Roanoke points might belong to an intermediate form bridging the transition from Roanoke to the later, small triangular Clarkesville style. Thus, Variant E may represent a form transitional between Yadkin Large Triangular and early Late Woodland triangular points of intermediate size in the Sandhills. The point was recovered at a depth of 20 to 30 cm bs in a Stage I shovel test that was later incorporated into SB2. (6) Pee Dee Serrated. This type has not been formally defined, but has been commonly used and applied ion the literature. Coe (1964:49) indicated that the form occurred in association with the Pee Dee Pentagonal type. It was described as a distinctive, small triangular point with concave blade edges that were commonly serrated. Later usage has come to distinguish non-serrated and serrated forms into respectively two types, Pee Dee Triangular and Pee Dee Serrated (Coe 1995:202-203). A review of the plate of examples in Coe (2005:202, Figure 10.8) indicates that Pee Dee Serrated is characterized by slightly concave, slightly concave and straight blade margins. Coe dates the types to Late Woodland/Mississippian period, ca. AD 1300. A single specimen (Figure 52:I, a300) of the type was recovered from investigations at 31HK2510. The specimen was whole and exhibited slightly convex and straight blade margins, the straight one of which was finely serrated by pressure flaking. The base was slightly concave. The specimen was fashioned from UER material. It was found at a depth of 0 to 10 cm bs in a Stage II shovel test belonging to SB6. Unifaces Unifaces consist of tools that are retouched along flake margins in a single direction (unifacial Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 180 Chapter 6. 31HK2510 edge shaping). Retouch was applied to shape a working edge or bit and, sometimes, a butt element for hafting. Limited bifacial retouch or resharpening is sometimes present along haft elements or working edges. Metric and attribute data for unifaces are presented in Appendix F. Twenty-five stone tools classified as unifaces were recognized in the collection (Table 38). These include 5 end scrapers, 19 side scrapers and a retouched flake. The specimens are described below under the rubric of the typological system developed by Coe (1964) and elaborated by Daniel (1998). (1) End Scrapers. Five end scraper specimens were identified in the 31HK2510 collection, of which one fragment was not identifiable to type. Daniel (1998:66) defines end scrapers as unifaces with “steep, regular unifacial retouch that forms a relatively narrow convex bit or working edge.” He further notes that the bit edge is commonly positioned in a transverse orientation to the long axis of the tool. Four main types and a series of subtypes have been recognized; including both hafted and unhafted forms. (a) Type II End Scrapers. A single Type IIb End Scraper (a150) was identified during the investigation (Figure 53:A). Coe (1964:76) defined the Type II End Scraper from his excavations at the Hardaway Site. The type was defined as hafted end scrapers made on flakes of variable shapes and sizes that were retouched only along the narrowest margins to form scraping edges. He recognized two varieties that Daniel (1998:72) later formalized. Type IIa End Scrapers consisted of large, thick irregular flakes that were casually shaped at one end to produce a scraper edge. Cutting edges were always irregular in marginal outline and sharp, suggesting to Coe that they were used to work hard material such as wood or bone. They were primarily confined to Level II at the Hardaway Site, which was primarily associated with Kirk Stemmed and Serrated and Stanly Stemmed points. Type IIb End Scrapers were made on thin, narrow, prismatic flakes. The scraping edge was formed along the end opposite the striking platform. This variety was not numerous, but it had the same stratigraphic association as the Type IIa variety. Daniel (1998:72-73) expanded the type definitions to also include specimens with limited lateral margin retouch. Specimen a150 was made of a fine-grained UER material. The bit edge of the tool was actually constructed on the platform of the flake blank. The use of the platform to make a bit edge was not recognized in the Hardaway collection, but the specimen is included here as an example of a Type IIb End Scraper because all of its other morphological characteristics conform to the type description. The use of the platform of a blank to form a bit edge was also noted at 31HK2502, where a much thicker flake was treated in this manner to produce a Type IIb End Scraper. The use of a platform to create a bit edge on a Type IIb End Scraper has also been documented at 31HT844 within the confines of Fort Bragg (Cable and Cantley 2006:311). Retouch on other margins of the tool was limited to one lateral edge, presumably to accommodate hafting. The bit is slightly convex in contour and the working edge exhibits extensive dulling from use. Minimal bifacial resharpening scars were observed on the ventral surface of flake along one end of the bit edge. A graver spur occurs along the juncture of the bit edge and the retouched lateral edge, opposite the bifacially reworked end of the bit. The specimen was recovered from a Stage I shovel test (N410/E530) at a depth of 30 to 40 cm bd. (b) Type III End Scrapers. The bit of a Type III End Scraper (a129) made of white quartz (Figure 53:B) was recovered in a Stage I shovel test in what would later become SB4. It was found at a depth of 30 to 40 cm bs in association with a cache or cluster of white quartz cores (see discussion above). Coe (1964:76) defined this type as a rough and larger version of the finely made Type I End Scraper, or “tear-drop” end scraper. Daniel (1998:75) notes that Type III End Scrapers at Hardaway exhibit the widest median bit value (42.4 mm) of all end scrapers, suggesting to him that this tool needed more lateral and dorsal retouch to form a functional haft element. The bit width of the specimen from 31HK2510 achieved a maximum dimension of 43.95 mm, which is well within the range for the type. The haft element, or butt end, of the specimen was broken off in a step fracture termination, suggesting that it may have been recycled as a potential core. The remaining portions of both lateral margins exhibited retouch for accommodating a haft. The bit end was steeply retouched forming a convex working surface with an edge approaching 90°. One end of the working edge had been severely undercut and was no longer service- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 181 355 359 105 230 295 405 541 245 361 445 460 557 271266a567 271266a156 271266a379 271266a484 271266a657 271266a910 271266a406 271266a570 271266a741 271266a766 271266a942 687 271266a1213 271266a562 327 271266a524 205 297 271266a486 271266a320 103 271266a151 97 51 271266a72 143 554 271266a933 271266a219 338 271266a543 271266a138 91 282 271266a468 103 271266a150 271266a129 BAG SPECIMEN NO. Retouched Flake Side Scraper Fragment Side Scraper Fragment Side Scraper Fragment Side Scraper Fragment Type IV Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type I Side Scraper Type I Side Scraper Type I Side Scraper Type I Side Scraper Type I Side Scraper End Scraper Fragment Type IV End Scraper Type IV End Scraper Type III End Scraper Type IIb End Scraper UNIFACE/RETOUCHED FLAKE TYPE PORTION BLANK TYPE Fragment Fragment Fragment Lateral Section Lateral Section Midsection Whole Fragment Whole Whole Whole Lateral Section Core Flake Blade Flake Core Flake Core Flake Core Flake Core Flake Trapezoidal Trapezoidal Lenticular Lenticular Trapezoidal Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Fragment Lenticular Trapezoidal Lenticular Lenticular Trapezoidal Trapezoidal Lenticular Lateral Section Core Flake Core Flake Core Flake Core Flake Core Flake Levallois? PlanoConvex PlanoConvex CROSS-SECTION Fragment Fragment Fragment Fragment Midsection Whole Bit Fragment Whole Whole Bit End Whole Uwharries Western Rhyolite Uwharries Western Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Crystal Quartz Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) White Quartz Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Eastern Rhyolite White Quartz Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Type I Rhyolite Tuff White Quartz Uwharries Eastern Rhyolite LITHIC RAW MATERIAL TYPE RETOUCH LENGTH (mm) (14.24) (16.40 (7.78) (5.49) 10.70 (12.25) 9.89 (12.72) 13.24 15.21 35.95 (20.88) (13.98) (15.69) (17.02) (27.93) 24.38 32.37 (6.37) 26.67 (4.14) 28.53 35.81 43.95 20.59 35° 64° 61* 48° 60° 60° 64° 59° 69° 54° 60° 73° 53° 35° 50° 72° 67° 78° 53° 69° 56° 75° 65° 87° 68° MINIMUM EDGE ANGLE SCRAPERS AND RETOUCHED FLAKES: 31HK2510 44° 65° 99° MAXIMUM EDGE ANGLE NUMBER OF RETOUCHED EDGES (1) (1) (1) (1) (1) 2 1 (1) 1 1 2 1 (1) (1) (1) (1) 1 (1) (1) 1 (1) 1 1 3 2 6.66 (6.75) (3.17) (3.13) (5.58) 7.44 4.25 (2.41) 1.52 2.46 5.29 (5.94) 4.90 (3.68) 3.74 (12.10) 18.85 12.32 7.79 8.88 (3.90) 12.99 4.99 15.10 4.08 MAXIMUM THICKNESS (mm) 182 (26.26) (15.80) (13.15) (11.84) (13.90) (20.08) 16.08 (14.07) 20.35 20.43 37.63 (24.67) (18.10) (16.43) (20.97) (27.93) (32.60) (51.42) (15.84) 31.04 (14.91) 33.95 36.44 (34.09) 21.60 MAXIMUM LENGTH (mm) Table 38. Summary Data for Scrapers and Retouched Flakes, 31HK2510. MAXIMUM WIDTH (mm) 30.02 (12.56) (16.63) (9.68) (15.65) 26.79 18.68 (7.93) 14.84 12.10 27.46 (20.70) (16.19) (14.21) (20.62) (19.96) (36.12) (44.13) (20.69) 24.57 (7.05) 45.88 29.71 41.75 20.59 WEIGHT (gm) 7.41 1.06 0.84 0.26 0.84 5.21 1.52 0.21 0.50 0.72 4.72 2.04 1.22 0.77 1.51 5.30 10.04 21.09 2.89 8.58 0.38 21.91 6.10 29.85 1.66 Maybe Same Tool as a910, Bifacial Retouch Maybe Same Tool as a942, Bif Edge Opposite Scraper Slightly Concave Working Edge Cortical Face Usedc as Base of Bit Two Graver Spurs Bit Edge Dulled, 1 Lat Edge Retouched, 1 Graver Spur NOTES Chapter 6. 31HK2510 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC Chapter 6. 31HK2510 Figure 53. Unifaces, Site 31HK2510. A: Type IIb End Scraper (a150), B: Type III End Scraper (a129), C-D: Type IV End Scrapers (a468 and a543 respectively), E-G: Type I Side Scrapers (a72, a486, a524 respectively), H: Type IIa Side Scraper (a567), I-J: Type IIb Side Scrapers (a379 and a657 respectively), K: Type IV Side Scraper (a910), L: Retouched Flake. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 183 Chapter 6. 31HK2510 able as a scraper edge. Both ends of the bit exhibit short projections that may have functioned as graver spurs. Although edge rounding was a common occurrence on most end scraper types at the Hardaway Site, Daniel (1998:79-83) noted that none of the Type III end scrapers exhibited this wear pattern. The specimen from 31HK2510 exhibited a sharp bit edge, as well. Daniel suggests that edge dulling and rounding on end scrapers may evidence the scraping of, in particular, dry hides. Other functions proffered for scrapers not exhibiting edge rounding were scraping of resistant material. However, since some examples within each of the other end scraper types also failed to exhibit edge dulling, it is likely that much of this variation is explained by the state of resharpening at the time of discard. Coe (1964:76) observed that all of the edges on Type III End Scrapers were irregular and sharp. Scraping of hard surfaces would not maintain a sharp edge. (c) Type IV End Scrapers. Two Type IV End Scrapers were identified in the 31HK2510 collection. Daniel (1998:75-76) extracted this type from the Hardaway scraper collection to segregate a group of informal end scrapers that he believed were handheld during use. Type IV End Scrapers are predominantly unshaped with the exception of the bit end and they exhibit wide bits. He notes, however, that approximately one half of the specimens displayed restricted retouch on edge margins near the bit. Daniel further observes that blank selection for Type IV End Scrapers was quite variable, suggesting that few design restrictions were placed on their manufacture. Moreover, bit placement was, most often, lateral to the striking platform of the blank, not opposite to it. The overall morphological characteristics suggest to Daniel that the tool type was manufactured on an ad-hoc basis to be used in an immediate and specific task. In other words, this type was considered to represent an expedient end scraper form. Specimen a468 was made of Type I Rhyolite and the dorsal face of the blank was nearly completely covered in cortex (Figure 53:C). The bit edge was created opposite the striking platform of the blank, which was derived from a directional core. In this characteristic, the tool is more like a Type I or II end scraper, but no attempt was made to retouch and shape the lateral margins or butt end of it. The working edge measured 35.81 mm in width and exhibited the careful and fine retouch of Type I and II end scrapers. The working edge was convex in outline and its central portion displayed extensive edge dulling from use. It was recovered at a depth of 70 to 80 cm bs in a Stage II shovel test excavated at N405/E520. Specimen a543 was found in a Stage II shovel test excavated in SB4 at a depth of 30 to 40 cm bs. It was made from a large core flake composed of USR (1) material (Figure 53:D). The bit was formed opposite the striking platform of the flake blank and the width of retouch was relatively narrow, extending over a distance of only about 28.53 mm. One lateral margin was minimally retouched on the ventral face of the flake, opposite the direction of retouch that formed the bit edge. This lateral margin may have served as a cutting edge as it would not have accommodated a haft. However, informal notches made by single percussion flakes on each of the lateral margins may have provided features to attach a haft. The working edge was moderately dulled from use. (2) Side Scrapers. Nineteen side scraper specimens were identified in the 31HK2510 collection, of which 15 were sufficiently complete to type further. Daniel (1998:6683-84) defines side scrapers as unifaces containing “one or more unifacially retouched working edges that parallel the long axis of the flake blank.” Four side scraper types and numerous subtypes were identified by Daniel in the Hardaway scraper collection. Side scrapers from 31HK2510 can be placed in Types I, II and IV. (a) Type I Side Scrapers. Five Type I Side Scrapers were identified in the collection (Figure 53:E-G). Coe (1964:77) defined this scraper type as being made from large, wedge-shaped flakes produced from directional cores. One of the long sides of the flake blank was subjected to steep percussion retouch, while one or both ends of the blank also received retouch shaping. Daniel (1998:84) further adds that the bit edges were generally convex to straight in shape and cross-sections appeared wide and flat (lenticular). Type I Side Scrapers were equally distributed between Levels II, III and IV at the Hardaway Site, in association with Kirk and Palmer Corner Notched and Hardaway Side Notched points and Hardaway Blades. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 184 Chapter 6. 31HK2510 All of these specimens exhibit substantial working edges and dulling along the full length of the retouched margins. Edge angles were steep, ranging between 53° and 72° and edge outlines ranged from straight to convex. Daniel (1998:84) suggests that this scraper type might have been hafted, but haft modification away from the working edge is not commonly noted. The specimens from 31HK2510 were recovered in relatively deep vertical positions consistent with an Early Archaic association. Three specimens (a151, a524 and a1213) were found between 30 and 40 cm bs (see Figure 53:G for a524), one (a486, Figure 53:F) was recovered at 50 to 60 cm bs and one (a72, Figure 53:E) was positioned at 60 to 70 cm bs. (b) Type II Side Scrapers. Building on Coe’s (1964:79) original description for the Type II Side Scraper, Daniel (1998:84-93) separated individual specimens further into two types based on the size of the flake blanks used to manufacture the tools. Type IIa refers to the classic description of a Type II Side Scraper. That is, scrapers made on large, irregular flakes whose bit edges were retouched without concern for regularizing the marginal outline. Daniel’s Type IIb was reserved for very small flakes, usually of less than 30 mm in length. The generally thin edges on these scrapers suggest that they were used for more delicate uses and on softer materials than Type I Side Scrapers. Coe (1964:79) observed that Type II Side Scrapers were strongly associated with the Palmer and Hardaway levels (III and IV) at the Hardaway Site. Five Type IIa Side Scrapers were recognized in the 31HK2510 collection (Figure 53:H). They were generally manufactured on core flakes from directional cores. Angles on retouched edge margins ranged between 35° and 73°. One specimen exhibited two retouched edges (a567), while the remainder displayed only one. All of the working edges were delicately retouched by pressure flaking and at the time of their discard, all bit margins exhibited light dulling. Individual specimens were predominantly recovered from vertical positions between 40 and 70 cm bs, although one specimen (a320) was found at only 10 to 20 cm bs. The great depths of the Type IIa scrapers is in line with an Early Archaic association. Four Type IIb Side Scrapers were recovered (Figure 53:I-J). These specimens were fashioned from small, thin flakes that were predominantly derived from directional cores. Bit margins were delicately retouched with pressure flaking and working edge angles ranged between 54° and 69°. Each specimen exhibited only one retouched lateral edge and bit margins were lightly dulled. Type IIb Side Scrapers were found at higher vertical positions than the other scraper types at 31HK2510. The sample is too small to demonstrate an actual difference in vertical distributions, but the sample of specimens were found between 0 and 40 cm bs and three were found above 30 cm bs. (c) Type IV Side Scrapers. Daniel (1998:95) recognized a previously undefined side scraper type from the Hardaway collection that he referred to as a Type IV Side Scraper. This scraper type was very similar in size and bit morphology to the Type I Side Scraper. However, in these specimens the side opposite the scraper edge is bifacially retouched. Daniel surmises that the bifacial retouch was applied to prepare the scraper for hafting. The fact that the bifacial edge was oriented parallel to the bit edge suggested to Daniel that the scraper was hafted in such a way as to allow the haft to be held by both hands in a draw-knife fashion. However, the bifacial edge on the specimen from 31HK2510 (a910, Figure 53:K) would not have required thinning to accommodate a haft. It is possible that the bifacial retouch was meant to “back” the tool to prepare it for hafting. Alternatively, this specimen the bifacial edge on this tool might have been used for cutting in the operation of a non-hafted, hand-held tool. Retouch on the scraper bit displays a denticulated outline, which may suggest that the tool should be reclassified as a denticuate. The item is too fragmentary to resolve this issue, however. It was found in a high vertical position, at a depth of 10 to 20 cm, in a Stage III shovel test in SB17. (3) Other Retouched Flake Tools. One other retouched flake tool was identified in the collection (a942, Figure 53:L). It exhibited a single sharp edge with informal bifacial retouch along the margin. It was recovered very near the Type IV Side Scraper (a910, Figure 53:K) discussed above in SB17 and it appears to be part of the same tool. The raw material is identical and the general morphology of the dorsal surface of the flake matches, but the items could not be precisely refit. It was found in a shovel test about 2.5 m distant from scraper, but at a vertical depth of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 185 Chapter 6. 31HK2510 40 to 50 cm bs. The scraper fragment was found at 10 to 20 cm bs, suggesting that the area of SB17 has been subjected to a good deal of vertical displacement of artifact bearing deposits. Utilized Flakes Twenty-eight non-retouched flakes exhibited wear patterns on one or more edges, indicating their use as expedient tools (Table 39). Metric and attribute data for utilized flakes are presented in Appendix G. All of these specimens exhibited edge attrition and bifacial nibbling scars indicative of cutting functions. Utilized edge angles fell between 18°and 60°. Most of the specimens appear to have been derived from directional cores (i.e. core flakes and blade flakes), but one represented a large elongated FBR similar in appearance to a blade flake. Aphyric raw material types (USR (1), USR (2) and Type I Rhyolite Tuff) dominate the collection. Only three specimens were made of porphyritic raw material types (UER and MMR). Use wear would be easier to detect on these microcrystalline rock types, however, which suggests that the porphyries might be under-represented because of observational difficulties. Miscellaneous Stone Tools Several non-chipped stone, lithic artifacts were recovered from the investigation at 31HK2510. These consisted of two grinding stone/abrader fragments, a metasedimentary slab or tabular fragment, a hammer stone spall and a hammer stone/anvil stone cobble tool. Metric and attribute data for these artifacts are presented in Appendix H. (1) Grinding Stones/Abraders. Two fragments of specimens that appear to have grinding or abrading facets and dull polish were recovered during the investigation. One specimen (a849, Figure 54:A) of arenite exhibited what appears to be a flat surface with grinding facets. It was recovered at a depth of 40 to 50 cm bs in a Stage III shovel test excavated in SB9. It measured 42.75 x 74.10 x 38.47 mm and weighed 148.55 gm. The faceted zone, which was only partially represented on the fragment, measured about 75 x 25 mm. The other specimen (a341) was a small piece of sandstone conglomerate with an extremely flat and faceted surface on one side. It weighed 7.16 grams and was found in a Stage II shovel test (N365/E495) at a depth of 30 to 40 cm bs. Although such items could have been used to grind vegetal material, they more likely represent abraders for stone tool manufacturing. Similar ferruginous sandstone abraders with flat working faces were commonly recovered in the Dalton deposits at the Brand Site in northeast Arkansas (Goodyear 1974:69-73). Goodyear suggested that the flat abraders, as opposed to the grooved ones, functioned in bone working, but they may also have been used to grind the faces of stone tools such as adzes. Daniel (1998:116-117) has proposed similar functions for the abraders in the Hardaway collection. The depth at which the abraders were recovered at 31HK2510 indicates the likelihood that they are associated with Archaic occupations. (2) Slab/Tabular Fragment. A large slab-like river cobble fragment (a580, Figure 54:B) of greenish metasedimentary material was recovered in a Stage III shovel test in SB1 from a depth of 40 to 50 cm bs. The relative fineness of the grain suggests that the material can be further classified as metasiltstone. It measured 16.11 x 93.60 x 70.65 mm and weighed 173.97 gm. The dorsal surface of the fragment consisted of water worn cortex, while the ventral face had been sheared across a laminar fracture plane. Use striations or any other evidence of modification was not apparent. The item may have been stored as a core at the site for future use. (3) Hammer Stone Spall. A flake of finegrained quartzite (a1058, Figure 54:C) in which the dorsal surface consisted of cobble cortex was recovered at a depth of 20 to 30 cm bs in a Stage III shovel test excavated in SB13. Battering wear was present on the curve of the dorsal surface, indicating that the flake probably cleaved off a hammer stone during percussion strikes. The ventral surface of the specimen did not exhibit a bulb of percussion or a platform. The spall weighed 13.68 gm. (4) Hammer Stone/Anvil Stone. A large river cobble of fine-grained quartzite with evidence of hammer battering, grinding and anvil marks (a288, Figure 55) was found in a Stage II shovel test in SB 7 at a depth of 30 to 40 cm bs. The cobble was flat, measuring 57.40 mm in thickness and 140.23 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 186 Chapter 6. 31HK2510 Table 39. Summary Data for Utilized Flakes, 31HK2510 MAXIMUM THICKNESS (mm) MAXIMUM LENGTH (mm) MAXIMUM WIDTH (mm) WEIGHT (gm) EDGE ANGLE 2.52 35.88 18.52 1.65 37° 13.76 24.98 40.28 10.82 42° NO. UTILIZED EDGES UTILIZED FLAKE FUNCTION UTILIZED FLAKES: 31HK2510 LITHIC RAW MATERIAL TYPE 271266a41 27 Whole Cutting Blade Flake 1 Uwharries Southern Rhyolite (1) 271266a118 83 Whole Cutting Core Flake 1 Uwharries Eastern Rhyolite 271266a181 120 Fragment Cutting Blade Flake 1 Uwharries Southern Rhyolite (1) 3.08 (20.23) 12.69 0.72 43° 271266a195 128 Whole Scraping Core Flake 1 Uwharries Southern Rhyolite (2) 7.21 18.08 27.16 3.39 60° 271266a197 129 Fragment Cutting Blade Flake 2 Uwharries Southern Rhyolite (1) 2.58 (20.95) 13.02 0.73 24° 271266a276 181 Whole Cutting Blade Flake 1 Mill Mountain Rhyolite 4.72 (32.20) 18.04 2.41 41* 271266a279 182 Fragment Cutting Blade Flake 2 Type I Rhyolite Tuff 2.69 (29.58) 14.26 1.1 33° 271266a290 186 Fragment Cutting Blade Flake 1 Type I Rhyolite Tuff 4.15 (21.28) 16.65 1.47 37° 271266a357 226 Whole Cutting Core Flake 1 Uwharries Southern Rhyolite (2) 4.28 27.64 22.38 2.91 29° 271266a432 265 Whole Cutting FBR 1 Uwharries Southern Rhyolite (2) 2.31 20.64 17.92 0.82 36° 271266a496 305 Whole Cutting Core Flake 1 Uwharries Southern Rhyolite (1) 3.38 26.86 13.70 0.97 40° 271266a590 378 Whole Cutting Core Flake 271266a592 379 Fragment Cutting 271266a599 382 Fragment Cutting SPECIMEN # BAG CONDITION BLANK TYPE 1 Uwharries Southern Rhyolite (1) 5.03 22.85 20.52 2.09 36° 1 Uwharries Southern Rhyolite (1) 2.47 (17.99) (11.72) 0.36 19° Core Flake 2 Type I Rhyolite Tuff 5.52 (18.09) 16.79 1.7 28° FBR/Blade Flake 1 Type I Rhyolite Tuff 4.86 42.29 16.72 2.65 23° 1 Type I Rhyolite Tuff 2.62 (16.58) 9.76 0.4 27° 3 Type I Rhyolite Tuff 5.47 28.11 27.01 3.03 21° 2 Type I Rhyolite Tuff 3.83 (20.21) 25.75 1.86 33° Type I Rhyolite Tuff 3.25 18.63 24.37 1.00 31° 271266a613 387 Whole Cutting 271266a718 435 Fragment Cutting 271266a725 441 Whole Cutting 271266a777 469 Fragment Cutting 271266a792 476 Whole Cutting Core Flake 1 271266a797 480 Whole Cutting Core Flake 1 Uwharries Southern Rhyolite (1) 6.16 33.48 20.68 3.7 33° 271266a915 543 Whole Cutting 1 Type I Rhyolite Tuff 4.05 (29.56) 9.56 1.09 37° 271266a934 554 Whole Cutting Core Flake 2 Type I Rhyolite Tuff 2.91 41.95 24.08 1.90 18° 271266a945 558 Whole Cutting Core Flake 1 Type I Rhyolite Tuff 7.91 17.10 18.07 1.54 52° 271266a948 560 Whole Cutting Core Flake 1 Type I Rhyolite Tuff 2.94 18.87 11.78 0.54 26° 271266a1020 591 Whole Cutting Core Flake 2 Type I Rhyolite Tuff 3.33 29.04 33.47 3.08 28° 271266a1027 592 Whole Cutting Core Flake 2 Type I Rhyolite Tuff 5.37 45.72 31.32 6.79 30° 271266a1082 633 Fragment Cutting 1 Uwharries Southern Rhyolite (2) 4.04 (18.73) (12.43) 1.44 22° 271266a1212 686 Fragment Cutting 1 Uwharries Eastern Rhyolite 2.48 (18.89) 18.68 4.27 33° Core Flake Core Flake Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 187 Chapter 6. 31HK2510 Figure 54. Miscellaneous Stone Tools, Site 31HK2510. A: Grinding Stone (a849), B: Slab (a580), C: Hammer Stone Spall (a1058). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 188 Chapter 6. 31HK2510 Figure 55. Hammer/Anvil Stone (a288), Site 31HK2510. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 189 Chapter 6. 31HK2510 x 100.10 mm in horizontal dimensions. It weighed 1180.58 gm. Both ends of the tool exhibited battering marks and flake scars from percussive edge collapse. Linear abrasion zones were present on the two lateral edges of the tool and multiple anvil marks and pitting were present on the centers of both faces and on the two lateral edges as well. These wear patterns are indicative of core reduction and chipped stone tool manufacture. Daniel (1998:116) reports similar cobble tools from the Hardaway collection and he further suggests that these tools may have been used in bone, meat or nut processing. Precontact Ceramics A small collection of precontact ceramics was recovered during the investigation at 31HK2510. It consisted of 18 sherds, including representatives from two recognized ceramic series, New River and Hanover. Sixteen individual vessels were recognized in the analysis. Metric and attribute data on precontact ceramics are contained in Appendix I. Monitoring of temper types within series has proceeded in an ad hoc fashion in earlier investigations (see Cable and Cantley 2005a, b, 2006, Cable 2010), but the enormity of the inventory of temper constituents and various compositional combinations seen in ceramic collections makes it imperative that in the future a universal system be developed to accommodate cross-project comparability. Formerly, temper classes were assigned letter designations as they were historically recognized within each series, creating a confusing list of different letter designations for the same temper type within each series. Here, temper constituent designations are standardized into a single classificatory system. The letters “a” and “b” are reserved for the basic paste representatives diagnostic of each series as described previously. The remaining letter designations note the addition of other temper constituents to these basic paste categories as described below: (c) Crushed Arenite (d) Granule to Pebble Sized Angular/Subangular Quartz Sand and/or Crushed Quartz (e) Crushed and Ground Indurated Granite (f) Medium to Granule Sized Angular/Subangular Feldspar Sand More than one of these temper constituents was sometimes present in a single sherd. In these cases, the constituent designations would include combinations of letter designations (e.g. “c/e” or “e/f”, “d/e”, “d/f”, etc.). During the course of the analysis it also became apparent that some of the pottery exhibited extremely hard pastes, much harder, in fact, than the typical paste previously defined for each series. A hardness scale consisting of three attribute states was developed to describe this variation. These consisted of: (1) friable, (2) compact and (3) hard. The attribute state of “friable” corresponded to sherds that were easily crushed on their edges with only minimal application of pressure with the thumb. A “compact” state was identified in instances where moderate pressure with the thumb was required to break the edge of the sherd. Finally the “hard” state was reserved for those sherds whose edges were not easily broken even with extreme thumb pressure. Generally, pottery achieves greater hardness when one or more of three conditions prevail (Shepard 1954:114). These are: (1) the pottery is made from low-fusing, dense-firing clay, (2) the pottery is fired at a high temperature and/or (3) the pottery is fired in an atmosphere promoting vitrification. It is common to observe in the local ceramic sequences of the Southeast, a progression from friable ceramics to much harder, more durable pottery in the later stages of cultural development. Clearly, Mississippian potters had developed technological strategies to produce extremely hard, sometimes vitrified pottery and it is quite possible that this advancement was meant to adapt to a more sedentary, agricultural life way in which durability and longevity of pots would have been advantageous. Similar adaptive technologies may have been incorporated into sedentary, agricultural societies peripheral to the Southeast Appalachian Mississippian culture as well, and for the same reasons. Currently, a sophisticated method for measuring hardness in “primitive” pottery does not exist due to its porosity and heterogeneity of paste composition. The “thumb” test, although somewhat imprecise, provides a relatively objective and empirical basis for identifying relative hardness in precontact pottery assemblages. For an extended discussion of hardness in precontact ceramics see Chapter 4. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 190 Chapter 6. 31HK2510 This method of analysis resulted in segregating extremely hard or vitrified pottery that would have traditionally been placed in the Cape Fear (sandtempered), Hanover (sand- and grog-tempered) and Yadkin (crushed rock- and granule sand-tempered) series. As a means of preserving data to test the “hardness trajectory” hypothesis, these hard examples were separated into new series variants distinguished by the Roman numeral suffix “III.” Thus, extremely hard variants of the series are respectively referred to here as Hanover III, Cape Fear III and Yadkin III. If the efficacy of the hypothesis is later confirmed, then it would be appropriate to assign new series names to the Cape Fear III and Yadkin III variants since they are generally viewed as Middle Woodland series. The ceramics from each series are described below. New River Series New River series ceramics are associated with the Early Woodland period in the region. The series was originally defined by Loftfield (1976) who characterized it as a coarse sand-tempered pottery of compact composition and gritty consistency characterized by cord marked, fabric impressed, thong simple stamped and plain exterior surface treatments. Correlated series of the Early Woodland period include Lenoir (Crawford 1966) and Deep Creek (Phelps 1983). The New River series in the Fort Bragg area is poorly understood. Herbert (2003:156) indicates that both net-impressed and parallel/overstamped cord marked types are recognized, while Cable and Cantley (2005a, b) have also recognized simple stamped, fabric impressed and plain surface treatments. The predominant classificatory criterion used to identify the series is an abundance of quartz sand temper as Loftfield (1976) prescribed. Herbert (2003:156) proposes that an arbitrary cut-off at about 15% sand temper density should be used to distinguish Cape Fear (< 15%) and New River (>15%) in the Fort Bragg area. Cable and Cantley (2005a, b) have recognized two major paste variants in Fort Bragg assemblages. These are variants Ia and IIa. Variant Ia consists of sherds dominated by abundant medium to very coarse subangular to rounded quartz sand grains in a gritty, friable paste matrix. Variant IIa contains of a fine, compact, relatively harder paste matrix with abundant to moderately dense medium to very coarse quartz sand inclusions. Variant IIa may correlate with the traditionally defined Cape Fear series. Feldspar sand is sometimes found in variant Ia paste, and it is designated New River If. Only one New River series sherd was recognized in the collection from 31HK2510, a New River IIa sherd with an indeterminate and eroded exterior surface. It was found at a depth of 40 to 50 cm bs in a Stage II shovel test (N415/E531). This vertical poisition is unusually low for ceramics, suggesting either displacement due to recovery error or disturbance process. The paste of the specimen (a564) was compact and gritty to the feel. Sand inclusions in the matrix predominantly consisted of medium sized, rounded to sub-rounded clear quartz grains. Hanover Series South (1960:16-17; 1976: 28) defined the Hanover series from surface collections made in 1960 on sites located between Wilmington, NC and Myrtle Beach, SC. Pottery affiliated with this series is tempered with aplastic clay lumps, most of which consist of grog fragments from crushed sherds. A correlate on the north coastal plain of North Carolina is referred to as the Carteret series (Loftfield 1979:154157). Cord marked and fabric impressed exterior surface treatments were included in the original series description, but other grog-tempered types have been recognized subsequently. Herbert (2003:75) has identified minority proportions of check stamped, plain (smoothed), net impressed, simple stamped, punctate, and random straw bundle punctate surface treatments. The latter two types are traditionally subsumed under the Refuge series on the central South Carolina coast where grog/clay temper occurs as a major temper variant (see Cable 2002: 195-201). A great deal of variability in paste composition has been noted in Hanover assemblages, primarily related to differences in quartz sand contributions, clay versus grog distinctions, and the density and size of clay/grog particles (see Cable et al. 1998; Herbert 2003; 74-75). Herbert (2003:191-192) recognizes two broad variants that he believes are sequential. The earliest variant, Hanover I, is primarily sandtempered with minor amounts of finely crushed grog. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 191 Chapter 6. 31HK2510 Surface treatments attributed to Hanover I pottery consist of cord marking, fabric impressing and check stamping. He estimates an age range of AD 400-800 for Hanover I pottery. The later variant, Hanover II, is inferred to correlate with a Late Woodland temporal range (AD 800 through 1500). Hanover II pottery is characterized by pastes with only small amounts of sand and abundant grog particles. The dominant exterior surface treatment of the later variant is fabric impressing. There are three major paste variants found in Hanover pottery at Fort Bragg based on relative hardness. Hanover I is characterized by a friable paste, Hanover II exhibits a compact paste and Hanover III contains a hard to vitrified paste. Within each of these relative hardness groupings are two primary temper constituent designations. Designation “a” refers to pastes with sparse to moderate amounts of fine to medium quartz sand and abundant medium to large grog particles. Designation “b” includes pastes with moderate to abundant amounts of medium to coarse quartz sand and sparse to moderate densities of medium to large grog particles. Other temper constituents are sometimes added to these basic paste variants. These include crushed arenite (“c”), granule to pebble sized angular to subangular quartz sand and, rarely, crushed quartz (“d”), crushed and ground indurated granite (“e”) and medium to granule sized subangular feldspar sand (“f”). Crushed arenite and indurated granite almost exclusively occur in Hanover III paste. Indurated granite temper was originally identified as sandstone conglomerate because of similar texture characteristics, but thin-sectioning of rock samples indicated that most of the temper in this category corresponds to a fine grained, indurated granite (see Appendix K). Some of the temper in this category may actually represent sandstone conglomerate, but this was not differentiated in the ceramic analysis due to the late submittal of samples. No examples of the Hanover I series were recognized in the collection. This series is thought to occur early in the Middle Woodland sequence in the region and this contention is supported by a greater incidence of check stamping. Hanover II series sherds comprise about 61 percent of the sherd collection from the site. Hanover II is dominated by fabric impressed surface treatments and is generally viewed as a late Middle Woodland and Late Woodland cor- relate. Hanover III makes up about 33 percent of the sample. It is hypothesized that this series is associated with the later part of the Late Woodland period in the region. It typically is characterized by parallel stamped cord marked and fabric impressed surface treatments. Recognizable surface treatments on the Hanover II sherd sample consisted entirely of fabric impressing (Table 40, Figure 56:A-B). Two of the sherds displayed clear enough impressions to discern flexible warp and medium weft patterns. All but one of the sherds belonged to the Hanover IIa paste grouping. This grouping consisted of compact fine pastes with sparse medium to coarse quartz sand inclusions and moderate to large particles of grog. Mean sherd wall thickness for the sample was 7.40 mm (SD = 1.11 mm). All of the sherds represented body portions of vessels and very little vessel form information was revealed. Undulating or irregular smoothed interiors suggests that the predominance of the sample consisted of open mouth jars. The Hanover III sample from 31HK2510 consisted of six sherds from six different vessels (Table 40). Recognizable surface finishes were dominated by cord marking (n = 3), all of which were characterized by parallel stamping patterns (Figure 56:C-E). Cord impressions were not particularly fine, ranging between 1.19 and 1.99 mm in diameter. One sherd displayed a fabric impressed surface treatment. All but one of the sherds contained IIIa paste, which is characterized by a hard to vitrified, fine background clay with sparse inclusions of medium to very coarse quartz sand inclusions and abundant medium to large pieces of grog. One of the cord marked sherds exhibited a IIIe paste, similar to variant IIIa in all respects except that abundant, medium to granule-sized particles of crushed and ground indurated granite were added (Figure 56:E). Mean sherd wall thickness for the sample was 6.53 mm (SD = 1.44 mm). Again, all of the sherds represented body portions of vessels and consequently very little vessel form information was revealed. Undulating or irregular smoothed interiors suggests that the predominance of the sample consisted of open mouth jars. Historic Artifacts Eleven historic artifacts were recovered Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 192 Chapter 6. 31HK2510 Table 40. Precontact Ceramic Inventory, 31HK2510. 271266p8 6 Hanover IIa Fabric Impressed 271266p9 7 Hanover IIa Fabric Impressed 271266p29 17 Hanover IIa Fabric Impressed 271266p201 132 Hanover IIa Fabric Impressed 271266p446 270 VESSEL NO. Hanover IIa Fabric Impressed Body 2 7.76 13 Body 4 7.03 2 Body 3 7.56 2 Body 5 7.65 3 Smoothed Body 2 7.19 5 Smoothed Body 2 10.5 9 Smoothed Irregular Body 2 7.53 14 Smoothed Flexible/ Medium Flexible/ Medium Smoothed Irregular Smoothed Irregular Smoothed Irregular SHERD TYPE INTERIOR FINISH STAMP PATTERN WARP/WEFT 357 WALL THICKNESS (mm) 271266p564 SIZE New River IIa indeterminate BAG NO. CERAMIC TYPE SPECIMEN NO. PRECONTACT CERAMICS: 31HK2510 Hanover IIa Fabric Impressed 271266p581 371 Hanover IIa Fabric Impressed 271266p1016 589 Indeterminate Body 1 6.84 16 Hanover IIa indeterminate 271266p318 205 Smoothed Body 2 6.78 7 Hanover IIa indeterminate 271266p497 306 Smoothed Body 3 6.93 10 Hanover IIa indet. decorated 271266p498 306 Eroded Body 2 6.25 10 Hanover IIb indeterminate 271266p516 320 Smoothed Body 2 7.09 11 Hanover IIIa Cord Marked 271266p7 Smoothed Body 2 6.3 1 Body 2 Body 2 4.83 6 Body 3 5.66 8 Body 2 7.33 4 Body 4 8.52 12 5 Hanover IIIa Cord Marked 271266p719 436 Hanover IIIa Fabric Impressed 271266p317 205 Hanover IIIa indeterminate 271266p340 219 Hanover IIIa indet. decorated 271266p148 102 Hanover IIIe Cord Marked 271266p520 322 during the investigation. Most of these appear to have been deposited during military training exercises. The inventory consists of full metal jacket bullet heads, M14/M16 bullet casings, tin and aluminum can parts, white ware and light green bottle glass (Table 41). Although the majority of these items were found in the upper 10 to 20 cm of deposit in the plow zone, the tin and aluminum can parts were found between 20 and 30 cm bs. Parallel Parallel Smoothed Irregular Smoothed Irregular Eroded Parallel Smoothed Irregular Smoothed Irregular 15 OCCUPATION PATTERNS Diagnostic markers of culture-chronological association were relatively numerous in the 31HK2510 artifact assemblage. The dominance of lithic artifacts and Archaic tool types indicates that the greatest proportion of the occupation debris was deposited during the Archaic period. Recovered pro- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 193 Chapter 6. 31HK2510 Figure 56. Precontact Ceramics, Site 31HK2510. A-B: Hanover IIa Fabric Impressed (p8 and p29 respectively), C-D: Hanover IIIa Cord Marked (p7 and p719 respectively), E: Hanover IIIe Cord Marked (p520). Table 41. Historic Artifact Inventory, 31HK2510. HISTORIC ARTIFACT INVENTORY: 31HK2510 HISTORIC ARTIFACT TYPE LEVELS 1 2 Aluminum Can Pop Top Ring 2 2 FMJ Bullet Head 2 Light Green Bottle Glass 2 2 M14/M16 Bullet Casing 1 1 1 3 Tin Can Body Fragment 1 1 Tin Can Rim 1 1 White Ware, Plain GRAND TOTAL 3 GRAND TOTAL 1 6 1 1 4 11 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 194 Chapter 6. 31HK2510 jectile points document Early Archaic (i.e., a Small Dalton, three palmer I Corner Notched and a LeCroy Bifurcate Stem) and Middle Archaic (i.e., a Morrow Mountain I Stemmed). In addition, a wide range of scraper types and specialized unifacial tools evidence especially Early Archaic occupation, although some of the forms have also been recognized in Middle Archaic deposits. Finally, a limited precontact ceramic collection evidences Woodland period occupation. This evidence was supplemented by three late precontact projectile points, including two Yadkin Large Triangular points and a Pee Dee Serrated point. The site is characterized by moderate artifact density, suggesting frequent short-term reoccupation patterns. Artifact density in positive Stage I and II shovel tests (n=170) equaled a mean of 3.02 artifacts (SD = 3.74), indicating that two-thirds of the positive tests yielded densities between one and seven artifacts. The modal positive shovel test outcome across the sample was one artifact. Mean artifact density for the entire site area (including negative Stage I and II shovel tests) was 1.29 artifacts (SD = 2.86). Metavolcanic debitage recovered during Stage I shovel testing was widely distributed across the terrace (Figure 57), while quartz debitage presented clusters of limited extent (Figure 58). Vertical distributions by raw material type indicate that debitage is centered between 10 and 50 cm bs (Table 42). The vertical profile suggests that deposits exhibit variable vertical positions and most likely are composed of a wide range of cultural associations. From what we have learned about the vertical distributions of cultural-chronological occupations on sites in the Sandhills (see Cable 2010; Cable and Cantley 2005a, b, 2006), this distribution indicates that the great majority of the occupations are associated with the Archaic period. Examination of the vertical distribution of Archaic projectile point styles and scrapers, however, shows that diagnostic Early and Middle Archaic stone tools are distributed somewhat lower than the debitage (Table 43). The stone tool distribution is centered between 30 and 40 cm bs, whereas debitage is most frequent between 20 and 30 cm bs. Seventy-two percent of the Middle and Early Archaic stone tools are situated below 30 cm bs, while only 54 percent of the debitage occupies a vertical position below 30 cm bs. This might suggest that a great deal more Woodland and Late Archaic occupation is present at the site than the frequency of recovered diagnostic stone tools and ceramics would imply. However, The skewed distribution may be the consequence of multiple factors. Throughout the region, Early Archaic occupations are overwhelmingly composed of short-term occupations that resulted in quantitatively low debitage production, while Middle and Late Archaic occupations are represented by higher proportions of longer-term residences that tend to create highdensity debitage concentrations. This alternative might suffice to fully explain the skewed distribution had a higher proportion of the recovered tool diagnostics belonged to the later Archaic phases. The overwhelming majority of projectile points and most of the scrapers in the inventory are associated with the Early Archaic period, but the vertical distributions of the chipped stone tools show that many are positioned higher in the soil matrix than would be expected (Table 43). Most notably “out of position” is the Dalton Adz and the Small Dalton point. Some of the scrapers occupy higher positions as well. Most of these “out of position” tools are distributed linearly along the mid-slope of the terrace above the two-track road (Figure 59), which suggests that depositional processes may contribute to the higher position of debitage as well. Figure 60 compares the stone tool distributions to a contour map of soil depth above the Bt-horizon as reflected by Stage I and II shovel test data. The crest of the terrace and the slope toe are characterized by deep solumns extending to depths of 80 cm bs to more than 100 cm bs, while the midslope displays much shallower ones, predominantly ranging between only 30 and 60 cm bs in depth. This is precisely the kind of soil displacement pattern that is common to alluvial fans. The main force of erosion occurs at mid-slope, and not at the top of a ridge, resulting in sediment transport to the slope toe creating an alluvial fan (Foss, pers. comm.). The potential impact of this factor can be further recognized by viewing the vertical debitage distributions segregated into groups consisting of maximum depths of above and below 60 cm bs (Tables 44 and 45). Approximately 72 percent of the debitage in shovel tests excavated on the mid-slope is positioned above 30 cm bs, while 60 percent of debitage on the top of the ridge and on the toe slope is situated below Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 195 Chapter 6. 31HK2510 530 Positive ST Negative ST 520 Two-Track Road oc kf ish R Springhead 510 Cr eek 73.0 500 .0 74 490 .5 73 480 470 31HK2510 460 450 440 430 73.0 420 Flood Plain 71 .5 410 400 0 72 .5 74. 390 72.0 380 370 74.0 360 350 N 340 0 40 m 330 320 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 57. Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31HK2510 (Contours = 1 piece of debitage). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 196 Chapter 6. 31HK2510 530 Positive ST 520 Negative ST Two-Track Road oc kf ish R Springhead 510 Cr eek 73.0 500 .0 74 490 31HK2510 .5 73 480 470 460 450 440 430 73.0 420 Flood Plain 71 .5 410 400 0 72 .5 74. 390 72.0 380 370 74.0 360 350 N 340 0 40 m 330 320 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 58. Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31HK2510 (Contours = 1 piece of debitage). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 197 Chapter 6. 31HK2510 Table 42. Debitage Inventory by Level for Stage I and II Shovel Tests, 31HK2510. DEBITAGE BY LEVEL, STAGE I AND II SHOVEL TESTS*: 31HK2510 LEVELS RAW MATERIAL TYPE 1 2 3 4 5 6 2 1 13 9 5 1 Type I Rhyolite Tuff 1 9 17 14 17 11 Uwharries Eastern Rhyolite 3 4 7 10 Uwharries Southern Rhyolite (1) 1 3 9 8 14 4 Uwharries Southern Rhyolite (2) 4 11 37 26 13 11 2 2 12 15 8 6 3 1 22 25 8 8 9 1 2 Mill Mountain Rhyolite Uwharries Western Rhyolite White Quartz 13 Fine Grained Quartzite 8 9 10 31 3 72 24 2 5 41 1 1 2 Metasedimentary, gray 1 Slate 2 88 1 1 5 1 1 24 49 110 48 1 indet. metavolcanic GRAND TOTAL 7 GRAND TOTAL 1 120 89 76 43 9 3 8 1 422 *Contents of Shovel Test Excavated at N380.3/E495.3 Eliminated Due to Combined Levels. 30 cm bs. Deflation of the mid-slope, then, appears to explain a significant portion of the instances of “out of position” stone tools and the skewed debitage profile. Thirty-eight percent of the debitage in the latter sample is actually situated below 40 cm bs, which attests to the relatively dense Early Archaic occupation at the site evidenced by the stone tool inventory. Two areas of primary Early Archaic occupation are identified by scraper distributions. One is on the northern end of the terrace top and the other is situated on the slope toe (see Figures 59 and 60). Scraper density is somewhat exaggerated on the terrace top locality because Stage III and IV contexts were also used to map tool distributions. Both localities display deep debitage profiles (Table 46), and even though this outcome might be predicted because of great soil depth, the concentration of Early Archaic tool types suggests that most of the debitage is also affiliated with Early Archaic occupations. Approximately 68 and 84 percent of the debitage, respectively, in the north terrace and slope toe localities is positioned below 30 cm bs. The slope toe is characterized by very low-density occupations that are deeply buried and it may contain almost exclusively Early Archaic deposits. Moreover, the influx of sediment from alluvial fan formation may have served to bury and protect these occupations from significant cultural deposit mixing. Precontact ceramics, as stipulated by the vertical coherence model, are positioned higher overall than lithic debitage distributions (Table 47). Their frequencies peak in Level 2 (10-20 cm bs) and 71 percent of the recovered ceramic collection occurs between Levels 1 and 3 (0 to 30 cm bs). Sample sizes are too small to reliably comment on inter-series vertical positions. Twenty-nine percent of the collection, however, is “out of position.” In this case, the positions of sherds are unusually low relative to expectations. It is likely that these specimens have been displaced through a number of potential processes, including recovery error, bioturbation, military training impacts and possibly cultural pit feature construction. Three of these “out of position” ceramics are situated on the slope toe and their low vertical positions are likely the consequence of increased depo- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 198 Chapter 6. 31HK2510 Table 43. Diagnostic Archaic Stone Tool Inventory, All Proveniences, 31HK2510. DIAGNOSTIC ARCHAIC STONE TOOLS, ALL PROVENIENCES: 31HK2510 LEVEL STONE TOOL TYPES 1 2 3 Dalton Adz 4 5 6 7 8 GRAND TOTAL 9 1 1 End Scraper Fragment 1 1 LeCroy Bifurcated Stem 1 1 Morrow Mountain I Stemmed 1 1 Palmer I Corner Notched 3 3 Retouched Flake 1 Side Scraper Fragment 1 1 1 1 Small Dalton Point 1 4 1 1 Type I Side Scraper 3 Type IIa Side Scraper 1 1 2 Type IIb End Scraper 1 5 2 5 1 1 1 4 Type III End Scraper 1 1 Type IV End Scraper 1 Type IIb Side Scraper 1 2 Type IV Side Scraper 1 2 1 GRAND TOTAL 2 1 3 4 12 4 1 4 1 1 32 Table 44. Vertical Distribution of Debitage in Mid-slope Stage I and II Shovel Tests, 31HK2510. VERTICAL DISTRIBUTION OF DEBITAGE IN MID-SLOPE SHOVEL TESTS (STAGE I&II): 31HK2510 LEVELS RAW MATERIAL TYPE Mill Mountain Rhyolite 1 2 3 4 5 1 1 3 5 1 1 2 Type I Rhyolite Tuff Uwharries Eastern Rhyolite 3 1 Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) 2 5 14 11 19 8 9 10 GRAND TOTAL 3 4 4 12 2 1 11 1 3 Slate GRAND TOTAL 4 7 11 1 2 Uwharries Western Rhyolite White Quartz 6 3 10 3 24 14 1 33 1 75 1 1 6 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 199 Chapter 6. 31HK2510 Table 45. Vertical Distribution of Debitage for Stage I and II Shovel Tests Minus the Mid-slope Sample, 31HK2510. VERTICAL DISTRIBUTION OF DEBITAGE IN SHOVEL TESTS MINUS MID-SLOPE (STAGE I&II): 31HK2510 LEVELS RAW MATERIAL TYPE 1 2 3 4 5 6 10 4 4 1 14 17 11 7 8 Mill Mountain Rhyolite 1 Type I Rhyolite Tuff 1 8 15 3 3 6 Uwharries Southern Rhyolite (1) 1 3 8 8 14 4 Uwharries Southern Rhyolite (2) 2 9 33 24 12 11 2 2 9 15 8 6 3 1 8 15 5 8 8 1 2 Uwharries Eastern Rhyolite Uwharries Western Rhyolite White Quartz 8 Fine Grained Quartzite 10 20 3 69 12 2 5 40 1 1 2 Metasedimentary, gray 1 13 30 96 sition rates in the alluvial fan. The remaining two sherds are situated on the terrace top, however, and their depth is more likely attributable to cultural displacement processes or bioturbation. The series are widely distributed across the terrace and slope toe, suggesting that Woodland occupation of the site consisted of short-term camp reoccupations. Nine separate Hanover II clusters, six Hanover III clusters and one New River II cluster are represented in the sample (Figure 61). In addition, three Woodland period points were recovered in the southern portion of the terrace, suggesting that this area may contain the highest density of Woodland occupation. Woodland period lithic and ceramic concentrations are generally “paired” and adjacently located features in short-term camps and their distributions tend to be exclusive of one another in shovel tests. The point styles indicate that both Middle and Late Woodland occupations are present at the site, an inference that is also supported by the identification of Hanover II and III ceramic series. The latter series is hypothesized to occur later in the Late Woodland sequence, which would suggest that it is associated with the Pee Dee Serrated point style. 2 99 45 55 1 indet. metavolcanic GRAND TOTAL 9 GRAND TOTAL 1 1 5 1 75 70 42 9 3 8 1 347 A more detailed appreciation for precontact land-use patterns evidenced at the site can be derived from a consideration of occupation type distributions and associations. Currently, four occupation types based on chipped stone criteria have been recognized on sites in the Fort Bragg area (see Cable and Cantley 2005c, 2006): (1) Type I occupations consist of a debitage concentration and an associated tool cluster situated at a confined location along its periphery. It is inferred that the tool cluster was formed around a hearth and in several instances calcined bone fragments have been recovered from the deposit delimited by a tool cluster. It is also believed that ephemeral shelters were situated adjacent to these tool clusters and opposite the concentrations. Debitage concentrations appear to represent lithic reduction loci established away from a shelter or sleeping area where tools were manufactured during the stay at the camp. These concentrations are generally composed of single lithic raw material types, and the associated tool clusters commonly consist of manufacturing rejects and discarded broken or worn-out tools made of the same raw material that comprises the associated debitage concentration. Although many Type I residences occur in isolation and appear to be the by-products from single nuclear or small extended family occupa- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 200 Chapter 6. 31HK2510 530 520 Two-Track Road oc kf ish R Springhead 510 Cr eek 73.0 500 .0 74 490 .5 73 480 Palmer I 470 31HK2510 460 Palmer I 450 Small Dalton 440 430 Morrow Mountain I 420 73.0 Flood Plain 71 .5 410 Dalton Adz 400 0 72 .5 74. 390 Palmer I LeCroy 72.0 380 370 74.0 360 Projectile Point Below 30 cm bs 350 Projectile Point Above 30 cm bs N 340 Scraper Below 30 cm bs 0 Scraper Above 30 cm bs 40 m Dalton Adz (Above 30 cm bs) 330 320 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 59. Locations of Stone Tools of Probable Early Archaic Association, 31HK2510. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 201 Chapter 6. 31HK2510 530 520 Two-Track Road R oc kf ish 510 Cr eek 500 490 480 470 Palmer I 31HK2510 460 Palmer I 450 Small Dalton 440 430 Morrow Mountain I 420 Flood Plain 410 Soil Depth Above Bt-horizon 400 Dalton Adz > 90 cm bs Palmer I 80-90 cm bs 390 70-80 cm bs LeCroy 60-70 cm bs 380 50-60 cm bs 40-50 cm bs 370 30-40 cm bs 360 Projectile Point Below 30 cm bs 350 Projectile Point Above 30 cm bs Scraper Below 30 cm bs N 340 0 Scraper Above 30 cm bs 40 m Dalton Adz (Above 30 cm bs) 330 320 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 60. Comparison of Soil Depth and Early Archaic Stone Tool Distributions, 31HK2510. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 202 Chapter 6. 31HK2510 Table 46. Comparison of Vertical Distribution of Debitage Between Early Archaic Localities, 31HK2510. COMPARISON OF VERTICAL DISTRIBUTION OF DEBITAGE BETWEEN EARLY ARCHAIC LOCALITIES: 31HK2510 NORTHERN TERRACE TOP LOCALITY (N440-490/E460-500) LEVELS RAW MATERIAL TYPE 1 2 3 indet. metavolcanic 4 5 6 7 8 9 10 1 Mill Mountain Rhyolite 1 Type I Rhyolite Tuff 1 Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) 2 1 2 3 2 3 10 9 8 8 1 1 2 1 3 5 8 2 3 18 10 3 6 7 10 2 2 Uwharries Western Rhyolite GRAND TOTAL 2 33 4 White Quartz 2 1 1 1 2 6 GRAND TOTAL 5 6 35 41 28 24 19 2 44 1 22 1 2 14 2 2 8 9 145 SLOPE TOE/ALLUVIAL FAN LOCALITY (N400-450/E520-540) RAW MATERIAL TYPE LEVELS 1 2 3 4 5 6 Type I Rhyolite Tuff 1 Uwharries Eastern Rhyolite 1 4 1 3 2 4 2 2 1 2 1 1 3 2 4 9 14 6 Uwharries Southern Rhyolite (1) 1 Uwharries Southern Rhyolite (2) 1 3 Uwharries Western Rhyolite White Quartz 1 GRAND TOTAL 2 1 7 10 GRAND TOTAL 1 5 2 2 3 9 1 1 18 4 8 2 6 1 45 Table 47. Vertical Distribution of Woodland Ceramics and Projectile Points, All Proveniences, 31HK2510. WOODLAND CERAMICS AND PROJECTILE POINTS, ALL PROVENIENCES: 31HK2510 LEVELS SERIES/TYPE 1 Hanover II Hanover III 1 2 3 4 5 6 4 3 1 2 1 3 1 New River II Pee Dee Serrated GRAND TOTAL 6 1 1 1 1 1 1 2 11 1 1 Yadkin Large Triangular, Variant B Yadkin Large Triangular, Variant E GRAND TOTAL 8 5 1 1 4 1 21 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 203 Chapter 6. 31HK2510 530 520 Two-Track Road oc kf ish R Springhead 510 Cr eek 73.0 500 .0 74 490 .5 73 480 470 31HK2510 460 450 440 430 73.0 420 Flood Plain .5 Yadkin L. Tr., Var. B 410 71 Pee Dee Serrated 400 0 72 .5 74. 390 72.0 Yadkin L. Tr., Var. E 380 370 74.0 360 New River II Hanover II Hanover III 350 Woodland Projectile Point N 340 0 40 m 330 320 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 61. Woodland Period Artifact Distributions, 31HK2510. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 204 Chapter 6. 31HK2510 tions, others appear to represent multi-family occupations that extend beyond the ability of a single shovel test to identify. (2) Type II residences are similar in spatial organization to Type I residences, but they exhibit much higher tool-to-debitage ratios and may commonly be comprised of diverse lithic raw materials. Lithic reduction on these sites consisted of tool maintenance and late stage shaping of tools from flake blanks or reduced biface cores. Consequently, debitage density is much lower than that of Type I debitage concentrations and the scatters consist primarily of late stage reduction debitage. Type II occupations are inferred to be the output from single or small, multiple household aggregations and are thought to represent the residue from High Technology Forager (HTF) residences (Spiess 1984; Todd 1983:231-233). The HTF model proposes a specialized forager adaptation combining high logistical and high residential mobility into a single settlement system characterized by highly curated technologies. Type II residences have thus far only been assigned to the Early Archaic period at Fort Bragg. The lithic raw material diversity characteristic of Type II residences suggests that these occupations represent the aggregation of social units that may have been seasonally dispersed, transporting a wide range of lithic raw materials from different sources. (3) Type III occupations are inferred to represent logistical camps and very short-term residences characterized by low-density debitage scatters and sporadic tool discard. (4) Type IV occupations represent extraction or processing loci that are not believed to have involved, necessarily, overnight stays. They are difficult, however, to distinguish from Type III occupations, even in situations of greater excavation exposure. It is difficult to distinguish effectively between these various occupation types at shovel test intervals of greater than 1.25 m, but a less precise classification can be used at greater shovel test intervals to aid in population projections. High-density debitage concentrations tend to indicate Type I residences, while low-density scatters tend to signal short-term residences, special purpose camps and locations (i.e. Type II, III and IV occupations). Highdensity debitage concentrations are temporarily stipulated to correspond to shovel test outcomes of five or more pieces of debitage of a specific raw material subtype. Eventually, it will be necessary to confirm this arbitrary threshold by examining shovel test counts from a large sample of known high-density debitage concentrations. Deposits containing less than five pieces of debitage are viewed as more likely to represent low-density debitage scatters associated with the other three site types. Since shovel tests can intersect the peripheral zones of high-density debitage concentrations where debitage densities are much lower, however, it is likely that Type I residential occupations will be under-represented in macrointerval shovel test data. Because it is difficult to determine whether a single occupation is represented in more than one shovel test, in the cases of multi-family residences or associated special activity zones on the periphery of a camp, the term “element” is preferred over the term “occupation” to refer to the occupational data in shovel tests. At a sampling interval of 5-m or 10-m intervals, we are assured that the presence of an element will not be repeated in adjacent shovel tests. Elements are identified by lithic raw material subtypes and/or by precontact sherd scatters associated with single vessels. This methodology runs the risk of overestimating occupations because more than one raw material subtype may comprise a single occupation. The term “element” recognizes this fact, focusing on depositional events and episodes rather than entire occupations. Elements can represent segments of contemporaneous multi-family occupations, separate occupations of various function, or contemporary members of multiple raw material reduction episodes. The current methodology for identifying elements in shovel tests consists of running cross-tabs of lithic raw material subtypes or subtype groups by shovel test number in a spreadsheet. Once the crosstab is generated, the sum of each lithic raw material subtype in a shovel test is identified as an element. In some instances, the vertical separation between individual items of the same lithic subtype can be so great as to suggest the presence of distinct elements. Due to the vagaries of bioturbation and possible recovery errors in which higher positioned material can be included in lower levels through sidewall displace- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 205 Chapter 6. 31HK2510 ment, however, the lower item(s) is assigned to the higher positioned item(s) to define a single occupation element Elements have horizontal dimensions and, as a result, sample units that are smaller than those dimensions have the benefit of increased sample efficiency (see Cable and Donaldson 1988; Rice 1987; Rice and Plog 1983). The typical element found at Fort Bragg has a diameter of about 3 m. Thus, a shovel test of .09 m2 samples only .09 percent of the area of a 10 m-square, but it samples 9.0 percent of the theoretical mean element space of 9.0 m2. Consequently, a 10 m-interval shovel test pattern actually samples about 9.0 percent of the available element space at a site, while a 5 m-interval shovel test pattern samples about 36.0 percent of the available element space. A mean estimate of the population of elements at shovel test intervals of 10 m and 5 m, then, requires a simple arithmetic calculation of counting the occupation elements identified in shovel tests and dividing this figure respectively by 9.0 percent and 36.0 percent. Although Stage I shovel test samples have been relied upon in the past to produce population projections because of their even and complete coverage, a concerted effort was made in the DO5 package to extend 5 m-interval shovel tests across the entire site area so that advantage could be taken of greater sample density. Following the procedures outlined above, the procedure succeeded in identifying 14 Type I residence elements, and 274 indeterminate (Type II, III and IV) elements in the combined Stage I and II Shovel Test data (Table 48). The Uhwarries Southern Rhyolite sample was segregated into two raw material groups (i.e., the R3m and R3t groups as discussed in the raw material description section of the chapter). It can further be estimated that the site contains a mean of about 800 elements, of which approximately 39 represent Type I residential elements. Stone tools were counted as elements of a particular lithic raw material subtype when found in isolation from debitage in a shovel test. Thirtysix isolated tools or tool clusters were counted in this census. Of interest is that 14 of these isolated tool finds consisted of scrapers, while a Palmer I Corner Notched and a Dalton Adz were also found in isolation. Isolation for the purposes of element population estimates means only that debitage matching the isolated tool was not recovered in the shovel test. This might suggest again, that a number of the Early Archaic elements represent members of Type II occupations, which are characterized by low-density debitage scatters of multiple raw material types and relatively high tool discard rates. The proportional representation of Type I elements of 4.9 percent at 31HK2510 is similar to the proportion calculated for nearby 31HK2502, which appears to have a nearly identical occupational record. This is a relatively low percentage, indicating that the site was not a preferred location from longer-term, or more intensive residential occupation. In general, the landform on which the site rests was used for special purpose camps and extraction loci for groups that rarely stayed longer than several days. Expanded over nearly the full range of human occupation (about 10,000 years), the population estimate indicates that precontact groups, on average, visited the site a maximum of once every 12.5 years. The occupation, however, was probably not evenly distributed through time. The site was no doubt used intensively by specific culture-historic groups, probably for strings of continuous years punctuated by abandonments in response to long-term land-use patterns dependent upon resource depletion cycles. The site appears to have been most intensively utilized during the Early Archaic period. Later occupation was less intense and primarily consisted of Middle Archaic and Woodland components. Due to the less constrained nature of the landform to provide suitable locations for camping, 31HK2510 was characterized by a much lower incidence of superimposition of separate occupations than 31HK2502. Mean element occurrence in the combined Stage I and II sample was 1.62 elements (SD = 0.86), compared with a mean of 2.24 elements (SD = 1.11) for 31HK2502. Only 42 percent of shovel tests contained two or more elements, compared to 72 percent of shovel tests at 31HK2502. Moreover, only 15 percent of the shovel tests contained more than two occupation elements, compared to 35 percent of the shovel tests at 31HK2502. Although there is not a one-to-one correspondence between elements and individual occupations, the data clearly point to a much lower level of deposit superimposition at 31HK2510. Just as at 31HK2502, though, the actual magnitude of superimposition may be overestimated Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 206 Chapter 6. 31HK2510 Table 48. Recognized Elements for the Stage I and II Shovel Test Sample, 31HK2510. IDENTIFIED ELEMENTS FROM THE STAGE I AND II SHOVEL TEST SAMPLE: 31HK2510 RAW MATERIAL TYPES TYPE I TYPE II, III, IV GRAND TOTAL PERCENTAGE OF ELEMENTS Mill Mountain Rhyolite 1 22 23 8.0 Type I Rhyolite Tuff 3 36 39 13.5 Uwharries Eastern Rhyolite 1 19 20 6.9 Uwharries Southern Rhyolite (1) 1 39 40 13.9 Uwharries Southern Rhyolite (2) 3 73 76 26.4 Uwharries Western Rhyolite 1 34 35 12.2 White Quartz 4 51 55 19.1 GRAND TOTAL 14 274 288 100.0 due to the potential for many of the Early Archaic occupations to consist of multiple elements from single Type II residences. This issue will be further developed in the next section, which examines the results of the Stage III Sample Block investigations. Not included in the population estimate was the Woodland period material. Thirteen individual ceramic vessels are represented in the Stage I and II shovel test sample. Since we can expect most sherds from single vessels to be contained within an area of about 2 m in diameter, we can estimate that there are a mean of 81 individual vessel scatters present at the site. Since Woodland period occupations generally contain paired debitage scatters and sherd clusters, it is likely that some of the elements identified in the lithic subtype tally are Woodland in association. Thus to avoid the issue of double counting, ceramics were not included in the element population estimate presented above. It is roughly estimated from this calculation, however, that about 10 percent of the occupation elements represented at 31HK2510 are affiliated with Woodland period occupations. The Stage I and II shovel test sample succeeded in identifying New River II, Hanover II and Hanover III series. Using the proportions of individual vessels associated with each of these series it can be inferred that about 8 percent of the Woodland occupation is affiliated with the Early Woodland period (New River II), approximately 54 percent with the Middle to Late Woodland transition (Hanover II) and about 38 percent with the later Late Woodland period (Hanover III). Numerous studies have documented vertical patterning in the relative depth of deposits in the unconsolidated sandy soils of the Coastal Plain (Cable and Cantley 2005 b, 2006; Michie 1990). The actual depth patterns vary according to the character of the deposit. Those deposits with higher B-horizons tend to have compressed sequences, while those in deep C-horizon and E-horizon sediments have more discreet and vertically expansive sequencing. Michie’s (1990) C-horizon model developed in the Waccamaw Neck area on the central coast of South Carolina begins with Mississippian and Woodland materials positioned in the upper 30 cm of sediment. This zone is underlain by ceramic Late Archaic occupations situated at 28 to 35 cm bs, Middle Archaic horizons at 35 to 55 cm bs, and Early Archaic cornernotched and side-notched components at 55 to 62 cm bs. This model is generally applicable to sites on Fort Bragg. However, the deposits at Fort Bragg appear Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 207 Chapter 6. 31HK2510 to be somewhat more compressed, judging by the vertical positions of samples of diagnostic artifacts. At Fort Bragg three vertical groupings have been developed for deep Sandhills sites: (1) Late Archaic/ Middle Archaic between 15 and 40 cm bs, (2) Middle Archaic/Early Archaic between 35 and 50 cm bs, and (3) Early Archaic below 50 cm bs. The divisions are broadly constructed and realistically anticipate a certain degree of vertical mixing and overlap between the various Archaic periods. This vertical model can be of great utility in reconstructing the occupation history of sites like 31HK2510 where the predominant artifact class is non-diagnostic lithic chipped stone debitage. The vertical positions of the various elements recognized in the Stage I and II shovel test sample provide a rough picture of the representation of occupation periods at the site (Table 49). Most elements were situated in a single level, but some, especially the higher density elements, spanned several levels. In the latter cases, the central tendency of the vertical distribution of a particular element was used to estimate its position. Vertical data from the shovel test sample was not precisely congruent with the model ranges discussed above. However, the following adaptation was developed to correlate culture-historic association with level data: (1) 0-20 cm bs, Woodland, (2) 20-40 cm bs, Woodland/Late Archaic/Middle Archaic, (3) 40-50 cm bs, Middle Archaic/Early Archaic and (4) 50 to 70 cm bs, Early Archaic. Approximately 84 percent of the occupation elements are positioned below 20 cm bs. In light of the estimates for Woodland period occupation provided by the mean number of vessels present, it is likely that almost all of these elements are Archaic in association. Definitive evidence of a Late Archaic presence was not produced during the investigation, suggesting that the Late Archaic contribution to the occupation history of the site was minimal. About 18.4 percent of the occupation is unambiguously Early Archaic, but much of the occupation positioned between 40 and 50 cm bs is most likely also affiliated with Early Archaic phases. Given the problems of deposit deflation on the midslope, it would appear that Early Archaic occupation probably exists between 20 and 40 cm bs as well. It is difficult to provide a firm estimate for Early Archaic representation because of this problem, but judging by the preponderance of Early Archaic diagnostics, it is likely that it comprises as much as 50 percent of the occupation at the site. The remaining30 to 35 percent of the elements are most likely affiliated with the Middle Archaic. Table 49. Inferred Culture-Historic Associations of Recognized Lithic Elements, 31HK2510. INFERRED ASSOCIATIONS OF RECOGNIZED LITHIC ELEMENTS: 31HK2510 GRAND TOTAL 6 4 5 11 2 16 47 16.3 14 14 9 15 39 21 19 131 45.5 Middle/Early Archaic 5 11 7 13 10 4 7 57 19.8 Early Archaic/Paleoindian 1 8 7 16 8 13 53 18.4 23 39 40 76 35 55 288 100.0 Woodland/Late Archaic/Middle Archaic GRAND TOTAL 20 PERCENTAGE WHITE QUARTZ UWHARRIES WESTERN RHYOLITE UWHARRIES SOUTHERN RHYOLITE (2) UWHARRIES SOUTHERN RHYOLITE (1) UWHARRIES EASTERN RHYOLITE 3 Woodland TYPE I RHYOLITE TUFF INFERRED ASSOCIATION MILL MOUNTAIN RHYOLITE ELEMENTS Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 208 Chapter 6. 31HK2510 Type I elements are primarily concentrated above 40 cm bs, where they are inferred to be associated most strongly with Woodland and Middle Archaic phases (Table 50). Approximately 71 percent of the Type I residences are positioned within this vertical range. The remaining Type I residences are situated at depths more consistent with Early Archaic occupation. The concentration of Type I elements in the upper 40 cm may also contribute to the skewed vertical debitage distributions as discussed earlier. Type I elements are characterized by much greater densities of debitage. Overall, the occupation history and element representation is nearly identical to the patterns delineated for 31HK2502. SAMPLE BLOCK INVESTIGATIONS Eighteen shovel test outcomes from the Stage I and II sample were targeted for close-interval shovel testing to further elaborate the character of the occupations extant at the site. In general, a series of eight additional shovel tests was excavated at 1.25 m-intervals surrounding the target shovel test. These are referred to as sample blocks and their locations are illustrated in Figure 48. Sample Block (SB) locations were evenly distributed across the site to achieve a representative picture of the settlement record. Tables 51 through 53 display the vertical distributions of debitage by raw material type for each of the sample blocks. Most of the sample blocks yielded debitage profiles concentrated between 20 and 50 cm bs, in what would be primarily considered the Archaic zone of the site. Debitage profiles and stone tool distributions (Table 54) indicate that Early Archaic occupation is most likely predominant in SB1, SB3, SB4, SB 5, SB8, SB9, SB11, SB12, SB15 and SB17. Middle Archaic occupation is confirmed in only one of the sample blocks, SB7, where two Morrow Mountain I Stemmed point was recovered at a depth of 30 to 40 cm bs. It is likely, however, that Middle Archaic occupation is dominant in a number of the sample blocks lacking diagnostic tools, possibly including SB7, SB13 and SB18. Woodland occupation was confirmed by projectile point styles in SB6 (Late Woodland), SB2 (Middle Woodland) and SB18 (Middle Woodland). Debitage profiles suggest that Woodland occupation was dominant in SB2 and SB16. Six precontact sherds, all Hanover II series, were recovered in sample blocks (i.e. SB1, SB7, Table 50. Inferred Culture-Historic Associations of Recognized Type I Elements, 31HK2510. INFERRED ASSOCIATIONS OF RECOGNIZED TYPE I ELEMENTS: 31HK2510 Woodland/Late Archaic/Middle Archaic 1 2 Middle/Early Archaic 2 1 Early Archaic/Paleoindian GRAND TOTAL 1 3 2 14.3 2 8 57.1 2 14.3 2 14.3 14 100.0 1 1 1 3 1 4 PERCENTAGE 2 1 1 1 GRAND TOTAL Woodland WHITE QUARTZ UWHARRIES WESTERN RHYOLITE UWHARRIES SOUTHERN RHYOLITE (2) UWHARRIES SOUTHERN RHYOLITE (1) UWHARRIES EASTERN RHYOLITE TYPE I RHYOLITE TUFF INFERRED ASSOCIATION MILL MOUNTAIN RHYOLITE ELEMENTS Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 209 Chapter 6. 31HK2510 Table 51. Vertical Distribution of Debitage, Sample Blocks 1—6, 31HK2510. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 1-6:31HK2510 SB1 RAW MATERIAL LEVEL 1 2 3 Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) GRAND TOTAL 4 5 6 7 1 3 2 6 2 1 3 1 1 1 5 6 7 8 5 6 7 8 1 5 6 1 1 2 7 8 8 SB2 RAW MATERIAL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 1 2 1 1 2 1 2 2 5 3 4 1 1 2 2 5 1 2 1 2 3 4 SB4 1 2 3 4 5 Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 2 3 5 6 7 2 2 GRAND TOTAL 1 4 1 3 9 LEVEL 1 2 3 4 1 1 1 1 1 1 4 2 2 1 10 1 13 2 2 4 21 3 11 2 1 39 SB6 RAW MATERIAL 6 1 3 1 1 6 1 SB5 indet. metavolcanic Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite GRAND TOTAL GRAND TOTAL 2 6 8 LEVEL Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Western Rhyolite White Quartz GRAND TOTAL RAW MATERIAL GRAND TOTAL 1 2 3 4 4 14 LEVEL Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL RAW MATERIAL 1 GRAND TOTAL 1 6 4 11 LEVEL SB3 RAW MATERIAL 8 2 22 4 2 5 6 7 GRAND TOTAL 1 2 43 7 16 9 1 79 LEVEL 1 2 3 4 2 19 21 2 1 9 12 1 1 4 6 12 1 1 8 GRAND TOTAL 1 3 7 35 46 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 210 Chapter 6. 31HK2510 Table 52. Vertical Distribution of Debitage, Sample Blocks 7—12, 31HK2510. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 7-12:31HK2510 SB7 RAW MATERIAL LEVEL 1 Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 2 3 4 2 14 1 9 2 1 5 1 1 1 17 3 4 1 1 19 SB8 RAW MATERIAL 1 2 4 1 1 6 2 3 2 14 5 1 1 2 2 6 1 8 2 3 16 1 1 1 19 2 4 1 13 18 26 3 2 4 5 6 7 8 21 1 21 2 8 3 1 22 2 25 18 3 1 2 24 12 5 6 7 1 3 1 3 1 6 1 8 6 7 8 1 1 2 3 1 5 6 GRAND TOTAL 1 31 1 5 16 2 4 60 GRAND TOTAL 1 48 5 9 1 64 GRAND TOTAL 1 73 9 2 4 89 LEVEL 1 2 3 4 SB11 8-10 GRAND TOTAL 1 10 1 12 LEVEL 1 2 Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 3 4 5 1 1 1 1 1 1 SB12 8 5 Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite GRAND TOTAL Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 7 4 SB10 RAW MATERIAL 8 LEVEL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL RAW MATERIAL 7 3 SB9 RAW MATERIAL 6 LEVEL Rose Quartz Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL RAW MATERIAL 5 3 3 2 1 1 6 7 GRAND TOTAL 1 2 5 8 LEVEL 1 2 3 4 5 3 5 3 5 8 1 9 2 2 8 GRAND TOTAL 16 3 19 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 211 Chapter 6. 31HK2510 Table 53. Vertical Distribution of Debitage, Sample Blocks 13—18, 31HK2510. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 13-18:31HK2510 SB13 RAW MATERIAL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL LEVEL 1 2 3 4 4 17 1 3 4 2 23 4 10 1 1 1 1 6 6 7 8 5 6 7 8 4 5 6 7 8 5 2 1 3 6 1 1 5 6 7 8-10 6 7 8 1 2 1 11 3 1 5 6 7 2 SB14 RAW MATERIAL 5 1 1 2 LEVEL 1 2 3 4 GRAND TOTAL SB15 1 Type I Rhyolite Tuff Uwharries Eastern Rhyolite White Quartz GRAND TOTAL 1 2 3 1 1 4 9 SB16 GRAND TOTAL 8 1 9 18 LEVEL RAW MATERIAL 1 Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 1 2 3 1 1 2 9 12 1 3 5 4 1 1 1 2 1 2 SB17 1 2 GRAND TOTAL 1 1 2 1 3 15 23 LEVEL RAW MATERIAL 1 2 3 4 5 Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 4 5 12 10 6 1 1 1 3 1 4 4 1 9 1 14 5 1 3 19 SB18 Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL GRAND TOTAL 0 LEVEL RAW MATERIAL RAW MATERIAL GRAND TOTAL 25 2 6 1 1 7 42 GRAND TOTAL 37 1 8 11 4 1 62 LEVEL 1 1 1 2 3 4 1 10 2 4 2 5 1 2 6 3 7 21 6 1 5 1 7 24 5 11 8 GRAND TOTAL 15 2 18 2 4 22 63 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 212 Chapter 6. 31HK2510 Table 54. Vertical Distribution of Stone Tools by Sample Block, 31HK2510. VERTICAL DISTRIBUTION OF STONE TOOLS IN SAMPLE BLOCKS: 31HK2510 STONE TOOL TYPE LEVELS 1 2 3 4 5 6 7 8 GRAND TOTAL SB1 Palmer I Corner Notched (USR2) 1 Slab/Tabular Core (Metased) 1 1 1 SB2 Utilized Flake (Type I RT) 1 1 Yadkin Lg. Tri., Var. E (USR2) 1 1 Projectile Point Frag. (UER) 1 1 Type III Biface/Preform (USR1) 1 1 SB3 Projectile Point Frag.(WQtz) 1 1 Type I Side Scraper (WQtz) 1 1 Side Scraper Frag. (UWR) 1 1 SB4 Multidirectional Core (WQtz) 1 1 Unidirectional Core (WQtz) 1 1 Unidirectional Core (WQtz) 1 1 Type III End Scraper (WQtz) 1 Utilized Flake (USR1) 1 1 1 2 SB5 Small Dalton Point (USR1) 1 Utilized Flake (Type I RT) 1 Type IIb Side Scraper (CQtz) 1 1 1 Utilized Flake (Type I RT) 1 1 1 SB6 Pee Dee Serrated (UEW) 1 Flake Blank (USR2) 1 1 1 Dalton Adz (Type I RT) 1 1 Type III Biface Frag.(USR2) 1 1 Projectile Point Frag. (USR2) 1 1 SB7 Side Scraper Frag. (USR1) 1 1 Type III Biface Frag. (USR2) 1 Utilized Flake (Type I RT) 1 1 1 Hammer Stone/Anvil (Qtzite) 1 1 Morrow Mountain I St. (USR2) 1 1 Type II Biface Frag. (Type I RT) 1 1 Utilized Flake (Type I RT) 1 1 Biface Frag. (USR1) 1 1 Biface Frag. (USR2) 1 1 SB8 Utilized Flake (Type I RT) Utilized Flake (USR1) 1 1 1 1 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 213 Chapter 6. 31HK2510 Table 54. Vertical Distribution of Stone Tools by Sample Block, 31HK2510 (Continued). VERTICAL DISTRIBUTION OF STONE TOOLS IN SAMPLE BLOCKS: 31HK2510 STONE TOOL TYPE LEVELS 1 2 3 4 5 6 7 8 GRAND TOTAL SB9 1 Palmer I Corner Notched (USR2) Grinding Stone? (Fe Qtzite) 1 1 1 SB11 Type IV End Scraper (USR) 1 1 SB12 1 Projectile Point Frag. (USR2) 1 2 1 Type I Biface Frag. (USR2) 1 1 Type IIa Side Scraper (USR2) 1 SB13 Hammer Stone Spall (Qtzite) 1 1 SB15 Utilized Flake (USR2) 1 1 1 Core Fragment (WQtz) 1 SB17 Type IV Side Scraper (USR2) 1 1 Utilized Flake (Type I RT) 1 1 Biface Frag. (USR2) 1 1 1 Projectile Point Frag. (USR1) 1 Projectile Point Frag. (USR2) 1 1 Palmer I Corner Notched (USR2) 1 1 Utilized Flake (Type I RT) 1 1 End Scraper Frag. (Type I RT) 1 1 Utilized Flake (Type I RT) 1 Retouched Flake (UWR) 1 1 1 1 Utilized Flake (Type I RT) 1 SB18 Core Frag. (WQtz) 1 Yadkin Lg. Tri., Var. B (UWR) 1 1 1 Projectile Point Frag. (UER) 1 1 Core Frag. (WQtz) 1 1 Utilized Flake (Type I RT) 1 Utilized Flake (Type I RT) SB10, SB14 and SB18). In only SB18, however, was there firm evidence of debitage in the Woodland zone of the deposit. Earlier, it was projected that portions of a mean of 81 individual ceramic vessels exist on the site, based on Stage I and II shovel test results. There is good reason to believe, however, that this method of calculation underestimates the actual vessel popu- 1 1 1 lation, as was discussed earlier in Chapter 5. It was also concluded from that discussion that it is unlikely that many full pot concentrations survived on these sites due to recycling of large sherds and off-site transport of the fragments. An idea of the magnitude of the sherd population at the site can be supplied by calculating a straight area mean. The sample of 399 Stage I and II shovel tests sums to an area of 35.98 m2. The site area is approximately 11,200 m2, which Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 214 Chapter 6. 31HK2510 yields an area sample of 0.32 percent. Fifteen sherds, weighing 63.16 gm, were recovered from the sample, resulting in a population mean estimate of sherds of 4,688 and a total sherd weight of 19.7 kg. If whole pots were entirely represented in the assemblage, it is doubtful that this total would be sufficient to cover the weight of only 15 typical open-mouth jars from the region (see Herbert and Irwin 2003:2). Given an average weight of 3 pounds (1,377 gm) for a typical pot, if there were 81 whole pots present on the site, a mean population weight of 112 kg would be required. Consequently, only about 18 percent of the expected vessel weight relative to the estimate of the vessel population is present on the site. Similar results were obtained at 31HK2502 where only about 10 percent of the expected sherd weight was found in the sample. This suggests that the intensity of use recycling is great on Sandhills sites and that large sherds from broken pots were probably curated and stored in caches or transported upon leaving camps, while smaller sherds were probably left at the point of breakage. If so, sherds from the same pot might be spread across multiple sites and a high degree of displacement and disorganization should be evident in the spatial distribution of ceramics on sites with short-term occupation records. SB 14 was established around a shovel test (N490/E460) that yielded a relatively large sherd of Hanover IIa Fabric Impressed weighing 11.93 gm to specifically investigate one of these sherd occurrences on the site. Eight shovel tests were excavated around the sherd at 1.25 m-intervals, none of which yielded additional sherds or even any lithic artifacts. Another shovel test (N410/E525) anchoring SB10 produced two indeterminate Hanover IIa sherds from the same vessel. Again, the eight shovel tests excavated around the anchor failed to yield not only additional sherds from the same vessel, but also no additional sherds were found at all. Three other shovel tests excavated respectively in SB1, SB7 and SB18 produced single sherds, but none of the member shovel tests yielded ceramics at all. Clearly, these sherd finds are not representative of pot breaks, nor do they evidence partial vessel aggregations. In contrast, all of these instances are consistent with the interpretation of limited primary refuse discard events in which the remaining item was broken of a larger sherd or partial vessel during use. All of the sample blocks contain a wide variety of lithic raw material types. Much of this heterogeneity can be attributed to occupation superimposition, but some of it may result from site functional variability. Type II residences characteristically contain heterogeneous lithic raw material profiles and the high density of scraper forms in the stone tool inventory suggests that many of the occupations at the site may represent Early Archaic Type II residences. The sample block frames generally cover an area of only about 2.5 m square, which is smaller than the average element size. Moreover, it would be uncommon for the sample block to expose nearly whole elements. However, a method was devised during the DO4 investigation (Cable 2010) to identify and distinguish Type I and III elements that are mostly contained within the extent of a sample block. Three measures related to shovel test outcomes were developed to describe this variability (Table 55). These are: (1) number of positive shovel tests within which the element occurs, (2) number of shovel tests with greater Table 55. Shovel Test Outcome Model for the Purpose of Identifying Element Types INFERRED ELEMENT TYPE NO. POSITIVE STPS NO. STPS WITH ≥ 5 DEBITAGE MEAN DEBITAGE/ POSITIVE TEST TYPE I HIGH HIGH HIGH TYPE I OR TYPE III HIGH MODERATE MODERATE TYPE III MODERATE LOW LOW Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 215 Chapter 6. 31HK2510 than or equal to 5 pieces of debitage and (3) mean debitage frequency per positive test. Since Type I residential elements are characterized by high density debitage concentrations they are expected to yield high outcome values for each of these measures. Type III residences, by contrast, are defined by the presence of low-density debitage scatters of about the same size as Type I high density scatters. Consequently, they should be expected to measure high or moderate on the number of positive shovel tests within which they are represented, but low on the number of shovel tests yielding five or more pieces of debitage of a specific raw material type and low on the mean frequency of debitage per positive shovel test. Given the structure of the sample blocks, there are generally nine outcomes (shovel tests) to measure. Some elements present measures that are intermediate between the two extremes and these can only be classified as Type I/III. Type II residential elements would be expected to reflect similar patterns to Type III residences, but they should contain multiple raw material types and a greater density of tools. Clearly, the identification of Type II residences in sample blocks is subjective and it would require additional confirmation from expanded sample blocks or test unit excavation to firmly identify such elements. Table 56 summarizes the result of applying the shovel test model to the elements defined by lithic raw material type within each sample block. Generally, only elements that were represented by at least three shovel tests were analyzed, as those with lower representation are regarded as insufficiently complete or ineffectively sampled in the sample block frame. As established by the Stage I and II element study, Type III elements predominate in the sample blocks. Seventy percent (n = 33) of the elements were classified as Type III, short-term residences. These elements are characterized by variably moderate to high shovel test representation (3 to 6 shovel tests), low frequencies of shovel tests containing greater than 5 pieces of debitage (0 to 1 shovel test) and debitage frequency means in positive shovel tests of less than 3.0. Type I elements were more numerous than expected, numbering eight of the 47 elements (17 percent) recognized for this study. These elements commonly had high shovel test representation (4 to 9 shovel tests), greater numbers of high debitage density shovel test outcomes (2 to 4 shovel tests) and high debitage mean densities, which ranged between 3.13 and 17.67 pieces of debitage. Seventeen percent (n = 8) of the elements exhibited intermediate characteristics and could not be differentiated. Cultural-chronological associations were inferred from depth profiles and from diagnostic stone tools that were matched to the raw material composition of the debitage concentrations. Sometimes diagnostic stone tool associations supplanted inferences based on depth patterns. In particular, some elements that occupied high vertical positions were assigned to the Early Archaic period based on diagnostic stone tool associations. This adjustment was found to be necessary in SB17. Most of the Early Archaic deposits, however, had expectedly deep debitage profiles consistent with associated scraper and projectile point types. Thirty-one Early Archaic, 11 Middle Archaic/ Early Archaic, 1 Middle Archaic, 5 Middle Archaic/ Woodland and 4 Woodland elements were identified in the study. Early Archaic elements included a Palmer I Corner Notched Type III element of USR (2) situated between 30 and 60 cm bs in SB1, a Small Dalton Type I/III element of USR (1) positioned at 20 to 40 cm bs in SB5, a Palmer I Corner Notched Type III element of USR (2) in SB9 at a depth of 30 and 60 cm bs and a Palmer I Corner Notched Type III element of USR (2) situated at 20 to 40 cm bs in SB17 (Table 56). Candidates for Type II elements were recognized in SB1, SB3, SB4, SB5, SB8, SB9, SB11, SB12, SB15, SB 17 and SB18 (see Table 56). These included all of the diagnostic Early Archaic Type III elements identified above. In these instances, groups of Type III elements could hypothetically belong to single Type II residences. Since such elements have only been positively associated with the Early Archaic period, only elements of this inferred affiliation were considered for this procedure. Type II residences are characterized by low-density scatters of debitage composed of multiple raw material types and relatively high densities of stone tools of heterogeneous raw material composition. It is suggested that these elements represent multiple family units that came together for short durations to conduct activities such as communal hunts that resulted in abnormally high tool discard rates related to processing animals. Each of the identified sample blocks meets these criteria. High numbers of tools are found with multiple “Type Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 216 Chapter 6. 31HK2510 Table 56. Inferred Element Types by Sample Blocks, 31HK2510. MEAN DEBITAGE/ + SHOVEL TEST 4 0 1.50 III Type II EA Uwharries Southern Rhyolite (2) 4 0 1.00 III Type II EA ASSOCIATION NO. STs WITH ≥ 5 DEBITAGE Uwharries Southern Rhyolite (1) ELEMENTS INFERRED ELEMENT TYPE NO. OF STs REPRESENTED TYPE II CANDIDATE INFERRED ELEMENT TYPES IDENTIFIED IN SAMPLE BLOCKS: 31HK2510 STONE TOOLS SB1 Metasedimentary Palmer I Corner Notched Slab/Tabular Core SB2 Type I Rhyolite Tuff 2 0 1.00 III Wood/MA Utilized Flake Uwharries Southern Rhyolite (1) 3 0 1.00 III Wood Type III Biface/Preform Yadkin, Lg. Tri., Var. E Uwharries Southern Rhyolite (2) 4 0 1.00 III MWood White Quartz 3 0 1.33 III Wood/MA Uwharries Eastern Rhyolite Projectile Point frag. SB3 White Quartz 6 0 1.50 III Uwharries Western Rhyolite Type II EA Projectile Point frag., Type I Side Scraper Type II EA Side Scraper frag. Type II EA Directional Cores (n=3), Type III End Scraper Type II EA Utilized Flakes (n=2) EA Utilized Flakes (n=2) SB4 White Quartz 3 0 1.00 III Uwharries Southern Rhyolite (1) SB5 Type I Rhyolite Tuff 8 3 5.38 I Uwharries Eastern Rhyolite 4 0 1.75 III Type II EA Uwharries Southern Rhyolite (1) 5 2 3.20 I/III Type II EA Uwharries Southern Rhyolite (2) 4 0 2.25 III Type II EA Type II EA White Quartz Small Dalton Point Type IIIb Side Scraper SB6 Uwharries Southern Rhyolite (1) 2 0 1.50 III Indet. Uwharries Southern Rhyolite (2) 6 0 1.17 III Indet. White Quartz 8 3 4.38 I Indet. Flake Blank, Type III Biface frag.,Projectile Point frag. Type I Rhyolite Tuff EA Dalton Adz Uwharries Eastern Rhyolite LWood Pee Dee Serrated EA/MA Utilized Flakes (n=2), Type II Biface frag. SB7 Type I Rhyolite Tuff 5 2 6.20 I Uwharries Southern Rhyolite (1) 5 0 1.00 III EA Side Scraper frag., Biface frag. MA Morrow Mountain I Stemmed, Type III Biface frag., Biface frag. Uwharries Southern Rhyolite (2) 4 2 4.00 I Uwharries Western Rhyolite 2 0 1.00 III EA/MA White Quartz 4 0 1.00 III EA/MA Fine-grained Quartzite Hammer Stone/Anvil SB8 Type I Rhyolite Tuff 6 3 6.00 I Uwharries Southern Rhyolite (1) 3 0 1.67 III Uwharries Southern Rhyolite (2) 5 1 1.80 III EA Utilized Flake Type II EA Utilized Flake Type II EA Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 217 Chapter 6. 31HK2510 Table 56. Inferred Element Types by Sample Blocks, 31HK2510 (Continued). 2 14.60 I 4 1 2.25 III Type II EA Uwharries Southern Rhyolite (2) 2 0 1.00 III Type II EA White Quartz 3 0 1.33 III Type II EA ASSOCIATION 5 Uwharries Southern Rhyolite (1) ELEMENTS INFERRED ELEMENT TYPE MEAN DEBITAGE/ + SHOVEL TEST Type I Rhyolite Tuff NO. OF STs REPRESENTED NO. STs WITH ≥ 5 DEBITAGE TYPE II CANDIDATE INFERRED ELEMENT TYPES IDENTIFIED IN SAMPLE BLOCKS: 31HK2510 STONE TOOLS SB9 EA Ferrous Quartzite Palmer I Corner Notched Grinding Stone SB10 Uwharries Southern Rhyolite (2) 5 1 2.00 III EA Uwharries Southern Rhyolite (2) 2 0 1.00 III Type II EA White Quartz 4 0 1.25 III Type II EA Type II EA Type IV End Scraper Projectile Point frag., Type I Biface frag., Type IIa Side Scraper SB11 Uwharries Southern Rhyolite (1) SB12 Uwharries Southern Rhyolite (2) 4 1 4.00 I/III Type II MA/EA White Quartz 2 0 1.50 III Type II MA/EA Mill Mountain Rhyolite 9 2 2.78 I/III MA/EA Uwharries Eastern Rhyolite 3 0 2.00 III MA/EA White Quartz 4 0 1.75 III MA/EA SB13 Fine-grained Quartzite Hammer Stone Spall SB15 Type I Rhyolite Tuff 6 0 1.33 III Type II EA White Quartz 4 1 2.25 III Type II EA Uwharries Southern Rhyolite (2) Core frag. Utilized Flake SB16 Uwharries Southern Rhyolite (2) 3 0 1.00 III Wood/MA White Quartz 4 1 3.75 I/III Wood/MA SB17 Type I Rhyolite Tuff 11 3 3.37 I/III Type II EA Utilized Flake (n=4), End Scraper frag., Uwharries Southern Rhyolite (1) 6 0 1.33 III Type II EA Projectile Point frag. Uwharries Southern Rhyolite (2) 6 0 1.83 III Type II EA Type IV Side Scraper, Biface frag., Projectile Point frag., Palmer I Corner Notched Uwharries Western Rhyolite 3 0 1.33 III Type II EA Retouched Flake Mill Mountain Rhyolite 4 2 3.75 I/III MA/EA Uwharries Eastern Rhyolite 6 2 3.00 I/III MA/EA Uwharries Southern Rhyolite (2) 2 0 2.00 III White Quartz 7 1 3.14 I/III SB18 MA/EA Uwharries Western Rhyolite Type I Rhyolite Tuff Projectile Point frag. Wood/MA Type II Core frag. (n=2) MWood Yadkin Lg. Tri., Var B; EA Utilized Flakes (n=2) Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 218 Chapter 6. 31HK2510 III” debitage scatters of a depth or association to be consistent with an Early Archaic affiliation. Type II elements are rare and not well understood in the region. A similar magnitude of Type II candidates of Early Archaic age were identified at 31HK2502, but at 31HK2510, the intensity of superimposition is much lower and the clarity of the deposits is much greater. This is very effectively demonstrated by the deeply isolated and narrow debitage profiles seen in SB1, SB3, SB4, SB8, SB9 and SB10 (see Tables 51 and 52). These locations would supply ideal situations in which to evaluate and define Type II and Type III Early Archaic elements. TEST UNITS At the time of fieldwork, the character of the Early Archaic occupation was not fully appreciated, as many of the scrapers were not recognized until the analysis phase, just as was the case at 31HK2502. Nevertheless, deposits were encountered that provide an opportunity to evaluate the possibility that Type II elements exist at 31HK2510. The number of test units would have been increased, had the potentially high Type II element frequency been realized at the time of fieldwork. Four 1-x-1-m test units were excavated at 31HK2510 (Table 57). Each unit was quartered and excavated in arbitrary 10 cm levels and each was terminated near the top of the Bt-horizon. The precise location of each test unit was determined after closeinterval shovel testing was deployed around targeted shovel tests. The test units were placed in SB4 (TU1 and TU4), SB6 (TU2) and SB9 (TU3), all on the southern end of the site (Figure 49). SB4 produced an unusual assemblage of four quartz tools (Type III End Scraper) and cores (directional cores). Two contiguous units, TU1 and TU4, were ultimately excavated here to better define the surrounding artifact patterns associated with what initially appeared to have been a cache. SB6 produced a Dalton Adz and a relatively dense white quartz debitage concentration. TU2 was placed here to further explore the relationship between the adz and the debitage concentration. Finally, SB9 produced evidence of a dense debitage concentration of Type I Rhyolite Tuff that was interpreted to represent a Type I element. Several Type I elements composed of Type I Rhyolite had been identified in three other sample blocks (i.e. SB5, SB7 and SB8) and TU3 was excavated to more fully investigate the structure, function and cultural association of one of these. A Palmer I Corner Notched point made of USR(2) was also recovered from SB9, in the same vertical position as the Type I Rhyolite concentration. Because the raw material did not match, the Palmer I point was assumed not to be associated with the concentration. This relationship was further explored as well by the TU3 excavation. The results for each test unit are presented below. TABLE 57. Summary Data for Test Units, 31HK2510 SUMMARY OF TEST UNITS: 31HK2510 TEST UNIT SIZE NORTH EAST NO. LEVELS MAX. DEPTH (cm bd) VOLUME 3 (m ) ARTIFACT DEPTH (cm bd) TU1 1x1m 370.25 480.25 5 50 0.50 50 TU2 1x1m 400.00 495.50 4 40 0.40 40 TU3 1x1m 389.25 486.50 7 70 0.70 70 TU4 1x1m 370.25 480.25 5 50 0.50 50 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 219 Chapter 6. 31HK2510 Test Units 1 and 4 (SB 4) TUs 1 and 4 were placed contiguous to one another at, respectively, grid coordinates N370.25/ E480.25 (NE corner) and N370.25/E481.25 (NE corner) in the east central portion of SB4. Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of the unit (Figure 62). Three soil layers were exposed in the wall profiles, not including the upper humic zone of the Ap-horizon which was charcoal laden from prescribed burning activities. The Ap-horizon was positioned in the upper 12 to 15 cm of the solum. It was composed of brown (10YR 5/3) sand. Its contact with the underlying E-horizon was regular and relatively even, indicating that it represented a plow zone. Plow scars, however, were not observed in profile or from horizontal exposure. The fine sand E-horizon transitioned from a darker (10YR 7/4) to lighter (10YR 7/3) shade of very pale brown at about 45 cm bd. The upper zone is identified as the E1-horizon, while the lower zone is referred to as an E2-horizon. The units were terminated at 50 cm bd, approximately 7 cm above the Bt-horizon, because the objectives of the excavation had been met and further downward extension of the units was not considered necessary. SB 4 was established around a Stage I shovel test that yielded three white quartz directional cores (Figure 50:A-C) and a white quartz Type III End Scraper (Figure 53:B). These items were removed in a single shovel test scoop at a depth of 30 to 40 cm bs, which suggested that they may have been cached and that additional members of the cache would likely be found in the immediate vicinity of the shovel test (N370/E480). Additional tools or cores that might have been associated with a cache were not recovered from the eight close interval shovel tests forming SB4. Artifact density was extremely low, consisting of nine pieces of debitage of rhyolite and white quartz (Table 51) and two utilized flakes of USR (1) (Table 54). Debitage distributions suggested that the presence of portions of several Type III elements, but the primary element represented in the area of SB4 was a white quartz scatter potentially associated with the tool and core cluster by virtue of matching raw material composition (see Table 56). The material in the sample block was situated between 40 and 60 cm bs, suggesting that the area contained an Early Archaic occupation surface with little or no mixing from occupations of later phases. The Early Archaic affiliation of the surface is further supported by the Type III End Scraper from the Stage I shovel test. Based on artifact distributions in SB4, TU1 was placed in the central portion of the sample block immediately over the area of the Stage I shovel test. A number of additional white quartz stone tools and cores were recovered (Table 58), including a Type I Side Scraper fragment, two split cores (Figure 50:DE) and a core fragment. The spatial distributions of the quartz tools and cores in TU1, it would appear most likely that the end scraper and cores found in the shovel test are not members of a cache, but part of a tool cluster associated with the living floor of a short term camp (Figure 63). TU4 was placed immediately east of TU1 to search for additional tools and to more fully trace the distribution of the quartz debitage scatter. Additional white quartz tools and cores were not recovered from the excavation of TU4, but more quartz debitage was encountered in the southern part of the unit. Density contours of debitage from test unit quadrants and shovel tests within the sample block (densities were standardized for 0.25 m2 areas), a white quartz debitage scatter was defined measuring a little more than 2 m in length and about 1 m in width. The white quartz tools and cores were confined to an area of about 0.6 m in diameter on the west side of the scatter, suggesting that this area served as a central activity zone organized around a hearth. In addition to the white quartz element, very low-density debitage scatters of UER, USR (1), USR (2), UWR and Mill Mountain Rhyolite were also detected in SB4 and in the test unit excavations (Table 58). In addition, stone tools with matching raw material were identified for USR (2), consisting of two utilized flakes and a projectile point fragment, and UER, represented by a utilized flake (Figure 63). Of course, each of these elements may represent separate occupations, but there is equally strong evidence that most of them belong to a Type II residence or residences. Type II residences are similar in spatial organization to Type I residences, but they exhibit much higher tool-to-debitage ratios and may commonly be comprised of diverse lithic raw materials. It is speculated that raw material heterogeneity is a function of groups travelling from different locales to aggregate at selected places on the landscape at specific and planned times. In these situations, it is Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 220 Chapter 6. 31HK2510 SOUTH WALL cm bd 0 H Lev 1 Ap 10 Lev 2 20 Tree Roots Lev 3 30 E1 Lev 4 40 E2 TU4 Lev 5 50 TU1 EN369.25/E481.25 N369.25/E479.25 0 2m Soil Strata H: Very Dark Gray (10YR 3/1) Prescribed Burn Charcoal and Loamy Sand A: Brown (10YR 5/3) Sand Plow Zone E1: Very Pale Brown (10YR 7/4) Fine Sand E2: Very Pale Brown (10YR 7/3) Fine Sand Figure 62. Profile Drawing, TU1 and TU4, South Wall, 31HK2510. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 221 Chapter 6. 31HK2510 Table 58. Stone Tools and Debitage by Level for TUs 1 and 4 and SB 4, 31HK2510. STONE TOOLS AND DEBITAGE BY LEVEL FOR TUs 1 AND 4 AND SB4: 31HK2510 TU1 RAW MATERIAL TYPE Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 1 2 1 1 2 TU4 RAW MATERIAL TYPE Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 1 2 1 1 2 SB4 RAW MATERIAL TYPE Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite 1 2 DEBITAGE LEVELS 3 4 5 1 1 1 1 6 9 3 6 10 DEBITAGE LEVELS 3 4 5 1 1 1 3 1 1 2 2 1 4 4 4 10 7 DEBITAGE LEVELS 3 4 5 1 STONE TOOLS LEVEL 4 GRAND TOTAL 1 1 2 17 21 UF I SS, Split Core (2), Core Fr., Biface Fr. STONE TOOLS LEVEL 3 GRAND TOTAL 1 2 4 6 10 23 6 1 3 1 GRAND TOTAL 1 4 1 White Quartz 2 1 3 GRAND TOTAL 3 6 9 PPK Fr. STONE TOOLS LEVELS 4 5 6 UF UF M Dir Core, Uni Core (2) III ES Core Fr.: Core Fragment, Biface Fr.: Biface Fragment, III ES: Type III End Scraper, I SS: Type I Side Scraper, M Dir. Core: Multidirectional Core, PPK Fr.: Projectile Point Fragment, Uni Core .: Unidirectional Core, UF: Utilized Flake likely that individual groups will be provisioned with a different assemblage of raw material types, even if the groups were foraging within the same general region. Lithic reduction on these sites consisted of tool maintenance and late stage shaping of tools from flake blanks or reduced biface cores. Consequently, debitage density is much lower than that of Type I debitage concentrations and the scatters consist primarily of late stage reduction debitage. Type II elements are distinguished from Type III elements by tool density. The former is characterized by high tool density, while the latter has few discarded tools relative to debitage. Type II occupations are inferred to represent the output from single or small, multiple household aggregations and are thought to constitute the residue from High Technology Forager (HTF) residences (Spiess 1984; Todd 1983:231-233). The HTF model proposes a specialized forager adaptation combining high logistical and high residential mobility into a single settlement system characterized by highly curated technologies. Each of the identified elements are characterized by small core flakes and BTFs and only small to moderately sized FBRs, which is not indicative of major tool production episodes, as one would expect for a Type I residence. The total weight of all debitage recovered from SB4 and TUs 1 and 4 is only 25.89 gm, which is only about 0.9 ounces. The raw material weight tied up in tools and cores far exceeds the associated debitage. The combined weight of all tools and cores is 321.54 gm, which is more than twelve times that of debitage weight. These characteristics are well within the expectations for Type II residences and, although the exposure is not sufficient to demonstrate this contention, there is a strong possibility that all of the elements are members of at least one Type II residence. Because of the heterogeneous nature of the raw material composition of such elements, the degree to which occupational deposits are mixed cannot be objectively evaluated without seeing larger distributional patterns in the area of SB4. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 222 Chapter 6. 31HK2510 Sample Block 4 479.25E 481.25E 370.25N 5 3 7 TU1 TU4 370.25N 5 369.25N 369.25N 479.25E 481.25E N 0 2m WHITE QUARTZ TOOLS Directional Core Split Core RHYOLITE TOOLS SB4 Shovel Test Type III End Scraper Utilized Flake (URS1) Type I Side Scraper Utilized Flake (UER) Core Fragment Projectile Point Fragment (URS2) 5 White Quartz Debitage Density (Contours = 1 item per 0.25 m2) Figure 63. Early Archaic Living Floor, Sample Block 4, 31HK2510. Test Unit 2 (SB6) TU2 was placed at grid coordinates N400.0/ E495.5 (NE corner) in the east central portion of SB6. Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of the unit (Figure 64). Four soil layers were exposed in the wall profiles, not including the upper humic zone of the Ap-horizon, which was charcoal laden from prescribed burning activities. . The Ap-horizon was positioned in the upper 15 cm of the solum. It was composed of brown (10YR 5/3) sand. Its contact with the underlying E-horizon was regular and relatively even, indicating that it represented a plow zone. Plow scars, however, were not observed in profile or from horizontal exposure. The fine sand E-horizon transitioned from a lighter (10YR 7/3) to darker (10YR 7/4) shade of very pale brown at about 32 cm bd. The upper zone is identified as the E1-horizon, while the lower zone is referred to as an E2-horizon. The unit was terminated Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 223 Chapter 6. 31HK2510 NORTH WALL cm bd 0 Shovel Test H H Tree Roots Ap H Lev 1 10 Lev 2 TR 20 E1 TR E2 Lev 3 30 Lev 4 Bt N400.0/E494.5 TU2 0 40 N400.0/E495.5 1m Soil Strata H: Very Dark Gray (10YR 3/1) Prescribed Burn Charcoal and Loamy Sand Ap: Brown (10YR 5/3) Sand Plow Zone E1: Very Pale Brown (10YR 7/3) Fine Sand E2: Very Pale Brown (10YR 7/4) Fine Sand TR: Burned Root Disturbance Bt: Light Brown (7.5YR 6/4) Sandy Clay Loam Figure 64. Profile Drawing, TU2, North Wall, 31HK2510. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 224 Chapter 6. 31HK2510 quantity of quartz debitage is larger by a factor of at least 1.5. A total weight value of 215 gm, however, is equivalent to at least 3 directional cores like the ones recovered in SB4, which average 54 gm per core. The recognizable debitage type inventory for the concentration consists of 44 percent core flakes, indicating that the debris was generated primarily through the reduction of directional cores. Moreover, only a single tool of matching raw material was recovered, a projectile point tip. These are clearly the characteristics of a Type I element and stand in stark contrast to the more common Type II and III elements at the site. at 40 cm bd, approximately 3 cm into the light brown (7.5YR 6/4) sandy clay loam Bt-horizon. The unit was placed in the area of the mid-slope of the terrace, where shovel test data show conditions of soil deflation, probably related to alluvial fan formation. SB6 was excavated around a Stage II shovel test yielding a Dalton Adz, a Pee Dee Serrated point and a relatively dense white quartz debitage concentration in the upper 30 cm of deposit (Table 59). The Pee Dee Serrated projectile point was made of UER and was recovered in the upper 10 cm of deposit. White quartz debitage, which was also positioned high in the deposit, was widely distributed across SB6 (Table 56) and displayed the apparent characteristics of a Type I element. Excavation of TU2 succeeded in recovering an additional large sample of white quartz debitage situated between 0 and 20 cm bs. Nearly 42 gm of white quartz debitage was recovered in the SB6 investigation. It is estimated that about 215 gm (7.5 ounces) of quartz debitage is contained within the sample frame, based solely on the shovel tests contents in the 2.5 m square area (6.25 m2) of SB6. The concentration extends beyond the confines of the sample block, however, and it is likely that the actual The age of the Type I element is not clear, but available evidence suggests that it is Late Woodland in affiliation and probably associated with the Pee Dee Serrated point found in the same general vertical position. The linkage between the two is supplied by the quartz projectile point tip (a1101). The morphology of the tip (Figure 65) is consistent with attenuated forms of Pee Dee Triangular and Pee Dee Serrated points (see Coe 1995:202-203). It was small, delicately flaked and exhibited a narrow blade that appeared to be turning outward at the posterior end to complete a recurved blade outline common to TABLE 59. Stone Tools and Debitage by Level for TU 2 and SB 6, 31HK2510. STONE TOOLS AND DEBITAGE BY LEVEL FOR TU2 AND SB6: 31HK2510 TU2 RAW MATERIAL TYPE indet. metavolcanic Mill Mountain Rhyolite Slate Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz Grand Total 1 13 13 2 1 1 1 2 1 3 1 20 30 SB6 RAW MATERIAL TYPE Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 1 2 19 21 2 2 1 9 12 DEBITAGE LEVELS 3 4 GRAND TOTAL 1 2 1 2 1 11 2 33 53 1 6 1 2 8 2 DEBITAGE LEVELS 3 4 5 1 1 4 6 1 12 1 6 1 3 PPK Fr. GRAND TOTAL 1 3 7 35 46 STONE TOOLS LEVELS 2 1 FB STONE TOOLS LEVELS 3 Dalton Adz III Bif Fr., PPK Fr. Dalton Adz: Dalton Adz, PPK Fr.: Projectile Point Fragment, III Bif: Type III Biface Fragment Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 225 Chapter 6. 31HK2510 both types. The blade is only 7.15 mm wide at the posterior end of the tip, which is 15.08 mm long. The recovered Pee Dee Serrated point and the white quartz tip were found not more than 0.50 to 0.75 m from one another. Tool to debitage ratios for weight (0.02:1) and count (0.03:1) incorporating both projectile point specimens into the calculation are low, and consistent with the characteristics of a Type I element. ing the Dalton Adz were recovered. These were positioned between 10 and 30 cm bs/bd, which is consistent with the position of the adz in this deflated area of the site. Even with compressed deposits, however, it would be possible to analytically segregate the inferred Late Woodland period Type I element from the earlier Archaic occupations. Ceramics were not recovered in the sample, but ceramic and lithic distributions in Woodland period occupations tend to be paired and mutually exclusive. Ceramics were recovered from a shovel test 5 m south of the SB4 anchor test. Individual sherds from two separate vessels were represented, a Hanover IIa indeterminate sherd and a Hanover IIIa Fabric impressed sherd (Figure 61). The latter is of a hypothetical age to be associated with the Pee Dee Test Unit 3 (SB 9) Figure 65. Pee Dee Projectile Points, TU2, 31HK2510. Serrated point and the white quartz Type III element. Whether ceramics are associated with each Woodland period lithic deposit, however, is not yet determined. Other stone tools recovered in the SB6 area included the lateral section of a Type III Biface fragment and a non-diagnostic projectile point base fragment, each made of USR (2). Both tools were relatively large and most probably not associated with the white quartz debitage concentration. Small amounts of matching debitage were recovered in the area, suggesting that these items comprise a portion of a Type II or III element of unknown affiliation. Only three matching pieces of debitage (Type I Rhyolite) match- TU3 was placed at grid coordinates N389.25/ E486.50 (NE corner) in the southeast quadrant of SB9. Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of the unit (Figure 66). Four soil layers were exposed in the wall profiles, not including the upper humic zone of the Ap-horizon, which was charcoal laden from prescribed burning activities. The Ap-horizon was positioned in the upper 15 cm of the solum. It was composed of brown (10YR 5/3) sand. Its contact with the underlying E-horizon was regular and relatively even, indicating that it represented a plow zone. Plow scars, however, were not observed in profile or from horizontal exposure. A tree root burn was present in the very northern end of the unit and it partly truncated by the plow zone. Root stains extended down across most of the vertical profile in the northeast corner of the unit. Much of the organic matter in the stains had deteriorated, suggesting that it represented a relatively old event. The lower Ap-horizon across most of the unit transitions into the E1-horizon at about 15 cm bd. The fine sand E-horizon transitioned from light yellowish brown (10YR 6/4) to very pale brown (10YR 7/4) at 40 to 45 cm bd. The upper zone is identified as the E1-horizon, while the lower zone is referred to as an E2-horizon. The unit was terminated at 70 cm bd, approximately 4 cm above the Bt-horizon. SB9 produced evidence of a dense debitage concentration of Type I RT that was interpreted to represent a Type I element. Several Type I elements composed of Type I Rhyolite had been identified in four other sample blocks (i.e. SB5, SB7, SB8 and SB17) and TU3 was excavated to more fully investigate the structure, function and cultural association of one of these. A Palmer I Corner Notched point (a454, Figure 52:C) made of USR (2) was also recovered from SB9, in the same vertical position as the Type I Rhyolite concentration. Because the raw material Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 226 Chapter 6. 31HK2510 EAST WALL cm bd 0 H Ap Tree Roots Lev 1 10 Lev 2 20 Lev 3 E1 30 E1 Lev 4 TR 40 Lev 5 50 TR E2 Lev 6 60 Lev 7 70 N389.25/E486.50 TU3 0 N388.25/E486.50 1m Soil Strata H: Very Dark Gray (10YR 3/1) Prescribed Burn Charcoal and Loamy Sand Ap: Brown (10YR 5/3) Sand Plow Zone E1: Light Yellowish Brown (10YR 6/4) Fine Sand E2: Very Pale Brown (10YR 7/4) Fine Sand TR: Burned Root Stain Figure 66. Profile Drawing, TU3, East Wall, 31HK2510. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 227 Chapter 6. 31HK2510 did not match, the Palmer I point was considered not likely associated with the concentration. This relationship was further explored as well. The Type I RT debitage concentration was centered between 30 and 60 cm bs in the shovel test sample (Table 60). The Palmer I Corner Notched point was found between 30 and 40 cm bs and was matched to two pieces of USR (2) debitage, suggesting that it represented a Type III element of unknown association with the Type I RT debitage. Other low-density Type III elements were identified from shovel testing, including MMR, USR (1) and white quartz. All of these were positioned relatively low in the matrix, in the same general vertical zone as the Palmer point and the Type I Rhyolite. This indicated that the occupation here appeared to constitute an Early Archaic “living surface,” but the cultural and functional relationships between the various elements was unclear. In addition to debitage, a arenite abrader (Figure 54:A) was recovered between 40 and 50 cm bs and is also associated with the Early Archaic zone. TU3 was excavated between 0.5 m and 1.5 m west and slightly south of the anchor shovel test that produced the Palmer point. This location was near the inferred edge of the Type I RT debitage concentration on the western side of the sample block and the unit was placed here in hopes of finding an associated tool cluster that might be instrumental in ascertaining the culture-chronological affiliation of the Type I element. Along with a large collection of Type I RT debitage a matching LeCroy Bifurcate Stem projectile point (a1176, Figure 52:E) was recovered at a depth of 40 to 50 cm bd. Chapman (1985:Table 7-1) dates the type to 6500 to 5800 radiocarbon years BC, which indicates that it post-dates Palmer/Kirk Corner Notched and St. Albans Side Notched and pre-dates Kirk Stemmed and Serrated. This confirms a sequence of occupation in the Early Archaic zone of SB9 and it provides support for the idea that the other Type I RT debitage concentrations identified in sample blocks at the site may also be affiliated with Bifurcate Tradition occupations. The Type I RT debitage from SB9 exhibits characteristics that conform to the overall Type I element model. In total, 14.72 grams of debitage of this material was collected in shovel tests and it can be estimated from this sample that approximately 114 gm of debitage is contained within the frame of the sample block. This would equal the weight of one Table 60. Stone Tools and Debitage by Level for TU 3 and SB 9, 31HK2510. STONE TOOLS AND DEBITAGE BY LEVEL FOR TU3 AND SB9: 31HK2510 TU3 RAW MATERIAL TYPE Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 1 1 1 2 1 2 3 7 1 6 20 7 GRAND TOTAL 6 3 119 6 3 1 1 127 7 GRAND TOTAL 2 1 2 1 4 8 SB9 RAW MATERIAL TYPE Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz Ferrous Quartzite GRAND TOTAL DEBITAGE LEVELS 4 5 1 42 41 1 2 3 1 5 6 43 42 DEBITAGE LEVELS 4 5 21 1 22 22 6 8 3 1 2 18 3 1 2 1 73 9 2 4 25 24 12 89 21 2 STONE TOOLS LEVELS 2 5 LeCroy Core Fr. (3) 4 STONE TOOLS LEVELS 5 6 PI CN Abrader Abrader: Abrader, Core Fr: Core Fragment, PI CN: Palmer I Corner Notched, LeCroy: LeCroy Bifurcate Stem Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 228 Chapter 6. 31HK2510 larger or two smaller directional cores of the kind described for SB4. Cortex was present on 6.9 percent of the debitage and core flakes comprised 22.6 percent of the shovel test sample. The average weight of the debitage was low, about 0.20 gm per piece. This would suggest that the LeCroy group that created the debitage participated in the reduction of two or more prepared directional cores. The chances are good that some of the raw material left the site in flake blank or directional core packages, indicating that it is likely that more than two cores were initially brought to the site. The characteristics of the other Type I RT elements are similar (Table 61). Total estimated debitage weight within the sample frames ranges between 97 and 164 gm, and all profiles would suggest directional core reduction to produce blanks for tools. All of the other elements yielded tools, primarily utilized flakes detached from directional cores. Other tools present included a Type III biface fragment and a small portion of an end scraper bit. Had these represented Type II elements, it is likely that many more scrapers would have been found. Tool to debitage ratios are low to moderate by count, ranging from 0.01:1 to 0.13:1, although the weight ratios are higher. Most of the matching tools, however, are utilized flakes. When only the formal tools are used in the calculation, the ratio for all five elements combined is 0.11:1. The culture-chronological affiliations of the other four Type I RT, Type I elements is clearly not determined by the investigation. The overall structural similarity of the deposits and the fact that Bifurcate Tradition points are strongly associated with Type I Rhyolite Tuff (see Cable and Cantley 2005a, b; Cable and Cantley 2006; Cable 2010), supports this overall cultural assignment. The depth at which the elements were found is also consistent with a middle Early Archaic age. The Type I RT elements in SB5 and SB17 are situated within several meters of one another (see Figure 48) and they may represent a multiple social unit camp. It is probable that long-term land-use patterns shifted throughout the Archaic in ways that should be evident in the nature and density of occupation types on individual sites. Two major Early Archaic settlement themes on 31HK2510 have been identified. The earliest one consists of Type II and Type III residences dating to the Dalton, Hardaway and Palmer phases and the later theme is characterized by short-term Type I residences established by later Early Archaic phases, in particular Bifurcate Tradition components. Type I RT occurs as lowerdensity Type II or Type III elements, as well, and it is likely that these elements have a wider culturalchronological occurrence throughout the Early Archaic period. Table 61. Debitage Characteristics of Type I Elements Composed of Type I RT Material, 31HK2510. DEPTH (cm bs) FREQUENCY OF DEBITAGE WEIGHT (gm) AVERAGE WEIGHT/ DEBITAGE PIECE (gm) ESTIMATED WEIGHT OF ALL DEBITAGE (gm) % CORE FLAKES % CORTEX DEBITAGE CHARACTERISTICS OF TYPE I RT, TYPE I ELEMENTS: 31HK2510 SB5 20-50 43 12.58 0.29 97.07 12.5 0.0 Utilized Flake (n=2) SB7 20-40 31 12.59 0.41 97.15 47.1 6.5 Utilized Flake (n=3), Type III Biface Frag. SB8 30-60 48 21.37 0.45 164.89 31.6 27.1 Utilized Flake SB9 30-60 73 14.72 0.20 113.58 22.6 6.9 LeCroy Bifurcate Stem SB17 20-40 48 17.08 0.36 131.79 16.7 10.4 Utilized Flake (n=4), End Scraper Frag. TYPE I ELEMENTS (TYPE I RT) MATCHING TOOLS Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 229 Chapter 6. 31HK2510 EVALUATION Site 31HK2510 is situated on a ridge terrace immediately adjacent to Rockfish Creek in Hoke County. The broad and relatively homogeneous nature of the landform created relatively unconstrained access to suitable locations for camping. Ultimately, this resulted in low incidences of superimposition by separate occupations. Only 15 percent of the shovel tests contained more than two occupation elements. As a consequence, the archaeological deposits are not significantly mixed with another and a high degree of vertical and horizontal clarity of occupations exists. The Archaic occupation of the site is primarily associated with the Middle and Early Archaic periods. The dominant settlement themes during this time was short-term residential and special purpose camps occupied for periods of only several weeks at the most. Long-term land-use probably involved serial reoccupations by family groups for several years punctuated by extended periods of abandonment as resource depletion reached a point where the foraging radius could no longer produce sufficient yields. Such a scenario would create an archaeological record that might be characterized by the reoccupation of the same camp locations on the landform by lineally related and/or demographically cycling family units for a period of years until the settlement continuity was broken by longer periods of abandonment. Test unit excavation demonstrated that the vertical separation between Archaic deposits is compressed and not as clear as seen on some uplands sites in the region. However, it also showed that individual occupational events could be analytically recognized and segregated. Debitage and tools could be easily matched by raw material type and the elements of individual camps such as tool clusters have been preserved. Archaic occupations consist primarily of Type I, II and III elements. These elements are characterized by very short-term occupation and they are difficult to distinguish in limited exposures because they share many compositional qualities. Both are characterized by limited debitage weight due to reduction strategies focused on tool maintenance and flake blank production to fashion expedient unifacial tools and un-retouched flake tools, although Type I residences can successfully be differentiated by pro- jecting the total weight of debitage associated with individual debitage scatters and concentrations, as was demonstrated in the TU2 and TU3 investigations. Here clear Type I elements were differentiated that associate with Late Woodland and Bifurcate Tradition phases. Type II residences are characterized by multiple elements of heterogeneous raw material composition, high stone tool frequencies and high tool diversity. Only a few Type II residences have been potentially identified in the region because of the necessity to expose large areas to confirm their existence. The best potential example (Figure 67) of a Type II element is supplied by an extensive closeinterval shovel test block and associated test unit excavation at 31HK1214 (see Cable and Cantley 2005b:204-210). Occupation HK1214.SB6.R2/R3.1 was comprised of low density scatters of Wolf Den Mountain Rhyolite (UWR), Type I Rhyolitic Tuff and Quartz and a wide array of chipped stone tools. The tool inventory consisted of two Palmer CornerNotched points, five flake tools, two Type II biface fragments, and a type II end scraper of UWR and USR (1) material; two flake blanks, two flake tools, one serrated projectile point tip, and a Type II biface of quartz; and a Type III side scraper and two flake tools of Type I Rhyolitic Tuff. Tool clusters were identified at either end of the debitage scatter, which suggests the presence of at least two social units, either nuclear family-like units or specially comprised task groups. Generally, it is assumed that tools will be discarded at relatively low rates, especially curated ones, which would imply that Type II residences involved lengthy rather than short duration stays. Cable (1996), however, has argued that high discard rates for curated tools can result from bulk processing activities such as the butchering and skinning of ungulates from mass kills. This would require aggregations of deer populations in the Southeast, which tend to occur in winter months when snow accumulations drive deer to gather in areas of high canopy cover where browse material is left exposed and optimal bedding areas exist (Gates and Harman 1980). Certainly, the Early Holocene environment of the Carolinas was far from a boreal forest, but the characteristic that may link these disparate cultural systems is an economic focus on highly mobile and dispersed ungulate populations, in this case deer. Opportunities for such aggregations probably waned with time and this may explain why Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 230 Chapter 6. 31HK2510 Flake Tool (R3B) (Fig. 24:J) Flake Blank (Q4) Flake Blank (Q4) (Fig. 14:H) Type III Side Scraper (Fig. 22:H) Flake Tool (R3A) (Fig. 23:F) 592.5 Flake Tool (Q4) Type II Biface Frag (R3D) Type II Biface (R3D) Flake Tool (R3F) (Fig. 24:I) Flake Tool (R3F) Palmer I CornerNotched(R3E) (Fig. 17:C) TU5 Serrated PPK Tip (Q4) Type II Biface Frag. (Q4) Flake Tool (Q4) (Fig. 24:F) Flake Tool (R3B) Type II End Scraper (R3A) (Fig. 21:H-I) Flake Tool (R2A) (Fig. 24:G) 590.0 Palmer II Corner-Notched (R3E) (Fig. 17:D) Morrow Mt. II Stemmed (R7) (Fig. 18F) 587.5 540.0 Flake Tool (R2) 537.5 0 542.5 N 2.5 meters Sample Block 6: 31HK1214 Quartz (Q4) Debitage (cont=1, begins @ 2, max=6) R2 Group Debitage (cont=1, begins @ 2, max=7) R3 Group Debitage (cont=1, brgins @ 2, max=7) R7 Group Debitage (cont=1, max=2) Figure 67. Hypothesized Type II Occupation at 31HK1214 (from Cable and Cantley 2005b). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 231 Chapter 6. 31HK2510 Type II occupations have not yet been recognized for deposits of later age. item was broken of a larger sherd or partial vessel during use. It is estimated that about a tenth of the occupation that occurred at the site was associated with the Woodland period. A contracted history of Woodland occupation was documented, principally ranging between the late Middle Woodland and Late Woodland periods. These occupations appear to have involved short-term seasonal occupation of the landform to exploit floodplain resources. Precontact ceramics were distributed across the entire landform, but a high degree of disorganization and displacement was demonstrated from their analysis. No potential whole pot deposits were identified and 16 individual vessels were recognized in the total inventory of 18 sherds. It was projected that portions of 81 individual vessels have been deposited at the site, while only about 18 percent of the expected weight of these vessels was present. This suggests that the intensity of Woodland use recycling is great on Sandhills sites and that large sherds from broken pots were probably curated and stored in caches or transported upon leaving camps, while smaller sherds were probably left at the point of breakage. If so, sherds from the same pot might be spread across multiple sites and a high degree of displacement and disorganization should be evident in the spatial distribution of ceramics on sites with short-term occupation records. Results in SB6, however, demonstrated that Woodland period Type I elements consisting of highdensity debitage concentrations are present on the site. These residues are sometimes hard to identify, but probably constitute the best avenue from which to understand the character of Woodland occupation on Sandhills sites in the uplands. The organization and structure of camps will be more evident from lithic features because they were formed almost exclusively as primary refuse on a living floor and will not be subject to same displacement and disorganization seen in ceramic distributions. SB14 was established around a shovel test (N490/E460) that yielded a relatively large sherd of Hanover IIa Fabric Impressed weighing 11.93 gm to specifically investigate one of these sherd occurrences on the site. Eight shovel tests were excavated around the sherd at 1.25 m-intervals, none of which yielded additional sherds or even any lithic artifacts. Another shovel test (N410/E525) anchoring SB10 produced two indeterminate Hanover IIa sherds from the same vessel. Again, the eight shovel tests excavated around the anchor failed to yield not only additional sherds from the same vessel, but also no additional sherds were found at all. Three other shovel tests excavated respectively in SB1, SB7 and SB18 produced single sherds, but none of the member shovel tests yielded ceramics at all. Clearly, these sherd finds are not representative of pot breaks, nor do they evidence partial vessel aggregations. In contrast, all of these instances are consistent with the interpretation of limited primary refuse discard events in which the remaining Site 31HK2510 is recommended eligible for inclusion on the National Register of Historic Places under criterion d, “that have yielded, or may be likely to yield, information important in prehistory or history’ in accordance with 36CFR60.4. First, because of the documented low superimposition intensity, the site supplies a unique opportunity to examine in detail the nature of short-term, Archaic occupation in the stream microenvironment with little impact from deposit mixing. The Archaic deposits contain intact Type I, II and III residential elements that were focused on the exploitation of floodplain resources. Only broad exposures will provide a basis for distinguishing between Type II and Type III elements and this work would advance our understanding of the range of variability that may be manifest in these types of occupations. A total of 12.3 kg of arenite, sandstone conglomerate and quartz rock was recovered during the investigation. Most of this material was likely used in the construction of rock hearths across the landform and higher densities of rock may point to locations containing intact cultural features capable of yielding subsistence data and possibly absolute chronometric samples. Generating a large sample of occupations across the landform would also provide the opportunity to consider the types of reoccupation patterns manifest in the cultural deposits. The site may simply be composed of unrelated, random occupations, but we may find serially related reoccupation events that would allow us to consider more elaborate long-term land-use patterns of hunter-gatherer groups in the region. The deposits appear capable of producing intact Woodland camps, subsistence remains and possibly dating information as well. Low Woodland Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 232 Chapter 6. 31HK2510 period occupation intensity should also provide the opportunity to elaborate ceramic discard patterns and to explore ceramic sequencing issues in the Sandhills. The Type I elements composed of Type I RT may provide an opportunity to specifically investigate the settlement patterns of the Bifurcate Tradition and to contrast them with the earlier Type II and Type III occupation themes characteristic of the early end of the Early Archaic sequence at the site. Of interest in this regard is the question of whether Bifurcate occupations in the region represent continuity with the earlier end of the Early Archaic sequence or shortterm intrusions from the Appalachian Mountains (see Broyles 1966; Chapman 1975). This question has never been sufficiently addressed in the North Carolina Piedmont and Coastal Plain regions. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 233 Chapter 6. 31HK2510 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 234 Chapter 7. 31HK2521 Site 31HK2521, the Quewhiffle Site, is located in Hoke County in the western portion of Fort Bragg (Figure 68). The site can be reached by traveling approximately 0.12 miles north-northeast on Quewhiffle Road from its intersection with Moore County Road (Figure 69). At that point, turn east onto Firebreak Road 31 and travel 0.3 miles to a dirt road turn-off on the north side of the road. The site is situated at the end of this dirt road, approximately 0.1 miles from its intersection with Firebreak Road 31 (Figure 70). The SCS soil survey for Hoke County, NC (Hudson 1981:81) classifies the landform as Blaney loamy sand (2 to 8 percent slopes), a well drained soil that typically forms on side slopes and narrow upland ridges. 31HK2521 rests on the top of a ridge toe overlooking Rockfish Creek. An intermittent drainage borders the site on its east flank and flows into the creek to the southeast of the site. In addition, a larger tributary named Calf Branch enters the creek on the bank opposite the site. Finally, the confluence of Rockfish Creek and Piney Bottom Creek is situated approximately 0.8 km (0.5 miles) north of the site. Although it has been subject to tree thinning, the surface of the site is relatively free of modern disturbances. of the ridge is characterized by a slope of only 2.5 percent. An apparent bulldozer cut measuring about 3.5 x 15 m was situated at the far northeast corner of the site where the nose of the ridge descends into the floodplain. Vegetation across the site consisted of a thinned stand of young long-leaf pines and an understory of turkey oak and wiregrass. Maximum site dimensions were established at 100 m north-south x 80 m east-west, covering an area of approximately 1.5 acres (6,100 m2). Elevations within the site boundaries ranged between 72.0 m amsl near the base of the ridge toe to about 74.5 m amsl on the apex of ridge top (Figure 71). The flat ridge top drops rather precipitously into the floodplain of the creek (Figures 72 and 73). Slope from the edge of the ridge top to the floodplain is approximately 16 percent, while the top Grunden and Ruggiero (2006:404) described three soil zones exposed in shovel tests and in a 50 cm x 50 cm test unit excavated near the site datum. The upper zone was composed of dark gray sand and extended to a depth of 10 cm bs. The intermediate zone consisted of a light yellowish brown sand that terminated between 65 and 85 cm bs at the contact PREVIOUS RESEARCH Site 31HK2521 was originally recorded by TRC Garrow Associates, Inc in the spring of 2003 during a Phase I survey of portions of the DD3 Training Area (Grunden and Ruggiero 2006:418-420). The site boundaries established at that time correspond closely to those later defined by PRI (Figure 74). Phase I site definition proceeded through the deployment of 15 m-interval shovel tests. Forty-one shovel tests were excavated in total, of which 15 were positive. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 235 Chapter 7. 31HK2521 656000 657000 656000 657000 3885000 3885000 3886000 655000 3886000 654000 3884000 3883000 3882000 3882000 3883000 3884000 31HK2521 654000 655000 Figure 68. Location of Site 31HK2521 (McCain, NC 7.5’ USGS Quadrangle) Scale: 1:24,000. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 236 Chapter 7. 31HK2521 654800 655200 655400 655600 3885000 3884800 3884800 3885000 3885200 655000 3885200 654600 R oc k fi sh 3884600 ek hif fle Ro ad 3884600 C re 3884400 3884400 Qu ew 31HK2521 31 reak 3884200 3884200 Fireb eak 32 Mo 3883800 or 654600 654800 eC ou nt yR 3884000 3883800 3884000 Firebr oa d 655000 655200 655400 655600 Figure 69. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31HK2521 (Scale: 1 inch = 200m, Projection: UTM, NAD83). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 237 Chapter 7. 31HK2521 Figure 70. View of Site 31HK2521, Looking Northeast from a Point Southwest of the Site Boundary. with olive yellow sandy clay substrate. The recovered artifact assemblage (n=116) consisted of 108 lithic artifacts and 8 precontact sherds. Four ceramic series were identified in the sherd collection, including Cape Fear, Hanover, New River and Yadkin. From this inventory, the surveyors inferred an Early to Middle Woodland occupation span. Most of the lithic assemblage consisted of rhyolite (n=90), quartz (n=15) and metavolcanic (n=1) tertiary flakes. In addition a chert biface and one tertiary flake were recovered from a single shovel test at a depth of 40 to 60 cm bs. Most of the lithic artifacts were recovered between 20 and 85 cm bs, suggesting the presence of a substantial Archaic occupation. It was noted that the site had not been seriously impacted by military operations or other historic activities and the deposits retained a great deal of integrity. It was also considered likely that cultural features were preserved in the deposit matrix. The surveyors recommended the site potentially eligible for inclusion on the National Register of Historic Places for its ability to inform on Woodland period site structure, function and adaptation in the Sandhills. FIELDWORK OVERVIEW Phase II fieldwork was implemented in four stages, each stage building on the information generated from the previous stage. Stage I investigation consisted of a program of 10 m-interval shovel tests that were excavated to establish firm site boundaries. Site boundaries were determined by the documenta- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 238 Chapter 7. 31HK2521 72 Bulldozed Trench .0 540 .0 73.0 75 74.0 520 500 31HK2521 73 .0 480 72.0 74 .0 460 440 420 N 40 m 0 75. 0 400 420 440 460 480 500 520 540 560 Figure 71. Site 31HK2521, Base Map. tion of two consecutive negative shovel tests in all directions on a 10-m grid, or in some cases when wetland was encountered. Once artifact distributions from the Stage I program were mapped and evaluated, the Stage II investigation involved the excavation of closer-interval shovel tests at 5 m-intervals in most of the artifact bearing areas of the site. Stage III investigations involved the deployment of closeinterval shovel tests of 1.25 m around the perimeter of targeted shovel tests yielding deposits deemed of theoretical importance for evaluating the site. These close-interval shovel test locations were identified as sample blocks and generally consisted of eight shovel tests excavated around a targeted Stage I or Stage II shovel test. The sample blocks were also instrumental in guiding the precise placement of 1-x-1-m test units, which constituted the final stage of the investigation. Test units were placed in locations judged to provide optimal opportunities to reconstruct the occupation history of the site and to assess the integrity and scientific importance of the deposits. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 239 Chapter 7. 31HK2521 Figure 72. View of Site 31HK2521, Looking Northeast onto the Apex of Ridge from N480/E460. Stage I investigations were initiated by relocating the datum established by Grunden and Ruggiero (2006) and laying out a 10-m interval grid with a total station. The datum was placed at survey grid location N515/E500, between survey transect shovel tests T81-1 and T81-2. The datum was subsequently reassigned the Phase II grid coordinates of N500/ E500. Relative topographic elevations at each 10 m grid intersection were recorded with the total station and marked with a pin flag. The elevation values (in meters) illustrated on the base map (Figure 71) are approximated real elevations based on the correlation of GPS points with a geo-referenced copy of the USGS, McCain NC 7.5-minute topographic quadrangle map. Cultural and landform features were also mapped and additional elevations were taken when appropriate to produce the topographic map. Shovel tests measured 30-x-30-cm and were square in shape. Excavation proceeded in 10 cm arbitrary levels measured from the surface. Eighty-two Stage I shovel tests were excavated at 10 m-intervals to define the limits of the site (Figure 75). Forty-two of the tests were positive. Maximum site boundaries were established at 100 m x 80 m, covering an area of about 6,100 m2, or approximately 1.5 acres. Mean shovel test depth for the Stage I sample was 59.76 cm bs (SD=21.12) with a mode of 60 cm bs and a range of 15 to 100 cm bs. Excavation was terminated in the upper 5 to 10 cm of the substrate. The SCS soil survey for Hoke County (Hudson 1984:31) maps the area of the site as Blaney loamy sand, which is a well-drained upland soil situated on side slopes and narrow ridges. The on-site Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 240 Chapter 7. 31HK2521 Figure 73. View of Site 31HK2521, Looking North Toward Rockfish Creek Floodplain from the Apex of the Ridge at N520/E500. soil profile corresponds well to the typical Blaney pedon. The A-horizon extends to a depth of 15 cm bs. A charred mixture of prescribed burn charcoal and sand occurs as a 5 cm lens resting on top of the A-horizon. The light coloration of the A-horizon suggests that it may never have been plowed below the impact zone of prescribed burning. The A-horizon is composed of pale brown (10YR 6/3) to light brownish yellow (10YR 6/4) sand. The upper E-horizon (E1) layer, when present, is composed of very pale brown (10YR 7/3) to pale yellow (2.5YR 7/3) sand. The lower E2-horizon sand is slightly lighter, but exhibits the same range of hue. The principal difference between the two layers is increased sand coarseness in the E2-horizon. The layers transition at about 30 to 35 cm bs. The Bt-horizon, which typically begins between 50 and 100 cm bs across the site, consists of brownish yellow (10YR 6/8) to olive yellow (2.5YR 6/6) sandy clay loam of dense and compact structure. The Stage I sample provided a basis for defining a Stage II sample frame within which to deploy a 5 m-interval shovel test grid. This guaranteed that the artifact bearing locations of the site would receive systematic sampling at a relatively intensive sample density, so that a detailed picture of the occupation history of the site could be generated. One hundred thirty-three Stage II shovel tests were excavated at 5 m intervals within the sample frame (Figure 76). Eighty-eight were positive. Mean shovel test depth for the Stage II sample was 63.46 cm bs (SD=18.70) with a mode of 70 cm bs and a range of 20 to 110 cm bs. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 241 - 545 - 530 - 515 - 500 - 485 - 470 - 455 - 440 Chapter 7. 31HK2521 - 545 80-2 - 530 81-1 82-1 - 515 82-J1 81-2 - 500 82-2 - 485 - 470 81-3 - 455 West Wall 2.5Y 4/1 Dark gray Sand Centimeters 0 10 - 440 20 2.5Y 6/4 Light yellowish brown Sand 30 TR 80 TR 81 TR 82 2.5Y 6/6 Olive yellow Sandy clay to Clay S GPS datum/Positive ST Negative transect ST Positive ST Positive transect ST Negative ST Site boundary Surface positive (if any) Contour (approximate) 50 cm x 50 cm test unit Wetland (if any) Shovel test pits shown are not to scale N Scale in Meters 0 15 30 45 Figure 218. Site Plan of 31HK2521. Figure 74. Site 31HK2521, Phase I Survey Sketch Map, TRC (from Grunden and Ruggiero 2006). 419 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 242 Chapter 7. 31HK2521 550 72 Bulldozed Trench .0 540 .0 73.0 75 530 74.0 520 510 500 490 31HK2521 73 .0 480 470 72.0 74 .0 460 450 440 430 Positive ST 420 Negative ST N 410 40 m 0 75. 0 400 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 75. Site 31HK2521, Stage I Shovel Test Sample. The Stage III shovel test sample was distributed in 18 locations anchored by Stage I or II shovel tests containing data relevant to evaluating the occupational history and integrity of the site matrix (Figure 77). One hundred forty-three Stage III shovel tests were excavated at typically 1.25 m-intervals around targeted Stage I and II shovel tests. One hundred thirty-three of these tests yielded artifacts. Mean shovel test depth for the Stage III sample was 69.28 cm bs (SD=19.56) with a mode of 90 cm bs and a range of 40 to 100 cm bs. Focus on the top of the ridge during the Stage II and III excavations resulted in slightly deeper maximum mean depths over the Stage I sample. Once the results of the Stage III shovel test sample were analyzed, specific locations yielding shovel test outcomes of interest to further evaluating Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 243 Chapter 7. 31HK2521 550 72 Bulldozed Trench .0 540 .0 73.0 75 530 74.0 520 510 500 490 31HK2521 73 .0 480 470 72.0 74 .0 460 450 440 430 Positive ST Negative ST 420 N 410 40 m 0 75. 0 400 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 76. Site 31HK2521, Stage I and II Shovel Test Sample. the site were selected for test unit excavation. Stage III shovel testing around targeted shovel test outcomes provided the basis for precise placement of test units. Eleven 1 m-x-1-m test units were excavated in total and their locations are illustrated in Figure 78. Blocks of four 1 x 1 m test units were established at two locations to follow out sherd aggregations. ARTIFACT INVENTORY Chipped stone debitage and tools dominated the artifact inventory from 31HK2521. The inventory consisted of 3,972 items including 2,319 pieces of lithic debitage, 72 chipped stone tools, 26 cores, 1 worked schist fragment, 5 hammer stones, 1 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 244 Chapter 7. 31HK2521 550 72 Bulldozed Trench .0 540 .0 73.0 75 530 74.0 2 520 15 17 11 16 4 510 3 14 500 8 9 10 7 18 490 12 13 31HK2521 480 6 73 .0 5 470 74 460 72.0 1 .0 450 440 430 Positive ST Negative ST 420 N 410 40 m 0 75. 0 400 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 77. Site 31HK2521, Stage III Shovel Tests and Sample Block Locations. hammer stone/anvil, 3 pebbles/tools, 1 nutting stone, 1 abrader/grinding stone, 1 soapstone vessel fragment, 615 rocks/fire-cracked rocks, 2 pieces of float, 917 precontact ceramics and 8 perishable items (i.e. 7 faunal bone fragments and 1 piece of wood charcoal). Appendix B contains the artifact database for the project, while appendices C through I provide additional data on lithic chipped stone and ground stone tools and precontact ceramics. Descriptions of the various artifact and lithic raw material types are presented below. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 245 Chapter 7. 31HK2521 550 72 Bulldozed Trench .0 540 .0 73.0 75 530 74.0 Block 3 520 2 7 Block 2 5 8 10 6 510 3 Block 1 500 9 4 1 11 490 31HK2521 73 .0 480 470 72.0 74 .0 460 450 440 430 420 N 410 40 m 0 75. 0 400 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 78. Site 31HK2521, Test Unit Locations. Lithic Raw Material Types Lithic raw material identification proceeded on two different levels. At the macro-scale, individual specimens were typed in accordance with major hard rock geological terminology to achieve comparability with previous projects at Fort Bragg. At the micro-level, however, an attempt was made to further partition some of the hard rock types into “core groupings” to distinguish between specimens derived from single hard rock types that may have originated from different cores. This was done to facilitate spatial analysis in close-interval shovel test grids where overlapping of distinct deposits or complex multi-household occupations were suspected. Core grouping analysis was undertaken principally in metavolcanic types where subtle differences in patina coloration, core color, phenocryst size and density, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 246 Chapter 7. 31HK2521 and matrix texture was easily discerned and monitored. Descriptions of the various rock types and the associated core grouping sub-types recognized in the collection are presented below. Metavolcanic Types Metavolcanic types, as a group, comprise the majority of the chipped stone collection from the site (Table 62). As a group, they represent 57.0 percent (n=1378) of the total chipped stone artifact inventory (i.e. debitage and chipped stone cores and tools). The best reflection of actual representation, however, is supplied by the combined Stage I and II shovel test samples, which is unbiased in terms of spatial coverage. In this sample, metavolcanic material makes up 70.6 percent of the chipped stone, an indication that the Stage III and IV investigations favored contexts that contained quartz debitage. Descriptions of the five identified metavolcanic types and associated core group subtypes follow. The typological system varies somewhat from earlier reports in response to the recent stone quarry sourcing study for the Carolina Slate Belt conducted by Steponaitis et al. (2006). (1) Mill Mountain Rhyolite (MMR). MMR was not clearly differentiated in the sourcing study (Steponaitis et al. 2006), but it is a readily recognizable type in the assemblages from Fort Bragg. Mill Mountain is located between the Uwharries Southern and Uwharries Eastern source areas on the opposite bank of the Yadkin River from Table Top Mountain. It was not included in either of these source areas, however, because of the presence of distinctive clear, glassy quartz phenocrysts. Daniel and Butler (1996:18) describe this type as a medium gray, aphanitic, rhyolite porphyry with sparse, glassy phenocrysts of quartz generally less than 1 mm in diameter. Very fine-grained, disseminated grains of pyrite are also present. Thin sections indicate a microcrystalline matrix composed primarily of feldspar and quartz, with some biotite, chlorite, and disseminated pyrite. The specimens assignable to this type from the 31HK2521 collection are characterized by a dark reddish gray to moderate brown coloration when not patinated. Four subtypes were assigned to the type, based primarily on the completeness of the observation field and patination. Subtype R8 corresponds to specimens that exhibit sparse glassy quartz phenocrysts only. Subtypes R8Hhf, R8Hhm and R8Hhc are fine-, medium- and coarse-grained specimens respectively that exhibit sparse glassy quartz, white subhedral quartz phenocrysts and, rarely, feldpsar phenocrysts. Although these phenocryst types were not included in the type description, the subtypes containing them were included as a variant of Mill Mountain Rhyolite due to the co-occurrence with glassy quartz phenocrysts. Mill Mountain Rhyolite makes up 4.8 percent of the combined Stage I and II chipped stone sample. (2) Type I Rhyolite Tuff (Type I RT). An exact correlate of this type was not discernible in the sourcing study (Steponaitis et al. 2006). Benson (1999:30) describes it as fine-grained dark green to gray material that looks and feels very similar to Piedmont chert. He noted that plagioclase phenocrysts were sporadically present. Five subtypes were identified in the 31HK2521 collection. Subtype R2 is only rarely identified in assemblages on Fort Bragg. It appears very similar to Subtype R3t, but its coloration can be described as a dusky blue green. The texture is microcrystalline and the edges are semi-translucent. Subtype R2A specimens exhibit a pale green patina, but broken edges reveal that the underlying material corresponds to R2. Darker green splotches and swirls on the faces of these specimens represent more resistant matrix. The matrix is aphantitic and microcrystalline in texture and is highly isotropic in fracture characteristics. Also recognized were subtypes R2Af, R2D and R2Id. Subtype R2Af includes patinated specimens that exhibit flow bands or laminations. Subtype R2D is identical to R2A, but also contains rectangular, euhedral biotite mica crystals. Finally, Subtype R2Id represents an extremely fine or microcrystalline material that contains sparse, small feldspar phenocrysts and needle-like aggregates of metamorphic stilpnomelane. It weathers to a dusky to moderate yellow color. The origin of Type I Rhyolite Tuff is not well understood and it is possible that it represents a very fine-grained metamudstone similar to some of those described for the Chatham Pittsboro source (see Moore and Irwin 2006:27-28; Stoddard 2006:57). In support of this is the occurrence of what appear to be bedding laminations in some of the patinated material. Moreover, the color of the patinated specimens Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 247 Chapter 7. 31HK2521 Table 62. Lithic Raw Material Frequencies by Chipped Stone Tool Class, 31HK2521. LITHIC RAW MATERIAL FREQUENCIES BY CHIPPED STONE TOOL CLASS: 31HK2521 20 Mill Mountain Rhyolite 61 1 1 Type I Rhyolite Tuff 149 1 2 1 Uwharries Eastern Rhyolite 187 4 2 Uwharries Southern Rhyolite (1) 159 2 1 UNIFACES CORES Hydrated Metavolcanic UTILIZED FLAKES BIFACES DEBITAGE RAW MATERIAL TYPES PROJECTILE POINTS CHIPPED STONE TOOL CLASSES GRAND TOTAL Metavolcanic 20 1 64 2 3 158 1 1 2 197 1 2 7 172 Uwharries Southern Rhyolite (2) 588 3 4 6 8 6 615 Uwharries Western Rhyolite 144 2 1 1 2 2 152 1 15 7 1 Quartz Crystal Quartz 27 White Quartz 972 1 28 3 999 Rose Quartz 1 1 Metasedimentary, Green 6 6 Metasedimentary, Tan 1 1 Other Material Schist GRAND TOTAL 4 2319 4 14 26 17 16 25 2417 is similar to that reported for the Chatham Pittsboro source. Type I Rhyolite Tuff makes up 11.2 percent of the combined Stage I and II chipped stone sample. (3) Uwharries Eastern Rhyolite (UER). Stoddard (2006:52) describes the specimens from this sourcing area as light to dark gray metadacite porphyry or crystal lithic tuff. All samples collected contained plagioclase feldspar and white quartz phenocrysts, ranging from less than 2 percent to 7 percent phenocryst density. Common metamorphic minerals within the matrices include biotite, stilpnomelane and calcite. Similar dacites and rhyolites are typical of the Uhwarries Southeastern source (Stoddard 2006:55-56) as well, but many of these specimens contain silica levels beyond the range typical of igneous rocks. Macroscopic criteria for distinguishing between lithic material from these two sources does not at present exist. Consequently, specimens from both source areas may be present in the analyzed lithic assemblages at Fort Bragg. Typical specimens from 31HK2521 are characterized by dark gray (nonpatinated) and grayish green (patinated) matrices with splotches of silicate and variable percentages (2 to 4 percent) of white quartz and plagioclase feldspar phenocrysts. Some specimens exhibit purplish to brown background matrices. Quartz phenocrysts predominate over feldspar in the examples from the site. Type Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 248 Chapter 7. 31HK2521 I Rhyolite, which has been distinguished separately in previous reports (Cable and Cantley 2005b, 2006, 2010), is also subsumed under this type. Three subtypes were identified in the collection: R8Ihf, R8Ihm, R8Ihc, all of which contained white quartz and plagioclase feldspar phenocrysts in variable percentages. They are distinguished from each other by differences in grain. Respectively, the subtypes correspond to fine-, medium- and coarsegrained specimens. UER makes up 9.4 percent of the combined Stage I and II chipped stone sample. (4) Uwharries Southern Rhyolite (USR). This type includes sources found on Morrow Mountain and surrounding areas, most notably Table Top Mountain. Daniel and Butler (1996:10-13) describe this type as a dark gray, aphanitic, aphyric rhyolite, that commonly exhibits flow-banding. Weathering tends to bring out the flow lines visually, which generally alternate in a pattern of light and dark gray. Although the matrix is usually homogeneous, some specimens exhibit small spherulites of less than 1 mm in diameter (see Stoddard 2006:52). Thin sections indicate a microcrystalline intergrowth of feldspar and quartz, with minor biotite and chlorite elements. The individual minerals are difficult to distinguish and strings of dark minerals can occur along fracture planes. The specimens identified as USR in the 31HK2521 assemblage are composed of two larger groupings, the 3m and the R3/R3t groups. The 3m Group corresponds to material that typically occurs on Morrow Mountain and includes a series of subtypes that describe variation in flow banding, texture and patination. Subtype R3m is non-patinated and it exhibits a dark gray fine-grained matrix devoid of phenocrysts, while R3mf corresponds to flow banded specimens of R3m material. Subtypes R3mD, R3mDf and R3mDs are patinated and range in color from grayish green to greenish gray to moderate yellow or tan. R3mD and R3mDf represent patinated specimens of R3m and R3mf respectively. R3mDs exhibits blotches and linear arrays of silicate containing biotite and perhaps chlorite elements, respectively. They too, appear to be patinated forms of R3m and R3mf. The R3/R3t group is composed of an extremely fine grained or microcrystalline dark gray matrix that is semi-translucent when viewed on attenuated edges. Finely disseminated feldspar is commonly seen on exposed surfaces, but the small blotches do not form phenocrysts. R3 and R3t are non-patinated, while R3p, R3pat and R3patf are fully patinated subtypes. All three are tan in color, but R3p also exhibits finely disseminated biotite specks and R3patf exhibits flow banding. Finally, R3tf constitutes a flow-banded version of R3t. The major distinctions between R3 and R3t specimens are differences in grain and translucency. The former is slightly coarser and is not translucent. USR makes up 36.1 percent of the Stage I and II chipped stone sample. The R3/R3t group, or USR (1), makes up 8.6 percent of the Stage I and II sample, while the R3m group, or USR (2), comprises 27.5 percent. (5) Uwharries Western Rhyolite (UWR). This source area includes specimens from Wolf Den Mountain and Falls Dam (Stoddard 2006:52). This material is described as gray to black microcrystalline felsic volcanic rock that contains rounded plagioclase phenocrysts and glomerocrysts. Green biotite and pale green amphibole are dispersed locally in the matrix and spherulites are sometimes present. Three subtypes were identified in the collection. The main subtype, R3mp, has been patinated to a grayish green or greenish gray color. Non-patinated specimens may have been assigned to one of the other Uwharrie source areas in the analysis, but the distinctive glomerocyst structure suggests that this was not a major problem. Due to patination, most of the specimens in the collection were characterized by fine to medium texture. A single Subtype was identified, R3mp. UWR makes up 7.9 percent of the combined Stage I and II chipped stone sample from 31HK2521. (6) Indeterminate Metavolcanic. A small amount of the metavolcanic material identified in the analysis could not be further classified to type due to excessive hydration. These were simply classified as “indeterminate metavolcanic.” Only eight pieces of debitage were placed in this category in the Stage I and II sample. Quartz Types Pure quartz forms as veins in igneous and metamorphic rock formations characterized by slow crystallization. This generally produces anhedral, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 249 Chapter 7. 31HK2521 milky structures (Spock 1962). In some cases, however, crystallization occurs earlier in the sequence, producing clear crystals with euhedral structure. Both vein and euhedral crystal quartz were utilized by precontact groups in the project area, but most of it appears to have come from river gravel deposits. A great deal of variation was evident in the vein quartz category ranging from mixed clear and cloudy exposures to opaque milky white. Since these characteristics appeared to be repeatedly present on single cores when debitage concentrations were analyzed, all of this variation was subsumed under the single white quartz subtype Q5. The category “crystal quartz” (CQ) was retained for instances where glasslike examples and crystal facets strongly suggested a euhedral crystal origin. “Rose quartz” was reserved for quartz specimens with a high degree of rubification, a commonly observed attribute in quartz cobbles. White quartz represented 28.8 percent of the combined Stage I and II chipped stone sample, while only one piece of crystal quartz (0.1 percent) was recovered in the sample. Rose quartz was not represented in the Stage I and II sample. In addition to debitage, quartz cobbles and fragments of various types were recovered from the investigation. These included 25 pieces of fire-cracked rock, 2 whole cobbles, 3 cobble fragments and 2 rock fragments of white quartz and a whole rose quartz cobble without detectable usewear. Finally, eight pebble and cobble tools composed of quartz were identified. These included two white quartz and one rose quartz pebbles/tools, one white quartz hammer stone, two rose quartz and one yellow quartz hammer stones and an abrader/grinding stone of rose quartz. All of these items were characterized by extensive fracture planes. and irregular nodules. The exterior of the nodules consist of fine-grained ruddy brown precipitate while the centers consist of dense, reddish brown to purple, densley cemented sandstone. The material was commonly used for rock hearths by the precontact inhabitants of Fort Bragg. Four hundred and thirty-six pieces of arenite were collected during the investigation, totaling 10.73 kg. (2) Chert-cemented Sandstone. A single item, a nutting stone fragment (a1133), was made from a tan sandstone conglomerate composed of coarse, well sorted sand cemented together with chert precipitate. The material is hard and resistant. (3) Grano-diorite. A hammer stone/anvil (a2286) and a large hammer stone (a825) were made from flattened, oblong river cobbles composed of a light colored, fine-grained, dense igneous rock. The material is tentatively identified as grano-diorite. Grano-diorite is gradational between granite and diorite and is formed when plagioclase replaces orthoclase in sufficient proportions during magma cooling (Spock 1962:58). Several other rock types were recognized in the 31HK2521 collection. Descriptions of each are presented below. (4) Metasedimentary. Seven pieces of debitage and two rock fragments of fine to medium grained, homogeneous metasedimentary of green and tan color were recovered during the investigation. Although the material occurs only infrequently, it was apparently used to manufacture a wide range of tool types in the region. At nearby 31HK2502, a Morrow Mountain II Stemmed base (271265a344), a Type III Side Scraper (271265a1166) and a fragment of a grooved abrader (271265a325) were all made from this material. The relative coarseness of the grain suggests that the material can be further classified as metasiltstone or fine metasandstone. Numerous source areas in the Durham-Wadesboro Basin contain similar rock types, including the Chatham Pittsboro, Person County and Chatham Siler City (Stoddard 2006:57-63). (1) Arenite. This material has been previously identified as ferruginous sandstone, but thinsectioning of a piece of this material indicates that a more accurate characterization is arenite, or hematite cemented quartz sandstone (Appendix K). Three pieces of arenite were classified as fire-cracked rock. It occurs in a variety of shapes including flattened (5) Petrified Wood. Two pieces of talc-like rock were recovered that appeared to exhibit woodgrain contouring, suggesting that they might represent petrified wood. Petrified wood has been reported from Uwharrie National Forest in nearby Montgomery County. Petrified wood might be expected to occur in locations across the Triassic Basin, in the Minority Types Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 250 Chapter 7. 31HK2521 same formations that produce the abundant sandstone conglomerates and arenite that are widespread throughout the Sandhills (Rogers 2006). (6) Sandstone Conglomerate. Also abundantly present in the deposit were chunks of sandstone conglomerate containing a wide range of sand and pebble sizes. Thin-sectioning of a piece of this material determined that individual grains of sand were cemented with clays (wacke) and exhibited udulatory extinction indicating that it originated from metamorphic rock (Appendix K). Occasionally interfaces with arenite could be observed on the chunks, indicating that the two materials were derived from the same formations (e.g. Middendorf). The matrix of the conglomerate consistently exhibited a grayish orange to dusky yellow coloration. Chunks of arenite were sometimes viewed in the matrix as well. This material primarily appears to have been used to make rock hearths. One hundred forty-two pieces of sandstone conglomerate were collected, summing to about 4.08 kg. (7) Schist. Four pieces of chunky debitage flake fragments and a worked tool fragment (a1150) were made of a soft, laminar, highly micaceous material identified as schist. The material was yellowish brown to gray in color with dark gray and reddish brown streaks. It did not have a “soapy” feel, although it was clearly of hardened talc-like consistency. This might suggest that material may be more properly identified as a pyrophyllite. (8) Steatite. A single sherd (a492) from a steatite bowl was recovered in a shovel test during the investigation. The material was light gray in color. Steatite vessel fragments are sporadically found on Fort Bragg (see McNutt and Gray 2007; Millis et al. 2010; Ruggiero 2003). The material is commonly found at locations in the Piedmont, although the nearest known quarries are in South Carolina and south central Virginia (see Millis et al. 2010:460-461; Truncer 2004). (9) Float. Two small water worn nodules of lithic material were recovered that are worth mentioning because they were either brought to the site by precontact groups or were transported and deposited during flood events. One piece was a white quartz pebble, while the other consisted of a hydrated metavolcanic rock. If the metavolcanic “float” was brought to the site by natural forces or as unintentional inclusions within rock hearth bundles, it would suggest that Uwharries metavolcanic material could be present in stream terrace formations in the area or in exposures of Triassic Basin sediments crosscut by streams. Rogers (2006) discusses the distribution of Carolina Slate Belt elements and the Triassic Basin in locations near Fort Bragg. Lithic Artifacts Basic definitions of the various lithic artifact classes were presented in Chapter 4. Here, specific descriptions of the 31HK2521 lithic collection are presented. Debitage The debitage classification is adapted from formal (Bradley 1973; Frison and Bradley 1980; House and Wogaman 1978; Newcomer 1971), attribute (Moore 2002; Shott 1994), and mass (Ahler 1989; Moore 2002) analysis approaches. Debitage classes were devised to identify both reduction stages and reduction/production systems (ie. biface core reduction, directional core reduction, and flake blank production). This classification was supplemented with the recording of three attributes: (1) size class, (2) percent cortex, and (3) condition. Table 63 presents the complete inventory (all field stages) of debitage type by raw material type from 31HK2521. Biface reduction types (FBRs and BTFs) comprise the majority of identifiable debitage, but directional core reduction as represented by core flakes, blade flakes and core rejuvination flakes is significantly high. Directional core debitage comprises 26 to 58 percent of the identifiable debitage types (i.e. blade flakes, core flakes, core rejuvination flakes, FBRs and BTFs) within the metavolcanic and white quartz samples. The distribution of cortex on dorsal faces and platforms of debitage provides a basis for further evaluating the nature of reduction strategies associated with the various raw material types (Table 64). Dibble et al. (2005:550) indicate that raw nodule Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 251 Chapter 7. 31HK2521 Table 63. Distribution of Debitage Type by Raw Material Type, 31HK2521. DISTRIBUTION OF DEBITAGE TYPE BY RAW MATERIAL TYPE: 31HK2521 CORTICAL CHUNK FLAKE FRAGMENT FBR CORE REJUVINATION FLAKE CORE FLAKE CHUNK BTF RAW MATERIAL TYPES BLADE FLAKE DEBITAGE TYPES 4 8 2 20 24 12 61 57 29 149 71 35 187 64 23 159 GRAND TOTAL METAVOLCANIC Hydrated Metavolcanic 3 3 Mill Mountain Rhyolite 4 20 1 23 1 Type I Rhyolite Tuff 38 Uwharries Eastern Rhyolite 30 50 Uwharries Southern Rhyolite (1) 46 26 Uwharries Southern Rhyolite (2) 122 Uwharries Western Rhyolite 1 1 1 96 1 257 111 588 13 34 1 73 23 144 1 3 QUARTZ Crystal Quartz 23 Rose Quartz White Quartz 27 1 134 1 210 3 593 31 972 4 1 6 OTHER TYPES Metasedimentary, Green 1 Metasedimentary, Tan 1 1 Schist 4 4 Grand Total 391 2 reduction from bifaces should yield between about 6 and 21 percent flakes with cortex for any single episode. It is likely that similar proportions characterize directional core reduction from raw nodules. MMR, and UER and USR (1) exhibit percentages (8.2 to 16.4 percent) within the range of raw nodule reduction. All of these types have correspondingly high core flake percentages, indicating that the predominant strategies in these cases involved raw nodule and prepared block core reduction to produce flake blanks. Interestingly, all of the remaining lithic raw material types also exhibit high directional core debitage 1 466 7 5 1175 267 2319 percentage, which suggests that prepared block core reduction was an important component in all of these provisioning systems. Bifacial reduction appears to have been primarily limited to shaping flake blanks produced from directional cores. Cores Metric and attribute data on cores are provided in Appendix C. Three classes of cores were identified in the 31HK2510 collection: (1) Direc- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 252 Chapter 7. 31HK2521 Table 64. Percent Cortex on Debitage and Percent Core Flakes by Raw Material Type, 31HK2521. PERCENT CORTEX AND CORE FLAKES BY RAW MATERIAL: 31HK2521 RAW MATERIAL TYPE TOTAL DEBITAGE Mill Mountain Rhyolite PERCENT W/CORTEX PERCENT CORE FLAKES 61 8.2 56.8 Type I Rhyolite Tuff 149 4.0 27.2 Uwharries Eastern Rhyolite 187 10.2 43.5 Uwharries Southern Rhyolite (1) 159 16.4 27.4 Uwharries Southern Rhyolite (2) 588 3.6 29.6 Uwharries Western Rhyolite 144 2.8 49.3 White Quartz 968 1.5 56.3 TOTAL 2256 tional Cores, (2) Core Fragments, (3) Natural Cores and (4) Flake Blanks. Each class is discussed below (Table 65). (1) Directional Cores. Directional, or block, cores consist of relatively large masses of material that exhibit flake scars running in one or more directions from flat striking platforms rather than from biface edges. The platform angle approximates 90°, but specimens nearing exhaustion sometimes exhibit more acute angles. Directional cores can be further divided into those with scars emanating from only one platform (unidirectional) and cores with multiple platforms (multidirectional). Generally, multidirectional cores represent the final stage of core use as new platforms are formed to extend their lives. In addition, directional cores are sometimes split in two in the later life stages to create thick core flake blanks, either by intention or because of a flake production error. These are called “split cores.” Three directional cores were recognized in the collection, all of which were made of quartz. These consisted of a unidirectional core (Figure 79:A), an exhausted directional core (Figure 79:B) and a fragmentary directional core (Figure 79:C). All of these cores appear to be “spent,” as it would be difficult to remove additional flakes that would be serviceable as tool blanks. Core thickness, the distance from the top of the platform to the base of the core, would define the maximum available length for the manufacture of flake blanks. On these cores, thickness ranged between 27 to 33 mm, while the flake blanks in the collection from 31HK2521 range between 32 and 43 mm in length. Specimen a989 (Figure 79:A) was fire-cracked and recovered in two pieces, indicating that it had probably been recycled into hearthstone in its final discard state. (2) Core Fragments. Angular pieces of raw material containing portions of one or more flake scars suggestive of directional core reduction, but which lacked striking platforms, were classified as core fragments. Ten such fragments were recognized in the collection, nine of which were made of white quartz (Figure 79:D). One specimen was manufactured from crystal quartz. All of these appear to represent fragments from directional or block core fragments. (3) Natural Cores. Natural cores are unmodified nodules or masses of lithic raw material that were presumably transported to a site for future use as a core. Two such specimens were recovered during the investigation. Specimen a495 was a fragment from a small, tabular core of USR (1) material that could have originated at a quarry outcrop (Figure 79:E). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 253 Chapter 7. 31HK2521 TABLE 65. Summary Data for Cores, 31HK2521. CORES: 31HK2521 SPECIMEN # BAG CORE TYPE PORTION FLAKE TYPE LITHIC RAW MATERIAL TYPE MAXIMUM THICKNESS (mm) MAXIMUM LENGTH (mm) MAXIMUM WIDTH (mm) WEIGHT (gm) 271267a2082 820 Directional Core Fragment White Quartz 27.87 31.87 21.04 20.26 271267a1098 487 Directional Core Whole White Quartz 30.33 30.20 28.95 32.56 271267a989 447 Unidirectional Core Whole White Quartz 33.04 58.85 30.62 68.57 271267a139 69 Core Fragment Fragment White Quartz 11.78 25.75 24.49 9.6 271267a593 297 Core Fragment Fragment White Quartz 16.15 31.62 12.94 8.36 271267a673 331 Core Fragment Fragment Crystal Quartz 11.46 27.17 15.12 4.59 271267a745 363 Core Fragment Fragment White Quartz 11.55 13.70 11.39 2.6 271267a969 440 Core Fragment Fragment White Quartz 18.14 24.84 10.71 5.01 271267a1171 521 Core Fragment Fragment White Quartz 18.57 22.80 16.96 6.18 271267a1551 674 Core Fragment Fragment White Quartz 26.71 55.28 25.97 19.15 271267a1920 781 Core Fragment Fragment White Quartz 20.19 17.42 12.46 4.61 271267a2115 825 Core Fragment Fragment White Quartz 21.85 34.62 21.18 24.53 271267a2528 909 Core Fragment Fragment White Quartz 13.07 30.69 18.93 7.03 271267a495 237 Natural Core Fragment Uwharries Southern Rhyolite (1) 14.21 24.86 27.15 9.85 271267a2521 907 Natural Core Whole Uwharries Eastern Rhyolite 25.59 89.81 63.45 206.5 271267a387 196 Flake Blank Fragment Core Flake Uwharries Southern Rhyolite (2) 16.56 41.02 47.30 22.25 271267a394 199 Flake Blank Whole Core Flake Uwharries Southern Rhyolite 10.33 32.13 34.87 15.13 271267a550 270 Flake Blank Whole Core Flake White Quartz 6.99 42.75 22.26 8.85 271267a916 422 Flake Blank Whole Core Flake Uwharries Western Rhyolite 6.67 33.07 39.10 11.78 271267a1173 521 Flake Blank Whole FBR Mill Mountain Rhyolite 5.79 40.53 36.90 8.8 271267a1212 537 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (2) 3.80 37.64 24.89 3.72 271267a1415 624 Flake Blank Whole Core Flake White Quartz 9.11 29.15 30.83 9.39 271267a1447 637 Flake Blank Whole Core Flake Uwharries Eastern Rhyolite 5.62 35.06 25.02 5.37 271267a1501 653 Flake Blank Whole Blade Flake Type I Rhyolite Tuff 4.04 26.25 12.49 0.85 271267a1797 760 Flake Blank Fragment Blade Flake Uwharries Southern Rhyolite (2) 2.78 39.77 16.24 1.94 271267a2268 846 Prismatic Blade Fragment Type I Rhyolite Tuff 3.07 (26.86) 15.48 1.47 Core Flake The other natural core (a2521) was composed of UER material (Figure 79:F). Its rounded and flat aspect appears to evidence detritus from talus slope wash. The specimen originated either at a Piedmont quarry or in Triassic Basin deposits. (4) Flake Blanks. Flake blanks were removed from a core for the purpose of further modification or use as a tool (Bradley 1973:6). However, they do not exhibit obvious macroscopic evidence of use or modification. Eleven specimens fit the physical requirements of this definition. One of these was derived from a large FBR (Figure 79:H), nine represent core flakes/blade flakes (Figure 79:G, I) and one represents the proximal end of a prismatic blade (Figure 79:J). Prismatic blades are specialized flakes removed from polyhedral blade cores (Crabtree 1972:86). They are generally triangulate to trapezoidal in cross-section and the dorsal surface exhibits two or three prism-like facets or scars from earlier blade removal from the core. Although prismatic blades manufactured from polyhrdral obsidian cores are a common occurrence on Mesoamerican sites, prismatic blades found in the Southeast United States are more likely associated with Paleoindian occupation (Collins 2005). The specimen (a2268) from 31HK2521 is a classic prismatic blade, it is trapezoidal in section, displays three facets on its dorsal surface and the platform angle is nearly 90°. The margins did not show evidence of use wear, suggesting Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 254 Chapter 7. 31HK2521 Figure 79. Cores, Site 31HK2521. A: Unidirectional Core (a989), B-C: Directional Core Fragments (a1098 and a2082 respectively), D-E: Core Fragments (a2115 and a495 respectively), F: Natural Core (a2521), G-I: Flake Blanks (a387, a1173 and a916 respectively), J: Prismatic Blade (a2268). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 255 Chapter 7. 31HK2521 that it may have been broken during manufacture. Although its association with Paleoindian occupation is likely, it was found at a depth of only 20 to 30 cm bd in TU1. Bifaces Fourteen biface fragments were recovered during the investigation (Table 66). Bifaces are derived from both large mass packages (i.e. unmodified and/or prepared cores) and flake blanks. Determination that a biface was derived from a flake blank was made when flake characteristics were observed on the item, such as remnant striking platforms, flake curvature, or bulbs of percussion from flake detachment. The bifaces in the collection do not have sufficient mass to suggest that they were derived from large biface cores. Instead, they all appear to have been produced from flake blanks of varying sizes. Each biface category as adapted from Daniel (2002:51-54) is described below. (1) Type I Bifaces. Daniel (2002:51) describes Type I Bifaces as irregularly shaped flake blanks and core masses with large conchoidal scars forming at least one section of a sinuous edge along the core edge. No evidence of secondary thinning scars is observable and a relatively high thickness to width ratio obtains. Three Type I Biface fragments were recognized in the collection. One specimen (a2165) made of UER was derived from a large flake blank, or what some might refer to as a quarry flake (Figure 80:A). The other two specimens (Figure 80:B-C) were made on smaller blanks, the conchoidal flake features of which are still partly exposed. These specimens (a517 and a908) were made respectively of UER and MMR. All of the Type I Bifaces were discarded upon breakage during manufacture, as reflected by step and hinge fracture terminations across the blade faces. (2) Type II Bifaces. These bifaces, in the main, represent intermediate stage preform rejects. They are roughly flaked and exhibit only large, conchoidal flake scars. No edge trimming is evident, but an ovate to lanceolate outline is generally recognizable and the perimeter of the edge margin is entirely formed. Facial retouch, however, may not be complete and features of the original blank are sometimes evident. Five Type II Biface fragments were identified in the collection (Figure 80:D). The illustrated specimen (a82) was manufactured from UER and was made on a large “quarry” flake. The remaining specimens primarily represent lateral section fragments with step terminations. (3) Type III Bifaces. Type III Bifaces represent late stage preform rejects. The edges exhibit secondary thinning and shaping scars and facial retouch is globally distributed. Outline shapes are well defined and reflect ovate and lanceolate forms. Most of these specimens exhibit step and hinge fractures that terminated production. Type III Bifaces tend to exhibit the lowest ratios of thickness-to-width. Sometimes features of diagnostic projectile point haft elements are present, allowing correlation with specific culture-historic types. In these cases, specimens are differentiated and identified as “Preforms.” Five Type III bifaces and biface fragments were identified in the collection. The fragments consisted of tips and lateral sections that exhibited well thinned bifacial margins that had they been more complete might have qualified as projectile point fragments (Figure 80:E). Two whole, ovate Type III Bifaces were recovered in the collection (Figure 80:F-G). Both exhibit convex basal outlines, but otherwise share the morphological characteristics of what Coe (1964:6465) described as Hardaway Blades. This type was recovered in the lowest levels of the Hardaway Site in primary association with the Hardaway Dalton point style. Coe’s illustrated examples range from deeply concave bases with flaring ears similar in form to the Quad point to less elaborate ovate to lanceolate forms with slightly concave to straight bases. Similar Quad-like, ovate bifaces with flared ears were recovered in the lower levels at the Haw River Site as well (Claggett and Cable 1982). Moreover, excavations at 31HT690 on Fort Bragg have produced a number of well made ovate and squat lanceolate Type III Bifaces at the base of the Early Archaic block (Cable 2008:147-155). These specimens exhibited both concave and convex bases, suggesting that basal concavities were formed during the final stages of manufacture, which may have been aided by basal fluting as a thinning strategy. Specimen a238 (Figure 80:F) from 31HK2521 exhibited a basal thinning flute on one face. Both specimens were found at depths consistent with an Early Archaic or Paleoindian association. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 256 Chapter 7. 31HK2521 TABLE 66. Summary Data for Bifaces, 31HK2521. BIFACES: 31HK2521 SPECIMEN # BAG TYPE PORTION SHAPE ASSOCIATION LITHIC RAW MATERIAL TYPE Uwharries Eastern Rhyolite Maximum Thickness (mm) Maximum Width (mm) Weight (gm) (43.88) (46.14) 35.43 271267a517 245 Type I Biface Fragment End Missing 271267a908 419 Type I Biface Fragment End Mill Mountain Rhyolite 10.78 (22.54) (34.00) 7.49 271267a2165 830 Type I Biface Fragment End Missing Ovoid Uwharries Eastern Rhyolite 25.07 (73.36) (38.65) 62.69 271267a82 38 Type II Biface Fragment End Ovate Uwharries Eastern Rhyolite 13.28 (45.30) (47.05) 25.05 271267a965 438 Type II Biface Fragment Lateral Section Uwharries Eastern Rhyolite 7.54 (28.08) 11.34 1.8 271267a1419 627 Type II Biface Fragment Lateral Section Uwharries Southern Rhyolite (2) 9.13 (43.24) (19.38) 4.96 271267a1572 682 Type II Biface Fragment Lateral Section White Quartz 7.77 (21.27) (28.58) 3.01 271267a2130 826 Type II Biface Fragment Lateral Section Uwharries Western Rhyolite 7.18 (27.95) (12.50) 2.21 271267a179 92 Type III Biface Fragment Upper Blade Tip Uwharries Southern Rhyolite (2) 4.95 (17.60) (18.69) 1.83 271267a209 106 Type III Biface Fragment Upper Blade Tip Type I Rhyolite Tuff 3.52 (17.94) (16.41) 0.85 271267a238 121 Type III Biface Whole Ovate Uwharries Southern Rhyolite (1) 13.57 51.63 33.44 21.68 271267a249 125 Type III Biface Whole Ovate Uwharries Southern Rhyolite (1) 5.71 40.21 25.32 7.9 271267a681 336 Type III Biface Fragment Lateral Section Uwharries Western Rhyolite 5.83 (38.60) (15.76) 3.55 271267a1142 505 Hafted Biface Tool Tip Missing Uwharries Southern Rhyolite (2) 9.95 (37.52) 20.54 6.25 Ovate Elongate Hardaway Blade? Hardaway Blade? 15.19 Maximum Length (mm) Specimen a249 (Figure 80:G) was recovered at a depth of 50 to 60 cm bs in a Stage II shovel test excavated at N470/E485, while Specimen a238 was found at a depth of 30 to 40 cm bs in a Stage I shovel test excavated at N460/E490. These locations are within 11 m of one another, which suggests that this area of the site could contain a Late Paleoindian occupation. The prismatic blade discussed earlier is also an indication of Paleoindian occupation, but it was found in TU1, approximately 60 m to the northeast. (4) Hafted Biface. A single, bifacially flaked tool was identified in the collection that could not be successfully assigned to a projectile point style, although it probably functioned as one (Figure 80:H). The specimen (a1142) was made on a core flake of USR (2) material. The blade was well shaped and regularized and exhibited a narrow, convex outline. The base, on the other hand, consisted of the unmodified platform of the flake blank, which retained cortex on the dorsal face. Wide notches above the unmodified portion of the make shift stem served as the point of attachment for the haft. The tool was found at a depth of 10 to 20 cm bs in a Stage III shovel test excavated in SB4. Projectile Points Projectile points are bifacially flaked tools with retouched haft elements. In addition, projectile point fragments consist of specimens with the fine bifacial retouch characteristic of finished projectile points, but not necessarily exhibiting evidence of a haft element. The collection from 31HK2521 consisted of 8 diagnostic projectile points and 9 fragments (Table 67). Metric and descriptive data on the projectile points can be found in Appendix E. (1) Palmer I Corner Notched. Coe (1964:67) described the Palmer Corner Notched type as a “small corner-notched blade with a straight, ground base and pronounced serrations.” Blade edges were generally straight, but were secondarily both incurved and excurvate. Bases were commonly straight and typically ground. Serrations were at times deep, averaging about 3 mm in width and 5 to 7 mm in length. The width of the shoulder barbs usually exceeded that of the base. Metric dimensions ranged between 28 and 60 mm in length, 15 and 25 mm in width, and 5 to 12 mm in thickness. At the Hardaway Site, Palmer Corner-Notched points were primarily associated with Level III, sandwiched between Hardaway SideNotched and Hardaway-Dalton styles in Level 4 and Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 257 Chapter 7. 31HK2521 Figure 80. Bifaces, Site 31HK2521. A-C: Type I Biface Fragments (a2165, a517 and a908 respectively), D: Type II Biface Fragment (a82), E: Type III Biface Fragment (a209), F-G: Type III Bifaces, Hardaway Blades? (a238 and a249 respectively), H: Hafted Biface Tool (a1142). Kirk Serrated and Kirk Corner-Notched in Level II. As an outcome of the Haw River excavations, Cable (1982) further divided the Palmer type into three variants, Palmer I, II and III. The Palmer I style is smaller, typically exhibits ground bases and was found in a slightly lower vertical position at Haw River than the larger variants described as Palmer II and Palmer III. The latter is the largest of the three variants and is characterized by a low incidence of basal grinding. It would correspond closely to the ideal type description for the Kirk Corner Notched type (Coe 1964:69-70). Palmer II is stylistically intermediate and it may overlap the temporal distributions of the other variants. It is larger than Palmer I specimens, but is characterized by a high incidence of basal grinding. Others suggest that all of this variation is largely a consequence of life-stage variation rather than temporal differences (Daniel 1998; Sassaman 2002). Although this may be the case for Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 258 BAG 220 431 272 756 363 818 340 239 128 178 263 425 451 721 779 815 845 SPECIMEN No. 271267a450 271267a951 271267a552 271267a1768 271267a742 271267a2060 271267a686 271267a507 271267a255 271267a342 271267a540 271267a927 271267a998 271267a1648 271267a1897 271267a2038 271267a2258 PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment Badin Crude Triangular, Var C Morrow Mountain I Stemmed Palmer I Corner Notched Palmer I Corner Notched Undefined Expended Stemmed Undefined Expanded Stemmed Undefined Rudimentary Stemmed Yadkin Large Triangular, Var A TYPE Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) 0.71 Uwharries Southern Rhyolite (2) Tip 0.70 White Quartz 8.22 White Quartz Lateral Section 0.67 5.18 2.68 1.05 9.35 0.34 8.87 Blade White Quartz Uwharries Southern Rhyolite (1) Blade Midsection Tip White Quartz Upper Blade Tip Blade White Quartz Lateral Section 3.29 4.35 White Quartz 3.68 Uwharries Southern Rhyolite (2) 6.99 2.31 5.08 6.86 WEIGHT (gm) Type I Rhyolite Tuff Uwharries Western Rhyolite Uwharries Southern Rhyolite (1) Uwharries Eastern Rhyolite (f) White Quartz LITHIC RAW MATERIAL TYPE Whole Tip Missing Whole Base Whole Whole Whole Lateral Section PORTION MAXIMUM THICKNESS (mm) 4.91 7.45 9.73 5.18 7.33 7.12 6.33 6.27 4.58 7.44 5.58 8.59 5.78 9.00 4.38 7.60 9.12 MAXIMUM LENGTH (mm) (12.27) (11.22) (40.66) (19.00) (30.58) (27.92) (14.60) (45.71) (12.66) (46.01) (28.05) 31.04 (28.94) (39.72) 33.66 33.30 (35.11) (27.07) 13.05 9.35 (18.44) 16.96 BASAL WIDTH (mm) PROJECTILE POINTS: 31HK2521 TANG WIDTH (mm) 22.04 11.83 8.53 16.96 13.80 20.38 29.87 18.42 18.10 19.74 25.71 19.38 19.27 SHOULDER WIDTH (mm) Table 67. Summary Data for Projectile Points, 31HK2521. TANG LENGTH (mm) 4.38 5.18 6.05 9.62 8.24 8.28 BLADE LENGTH (mm) 25.86 25.42 25.0 BASAL GRINDING Present Present BLADE SERRATION Present Present EDGE ANGLE 50° 67° 71° 60° 53° 46° 52* 53° 60° 61° 55° 66° 63° 65° 46° 82° 70° Slightly Concave Straight Slightly Convex Straight Slightly Convex Slightly Convex BASE SHAPE Chapter 7. 31HK2521 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 259 Chapter 7. 31HK2521 Palmer II and III, this explanation does not take into account the much narrower tang widths of Palmer I specimens. The time range of this style is generally accepted at 9500–8900 BP (Sassaman et al. 2002:10). Two projectile point specimens from 31HK2521 were identified as Palmer I Corner Notched. Specimen a552 (Figure 81:A) was found between 10 and 20 cm bs in a Stage II shovel test excavated in SB14. It was manufactured from USR (1) and the base was slightly convex and lightly ground. The central portion of the base had been spalled off because of use or repair and grinding was not evident in this region of the basal margin. The blade margins were straight to slightly convex and the remnants of serration can be seen on both blade edges. Specimen a1768 (Figure 81:B) was missing its tip and one of its basal ears. The base was heavily ground and it had taken on a slightly concave outline due to episodes of basal thinning. It was recovered from TU4 where it was piece plotted at a depth of 39 to 40 cm bd. It was made from UER material with exclusively feldspar phenocrysts. (2) Morrow Mountain I Stemmed. A single Morrow Mountain I Stemmed point (a951) was identified in the collection from 31HK2521 (Figure 81:C). The Morrow Mountain I Stemmed type was originally defined by Coe (1964:37-43) from his excavations at the Doershuk Site and it is similar in morphology to the Morrow Mountain II Stemmed point. Both styles were characterized by contracting stems, but two differences were cited. The blade is proportionately longer and narrower and a noticeable angle occurs at the shoulder/stem juncture in the Morrow Mountain II Stemmed type. The shoulder of the Morrow Mountain I style, by contrast, curved into the stem without a break in the marginal contour. There is some evidence to suggest that Morrow Mountain II occurred at a slightly higher stratigraphic/vertical position than Type I in the Doershuk Site excavations, which led Coe (1964:43) to tentatively posit that the former was “transitional and survived to a later date.” This type is similar in length to the Morrow Mountain II Stemmed type (30–80 mm), but is noticeably broader at the shoulder (18–30 mm). Specimen a951 was whole and occupied a position on the short end of the range of length measurements reported for the type. Its maximum length was 33.66 mm. Slight shifts in the orientation of the upper stem margins are observable near the juncture with the blade, but the shift was not deemed sufficiently distinct to merit assignment of the point to the Morrow Mountain II style. The blade margins were out-curved or convex in outline. Retouch along the blade margins had created steep edge angles in the range of 75° to 82°. The point was found in a Stage III shovel test excavated in SB7 at a depth of 20 to 30 cm bs. (3) Badin Crude Triangular Variants. Coe (1964:45) originally defined the type from excavations conducted at the Doerschuk Site. It was described as a “large, crudely made triangular point.” The specimens from Doerschuk were manufactured exclusively of direct percussion and were not finished out with secondary pressure flaking. Coe entertained the hypothesis that Badin Crude Triangular points represented unfinished Yadkin Large Triangular points, which were much more finely flaked. However, he noted that Yadkin points occupied a somewhat higher stratigraphic position than Badin points at the Doerschuk Site, leading him to conclude that the two styles were discrete types related to each other in a chronological sequence. The modal outline of the Badin type consists of a triangular-shaped blade with a concave base, but reference to the Badin point photographic plate (Coe 1964:Figure 41) reveals three shapes: (1) excurvate to straight sided triangular with straight bases, (2) excurvate to straight sided triangular with concave bases and (3) lanceolate with concave bases. Although this variation may simply reflect life cycle stages within a single constellation of typological traits, it is just as possible that it represents a chronological sequence of design style changes. If the stratigraphic and spatial dimensions of this variation are not monitored through typological recognition, then this hypothesis will remain inadequately addressed. For this reason, it has been proposed (Cable 2010:150-151) that each of these shape categories be recognized as separate variants of the larger Badin Crude Triangular type: (1) Variant A: excurvate to straight sided triangular/straight base, (2) Variant B: excurvate to straight sided triangular/ concave base and (3) Variant C: lanceolate/concave base. A single white quartz specimen (a450) found during the investigation corresponds to Variant C of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 260 Chapter 7. 31HK2521 Figure 81. Projectile Points, Site 31HK2521. A-B: Palmer I Corner Notched (a552 and a1768 respectively), C: Morrow Mountain I Stemmed (a951), D-E: Unidentified Stemmed Midsections (a998 and a342 and respectively), F: Badin Crude Triangular, Variant C (a450), G: Yadkin Large Triangular, Variant A (a507), H: Projectile Point Upper Blade (a1897), I-J: Unidentified Expanded Stemmed (a742 and a2060 respectively), K: Unidentified Rudimentary Stemmed (a686). the Badin Crude Triangular type (Figure 81:F). It is a lanceolate shaped point with excurvate blade edges in which the maximum width occurs about one-third distance from the base. The base is straight rather than concave as is more common in this variant, but the maximal width position substantially above the base makes it more appropriate to place it within the Variant C range of morphological variation. The blade faces exhibited evidence of only percussion flaking. Attempts to further shape the margins with secondary percussion flaking or pressure flaking were not evidenced. The original margin on one side of the blade had been broken off in a step fracture, but the break did not seriously impact the functionality of the tool. The specimen was found in a Stage II shovel test in SB18 at a depth of 30 to 40 cm bs. Variant C most Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 261 Chapter 7. 31HK2521 closely resembles the Nolichucky and Greenville types of eastern Tennessee (Kneberg 1957:64-65) and the Copena point of the Mid-South (Webb and DeJarnette 1942:37; Cambron and Hulse 1969:25), which date to the time range of 3,000 to 1,500 BP. (4) Yadkin Large Triangular. Coe (1964:4549) defined the Yadkin Large Triangular type from excavations conducted at the Doerschuk Site. The type was described as “a large, symmetrical, and well made triangular point.” Bases were usually concave and blade margins ranged from excurvate to straight to incurvate. Coe (1964:45) observed that all specimens were finished by careful pressure flaking. As is true of the Badin type, a great deal of morphological variation is contained within the Yadkin Large Triangular type description. Although this variation may simply reflect life cycle stages within a single constellation of typological traits, it is just as possible that it represents a chronological sequence of design style changes. If the stratigraphic and spatial dimensions of this variation are not monitored through typological recognition, this hypothesis will remain inadequately addressed. For this reason, Cable (2010:416) has proposed that various shape configurations of the Yadkin type be segregated into several variants, as was also done for the Badin type. Four variants are identified in the plate of Yadkin points in the Doerschuk Site report (Coe 1964: Figure 42): (1) Variant A: triangular blade w/excurvate blade margins and concave base, (2) Variant B: triangular blade with straight blade margins and concave base, (3) Variant C: triangular blade with incurvate blade margins and concave base, and (4) Variant D: Lanceolate blade with excurvate blade margins and concave bases with side notched ears. A single Yadkin Large Triangular point was identified in the collection (Figure 81:G). It exhibited slightly convex blade margins and a concave base, which is consistent with the definition of Variant A. The point was carefully flaked with primary and secondary percussion shaping. Both basal ears and the very distal end of the tip had been broken off. The specimen was found at a depth of 25 to 35 cm bs in a Stage III shovel test excavated in SB13. (5) Undefined Expanded Stemmed. Two points with short, expanding stems were recovered from the investigations at 31HK2521 (Figure 81:I-J). Existing typological correlates for these points have not yet been defined for the region. South (2005:64) has described a small, stemmed point form his work in the Roanoke River Reservoir that he named the Thelma point that may be somewhat analogous. These points range from 27-41 mm in length and 16 to 21 mm in width, which is within the general size range of the two specimens from 31HK2521 (Table 67). These points appear to be associated with the Vincent phase, which is assigned to the Early Woodland period. However, the morphology of the haft of the Thelma point differs in significant ways. The Thelma points have much longer stems, predominantly ranging between 10 and 13 mm, while the tangs of the specimens from 31HK2521 are quite squat, ranging between 5.18 and 6.05 mm. Moreover, although some examples of the Thelma point have broadly side notched stems, most are straight with obtuse blade/haft junctures. The specimens from 31HK2521 are strongly corner notched and the shoulders are slightly barbed. Although they may represent a transition from the Savannah River Stemmed style to the large triangular points of the Early Woodland period as South suggests for the Thelma point, the specimens from 31HK2521 are smaller than typical Yadkin and Badin points and they exhibit much finer pressure flaking along the blade margins. These characteristics suggest that these points may occur later in the Woodland sequence and may co-occur with later triangular point types. Millis et al. (2010:416-419) have provided an overview of small, stemmed projectile points from Fort Bragg that are inferred to represent Woodland types. These include Woodland Contracting Stemmed, Woodland Expanded Stemmed and Woodland Straight Stemmed. The expanded stemmed category approximates the size parameters of the specimens from 31HK2521, although a great deal of morphological variation appears to exist in the group. However, one illustrated point (Millis et al. 2010:Figure 20.14c) approaches the morphology of the specimens. More information is needed before the chronological and cultural position of this point type can be more firmly established. Specimen a742 (Figure 81:I) was made of microcrystalline Type I RT. The margins are parallel from the shoulder until about two-thirds of the distance to the tip, at which point one margin appears Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 262 Chapter 7. 31HK2521 to round inward to form a stub-nosed tip similar to that illustrated by the whole specimen (Figure 81:J). The shoulders are barbed and the corner-notched stem expands slightly toward the base. The stem is extremely narrow, extending no more than about 8.5 mm. The blade margins appear to have been pressure flaked. The base is slightly convex in outline. It was found in a Stage III shovel test in SB10 at a depth of 10 to 20 cm bs. Specimen a2060 (Figure 81:J) was whole and exhibited a wider tang relative to shoulder width than Specimen a742. Blade margins were carefully flaked by percussion and the edges were relatively steep. At a point approximately twothirds distant from the shoulder, the blade margins cut inward to form a shortened tip. The base was straight and the tang was formed by carefully chipped corner notches. The point was recovered at a depth of 20 to 30 cm bd in TU11. (6) Undefined Rudimentary Stemmed. A single specimen (a686) with a very short contracting stem could not be correlated with a known type in the region (Figure 81:K). Some forms of the Randolph Stemmed type exhibit contracting stems, but these points are narrow, thick, poorly flaked and typically made from old flakes (Coe 1964:49-50). The specimen from 31HK2521 is thin, relatively well made and exhibits an extremely shortened stem, unlike the Randolph Stemmed type. Perhaps a correlate for the specimen is the Ebenezer point from Tennessee and Alabama, which is associated with Early to Late Woodland contexts (Cambron and Hulse 1975:42). The Ebenezer point is a small point with a short rounded and contracting stem. Blade edges are usually excurvate and the cross-section is biconvex. In all of these characteristics, the specimen from 31HK2521 conforms. Anderson et al. (1982:156-157) describe an identical point form from the Mattassee Lake sites in the middle Santee River Valley that they identified as a Group 8 projectile point. They do not suggest an age for the point based on evidence from their own work, but they note that Trinkley (1981) described a similar point style from a site (38LX5) near Columbia, SC, which he named Deptford Stemmed. The implication of the name is that the point is associated with the Middle Woodland period. Millis et al. (2010:418) have identified a number of presumed Woodland-affiliated contracting stem points at Fort Bragg, but the group is typologically heterogeneous and in need of further analytical separation. The Undefined Rudimentary Stemmed type from 31HK2521 exhibits a short tang of 4.38 mm in length and the stem breaks slightly at the should juncture. The blade margins have been carefully retouched by delicate percussion flaking. One blade margin is excurvate, while the other is incurvate in an equal arc. The tip was snapped off, probably during use. Had the base been straight, the size and workmanship of the point would be consistent with the Uwharrie (Coe 1964:121), Clements or larger Roanoke triangular types (South 2005:62). The specimen was recovered in a Stage III shovel test in SB8 at a depth of 20 to 30 cm bs. (7) Projectile Point Fragments. Although most of the projectile point fragments consist of small tips, lateral sections and midsections, three of them are larger and contain some diagnostic information. Specimen a998 (Figure 81:D) consisted of a large midsection fragment that retained the haft/ shoulder junction. The blade was carefully shaped by percussion flaking and the stem appeared to contract below the shoulder. The size and morphological characteristics suggest that the specimen may represent a Morrow Mountain Stemmed point fragment. It was found at a depth of 10 to 20 cm bs in a Stage III shovel test located in SB6. Specimen a342 (Figure 81:E) is also a large midsection containing the haft/ shoulder juncture. The stem again is contracting below the shoulder, but the overall size of the point and the somewhat cruder flaking are more in keeping with the qualities of a Savannah River Stemmed point. It was also found in SB6 in a Stage III shovel test at a depth of 20 to 30 cm bs. Finally, a large white quartz upper blade fragment (Figure 81:H, Specimen a1897) suggests an Archaic affiliation. It was found in TU5 at a depth of 0 to 10 cm bd. Unifaces Unifaces consist of tools that are retouched along flake margins in a single direction (unifacial edge shaping). Retouch was applied to shape a working edge or bit and, sometimes, a butt element for hafting. Limited bifacial retouch or resharpening is sometimes present along haft elements or working edges. Metric and attribute data for unifaces are presented in Appendix F. Twenty-five stone tools classified as unifaces were recognized in the collec- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 263 Chapter 7. 31HK2521 tion (Table 68). These included 5 end scrapers, 10 side scrapers, a pointed scraper, two indeterminate scraper fragments, two denticulates, three gravers, a retouched flake and a spoke shave. The specimens are described below under the rubric of the typological system developed by Coe (1964) and elaborated by Daniel (1998). (1) End Scrapers. Six end scraper specimens were identified in the 31HK2521 collection, of which one fragment was not identifiable to type. Daniel (1998:66) defines end scrapers as unifaces with “steep, regular unifacial retouch that forms a relatively narrow convex bit or working edge.” He further notes that the bit edge is commonly positioned in a transverse orientation to the long axis of the tool. Four main types and a series of subtypes have been recognized; including both hafted and unhafted forms. Specifically identified during the investigation at 31HK2521 were Type Ia, Type Ib, Type IIa and Type IIb end scrapers. These are described below. (a) Type I End Scrapers. Two Type I End Scrapers were identified in the collection. These are distinctive unifacial tools commonly found in both Paleoindian and Early Archaic tool kits in North America (Collins 2005). Coe (1964:76) observed that the Type I End Scrapers varied little from the descriptions of similar tools found at the Lindenmeier, Quad, Williamson, Shoop and Bull Brook Paleoindian sites. He recognized two forms in the Hardaway collection that Daniel (1998:66-68) later formalized into two subtypes. Type Ia included triangular to trapezoidal shaped end scrapers with wellexecuted unifacial retouch on lateral and distal edges. Type Ib End Scrapers are distinguished by additional facial retouch across the dorsal surface of the flake blank that produced a smooth and rounded contour resulting in a ‘tear-drop” plan view. Specimen a2181 was assigned to the Type Ia subtype because of a lack of extensive cross-face retouch on the dorsal surface (Figure 82:A). It was made of USR (1) material and it was recovered at a depth of 20 to 30 cm bs in TU10. Both ends of the tool appear to have been made into bits. The primary bit was manufactured on the highest part of the dorsal surface of the flake and exhibits broad pressure flakes that run to the top of the central keel of the flake. The working edge, which exhibits heavy edge dulling and polish from use, has been undercut by a series of step fractures that extend not more than 4 mm up the face. Once the undercutting had progressed to this point, it would appear that the obverse end, which had served as the original but, was converted to a scraper edge. This face was made on a broken surface and marginal retouch extended only about 3 mm above the working edge. It too, exhibited edge dulling and polish from use. Both unifacial lateral margins of the tool were heavily retouched or “backed” to accommodate hafting. Specimen a376 (Figure 82:B) was classified as a Type Ib End Scraper. It was made on a thin, trapezoidal core flake of USR (1) material. The bit had been shaved off by a shovel during recovery operations, but the careful unifacial retouch typical of these scrapers was evident on both lateral edges. All of the edges had been subject to hydration erosion. The bit appears to have been convex in outline and the tool was triangular to “tear-drop” in shaped in plan view. It was recovered at a depth of 40 to 50 cm bs in a Stage II shovel test that anchored SB8. (b) Type II End Scrapers. Coe (1964:76) defined the Type II End Scraper from his excavations at the Hardaway Site. It included hafted end scrapers made on flakes of variable shapes and sizes that were retouched only along the narrowest margins to form scraping edges. He recognized two varieties that Daniel (1998:72) later formalized. Type IIa End Scrapers consisted of large, thick irregular flakes that were casually shaped at one end to produce a scraper edge. Cutting edges were always irregular in marginal outline and sharp, suggesting to Coe that they were used to work hard material such as wood or bone. They were primarily confined to Level II at the Hardaway Site, which was primarily associated with Kirk Stemmed and Serrated and Stanly Stemmed points. Type IIb End Scrapers were made on thin, narrow, prismatic flakes. The scraping edge was formed along the end opposite the striking platform. This variety was not numerous, but it had the same stratigraphic association as the Type IIa variety. Daniel (1998:72-73) expanded the type definitions to also include specimens with limited marginal retouch on lateral edges. Three specimens were identified as Type II end scrapers in the 31HK2521 collection. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 264 566 335 368 663 898 409 919 427 842 59 292 444 575 271267a1279 271267a679 271267a755 271267a1527 271267a2486 271267a873 271267a2558 271267a938 271267a2225 271267a115 271267a585 271267a982 271267a1301 645 443 521 271267a1477 271267a981 271267a1172 33 423 271267a919 271267a69 748 271267a1719 17 544 271267a1233 271267a35 88 498 271267a1122 191 271267a376 271267a172 833 BAG 271267a2181 SPECIMEN NO. Spoke Shave Retouched Flake Graver Graver Graver Denticulate Denticulate Scraper Fragment Scraper Fragment Side Scraper Fragment Type III Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type I Side Scraper Type I Side Scraper End Scraper Fragment Type IIb End Scraper Type IIa End Scraper Type IIa End Scraper Type Ib End Scraper Type Ia End Scraper UNIFACE/RETOUCHED FLAKE TYPE PORTION Fragment Fragment Fragment Fragment Fragment Fragment Whole Fragment Fragment Fragment Whole Fragment Fragment Fragment Fragment Whole Whole Whole Whole Fragment Whole Whole Fragment Whole Whole BLANK TYPE Core Flake Core Flake Blade Flake Core Flake Blade Flake Core Flake FBR Core Flake Core Flake Core Flake Core Flake Core Flake Core Flake Core Flake CROSS-SECTION Uwharries Southern Rhyolite (1) Uwharries Western Rhyolite PlanoConvex Mill Mountain Rhyolite Uwharries Western Rhyolite Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Type I Rhyolite Tuff Type I Rhyolite Tuff Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz White Quartz Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Trapezoidal Trapezoidal Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular White Quartz Uwharries Eastern Rhyolite PlanoConvex Trapezoidal Crystal Quartz Trapezoidal Uwharries Eastern Rhyolite ConvexTrapezoidal Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) PlanoConvex Lenticular Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Lenticular PlanoConvex LITHIC RAW MATERIAL TYPE BIT RETOUCH LENGTH (mm) 17.17 (25.55) (7.45) (5.89) 7.96 (15.64) 32.76 (13.10) (11.62) (11.43) 25.07 9.48 (17.45) (20.72) (23.20) (10.26) (12.42) 33.15 25.44 (13.46) 15.88 22.48 16.01 (19.22) 19.91 21.09 (7.73) (12.48) (3.73) (6.37) (6.26) 30.62 22.08 17.54 12.14 36.23/29.26 23.54 36.64/34.42/ 18.79 OTHER BIT RETOUCH LENGTH (mm) SCRAPERS AND RETOUCHED FLAKES: 31HK2521 72° 37° 70° 42° 22° 63° 46° 77° 72° 58° 68° 67° 60° 59° 75° 56° 44° 65° 84° 44° 71° 65° 78° MAXIMUM EDGE ANGLE 75° 70° 50° 99° 4 NUMBER OF RETOUCHED EDGES 2 1 2 2 1 (1) 1 (1) (1) 1 1 2 2 (1) 2 (1) 1 2 2 2 (2) 3 2 7.47 7.09 6.69 5.36 4.07 5.23 3.52 6.63 3.40 3.68 3.53 3.71 2.74 5.70 6.35 4.17 4.88 10.46 8.71 9.51 7.98 8.50 10.77 4.83 13.18 MAXIMUM THICKNESS (mm) Table 68. Summary Data for Scrapers and Retouched Flakes, 31HK2521 MINIMUM EDGE ANGLE 40.96 MAXIMUM LENGTH (mm) 55.28 (26.46) (16.22) (9.32) (22.82) (17.16) 36.52 (15.90) (11.78) (12.18) 28.94 (22.93) (19.80) (34.41) (28.33) 31.70 41.65 45.53 35.27 (16.58) 25.50 43.54 (43.95) 30.99 23.09 MAXIMUM WIDTH (mm) 7.44 (20.63) (18.24) (15.98) (28.82) 16.80 30.69 (9.87) (11.24) (10.05) 14.68 14.59 (21.83) (20.47) (33.41) (17.63) 19.28 30.23 23.59 (13.63) 31.28 33.67 (34.10) 24.96 WEIGHT (gm) 6.14 2.44 1.15 0.44 1.91 1.20 3.61 0.80 0.33 0.33 1.34 1.58 1.21 3.55 5.27 2.80 3.43 14.80 8.82 2.51 4.84 12.71 13.43 4.03 15.26 Bifacially Retouched Margin Utilized Margin Opposite Graver Alternate Face Retouch, Opposite Edge Utilized Alternate Face Retouch Flake from MD Core NOTES Chapter 7. 31HK2521 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 265 Chapter 7. 31HK2521 Figure 82. Unifaces, Site 31HK2521. A: Type Ia End Scraper (a2181), B: Type Ib End Scraper (a376), C-D: Type IIa End Scrapers (a172 and a1122 respectively), E: Type IIb End Scraper (a1233), F: End Scraper Fragment? (a1719), G-H: Type I Side Scrapers (a919 and a1279 respectively), I-J: Type IIa Side Scrapers (a679 and a1527 respectively), K-M: Type III Side Scrapers (a873, a2558 and a938 respectively), N-O: Denticulates (a982 and 1301 respectively), P: Graver (a35), Q: Retouched Flake (a981), R: Spoke Shave (a1172). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 266 Chapter 7. 31HK2521 (2) Side Scrapers. Ten side scraper specimens were identified in the 31HK2521 collection, of which nine were sufficiently complete to type further. Daniel (1998:6683-84) defines side scrapers as unifaces containing “one or more unifacially retouched working edges that parallel the long axis of the flake blank.” Four side scraper types and numerous subtypes were identified by Daniel in the Hardaway scraper collection. Side scrapers from 31HK2521 can be placed in Types I, II and III. (a) Type I Side Scrapers. Two Type I Side Scrapers were identified in the collection. Coe (1964:77) defined this scraper type as being made from large, wedge-shaped flakes produced from block (directional) cores. One of the long sides of the flake blank was subjected to steep percussion retouch, while one or both ends of the blank also received retouch shaping. Daniel (1998:84) further adds that the bit edges were generally convex to straight in shape and cross-sections appeared wide and flat (lenticular). Type I Side Scrapers were equally distributed between Levels II, III and IV at the Hardaway Site, in association with Kirk and Palmer Corner Notched and Hardaway Side Notched points and Hardaway Blades. Specimen a919 (Figure 82:G) was manufactured on a core flake of UER material. Two edges received unifacial retouch. One margin was retouched on the dorsal surface of the flake, producing a steep working edge averaging about 84°. An adjacent edge exhibited retouch on the ventral surface of the flake. This edge was more acute and the edge was thinner, suggesting that it may have been purposed for delicate tasks such as light scraping and cutting. The tool was found at 30 to 40 cm bs in a Stage III shovel test excavated in SB9. Specimen a1279 (Figure 82:H) was made of white quartz. Retouch was placed on both lateral edges of the flake, one exhibiting a convex outline and one a slightly concave one. Edge angles ranged between 50° and 65° and the substantial nature edges suggest that the tool was used to scrape hard materials. It was found in a Stage III shovel test excavated in SB15 at a depth of 10 to 20 cm bs. (b) Type II Side Scrapers. Building on Coe’s (1964:79) original description for the Type II Side Scraper, Daniel (1998:84-93) separated indi- vidual specimens further into two types based on the size of the flake blanks used to manufacture the tools. Type IIa refers to the classic description of a Type II Side Scraper. That is, scrapers made on large, irregular flakes whose bit edges were retouched without concern for regularizing the marginal outline. Daniel’s Type IIb was reserved for very small flakes, usually of less than 30 mm in length. The generally thin edges on these scrapers suggest that they were used for more delicate uses and on softer materials than Type I Side Scrapers. Coe (1964:79) observed that Type II Side Scrapers were strongly associated with the Palmer and Hardaway levels (III and IV) at the Hardaway Site. Four Type IIa Side Scrapers were recognized in the 31HK2521 collection (Figure 82:I-J). Most were manufactured on core flakes from directional cores, although one specimen (a755) was made on a large FBR. Angles on retouched edge margins ranged between 56° and 75°. One specimen exhibited two retouched edges (a1527), while the remainder displayed only one. All of the working edges were delicately retouched by pressure flaking and at the time of their discard, all bit margins exhibited light dulling. Individual specimens were predominantly recovered from vertical positions between 30and 60 cm bs, although one specimen (a755) was found between 10 to 20 cm bs. In general, the depth of the Type IIa Side Scraper proveniences is in line with an Early Archaic association. Two Type IIb Side Scrapers were identified in the collection (Figure 82:K-L). These specimens were fashioned from small, thin flakes that were predominantly derived from directional cores. Bit margins were delicately retouched with pressure flaking and working edge angles ranged between 60° and 67°. Both specimens exhibited two retouched edges and bit margins were lightly dulled. The specimens were found at variably different vertical depths. Specimen a2558 (Figure 82:K) was recovered at a depth of 60 to 70 cm bs in a Stage III shovel test excavated in SB11, while specimen (a873) was encountered at a depth of only 20 to 30 cm bs in a Stage III shovel test located in SB9. (c) Type III Side Scrapers. One Type III Side Scraper was identified in the collection from 31HK2521 (Figure 82:M). Type III Side Scrapers Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 267 Chapter 7. 31HK2521 were made on thin and narrow flakes (Coe 1964:79). Retouch occurred along the edges parallel to the long axis of the flake and generally involved reshaping of only the working edges. No accommodations for hafting were present. This scraper type was found predominantly in Levels II and III at Hardaway, levels associated with Palmer through Morrow Mountain occupation (Coe 1964:73). Very few of these forms were found at Doerschuk, however, which suggests that their primary association may be with Early Archaic phases. Daniel (1998:93) indicates that the Type III Side Scrapers from the Hardaway collection were made on bladelike flakes, which, in this report, are referred to as blade flakes. The specimen (a938) from 31HK2521 displayed retouch along only one lateral margin. It was manufactured from USR (1) material. The average working edge angle was 68°. It was recovered from a Stage III shovel test in SB7 at a depth of only 10 to 20 cm bs. (b) Gravers. Gravers represent flakes with short unifacially flaked projections (Daniel 1998:104; Goodyear 1974:55). It is generally held that gravers served as engravers or points for slotting bone and wood. Three flakes exhibited minimally retouched projections suggestive of graver points. Specimen a35 (Figure 82:P) was made of USR (2) material and was recovered in a Stage I shovel test at a depth of 50 to 60 cm bs. Specimens a69 and a1477 also displayed utilized cutting edges situated on one lateral edge opposite the placement of the graver projections. The former was made of UWR material and was found at a depth of 50 to 60 cm bs in a Stage I shovel test. The latter specimen was also recovered between 50 and 60 cm bs, in a Stage III shovel test located in SB18. It was made of MMR material. Gravers are typically associated with Early Archaic assemblages (Goodyear 1974:53) and the depth at which these specimens were found is consistent with this observation. (3) Other Retouched Tools. In addition to the more conventional scraper types, seven additional retouched tools were recognized in the 31HK2521 collection. These consist of two denticulates, three gravers, a retouched flake and a spoke shave. Each of these is described below. (c) Retouched Flake. Specimen a981 (Figure 82:Q) exhibited a single sharp edge with informal bifacial retouch along the margin. It was manufactured from USR (1) material. The tool was found at a depth of 40 to 50 cm bs in a Stage III shovel test located in SB7. (a) Denticulates. Denticulates represent flakes that exhibit a series of regularly spaced notches/ projections along a working edge creating a saw-like appearance (Daniel 1998:103). The morphology of the working edge suggests that the tool functioned as a knife or saw for processing materials of soft to moderate hardness. Two denticulates were recovered in the collection from 31HK2521. Specimen a982 (Figure 82:N) was manufactured on a thin core flake of USR (1) material. The denticulated edge consists of six regularly spaced short projections made by concave pressure retouch to one face of the flake. The acute edge angle (46°) formed along the span of retouch and the thin nature of the edge suggest that the tool was used to cut soft tissue like meat. It was recovered at a depth of 10 to 20 cm bs in a Stage III shovel test located in SB7. The other denticulate was fragmentary and exhibited much larger, thicker and rounded teeth (Figure 82:O). The structure of this edge suggests that this tool was used to saw hard, resistant material such as bone or wood. It was made of USR (2) material and was found at a depth of 20 to 30 cm bs in a Stage III shovel test excavated in SB13. (d) Spoke Shaves. One specimen (a1172) in the collection corresponds to the definition of a spoke shave (Figure 82:R). Typically, unifacial retouch is applied to a natural concavity on a flake margin to produce a strong working edge from which to scrape or shave narrow convex surfaces such as spear or arrow shafts. Goodyear (1974:50-53) identified both hafted and unhafted spoke shaves at the Brand Site, and Daniel (1998:104-105) recognized two hafted specimens in the Hardaway collection. Goodyear believes that most of the spoke shaves he recovered at the Brand Site were associated with the Dalton occupation. The specimen from 31HK2521 is elongate and may have been hafted. The tool displays three retouched concavities that could have been used as spoke shave edges. A wide projection located between the two main concavities appears to represent a graver spur, which is a common characteristic of hafted spoke shaves at the Brand Site (Goodyear 1974:51-52). This phenomenon was also noted on the two hafted spoke shaves Daniel recognized in the Hardaway collection. The 31HK2521 specimen was manufactured from UWR material. It was recovered Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 268 Chapter 7. 31HK2521 from a Stage III shovel test in SB3 at a depth of 10-20 cm bs. This vertical position is not consistent with an Early Archaic association, but the variable depths at which Early Archaic tools are found at the site suggests that the substantial Woodland period occupation has resulted in significant vertical displacement of earlier deposits. Utilized Flakes Sixteen non-retouched flakes exhibited wear patterns on one or more edges, indicating their use as expedient tools (Table 69). Metric and attribute data for utilized flakes are presented in Appendix G. All of these specimens exhibited edge attrition and bifacial nibbling scars indicative of cutting functions. Utilized edge angles fell between 24°and 64°. Most of the specimens appear to have been derived from directional cores (i.e. core flakes and blade flakes), but one (a967) represented a large FBR. Aphyric metavolcanic material types dominate the collection. Only one specimen of porphyritic rhyolite (UER) and one of white quartz were identified. Use wear would be easier to detect on the microcrystalline, aphyric rock types, however, which suggests that the porphyries and quartz might be under-represented because of observational difficulties. Miscellaneous Stone Tools Several non-chipped stone, lithic artifacts were recovered from the investigation at 31HK2521. These consisted of a grinding stone/abrader, five hammer stones, a hammer stone/anvil, a nutting stone, three pebbles/tools, a steatite bowl fragment and a worked schist tool. Metric and attribute data for these artifacts are presented in Appendix H. Each of these categories is discussed below. (1) Grinding Stones/Abraders. A single ground stone fragment (a2039) made from a rose quartz cobble was identified in the collection (Figure 83:A). The quartz material is extremely fractured and it is possible that a better description of the raw material type would be a cemented quartzite composed of large quartz sand grains. The artifact measures about 51 x 78 x 71 mm. Both faces exhibit obvious grinding wear and the ground edge between the two faces forms a regular arc. The regularity of the edge and the heavily ground nature of the faces suggest that this fragment may have been used to grind seeds and other vegetal remains. It was recovered at a depth of 20 to 30 cm bd in TU11. (2) Hammer Stones. Five cobble hammer stones were identified in the collection from 31HK2521. Specimen a1357 (Figure 83:B) was made from a fractured rose quartz cobble that exhibited battering on both ends. It weighed 137.59 gm and was found at a depth of 30 to 40 cm bs in a Stage III shovel test excavated in SB13. Specimen a1799 (Figure 83:C) was a yellow quartz cobble fragment that exhibited battering on two different edge locations containing cobble cortex. One side of the cobble appears to have been flaked, and may have been minimally used as a flake core after its primary function as a hammer stone ended. It weight 115.78 gm and was found at a depth of 40 to 50 cm bd in TU4. Specimen a2053 (Figure 83:D) was made from a rose quartz cobble and was represented by only an end portion of the original hammer stone. Battering wear was extensively distributed on the cobble cortex surface at the end of the fragment, as the battered face had been worn flat from use. The specimen weighed 140.27 gm. It was recovered from TU11 at a depth of 20 to 30 cm bd. Specimen a668 (Figure 83:E) represented a fragment from a large quartz pebble with “eggshell” cortex. Battering was observed on one end of the fragment. The fragment, which represents a little less than half of the original pebble, weighed 25.73 gm. The small size of the specimen indicates that it was used for final marginal retouch on tools. It was found at a depth of 10 to 20 cm bs in a Stage III shovel test excavated ion SB8. Finally, Specimen a825 (Figure 84:A) was a large (1723.65 gm) cobble of dense igneous rock identified as grano-diorite. Battering occurred around the edges of the tool. Its weight suggests that it might have been used in conjunction with an anvil stone to detach flakes from a core. It was recovered from a Stage III shovel test in SB 10 between 30 to 40 cm bs. (3) Hammer Stone/Anvil Stone. A grano-diorite cobble (Specimen a2286, Figure 84:B) exhibiting intermittent edge battering and an avil depression on one face was recovered at a depth of 25 to 30 cm bd in TU1. The anvil depression was v-shaped and extended to a depth of about 5 mm below the general Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 269 Chapter 7. 31HK2521 surface contour of the cobble. It measured approximately 20 mm in diameter. In all of these characteristics, the depression has a form consistent with use as an anvil for reducing stone cores. The specimen weighed 816.47 gm. It was found approximately 7 m from the large grano-diorite hammer stone (a825) discussed above and it is possible that the two items were used together. (4) Nutting Stone. A split fragment of a nutting stone (Specimen a1133, Figure 85:A) made of chert-cemented sandstone was recovered at a depth of 20 to 30 cm bs in a Stage III shovel test excavated in SB4. The rounded and regular contour of the depression found in the center of one of the faces of the tool indicates that it was used to crack and process nuts and seeds. The item was pecked and ground along the edges to produce a regularized ovoid plan view. The tool was split in the approximate center of the depression, which may have resulted in its ultimate discard. The depression measured approximately 37 mm in diameter and extended to a maximum depth of about 6 mm below the general contour of the face of the tool. It weighed 212.50 gm. (5) Pebbles. Three quartz pebbles were recovered that exhibited no evidence of edge battering (Figure 85:B-D). They may have been stockpiled for use as finishing hammer stones like Specimen a668 discussed earlier. However, it is possible that the pebbles were used to polish and smooth pottery during manufacture. The site contains a large amount of Woodland period pottery, although no evidence of manufacturing was found during the investigation. Examination of the cortical rinds of the pebbles was also inconclusive, although patches of dulled polish could be observed on all of the specimens. This kind of wear could be as easily caused by natural stream Table 69. Summary Data for Utilized Flakes, 31HK2521. WEIGHT (gm) EDGE ANGLE Core Flake LITHIC RAW MATERIAL TYPE MAXIMUM WIDTH (mm) BLANK TYPE MAXIMUM LENGTH (mm) CONDITION MAXIMUM THICKNESS (mm) BAG NO. UTILIZED EDGES SPECIMEN # UTILIZED FLAKE FUNCTION UTILIZED FLAKES: 31HK2521 1 Uwharries Southern Rhyolite (1) 6.74 (29.97) (16.31) 3.17 32° 2 Type I Rhyolite Tuff 7.68 (24.95) (10.74) 2.07 45° 271267a57 27 Fragment Cutting 271267a230 115 Fragment Cutting 271267a254 128 Whole Cutting Blade Flake 2 Uwharries Southern Rhyolite (2) 6.40 39.21 18.59 4.59 43° 271267a967 439 Fragment Cutting FBR 1 Uwharries Southern Rhyolite (1) 4.00 (29.03) (32.09) 2.42 34° 271267a1003 453 Fragment Cutting BTF 1 Type I Rhyolite Tuff 2.05 (22.61) (13.02) 0.63 35° 271267a1174 521 Fragment Cutting 1 Uwharries Southern Rhyolite (2) 7.37 (40.50) (22.57) 5.57 42° 271267a1252 551 Fragment Cutting Blade Flake 2 Uwharries Southern Rhyolite (2) 5.75 (41.85) 25.89 5.95 24° 271267a1253 552 Fragment Cutting Blade Flake 1 Uwharries Southern Rhyolite (2) 3.43 (42.70) 23.10 3.37 27° 271267a1329 586 Fragment Cutting Core Flake 1 Uwharries Southern Rhyolite (2) 6.04 (21.64) (24.75) 2.41 37° 271267a1377 605 Fragment Cutting 2 Uwharries Southern Rhyolite (2) 9.00 35.03 20.03 6.71 52° 271267a1465 643 Fragment Cutting 1 Uwharries Eastern Rhyolite 7.93 (45.46) (20.80) 5.76 36° 271267a1487 649 Whole Cutting Blade Flake 1 Uwharries Western Rhyolite 2.53 36.52 11.01 1.19 30° 271267a2061 818 Whole Cutting Core Flake 1 White Quartz 8.71 (26.36) (28.54) 6.19 64* 271267a2157 829 Fragment Cutting 2 Uwharries Western Rhyolite 16.19 (54.90) (22.37) 12.57 44° 271267a2344 869 Whole Cutting Core Flake 1 Uwharries Southern Rhyolite (2) 7.80 36.53 21.38 4.65 45° 271267a2586 936 Whole Cutting Blade Flake 1 Uwharries Southern Rhyolite (2) 3.60 37.62 15.60 2.11 40° Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 270 Chapter 7. 31HK2521 abrasion as ceramic manufacturing processes. Specimen a 479 (Figure 85:B) was recovered between 10 and 30 cm bs in a Stage II shovel test. It weighed 49.46 gm. Specimen a917 (Figure 85:C) weighed 44.19 gm and was found in a Stage III shovel test excavated in SB9 at a depth of 20 to 30 cm bs. Finally, Specimen a828 was also found in a Stage III shovel test in SB9 at a depth of 20 to 30 cm bs. It weighed 18.74 gm. (6) Steatite Bowl Fragment. A single steatite bowl fragment (Specimen a492, Figure 85:E) was recovered between 30 and 40 cm bs in a Stage II shovel test in SB4. The talc-schist material was medium light gray in color and porous. Steatite vessel fragments are sporadically found on Fort Bragg (see McNutt and Gray 2007; Millis et al. 2010; Ruggiero 2003). The material is commonly found at locations in the Piedmont, although the nearest known quarries are in South Carolina and south central Virginia (see Millis et al. 2010:460-461; Truncer 2004). The interior of the sherd was smoothly finished, although the regular contours were broken by numerous open pockets. The exterior, by contrast, was undulating and irregular. The sherd wall was 16.91 mm thick and the item weighed 18.19 gm. (7) Worked Schist. A single fragment of talc-schist (Specimen a1150, Figure 85:F) exhibited edges that were unifacially flaked by percussion. The original artifact was much larger than the remaining fragment. Roughly flaked, laminated raw material like this have been used as knives to cut and saw semi-resistant plant material in certain archaeological cultures. Large flat phyllite knives were used to cut and harvest agave, for instance, in southern Arizona during precontact times (Fish and Fish 1994). The fragment weighs 48.48 gm. It was recovered between 30 and 40 cm bs in a Stage III shovel test excavated in SB4. Precontact Ceramics An unusually large collection of precontact ceramics was recovered during the investigation at 31HK2521. It consisted of 910 sherds, including representatives from four series, including New River, Hanover, Yadkin and Cape Fear. Two hundred and forty individual vessels were recognized in the analysis. Metric and attribute data on precontact ceramics are contained in Appendix I. Although monitoring of temper types within series has proceeded in an ad hoc fashion in earlier investigations (see Cable and Cantley 2005a, b, 2006, Cable 2010), the enormity of the inventory of temper constituents and various compositional combinations seen in ceramic collections makes it imperative that in the future a universal system be developed to accommodate cross-project comparability. Formerly, temper classes were assigned letter designations as they were historically recognized within each series, creating a confusing list of different letter designations for the same temper type within each series. Here, temper constituent designations are standardized into a single classificatory system. The letters “a” and “b” are reserved for the basic paste representatives diagnostic of each series as described previously. The remaining letter designations note the addition of other temper constituents to these basic paste categories as described below: (c) Crushed Arenite (d) Granule to Pebble Sized Angular/Subangular Quartz Sand and/or Crushed Quartz (e) Crushed and Ground Indurated Granite (f) Medium to Granule Sized Angular/Subangular Feldspar Sand More than one of these temper constituents was sometimes present in a single sherd. In these cases, the constituent designations would include combinations of letter designations (e.g. “c/e” or “e/f”, etc.). During the course of the analysis it also became apparent that some of the pottery exhibited extremely hard pastes, much harder, in fact, than the typical paste previously defined for each series. A hardness scale consisting of three attribute states was developed to describe this variation. These consisted of: (1) friable, (2) compact and (3) hard. The attribute state of “friable” corresponded to sherds that were easily crushed on their edges with only minimal application of pressure with the thumb. A “compact” state was identified in instances where moderate pressure with the thumb was required to break the edge of the sherd. Finally the “hard” state was reserved for those sherds whose edges were not easily broken even with extreme thumb pressure. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 271 Chapter 7. 31HK2521 Figure 83. Grinding and Hammer Stones, Site 31HK2521. A: Abrading/Grinding Stone (Ceramic Vessel Shaper?) (a2039), B-E: Hammer Stones (a1357, a1799, a2053 and a668 respectively). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 272 Chapter 7. 31HK2521 Figure 84. Hammer/Anvil Stones, Site 31HK2521. A-B: (a825 and a2286 respectively). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 273 Chapter 7. 31HK2521 Figure 85. Miscellaneous Stone Tools, Site 31HK2521. A: Nutting Stone (a1133), B-D: Pebbles (a470, a917 and a828 respectively), E: Steatite Vessel Fragment (a492), F: Worked Schist (a1150). Generally, pottery achieves greater hardness when one or more of three conditions prevail (Shepard 1954:114). These are: (1) the pottery is made from low-fusing, dense-firing clay, (2) the pottery is fired at a high temperature and/or (3) the pottery is fired in an atmosphere promoting vitrification. It is common to observe in the local ceramic sequences of the Southeast, a progression from friable ceramics to much harder, more durable pottery in the later stages of cultural development. Clearly, Mississip- pian potters had developed technological strategies to produce extremely hard, sometimes vitrified pottery and it is quite possible that this advancement was meant to adapt to a more sedentary, agricultural life way in which durability and longevity of pots would have been advantageous. Similar adaptive technologies may have been incorporated into sedentary, agricultural societies peripheral to the Southeast Appalachian Mississippian culture as well, and for the same reasons. Currently, a sophisticated method for Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 274 Chapter 7. 31HK2521 measuring hardness in “primitive” pottery does not exist due to its porosity and heterogeneity of paste composition. The “thumb” test, although somewhat imprecise, provides a relatively objective and empirical basis for identifying relative hardness in precontact pottery assemblages. For an extended discussion of hardness in precontact ceramics see Chapter 4. This method of analysis resulted in segregating extremely hard or vitrified pottery that would have traditionally been placed in the Cape Fear (sandtempered), Hanover (sand- and grog-tempered) and Yadkin (crushed rock- and granule sand-tempered) series. As a means of preserving data to test the “hardness trajectory” hypothesis, these hard examples were separated into new series variants distinguished by the Roman numeral suffix “III.” Thus, extremely hard variants of the series are respectively referred to here as Hanover III, Cape Fear III and Yadkin III. If the efficacy of the hypothesis is later confirmed, then it would be appropriate to assign new series names to the Cape Fear III and Yadkin III variants since they are generally viewed as Middle Woodland series. Description of the sherd collection is presented below, organized by series. New River Series New River series ceramics are associated with the Early Woodland period in the region. The series was originally defined by Loftfield (1976) who characterized it as a coarse sand-tempered pottery of compact composition and gritty consistency characterized by cord marked, fabric impressed, thong simple stamped and plain exterior surface treatments. Correlated series of the Early Woodland period include Lenoir (Crawford 1966) and Deep Creek (Phelps 1983). The New River series in the Fort Bragg area is poorly understood. Herbert (2003:156) indicates that both net-impressed and parallel/overstamped cord marked types are recognized, while Cable and Cantley (2005a, b) have also recognized simple stamped, fabric impressed and plain surface treatments. The predominant classificatory criterion used to identify the series is an abundance of quartz sand temper as Loftfield (1976) prescribed. Herbert (2003:156) proposes that an arbitrary cut-off at about 15% sand temper density should be used to distinguish Cape Fear (< 15%) and New River (>15%) in the Fort Bragg area. Cable and Cantley (2005a, b) have recognized two major paste variants in Fort Bragg assemblages. These are variants Ia and IIa. Variant Ia consists of sherds dominated by abundant medium to very coarse subangular to rounded quartz sand grains in a gritty, friable paste matrix. Variant IIa contains a fine, compact, relatively harder paste matrix with abundant to moderately dense medium to very coarse quartz sand inclusions. Variant IIa may correlate with the traditionally defined Cape Fear series. Feldspar sand is sometimes found in variant Ia paste, and it is designated New River If. Thirty-two New River series sherds were recognized in the collection from 31HK2521 (Table 70). The New River Ia sample consisted of eighteen sherds. Recognized surface treatments consisted of four net impressed, one cord marked and two plain sherds. The knotted Net Impressed sherds derived from two vessels (Figure 86:A-B), both of which displayed the knotted type of net impressing. The cord marked sherd displayed parallel stamping and the average cord diameter was 1.60 mm (Figure 86:D). This particular specimen contained a moderate amount of granule sized subangular sand, which was a common occurrence in the New River collection from the site. Plain surfaces were identified on one body and one base sherd. In addition, a single New River sherd with subangular feldspar (If) sand was identified. It displayed a net impressed surface, but the net was pressed deep into the clay obscuring knot impressions (Figure 86:C). Vessel wall thickness in the New River I sample was variable, ranging from 6.13 to 13.84 mm with a mean of 8.06 mm (SD=3.28 mm). The New River IIa sample included cord marked (n=4), fabric impressed (n=2), check stamped (n=1) and plain (2) surface treatments. The sample totaled 13 sherds in all. The cord marked sherds derived from three separate vessels, all of which displayed parallel cord orientations (Figure 70:E). Impressions ranged between 1.15 and 1.64 mm in diameter. One fabric impressed sherd exhibited medium weft cordage interwoven around a rigid warp (Figure 86:F). A single checked stamped sherd was assigned to the New River series (Figure 86:H). Checks were square and averaged about 4.5 mm on Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 275 Chapter 7. 31HK2521 a side. It was identical to the check stamped pottery of the Hanover I subseries, but lacked grog tempering. A single rim was recovered in the New River II collection, a direct, simple form with a rounded lip, suggestive of a hemispherical bowl (Figure 87). The advanced character the identified surface treatments indicates that New River II younger than New River I and perhaps correlated with the Cape Fear series. The material was not identified as Cape Fear, however, because sand density and relative coarseness was not easily discriminated from the New River I sample. The two subseries were distinguished solely on the basis hardness. New River I was much more friable and brittle, while New River II paste was more compact. Hanover Series South (1960:16-17; 1976: 28) defined the Hanover series from surface collections made in 1960 on sites located between Wilmington, NC and Myrtle Beach, SC. Pottery affiliated with this series is tempered with aplastic clay lumps, most of which consist of grog fragments from crushed sherds. A correlate on the north coastal plain of North Carolina is referred to as the Carteret series (Loftfield 1979:154157). Cord marked and fabric impressed exterior surface treatments were included in the original series description, but other grog-tempered types have been recognized subsequently. Herbert (2003:75) has identified minority proportions of check stamped, plain (smoothed), net impressed, simple stamped, punctate, and random straw bundle punctate surface treatments. The latter two types are traditionally subsumed under the Refuge series on the central South Carolina coast where grog/clay temper occurs as a major temper variant (see Cable 2002: 195-201). A great deal of variability in paste composition has been noted in Hanover assemblages, primarily related to differences in quartz sand contributions, clay versus grog distinctions, and the density and size of clay/grog particles (see Cable et al. 1998; Herbert 2003; 74-75). Herbert (2003:191-192) recognizes two broad variants that he believes are sequential. The earliest variant, Hanover I, is primarily sandtempered with minor amounts of finely crushed grog. Surface treatments attributed to Hanover I pottery consist of cord marking, fabric impressing and check stamping. He estimates an age range of AD 400-800 for Hanover I pottery. The later variant, Hanover II, is inferred to correlate with a Late Woodland temporal range (AD 800 through 1500). Hanover II pottery is characterized by pastes with only small amounts of sand and abundant grog particles. The dominant exterior surface treatment of the later variant is fabric impressing. Cable and Cantley (2005a, b) have followed Herbert (2003) in describing Hanover series ceramics in the Fort Bragg area. However, in an attempt to more fully describe the temper and paste variation, these basic groupings have been expanded to include sub-variants. As research has progressed, the number of sub-variants has increased to accommodate an ever-expanding range of temper constituents and paste qualities. There are three major paste variants found in Hanover pottery at Fort Bragg based on relative hardness. Hanover I is characterized by a friable paste, Hanover II exhibits a compact paste and Hanover III contains a hard to vitrified paste. Within each of these relative hardness groupings are two primary temper constituent designations. Designation “a” refers to pastes with sparse to moderate amounts of fine to medium quartz sand and abundant medium to large grog particles. Designation “b” includes pastes with moderate to abundant amounts of medium to coarse quartz sand and sparse to moderate densities of medium to large grog particles. Other temper constituents are sometimes added to these basic paste variants. These include crushed arenite (“c”), granule to pebble sized angular to subangular quartz sand and, rarely, crushed quartz (“d”), crushed and ground indurated granite (“e”) and medium to granule sized subangular feldspar sand (“f”). Crushed arenite and indurated granite almost exclusively occur in Hanover III paste. Indurated granite temper was originally identified as sandstone conglomerate because of similar texture characteristics, but thin-sectioning of rock samples indicated that most of the temper in this category corresponds to a fine grained, indurated granite (see Appendix K). Some of the temper in this category may actually represent sandstone conglomerate, but this was not differentiated in the ceramic analysis due to the late submittal of samples. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 276 Chapter 7. 31HK2521 Table 70. New River Ceramic Inventory, 31HK2521. INVENTORY OF NEW RIVER SERIES CERAMICS: 31HK2521 CERAMIC TYPE VESSEL PORTION BASE BODY RIM RESIDUAL GRAND TOTAL New River Ia Cord Marked 1 New River Ia Indet. 9 New River Ia indet. decorated 1 1 New River Ia Net Impressed 4 4 1 2 New River If Net Impressed 1 1 New River IIa Check Stamped 1 1 New River IIa Cord Marked 4 4 New River IIa Fabric Impressed 2 2 New River IIa indet. 3 3 New River IIa Indet. stamped 1 1 New River IIa Plain 1 1 29 1 New River Ia Plain GRAND TOTAL 1 1 (1) Hanover I. Seventy sherds assignable to the Hanover I subseries were identified in the analysis (Table 71). Check stamped and linear check stamped surface treatments dominate, comprising 38.8 percent of the collection. Fabric impressed and cord marked are equal subdominants, comprising respectively 28.6 and 26.5 percent of the collection. Minority types included net impressed (4.1 percent) and plain (2 percent). Hanover Ib paste makes up 77 percent (n=54) of the collection. This paste is nearly identical to New River I and II. Subangular quartz sand is abundant and ranges from medium to very coarse in grain size. Grog tempering is variably dense and grog particles are small to medium in size. Hardness ranges from friable to slightly compact. Hanover Ia paste is more compact, quartz sand inclusions are less dense and grog particles are larger, although of only moderate density. Hanover Ia paste makes up 14.3 percent (n=10) of the collection. Finally, Feldspar sand in combination with quartz sand was found in 8.6 percent (n=6) of the collection. This is referred to as Hanover If paste. Check stamping (Figure 88:A-G) ranges from faint to clear land impressions and from carefully 1 1 10 2 1 32 applied check panels to over-stamped areas. Linear check stamping, which consists of repeated parallel arrangements of paired longitudinal lands with series of finer perpendicular transverse lands running between the framing lands (Caldwell and Waring 1968:116), was not common in the Hanover I sample. One specimen of Hanover Ib Linear Check Stamped was identified (Figure 88:G). The original stamp pattern was over-stamped at an oblique angle. Cord marked patterns (Figure 88:H-K) consisted of both parallel (n=6) and perpendicular/oblique (n=4). Cord impression diameter ranged between 0.75 and 2.32 mm with a mean of 1.19 mm (SD=0.48 mm). The fabric impressed specimens were dominated by flexible warps with coarse weave (Figure 88:L). Eightynine percent (n=8) of the recognizable patterns were made with flexible fabric, while one specimen displayed a rigid/coarse weave pattern. Net impressed patterns appear to have been made with knotted net, but the impressions are too deep on the two examples to confirm this (Figure 88:M). Three rims were present in the collection (Figure 87). These consisted of a restricted thickened rim (p114) with a flat lip (Figure 88:E), suggestive of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 277 Chapter 7. 31HK2521 Figure 86. New River Series Ceramics, Site 31HK2521. A-B: New River Ia Net Impressed (p83 and p151 respectively), C: New River If Net Impressed (p1414), D: New River Ia Cord Marked (p123), E: New River IIa Cord Marked (p2101), F: New River IIa Fabric Impressed (p604), G: New River IIa Plain (a1751), H: New River IIa Check Stamped (p655). a constricted orifice, short-necked jar and two direct simple rims. One of these (p627) was thinned by an indentation to the outside of the vessel wall near the lip. The direct simple rims appear to represent straight-sided, open mouth jars. Vessel wall thickness for the Hanover I sample was 8.37 mm (SD=1.04 mm). (2) Hanover II. One hundred fifty-six sherds assignable to the Hanover II subseries were identified in the analysis (Table 72). Cord Marked and Check Stamped/Linear Check Stamped Check surface treatments dominate, comprising respectively 38.3 (n=41) and 34.6 (n=37) percent of the collection. Fabric impressed surface treatments made up Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 278 Chapter 7. 31HK2521 New River Rims Hanover I Rims Cape Fear Rims p627 p1751 Direct Simple/ Indeneted Direct Simple p525 p1893 Direct Simple p114 Restricted Thickened Hanover II Rims p2301 p576 p2583 p1865 Direct Simple/ Indented p1596 Direct Thickened Restricted Thickened p144 p2015 p1651 Direct Simple p1629 Restricted Simple Hanover III Rims p2010 p2341 p45 p2065 Restricted Bevelled p46 p2107 p489 Direct Thinned p2135 Direct Thickened P1895 p625 Incurved Simple 0 p2124 1 inch 0 p1978 2 cm p47 Direct Simple Yadkin III Rims p735 p2358 p1198 p1644 p733 Direct Thinned p2198 p793 p1905 Recurved Flared p2236 p2372 Restricted Bevelled Direct Bevelled p1904 p2125 Direct Simple Figure 87. Rim Profiles, Site 31HK2521. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 279 Chapter 7. 31HK2521 25.2 percent (n=27) of the recognizable surface treatments. Minority types included incised over unidentified textile impressed (0.9 percent) and plain (0.9 percent). Hanover IIa paste (larger grog particles and more sparse quartz sand) makes up 59 percent (n=92) of the collection, while Hanover IIb (smaller grog particles and more abundant quartz sand) comprises only 16 percent (n=25) of the sample. Hanover IId paste, which is identical to Hanover IIa paste with the addition of moderate densities of granule- to pebble-sized subangular white quartz sand and crushed quartz, constitutes a secondary paste dominant, representing 25 percent (n=39) of the sample. This paste is nearly identical to New River I and II. Subangular quartz sand is abundant and ranges from medium to very coarse in grain size. Hanover II paste is more compact than the paste that characterizes the more friable Hanover I subseries. Check stamped surfaces exhibit the same range and quality of application described for the Hanover I subseries (Figure 89). However, Linear Check Stamping is more abundant in the Hanover II sample (Figure 89:I-J). It makes up 18.9 percent of the Hanover II check stamped collection, while it comprises only 5.3 percent of the Hanover I check stamped sample. Cord marked sherds (Figure 90:CD) are dominated by perpendicular/oblique stamp patterns, which make up 74.3 percent of the recognizable patterns. Cord impression diameter ranged between 0.25 and 1.64 mm with a mean of 0.97 mm (SD=0.30 mm). Eighty percent of the recognizable patterns on the Fabric impressed specimens are represented by rigid warps and medium weave fabrics (Figure 90:E-F). Two Hanover II sherds displayed both check stamped and fabric impressed surface treatments (Figure 89:C). One Hanover II sherd (Figure 90:J) exhibited an incised line over a faint and unrecognizable textile impression. The cross-section of the incised line was u-shaped. Incising of a similar structure over fabric and net impressing begins around A.D. 800 in the Uwharrie phase of the southern Piedmont (Ward and Davis 1999:100). Five different rim types were recognized in the Hanover II sample (Figure 87). Restricted thickened rims with flat lips were found respectively on a check stamped sherd and a cord marked sherd. The checked stamped rim exhibits a short neck on a constricted orifice jar of globular body shape (Figure 89:B). The cord marked rim does not exhibit a neck, suggesting a restricted orifice jar with a body shaped like a truncated oval. Two rims have thickened rims and direct orientations. Both sherds show slightly rounded bodies below the lip, suggesting that they represent hemispherical or semi-hemispherical jars or bowls. One of these was check stamped (Figure 89:A), while the other had an indeterminate surface treatment. One rim appears to represent a relatively large jar with a slightly restricted orifice. The orientation was difficult to read on this sherd and it is possible that it represents recurved, slightly flared rim jar. The sherd exhibits a cord marked exterior (Figure 90:B). The final four specimens are simple direct rims suggestive of straight-sided jars or hemispherical bowls. Three of these have fabric impressed exteriors (Figure 90:E-F and H), while one (p576) displayed a check stamped treatment. Two of these rims had thinning indentations on the interior of the vessel just below the lip. A single conoidal base sherd and two sherds that appear to represent rounded or flattened bases were identified in the Hanover II sample, suggesting that both pointed and flattened bases were represented in the vessel assemblage. Vessel wall thickness for the Hanover II sample was 8.32 mm (SD=1.50 mm), which is nearly the same as the Hanover I sample. There is some variation in the mean vessel wall thickness of the three major surface treatments, however. Check stamped and linear check stamped are thickest, with a mean of 8.55 mm (SD=1.02 mm). Cord marked and fabric impressed sherds are thinner, averaging respectively 7.87 mm (SD=0.73 mm) and 7.90 mm (SD=1.30 mm). This may simply reflect functional differences, but the association of surface treatments by Hanover II paste sub-variant exposes what might represent a fine-grained seriation within the Hanover II ceramic phase (Table 73). The stronger association of Hanover IIb paste with check stamping suggests that this pottery type’s popularity may have waned through the phase. Fabric impressed surface finishes are primarily represented by Hanover IIa paste, while cord marking displays a much stronger association with Hanover Id paste, which contains significant amounts of granule- and pebble sized white subangular quartz. These associations might suggest that the popularity of both fabric impressing and cord marking increased through the phase. Fabric impressing occupies an intermediate position in this seriation. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 280 Chapter 7. 31HK2521 Table 71. Hanover I Ceramic Inventory, 31HK2521. INVENTORY OF HANOVER I SUB-SERIES CERAMICS: 31HK2521 CERAMIC TYPE VESSEL PORTION BASE BODY RIM RESIDUAL GRAND TOTAL Hanover Ia Check Stamped 1 1 Hanover Ia Cord Marked 4 4 Hanover Ia Fabric Impressed 1 1 Hanover Ia Indet. 2 Hanover Ia Plain Hanover Ib Check Stamped 17 Hanover Ib Cord Marked Hanover Ib Fabric Impressed 1 3 1 1 2 19 8 8 13 13 Hanover Ib Indet. 9 Hanover Ib indet. stamped 2 2 Hanover Ib Lin. Check Stamped 1 1 Hanover Ib Net Impressed 1 1 Hanover If Cord Marked 1 1 Hanover If Indet. 3 Hanover If Net Impressed 1 GRAND TOTAL 2 (2) Hanover III. Three hundred and ninety sherds assignable to the Hanover III subseries were identified in the analysis (Table 74). Fabric impressed surface treatments overwhelmingly dominated the collection, comprising 86.3 percent (n=321) of the sample. Cord marking was present on 10 percent (n=37) of the collection, while minor surface treatments consisted of linear check stamped, plain, dowel impressed and wide simple stamped. These minor surface treatments together made up 3.8 percent of the sample. These percentages are misleading, however, because approximately two-thirds (n=260) of the fabric impressed sherds derived from only three partial vessel aggregations (V150, V154 and V155) recovered from test unit excavations. Fifty-eight vessels were identified 62 1 1 10 4 1 3 3 70 in the collection and if this data set is used to characterize the relative popularity of surface treatments in the Hanover III collection, a wholly different picture of proportional relationships emerges. Fifty-four of the vessels exhibited recognizable surface treatments. Fabric impressed exterior surfaces still dominate the collection, comprising 48.2 percent (n=26 vessels). However, cord marked exterior surfaces were found on 18 vessels, which represents 33.3 percent of the recognizable surface treatments on vessels. Dowel impressed surfaces occur on 5.6 percent (n=3 vessels), plain surfaces were recorded on 7.4 percent (n=4 vessels), wide simple stamped was identified on 3.7 percent (n=2 vessels) and linear check stamping was found on 1.9 percent (n=1 vessel) of the collection. Plain sherds tend to be prevalent on the lower Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 281 Chapter 7. 31HK2521 Figure 88. Hanover I Series Ceramics, Site 31HK2521. A-F: Hanover Ib Check Stamped (p364, p1702, p117, p2438, p114 and p458 respectively), G: Hanover Ib Linear Check Stamped (p1740), H-K: Hanover Ib Cord Marked (p1092, p1093, p277 and p1097), L: Hanover Ib Fabric Impressed (p617), M: Hanover If Net Impressed (p631). portions of decorated vessels, so the plain percentages may be over-represented. Fabric impressed weave patterns were evenly divided in the vessel collection (Figures 91-93). Rigid warps occurred on 55 percent of the recognizable patterns while flexible warps occurred on 45 percent. Wefts were skewed toward medium and fine weaves. Each of these textures occurred on 40.1 percent of the recognizable patterns on vessels. The remaining 18.2 percent of the vessel assemblage was characterized by coarse wefts. Stamp patterns on cord marked vessels (Figures 94 and 95) consisted of 50 percent (n=9 vessels) perpendicular/oblique, 33.3 percent parallel Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 282 Chapter 7. 31HK2521 Table 72. Hanover II Ceramic Inventory, 31HK2521. INVENTORY OF HANOVER II SUB-SERIES CERAMICS: 31HK2521 CERAMIC TYPE VESSEL PORTION BASE BODY Hanover IIa Check Stamped 9 Hanover IIa Cord Marked Hanover IIa Fabric Impressed NECK 1 RIM RESID. 2 12 12 1 GRAND TOTAL 12 19 3 23 Hanover IIa Fab.Imp./Check St. 2 2 Hanover IIa Incised/indet. Textile 1 1 Hanover IIa indet. 1 17 Hanover IIa indet. decorated 7 6 13 Hanover IIa indet. Textile 2 2 Hanover IIa Lin. Check Stamped 4 4 Hanover IIa Plain 4 1 Hanover IIb Check Stamped 22 1 11 1 1 13 Hanover IIb Fabric Impressed 3 Hanover IIb indet. 1 Hanover IIb indet. decorated 4 4 Hanover IIb Lin. Check Stamped 3 3 Hanover IId Check Stamped 3 3 Hanover IId Cord Marked 3 1 27 2 2 29 Hanover IId Fabric Impressed 1 1 Hanover IId indet. 2 2 Hanover IId indet. decorated 1 1 Hanover IId indet. textile 3 3 GRAND TOTAL 10 and 16.7 percent with vessels exhibiting both stamp patterns on individual sherds. Cord widths averaged 1.20 mm in diameter (SD=0.37 mm) and ranged between 0.79 to 2.27 mm, which may represent a slight increase in cord diameters from the Hanover II sample mean of 0.97 mm. A small number of sherds displayed an impression that appears to have been made with a cord wrapped dowel or by the edge 131 2 9 4 156 of a cord wrapped paddle (Figure 95:E). These are referred to here as dowel impressed to differentiate them from classic cord marked treatments and from rigid warp fabric impressed patterns, which they resemble on a superficial basis. A single linear check stamped (Figure 96:A-C) vessel was assigned to the Hanover III subseries. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 283 Chapter 7. 31HK2521 Figure 89. Hanover II Check Stamped, Site 31HK2521. A-B: Hanover IIa Check Stamped (p1596 and p1865 respectively), C: Hanover IIa Check Stamped/Fabric Impressed (p899), D: Hanover IIa Check Stamped (p414), E: Hanover IId Check Stamped (p1749), F-H: Hanover IIa Check Stamped (p620, p2234, and p650 respectively), I-J: Hanover IIa Linear Check Stamped (p1240 and p1246 respectively). Finally, three sherds from two separate vessels displayed wide, parallel lands that appear to have been made from carved paddles (Figure 96:D-F). The lands are squared off creating a design resembling complicated or check stamped exterior surfaces. The sherds were classified as wide simple stamped, however, because finer, transverse lands that would have completed a linear check stamped pattern were not present and only parallel lands were observed. A visually similar type of simple stamping that Boudreaux (2005:46, 86-87) refers to as Large Simple Stamped occurs in Leak phase assemblages in the Upper Pee Dee River valley by about AD1300. Given the context of the simple stamped sherds from 31HK2521, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 284 Chapter 7. 31HK2521 Figure 90. Hanover II Cord Marked, Fabric Impressed and Incised, Site 31HK2521. A-D: Hanover IId Cord Marked (p1631, p1629, p2293 and p1717 respectively), E-I: Hanover IIa Fabric Impressed (p414, p1651, p1715, p2015 and p1053 respectively), J: Hanover IIa Incised Over/Indeterminate Textile Impressed. a date of post-AD 1300 would seem too late. The rim form associated with Specimen p489 (Figure D-E), in fact, is similar to some of the rims affiliated with Hanover I and II check stamped types. It has a thickened, flat rim with a direct or straight neck orientation. South (2005:28) indicates that some Gaston Simple Stamped sherds from the Roanoke Valley exhibit similarly wide, squared-off lands that appear to have been made with a carved paddle, although the majority of the type was made using a thong or sinew wrapped paddle. The rim treatments on Gaston Simple Stamped, however, exhibit a wide range of Lamar-like decorative elements, including incidental hollow reed punctations and folded and applique rim strips and much of the Gaston series pottery clearly occurs after AD1600. However, the transition from the earlier cord marked dominated assemblages (i.e. Clements) to Gaston is not well understood and it is possible that simple stamped pottery arose much earlier in the Roanoke Valley than is currently accepted. Herbert’s (2003:194-201) review of the Cashie and Townsend series in the North Carolina Tidewater region indicates a similar uncertainty over the earliest appearance of simple stamped types, but it is possible that they emerge in the early part of the Late Woodland period. There exist a wide range of temper types in the Hanover III collection from 31HK2521. The most interesting development in this regard is the addition of large pieces of crushed arenite and especially crushed and ground indurated granite. Arenite is commonly found on precontact sites on Fort Bragg and although most of it was used for hearth Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 285 Chapter 7. 31HK2521 Table 73. Percentage of Surface Treatment Types by Hanover II Paste Subvariants, 31HK2521.* PERCENTAGE OF SURFACE TREATMENTS BY HANOVER II PASTE SUBVARIANTS SURFACE TREATMENT Check Stamped Fab. Imp./Check St. PERCENT HANOVER II SUBVARIANTS Hanover IIa Hanover IIb Hanover IId 45.7 45.7 8.6 100.0 Fabric Impressed 85.7 Cord Marked 29.3 GRAND TOTAL 53 GRAND TOTAL 35 1 10.7 19 3.6 28 70.7 41 33 105 * Data Generated from Vessel Counts Rather than Raw Sherd Frequencies. rocks and heating stones, a secondary use appears to have been to produce temper. Indurated granite is only rarely found in naturally occuring rock form, but it represents a major source of temper. Along with grog temper, approximately 38.2 percent of the Hanover III vessel sample contains crushed arenite, crushed and ground indurated granite, or both. Particles of this material range from medium to granule or pebble in grain-size, similar in visual effect and probably technological function to grog. Most of these specimens also include sparse to moderate amounts of granule to pebble sized subangular white quartz sand. The latter material occurred rarely in isolation from arenite and inurated granite particles (n=2 vessels) and these were assigned paste category IIId. Breaking out sherds with crushed rock and granule sized subangular quartz from the more common Hanover IIIa and IIIb paste variants, demonstrates an almost identical representation of surface treatments (Table 75). Fabric impressed exteriors were most abundant in each grouping, ranging between 46.9 and 52.2 percent of the vessel sample. Cord marking constituted the major subdominant, ranging between 31.3 and 34.8 percent of the vessel sample. Clearly, the two groupings could be drawn from the same, contemporaneous population of vessels, although slightly higher proportions of fabric impressing and the absence of linear check stamping in the crushed rock grouping might evidence a fine-grained seriation within the Hanover III phase. Drawing on ceramic patterns from neighboring regions, it is likely that the Hanover III phase post-dates AD 800. Two lines of evidence have already been mentioned in this regard. The presence of incising in the Hanover II sample suggests linkages with the Uwharrie phase of the southern Piedmont, which begins at about AD 800 (see Ward and Davis 1999:100). Second, simple stamping, which occurs in low proportions in Hanover III, emerges after AD 800 in the Tidewater region of North Carolina and in the Roanoke River Valley and similar carved paddle simple stamped with broad grooves and lands is found in the Leak phase of the southern Piedmont Mississippian assemblages around Town Creek after AD 1300. Another line of evidence that supports this chronological position concerns the finishing characteristics on vessel interiors. Perhaps the most striking and diagnostic technique is serrated scraping, which was found on the interior of a Hanover IIIe Fabric Impressed partial vessel aggregation (see Figure 93:F). One hundred and ninety-nine sherds of this vessel aggregate (V155) were recovered during the investigation, almost all of which exhibited this scraping technique. Several sherds near the base of the vessel, however, were smooth. This particular type of interior finishing is characteristic of Uwharrie, Dan River and Gaston phase ceramics in the eastern Piedmont (South 2005:28; Ward and Davis 1999:101, 106). This technique is diagnostic of pottery postdating AD 800 in these regions. Serrated scraping is not common in Fort Bragg assemblages, but a Hanover IIIa sherd of indeterminate exterior surface treatment was also finished with serrated scraping (Figure 31:E). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 286 Chapter 7. 31HK2521 Table 74. Hanover III Ceramic Inventory, 31HK2521. INVENTORY OF HANOVER III SUB-SERIES CERAMICS: 31HK2521 CERAMIC TYPE VESSEL PORTION BASE BODY RIM RESID. GRAND TOTAL Hanover IIIa Cord Marked 17 1 Hanover IIIa Fabric Impressed 58 3 Hanover IIIa indet. 1 1 Hanover IIIa indet. decorated 4 Hanover IIIa indet. textile 1 Hanover IIIa Linear Check Stamped 4 4 Hanover IIIa Plain 1 1 Hanover IIIa Wide Simple Stamped 1 Hanover IIIb Cord Marked 6 Hanover IIIb Dowell Impressed 1 1 2 Hanover IIIb Fabric Impressed 4 2 6 1 2 Hanover IIIb indet. 1 18 2 63 2 4 1 2 1 2 6 Hanover IIIb indet. decorated 2 2 Hanover IIIb indet. textile 2 2 1 2 1 1 Hanover IIIb Plain 1 Hanover IIIc Fabric Impressed Hanover IIIc/e Dowell Impressed 1 Hanover IIIc/e Fabric Impressed 1 4 Hanover IIId Fabric Impressed 31 4 2 33 Hanover IIId indet. 1 1 Hanover IIId Wide Simple Stamped 1 1 Hanover IIIe Cord Marked Hanover IIIe Fabric Impressed 12 13 202 214 Hanover IIIe indet. 2 2 Hanover IIIe indet. decorated 1 1 Hanover IIIe Plain 1 1 GRAND TOTAL 12 1 15 359 14 2 390 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 287 Chapter 7. 31HK2521 Figure 91. Hanover III Fabric Impressed, Site 31HK2521. A-B: Hanover IIIa Fabric Impressed (p46 and p45 respectively), C-D: Hanover IIIb Fabric Impressed (p2107 and p2124 respectively), E: Hanover IIIa Fabric Impressed (p47), F: Hanover IIIe Fabric Impressed (p1073), G: Hanover IIIa Fabric Impressed (p2588), H: Hanover IIIe Fabric Impressed (p1112), I Hanover IIIb Fabric Impressed (p176), J: Hanover IIIe Fabric Impressed (p1998), K: Hanover IIIa Fabric Impressed (p346), L: Hanover IIIb Fabric Impressed (p2238). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 288 Chapter 7. 31HK2521 Figure 92. Hanover IIId Fabric Impressed (p1978), Vessel 154, Site 31HK2521. Two other, more common, interior surface finishes were observed on Hanover III sherds that also appear to have diagnostic dating potential. One finish can be described as “floated.” The resultant surface is lustrous and appears to have been formed by polishing with the hands or with an implement such as a polishing stone while the pot was still wet. This brought suspended clay particles from the paste up onto the surface, creating a “floated” surface resembling a thin slip. Vessel interiors were variably smoothed, some approximating a regularized contour (Figure 93:D, 95:F), while lumps of temper and surface undulations protruded on others (Figure 95:B, D). Fingerprints are preserved in two places on the interior of a Hanover IIIa Cord Marked rim (Figure 95:A-B), a clear indication that the interior was wet during the application of the interior surface technique. Floating appears on Hanover II interiors as well, but in these cases the surfaces were polished to a luster and the interiors were subject to dissolving in fine particles with exposure to fluids. The floated surface on the single Linear Check Stamped sherd with Hanover III paste is not polished, although its interior was smoothed to a regular contour (Figure 96:B). Fine scraping stria can be seen on this specimen, suggesting that it was finished after it had partially dried. Phelps and Heath (1998:12) note that the interiors of most Cashie I ceramics are floated, which results visually in subduing the effects of protruding granule- and pebble-sized inclusions. Radiocarbon Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 289 Chapter 7. 31HK2521 Figure 93. Hanover IIIa Fabric Impressed Interior Finish Treatments, Site 31HK2521. A-B: Hanover IIIa Fabric Impressed (A: Exterior, B: Interior, Ground Smooth), C-D: Hanover IIIe Fabric Impressed (C: Exterior, D: Interior, Floated), E-F: Hanover IIIe Fabric Impressed (E: Exterior, F: Interior, Serrated Scraped). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 290 Chapter 7. 31HK2521 Figure 94. Hanover III Cord Marked, Site 31HK2521. A-B: Hanover IIIa Cord Marked (p408 and p600 respectively), C-D: Hanover IIIe Cord Marked (p1241 and p2342 respectively), E-F: Hanover IIIa Cord Marked (p583 and p1987 respectively), G: Hanover IIIe Cord Marked (), H-I: Hanover IIIa Cord Marked ( p582 and p575 respectively), J: Hanover IIIb Cord Marked (p1646). dates for Cashie I place it between AD 985 and 1480 at the 1-sigma range of uncertainty (Eastman 1994a, b; Phelps and Heath 1998:6). South (2005:28) found that Gaston phase interiors were also sometimes “smoothed by use of a smooth stone or other object, so that the temper is pressed into the paste.” Approximately 33 percent of the identified Hanover III vessels display semi lustrous, floated interior surfaces. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 291 Chapter 7. 31HK2521 Figure 95. Hanover III Cord Marked and Paddle Edge Cord Marked Interior Finish Treatments, Site 31HK2521. A-B: Hanover IIIa Cord Marked (p2135, A: Exterior, B: Interior, Floated), C-D: Hanover IIIe Cord Marked (p1241, C: Exterior, D: Interior, Floated), E-F: Hanover IIIb Dowell Impressed/Paddle Edge Stamped (p625, E: Exterior, F: Interior, Floated). Another interior finish that appears to be characteristic of the Hanover III collection is referred to as “ground smooth.” The surfaces are extremely regular and exhibit very few undulations, but spaces are not always filled between temper particles (Figure 93:B). Instead, less resistant temper particles such as indurated granite and grog appear to have been ground into a flat contour. This suggests that interior irregularities were rubbed off or ground with a coarse or gritty implement such as an abrader to create a smooth but non-lustrous surface. Scraping marks are generally not observed on these interiors, further suggesting that grinding may have been aided by the addition of water. Shepard (1954:66) describes a technique employed by the Tzintzuntzan of Mesoamerica that might be analogous: “Next a ‘polishing’ stone of coarse volcanic rock is dipped in water and rubbed over the outside to level ridges and smooth the surface. The finishing process is finally completed by rubbing the surface with a wet rag.” Approximately 18 percent of the vessel collection exhibited ground smooth interior finishes. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 292 Chapter 7. 31HK2521 Figure 96. Hanover III Check Stamped and Simple Stamped, Site 31HK2521. A-B: Hanover IIIa Linear Check Stamped (p657, A: Exterior, B: Interior, Floated), C: Hanover IIIa Linear Check Stamped (p659), D-E: Hanover IIIa Wide Simple Stamped (p489, D: Exterior, E: Interior), F: Hanover IIId Wide Simple Stamped (p2585). There are artifacts in the stone tool collection from the site that could have served as pottery manufacturing implements. The three quartz pebbles (Figure 85:B-D) are of a size and texture to have served efficiently as polishing stones. Moreover, the quartz/quartzite grinding stone (Figure 83:A) is of the consistency to have been used as a surface grinder. The concave edge on the tool would have been especially functional as an exterior wall smoother and shaper on wet pots and a similar tool with a convex edge would have provided an excellent edge for shaping wet interior vessel walls and as a grinder to smooth dried vessel walls. No direct evidence of pottery manufacture such as kilns or manufacturing rejects was recovered from the site, but the fact that locally available arenite and possible indurated granite (stream transported?) were used as tempering agents on a number of the pots provides some indirect support for this hypothesis. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 293 Chapter 7. 31HK2521 Table 75. Percentage of Surface Treatment Types by Hanover III Paste Subvariants, 31HK2521.* REPRESENTATION OF HANOVER III PASTE VARIANTS AND SURFACE TREATMENTS: 31HK2521 CORD MARKED DOWEL IMPRESSED WIDE SIMPLE STAMPED 58.2 46.9 31.3 6.3 3.1 IIIc, IIIc/e, IIId, IIIe 41.8 52.2 34.8 4.3 4.3 GRAND TOTAL 100.0 PLAIN FABRIC IMPRESSED IIIa/b HANOVER III PASTE VARIANT GROUPS LINEAR CHECK STAMPED OVERALL PERCENT SURFACE TREATMENT TYPE 3.1 9.4 32 4.3 23 GRAND TOTAL 55 * Data Generated from Vessel Counts Rather than Raw Sherd Frequencies. Five different rim groupings were recognized in the Hanover III sample (Figure 87). Most of the rims had direct orientations from straight-sided or very slightly recurved vessel wall profiles. These include direct thickened, direct simple and direct thinned rim groups. The thickened rims expand on both interior and exterior sides and represent large open mouth jars. Exterior surface treatments of this rim group included wide simple stamped (Figure 96:E-F) and perpendicular cord marked (Figure 95:A-B). Parallel cord marking impressions were applied to the top of the lip on the cord marked specimen. Direct simple rims derive from smaller vessel forms that appear to represent hemispherical bowls or small, straight-sided open moth jars. The exteriors of both were fabric impressed (Figure 91:D-E) and each displayed a plain or smoothed band below the lip on the exterior. The largest rim group consists of direct rims that have been thinned on the interior contour beginning at a distance of about 8 to 10 mm below the lip. This gives a slightly flaring appearance to the lip termination, but the vessel walls are straight-sided to very slightly recurved. Most of these rims are thick and appear to represent large open mouth jars. Five of these were fabric impressed (Figure 91:A-C, Figure 92), one was dowel impressed (Figure 95:E-F), one was cord marked and one had an indeterminate exterior surface treatment. Two other, less frequently occurring rim groups were also identified. One form represented a restricted orifice jar with a lip oriented at an oblique angle to the rim line. Restricted orifice jars are more common in Hanover I and II, but the obliquely oriented lip appears to be a newly added trait. The other group consisted of a single incurved rim from a small vessel. In plan, the rim line is noticeably concave, suggesting a vessel form resembling a “seed-jar.” Vessel wall thickness for the Hanover III sample was 8.20 mm (SD=1.81 mm), which is nearly the same as the Hanover I and II samples as well. Means for cord marked (8.21 mm, SD=1.49 mm) and fabric impressed (8.13 mm, SD=1.84 mm) within the Hanover III sample were essentially the same as well. Minority surface treatments occurred too infrequently to generate meaningful vessel wall thickness statistics. Yadkin Series Coe (1964:30-32) defined Yadkin series pottery from his excavations at the Doershuck Site. Abundant, angular crushed quartz was the diagnostic trait of the series, which was characterized by three main surface treatments, fabric impressed, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 294 Chapter 7. 31HK2521 cord marked, and linear check stamped. Subsequent investigations at Town Creek (Coe 1995) also associated simple stamped exteriors with the series. In the North Carolina Sandhills, the definition of Yadkin has been expanded to include all types of crushed rock temper (Herbert 2003:58–59). Currently, examples of crushed granite, crushed quartzite or arenite, crushed and ground sandstone, subangular feldspar sand and granule/pebble subangular quartz sand have been identified in assemblages in the region. The presence of large, crushed and ground sandstone particles in sherd collections within Fort Bragg was first recognized during Phase II testing at 31MR326 and 31HT748 conducted by Palmetto Research (Cable and Cantley 2005b). Due to its texture and composition, it can be confused with grog and probably has been on occasion in the Fort Bragg area of the Sandhills. The material is light tan to grayish yellow when viewed in sherds and generally contains fine to medium grains of arenite that can become finely disseminated in the paste matrix. The two materials are found in the same geological deposits and large particles of each can co-occur in individual sherds. Significant quantities of the sandstone tempered sherds at 31MR326 also contained grog temper, which is also the case for the 31HK2521 ceramic collection. The Yadkin material from 31HK2521 exhibited hard paste; consequently, all of it was assigned to the Yadkin III subseries. One hundred eighty-seven Yakin III sherds were identified (Table 76). Fabric impressed and cord marked surface treatments were co-dominant, each representing 39.1 (n=45) percent of the recognizable surface treatment sample. An important sub-dominant was plain or smoothed surfaces, which comprised 15.7 percent (n=18) of the collection. Minor percentages were contributed by dowel impressed (2.6 percent, n=3), wide simple stamped (2.6 percent, n=3) and check stamped (0.9 percent, n=1). A slightly different picture emerges, however, when the surface treatment percentages are controlled by vessel representation. The Yadkin III ceramic inventory included 46 vessels with recognizable surface treatments. Within this sample, Fabric impressed surfaces occur on 47.8 percent (n=22) of the vessels, while cord marking was identified on 32.6 percent (n=15). Minority representation included 8.7 percent (n=4) plain, 4.4 percent (n=2) dowel impressed, 4.4 percent (n=2) wide simple stamped and 2.2 percent (n=1) check stamped. These percentages are nearly identical to the Hanover III surface treatment profile, suggesting that the two subseries may be relatively contemporaneous. Rigid warps were recorded on approximately 69 percent (n=9) of the fabric impressed vessel sample, compared with 31 percent (n=4) flexible warps (Figure 97). Medium wefts dominate the fabric impressed vessel sample, comprising about 62 percent (n=8) of the recognizable surface treatments. Fine wefts make up 31 percent (n=4) of the vessel sample and coarse wefts were found on only one vessel (approximately 8 percent). The Yadkin III sample has higher proportions of rigid warps and fine to medium wefts than Hanover III. Stamp patterns on cord marked vessels (Figure 98) consisted of 47 percent (n=7 vessels) perpendicular/oblique, 40 percent parallel (n=6 vessels) and 13 percent (n=2) of vessels exhibiting both stamp patterns on individual sherds. These patterns are similar to the Hanover III sample, although parallel patterns might be marginally more frequent. Cord impressions averaged 1.26 mm in diameter (SD=0.40 mm), and individual widths ranged between 0.62 and 2.13 mm. These statistics are nearly identical to the Hanover III cord marked sample. A small number of sherds displayed an impression that appears to have been made with a cord wrapped dowel or by the edge of a cord wrapped paddle (Figure 99:A-B). These are referred to here as dowel impressed to differentiate them from classic cord marked treatments and from rigid warp fabric impressed patterns, which they resemble on a superficial basis. A single check stamped (Figure 96:AC) vessel was assigned to the Yadkin series (Figure 99:F). It contained abundant granule- to pebble-sized crushed and ground indurated granite temper and exhibited an extremely hard paste. Finally, three sherds from two separate vessels were classified as wide simple stamped. One exhibits squared-off lands and grooves indicative of being made with a carved paddle stamp, like the examples described for Hanover III (Figure 99:E). The other specimen displayed broad, rounded lands indicative of being formed with a thong wrapped paddle (Figure 99:C-D). As discussed for the Hanover III samples, South (2005:28) indicates that Gaston Simple Stamped sherds from the Roanoke Valley exhibit Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 295 Chapter 7. 31HK2521 both carved paddle and thong wrapped paddle varieties. The rim treatments on Gaston Simple Stamped, however, possess a wide range of Lamar-like decorative elements, including incidental hollow reed punctations and folded and applique rim strips dating to the sixteenth and seventeenth centuries. However, the transition from the earlier cord marked dominated assemblages (i.e. Clements) to Gaston is not well understood and it is possible that simple stamped pottery arose much earlier in the Roanoke Valley than is currently accepted. Herbert’s (2003:194-201) review of the Cashie and Townsend series in the North Carolina Tidewater region indicates a similar uncertainty over the earliest appearance of simple stamped types, but it is possible that they emerge in small frequencies in the early part of the Late Woodland period. The Yadkin III collection from 31HK2521 exhibits a diverse range of temper constituents. The most interesting development in this regard is the addition of large pieces of crushed arenite and especially crushed and ground indurated granite. Arenite is commonly found on precontact sites on Fort Bragg and although most of it was used for hearth rocks and heating stones, a secondary use appears to have been to produce temper. Indurated granite is only rarely found in naturally occuring rock form, but it represents a major source of temper. Along with grog temper, approximately 38.2 percent of the Hanover III vessel sample contains crushed arenite, crushed and ground indurated granite, or both. grog in the Hanover III series and their presence in the Yadkin III series suggests partial contemporaneity and/or cultural continuity. Approximately 46.8 percent of the Yadkin III vessel assemblage is tempered with indurated granite (Table 77). Various combinations of temper constituents in which all share crushed arenite make up 22.6 percent of the sample, while granule- to pebble-sized subangular white quartz sand contributes 17.7 percent and feldspar sand makes up the remaining 12.9 percent of the sample. Respectively, 88 percent of feldspar-, 38 percent of arenite- and 47 percent of indurated granite-tempered sherds also contain granule- to pebble-sized subangular white quartz. As was the case for Hanover III, fabric impressed exteriors were most abundant, ranging between 45.8 and 60.0 percent of the vessel sample. No fabric impressed vessels were identified in the feldspar variant, but the sample size is too small to arrive at strong conclusions about the proportional make up of this variant. Potentially significant is the absence of cord marking in the arenite sample. If the proportion of fabric impressed surfaces is expected to increase with time, then this result could indicate that arenite-tempered material is positioned later in time. Although small (n=3), the Hanover III arenite-tempered sample is also devoid of cord marked vessels. Two vessels display fabric impressed exteriors and one exhibits a dowel impressed exterior treatment. Larger sample sizes would be needed to confirm this potential pattern. Interior finishes on the Yadkin III sample are similar to those described for the Hanover III sample, supporting the hypothesis that Yadkin III represents a Late Woodland series. Floated (Figure 97:G, I, Figure 98:J, Figure 99:I) and ground smooth (Figure 98:M, Figure 99:M) surfaces respectively occur on 25.8 and 22.6 percent of the Yadkin III vessel sample, proportions that compare well to the Hanover III sample. As noted earlier, Phelps and Heath (1998:12) found that the interiors of most Cashie I ceramics are floated, which results visually in subduing the effects of protruding granule- and pebble-sized inclusions. Similarly, floating may have served to cover-up the protruding contours of granule- and pebble-sized temper particles in the Hanover III and Yadkin III material from Fort Bragg. However, floated surfaces may also have been created to increase the water-holding capacity of vessels and to retard fine particle dissolution. Radiocarbon dates for Cashie I place it between AD 985 and 1480 at the 1-sigma range of uncertainty (Eastman 1994a, b; Phelps and Heath 1998:6). South (2005:28) found that Gaston phase interiors were also sometimes “smoothed by use of a smooth stone or other object, so that the temper is pressed into the paste.” The Yadkin III rim assemblage is generally similar to the Hanover III rims as well. Five different rim groupings were recognized in the Yadkin III sample (Figure 87). Most of the rims had direct orientations from straight-sided or very slightly recurved vessel wall profiles. These include direct thinned, direct simple and direct beveled rim groups. Direct simple rims derive from smaller vessel forms that appear to represent hemispherical bowls or small, straight-sided open moth jars. Five rims of this group were recognized, representing three vessels. One was fabric impressed (Figure 97:D-E), one was plain and Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 296 Chapter 7. 31HK2521 Table 76. Yadkin III Ceramic Inventory, 31HK2521. INVENTORY OF YADKIN III CERAMICS: 31HK2521 CERAMIC TYPE CERAMIC PORTION BASE BODY NECK RIM RESID. GRAND TOTAL Crushed Quartzite Yadkin IIIc Fabric Impressed 1 1 2 Yadkin IIIc indet. 1 1 Yadkin IIIc Indet. decorated 1 1 Yadkin IIIc indet. stamped 1 1 Yadkin IIIc Plain 2 Yadkin IIIc/f Fabric Impressed 2 1 3 Yadkin IIIc/f indet. 3 1 4 2 Crushed Quartzite/Sandstone Yadkin IIIc/e Dowell Impressed 2 2 Yadkin IIIc/e Fabric Impressed 2 2 Yadkin IIIc/e indet. 1 1 Yadkin IIIc/e indet. decorated 1 1 Yadkin IIIc/e Wide Simple Stamped 1 1 Yadkin IIIc/e/f Fabric Impressed 2 2 Yadkin IIIc/e/f Plain 1 1 Crushed Sandstone Yadkin IIIe Check Stamped Yadkin IIIe Cord Marked 1 1 1 25 5 31 Yadkin IIIe Fabric Impressed 27 3 30 Yadkin IIIe indet. 15 2 17 Yadkin IIIe indet. decorated 1 1 Yadkin IIIe indet. stamped 3 3 Yadkin IIIe indet. textile 3 3 Yadkin IIIe Plain 1 1 Yadkin IIIe Wide Simple Stamped 2 2 13 13 Granule/Pebble Subang. White Quartz Yadkin IIId Cord Marked Yadkin IIId Fabric Impressed 6 6 Yadkin IIId Indet. 24 24 Yadkin IIId indet. 1 1 Yadkin IIId indet. decorated 4 4 Yadkin IIId Plain 1 1 Feldspar Sand Yadkin IIIf Cord Marked 1 Yadkin IIIf Dowell Impressed Yadkin IIIf indet. 10 Yadkin IIIf Plain 11 GRAND TOTAL 1 1 1 170 1 10 2 1 13 13 2 187 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 297 Chapter 7. 31HK2521 Figure 97. Yadkin III Fabric Impressed, Site 31HK2521. A-E: Yadkin IIIe Fabric Impressed (p1153, p1929, p2412, p1905 and p1904 respectively), F-G: Yadkin IIIc/e Fabric Impressed (p1132, F: Exterior, G: Interior, Floated), H-I: Yadkin IIIc/f Fabric Impressed (p733, H: Exterior, I: Interior, Floated), J: Yadkin IIIc/e Fabric Impressed (p1562), K: Yadkin IIId Fabric Impressed (p51). the other was indeterminate. As was true of Hanover III, another popular group consists of direct rims that have been thinned on the interior contour beginning at a distance of about 8 to 10 mm below the lip. This gives a slightly flaring appearance to the lip termination, but the vessel walls are straight-sided to very slightly recurved. Most of these rims are thick and appear to represent large open mouth jars. Two of these displayed fabric impressed (Figure 97:H-I) exteriors, while three rims exhibited cord marked (Figure 98:A, D, E) exterior surface treatments. A single direct beveled rim was identified (Figure 98:C) on a cord marked sherd. The vessel wall was slightly out-curved, suggesting a medium sized, open moth oblong-shaped jar. A thick, beveled rim with a thickened or rolled lip and a restricted orifice (Figure 98:B) indicates a relatively large, neck-less, restricted orifice jar. This jar was cord marked as well. Finally, a thin, fabric impressed rim with a thickened or rolled lip and a recurved orientation appears to represent a small, slightly flaring jar. Vessel wall thickness for the Yadkin III sample was 7.90 mm (SD=1.28 mm), which is thinner than the Hanover III sample. Most of the difference appears to be explained by the thinner vessel wall means for the fabric impressed and plain subsamples. Respectively these surface treatments had means of 8.02 mm (SD=1.41 mm) and 7.49 mm (SD=1.59 mm). By contrast, the cord marked subsample was relatively thick with a mean of 8.31 mm (SD=1.15 mm). These differences may simply relate to functional variability in associated vessel forms, but it is also possible that chronological factors are operable. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 298 Chapter 7. 31HK2521 Figure 98. Yadkin III Cord Marked, Site 31HK2521. A-H: Yadkin IIIe Cord Marked (p2198, p2372, p2236, p1198, p1664, p1037, p1275 and p1168 respectively), I-J: Yadkin IIIe Cord Marked (p1247, I: Exterior, J: Interior, Floated), K: Yadkin IIIf Cord Marked (p1152), L-M: Yadkin IIIe Cord Marked (p392, L: Exterior, M: Interior, Floated). Cape Fear Series Stanley South (1976:18-20) first described the Cape Fear series, or ware-group as he described it, from surface collections made in 1960 on sites located between Wilmington, NC and Myrtle Beach, SC. He applied the term to virtually all of the sand-tempered cord marked, fabric impressed, and net impressed pottery within the collection. Many have suggested that the series was too broadly applied by South and Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 299 Chapter 7. 31HK2521 Figure 99. Yadkin III Cord-Wrapped Paddle Edge Stamped, Simple Stamped, Check Stamped and Plain, Site 31HK2521. A-B: Yadkin IIIc/e Dowel Impressed/Paddle Edge Stamped (p1376, A: Exterior, B: Interior), C-D: Yadkin IIIe Wide Simple Stamped (p2343, C: Exterior, D: Interior, Floated), E: Yadkin IIIc/e Wide Simple Stamped (p985), F: Yadkin IIIe Check Stamped (p2062), G: Yadkin IIIf Plain (p2154), H-I: Yadkin IIIc Plain (p192, H: Exterior, I: Interior, Floated). other investigations in North Carolina indicate that such material should be separated into a sequence of series based on sand inclusion coarseness (see Anderson and Logan 1981:107-108; Trinkley 1981:11). Coastal North Carolina sequences contain at least two sand grain size modes that are used to distinguish sand-tempered series (Phelps 1983). The Deep Creek series consists of pastes with abundant, coarse sand and it is chronologically correlated with Deptford. Other series (i.e. Mount Pleasant and Cape Fear) contain finer grained pastes and they are roughly dated to a post-Deptford or late Middle to Late Woodland context (i.e. AD 500-900 or 1500 to 1100 BP). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 300 Chapter 7. 31HK2521 Table 77. Percentage of Surface Treatment Types by Yadkin III Paste Subvariants, 31HK2521.* REPRESENTATION OF YADKIN III PASTE VARIANTS AND SURFACE TREATMENTS: 31HK2521 45.8 IIId (quartz) 17.7 50.0 40.0 IIIf (feldspar) 12.9 IIIc Comb. (arenite) 22.6 60.0 GRAND TOTAL (n) 100.0 22 33.3 16 4.2 4.2 33.3 10.0 10.0 2 2 1 GRAND TOTAL (n) PLAIN 45.8 CHECK STAMPED CORD MARKED 46.8 WIDE SIMPLE STAMPED FABRIC IMPRESSED IIIe (indurated granite) YADKIN III PASTE VARIANT GROUPS DOWEL IMPRESSED OVERALL PERCENT SURFACE TREATMENT TYPE 24 10.0 10 33.3 3 20.0 10 4 47 * Data Generated from Vessel Counts Rather than Raw Sherd Frequencies. Herbert (2003:156) distinguishes the sandtempered New River and Cape Fear series by quartz sand density in the North Carolina Sandhills. Cape Fear series material is characterized by low to moderate densities of quartz sand amounting to less than 15% of the paste body. In addition, he places all perpendicular stamped cord marked exterior surfaces in the Cape Fear series, while all net-impressed is assigned arbitrarily to New River. Cable and Cantley (2005a, b) have distinguished two series variants for Cape Fear in the Fort Bragg area based on quartz sand density and paste characteristics. Cape Fear I was characterized by a friable, gritty paste and a moderate density of medium to coarse quartz sand temper. Cape Fear II exhibited a finer, more compact paste with sparse medium to coarse quartz sand temper. Although the Cape Fear series sherds identified at 31HK2521 meet the general sand-grain size and density parameters discussed above, it possessed an extremely hard to vitrified paste. Consequently, it was all assigned to a third Cape Fear subseries, Cape Fear III, which is assumed to belong to the Late Woodland period. Only twenty-nine Cape Fear III sherds were recognized in the collection (Table 78). The majority of the sherds are cord marked, repre- senting 54.2 percent of the recognizable surface treatments. Plain surfaces comprise 25.0 percent and fabric impressed exteriors occur on 20.8 percent. The vessel sample, however, shows an even higher dominance of cord marking. Of the fourteen identified Cape Fear III vessels with recognizable surface treatments, nine displayed cord marking, four were plain and only one was fabric impressed. Respectively, then, cord marking represented 64.3 percent, plain surfaces were found on 28.6 percent and fabric impressed was recorded on only 7.1 percent of the vessel assemblage. These percentages represent a clear departure from the Hanover III and Yadkin III collections, suggesting temporal and/or cultural differences. Stamp patterns on cord marked vessels (Figure 100:A-I) consisted of 57 percent (n=4 vessels) parallel, 43 percent (n=3) exhibiting both parallel and oblique over-stamped patterns on individual sherds or on different fields of the same sherd. Parallel patterns were recorded on small sherds, while perpendicular/oblique patterns were observed on larger sherds, strongly suggesting that the modal stamping pattern of whole vessels would have been zoned fields of parallel stamping on one hand and oblique Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 301 Chapter 7. 31HK2521 cross stamping on the other. The two largest cord marked sherds in the collection clearly show this pattern (Figure 100:A, H). Cord impressions averaged 1.07 mm in diameter (SD=0.38 mm), and individual widths ranged between 0.59 and 1.59 mm. These data might suggest that cord marking on Cape Fear III vessels was somewhat finer than that found on Hanover III and Yadkin III specimens. These series produced mean cord widths between 1.20 and 1.26 mm and maximum diameters ranged between 2.13 and 2.27 mm. The fabric impressed sherds originated from a single vessel (Figure 100:J-K), but weave patterns were not recognizable. The plain sherds (Figure 100:L-M) were small, but all were only roughly finished with individual temper grains protruding along the exterior surfaces. Interior finishes on the Cape Fear III series resemble those found in the Hanover III and Yadkin III samples, with two differences. First, there appears to be a greater emphasis placed on creating regularized interior surfaces. Sixty-three percent of interiors are either floated or ground smooth in the Cape Fear III sample, while only 51 and 48 percent occurrences were recorded respectively for Hanover III and Yadkin III. Second, floated surfaces are more common in Hanover III and Yadkin III, while the reverse is true for Cape Fear III. In the latter sample, 44 percent of interiors are ground smooth (Figure 100:B, G, I, K, L), while only 19 percent are floated (Figure 100:D). Since Cape Fear III is characterized by finer temper particle sizes, the suggested function of reducing the visual impact of protruding temper particles by floating surfaces (see Phelps and Heath 1998:12) gains support. A single rim sherd was recovered in the Cape Fear III sample. The exterior of the sherd was cord marked and the rim form appears to represent a medium-sized jar with a recurved wall and a slightly flaring rim orientation (Figure 87; Figure 100:E). Vessel wall thickness for the sample was 7.66 mm (SD=1.27 mm), which is thinner than either the Hanover III or Yadkin III samples. This difference is particularly striking because the Yadkin III and Hanover III cord marked samples have respective mean wall thicknesses of 8.31 mm and 8.21 mm, while the Cape Fear III Cord Marked sherds have a mean of only 7.77 mm (SD=1.5 mm). Temper constituents in the Cape Fear III series are composed exclusively of clear and white, rounded quartz sand. The modal mixture consists of abundant fine to medium quartz sand and seventy-six percent of the sample contains only medium and fine sand grains (Table 79). A minority of specimens contained also contained larger sand grains, but most of these were also characterized by moderate to sparse density. These specimens were assigned to the Cape Fear III series based on two factors, the exclusive use of predominantly fine to medium quartz sand temper and hard to vitrified pastes. Cape Fear III series paste resembles most closely the paste varieties that characterize Mississippian series pottery as it is represented in the upper Pee Dee River Valley (Town Creek), the upper Wateree River Valley (Mulberry-Adamson) and the lower Santee River Valley (Jeremy-Tibwin). Whether there is a direct correlation with neighboring Mississippian cultures is an issue that cannot be resolved at present. However, there is a similar cord marked type, Savannah Creek Fine Cord Marked, which commonly occurs in Teal phase assemblages of the Town Creek sequence (see Oliver 1992:206). Heavy cross stamping at oblique angles is common to the type and cord diameters are narrow, ranging between 1 and 2 mm. Bordeaux (2005:92) assigns a temporal range of AD 1000 to 1150 for the Teal phase, which correlates with Savannah I assemblages across the South Appalachian Mississippian region. Although complicated stamped is the most frequent ceramic type in the Teal phase reconstruction, cord marking ranks second. Investigations in other regions indicate that an earlier, transitional Mississippian phase exists which is dominated by cord marked and/or simple stamped pottery. The Santee II phase on the lower Santee River in South Carolina is dominated by simple stamped pottery that is commonly cross stamped at oblique angles in a similar fashion to Savannah Creek Fine Cord Marked (see Anderson et al. 2002). A similar simple stamped ceramic phase known as the Camden phase underlies the Mississippian sequence in the upper Wateree Valley (Cable et al. 1998; Stuart 1975). The Savannah I phase at Irene is composed almost exclusively of a fine cord marked pottery known as Savannah Fine Cord Marked (Caldwell and McCann 1941). Small amounts of fabric impressed pottery has also been identified at the Teal site that may be associated with the earliest portions of the Teal phase Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 302 Chapter 7. 31HK2521 Table 78. Cape Fear III Ceramic Inventory, 31HK2521. INVENTORY OF CAPE FEAR III CERAMICS: 31HK2521 CERAMIC TYPE Cape Fear III Cord Marked PORTION BODY 12 RIM 1 GRAND TOTAL 13 Cape Fear III Fabric Impressed 5 5 Cape Fear III indet. 4 4 Cape Fear III indet.decorated 1 1 Cape Fear III Plain 6 6 GRAND TOTAL (Boudreaux 2005: 45). Finally, Cable (2011) has identified a ceramic phase transitional from Wilmington to Santee II (almost exclusively simple stamped pottery) that contains fabric impressed, cord marked and dowel impressed surface treatments in addition to simple stamping. In combination with the interior surface treatment data, then, the Cape Fear III series fits well into a transitional Mississippian phase format. We know that later Mississippian pottery is occasionally found in the Sandhills. For instance, a portion of a Pee Dee Complicated Stamped flare-rim jar exhibiting concentric circle motifs was recovered by TRC-Garrow during a Phase II investigation at 31CD1356 (Gray 2005:121-128, 306-309). This site is situated at the head of Stewart Creek in the southeastern part of Fort Bragg. The design and vessel form suggest an early to middle Mississippian affiliation. The sporadic and infrequent occurrence of Mississippian pottery in the North Carolina Sandhills suggests that Mississippian groups temporarily occupied the region, probably for brief logistical forays. Pee Dee Serrated and Small Triangular points are much more commonly found in the Sandhills than Mississippian pottery and this might suggest that much of the Mississippian utilization of the Sandhills involved hunting parties. The Cape Fear III series ceramics found at the Quewhiffle site probably do not represent cultural continuity with the Hanover III and Yadkin III series ceramics, which 28 1 29 appear to constitute the main elements of the local regional sequence. If this reconstruction is correct, it suggests that the Fort Bragg locality may have served as a culture contact zone during portions of the Late Woodland period In light of this discussion, it is likely that the Cape Fear series has probably been too broadly defined in the past to include all sparse to moderate sand-tempered post-New River pottery in the region. The differences noted in the Cape Fear cord marked to fabric impressed ratios between the Haag collection (Herbert 2003:69-70) and Stanly South’s (1960) survey collection should not be dismissed in this regard. South calculated a cord marked to fabricimpressed ratio of 1.6:1.0, which suggested to him that Cape Fear succeeded Hanover. Herbert, by contrast, generated a reverse ratio of 0.39:1.0 and concluded that Cape Fear preceded Hanover. A good deal of this disparity can probably be attributed to definitional issues. Herbert assigned only perpendicular stamped (cross stamped) cord marked specimens to Cape Fear, while sand-tempered parallel stamped cord marked and net-impressed were included exclusively within the New River series. However, the mere recognition of these differences points to the strong probability that unrecognized post-Hanover sand-tempered series exist in the regional record that might be derivative from Hanover III and Yadkin III. The McLean (MacCord 1966) and Buie Mound (Wetmore 1978) Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 303 Chapter 7. 31HK2521 Figure 100. Cape Fear Series Ceramics, Site 31HK2521. A-B: Cape Fear III Cord Marked (p2311, A: Exterior, B: Interior, Ground Smooth), C-D: Cape Fear III Cord Marked (p85, A: Exterior, B: Floated), E: Cape Fear III Cord Marked (p1893), F-G: Cape Fear III Cord Marked (p482, F: Exterior, G: Interior, Floated), H-I: Cape Fear III Cord Marked (p1738, H: Exterior, I: Interior, Ground Smooth), J-K: Cape Fear III Fabric Impressed (p2567, J: Exterior, K: Interior, Ground Smooth), L-M: Cape Fear III Plain (p2587, L: Exterior, M: Interior, Ground Smooth). ceramic assemblages, as discussed in Chapter 3 (see also the discussion in Herbert 2003:193), constitute two excellent candidates for Late Woodland and protohistoric sand-tempered series near Fort Bragg. The latter, however, may be linked to South Appala- chian cultures, while the former appears to constitute a viable candidate for development out of the local Hanover III/Yadkin III tradition. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 304 Chapter 7. 31HK2521 Table 79. Contingency Table of Sand Density and Grain Size, Cape Fear III Sample, 31HK2521. CONTINGENCY TABLE OF SAND DENISTY BY SAND GRAIN SIZE, CAPE FEAR III SAMPLE: 31HK2521 DOMINANT GRAIN SIZE Fine to Medium DENSITY ABUNDANT MODERATE 16 1 Medium Medium to Coarse 2 18 Concluding Remarks The ceramic assemblage recovered from the Quewhiffle Site provides a window into a large portion of the ceramic sequence of the North Carolina Sandhills. Table 80 displays surface treatment percentages for the identified ceramic sub-series as reflected by the vessel data set. This data set supplies a basic advantage over raw sherd counts, because it controls for problems of over-representation brought about by the recovery of large numbers of sherds from a limited number of pots. This data set was not recovered purely from a non-biased areal sample of the site since it includes sherds from test units and sample block shovel tests. However, it has the added advantage of supplying a larger sample size from which to construct proportional models of ceramic change than would be produced solely from the Stage I and II shovel tests. De-emphasis on high sherd volume vessel aggregates should offset problems with recovery bias. Although the ceramic subseries are organized in an inferred chronological sequence based on changes in paste technology and temper constituents, it is not assumed that the sequence as presented is complete. It is merely a reflection of the Woodland period occupation history of the Quewhiffle Site. Regional and chronological considerations of integration are discussed at length in the synthesis chapter. A broad outline of the inferred chronology is summarized here. New River I represents the Early SPARSE 4 21 1 1 4 6 Medium to Very Coarse GRAND TOTAL GRAND TOTAL 6 1 1 5 29 Woodland period, while New River II is viewed as an early Middle Woodland manifestation. Hanover I and II are correlated with the late Middle Woodland period. Hanover III and Yadkin III are viewed as contemporaneous based on exterior surface treatment characteristics and similarities in paste. For reasons enumerated in the synthesis chapter, Hanover III and Yadkin III are correlated with the early Late Woodland period, with the exception of arenite-tempered specimens of each series, which reflect a later pattern of exterior surface treatment representation. Finally, the Cape Fear III series is hypothesized to represent an intrusion of early or transitional Mississippian groups who might have occupied the area on a seasonal and intermittent basis. OCCUPATION PATTERNS Diagnostic markers of culture-chronological association were relatively numerous in the Quewhiffle site artifact assemblage. Unlike most sites in the interior uplands of Fort Bragg, this site produced an extensive Woodland period occupation reflected in both diagnostic projectile points and ceramics. Recovered projectile points document Early Archaic (i.e., two Palmer I Corner Notched) and Middle Archaic (i.e., a Morrow Mountain I Stemmed). In addition, a wide range of scraper types and specialized unifacial tools evidence especially Early Archaic occupation, although some of the forms have also Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 305 Chapter 7. 31HK2521 been recognized in Middle Archaic deposits. Woodland point styles include a Badin Large Triangular, Variant C (Early Woodland), a Yadkin Large Triangular, Variant A (Middle Woodland), an Undefined Rudimentary Stemmed and two Undefined Expanded Stemmed. The rudimentary stemmed point has affinities to points in South Carolina that are tentatively associated with Deptford. The smaller size of the point compared to the Yadkin Large Triangular point, suggests that it might be associated with terminal Middle Woodland occupation (i.e. Hanover I/II). The two expanded stemmed points appear to represent the range of variation that would be compatible within the rubric of a single type. The small size and careful shaping technique suggest that the point style is affiliated with a post-Archaic culture. It is not possible to establish its precise cultural association, but one likely option is that it is affiliated with the extensive early Late Woodland occupation (i.e. Hanover III/ Yadkin III). The ceramic analysis identified the presence of Early Woodland (New River I), early Middle Woodland (New River II), late Middle Woodland (Hanover I and II) and early Late Woodland (Hanover III/Yadkin III). In addition, the Cape Fear series is interpreted as evidence of a temporary intrusion of transitional Mississippian seasonal occupations into the region. The Quewhiffle site is characterized by high artifact density. Mean artifact density in positive Stage I and II shovel tests (n=130) was 6.38 artifacts (SD = 6.71), indicating that two-thirds of the positive tests yielded densities between one and 13 artifacts. Although the modal positive shovel test outcome across the sample was only one artifact, 53 percent of the positive tests contained four or more artifacts. The entire site area (including negative Stage I and II shovel tests) had a mean artifact density of 3.86 artifacts (SD = 6.08) per shovel test. These values are approximately two times greater than those from shovel testing at 31HK2510. Metavolcanic debitage recovered during Stage I shovel testing was widely distributed across the terrace (Figure 101). The densest clusters occurred between the center of the ridge top and the ridge nose on the northeast end of the site. Quartz debitage displayed a similar distributional pattern, although overall artifact density was lower (Figure 102). Vertical distributions by raw material type indi- cate that debitage is centered between 10 and 50 cm bs (Table 81). These profiles suggest that deposits exhibit variable vertical positions and most likely are composed of a wide range of cultural associations. From what we have learned about the vertical distributions of cultural-chronological occupations on sites in the Sandhills (see Cable 2010; Cable and Cantley 2005a, b, 2006), this distribution should indicate that most of the lithic debris is associated with the Archaic period. However, the unusually intense Woodland period occupation at the site suggests that significant mixing of cultural deposits has occurred in the upper 30 cm of soil. Table 82 displays the vertical distributions of the various ceramic sub-series for all proveniences. Ceramic deposits are concentrated between 0 and 30 cm bs/bd and are centered between 10 and 20 cm in Level 2. Ceramics found below 30 cm bs were probably re-deposited as a consequence of bioturbation or intrusive cultural features. Recovery error may also be a factor in these “out of position” instances. Intensive Woodland occupation appears to have displaced some Archaic deposits upward, as is demonstrated by the vertical distributions of Archaic stone tools (Table 83). Just under half of the Early and Middle Archaic diagnostics occupy higher than expected vertical positions. Diagnostic Woodland projectile points are found between 0 and 40 cm bs in what appears to represent a coherent vertical sequence of point styles (Table 83). The The Badin Crude Triangular, affiliated with the early Woodland period, was positioned lowest in the sequence, at a depth of 30 to 40 cm bs. The Yadkin Large Triangular point, probably an early Middle Woodland type, was recovered between 20 and 40 cm bs. The undefined stemmed points that are speculated to be affiliated with late Middle Woodland and early Late Woodland occupation were found slightly higher, between 10 and 30 cm bs/bd. Given the vertical patterning of the Woodland ceramics and diagnostic projectile points, it is likely that much of the lithic debris found in Levels 1 and 2 of the Stage I and II sample is affiliated with Woodland occupation and that a significant proportion of the debitage found in Level 3 may also be so associated. It is estimated, then, that as much as 30 to 35 percent of the lithic debris recovered from the shovel test sample belongs to Woodland period occupation. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 306 Chapter 7. 31HK2521 Table 80. Percentage of Exterior Surface Treatment Types by Ceramic Series Variant, 31HK2521*. PERCENTAGE OF EXTERIOR SURFACE TREATMENT TYPES BY CERAMIC SERIES VARIANT: 31HK2521 New River I Hanover II 27.3 31.8 9.1 13.6 40.9 1.9 34.6 50.0 5.8 34.0 46.8 6.4 64.3 7.1 58 75 22 6 GRAND TOTAL (n) PLAIN INCISED/INDET TEXT. WIDE SIMPLE STAMPED DOWEL IMPRESSED 31.8 2.1 5 28.6 4.6 Cape Fear III GRAND TOTAL (n) 42.9 27.3 Hanover III Yadkin III FABRIC IMPRESSED 16.7 14.3 9.1 CORD MARKED LINEAR CHECK STAMPED 50.0 New River II Hanover I CHECK STAMPED CERAMIC SERIES VARIANT NET IMPRESSED EXTERIOR SURFACE TREATMENT TYPE 33.3 6 14.3 7 22 2.3 5 2.3 44 3.9 3.9 52 4.3 8.5 47 28.6 14 14 192 4 1 * Data Generated from Vessel Counts Rather than Raw Sherd Frequencies. Vertical patterning associated with the ceramic sub-series is indistinct, but there appears to be a vertical separation between the earliest series (i.e. New River I and II and Hanover I) and the later ones (i.e. Hanover II and III and Yadkin III). Respectively 63 and 60 percent of the New River I and II collections occur below 20 cm bs/bd, while 44 percent of the Hanover I collection occurs at this depth range. The later series display higher vertical positions. Only 29 percent of Hanover II sherds occur below 20 cm bs/bd, while 27 percent and 24 percent respectively of the Hanover III and Yadkin III sherds occur at this depth range. The Cape Fear III series displays a somewhat lower than expected vertical profile. Approximately 40 percent of this series occurs below 20 cm bs. This might indicate an earlier age for this material than assigned in the ceramic descriptions, but the hardness of the paste and the interior surface treatment characteristics support a later chronological placement. The small size of the Cape Fear III sample makes inferences concerning its vertical position tentative. It is assumed, at present, to be penecontemporaneous with Hanover III and Yadkin III. Comparing the vertical distributions of the Woodland projectile point styles to those of the ceramic subseries, the Badin point would seem to associate with New River, the Yadkin point with either New River II or Hanover I and the undefined stemmed points with Hanover III and Yadkin III. The Stage I and II shovel test sample indicates that ceramic sub-series are distributed primarily along the edge of the ridge top overlooking the tributary drainage on the east side of the site and the floodplain of Rockfish Creek on the northern end. The New River series are represented by limited and scattered deposits suggestive of periodic, short-term site reoccupation by small social units on the order of specially comprised task groups or single nuclear families (Figure 103). Hanover I and II deposits Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 307 Chapter 7. 31HK2521 (Figures 104 and 105) appear to be somewhat more concentrated and these concentrations may have been produced by larger social units (i.e. single or multiple families) that may have occupied the site for longer spans of time within single seasons. The Hanover III and Yadkin III series, which are thought to be contemporaneous, are characterized by denser, more concentrated occupations that might reflect multi-household camps that were occupied for much of a single season (Figure 106). Finally, the Cape Fear III series displays a dispersed distribution of low-density outputs similar in overall structure to the New River distributions (Figure 107). It is hypothesized that these represent separate seasonal, short-term reoccupations of the site by specially comprised task groups or nuclear family units. A more detailed appreciation for precontact land-use patterns evidenced at the site can be derived from a consideration of occupation type distributions and associations. Currently, four occupation types based on chipped stone criteria have been recognized on sites in the Fort Bragg area (see Cable and Cantley 2005c, 2006): (1) Type I occupations consist of a debitage concentration and an associated tool cluster situated at a confined location along its periphery. It is inferred that the tool cluster was formed around a hearth and in several instances calcined bone fragments have been recovered from the deposit delimited by a tool cluster. It is also believed that ephemeral shelters were situated adjacent to these tool clusters and opposite the concentrations. Debitage concentrations appear to represent lithic reduction loci established away from a shelter or sleeping area where tools were manufactured during the stay at the camp. These concentrations are generally composed of single lithic raw material types, and the associated tool clusters commonly consist of manufacturing rejects and discarded broken or worn-out tools made of the same raw material that comprises the associated debitage concentration. Although many Type I residences occur in isolation and appear to be the by-products from single nuclear or small extended family occupations, others appear to represent multi-family occupations that extend beyond the ability of a single shovel test to identify. Woodland period Type I occupations are characterized by an additional element, partial vessel sherd aggregations. These occur adjacent to debitage concentrations and they are hypothesized to occupy a portion of the living floor associated with a hut. The model would stipulate that tool clusters are situated between the debitage concentrations and sherd aggregations. (2) Type II residences are similar in spatial organization to Type I residences, but they exhibit much higher tool-to-debitage ratios and may commonly be comprised of diverse lithic raw materials. Lithic reduction on these sites consisted of tool maintenance and late stage shaping of tools from flake blanks or reduced biface cores. Consequently, debitage density is much lower than that of Type I debitage concentrations and the scatters consist primarily of late stage reduction debitage. Type II occupations are inferred to be the output from single or small, multiple household aggregations and are thought to represent the residue from High Technology Forager (HTF) residences (Spiess 1984; Todd 1983:231-233). The HTF model proposes a specialized forager adaptation combining high logistical and high residential mobility into a single settlement system characterized by highly curated technologies. Type II residences have thus far only been assigned to the Early Archaic period at Fort Bragg. The lithic raw material diversity characteristic of Type II residences suggests that these occupations represent the aggregation of social units that may have been seasonally dispersed, transporting a wide range of lithic raw materials from different sources. (3) Type III occupations are inferred to represent logistical camps and very short-term residences characterized by low-density debitage scatters and sporadic tool discard. (4) Type IV occupations represent extraction or processing loci that are not believed to have involved, necessarily, overnight stays. They are difficult, however, to distinguish from Type III occupations, even in situations of greater excavation exposure. Isolated sherd aggregations are hypothesized to represent one kind of Type IV element associated with Woodland occupations. It is difficult to distinguish effectively between these various occupation types at shovel test intervals of greater than 1.25 m, but a less precise classification can be used at greater shovel test Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 308 Chapter 7. 31HK2521 550 72 Bulldozed Trench .0 540 .0 73.0 75 530 74.0 520 510 500 31HK2521 490 73 .0 480 470 72.0 74 .0 460 450 440 430 Positive ST Negative ST 420 N 410 40 m 0 75. 0 400 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 570 Figure 101. Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31HK2521 (Contours = 1 piece of debitage). intervals to aid in population projections. High-density debitage concentrations tend to indicate Type I residences, while low-density scatters tend to signal short-term residences, special purpose camps and locations (i.e. Type II, III and IV occupations). Highdensity debitage concentrations are temporarily stipulated to correspond to shovel test outcomes of five or more pieces of debitage of a specific raw material subtype. Eventually, it will be necessary to confirm this arbitrary threshold by examining shovel test counts from a large sample of known high-density debitage concentrations. Deposits containing less than five pieces of debitage are viewed as more likely to represent low-density debitage scatters associated with the other three site types. Since shovel tests can intersect the peripheral zones of high-density debitage concentrations where debitage densities are much lower, however, it is likely that Type I residential occupations will be under-represented in macrointerval shovel test data. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 309 Chapter 7. 31HK2521 72 Bulldozed Trench .0 540 .0 73.0 75 74.0 520 500 31HK2521 73 .0 480 72.0 74 .0 460 440 Positive ST 420 N Negative ST 0 40 m 0 75. 400 420 440 460 480 500 520 540 560 Figure 102. Density Distribution of Quartz Debitage, Stage I and II Shovel Test Sample, Site 31HK2521 (Contours = 1 piece of debitage). Because it is difficult to determine whether a single occupation is represented in more than one shovel test, in the cases of multi-family residences or associated special activity zones on the periphery of a camp, the term “element” is preferred over the term “occupation” to refer to the occupational data in shovel tests. At a sampling interval of 5-m or 10-m intervals, we are assured that the presence of an element will not be repeated in adjacent shovel tests. Elements are identified by lithic raw material subtypes and/or by precontact sherd aggregations asso- ciated with one or more vessels. This methodology runs the risk of overestimating occupations because more than one raw material subtype may comprise a single occupation. The “element” concept recognizes this fact, focusing on depositional events and episodes rather than entire occupations. Elements can represent segments of contemporaneous multifamily occupations, separate occupations of various function, or contemporary members of multiple raw material reduction episodes. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 310 Chapter 7. 31HK2521 Table 81. Debitage Inventory by Level for Stage I and II Shovel Tests, 31HK2521. DEBITAGE BY LEVEL, STAGE I AND II SHOVEL TESTS: 31HK2521 LEVELS RAW MATERIAL TYPE Mill Mountain Rhyolite 1 2 3 4 5 6 7 3 5 7 6 3 1 1 1 8 9 GRAND TOTAL 10 26 Type I Rhyolite Tuff 1 10 18 15 5 3 Uwharries Eastern Rhyolite 3 16 10 5 8 6 1 49 7 7 8 13 3 3 41 7 Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) 4 20 45 39 19 Uwharries Western Rhyolite 1 18 13 2 6 hydrated metavolcanic 1 4 2 1 White Quartz 4 18 59 37 Crystal Quartz 6 2 56 1 1 141 1 42 8 13 11 2 3 3 1 151 1 Schist 1 1 1 Metasedimentary, Green 1 GRAND TOTAL 1 17 98 162 114 68 1 31 11 9 6 1 517 Table 82. Vertical Distribution of Ceramic Sub-series, All Proveniences, 31HK2521. VERTICAL DISTRIBUTION OF CERAMIC SUB-SERIES, ALL PROVENIENCES: 31HK2521 CERAMIC SUBSERIES LEVELS 1 2 3 4 5 7 1 GRAND TOTAL New River I 4 3 7 3 New River II 2 4 8 1 Hanover I 3 36 26 3 1 1 70 Hanover II 24 87 34 7 3 1 156 Hanover III 94 202 77 12 1 3 Yadkin III 65 71 43 5 9 7 10 Residual Sherds 20 25 5 2 GRAND TOTAL 221 435 210 33 Cape Fear III 1 6 19 15 1 390 3 187 1 27 1 7 7 53 4 917 The current methodology for identifying elements in shovel tests consists of running crosstabs of lithic debitage raw material subtypes or subtype groups by shovel test number in a spreadsheet. Sherd aggregations are not used in these calculations because, in general, they are associated with debitage scatters and their inclusion would result in potentially double counting a good number of elements. Once the cross-tab is generated, the sum of each lithic raw material subtype in a shovel test is identified as an element. In some instances, the vertical separation between individual items of the same lithic subtype can be so great as to suggest the presence of distinct elements. Due to the vagaries of bioturbation and possible recovery errors in which higher positioned material can be included in lower levels through sidewall displacement, however, the lower item(s) is assigned to the higher positioned item(s) to define Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 311 Chapter 7. 31HK2521 Table 83. Diagnostic Stone Tool Inventory, All Proveniences, 31HK2521. DIAGNOSTIC STONE TOOLS, ALL PROVENIENCES: 31HK2521 DIAGNOSTIC STONE TOOLS LEVELS 2 3 3-4 4 5 6 7 GRAND TOTAL EARLY ARCHAIC PROJECTILE POINTS/BLADES Palmer I Corner Notched 1 1 Type III Biface (Hardaway Blade) 2 1 1 2 RETOUCHED TOOLS Denticulate 1 1 2 Graver 3 Retouched Flake Spoke Shave 3 1 1 1 1 END SCRAPERS End Scraper Fragment 1 Type Ia End Scraper 1 1 1 Type Ib End Scraper 1 Type IIa End Scraper 1 1 1 Type IIb End Scraper SCRAPER FRAGMENTS 2 1 1 1 1 2 SIDE SCRAPERS Side Scraper Fragment 1 Type I Side Scraper 1 1 1 Type IIa Side Scraper 1 Type IIb Side Scraper 1 Type III Side Scraper 2 2 1 4 1 1 2 1 MIDDLE ARCHAIC PROJECTILE POINTS Morrow Mountain I Stemmed 1 1 WOODLAND PROJECTILE POINTS Badin Crude Triangular, Var. C 1 Yadkin Large Triangular, Var. A 1 Undefined Rudimentary Stemmed Undefined Expanded Stemmed GRAND TOTAL a single occupation element. In cases where large debitage concentrations of the same lithic raw material type are found at distinctively separated levels, though, each is counted as separate elements. Elements have horizontal dimensions and, as a result, sample units that are smaller than those dimensions have the benefit of increased sample efficiency (see Cable and Donaldson 1988; Rice 1987; 1 1 1 1 1 1 2 9 8 1 7 2 6 2 35 Rice and Plog 1983). The typical element found at Fort Bragg has a diameter of about 3 m. Thus, a shovel test of .09 m2 samples only .09 percent of the area of a 10 m-square, but it samples 9.0 percent of the theoretical mean element space of 9.0 m2. Consequently, a 10 m-interval shovel test pattern actually samples about 9.0 percent of the available element space at a site, while a 5 m-interval shovel test pattern samples about 36.0 percent of the available element Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 312 Chapter 7. 31HK2521 72 Bulldozed Trench .0 540 .0 73.0 75 74.0 520 500 31HK2521 73 .0 480 72.0 74 .0 460 New River I Series 440 New River II Series Positive ST 420 N Negative ST 0 40 m 0 75. 400 420 440 460 480 500 520 540 560 Figure 103. Density Distribution of New River Series Ceramics, Stage I and II Shovel Test Sample, Site 31HK2521 (Contours = 1 sherd). space. A mean estimate of the population of elements at shovel test intervals of 10 m and 5 m, then, requires a simple arithmetic calculation of counting the occupation elements identified in shovel tests and dividing this figure respectively by 9.0 percent and 36.0 percent. Confidence intervals can be constructed around these means, but they have not been calculated in the Fort Bragg testing projects. Although Stage I shovel test samples have been relied upon in the past to produce population projections because of their even and complete coverage, a concerted effort was made in the DO5 package to extend 5 m-interval shovel tests across the entire site area so that advantage could be taken of greater sample density. Following the procedures outlined above, the method succeeded in identifying 21 Type I residence elements, and 204 indeterminate (Type II, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 313 Chapter 7. 31HK2521 72 Bulldozed Trench .0 540 .0 73.0 75 74.0 520 500 31HK2521 73 .0 480 72.0 74 .0 460 Hanover I Series 440 Positive ST 420 N Negative ST 0 40 m 0 75. 400 420 440 460 480 500 520 540 560 Figure 104. Density Distribution of Hanover I Series Ceramics, Stage I and II Shovel Test Sample, Site 31HK2521 (Contours = 1 sherd). III and IV) elements in the combined Stage I and II Shovel Test data (Table 84). The Uhwarries Southern Rhyolite sample was segregated into two raw material groups (i.e., the R3m and R3t groups as discussed in the raw material description section of the chapter). Uwharries Southern Rhyolite (2) and White Quartz together comprise 54.7 percent of the identified elements. It can further be estimated that the site contains a mean of about 625 elements, of which approximately 58 represent Type I residential elements. Stone tools were counted as elements of a particular lithic raw material subtype when found in isolation from debitage in a shovel test. Eleven isolated tools or tool clusters were counted in this census. As was true of the 31HK2502 and 31HK2510, probable Early Archaic artifacts form a large portion of the isolated Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 314 Chapter 7. 31HK2521 72 Bulldozed Trench .0 540 .0 73.0 75 74.0 520 500 31HK2521 73 .0 480 72.0 74 .0 460 Hanover II Series 440 Positive ST 420 N Negative ST 0 40 m 0 75. 400 420 440 460 480 500 520 540 560 Figure 105. Density Distribution of Hanover II Series Ceramics, Stage I and II Shovel Test Sample, Site 31HK2521 (Contours = 1 sherd). stone tools. These include a Type IIa End Scraper, a scraper fragment, two gravers and the two ovate Type III bifaces that resemble Hardaway Blades. Isolation for the purposes of element population estimates means only that debitage matching the isolated tool was not recovered in the shovel test. This might suggest again, that a number of the Early Archaic elements represent members of Type II occupations, which are characterized by low-density debitage scat- ters of multiple raw material types combined with high stone tool density. The proportional representation of Type I residential elements (9.3 percent) is about twice as high as it is at 31HK2502 and 31HK2510. This percentage is still low compared with a large segment of the Fort Bragg site assemblage, which have been demonstrated to contain as much as 20 to 30 percent Type Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 315 Chapter 7. 31HK2521 72 Bulldozed Trench .0 540 .0 73.0 75 74.0 520 500 31HK2521 73 .0 480 72.0 74 .0 460 Hanover III Series Yadkin III Series 440 Positive ST 420 Negative ST N 0 40 m 0 75. 400 420 440 460 480 500 520 540 560 Figure 105. Density Distribution of Hanover III and Yadkin III Series Ceramics, Stage I and II Shovel Test Sample, Site 31HK2521 (Contours = 1 sherd). I residences (see Cable 2010). The site was occupied primarily by social units for short-periods of time, but it will be shown that during certain intervals the site used more intensively and appears to have served as a residential hub. Expanded over nearly the full range of human occupation (about 10,000 years), the population estimate indicates that precontact groups, on average, visited the site a maximum of once every 44.4 years. The occupation, however, was probably not evenly distributed through time. The site was no doubt used intensively by specific culturehistoric groups, probably for strings of continuous years punctuated by abandonments in response to long-term land-use patterns dependent upon resource depletion cycles. The site appears to have been most intensively utilized during the Early Archaic and Woodland periods. Due to the constrained nature of the landform to provide suitable locations for camping, the Que- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 316 Chapter 7. 31HK2521 72 Bulldozed Trench .0 540 .0 73.0 75 74.0 520 500 31HK2521 73 .0 480 72.0 74 .0 460 Cape Fear III Series 440 Positive ST 420 Negative ST N 0 40 m 0 75. 400 420 440 460 480 500 520 540 560 Figure 107. Density Distribution of Cape Fear III Series Ceramics, Stage I and II Shovel Test Sample, Site 31HK2521 (Contours = 1 sherd). whiffle site was characterized by a high incidence of element superimposition. Mean element superimposition in the combined Stage I and II shovel test sample was 2.07 (SD = 1.10). This is lower than the incidence of superimposition at 31HK2502 (2.24 elements per positive shovel test), which was located on an even more restricted levee-like formation, but it is higher than 31HK2510 (1.62 elements per positive shovel test), which was situated on a more generic terrace edge with extensive stretches of flat topography. Just as at 31HK2502 and 31HK2510, the actual magnitude of superimposition may be overestimated due to the potential for many of the Early Archaic occupations to consist of multiple elements from single Type II residences. This issue will be further developed in the next section, which examines the results of the Stage III Sample Block investigations. Not included in the population estimate was the Woodland period ceramic collection. Seventy-six Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 317 Chapter 7. 31HK2521 individual ceramic vessels are represented in the Stage I and II shovel test sample. Since we can expect most sherds from single vessels to be contained within an area of about 2 m in diameter, the Stage I and II shovel tests would be expected to achieve 16 percent coverage for sherd aggregate space. This would yield a total vessel population estimate of 475. Although this seems on the face of it a very high number of vessels/sherd aggregates, it may actually underestimate the total number of vessels present at the site since 233 vessels were identified in total from the Phase II investigation. The relative enormity of this estimate, and since vessel scatters are sometimes paired with debitage scatters and concentrations, it is likely that significant numbers of elements identified in the lithic subtype tally are Woodland in association. Thus to avoid the issue of double counting, ceramics were not included in the element population estimate presented above. The Stage I and II shovel test sample succeeded in identifying New River I, New River II, Hanover I, Hanover II, Hanover III, Yadkin III and Cape Fear III series. Using the proportions of individual vessels associated with each of these subseries (Table 85) it can be inferred that 9.2 percent of the Woodland occupation is affiliated with the Early Woodland period (New River I), 5.3 percent with the early Middle Woodland (New River II), 35.2 percent with the late Middle Woodland period (Hanover I and II) and 40.8 percent with the early Late Woodland (Hanover III and Yadkin III). Cape Fear III, which is postulated to represent an Early Mississippian intrusion into the area, comprised 10.5 percent of the vessel sample. Numerous studies have documented vertical patterning in the relative depth of deposits in the unconsolidated sandy soils of the Coastal Plain (Cable and Cantley 2005 b, 2006; Michie 1990). The actual depth patterns vary according to the character of the deposit. Those deposits with higher B-horizons tend to have compressed sequences, while those in deep C-horizon and E-horizon sediments generally display more discreet and vertically expansive sequencing. Michie’s (1990) C-horizon model developed in the Table 84. Recognized Elements for the Stage I and II Shovel Test Sample, 31HK2521. IDENTIFIED ELEMENTS FROM THE STAGE I AND II SHOVEL TEST SAMPLE: 31HK2521 RAW MATERIAL TYPES TYPE I TYPE II, III, IV GRAND TOTAL PERCENTAGE OF ELEMENTS Mill Mountain Rhyolite 1 14 15 6.7 Type I Rhyolite Tuff 4 17 21 9.3 Uwharries Eastern Rhyolite 4 17 21 9.3 Uwharries Southern Rhyolite (1) 1 25 26 11.6 Uwharries Southern Rhyolite (2) 6 57 63 28.0 Uwharries Western Rhyolite 1 16 17 7.6 White Quartz 4 56 60 26.7 Crystal Quartz 1 1 0.4 Meta-sedimentary 1 1 0.4 204 225 100.0 GRAND TOTAL 21 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 318 Chapter 7. 31HK2521 Waccamaw Neck area on the central coast of South Carolina begins with Mississippian and Woodland materials positioned in the upper 30 cm of sediment. This zone is underlain by ceramic Late Archaic occupations situated at 28 to 35 cm bs, Middle Archaic horizons at 35 to 55 cm bs, and Early Archaic cornernotched and side-notched components at 55 to 62 cm bs. This model is generally applicable to sites on Fort Bragg. However, the deposits at Fort Bragg appear to be somewhat more compressed, judging by the vertical positions of samples of diagnostic artifacts. At Fort Bragg three vertical groupings have been developed for deep Sandhills sites: (1) Late Archaic/ Middle Archaic between 15 and 40 cm bs, (2) Middle Archaic/Early Archaic between 35 and 50 cm bs, and (3) Early Archaic below 50 cm bs. The divisions are broadly constructed and realistically anticipate a certain degree of vertical mixing and overlap between the various Archaic periods. This vertical model can be of great utility in reconstructing the occupation history of sites like 31HK2521 where the predominant artifact class is non-diagnostic lithic chipped stone debitage. The vertical positions of the various elements recognized in the Stage I and II shovel test sample allow a rough picture of the representation of occupation periods at the site (Table 86). Most elements were situated in a single level, but some, especially the higher density elements, spanned several levels. In the latter cases, the central tendency of the vertical distribution of a particular element was used to estimate its position. Vertical data from the shovel test sample was not precisely congruent with the model ranges discussed above. However, the following adaptation was developed to correlate culture-historic association with level data: (1) 0-20 cm bs, Woodland, (2) 20-40 cm bs, Woodland/Late Archaic/Middle Archaic, (3) 40-50 cm bs, Middle Archaic/Early Archaic and (4) 50 to 70 cm bs, Early Archaic. Approximately 77 percent of the occupation elements are positioned above 30 cm bs. In light of the estimates for Woodland period occupation provided by the mean number of vessels present, it is likely that at least two-thirds of these elements are affiliated with the Woodland period. This would suggest that Woodland period occupation is represented by 50 to 55 percent of the identified elements. Definitive evidence of a Late Archaic presence was not produced during the investigation, suggesting that the Late Archaic contribution to the occupation history of the site was minimal. About 9.8 percent of the elements are unambiguously Early Archaic. However, the large number of scrapers recovered from the investigation relative to Middle Archaic diagnostics (n=1), it is likely that most of the elements positioned between 40 and 50 cm bs are Early Archaic in affiliation. This would suggest that approximately 20 percent of the elements are Early Archaic in age. Thus the remaining 20 to 25 percent of the occupation at the site is probably associated with Middle and Late Archaic phases. Type I elements are primarily concentrated above 40 cm bs, where they are inferred to be associated most strongly with Woodland and Middle Archaic phases (Table 87). Approximately 80 percent of the Type I residences are positioned within this vertical range. The remaining Type I residences are situated at depths more consistent with Early Archaic occupation. The relative proportions of Type I elements by vertical age grouping is nearly identical to the overall element profile, suggesting that 40 to 45 percent (or 50 percent of the Type I residences positioned between 0 and 40 cm bs) of the Type I residences on the site are affiliated with Woodland period occupations. SAMPLE BLOCK INVESTIGATIONS Eighteen shovel test outcomes from the Stage I and II sample were targeted for close-interval shovel testing to further elaborate the character of the occupations extant at the site. In general, a series of eight additional shovel tests was excavated at 1.25 m-intervals surrounding the target shovel test. These are referred to as sample blocks and their locations are illustrated in Figure 77. An attempt was made to distribute the sample blocks (SBs) evenly across the site to achieve a representative picture of the settlement record, but locations along the northern edge of the ridge nose were favored to more fully investigate the Woodland period occupation. Tables 88 through 91 display the vertical distributions of debitage by raw material type for each of the sample blocks. Exposed vertical profiles fell into three depth groups. Nine sample blocks (SBs 1, 3, 5, 6, 7, 12, 15, 16 and 17) contained profiles primarily Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 319 Chapter 7. 31HK2521 Table 85. Ceramic Vessel Representation by Ceramic Sub-series for Stage I and II Shovel Tests, 31HK2521. REPRESENTATION OF IDENTIFIED VESSELS BY CERAMIC SUB-SERIES FOR STAGE I AND II SHOVEL TEST SAMPLE: 31HK2521 CERAMIC SUB-SERIES VESSEL COUNT PERCENT REPRESENTATION New River I 7 9.2 New River II 4 5.3 Hanover I 11 14.5 Hanover II 15 19.7 Hanover III 20 26.3 Yadkin III 11 14.5 Cape Fear III 8 10.5 GRAND TOTAL 76 100.0 confined to the upper 40 cm of deposit and these were inferred to represent Woodland and, secondarily, Late and Middle Archaic occupation. Six sample blocks (SBs 2, 8, 9, 10, 13 and 14) exhibited profiles primarily situated between 20 and 50 cm bs, which were viewed as representative of the full range of Archaic occupation. The remaining three sample blocks (SBs 4, 11 and 18) displayed debitage profiles concentrated between 40 and 70 cm bs. These were considered to represent, primarily, Early Archaic occupations. Several factors intervene, however, to complicate this vertical model. The diagnostic stone tool distributions reveal that Early Archaic diagnostics are situated in higher than expected positions in some of the sample blocks (Table 92). This is clearly illustrated for SBs 3, 7 and 15. They display debitage profiles positioned between 0 and 40 cm bs, but contain large numbers of Early Archaic affiliated stone tools as well. These particular SBs are situated on the edge of the ridge nose slope (Figure 77), which appears to have been subjected to deflationary processes (i.e. slope erosion and aeolian redeposition) that reduced the vertical separation between cultural deposits of differing ages. Each of these SBs also contains large numbers of Woodland ceramics in the upper 30 cm of deposit (Table 93), which evidence long and intensive records of Woodland occupation spanning portions of the Early, Middle and Late Woodland periods. An indication of the intensity of occupation is supplied by the vessel counts from the shovel test samples in these blocks, which range between 8 and 12. Intensive and complex Woodland period occupation was documented also in SBs 4, 8, 9, 13 and 16. Most of these blocks are situated farther up on the top of the ridge, suggesting that the ridge top was protected from deflationary processes (Tables 88-91 and Table 92). Woodland deposits interface with Archaic occupations between 30 and 40 cm bs in the ridge top blocks. Relatively unmixed Early Archaic occupations were identified in SBs 4, 11 and 18. It is estimated that a mean of 475 individual ceramic vessels are represented on the site, based on Stage I and II shovel test results. There is good reason to believe, however, that this method of calculation underestimates the actual vessel population, as was discussed in depth in Chapter 5. It was also concluded from that discussion that it is unlikely that many full pot concentrations survived on these sites due to recycling of large sherds and off-site transport of the larger vessel fragments. An idea of the magnitude of the sherd population at the site can be supplied by calculating a straight area mean. The sample of 215 Stage I and II shovel tests sums to an area of 19.35 m2. The site area is approximately 6,100 m2, which Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 320 Chapter 7. 31HK2521 Table 86. Inferred Culture-Historic Associations of Recognized Lithic Elements, 31HK2521. INFERRED ASSOCIATIONS OF RECOGNIZED LITHIC ELEMENTS: 31HK2521 UWHARRIES SOUTHERN RHYOLITE (1) UWHARRIES SOUTHERN RHYOLITE (2) UWHARRIES WESTERN RHYOLITE WHITE QUARTZ 10 5 12 3 10 Woodland/Late & Middle Archaic 20-40 6 13 8 10 39 8 37 Middle & Early Archaic 40-50 2 1 2 7 6 4 6 Early Archaic & Paleoindian 50-90 1 1 1 4 6 2 7 15 21 21 26 63 17 60 GRAND TOTAL 1 1 1 1 PERCENTAGE UWHARRIES EASTERN RHYOLITE 6 GRAND TOTAL TYPE I RHYOLITE TUFF 6 Woodland META-SEDIMENTARY MILL MOUNTAIN RHYOLITE 0-20 INFERRED ASSOCIATION CRYSTAL QUARTZ VERTICAL RANGE (cm bs) LITHIC ELEMENTS 52 23.1 122 54.2 29 12.9 22 9.8 225 100.0 yields an area sample of 0.32 percent. One hundred and twelve sherds, weighing 625.29 gm, were recovered from the sample, resulting in a population mean estimate of 35,000 sherds and a total sherd weight of 195.4 kg. If whole pots were entirely represented in the assemblage, it is doubtful that this total would be sufficient to cover the weight of only 142 typical open-mouth jars from the region (see Herbert and Irwin 2003:2). Given an average weight of 3 pounds (1,377 gm) for a typical pot, if there were 475 whole pots present on the site, a mean population weight of 654 kg would be required. Consequently, only about 30 percent of the expected vessel weight relative to the estimate of the vessel population is present on the site. Similar results were obtained at 31HK2502 and 31HK2510 where only about 10 and 18 percent, respectively, of the expected sherd weight was found in the sample. This suggests that the intensity of use recycling is great on Sandhills sites and that large sherds from broken pots were probably recycled for other uses (see David and Hennig 1972:21-21) and stored in caches or transported upon leaving camps, while smaller sherds were probably left at the point of breakage. If so, sherds from the same pot might be spread across multiple sites and a high degree of displacement and disorganization should be evident in the spatial distribution of ceramics on sites with short-term occupation records. All of the sample blocks contain a wide variety of lithic raw material types. Much of this heterogeneity can be attributed to occupation superimposition, but some of it may result from site functional variability. Type II residences characteristically contain heterogeneous lithic raw material profiles and the high density of scraper forms in the stone tool inventory suggests that many of the occupations at the site may represent Early Archaic Type II residences. The sample block frames generally cover an area of only about 2.5 m square, which is smaller than the average element size. Moreover, it would be uncommon for the sample block to expose nearly whole elements. However, a method was devised during the DO4 investigation (Cable 2010) to identify and distinguish Type I and III elements that are mostly contained within the extent of a sample block. Three measures related to shovel test outcomes were developed to describe this variability (Table 94). These are: (1) number of positive shovel tests within which the element occurs, (2) number of shovel tests with greater than or equal to 5 pieces of debitage and (3) mean debitage frequency per positive test. Since Type I residential elements are characterized by high density debitage concentrations they are expected to yield high outcome values for each of these measures. Type III residences, by contrast, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 321 Chapter 7. 31HK2521 Table 87. Inferred Culture-Historic Associations of Recognized Type I Elements, 31HK2521. INFERRED ASSOCIATIONS OF RECOGNIZED TYPE I LITHIC ELEMENTS: 31HK2521 Middle & Early Archaic 40-50 Early Archaic & Paleoindian 50-90 GRAND TOTAL 1 META-SEDIMENTARY 6 28.6 11 52.4 4 1 1 2 9.5 1 2 9.5 21 100.0 4 are defined by the presence of low-density debitage scatters of about the same size as Type I high density scatters. Consequently, they should be expected to measure high or moderate on the number of positive shovel tests within which they are represented, but low on the number of shovel tests yielding five or more pieces of debitage of a specific raw material type and low on the mean frequency of debitage per positive shovel test. Given the structure of the sample blocks, there are generally nine outcomes (shovel tests) to measure. Some elements present measures that are intermediate between the two extremes and these can only be classified as Type I/III. Type II residential elements would be expected to reflect similar patterns to Type III residences, but they should contain multiple raw material types and a greater density of tools. Clearly, the identification of Type II residences in sample blocks is subjective and it would require additional confirmation from expanded sample blocks or test unit excavation to firmly identify such elements. Table 95 summarizes the results of applying the shovel test model to the elements defined by lithic raw material type within each sample block. Generally, only elements that were represented by at least three shovel tests were analyzed, as those with lower representation are regarded as insufficiently complete or ineffectively sampled in the sample block frame. As established by the Stage I and II element study, 4 1 6 4 CRYSTAL QUARTZ WHITE QUARTZ 1 2 1 1 UWHARRIES WESTERN RHYOLITE 1 UWHARRIES SOUTHERN RHYOLITE (2) UWHARRIES SOUTHERN RHYOLITE (1) 3 PERCENTAGE 20-40 1 GRAND TOTAL Woodland/Late & Middle Archaic UWHARRIES EASTERN RHYOLITE 0-20 TYPE I RHYOLITE TUFF Woodland MILL MOUNTAIN RHYOLITE INFERRED ASSOCIATION VERTICAL RANGE (cm bs) TYPE I LITHIC ELEMENTS 1 4 Type III elements predominate in the sample blocks, comprising 71.6 percent (n=53) of the sample. These elements are characterized by variably moderate to high shovel test representation (3 to 6 shovel tests), low frequencies of shovel tests containing greater than 5 pieces of debitage (0 to 1 shovel test) and debitage frequency means in positive shovel tests of less than 3.0. Stone tools are scarce. Type I elements were slightly more numerous than was predicted by the Stage I and II sample, but anticipated by an acknowledgement of the limitations of the method. Type I elements identified in the sample blocks numbered thirteen of the 74 identified elements (17.6 percent). These elements commonly had high shovel test representation (4 to 9 shovel tests), greater numbers of high debitage density shovel test outcomes (2 to 4 shovel tests) and high mean debitage densities, which ranged between 3.67 and 21.00 pieces of debitage. Approximately 10.8 percent (n = 8) of the elements exhibited intermediate characteristics and could not be further differentiated. Cultural-chronological associations were inferred from depth profiles and from diagnostic stone tools that were matched to the raw material composition of the debitage concentrations. Sometimes diagnostic stone tool associations supplanted inferences based on depth patterns. In particular, some elements that occupied high vertical positions were assigned Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 322 Chapter 7. 31HK2521 Table 88. Vertical Distribution of Debitage, Sample Blocks 1—5, 31HK2521. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 1-5:31HK2521 SB1 LEVEL RAW MATERIAL 1 2 Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 1 1 2 6 2 5 1 2 10 3 4 1 2 1 1 2 6 4 5 SB2 RAW MATERIAL LEVEL 1 2 Type I Rhyolite Tuff Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 3 4 5 1 1 3 5 1 6 7 2 2 SB3 RAW MATERIAL hydrated metavolcanic Metasedimentary, Green Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL Crystal Quartz hydrated metavolcanic Metasedimentary, Green Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 2 3 4 hydrated metavolcanic Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 7 8 5 6 7 8/9 1 1 1 1 1 1 2 6 7 44 49 1 1 2 5 1 29 39 1 1 1 13 15 9 10 7 7 1 4 1 2 GRAND TOTAL 1 2 11 14 1 1 4 5 6 7 8 1 5 1 2 8 GRAND TOTAL 1 1 1 2 2 4 11 1 110 133 LEVEL 1 2 3 1 2 1 1 1 1 3 3 7 2 1 18 3 5 1 2 11 7 9 1 1 22 1 14 SB5 RAW MATERIAL 6 LEVEL SB4 RAW MATERIAL GRAND TOTAL 2 1 10 1 2 10 26 5 1 13 2 9 2 1 1 4 29 1 4 6 7 8/9 GRAND TOTAL 1 2 1 1 5 34 7 39 2 9 101 LEVEL 1 2 1 1 1 3 4 5 1 8 2 2 13 4 2 1 9 1 13 8 3 25 5 1 1 7 1 GRAND TOTAL 1 2 1 1 31 12 7 55 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 323 Chapter 7. 31HK2521 Table 89. Vertical Distribution of Debitage, Sample Blocks 6—9, 31HK2521 VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 6-9:31HK2521 SB6 RAW MATERIAL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL LEVEL 1 2 3 4 2 2 3 2 5 3 2 10 5 22 1 2 3 1 3 1 5 10 4 5 1 1 2 3 3 9 14 5 13 2 1 1 3 2 1 3 13 2 3 4 4 SB8 Crystal Quartz Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 6 GRAND TOTAL 2 1 6 3 18 9 39 LEVEL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL RAW MATERIAL 6 2 SB7 RAW MATERIAL 5 7 8/9 8/9 1 1 2 3 2 5 6 7 2 1 4 17 1 1 1 3 3 1 2 5 6 7 8/10 1 1 1 GRAND TOTAL 3 1 10 7 6 1 17 45 LEVEL 1 1 2 1 1 10 1 1 3 1 1 3 2 18 7 1 2 7 1 1 19 SB9 3 4 5 1 GRAND TOTAL 2 1 22 2 8 18 4 9 66 LEVEL RAW MATERIAL 1 2 hydrated metavolcanic Metasedimentary, Green Mill Mountain Rhyolite Schist Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 1 3 4 1 1 1 1 2 1 1 1 2 7 3 2 5 2 13 3 7 3 16 2 2 33 3 2 2 7 2 15 4 1 1 3 2 6 GRAND TOTAL 2 1 3 1 10 10 8 35 2 6 78 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 324 Chapter 7. 31HK2521 Table 90. Vertical Distribution of Debitage, Sample Blocks 10—14, 31HK2521 VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 10-14:31HK2521 SB10 RAW MATERIAL LEVEL 1 hydrated metavolcanic Metasedimentary, Green Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 2 3 4 1 2 2 1 1 2 2 1 1 4 2 5 3 Mill Mountain Rhyolite Eastern Uwharries Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 2 11 2 6 3 1 4 3 19 18 Mill Mountain Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 2 3 4 1 1 2 1 1 5 1 1 9 1 1 1 3 5 7 4 2 1 9 Crystal Quartz hydrated metavolcanic Metasedimentary, Green Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 2 Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 8 1 1 4 3 2 2 4 17 1 1 1 5 6 7 3 3 5 1 3 1 4 6 6 7 6 7 3 4 5 4 GRAND TOTAL 3 1 4 5 15 11 11 10 9 69 GRAND TOTAL 1 4 1 26 5 4 41 11 11 11 11 5 5 GRAND TOTAL 1 1 1 54 57 1 1 1 14 16 13 14 LEVEL 1 2 3 1 2 2 3 4 3 3 1 2 17 5 6 2 6 9 28 4 5 1 1 1 4 5 3 9 3 2 27 2 2 1 5 9 1 20 SB14 RAW MATERIAL 8 LEVEL SB13 RAW MATERIAL 7 LEVEL 1 SB12 RAW MATERIAL 6 1 SB11 RAW MATERIAL 5 1 2 3 8/9 GRAND TOTAL 1 1 1 14 18 9 27 22 5 98 LEVEL 1 2 3 4 1 1 1 1 6 2 2 1 8 2 5 5 1 3 1 1 6 6 1 1 2 7 GRAND TOTAL 2 2 13 1 5 23 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 325 Chapter 7. 31HK2521 Table 91. Vertical Distribution of Debitage, Sample Blocks 15—18, 31HK2521 VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 15-18:31HK2521 SB15 RAW MATERIAL Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL LEVEL 1 2 3 4 1 1 3 2 6 10 1 4 7 8 5 1 8 14 1 2 hydrated metavolcanic Mill Mountain Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 1 2 1 1 6 11 1 2 3 4 5 6 2 2 2 2 7 8/9 1 1 5 6 4 6 GRAND TOTAL 2 1 10 2 26 41 GRAND TOTAL 3 3 2 1 20 29 LEVEL 1 1 1 2 3 4 1 5 1 4 1 2 4 2 14 1 2 17 3 28 GRAND TOTAL 3 9 1 4 21 6 44 5 1 SB18 hydrated metavolcanic Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 1 8/9 1 SB17 RAW MATERIAL 1 3 4 7 LEVEL RAW MATERIAL hydrated metavolcanic Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 6 1 SB16 RAW MATERIAL 5 LEVEL 1 2 3 4 1 1 6 2 1 2 1 14 3 2 24 3 7 13 32 5 1 1 3 10 8 2 25 6 7 8/10 5 3 1 1 9 1 1 5 3 9 5 23 GRAND TOTAL 2 7 9 3 15 44 29 5 114 to the Early Archaic period based on diagnostic stone tool associations. This adjustment was found to be necessary in a relatively large number of blocks (i.e. SBs 2, 3, 4, 7, 10, 11, 14 and 15. This incongruity is viewed to be the result of primarily two factors, deflation from slope erosion and Woodland period intrusions into earlier Archaic deposits resulting in upward movement of portions of the occupation floors. Twenty-eight Early Archaic, twelve Middle Archaic/Early Archaic, one Middle Archaic, one Late Archaic, six Late Archaic/Middle Archaic, nine Middle Archaic/Late Archaic Woodland and eleven Woodland lithic elements were identified in the study. Candidates for Type II elements were recognized in SBs 2, 3, 7, 8, 9, 10, 13, 14, 15 and 18 (see Table 95). In these instances, groups of Type III elements could hypothetically belong to single Type II residences. Since such elements have only been positively associated with the Early Archaic period, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 326 Chapter 7. 31HK2521 Table 92. Vertical Distribution of Stone Tools by Sample Block, 31HK2521. VERTICAL DISTRIBUTION OF STONE TOOLS IN SAMPLE BLOCKS: 31HK2521 STONE TOOL TYPE LEVELS 1 2 3 4 5 6 7 8 ASSOCIATION SB1 Flake Blank 1 SB2 Type IIa End Scraper 1 Early Archaic SB3 Core Frag. 1 Flake Blank 1 Scraper Frag. 1 Early Archaic Spoke Shave 1 Early Archaic 1 Type IIb End Scraper 1 Utilized Flake Early Archaic 1 SB4 Directional Core 1 Hafted Biface Tool 1 Nutting Stone 1 Type IIa End Scraper 1 Type III/IV Biface Frag. 1 Worked Schist 1 Early Archaic SB5 Flake Blank 1 Type II Biface Frag. 1 SB6 PPK Fragment 1 1 Utilized Flake 1 SB7 Core Fragment 1 Denticulate 1 Flake Blank Early Archaic 1 Morrow Mountain I St. 1 PPK Fragment Middle Archaic 1 Retouched Flake 1 Type II Biface Frag. 1 Type III Side Scraper 1 Early Archaic Unidirectional Core 1 Utilized Flake 1 SB8 Core Frag. Hammerstone 1 1 Type Ib End Scraper 1 Type IIa Side Scraper 1 Type III Biface Frag. Undef, Rudimentary St. Early Archaic Early Archaic 1 1 Middle Woodland? only elements of this inferred affiliation were considered for this procedure. If groups of Type III Early Archaic elements belong to single occupations, the representation of Early Archaic occupations would be significantly diminished in the sample block inven- tory. Type II residences are characterized by lowdensity scatters of debitage composed of multiple raw material types and relatively high densities of stone tools of heterogeneous raw material composition. It is suggested that these elements represent multiple Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 327 Chapter 7. 31HK2521 Table 92. Vertical Distribution of Stone Tools by Sample Block, 31HK2521 (Continued). VERTICAL DISTRIBUTION OF STONE TOOLS IN SAMPLE BLOCKS: 31HK2521 STONE TOOL TYPE LEVELS 1 2 3 4 5 6 7 8 ASSOCIATION SB9 Flake Blank 1 Pebble 2 Type I Biface Frag. 1 Type I Side Scraper 1 Type IIb Side Scraper Early Archaic 1 Early Archaic SB10 Core Frag. 1 Flake Blank 1 Hammerstone 1 Type IIa Side Scraper 1 Undef. Expanded St. Early Archaic 1 Late Woodland? SB11 Type IIb Side Scraper 1 Utilized Flake Early Archaic 1 SB12 Core Frag. 1 Soapstone Sherd Type II Biface Frag. 1 Late Archaic 1 SB13 Core Frag. 1 Denticulate 1 Hammerstone Early Archaic 1 Utilized Flake 1 Yadkin Lg. Triangular, Var A 1 1 Middle Woodland SB14 Palmer I Corner Notched 1 Early Archaic SB15 Scraper Frag. 1 Type I Side Scraper Utilized Flake Early Archaic 1 1 Early Archaic 1 SB16 PPK Frag. 1 SB18 Badin Crude Triangular, Var C 1 Flake Blank Graver 1 Early Archaic Type IIa Side Scraper 1 Early Archac Utilized Flake Early Woodland 1 family units that came together for short durations to conduct activities such as communal hunts that resulted in abnormally high tool discard rates related to processing animals. Each of the identified sample blocks meets these criteria. High numbers of tools are found with multiple “Type III” debitage scatters of a depth or association to be consistent with an Early Archaic affiliation. Type II elements are rare 2 and not well understood in the region. Similar magnitudes of Type II candidates of Early Archaic age were identified at 31HK2502 and 31HK2510. Disturbance processes have severely impacted the Early Archaic zone of 31HK2521, however. The heavily reoccupied nature of the Woodland zone also makes it difficult to associate particu- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 328 Chapter 7. 31HK2521 Table 93. Vertical Distribution of Ceramic Types in Sample Blocks, 31HK2521. VERTICAL DISTRIBUTION OF CERAMIC TYPES IN SAMPLE BLOCKS: 31HK2521 CERAMIC TYPE LEVELS 1 2 3 4 5 6 7 Vessel # 8 SB1 New River If Net Impressed 1 V133 SB2 1 Yadkin IIIe Fabric Impressed V118 1 Yadkin IIIf Cord Marked V117 SB3 V126 Hanover Ia Fabric Impressed 1 Hanover Ib Check Stamped 1 1 V24, V25 Hanover IIa Linear Check St. 1 1 V124 Hanover IIIe Cord Marked 2 Hanover IIa Check Stamped V122 1 V125 1 New River Ia Cord Marked Residual Sherd V234 1 Yadkin IIId Fabric Impressed 1 Yadkin IIId Indet. 1 Yadkin IIIe Cord Marked 2 V235 Yadkin IIIe indet. Yadkin IIIf Plain V235 1 4 V120, V125 1 V123 V121 1 SB4 Hanover Ib Cord Marked Hanover Ib Fabric Impressed Hanover IIa Fabric Impressed 2 2 V112 1 1 V200 1 V111 V113 1 Hanover IIb Fabric Impressed 1 Hanover IIIa Fabric Impressed 1 Hanover IIIb Fabric Impressed V115 1 Hanover IIIe Cord Marked 2 Hanover IIIe Fabric Impressed 2 Hanover IIIe indet. 1 V108 1 V15 4 V15 Residual Sherd 2 Yadkin IIIc/e Fabric Impressed 1 Yadkin IIIc/e indet. decorated V116 V228 1 Yadkin IIIc/e/f Fabric Impressed 1 Yadkin IIIe Fabric Impressed 1 V114 1 Yadkin IIIe indet. 1 V109, V110, V232 2 V109 SB5 New River Ia Indet. 2 V30 lar debitage scatters and concentrations with specific ceramic concentrations. A review of Table 93 shows that nearly all of the sample blocks are characterized by a wide range of sub-series and the associated shred aggregations at different locations appear to represent only small portions of vessels. The earlier vessel population projections and estimated sherd weight calculations point to a generally displaced and disorganized ceramic deposit and the results of the sample block investigations strongly support this inference. One of the objectives of the test unit investigation was to further evaluate the degree of ceramic vessel organization and completeness at the site. TEST UNITS Eleven 1-x-1-m test units were excavated at 31HK2521 (Table 96). Each unit was quartered and Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 329 Chapter 7. 31HK2521 Table 93. Vertical Distribution of Ceramic Types in Sample Blocks, 31HK2521 (Continued). VERTICAL DISTRIBUTION OF CERAMIC TYPES IN SAMPLE BLOCKS: 31HK2521 CERAMIC TYPE LEVELS 1 2 3 1 1 4 5 6 7 Vessel # 8 SB7 Hanover IIa Fabric Imp. Hanover IIIc/e Dowell Imp. 1 Hanover IIIc/e Fabric Imp. 4 V100 V105 V104 Hanover IIIe Cord Marked 2 V99, V103 New River IIa Cord Marked 1 V106 New River IIa indet. 1 V106 Yadkin IIIc/e Wide Simple St 1 Yadkin IIIc/e/f Fabric Imp. 1 V107 V101 Yadkin IIIe Cord Marked 2 Yadkin IIIe Plain 1 V38 1 V102 Hanover Ib Check Stamped 1 V35 Hanover IIa Fabric Impressed 3 V36 Hanover IIa indet. 1 Yadkin IIIf indet. SB8 Hanover IIa indet. decorated 6 1 Hanover IIIa Linear Check St. 3 2 Hanover IIIa Fabric Impressed 2 V80 1 V79 V166 SB9 Cape Fear III Cord Marked 1 Hanover IIa Cord Marked 1 Hanover IIa Fabric Impressed Hanover IIa indet. 1 Hanover IIa Linear Check St. 1 V97 1 2 V94, V95 1 1 2 V98 1 1 V95, V96 Hanover IIa Fab. Imp./Chk. St. V96 Hanover IIIa Cord Marked 1 V39 Hanover IIIa Plain 1 V40 1 V89, V93 Hanover IIIe indet. decorated 1 V90 New River Ia Indet. 1 V91 Hanover IIIe Cord Marked V92 1 Hanover Ia Indet. 1 excavated in arbitrary 10 cm levels. The units were terminated above the Bt-horizon to accommodate the exposure of larger horizontal areas to follow out the distributions of targeted sherd aggregates. The precise location of each test unit was determined after closeinterval shovel testing was deployed around targeted shovel tests. Test units were placed at four locations and were arranged in three blocks of contiguous test units along with one isolated unit (Figure 78). Block 1 consisted of two 1-x-1-m test units (TUs 4 and 9) placed in SB 8 to further investigate the structure and vertical integrity of deeply buried, “in position” Early Archaic elements. Block 2 consisted of a linear arrangement of four 1-x-1-m test units (TUs 1, 6, 10 and 11) placed across SB 3 to investigate an area containing a wide range of ceramic types and dense quartz debitage concentration apparently positioned in the Woodland zone of the site. Block 3 included four 1-x-1-m test units (TUs 2, 5, 7 and 8) arranged in a 2-x-2-m square. This block was placed over SB 16 on the tip of the ridge nose where an extensive census of vessels, 17 were identified in the shovel test sample, were documented representing New River, Hanover and Yadkin ceramic types. Both Blocks 2 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 330 Chapter 7. 31HK2521 Table 93. Vertical Distribution of Ceramic Types in Sample Blocks, 31HK2521 (Continued). VERTICAL DISTRIBUTION OF CERAMIC TYPES IN SAMPLE BLOCKS: 31HK2521 CERAMIC TYPE LEVELS 1 2 3 4 5 6 7 Vessel # 8 SB10 Hanover Ib Indet. 1 Hanover Ib indet. stamped 2 Hanover IIa indet. 2 Hanover IIa Plain V88 V88 1 New River Ia Indet. V87 1 Hanover IIIa Cord Marked V84 1 V83 1 New River Ia indet. decorated V83 1 Yadkin IIIc Fabric Impressed 2 Yadkin IIIc/f Fabric Impressed 1 V85 1 Yadkin IIIc/f indet. V81 1 V81 4 Yadkin IIId Indet. 1 Yadkin IIIe Cord Marked 1 Yadkin IIIe Fabric Impressed 1 Yadkin IIIe indet. 1 V82 V86 V37 V37 Yadkin IIIe indet. stamped V231 1 SB11 Cape Fear III Fabric Impressed 4 Hanover IIb indet. 1 V161 1 V163 Hanover IIb indet. decorated 3 V163 Hanover IIIa Fabric Impressed 1 V165 Hanover IIIb Fabric Impressed V162 1 Hanover IIId Wide Simple St. V164 1 SB12 Hanover IIa Check Stamped 1 Hanover IIIa indet. decorated 1 1 1 1 1 New River Ia Indet. 1 Yadkin IIIc/e indet. 1 1 V47 V139 1 1 Yadkin IIIc/e Fabric Impressed V138, V141 V140 Hanover IIIa Wide Simple St. Hanover IIIe Fabric Impressed V142 1 Hanover IIa indet. decorated V137 V135 Yadkin IIId Cord Marked 2 V136 Yadkin IIIf indet. 1 V48 and 3 were established to further evaluate the degree of ceramic vessel organization and completeness characterizing the ceramic deposits at the site. The non-sequential test unit numbering system associated with each block is an artifact of field decisions to expand investigations in each of these block areas at different times. Finally, TU3 was established in the area of SB 14 where a discrete Hanover Cord Marked sherd aggregate was identified in an effort to evaluate vessel organization at a location of lighter Woodland period deposit superimposition. The results for each test unit are presented below. Block 1 (TUs 4 & 9/SB 8) TUs 4 and 9 were placed contiguous to one another at grid coordinates N504.5/E515.5 (NE corner) and N505.5/E515.5 (NE corner), respectively, in the north central portion of SB 8. Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of the unit (Figure 108). Two soil layers were exposed in the wall profiles. The A-horizon was positioned in the upper 5 to 10 cm of the soil profile. The contact with the lower E-Horizon was irregular, suggesting that it was not representative Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 331 Chapter 7. 31HK2521 Table 93. Vertical Distribution of Ceramic Types in Sample Blocks, 31HK2521 (Continued). VERTICAL DISTRIBUTION OF CERAMIC TYPES IN SAMPLE BLOCKS: 31HK2521 CERAMIC TYPE LEVELS 1 2 3 4 5 6 7 Vessel # 8 SB13 Hanover IIa Cord Marked V131 2 Hanover IIa indet. decorated 1 V129 Hanover IIIa Cord Marked 1 V130 Hanover IIIe Cord Marked 1 V51 Hanover IIIe Plain 1 V50 Yadkin IIIc/e Dowell Impressed Yadkin IIIf indet. V132 1 V49 1 SB14 Hanover IIa Cord Marked Hanover IId Cord Marked V145 1 4 Yadkin IIIf Dowell Impressed 2 V146 1 V147 1 SB15 1 Hanover Ib Indet. 1 V237 1 V60 Hanover IIb Check Stamped 9 Hanover IIb Linear Check St. Hanover IIIa Cord Marked 2 1 New River Ia Indet. Yadkin IIIe Cord Marked Yadkin IIIe indet. stamped V124 2 V59 1 V127 V128 2 Yadkin IIIe indet. V61 Hanover Ib Cord Marked 1 2 of a plow zone. The A-Horizon was comprised of a pale brown (10YR 6/3) sand matrix that was partially stained a very dark grayish brown (10YR 3/2) from leached charcoal derived from prescribed burn episodes. Immediately below the A-Horizon was a 5 to 7 cm thick, strong brown (7.5YR 5/6) oxidized zone (E’) in the top of the E-Horizon that was formed during episodes of prescribed burning on the site. The E-Horizon extended below the base of the unit, which was terminated at 60 cm bd. Probing indicated that the Bt-Horizon was situated at a depth of 67 to 70 cm bd. The E-Horizon was comprised of light yellowish brown (10YR 6/4) sand. A large tree root stain composed of brown (10YR 5/3) loamy was encountered in the southeast quadrant of TU4 and was visible in the east wall profile of the block. The character of the fill suggests that the disturbance might have occurred in precontact times. The unit was terminated above the Bt-Horizon because the targeted Early Archaic deposits had been sufficiently recovered and the objectives of the investigation had been met. V236 SB 8 was established around a Stage II shovel test that yielded a Type Ib End Scraper (Figure 82:B) at a depth of 40 to 50 cm bs. Debitage from a diverse set of raw materials were spread between 20 and 80 cm bs in the shovel test and check stamped and fabric impressed sherds (n=3) from two Hanover vessels were recovered between 10 and 20 cm bs. Based on this evidence it was inferred that this location contained a complex superimposition of Woodland and Archaic components. Owing to its topographic position on top of the ridge, it was anticipated that investigation here would provide an opportunity to evaluate the vertical separation of chronologically ordered cultural debris in a favorable or optimal depositional setting. The recovery of the end scraper also suggested that investigation here might present a window into the nature and functional configuration of Early Archaic occupation at the site. The investigation in SB 8 succeeded in identifying five spatially extensive lithic debitage elements Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 332 Chapter 7. 31HK2521 Table 93. Vertical Distribution of Ceramic Types in Sample Blocks, 31HK2521 (Continued). VERTICAL DISTRIBUTION OF CERAMIC TYPES IN SAMPLE BLOCKS: 31HK2521 CERAMIC TYPE LEVELS 1 2 7 Vessel # 3 4 5 6 8 Hanover Ib Indet. 1 1 1 1 Hanover IIa Check Stamped 1 V69 1 V157, V180 SB16 Hanover Ib Fabric Impressed V67 1 Hanover IIa Fabric Impressed 1 Hanover IIa indet. 1 V68 Hanover IIa indet. decorated 1 Hanover IIb Check Stamped 1 Hanover IIIa Fabric Impressed 1 Hanover IIIb Cord Marked 1 2 2 Hanover IIId indet. 1 Hanover IIIe Fabric Impressed V149 V159 1 Hanover IIId Fabric Impressed V180 V158 1 Hanover IIIb Plain V67 1 1 V154 V154 3 1 V155 1 New River Ia Plain 1 New River IIa Cord Marked Residual Sherd 1 Yadkin IIId Indet. 1 Yadkin IIIe Cord Marked Yadkin IIIe Fabric Impressed 1 Yadkin IIIe indet. 1 1 V156 V160 1 1 1 V152 2 V148 1 V151 V153 SB17 Hanover IIa Check Stamped 1 Hanover IIb Check Stamped 1 V143 1 V144 SB18 Hanover IIIa Cord Marked V134 1 Yadkin IIIc/e Dowell Impressed V132 1 Table 94. Shovel Test Outcome Model for the Purpose of Identifying Element Types. INFERRED ELEMENT TYPE NO. POSITIVE STPS NO. STPS WITH ≥ 5 DEBITAGE MEAN DEBITAGE/ POSITIVE TEST TYPE I HIGH HIGH HIGH TYPE I OR TYPE III HIGH MODERATE MODERATE TYPE III MODERATE LOW LOW Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 333 Chapter 7. 31HK2521 Table 95. Inferred Element Types by Sample Blocks, 31HK2521. MEAN DEBITAGE/ + SHOVEL TEST 4 1 2.50 III W White Quartz 5 1 2.00 III W/LMA White Quartz 5 1 2.20 III Type I Rhyolite Tuff 1 0 1.00 III White Quartz 9 4 12.22 Uwharries Southern Rhyolite (1) 3 0 1.33 ASSOCIATION NO. STs WITH ≥ 5 DEBITAGE Uwharries Eastern Rhyolite ELEMENTS INFERRED ELEMENT TYPE NO. OF STs REPRESENTED TYPE II CANDIDATE INFERRED ELEMENT TYPES IDENTIFIED IN SAMPLE BLOCKS: 31HK2521 STONE TOOLS SB1 Flake Blank SB2 MA/EA x EA Type IIa End Scraper I W/LMA Core Frag. III LA/MA SB3 Uwharries Southern Rhyolite (2) 7 0 1.57 III Mill Mountain Rhyolite 1 0 1.00 III II LA/MA Flake Blank Flake Blank, Utilized Flake Type I Rhyolite Tuff 2 0 1.00 III II Scraper frag. Uwharries Western Rhyolite 1 0 1.00 III II Uwharries Eastern Rhyolite 2 0 1.00 III II Type I Rhyolite Tuff 4 0 1.25 III Uwharries Eastern Rhyolite 8 3 4.25 I EA Uwharries Southern Rhyolite (1) 1 0 3.00 III EA Uwharries Southern Rhyolite (2) 9 3 4.33 I White Quartz 6 0 1.50 III Spoke Shave EA Type IIb End Scraper SB4 MA/EA Type IIa End Scraper Type III/IV Biface frag., Hafted Biface Tool MA/EA Directional Core Chert Cemented Sandstone Nutting Stone Schist Worked Schist SB5 Uwharries Southern Rhyolite (2) 7 2 4.43 I W/LMA Uwharries Western Rhyolite 3 1 4.00 I/III W White Quartz 3 1 2.33 III W Uwharries Eastern Rhyolite 1 0 1.00 III W/LMA Uwharries Eastern Rhyolite 3 0 2.00 III LA/MA Uwharries Southern Rhyolite (2) 5 1 3.60 I/III LA/MA White Quartz 5 0 1.80 III W/LMA Type II Biface frag. Flake Blank SB6 PPK frag. Uwharries Southern Rhyolite (1) 2 0 1.50 III LA/MA PPK frag. Type I Rhyolite Tuff 1 0 1.00 III LA/MA Utilized Flake Uwharries Eastern Rhyolite 6 0 1.67 III II W/LMA Type II Biface frag. Uwharries Southern Rhyolite (1) 6 0 1.17 III II EA Type III Side Scraper, UF, RF, Denticulate Uwharries Southern Rhyolite (2) 3 0 2.00 III II EA Flake Blank White Quartz 5 1 3.40 III W/LMA Unidirectional Core, Core frag., PPK frag. MA Morrow Mountain I Stemmed SB7 Uwharries Western Rhyolite Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 334 Chapter 7. 31HK2521 Table 95. Inferred Element Types by Sample Blocks, 31HK2521 (Continued). MEAN DEBITAGE/ + SHOVEL TEST 6 2 3.67 I Uwharries Southern Rhyolite (1) 4 0 2.00 III x EA Type Ib End Scraper Uwharries Southern Rhyolite (2) 5 1 3.60 I/III x EA Type IIa Side Scraper Uwharries Western Rhyolite 4 0 1.00 III White Quartz 5 0 1.80 III ASSOCIATION NO. STs WITH ≥ 5 DEBITAGE Type I Rhyolite Tuff ELEMENTS INFERRED ELEMENT TYPE NO. OF STs REPRESENTED TYPE II CANDIDATE INFERRED ELEMENT TYPES IDENTIFIED IN SAMPLE BLOCKS: 31HK2521 STONE TOOLS SB8 MA/EA Type III Biface frag. W Undef. Rudimentary Stemmed, Core frag. Hammerstone Quartz Cobble SB9 Type I Rhyolite Tuff 5 1 2.00 III Uwharries Eastern Rhyolite 4 1 2.50 III x EA Uwharries Southern Rhyolite (1) 4 0 2.00 III x EA Uwharries Southern Rhyolite (2) 8 3 4.38 I EA White Quartz 6 0 1.00 III W/LMA Mill Mountain Rhyolite 2 0 1.50 III x EA Type I Biface frag. Uwharries Western Rhyolite 2 0 1.00 III x EA Flake Blank W/LMA Pebbles (2) W Undef. Expanded Stemmed Quartz Cobble Type I Side Scraper Type IIb Side Scraper SB10 Mill Mountain Rhyolite 3 0 1.33 III Type I Rhyolite Tuff 4 0 1.25 III Uwharries Eastern Rhyolite 3 1 5.00 I/III x EA Uwharries Southern Rhyolite (1) 5 1 2.20 III x EA Uwharries Southern Rhyolite (2) 6 0 1.83 III x EA Uwharries Western Rhyolite 4 0 2.50 III x EA White Quartz 5 0 1.80 III W Grano-Diorite Type IIa Side Scraper, Flake Blank Core Frag. Hammerstone SB11 Uwharries Southern Rhyolite (2) 6 3 4.33 I Uwharries Western Rhyolite 4 0 1.25 III Uwharries Southern Rhyolite (1) 1 0 1.00 III 6 3 9.00 I Utilized Flake EA Type IIb Side Scraper W Type II Biface frag., Core frag. LA Soapstone Vessel frag. SB12 White Quartz Steatite SB13 Type I Rhyolite Tuff 7 1 2.00 III x MA/EA Uwharries Eastern Rhyolite 6 1 3.00 III x MA/EA Uwharries Southern Rhyolite (1) 4 1 2.25 III x MA/EA Core frag. Uwharries Southern Rhyolite (2) 9 1 3.00 III x MA/EA Utilized Flake (2), Denticulate Uwharries Western Rhyolite 5 1 4.40 1/III x MA/EA White Quartz 5 0 1.00 III Quartz Cobble W Yadkin Large Triangular, Var. A Hammerstone Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 335 Chapter 7. 31HK2521 Table 95. Inferred Element Types by Sample Blocks, 31HK2521 (Continued). NO. STs WITH ≥ 5 DEBITAGE MEAN DEBITAGE/ + SHOVEL TEST INFERRED ELEMENT TYPE TYPE II CANDIDATE Uwharries Southern Rhyolite (2) 7 1 1.86 III x EA White Quartz 4 0 1.25 III x MA/EA Uwharries Southern Rhyolite (2) 6 0 1.67 III x EA Utilized Flake (2) White Quartz 7 2 3.71 I x EA Type I Side Scraper Type I Rhyolite Tuff 2 0 1.00 III x EA Scraper frag. 5 2 4.00 I MA/EA PPK frag. Type I Rhyolite Tuff 2 1 4.50 I/III W Uwharries Western Rhyolite 1 1 21.00 I W Type I Rhyolite Tuff 5 0 1.80 III x EA Uwharries Southern Rhyolite (1) 4 1 3.75 I/III x EA Uwharries Southern Rhyolite (2) 8 3 5.50 I x EA Uwharries Western Rhyolite 4 2 7.25 I x EA Utilized Flake White Quartz 3 0 1.33 III x EA Type IIa Side Scraper White Quartz 1 0 1.00 III W Badin Crude Triangular, Var C, Mill Mountain Rhyolite 1 1 7.00 I/III x EA Graver Uwharries Eastern Rhyolite 2 0 1.50 III x MA/EA Utilized Flake ELEMENTS ASSOCIATION NO. OF STs REPRESENTED INFERRED ELEMENT TYPES IDENTIFIED IN SAMPLE BLOCKS: 31HK2521 STONE TOOLS SB14 Palmer I Corner Notched SB15 SB16 White Quartz SB17 SB18 Flake Blank (Table 95). Most of these were classified as Type III debitage scatters and two produced Early Archaic scrapers matched to the USR(1) and USR(2) scatters. These latter two scatters were situated between 30 and 50 cm bs in vertical positions consistent with an Early Archaic association (Table 97). The vertical profiles of the various elements as reflected by debitage of specific raw material types, however, were widely dispersed within the sediment, suggesting a significant degree of deposit migration. A debitage concentration of Type I RT exhibited the characteristics of a Type I element. It was positioned between 20 and 40 cm bs, slightly higher than the positively identified Early Archaic elements (Tables 95 and 97) and it was assigned a Middle/Early Archaic age by virtue of this slight difference in vertical position. White quartz debitage was scattered between 20 and 80 cm bs (Table 97). This element was assigned a Woodland age based on the recovery of a projectile point classified as Undefined Rudimentary Stemmed (Figure 81:K). The wide vertical dispersion of the quartz profile, however, suggested that either extensive migration of deposits had occurred or that superimposed white quartz elements were present in the matrix at this location. Horizontal separation of white quartz debitage by level, however, was not readily identified. Ceramics recovered from SB 8 belonged to each of the three Hanover sub-series (Table 98). Five separate vessels were represented in the collection of 20 sherds. They were concentrated between 10 and 20 cm bs and appeared to contact Archaic deposits between 20 and 30 cm bs. None of the lithic elements could be associated with the ceramics based on vertical distributions (Table 97). The Type I RT element was found highest in the matrix, but it was clearly Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 336 Chapter 7. 31HK2521 Table 96. Summary Data for Test Units, 31HK2521. SUMMARY OF TEST UNITS: 31HK2521 SIZE NORTH EAST NO. LEVELS MAX. DEPTH (cm bd) VOLUME 3 (m ) ARTIFACT DEPTH (cm bd) TU3 1x1m 504.00 497.00 4 40 0.40 40 1 TU4 1x1m 504.50 515.50 6 60 0.60 60 1 TU9 1x1m 505.50 515.50 6 60 0.60 60 2 TU1 1x1m 509.00 529.50 4 40 0.40 40 2 TU6 1x1m 509.00 530.50 4 40 0.40 40 2 TU10 1x1m 509.00 528.50 4 40 0.40 40 2 TU11 1x1m 509.00 531.50 4 40 0.40 40 3 TU2 1x1m 517.00 535.50 4 40 0.40 40 3 TU5 1x1m 516.00 535.50 4 40 0.40 20 3 TU7 1x1m 517.00 536.50 4 40 0.40 40 3 TU8 1x1m 516.00 536.50 4 40 0.40 40 BLOCK TEST UNIT positioned below the ceramic zone between 20 and 40 cm bs. Based on artifact distributions in SB 8, TU4 was placed in the central portion of the sample block, immediately over the location of the Stage II anchor shovel test. TU9 was later appended to the north side of TU4 to follow out the Early Archaic deposits. Ceramics in the test units were centered between 10 and 20 cm bs, but a significant quantity occurred between 20 and 30 cm bs as well (Table 98). Forty-two percent of the sherds were recovered at depths below 20 cm bs. Eighteen separate vessels were identified in the sherd collection, indicating a great deal of disorganization and fragmentation. Schiffer (1987:282284) has suggested some indices (i.e. Completeness and Fragmentation Indices) to measure these effects, but they require a much larger coverage of deposits than we could achieve through limited testing to produce viable measures. Table 99 displays partial completeness percentages for the identified vessels in SB8 and Block 1 based on the average vessel weight of 1.377 kg suggested by Herbert and Irwin (2003:2). Most of the vessels are represented by less than 1 percent of the idealized vessel weight, while only three of the vessels achieved greater than 4 percent repre- sentation and no vessels achieved representation of greater than 9 percent. Only a much wider exposure would allow the full coverage necessary to generate full completeness percentages. However, as is shown in Table 99, only five vessels were represented in the broader SB 8 shovel test sample and all but two of these were found in only one shovel test. These data clearly demonstrate that the ceramic deposits at the site are characterized by a high degree of disorganization and fragmentation. This would suggest that the ceramic deposit was predominantly formed by a combination of primary, limited secondary and de facto refuse processes involving recycling of large sherds from broken pots and perhaps caching of vessels and fragments for use during site revisits. Moreover, each succeeding reoccupation would have contributed to the disorganization and fragmentation of the ceramic deposit. The low degree of completeness and high incidence of fragmentation also suggest that the method used to project vessel populations at these sites results in significant underestimation. That method assumes a relatively high degree of completeness and low incidence of fragmentation. An alterna- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 337 Chapter 7. 31HK2521 EAST WALL cm bd 0 A E’ Lev 1 10 Lev 2 Root 20 Lev 3 30 Lev 4 E 40 Lev 5 A’ 50 Lev 6 60 TU9 N506.5/E515.5 TU4 N505.5/E515.5 0 N504.5/E515.5 2m Soil Strata A: A’: E’: E: Pale Brown (10YR 6/3) Sand, Satined Very Dark Grayish Brown (10YR 3/2) in Part By Charcoal Leaching from Prescribed Burns Tree Root Stain, Brown (10YR 5/3) Loamy Sand Strong Brown (7.5YR 5/6) Oxidized Sand Light Yellowish Brown (10YR 6/4) Sand Figure 108. Profile Drawing, TU4 and TU9, East Wall, 31HK2521. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 338 Chapter 7. 31HK2521 tive method would be to use the area of the sample block as the basis for an areal sample fraction of vessels. Each sample block covers an area of 6.25 m2. Most of the ceramics are concentrated along the top of the ridge and along the northern and eastern slope, which covers an area of about 1,300 m2. SB 8 then, would constitute a 0.48 percent sample fraction of the ceramic deposit. If five vessels were found in the shovel tests, then this limited sample would project a total vessel population of 1,041 vessels for the site. This estimate is only a heuristic because the coverage is too limited to produce an accurate estimate for the entire site, but the magnitude of difference (more than two times higher) for this method versus the high completeness model should be appreciated. The method also assumes a complete recovery of vessels in the sample area, which we have seen is not the case. Block 1 covered an area of 2 m2, a sampling fraction of 0.15 percent of the primary ceramic deposit at the site. Since 18 vessels were identified in the block, a straight projection from this result would yield a population estimate of 12,000 vessels. The problem with this approach is that the full spatial extent of the vessels is not represented, which would lead to a difficult to control over-estimate. The data from Block 1, however, suggests that a true population mean of vessels deposited at the site lies somewhere between 1,000 and 12,000. Vertical distributions of lithic debitage in the test units mirrors the results derived from shovel testing (Table 97). Most of these debitage scatters are situated below the Woodland zone in the sample block and the recovery of additional diagnostics indicate that the UER and most of the white quartz debitage are affiliated with Early Archaic Type III elements. A Palmer I Corner Notched point (Figure 81:B) is matched with the UER scatter and a Type IIa Side Scraper (Figure 82:J) is matched with white quartz debitage. The earlier SB 8 investigation recovered Early Archaic scrapers matching the USR(1) and USR(2) debitage scatters as well. All of these elements are situated between 30 and 50 cm bs, which is consistent with the depth at which Early Archaic “living floors” would be expected. These Early Archaic elements may represent separate Type III occupations, but, just as we saw at 31HK2502 and 31HK2510, these deposits exhibit characteristics that would also be consistent with Type II residences. Type II residences are similar in spatial organization to Type I residences, but they exhibit much higher tool-to-debitage ratios and may commonly be comprised of diverse lithic raw materials. It is speculated that raw material heterogeneity is a function of dispersed groups travelling from different locales to aggregate at selected places on the landscape at specific and planned times. In these situations, it is likely that individual groups will be provisioned with a different assemblage of raw material types, even if the groups were foraging within the same general region. Lithic reduction on these sites consisted of tool maintenance and late stage shaping of tools from flake blanks or reduced biface cores. Consequently, debitage density is much lower than that of Type I debitage concentrations and the scatters consist primarily of late stage reduction debitage. Type II elements are distinguished from Type III elements by tool density. The former is characterized by high tool density, while the latter has few discarded tools relative to debitage. Type II occupations are inferred to represent the output from single or small, multiple household aggregations and are thought to constitute the residue from High Technology Forager (HTF) residences (Spiess 1984; Todd 1983:231-233). The HTF model proposes a specialized forager adaptation combining high logistical and high residential mobility into a single settlement system characterized by highly curated technologies. Each of the identified Early Archaic elements is characterized by small core flakes, BTFs and only small to moderately sized FBRs, which is a pattern typical of both Type II and III residences, but not Type I residences where major tool manufacturing activities transpired. This is true even for the USR(2) element, which was considered a potential Type I candidate prior to test unit excavation (see Table 95). The total weight of all debitage recovered from SB 8 and Block 1 inferred to be associated with the four Early Archaic elements is only 117 gm. It is further estimated, based on a 12.96 percent areal coverage by shovel tests, that a total of about 183 gm of debitage from these elements is present in the area covered by SB 8, which is only about 6.5 ounces. This combined weight would equal the weight of about three spent directional cores similar to the ones described at 31HK2510. Tool to debitage weight ratios are large, ranging between 0.1 and 0.4, but could not be used to distinguish between Type II and III elements. The Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 339 Chapter 7. 31HK2521 Table 97. Stone Tools and Debitage by Level for SB 8 and Block 1, 31HK2521. VERTICAL DISTRIBUTION OF DEBITAGE AND STONE TOOLS BY RAW MATERIAL TYPE FOR SB 8 AND BLOCK 1: 31HK2521 TU4 RAW MATERIAL TYPE LEVELS 1 Crystal Quartz 2 3 4 15 7 1 5 6 GRAND TOTAL 23 Schist 1 Type I Rhyolite Tuff 1 2 1 3 12 4 2 2 4 Uwharries Southern Rhyolite (1) 2 4 7 1 3 17 Uwharries Southern Rhyolite (2) 2 14 24 34 16 90 2 2 1 4 4 7 3 19 24 33 42 48 23 170 White Quartz 4 5 6 End Scraper frag. 2 Uwharries Eastern Rhyolite Uwharries Western Rhyolite LEVELS 2 Palmer I CN Flake Blank 4 Quartz Cobble Hammerstone GRAND TOTAL TU9 LEVELS TU9 Raw Material Type 1 2 3 Metasedimentary, Green 1 Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) 1 White Quartz GRAND TOTAL 5 6 1 Type I Rhyolite Tuff Uwharries Eastern Rhyolite 4 2 GRAND TOTAL LEVELS 2 4 5 6 1 2 3 5 3 1 6 1 1 1 11 13 2 3 5 7 2 19 2 3 11 22 7 46 2 2 10 4 7 25 11 12 38 34 18 115 Nat. Core Type IIa Side Scraper Core frag. SB 8 LEVELS SB 8 RAW MATERIAL Crystal Quartz 1/2 4 1 Mill Mountain Rhyolite Type I Rhyolite Tuff 3 5 7/9 2 3 22 1 2 Uwharries Eastern Rhyolite 1 1 Uwharries Southern Rhyolite (1) 1 2 4 Uwharries Southern Rhyolite (2) 3 7 5 White Quartz 1 4 1 8 1 2 18 1 1 4 1 1 9 Quartz Pebble GRAND TOTAL 3 4 5 2 1 2 LEVELS 2 1 7 1 GRAND TOTAL 1 10 Uwharries Western Rhyolite 6 Type IIa Side Scraper Undef. Rud. Stemmed Type Ib End Scraper Type III Biface frag. Core frag. Hammerstone 4 18 19 17 3 5 66 extensive distribution of Early Archaic tools across the site, combined with the information from Block 1, indicates that the Quewhiffle site contains an extensive Early Archaic zone comprised primarily of Type II and Type III occupations. Two high frequency elements appear to be situated above the Early Archaic zone in SB 8. These are the Type I RT debitage concentration, which was inferred to represent a Type I element from shovel test results and a crystal quartz debitage scatter was not well defined until the excavation of TU4 (Table 97). The Type I RT scatter is situated between 20 and 40 cm bs, a vertical position that appears higher than the Early Archaic occupations in the sample block, but situated lower than the Woodland zone. Although frequency distribution data indicate that this concentration would qualify as a Type I element, an examination of the debitage characteristics reveals that it resembles the patterns described for the Type II/Type III Early Archaic elements. The entire sample of Type I RT debitage collected during the investigation Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 340 Chapter 7. 31HK2521 Table 98. Vertical Distribution of Ceramics in SB 8 and Block 1, 31HK2521. VERTICAL DISTRIBUTION OF CERAMIC TYPES IN SB 8 AND BLOCK 1: 31HK2521 TU 4 CERAMIC TYPE Cape Fear III Cord Marked Hanover Ib Check Stamped Hanover Ib Linear Check Stamped Hanover IIa Cord Marked Hanover IIa Fabric Impressed Hanover IIa indet. Hanover IId Check Stamped Hanover IIIa Fabric Impressed New River IIa indet. New River IIa Plain GRAND TOTAL 1 1 2 1 5 5 6 16 3 7 1 16 GRAND TOTAL 3 7 1 1 2 3 3 11 1 2 34 1 2 2 4 1 1 LEVELS 3 4 2 2 1 1 1 VESSEL # V167 V35, V171 V171 V169 V36, V168 V80 V170 V166 V172 V172 TU 9 CERAMIC TYPE Hanover Ia Check Stamped Hanover Ia Cord Marked Hanover Ib Check Stamped Hanover Ib Cord Marked Hanover Ib Indet. Hanover IIa Fabric Impressed Hanover IId Fabric Impressed Hanover IId indet. Hanover IId indet. decorated Hanover IIIe Fabric Impressed Residual Sherd Yadkin IIIe Cord Marked Yadkin IIIe Fabric Impressed GRAND TOTAL 1 2 5 1 LEVELS 3 4 1 1 2 1 2 5 6 1 1 7 2 2 2 1 1 1 1 3 2 1 25 2 1 1 1 1 4 1 1 2 1 13 GRAND TOTAL 1 8 VESSEL # V226 V204 V35 V224 V224 V36 V222 V222 V225 V241 V223 SB 8 CERAMIC TYPE Hanover Ib Check Stamped Hanover IIa Fabric Impressed Hanover IIa indet. Hanover IIa indet. decorated Hanover IIIa Linear Check St. Hanover IIIa Fabric Impressed GRAND TOTAL 1 2 1 3 1 6 3 2 16 weighed only 8.37 gm with an average weight per piece of debitage equaling only 0.2 gm. Combined, these data suggest that it represents a Type III element of Middle or Early Archaic affiliation. The crystal quartz scatter is almost entirely contained within the southeast quadrant of TU4 and totals only 17.82 gm. Average weight per piece of debitage is relatively high, however, at 0.7 gm per LEVELS 3 4 1 2 1 3 1 5 6 GRAND TOTAL 1 3 1 8 5 2 20 VESSEL # V35 V36 V80 V79 V166 piece. Its limited extent and low total weight, though, suggest that it represents a Type III element. Unlike the other elements, however, its vertical position is squarely within the Woodland zone of the site, concentrating between 10 and 20 cm bs. A potential contradiction to this cultural assignment is presented by the recovery of what appears to be a small bit fragment from an end scraper made of this material. It was also found in the southeast quadrant of TU4 at Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 341 Chapter 7. 31HK2521 Table 99. Partial Completeness Percentages for Vessels in SB 8 and Block 1, 31HK2521. PARTIAL COMPLETENESS PERCENTAGES FOR VESSELS IN SB 8 AND BLOCK 1: 31HK2521 VESSEL NO. WEIGHT (gm) PERCENT COMPLETENESS OCCURRENCE IN SHOVEL TESTS 1 35 56.09 4.1 36 22.6 1.6 2 79 35.11 2.5 2 80 116.22 8.4 1 166 123.35 9.0 1 167 38.22 2.8 0 168 5.78 0.4 0 169 3.24 0.2 0 170 16.84 1.2 0 171 28.24 2.1 0 172 8.82 0.6 0 204 1.14 0.1 0 222 7.33 0.5 0 223 13.01 0.9 0 224 17.62 1.3 0 225 16.62 1.2 0 226 1.23 0.1 0 241 19.33 1.4 0 GRAND TOTAL 530.79 a depth of 10 to 20 cm bd. Its high vertical position, however, suggests that the crystal quartz element is more likely affiliated with the Woodland occupation at the site and that the classification of the tool as an end scraper fragment is probably in error. The specimen exhibits the typical steep, pressure flaked retouch commonly associated with end scraper bits, but similar flaking patterns are more easily created inadvertently on crystal quartz than almost any other type of material owing to its glassy isotrophic qualities. A final issue concerns the association of the Undefined Rudimentary Stemmed (Figure 81:K) point recovered between 20 and 30 cm bs in the shovel test sample of SB 8. A matching white quartz debitage scatter in the upper 30 cm of deposit was not positively identified during the investigation. The only white quartz scatter was positioned much lower and was associated with a side scraper, suggesting an Early Archaic affiliation. The projectile point, however, was found in a shovel test (N506.25/ E515) excavated at the very northern perimeter of SB 8, situated 0.75 m north of TU9. Therefore, its cultural affiliation is not effectively evaluated from existing information. Although it might represent an extremely small Morrow Mountain I Stemmed point, the size and morphology of the stem are more consistent with forms that are reported to have Middle to Late Woodland affiliations in other regions. Examples include the Group 8 points at the Mattassee Lake sites (Anderson et al. 1982:157-158) in the Santee River Valley of South Carolina and the Ebenezer point from Tennessee and Alabama (Cambron and Hulse 1975:42). The SB 8/Block 1 investigations succeeded in identifying a wide range of occupations in which Archaic components were vertically distinct from Woodland period occupations. SB 8 was situated on the top of the ridge in an area where vertical separation of deposits was most optimal. The Early Archaic Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 342 Chapter 7. 31HK2521 zone was situated between 30 and 50 cm bs/bd and included numerous Type II/III elements that can be analytically segregated by raw material type. The Woodland zone occupied the upper 20 to 30 cm of deposit and included a wide range of ceramic subseries. The ceramic deposit was characterized by a high degree of disorganization and fragmentation. Although none of the identified vessels achieved greater than about 9 percent representation of an idealized vessel weight, at least one intact debitage scatter (i.e. crystal quartz) of probable Woodland period affiliation was encountered. This suggests that the structure and organization of Woodland period occupations at the site may be preserved in lithic distributions, even if ceramic deposits are largely displaced. Block 2 (TUs 1, 6, 10 and 11/SB 3) Block 2 consisted of a linear arrangement of four contiguous 1-x-1-m test units placed across SB 3. The units were numbered 1, 6, 10 and 11 and were located along the N509 grid line between E527.5 and E531.5. Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner TU1 (Figure 109). Two soil layers were exposed in the wall profiles. The A-horizon was positioned in the upper 5 to 10 cm of the soil profile. The contact with the lower E-Horizon was irregular. The A-Horizon was comprised of a pale brown (10YR 6/3) sand matrix that was partially stained a very dark grayish brown (10YR 3/2) from leached charcoal derived from prescribed burn episodes. The E-Horizon extended below the base of the unit, which was terminated at 40 cm bd. Probing indicated that the Bt-Horizon was situated at depths of about of 65 to 75 cm bd. The E-Horizon was comprised of light brown (10YR 6/3) sand. Since the basic goal of work in Block 2 was to examine the nature of deposits primarily situated in the upper portion of the E-Horizon, the units were terminated at 40 cm bd to accommodate greater horizontal exposure. SB 3 was established around a Stage I shovel test that produced a dense white quartz debitage concentration centered at a depth of 10 to 30 cm bs (Table 88). This location was selected for closer examination because of the possible association of the debitage with Woodland period occupation, which was the major focus of investigation at the site. Debitage of other raw material types was present at lower densities and the recovery of a scraper fragment, a spoke shave and a Type IIb End Scraper in shovel testing suggested that much of this other material was associated with the Early Archaic period (Table 95). These tools were found high in the soil matrix and “out of position” for Early Archaic occupations. However, since the block was situated at the beginning of the ridge slope descent, it was anticipated that deposits would be compressed and that Woodland and Archaic debris would be mixed. Ceramics were nearly all recovered between 0 and 30 cm bs and they represented a wide range of Woodland period occupation including New River I, Hanover I, Hanover II, Hanover III and Yadkin III sub-series (Table 93). The SB 3 ceramic collection numbered 21 sherds, which represented fragments of 12 separate vessels. The first unit established in the block was TU1, which was placed about 1 m south and 0.5 m west of the Stage I anchor shovel test. The remaining units were gradually added to the block to continue following the white quartz debitage concentration. Test unit contents from top to bottom in Table 100 are arranged in order from west to east. The west end of the unit was situated at the highest topographic elevation and the most eastern unit, TU11, was farthest down slope and occupied the lowest elevation. Level 1 gradually attenuated to the east. Thus, Level 1 on the west is equivalent to Level 2 on the east in terms of depth below surface. The block managed to cross-cut most of the span of the white quartz debitage concentration, which was most dense in TU6. TU11 on the east end of the block produced a tool cluster of matching white quartz tools and cores. These items included a projectile point lateral section, a utilized flake and a directional core (Figure 79:C). Tool clusters tend to be located along the perimeter of Type I debitage concentrations and it is believed that they demarcate the general area of a residential cooking hearth. Also found in TU11 at the same vertical position, between 10 and 20 cm bs (20 to 30 cm bd), was an Undefined Expanded Stemmed projectile point (Figure 81:J) manufactured from USR (2). The age of this point is unknown, but it is speculated here that the type is affiliated with Middle or Late Woodland occupation. Its presence in TU11 suggests that it might also be a member of the tool cluster defined by the white quartz stone tools and cores. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 343 Chapter 7. 31HK2521 One hundred thirty-nine precontact ceramic sherds were recovered from Block 2 and SB 3 (Table 101). Ceramic vertical distributions mirror the patterns described for the white quartz debitage concentration. There is a gradual and artificial dip in level data to the east because a single vertical datum was maintained across the sloping topography. Once an adjustment is made to standardize the level data relative to depth below surface, the units consistently show ceramic densities centered between 0 and 20 cm bs. Only about 13 percent of the ceramic collection was found below 20 cm bs. Forty-four separate vessels were identified in the collection of only 139 sherds, demonstrating the same kind of fragmentation and displacement seen in the Block 1 data previously discussed. Table 102 displays partial completeness percentages for the identified vessels in SB3 and Block 2 based on the average vessel weight of 1.377 kg suggested by Herbert and Irwin (2003:2). Eighty percent of the vessels are represented by less than 2 percent of their idealized vessel weight, while only three of the vessels achieved greater than 3.7 percent representation and no vessels achieved representation of greater than 7.2 percent. Only a much wider exposure would allow the coverage necessary to generate full completeness percentages for each vessel, but it is clear that the ceramic deposits in this area of the site are characterized by high levels of disorganization. If that were not the case, the area of exposure would have produced much higher partial completeness percentages, as we should expect the sherds from most pot breaks to be situated within a 2 m diameter area. These results suggest that the ceramic deposit was predominantly formed by a combination of primary, limited secondary and de facto refuse processes involving recycling of large sherds from broken pots and perhaps caching of vessels and fragments for use during site revisits. Moreover, each succeeding reoccupation appears to have contributed to the disorganization and fragmentation of the ceramic deposit. The low degree of completeness also suggests that the method used to project vessel populations at these sites results in significant underestimation. That method assumes a relatively high degree of completeness and low incidence of fragmentation. An alternative method, discussed above, would be to use the area of the sample block as the basis for an areal sample fraction of vessels. SB 3 then, would constitute a 0.48 percent sample fraction of the ceramic deposit at the site which is contined primarily in an area of 1,300 m2. Since 12 vessels were identified in shovel tests in SB 3, then this limited sample would project an estimate of 2,500 vessels for the site. This estimate is only a heuristic because the coverage is too limited to produce an accurate estimate for the entire site, but the magnitude of difference (more than two times higher) for this method versus the high completeness model should be appreciated. The estimate from SB 8, which is situated in a lower ceramic density area of the site was only 1,041 vessels. The method also assumes a complete recovery of vessels in the sample area, which we have seen is not the case. Block 2 covered an area of 4 m2, a sampling fraction of 0.30 percent of the primary ceramic deposit at the site. Since 42 vessels were identified specifically in Block 2, a straight projection from this result would yield a population estimate of 14,000 vessels, which is comparable to the estimate of 12,000 vessels generated for the Block 1 excavation. The problem in application of this approach is that the full spatial extent of the vessels is not represented, which would lead to a difficult to evaluate over-estimate. Averaging the data from Blocks 1 and 2 and SB3 and SB 8, however, suggests that a true population mean of vessels deposited at the site lies somewhere between 1,770 and 13,000. The SB 3/Block 2 investigations documented the processes of deposit compression that occur at the top edges of the ridge slope. Early Archaic and Woodland period diagnostic artifacts came into contact at depths of 20 to 30 cm bs. Potentially affiliated with the Early Archaic were a Type Ia End Scraper recovered from Level 3 of TU10, a side scraper fragment found in Level 2 of TU1 and a Type IIb End Scraper, a spoke shave and a scraper fragment recovered between 10 and 20 cm bs in shovel tests. A prismatic blade fragment (Figure 79:J), which could have been deposited during Paleoindian occupation at the site, was positioned relatively high as well, between 20 and 30 cm bd in TU1. Block 2 investigations focused on the delineation of a dense, white quartz debitage concentration positioned primarily between 10 and 30 cm bs. Based on artifact density distributions the concentration appears to have reached maximum dimensions Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 344 Chapter 7. 31HK2521 EAST WALL cm bd 0 A 10 Root E TU10 TU1 N509.0/E525.5 TU6 N509.0/E527.5 0 20 E 30 40 Lev 1 Lev 2 Lev 3 Lev 4 TU11 N509.0/E529.5 4m Soil Strata A: E: Pale Brown (10YR 6/3) Sand, Satined Very Dark Grayish Brown (10YR 3/2) in Part By Charcoal Leaching from Prescribed Burns Light Brown (10YR 6/3) Sand Figure 109. Profile Drawing, TU1, TU6, TU10 and TU11, East Wall, 31HK2521. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 345 Chapter 7. 31HK2521 Table 100. Stone Tools and Debitage by Level for Block 2, 31HK2521. VERTICAL DISTRIBUTION OF DEBITAGE AND STONE TOOLS BY RAW MATERIAL TYPE FOR SB 3 AND BLOCK 2: 31HK2521 TU10 RAW MATERIAL TYPE 1 Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) 1 1 Uwharries Western Rhyolite White Quartz GRAND TOTAL 2 4 DEBITAGE LEVEL 2 3 1 4 2 1 4 STONE TOOLS LEVEL GRAND TOTAL 1 7 5 1 5 3 1 13 1 4 5 3 14 4 18 1 1 2 10 38 2 3 4 Type I Biface frag. Type Ia End Scraper Core frag. Type II Biface frag., Utilized Flake TU1 5 DEBITAGE LEVEL 2 3 1 3 1 1 1 2 2 6 4 1 1 34 20 14 1 4 1 1 5 17 2 73 7 48 20 104 RAW MATERIAL TYPE 1 Metasedimentary, Tan Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz Grano-Diorite GRAND TOTAL 2 4 STONE TOOLS LEVELS GRAND TOTAL 1 5 1 2 3 4 Prismatic Blade Side Scraper frag. PPK frag. Hammerstone/Anvil 29 TU6 RAW MATERIAL TYPE 1 Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 1 1 DEBITAGE LEVEL 2 3 1 1 4 2 95 129 96 136 4 1 1 STONE TOOLS LEVEL GRAND TOTAL 1 2 6 2 255 265 30 32 2 3 4 3 4 Utilized Flake TU11 RAW MATERIAL TYPE 1 Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz DEBITAGE LEVEL 2 3 2 1 2 4 2 3 5 1 STONE TOOLS LEVEL GRAND TOTAL 1 4 4 7 1 14 55 18 87 14 60 29 103 Quartz Cobble GRAND TOTAL 2 Utilized Flake Undef. Expanded St. PPK frag., Utilized Flake Hammerstone, Abrader/Grinder Directional Core frag. of about 4 m in diameter. A matching tool cluster was identified on the eastern perimeter of the concentration in TU11. Tool clusters are a common feature of Type I elements and they are speculated to represent the general area of cooking hearths, sometimes situated at the threshold of huts and living floor voids adjacent to concentrations. In addition, large concentrations such as this many times contain multiple tool clusters spaced at distances from one another, each speculated to represent distinct social units such as nuclear or extended family groups. The depth of the concentration suggests that it was formed during the Woodland period, but matching (i.e. white quartz) diagnostic point types were not recovered in the partial exposure of the feature. However, an Undefined Expanded Stemmed point (Figure 81:J) of USR (2) material was also recovered from TU11 at the same level as the quartz tools and debitage. The cultural affiliation of this point type is not yet established, but expanded and notched stem points similar to this one Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 346 Chapter 7. 31HK2521 Table 101. Vertical Distribution of Ceramics Block 2 Test Units, 31HK2521. VERTICAL DISTRIBUTION OF CERAMIC TYPES IN BLOCK 2 TEST UNITS: 31HK2521 TU10 CERAMIC TYPE LEVEL 1 Cape Fear III Cord Marked Hanover Ia Indet. Hanover IIa indet. Hanover IIb indet. Hanover IIIa Cord Marked Hanover IIIb Fabric Impressed New River IIa Cord Marked Residual Sherd Yadkin IIId Cord Marked Yadkin IIId Indet. Yadkin IIIe Cord Marked Yadkin IIIe indet. Yadkin IIIe indet. textile Yadkin IIIf indet. Yadkin IIIf Plain GRAND TOTAL 2 3 GRAND TOTAL 4 1 2 1 1 1 2 1 2 2 1 4 6 2 20 4 6 1 1 1 1 2 7 27 1 2 5 1 1 2 1 2 2 1 2 3 2 1 4 8 6 7 2 3 10 54 VESSEL NO. V198 V204 V203 V197, V199 V200 V242 V242 V119 V119 V119 V121 V121 TU1 CERAMIC TYPE LEVEL 1 Hanover Ia Cord Marked Hanover IIa Check Stamped Hanover IIb Fabric Impressed Hanover IIb Linear Check Stamped Hanover IIIa Cord Marked Hanover IIIa indet. Hanover IIIb Fabric Impressed Hanover IIIe Fabric Impressed Residual Sherd Yadkin IIId Cord Marked Yadkin IIId Indet. Yadkin IIIe Cord Marked Yadkin IIIe indet. textile Yadkin IIIf Plain GRAND TOTAL 2 3 GRAND TOTAL 4 3 1 3 1 1 1 1 1 1 1 8 3 11 7 1 1 41 1 1 1 1 1 1 6 1 2 3 1 2 8 3 1 1 1 1 1 21 16 4 VESSEL NO. V207 V209 V212 V124 V203 V203 V211 V208 V119 V119 V120, V205, V210 V206 V121 TU6 CERAMIC TYPE LEVEL 1 Cape Fear III Cord Marked Cape Fear III indet. Hanover Ia Plain Hanover Ib Fabric Impressed Hanover IIa Check Stamped Hanover IIb Check Stamped Hanover IIIa Cord Marked Hanover IIIb indet. textile Hanover IIIe Cord Marked New River IIa Indet. stamped Residual Sherd Yadkin IIId Cord Marked Yadkin IIIe Cord Marked Yadkin IIIe Fabric Impressed Yadkin IIIe indet. Yadkin IIIe indet. decorated Yadkin IIIe Wide Simple Stamped Yadkin IIIf indet. GRAND TOTAL 2 3 GRAND TOTAL 4 1 1 1 1 1 1 1 1 1 2 3 1 1 3 1 1 1 1 2 4 27 1 1 1 1 1 2 3 1 1 2 1 1 1 1 1 2 1 17 3 3 7 VESSEL NO. V221 V221 V193 V209 V214 V203 V220 V217 V215 V242, V233 V210 V123 V123 V123 V218 V121, V216, V219 TU11 CERAMIC TYPE Hanover Ib Fabric Impressed Hanover IIa Fabric Impressed Hanover IIb Fabric Impressed Hanover IIIa indet. textile Residual Sherd Yadkin IIIe Check Stamped Yadkin IIIe Cord Marked Yadkin IIIf Plain GRAND TOTAL LEVEL 1 2 3 GRAND TOTAL 4 7 1 1 1 2 1 1 1 2 1 14 2 8 1 1 2 1 1 2 1 17 VESSEL NO. V193, V196 V179 V194 V195 V240 V120 V121 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 347 Chapter 7. 31HK2521 have been associated with Woodland period phases in other regions. Sherd aggregations are generally members of Woodland period Type I elements, but these are found on the opposite side of tool clusters from debitage concentrations. Therefore, it is likely that most of the ceramics found in Block 2 are not associated with the debitage concentration and the tool cluster. The fact that lithic elements of probable Woodland age are intact suggests that the structural characteristics and spatial organization of Woodland period occupations are accessible for study, even if ceramic deposits are largely displaced and heavily impacted by recycling and reoccupation. investigation at the site. Debitage of other raw material types was present, but only in low frequencies. Shovel testing within the sample block area recovered quartz as low as 50 to 60 cm bs (Table 91), but sherds were also found at these depths in several of the shovel tests, indicating that later material intruded into earlier and lower deposits due to either cultural or natural processes of disturbance. The three shovel tests along the western margin of SB 16 and another on the northeast end documented sherds and quartz debitage that had been displaced downward (Figure 112). Discoloration that might evidence precontact cultural features could not be discerned in the shovel test profiles. Block 3 (TUs 2, 5, 7 and 8/SB 16) Block 3 was placed on the northeast side of SB 16 in an area that appeared to represent contact between a vessel aggregation characterized by large sherd sizes and the white quartz debitage concentration. Contacts between these types of features are common in Woodland period Type I elements and this is also the anticipated location of tool clusters believed to be associated with hearth areas. As was the case for the other blocks excavated at the site, a large number of vessels were represented relative to the sherd population (Table 93). Sixteen vessels were identified in the collection of 39 sherds, which were assigned to New River I, New River II, Hanover I, Hanover II, Hanover III and Yadkin III sub-series. Several of the vessel fragments found in shovel testing were characterized by larger sherd sizes, suggesting that some of the ceramic deposit here might be more organized and less displaced than had been encountered elsewhere. Block 3 consisted of a contiguous arrangement of four 1-x-1-m test units forming a 2-x-2-m square (Figure 78). The units were numbered 2, 5, 7 and 8 and were situated between N517 and N515 and E534.5 and E536.5 on the northeast perimeter of SB 16 (Figure 110). Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner TU2 (Figure 111). Two soil layers were exposed in the wall profiles. The A-horizon was positioned in the upper 5 to 10 cm of the soil profile and, just was true of the other test unit profiles the contact with the E-Horizon was irregular. The A-Horizon was comprised of a pale brown (10YR 6/3) sand matrix that was partially stained a very dark grayish brown (10YR 3/2) from leached charcoal derived from prescribed burn episodes. The E-Horizon extended below the base of the unit, which was terminated at 40 cm bd. Probing indicated that the Bt-Horizon was situated at depths of about of 55 to 80 cm bd. The E-Horizon was comprised of pale yellow (10YR 7/3) sand. Since the basic goal of work in Block 3 was to examine the nature of deposits primarily situated in the upper portion of the E-Horizon (i.e. Woodland occupation), the units were terminated at 40 cm bd to accommodate greater horizontal exposure. SB 16 was established around a Stage I shovel test that produced a dense white quartz debitage concentration centered at a depth of 10 to 30 cm bs and Hanover series sherds (Table 91). This location was selected for closer examination because of the possible association of the debitage with Woodland period occupation, which was the major focus of The quartz debitage scatter was centered to the south and west of Block 3 (Figure 113). Although excavations were not carried below 40 cm bd, vertical artifact distributions in Block 3 show clearly that quartz debitage is concentrated in the upper 30 cm of deposit (Table 103). In the test units (TUs 2 and 5) located at higher topographic elevations, all of the quartz debitage is contained in the upper 20 cm of deposit and it is strongly concentrated between 0 and 10 cm bd. Three white quartz tools and cores were recovered in a restricted area on the southwest margin of Block 3 and in a shovel test placed at N516.25/ E513.75. These items consisted of a core fragment and two non-diagnostic projectile point fragments, a tip and an upper blade section. The latter was rela- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 348 Chapter 7. 31HK2521 Table 102. Partial Completeness Percentages for Vessels in SB 3 and Block 2, 31HK2521. PARTIAL COMPLETENESS PERCENTAGES FOR VESSELS IN SB 3 AND BLOCK 2: 31HK2521 VESSEL NO. SHERD COUNT WEIGHT (gm) PERCENT COMPLETENESS 24 1 15.41 1.1 25 1 9.31 0.7 119 11 29.1 2.1 120 8 35.57 2.6 121 17 99.06 7.2 122 1 3.44 0.2 123 4 5.88 0.4 124 3 39.75 2.9 125 7 50.97 3.7 126 1 3.83 0.3 179 1 0.81 0.1 193 8 27.70 2.0 194 1 2.02 0.1 195 2 16.45 1.2 196 1 7.09 0.5 197 1 13.83 1.0 198 1 2.40 0.2 199 2 9.82 0.7 200 2 6.85 0.5 201 1 10.68 0.8 202 1 5.57 0.4 203 5 34.43 2.5 204 2 4.70 0.3 205 3 23.21 1.7 206 1 13.01 0.9 207 3 3.52 0.3 208 1 1.08 0.1 209 2 25.58 1.9 210 2 19.87 1.4 211 1 7.64 0.6 212 1 3.83 0.3 214 1 4.63 0.3 215 1 8.68 0.6 216 1 0.85 0.1 217 3 30.21 2.2 218 2 15.30 1.1 219 2 8.61 0.6 220 2 6.04 0.4 221 2 3.22 0.2 233 1 2.61 0.2 234 1 5.81 0.4 235 3 15.08 1.1 240 1 15.20 1.1 242 28 77.06 5.6 144 725.71 1.2 GRAND TOTAL Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 349 Chapter 7. 31HK2521 tively large (Figure 81:H), but it is within the size range of Woodland period projectile point types. The fragment is only roughly flaked and it may represent only a preform reject that was not yet thinned to its final form. The tight clustering of these tools on the perimeter of the debitage scatter is consistent with the physical and distributional characteristics of a tool cluster. Several of the vessels identified in the collection displayed larger than average sherd aggregations. Two of these were situated primarily in TU8 and were initially identified in a shovel test placed at N515/E536.5 (Figure 114). These consisted of a Hanover IIId Fabric Impressed open-mouth, straightsided jar (Figure 92) and a Hanover IIIe Fabric Impressed jar of uncertain orifice shape, but with a conoidal base (Figure 93:E-F). It is probable that both vessels, which were concentrated between 0 and 20 cm bd, extend outside of the coverage to the east (Table 104). The interior of Vessel 155 is scraped with a serrated tool, evidencing influences from the Late Woodland ceramic traditions of the eastern Piedmont and the Cashie area. Vessel 150, a Hanover IIIe Fabric Impressed jar (Figure 93:C-D), appears to represent only a vessel fragment as it was situated near the center of Block 3. One other vessel, V189, was represented by a significant aggregate of sherds. This was also a Hanover IIIe Fabric Impressed jar. It was situated exclusively in the southwest quadrant of TU8 and its distribution probably extends outside of Block 3 to the east. These aggregates, which may belong to a single, contemporary household vessel assemblage, are positioned adjacent to the quartz scatter and the tool cluster in a manner consistent with a Type I Woodland period residence. More exposure would be needed to better evaluate this hypothesis, but all of the identified elements exhibit the predicted spatial relationships necessary to fulfill the model. Although a much greater degree of organization characterizes the ceramic deposits in Block3, the general condition of high fragmentation and displacement seen in the other blocks is repeated. Three hundred and ninety-six precontact ceramic sherds were recovered from Block 3 and SB 16. Table 105 displays partial completeness percentages for the identified vessels based on the average vessel weight of 1.377 kg suggested by Herbert and Irwin (2003:2). In spite of the high partial completeness percentages for V154 and V155, the remainder of the vessel sample is characterized by less than 5.3 percent completeness. Eighty-three percent of the vessels are represented by less than 3 percent of their idealized vessel weight. Only a much wider exposure would allow the coverage necessary to generate full completeness percentages for each vessel, but it is clear that most of the identified vessels in this area of the site are characterized by high levels of disorganization. If that were not the case, the area of exposure would have produced much higher partial completeness percentages, as we should expect the sherds from most pot breaks to be situated within a 2 m diameter area. These results suggest that the ceramic deposit around Block 3 was also formed by a combination of primary, limited secondary and de facto refuse processes involving recycling of large sherds from broken pots and perhaps caching of vessels and fragments for use during site revisits. Moreover, each succeeding reoccupation appears to have contributed to the disorganization and fragmentation of the ceramic deposit. Of interest in this regard is the fact that the vessels with the highest percentage of completeness in Block 3 are associated with the early Late Woodland occupation, which appears to represent the latest intensive occupation span at the site. The Cape Fear III series may post-date most of the Hanover III/Yadkin III occupation, but it was characterized by much less intensive use of the site. As we have seen from discussions above, the low degree of completeness and high incidence of fragmentation suggest that the method used to project vessel populations at sites in the Sandhills results in significant underestimation. That method assumes a relatively high degree of completeness and low incidence of fragmentation. Two alternative methods have been offered for interpreting the results of the sample blocks and test unit excavations at the site. One method uses the area of the shovel test sample block as the basis for an areal sample fraction of vessels. Each sample block covers an area of 6.25 m2. Most of the ceramics at the site are concentrated along the top of the ridge and along the northern and eastern slope, which covers an area of about 1,300 m2. A sample block, then, would constitute a 0.48 percent sample fraction of the ceramic deposit. Since 16 vessels were identified in shovel tests in SB 16, this limited sample would project a total population of 3,300 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 350 Chapter 7. 31HK2521 BLOCK 3 517 TU2 TU7 TU5 TU8 North 516 515 SB 16 514 513 533 534 535 536 537 East N 0 1m Figure 110. Plan View of SB16 and the Block 3 Test Units, 31HK2521. vessels for the site. This projection is only a heuristic because the coverage is too limited to produce an accurate estimate for the entire site, but the magnitude of difference (more than seven times higher) for this method versus the high completeness model should be appreciated. Estimates for Sample Blocks 8 and 3 are both smaller, 1,041 and 2,500 respectively. This method assumes a complete recovery of vessels in the sample area, however, which we have seen is not the case, suggesting that it too will underestimate vessel populations. Block 2 covered an area of 4 m2, a sampling fraction of 0.30 percent of the primary ceramic deposit at the site. Since 42 vessels were identified specifically in Block 2, a straight projection from this result yielded a population estimate of 14,000 vessels, which is comparable to the estimate of 12,000 vessels generated for the Block 1 excavation. Block 3 produces an estimate similar to Block 1. A sample fraction of 0.30 for the 36 identified vessels in Block 3 yields a mean estimate of 12,000 vessels. Since the full spatial extent of the vessels is not represented in limited test unit excavations, this method tends to risk overestimating the vessel population at the site. Averaging the data from all three test unit blocks and the associated shovel tests, however, suggests that a true population mean of vessels deposited at the site lies somewhere between about 2,230 and 12,500. The SB 16/Block 3 investigations identified a hypothetical Type I Woodland period residential element consisting of a white quartz debitage scatter, a matching stone tool cluster and a cluster of relatively organized Hanover III Fabric Impressed vessel and partial vessel aggregates. These elements are spatially structured in a manner consistent with the Woodland period Type I element model. The debitage scatter and the vessel aggregates are tangent to one another, indicating segregated activity areas. The debitage scatter was the locus of stone tool manufacture and perhaps food preparation and cooking. The intrusive deposits identified in the line of shovel tests on the western side of SB 16 may represent pit fea- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 351 Chapter 7. 31HK2521 2 7 NORTH WALL cm bd 0 A Roots Lev 1 A E 10 Lev 2 A 20 Lev 3 E 30 Lev 4 40 Lev 5 50 TU2 N517.0/E534.5 TU7 N517.0/E535.5 0 N517.0/E536.5 2m Soil Strata A: E: Pale Brown (10YR 6/3) Sand, Satined Very Dark Grayish Brown (10YR 3/2) in Part By Charcoal Leaching from Prescribed Burns Pale Yellow (2.5YR 7/3) Sand Figure 111. Profile Drawing, TU2 and TU7, Block 2, North Wall, 31HK2521. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 352 Chapter 7. 31HK2521 517 TU7 TU2 516 TU5 North TU8 515 Intrusive Artifacts SB 16 514 513 533 534 535 536 537 East N 0 1m Figure 112. Locations of Intrusive Artifacts in Lower Levels, SB16, 31HK2521. tures associated with this activity zone, although soil discoloration was not observed. The more complete vessel aggregates are all members of the Hanover III sub-series and might constitute a contemporaneous assemblage of vessels and partial vessels associated with a single household unit. The area covered by the aggregates is relatively devoid of debitage and other non-ceramic artifacts. Only 12.45 gm of the total of 368.9 gm of fire-cracked rock and rock was recovered from TUs 7 and 8 in the area of primary vessel aggregate concentration. Voids such as this are commonly found with “living floors” or sleeping areas. The Woodland period Type I model suggests that such areas represent the former locations of huts and that such areas are commonly associated with ceramic vessels and partial vessels which were stored in such shelters. The remainder of the fire-cracked rock and rock is found in TU2 (128.36 gm) and the southern portion of TU5 and the anchor shovel test at N515/E535 (227.58 gm). The latter location is situated directly in the area of the tool cluster, adding support to the hypothesis that tool clusters demarcate former hearthing areas. The diversity of the ceramic collection recovered from Block 3 indicates that a high degree of Woodland period deposit mixing is extant. However, there are spatially discrete and organized structural elements of camps that are still preserved at this location and at other locations across the site. These deposits are capable of further elucidating the nature of Woodland period occupation at the Quewhiffle site and larger patterns of adaptation. TU 3 (SB 14) TU3 was placed at grid coordinates N504/ E497 (NE corner) in the southeast quadrant of SB14 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 353 Chapter 7. 31HK2521 BLOCK 3 517 Vertical Disturbances 516 PPK TIP TU2 TU7 TU5 TU8 North 5. 0 PPK UPPER BLADE CORE FRAG.. .0 15 515 SB 16 25.0 514 Quartz Debitage Density Contours (Counts per 0.25 m2) 513 533 534 535 536 537 East N 0 1m Figure 113. Distribution of Quartz Debitage and Tools, SB16 and TU Block 3, 31HK2521. (Figures 77 and 78). Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of the unit (Figure 115). Two natural soil layers were identified in the profile. The A-Horizon was comprised of a pale brown (10YR 6/3) sand matrix that was partially stained a very dark grayish brown (10YR 3/2) from leached charcoal derived from prescribed burn episodes. The E-Horizon extended below the base of the unit, which was terminated at 40 cm bd. Probing indicated that the Bt-Horizon was situated at depths of about of 55 to 65 cm bd. The E-Horizon was composed of very pale brown (10YR 7/3) sand. Sandwiched between these two zones was a loose white (5YR 8/1) sand zone that appears to represent sorted slope wash. A root hole extending to the base of the unit was filled with this sand, suggesting that it was filled in during a slope erosion event. The event pre-dated the prescribed burn event registered in the A-Horizon. Since the basic goal of work in TU3 was to examine the nature of deposits primarily situated in the upper portion of the E-Horizon (i.e. Woodland occupation), the unit was terminated at 40 cm bd to accommodate greater horizontal exposure across the site. SB 14 was established around a Stage II shovel test that produced a Palmer I Corner Notched point (Figure 81:A) made of USR (2) at a depth of only 10 to 20 cm bs. The original intent was to examine this Early Archaic element in a portion of the site that was characterized by low artifact density (Table 106). Subsequent, shovel testing in the sample block, however, revealed that the most distinctive characteristic of the sample area was a portion of a Hanover IId Cord Marked vessel (V146). Individual sherds from the vessel were small (Figure 90:A-B) and appeared to have undergone post-depositional processes of fragmentation, perhaps from heavy equipment compression of the upper soil zone during prescribed burn activities. The vessel fragment, however, was spatially concentrated due to low occupation intensity Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 354 Chapter 7. 31HK2521 Table 103. Stone Tools and Debitage by Level for SB 16 and Block 3, 31HK2521. VERTICAL DISTRIBUTION OF DEBITAGE AND STONE TOOLS BY RAW MATERIAL TYPE FOR SB 16 AND BLOCK 3: 31HK2521 TU2 DEBITAGE LEVEL 2 3 RAW MATERIAL TYPE 1 Mill Mountain Rhyolite Uwharries Eastern Rhyolite White Quartz GRAND TOTAL 1 4 6 11 GRAND TOTAL 4 STONE TOOLS LEVEL 1 1 4 7 12 1 1 TU5 DEBITAGE LEVEL 2 3 RAW MATERIAL TYPE 1 Mill Mountain Rhyolite Uwharries Southern Rhyolite (1) Uwharries Western Rhyolite White Quartz GRAND TOTAL GRAND TOTAL 4 1 1 1 1 1 35 38 1 6 7 29 31 STONE TOOLS LEVEL 1 PPK frag., Core frag. TU7 RAW MATERIAL TYPE 1 Uwharries Eastern Rhyolite White Quartz GRAND TOTAL 1 1 2 DEBITAGE LEVEL 2 3 1 2 8 3 8 GRAND TOTAL 4 1 1 STONE TOOLS LEVEL 1 2 12 14 TU8 DEBITAGE LEVEL 2 3 RAW MATERIAL TYPE 1 Mill Mountain Rhyolite Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL GRAND TOTAL 4 1 1 6 6 1 6 7 1 1 STONE TOOLS LEVEL 1 1 1 13 15 (Table 107). Test Unit 3 was placed over a Stage III shovel test excavated at N503.75/E496.5 that yielded seven sherds from this vessel. of the spatial association between vessel and partial vessel aggregates and “living floors” or sleeping areas located within huts or other types of shelters. The test unit produced an additional 22 sherds from this vessel as well as smaller portions of a Hanover IIa Cord Marked vessel (V213) (Figure 90:C). Most of the sherds of both vessels were concentrated between 10 and 20 cm bs. Excavation of the unit revealed that very little else was contained in the area of the sherd aggregates (Table 106), a condition that typifies sherd aggregates of any size when they are found at locations with only minor mixing from other occupations. Void spaces in association with sherd aggregates are speculated to occur because Although low occupation intensity has reduced the disorganization in the ceramic deposit here, vessel completeness percentages remain low (Table 108). It is doubtful that any of the vessels identified in the investigation of SB 14 and TU3 represent whole vessels. Instead, they appear to constitute fragmented portions of large sherds. Vessel 146 yielded a completeness percentage of only 11.6 percent. As was concluded for the other test unit investigations, the ceramic deposit around SB 14 was most likely formed by a combination of primary, limited Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 355 Chapter 7. 31HK2521 Partial Vessel Aggregates Contours = Grams per 0.25 m2 40 517 225 V150, Hanover IIIe Fabric Impressed TU7 TU2 V154, Hanover IIId Fabric Impressed 22 5 V155, Hanover IIIe Fabric Impressed 12 20 TU5 30 75 10 PPK TIP 516 5 40 175 Tool Cluster 225 5 5 5 CORE v FRAG.. 12 17 22 North 75 PPK UPPER BLADE TU8 515 Main White Quartz Debitage Concentration SB 16 514 513 533 534 535 536 537 East N 0 1m Figure 114. Spatial Relationships of Sherd Aggregations and the Quartz Debitage Element, SB16, 31HK2521. secondary and de facto refuse processes involving recycling of large sherds from broken pots and perhaps caching of vessels and vessel fragments for use during site revisits. Due to the low incidence of reoccupation, vessel population estimates based on the data from SB14 and TU3 is much lower than that derived from the more dense ceramic deposits along the edge of the slope. For disorganized ceramic deposits, the shovel test sample block method generates an estimate of 625 vessels for the entire site based on the data from SB14, while the test unit method produces an estimate of 5,000 vessels for the TU3 data. Incorporating these data into the running average, then, the true population mean of vessels deposited at the site probably lies somewhere between about 1,828 and 10,625. The SB 14/TU3 investigations isolated two Hanover II Cord Marked sherd aggregates in a location of low occupation intensity at the site. The sherd aggregates appear to be located in artifact voids, which is stipulated by the Woodland period Type I element model. Vessels and partial vessels are predicted to occur on “living floors” and sleeping areas within the arcs of huts and other shelters. Neither of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 356 Chapter 7. 31HK2521 Table 104. Vertical Distribution of Ceramics in Test Units of Block 3, 31HK2521. VERTICAL DISTRIBUTION OF CERAMIC TYPES IN TEST UNITS OF BLOCK: 31HK2521 TU2 CERAMIC TYPE Cape Fear III Plain Hanover Ib Indet. Hanover IIa Check Stamped Hanover IId indet. Hanover IId indet. textile Hanover IIIe Fabric Impressed Residual Sherd Yadkin IIId Fabric Impressed Yadkin IIId indet. decorated GRAND TOTAL LEVEL 1 1 1 2 1 3 5 5 1 2 21 2 3 4 1 1 1 1 GRAND TOTAL 1 1 3 1 3 5 5 2 2 23 TU5 CERAMIC TYPE Cape Fear III Cord Marked Cape Fear III indet. Cape Fear III Plain Hanover IIa indet. Textile Hanover IIIa Fabric Impressed Hanover IIIb Cord Marked Hanover IIIb indet. Hanover IIIb Plain Hanover IIIe Fabric Impressed Hanover IIIe indet. New River Ia Plain Residual Sherd Yadkin IIId Fabric Impressed Yadkin IIId indet. decorated Yadkin IIIe Fabric Impressed Yadkin IIIe indet. GRAND TOTAL LEVEL 1 2 1 2 1 3 1 1 1 25 1 1 4 1 1 14 59 2 1 3 4 GRAND TOTAL 3 1 2 1 3 1 1 1 26 1 1 4 1 1 17 1 65 1 3 1 6 TU7 CERAMIC TYPE LEVEL 1 Hanover Ib Fabric Impressed Hanover IIIe Fabric Impressed Yadkin IIIc/e/f Plain Yadkin IIId Indet. Yadkin IIId indet. Yadkin IIId indet. decorated Yadkin IIId Plain Yadkin IIIe Fabric Impressed GRAND TOTAL 2 1 9 1 1 1 1 1 2 17 3 1 2 4 GRAND TOTAL 2 11 1 1 1 1 1 3 21 1 4 TU8 CERAMIC TYPE Hanover IIa Fabric Impressed Hanover IIa Incised/indet. Textile Hanover IIa indet. Hanover IIa indet. Textile Hanover IIIa Cord Marked Hanover IIIa Fabric Impressed Hanover IIIa indet. Hanover IIIb Cord Marked Hanover IIIb indet. Hanover IIId Fabric Impressed Hanover IIIe Fabric Impressed New River Ia Indet. Residual Sherd Yadkin IIIe Fabric Impressed Yadkin IIIe indet. GRAND TOTAL LEVEL 1 2 3 2 4 1 2 2 1 5 1 1 1 9 28 1 22 117 2 38 1 11 2 1 42 159 46 1 GRAND TOTAL 2 1 2 1 5 3 1 1 1 33 183 1 11 2 1 248 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 357 Chapter 7. 31HK2521 Table 105. Partial Completeness Percentages for Vessels in SB 16 and Block 3, 31HK2521. PARTIAL COMPLETENESS PERCENTAGES FOR VESSELS IN SB 16 AND BLOCK 3: 31HK2521 VESSEL NO. 67 68 69 148 149 150 151 152 153 154 155 156 157 158 159 160 173 174 175 176 177 178 180 181 182 183 184 185 186 187 188 189 190 191 229 238 Unassigned GRAND TOTAL COUNT WEIGHT (gm) PERCENT COMPLETENESS 7 1 2 2 9 23 2 5 15 38 200 1 1 1 1 2 4 2 3 3 3 1 3 2 2 3 4 1 2 1 6 9 1 3 5 2 26 396 30.61 2.10 37.15 13.92 64.22 72.83 5.55 14.52 46.95 481.70 885.90 8.24 3.17 12.70 2.26 29.09 26.05 7.00 10.35 12.98 5.16 1.99 7.15 15.67 3.21 39.95 9.64 2.25 13.05 11.78 31.89 183.74 1.06 11.24 25.16 8.74 18.26 2157.23 2.2 0.2 2.7 1.0 4.7 5.3 0.4 1.1 3.4 35.0 64.3 0.6 0.2 0.9 0.2 2.1 1.9 0.5 0.8 0.9 0.4 0.1 0.5 1.1 0.2 2.9 0.7 0.2 0.9 0.9 2.3 13.3 0.1 0.8 1.8 0.6 n/a 4.3 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 358 Chapter 7. 31HK2521 NORT WALL cm bd 0 A Lev 1 10 Tree Roots Lev 2 A’ 20 E Lev 3 E 30 Lev 4 Tree Burn Stain TU3 N504.0/E496.0 0 40 N504.0/E497.0 1m Soil Strata A: A’: E: Pale Brown (10YR 6/3) Sand, Satined Very Dark Grayish Brown (10YR 3/2) in Part By Charcoal Leaching from Prescribed Burns White (5YR 8/1) Sand, Slope Wash Very Pale Brown (10YR 7/3) Sand Figure 115. Profile Drawing, TU3, North Wall, 31HK2521. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 359 Chapter 7. 31HK2521 Table 106. Stone Tools and Debitage by Level for SB 14 and TU 3, 31HK2521. VERTICAL DISTRIBUTION OF DEBITAGE AND STONE TOOLS BY RAW MATERIAL TYPE FOR SB 14 AND TU3: 31HK2521 SB 14 RAW MATERIAL TYPE 2 Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 3 1 1 1 6 2 1 8 DEBITAGE LEVEL 4 5 1 1 2 3 1 2 1 5 6 6 1 1 2 GRAND TOTAL 2 2 13 1 5 23 STONE TOOLS LEVEL 2 Palmer I Corner Notched TU3 RAW MATERIAL TYPE Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 2 1 1 3 DEBITAGE LEVEL 4 GRAND TOTAL 1 2 1 4 1 1 2 2 STONE TOOLS LEVEL 2 Table 107. Vertical Distribution of Ceramics in SB 14 and TU 3, 31HK2521. VERTICAL DISTRIBUTION OF CERAMIC TYPES IN SB14 AND TU3: 31HK2521 SB14 CERAMIC TYPE Hanover IIa Cord Marked Hanover IId Cord Marked Yadkin IIIf Dowell Impressed GRAND TOTAL 1 4 4 LEVEL 2 1 2 1 4 3 1 1 GRAND TOTAL 1 7 1 9 VESSEL NO. V145 V146 V147 TU3 CERAMIC TYPE Hanover IIa Cord Marked Hanover IIa indet. Hanover IId Cord Marked Residual Sherd GRAND TOTAL 1 LEVEL 2 6 20 2 28 3 1 2 3 GRAND TOTAL VESSEL NO. 6 1 22 2 31 V213 V213 V146 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 360 Chapter 7. 31HK2521 Table 108. Partial Completeness Percentages for Vessels in SB 14 and TU 3, 31HK2521. PARTIAL COMPLETENESS PERCENTAGES FOR VESSELS IN SB 14 AND TU3: 31HK2521 VESSEL NO. COUNT WEIGHT PERCENT COMPLETENESS 145 1 2.76 0.2 146 29 159.5 11.6 147 1 7.89 0.6 213 7 19.27 1.4 Residual 2 1.08 0.1 40 190.5 3.4 GRAND TOTAL the two more complete sherd aggregates was likely deposited as whole vessels. Instead, their last prediscard use appears to have been as recycled large sherds. Here again, intact Woodland period occupation elements were identified in the upper 20 to 30 cm of soil on the site. EVALUATION The Quewhiffle site, 31HK2521, is recommended eligible for inclusion on the National Register of Historic Places under criterion d, “that have yielded, or may be likely to yield, information important in prehistory or history’ in accordance with 36CFR60.4. The importance of the site is contained in the extensive Woodland period occupation. Only rarely are sites with such dense Woodland occupation encountered in the uplands of the North Carolina Sandhills. The fact that the site is situated on a ridge terrace of a major Sandhills stream, Rockfish Creek, should not be overlooked. However, the reasons for the dense Woodland occupation at this site in particular must go beyond this simple fact, as there are many sites along the creek without comparable Woodland occupation. Directly below the site is an extensive wetland created by a decreased stream gradient within a widened-out basin (Figure 68). Swampy and con- tinually wet soil patches such as this would have provided an opportune microenvironment to cultivate certain of the Eastern Agricultural Complex (Ford 1985:347-349) seed crops such as sunflower (Helianthus annus), sumpweed (Iva annua var. macrocarpa), goosefoot (Chenopodium bushianum), maygrass (Phalaris caroliniana), knotweed (Polygynum erectum L.), little barley (Hordeum pusillum Nutt.) and giant ragweed (Ambrosia trifida). The status of indigenous cultivation has not yet been elucidated or confirmed for the Sandhills, but sites such as Quewhiffle may supply this evidence. Numerous Late Woodland sites in the Coastal Plain and Piedmont have produced examples of indigenous seed crops (Scary and Scary 1997) and it is likely that these subsistence patterns were common to cultures throughout the Southeast Atlantic Slope by at least the Late Woodland time horizon. If the Quewhiffle site owes it’s dense Woodland occupation to indigenous seed crop cultivation, then we should not expect it to be unique. Agriculturalists and cultivators are brought to this adaptation by population packing (Binford 1983:210-213; Netting 1968, 1990, 1993) and we can expect the Sandhills streams to be dotted with Quewhiffle-like sites at locations of expanded floodplain development if Late Woodland populations in the region were cultivating crops. Site 31HK2502, located immediately south of the Quewhiffle site, may represent another residential node in this hypothesized Late Woodland horticultural settlement system along Rockfish Creek. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 361 Chapter 7. 31HK2521 A wider view of one of these linear stream settlement patterns is supplied by dense Woodland period sites along McPherson Creek and its tributaries (Cable and Cantley 2005a) on the east side of Fort Bragg. The various calculations of vessel populations generated for the Quewhiffle site indicate that portions of somewhere between 2,000 and 10,000 ceramic vessels ultimately came to rest there over the course of the Early, Middle and early Late Woodland periods. This suggests a relatively intense and longlived use of the site, probably on a returning seasonal basis. The shovel test and test unit exposures reveal that much of this occupation consisted of Type I elements, most likely representing household residential outputs of nuclear and extended families. Although the upper solumn has been disturbed by prescribed burn activities, the lack of a deep A-Horizon suggests that the site was never extensively plowed. Both sherd/vessel aggregates and lithic debitage elements in the upper 20 to 30 cm of deposit are well preserved. Major disturbances to these deposits appear to have been instigated by post-depositional events related to reoccupations of the site by precontact groups. Lithic elements are best preserved and provide an excellent basis for reconstructing the functional and spatial organization of the Woodland occupation. The fact that the horizontal distributions of the lithic elements are preserved indicates that the spatial and organizational patterns of individual Woodland period occupations can be reconstructed here in spite of the disorganization of the ceramic assemblage. It should be noted, however, that the monitoring of the spatial distributions of attributes (i.e. sherd size, erosional wear, etc.) sensitive to formation processes in sherd assemblages can yield important data on the structure of occupations and camp layout as well (see DeBoer and Lathrap 1979; Stanislawski 1978). Thus, even disorganized ceramic assemblages can aid in the reconstruction of site structure. The relatively greater organization of the Late Woodland ceramics, as demonstrated from the Block 3 investigation, suggests that the latest occupations will be best preserved. Data recovery at the Quewhiffle site can supply important additional information concerning the function of the various Woodland occupations, subsistence practices and larger changes in Woodland period adaptations and settlement. Finally, due to its tremendous variability in paste and temper characteristics, the ceramic data contained in the site have great potential for contributing to a more detailed understanding of the local chronological sequence (see Chapter 11 for an elaboration of this point). The high density of check stamping is unusual from the standpoint of recognized ceramic patterns in the region and suggests that the site contains an Early to Middle Woodland component that has not yet been well defined nor appreciated for the Sandhills. Moreover, the Hanover III and Yadkin III assemblages appear to express an early Late Woodland manifestation that has been previously unanticipated for the region. The presence of interior surface treatments (i.e. serrated scraped and floated) on these subseries that correlate with Late Woodland series (i.e. Uwharrie and Colington/Townsend) in neighboring regions provides additional support for placing these assemblages in the Late Woodland chronological position. Extended work at this site would provide a basis for deriving a much clearer picture of the chronological positions of these components through the extraction of TL-samples and probably charcoal samples for radiocarbon dating either from sooting on vessel interiors or from cultural features. Although cultural features were not positively identified during testing, the deep intrusion of ceramics into lower levels as documented in SB 16 demonstrates the feasibility that subsurface features are preserved at the site. The probability that subsurface features are present is also supported by the ceramic completeness values, which are characteristic of habitation sites rather than short-term temporary camps (see Chapter 11). It should be noted that the presence of cultural features is not well established by limited subsurface testing programs (Lightfoot 1986). Often times, when broad exposures are established over sites where features were not identified by testing or that show very little promise for such, large arrays of features are discovered. A case in point has recently been demonstrated by Archaeological Research Consultants in North Carolina (John Mintz, personal communication, August, 2011). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 362 Chapter 8. 31RH478 Site 31RH478 is located in Richmond County within the confines of Camp MacKall and just to the east of the airfield (Figure 116). The site can be reached by traveling approximately 0.2 miles west on Glider Road from 17th ABN DIV Road to a dirt road. Take the dirt road northward and after another turn to the left and then one to the right over a distance of about 0.1 miles the site is situated on the west side of a north-south running dirt road (Figure 117). The SCS soil survey for Richmond County, NC (Evans 1999) classifies the landform as Wakkula sand (WcB), an excessively drained soil that typically forms on upland interfluves and sand sheets in the Coastal Plain. The site rests on the edge of a broad ridge finger approximately 215 m from an unnamed tributary of Big Muddy Creek that now forms the northern extension of Moss Gill Lake. The confluence of the two drainages is approximately 0.9 km south of the site. Although the site has been subject to tree thinning and prescribed burning, the surface topography is relatively free of modern disturbances (Figure 118). Elevations within the site boundaries ranged between 100.5 m and 99.75 m amsl (Figure 119). The apex of the landform is situated slightly above and west of the site (Figure 120). Topography drops off gently to the east and south leading to the tributary drainage channel. Slope from west to east within the site limits is about 1.25 percent. Vegetation across the site consisted of a thinned stand of young longleaf pines and an understory of turkey oak and wiregrass. Much of the surface was exposed by a com- bination of wind erosion and xeric conditions from excessive soil drainage. Maximum site dimensions were established at 40 m x 80 m, covering an area of approximately 0.5 acres (2,100 m2). PREVIOUS RESEARCH Site 31RH478 was originally recorded by TRC Garrow Associates, Inc in the spring of 2001 during a Phase I survey of portions of Fort Bragg and the KK1 Tract at Camp MacKall (Ruggiero 2003:438-439). The site boundaries established at that time measured only 10 x 15 m (Figure 121). Phase I site definition proceeded through the deployment of 15 m-interval shovel tests around a positive survey transect shovel test. Including site discovery transect shovel tests, 17 shovel tests were excavated in total, two of which were positive. Ruggiero (2003:438) described three soil zones exposed in shovel tests. The upper zone was composed of very dark grayish brown sand extending to a depth of 10 cm bs. The intermediate zone was brownish yellow sand that contacted yellowish brown sand at about 40 cm bs. The latter zone continued lower than the base of shovel testing, which occurred at about 75 cm bs. All of the artifacts were recovered between 60 and 75 cm bs in the lowest stratum. Artifacts collected from shovel tests included nine pieces of porphyritic rhyolite debitage and two pieces of undefined rhyolite debitage. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 363 Chapter 8. 31RH478 640000 638000 639000 640000 3880000 3879000 3878000 3877000 31RH478 3877000 3878000 3879000 3880000 3881000 639000 3881000 638000 Figure 116. Location of Site 31RH478 (Pine Bluff, NC 7.5’ USGS Quadrangle) Scale: 1:24,000. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 364 Chapter 8. 31RH478 637800 638000 638200 638400 3878400 17 th AB 3878200 N Beaver DI V Ro ad 3878200 3878400 3878600 637600 3878600 637400 Mackall Airfield 3878000 er Glid d Roa 3877800 3877800 3878000 nd Po 31RH478 3877400 3877200 3877200 3877400 ke La 3877600 ill G 3877600 ss Mo 637400 637600 637800 638000 638200 638400 Figure 117. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31RH478 (Scale: 1 inch = 200m, Projection: UTM, NAD83). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 365 Chapter 8. 31RH478 Figure 118. View of Site 31RH478, Looking North to SB2 from N480/E500. Ruggiero (2003:438) concluded that the site appeared to represent a single occupational event. Furthermore shovel testing revealed that the debris from the occupation was in tact and free of modern disturbances due to its great depth. Due to the potential for the site to yield important data concerning the structure and organization of lithic camps and activity areas in the Sandhills from a pure and unmixed context, the site was recommended potentially eligible for inclusion on the National Register of Historic Places. FIELDWORK OVERVIEW Phase II fieldwork was implemented in four stages, each stage building on the information gener- ated from the previous stage. Stage I investigation consisted of a program of 10 m-interval shovel tests that were excavated to establish firm site boundaries. Site boundaries were determined by the documentation of two consecutive negative shovel tests in all directions on a 10 m grid, or in some cases when wetland was encountered. Once artifact distributions from the Stage I program were mapped and evaluated, the Stage II investigation involved the excavation of closer-interval shovel tests at 5 m- and 2.5 m-intervals around the few positive shovel tests identified in the Stage I sample. This constituted a departure from the approach taken at the other sites in the package, where an attempt was made to cover most of the site areas with 5 m-interval shovel tests. However, given the extremely low density of artifacts across 31RH478, it was felt that better use of the shovel test volume could be made through concentrated Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 366 Chapter 8. 31RH478 580 Scale 0 00 20 m oa tR Dir 100. 570 d .50 99 75 99. 99 .7 5 560 550 N .25 100 540 530 31RH478 520 10 0. 75 99.50 510 500 .25 100.5 99 0 490 480 0 100.0 470 99.75 460 99.50 450 .25 99 440 Di 450 460 470 480 ad 490 500 510 520 530 540 0 .5 98 o rt R 550 560 Figure 119. Site 31RH478, Base Map. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 367 Chapter 8. 31RH478 Figure 120. View of Site 31RH478, Looking Southwest from Dirt Road at N520/E540. investigation in limited locations. Stage III investigations involved the deployment of close-interval shovel tests of 1.25 m-intervals around the perimeter of targeted shovel tests yielding deposits deemed of theoretical importance for evaluating the site. These close-interval shovel test locations were identified as sample blocks. Again, the general approach across the site package was to establish sample blocks consisting of eight close-interval shovel tests spaces at 1.25 m-intervals around Stage I and II anchor tests. In the special case of 31RH478, however, large numbers of close-interval shovel tests were placed at only two locations to more fully define and map artifact distributions in these discrete deposits. The sample blocks were also instrumental in guiding the precise placement of 1 x 1 m test units, which constituted the final stage of the investigation. Test units were placed in locations judged to provide optimal opportunities to reconstruct the occupation history of the site and to assess the integrity and scientific importance of the deposits. Stage I investigations were initiated by relocating the datum established by Ruggiero (2003) and laying out a 10-m interval grid with a total station. The datum was easily relocated when the concrete footer noted on the survey site plan (Figure 121) was found. The grid coordinates (N500/E500) for the datum were maintained for the Phase II work. Relative topographic elevations at each 10 m-grid intersection were recorded with the total station and marked with a pin flag. The elevation values (in meters) illustrated on the base map (Figure 119) are approximated real elevations based on the correlation of GPS points with a geo-referenced copy of the USGS, Pine Bluff, NC 7.5-minute topographic quadrangle map. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 368 Chapter 8. 31RH478 E440 E410 E470 E500 E560 E530 E590 N590 Trail N560 T7 N530 concrete footer + N500 T6 N470 T5 N440 To Moss Gill Lake N410 Positive STP Positive transect STP Negative STP Negative transect STP Site boundary + GPS Contour (approximate) * Surface find (if any) Tests shown are not to scale 0 Meters 45 Figure 206. Site Plan of 31HK478. Figure 121. Site 31RH478, Phase I Survey Sketch Map, TRC (from Ruggiero 2003). 439 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 369 Chapter 8. 31RH478 Cultural and landform features were also mapped and additional elevations were taken when appropriate to produce the topographic map. Shovel tests measured 30-x-30-cm and were square in shape. Excavation proceeded in 10 cm arbitrary levels measured from the surface. Sixty-six Stage I shovel tests were excavated at 10 m-intervals to define the limits of the site (Figure 122). Only five of these were positive for precontact artifacts. Two shovel tests located outside of the established Phase II site boundaries produced twentieth-century historic artifacts, but these finds were not considered relevant to the task of evaluating the precontact deposits. Maximum site boundaries were established at 80 m x 40 m, covering an area of about 2,100 m2, or approximately 0.5 acres. Mean shovel test depth for the Stage I sample was 74.24 cm bs (SD=8.78) with a mode of 70 cm bs and a range of 50 to 100 cm bs. A strong sandy loam substrate was absent in the upper solumn of the site and shovel tests were generally terminated between 70 and 80 cm unless artifact occurrence dictated deeper excavation. The SCS soil survey for Richmond County (Evans 1999) maps the area of the site as Wakkula sand (WcB), an excessively drained soil that typically forms in sandy Coastal Plain sediments on upland interfluves and stream terraces. The on-site soil profile corresponds well to the typical Wakkula pedon. The A-Horizon represents a plow zone as is evidenced by its smooth and regular contact with the E-Horizon. The Ap-horizon extends to a depth of 10 to 15 cm bs and is composed of a grayish brown (10YR 5/2) sand. A charred mixture of prescribed burn charcoal and sand occurs as a 5 cm thick lens resting on top of the Ap-horizon and potions of the A-Horizon exhibit very dark gray (10YR 3/1) coloration as a result of leaching from the prescribed burn zone. The regularized contact of the base Below the A-Horizon rests an E-Horizon composed of light yellowish brown (10YR6/4) to brownish yellow (10YR 6/6) sand. It transitions into a strong brown (7.5YR 5/6) Bt-Horizon of compact loamy medium to coarse sand at 55 to 75cm bs. The Stage II shovel test sample consisted of 5 m- and 2.5 m-interval shovel tests placed around the five shovel tests producing precontact artifacts. Sixty-one Stage II shovel tests were excavated within the sample frame (Figure 123). Sixteen of these were positive. Mean shovel test depth for the Stage II sample was 63.46 cm bs (SD=18.70) with a mode of 70 cm bs and a range of 20 to 110 cm bs. The Stage III shovel test sample was distributed at two locations, SB1 and SB2, to further define the distributional and chronological characteristics of two dense debitage concentrations (Figure 124). Eighty-two Stage III shovel tests were excavated at typically 1.25 m-intervals around targeted shovel tests in the sample blocks. Thirty-eight of these tests yielded artifacts. Mean shovel test depth for the Stage III sample was 85.91 cm bs (SD=14.43) with a mode of 90 cm bs and a range of 40 to 100 cm bs. Once the results of the Stage III shovel test sample were analyzed, four test units were placed at specific locations within SB2 in attempt to delineate further the function and culture chronological association of one of these debitage concentrations (Figure 125). The horizontal extent of the debitage concentrations in both sample units had grown considerably as Stage III shovel testing progressed, but diagnostic artifacts had not been found. It was felt that focusing on one concentration with test units would increase our chances of finding diagnostic artifacts and also of reconstructing the spatial organization and structure of these debris scatters. In addition to the four 1-x1-m test units, four contiguous 0.5-m-x-0.5-m test units were excavated along the northern and eastern perimeter of TU4 to follow artifact patterning further. ARTIFACT INVENTORY Chipped stone debitage and tools dominated the artifact inventory from 31RH478. The inventory consisted of 784 items including 738 pieces of lithic debitage, 4 chipped stone tools, 4 cores, 20 rock and fire-cracked rock fragments and 18 twentieth century historic artifacts. Appendix B contains the artifact database for the project, while appendices C through H provide additional data on lithic chipped and ground stone tools. Descriptions of the various artifact and lithic raw material types are presented below. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 370 Chapter 8. 31RH478 580 20 m Dir oa tR 570 Scale 0 00 100. d .50 99 75 99. 99 .7 5 560 550 N .25 100 540 530 31RH478 520 10 0. 75 99.50 510 500 .25 100.5 99 0 490 480 0 100.0 470 99.75 460 99.50 450 Di Negative ST 450 460 470 480 ad 490 500 510 520 530 540 0 .5 98 o rt R 5 440 .2 99 Positive ST 550 560 Figure 122. Site 31RH478, Stage I Shovel Test Sample. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 371 Chapter 8. 31RH478 580 20 m Dir oa tR 570 Scale 0 00 100. d .50 99 75 99. 99 .7 5 560 550 N .25 100 540 530 31RH478 520 10 0. 75 99.50 510 500 .25 100.5 99 0 490 480 0 100.0 470 99.75 460 99.50 450 Di Negative ST 450 460 470 480 ad 490 500 510 520 530 540 0 .5 98 o rt R 5 440 .2 99 Positive ST 550 560 Figure 123. Site 31RH478, Stage I and II Shovel Test Sample. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 372 Chapter 8. 31RH478 580 20 m Dir oa tR 570 Scale 0 00 100. d .50 99 75 99. 99 .7 5 560 550 N .25 100 540 31RH478 530 520 10 SB1 0. 75 99.50 510 500 490 99 0 .25 SB2 100.5 480 0 100.0 470 99.75 460 99.50 450 Di Negative ST 450 460 470 480 ad 490 500 510 520 530 540 0 .5 98 o rt R 5 440 .2 99 Positive ST 550 560 Figure 124. Site 31RH478, Stage III Shovel Test Sample and Locations of Sample Blocks. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 373 Chapter 8. 31RH478 580 20 m Dir oa tR 570 Scale 0 00 100. d .50 99 75 99. 99 .7 5 560 550 N .25 100 540 530 31RH478 520 10 0. 75 99.50 510 SB2 500 TU4 480 .25 TU2 TU1 0 490 99 100.5 TU3 0 100.0 470 99.75 460 99.50 450 Di Negative ST 450 460 470 480 ad 490 500 510 520 530 540 0 .5 98 o rt R 5 440 .2 99 Positive ST 550 560 Figure 125. Site 31RH478, Test Unit Locations. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 374 Chapter 8. 31RH478 Lithic Raw Material Types Lithic raw material identification proceeded on two different levels. At the macro-scale, individual specimens were typed in accordance with major hard rock geological terminology to achieve comparability with previous projects at Fort Bragg. At the micro-level, however, an attempt was made to further partition some of the hard rock types into “core groupings” to distinguish between specimens derived from single hard rock types that may have originated from different cores. This was done to facilitate spatial analysis in close-interval shovel test grids where overlapping of distinct deposits or complex multi-household occupations were suspected. Core grouping analysis was undertaken principally in metavolcanic types where subtle differences in patina coloration, core color, phenocryst size and density, and matrix texture was easily discerned and monitored. Descriptions of the various rock types and the associated core grouping sub-types recognized in the collection are presented below. Metavolcanic Types Metavolcanic types, as a group, comprise the majority of the chipped stone collection from the site (Table 109). As a group, they represent 99.7 percent of the total chipped stone artifact inventory (i.e. debitage and chipped stone cores and tools). Descriptions of the four identified metavolcanic types and associated core group subtypes follow. Since the Stage I sample is so small and the Stage II sample is biased toward the positive Stage I shovel tests, an unbiased estimate of raw material type representation cannot be generated from the Phase II sample. The Stage I and II shovel test sample percentages are nevertheless reported, but should not be viewed necessarily as representative of the larger site area. The Stage I and II sample contains only 43 specimens. The typological system varies somewhat from earlier reports in response to the recent stone quarry sourcing study for the Carolina Slate Belt conducted by Steponaitis et al. (2006). (1) Mill Mountain Rhyolite (MMR). MMR was not clearly differentiated in the sourcing study (Steponaitis et al. 2006), but it is a readily recognizable type in the assemblages from Fort Bragg. Mill Mountain is located between the Uwharries Southern and Uwharries Eastern source areas on the opposite bank of the Yadkin River from Table Top Mountain. It was not included in either of these source areas, however, because of the presence of distinctive clear, glassy quartz phenocrysts. Daniel and Butler (1996:18) describe this type as a medium gray, aphanitic, rhyolite porphyry with sparse, glassy phenocrysts of quartz generally less than 1 mm in diameter. Very fine-grained, disseminated grains of pyrite are also present. Thin sections indicate a microcrystalline matrix composed primarily of feldspar and quartz, with some biotite, chlorite, and disseminated pyrite. The specimens assignable to this type from the 31RH478 collection are characterized by a dark reddish gray to moderate brown coloration when not patinated. Four subtypes were assigned to the type, based primarily on the completeness of the observation field and patination. Subtype R8 corresponds to specimens that exhibit sparse glassy quartz phenocrysts only. Subtypes R8Hhf, R8Hhm and R8Hhc are fine-, medium- and coarse-grained specimens respectively that exhibit sparse glassy quartz, white subhedral quartz phenocrysts and, rarely, feldpsar phenocrysts. Although these additional phenocryst types were not included in the type description, the subtypes containing them were included as a variant of Mill Mountain Rhyolite due to their co-occurrence with glassy quartz phenocrysts. Mill Mountain Rhyolite makes up 20.9 percent of the combined Stage I and II chipped stone sample. (2) Type I Rhyolite Tuff (Type I RT). An exact correlate of this type was not discernible in the sourcing study (Steponaitis et al. 2006). Benson (1999:30) describes it as fine-grained dark green to gray material that looks and feels very similar to Piedmont chert. He noted that plagioclase phenocrysts were sporadically present. Two subtypes were identified in the 31RH478 collection, R2A and R2Ah. The basic subtype is R2, which was not identified at the site. Subtype R2 is a microcrystalline dusky green texture material whose edges are semi-translucent. Subtype R2A specimens exhibit a pale green patina, but broken edges reveal that the underlying material corresponds to R2. Darker green splotches and swirls on the faces of these specimens represent more resistant matrix. The matrix is aphantitic and microcrystalline in texture and is highly isotropic in fracture characteristics. Also recognized was subtypes R2Ah, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 375 Chapter 8. 31RH478 Table 109. Lithic Raw Material Frequencies by Chipped Stone Tool Class, 31RH478. Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite 71 2 2 660 1 1 UTILIZED FAKE PROJECTILE POINT FRAGMENT. FLAKE BLANK CORE FRAGMENT BIFACE FRAGMENT RAW MATERIAL TYPE DEBITAGE LITHIC RAW MATERIAL FREQUENCIES BY CHIPPED STONE TOOL CLASS: 31RH478 GRAND TOTAL 1 74 1 3 1 1 664 Uwharries Southern Rhyolite (1) 3 3 White Quartz 2 2 GRAND TOTAL 738 a hydrated condition of R2A. The origin of Type I Rhyolite Tuff is not well understood and it is possible that it represents a very fine-grained metamudstone similar to some of those described for the Chatham Pittsboro source (see Moore and Irwin 2006:27-28; Stoddard 2006:57). In support of this is the occurrence of what appear to be bedding laminations in some of the patinated material. Moreover, the color of the patinated specimens is similar to that reported for the Chatham Pittsboro source. Type I Rhyolite Tuff makes up 7.0 percent of the combined Stage I and II chipped stone sample. (3) Uwharries Eastern Rhyolite (UER). Stoddard (2006:52) describes the specimens from this sourcing area as light to dark gray metadacite porphyry or crystal lithic tuff. All samples collected contained plagioclase feldspar and white quartz phenocrysts, ranging from less than 2 percent to 7 percent phenocryst density. Common metamorphic minerals within the matrices include biotite, stilpnomelane and calcite. Similar dacites and rhyolites are typical of the Uhwarries Southeastern source (Stoddard 2006:55-56) as well, but many of these specimens contain silica levels beyond the range typical of igneous rocks. Macroscopic criteria for distinguishing between lithic material from these two sources does not at present exist. Consequently, specimens from both source areas may be present in the analyzed 1 3 1 2 1 746 lithic assemblages at Fort Bragg. Typical specimens from 31RH478 are characterized by dark gray (nonpatinated) and grayish green (patinated) matrices with splotches of silicate and variable percentages (2 to 4 percent) of white quartz and plagioclase feldspar phenocrysts. Some specimens exhibit purplish to brown background matrices. Quartz phenocrysts predominate over feldspar in the examples from the site. Type I Rhyolite, which has been distinguished separately in previous reports (Cable and Cantley 2005b, 2006, 2010), is also subsumed under this type. Five subtypes were identified in the collection: R8Ihf, R8Ihf(2), R8Ihm, R8Ihc and R8Ic all of which contained white quartz and plagioclase feldspar phenocrysts in variable percentages. They are primarily distinguished from each other by differences in texture. Respectively, the subtypes correspond to fine-, fine- medium-, coarse- and coarse-grained specimens. R8Ihf(2) is further distinguished by the presence of only small, sparse phenocrysts. R8Ic is further characterized by extensive patination. UER makes up 69.8 percent of the combined Stage I and II chipped stone sample from the site. (4) Uwharries Southern Rhyolite (USR). This type includes sources found on Morrow Mountain and surrounding areas, most notably Table Top Mountain. Daniel and Butler (1996:10-13) describe this type as a dark gray, aphanitic, aphyric rhyolite, that commonly Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 376 Chapter 8. 31RH478 exhibits flow-banding. Weathering tends to bring out the flow lines visually, which generally alternate in a pattern of light and dark gray. Although the matrix is usually homogeneous, some specimens exhibit small spherulites of less than 1 mm in diameter (see Stoddard 2006:52). Thin sections indicate a microcrystalline intergrowth of feldspar and quartz, with minor biotite and chlorite elements. The individual minerals are difficult to distinguish and strings of dark minerals can occur along fracture planes. The specimens identified as USR in the 31RH478 assemblage are assigned to only one, USR (1), of the two groupings associated with the type. USR (1) is composed of R3/R3t subtypes, which exhibit an extremely fine grained or microcrystalline dark gray matrix that is semi-translucent when viewed on attenuated edges. Finely disseminated feldspar is commonly seen on exposed surfaces, but the small blotches do not form phenocrysts. Two subtypes were recognized in the 31RH478 collection, R3t and R3patf. The former is non-patinated, while the latter appears to represent R3t material that is patinated to a light tan color. USR (1) makes up 2.3 percent of the Stage I and II chipped stone sample. Quartz Types Pure quartz forms as veins in igneous and metamorphic rock formations characterized by slow crystallization. This generally produces anhedral, milky structures (Spock 1962). In some cases, however, crystallization occurs earlier in the sequence, producing clear crystals with euhedral structure. Both vein and euhedral crystal quartz were utilized by precontact groups in the project area, but most of it appears to have come from river gravel deposits. A great deal of variation was evident in the vein quartz category ranging from mixed clear and cloudy exposures to opaque milky white. Since these characteristics appeared to be repeatedly present on single cores when debitage concentrations were analyzed, all of this variation was subsumed under the single white quartz subtype Q5. The category “crystal quartz” (CQ) was retained for instances where glass-like examples and crystal facets strongly suggested a euhedral crystal origin. “Rose quartz” was reserved for quartz specimens with a high degree of rubification, a commonly observed attribute in quartz cobbles. Only white quartz was represented in the 31RH478 collection. Only two pieces of white quartz debitage were recovered during the investigation, neither of which occurred in the Stage I and II shovel test sample. In addition, 13 pieces of fire-cracked rock and one cobble were composed of white quartz. These items totaled 512.4 gm. Minority Types Several other rock types were recognized in the 31RH478 collection. Descriptions of each are presented below. (1) Arenite. This material has been previously identified as ferruginous sandstone, but thinsectioning of a piece of this material indicates that a more accurate characterization is arenite, or hematite cemented quartz sandstone (Appendix K). Three pieces of arenite were classified as fire-cracked rock. It occurs in a variety of shapes including flattened and irregular nodules. The exterior of the nodules consist of fine-grained ruddy brown precipitate while the centers consist of dense, reddish brown to purple, densley cemented sandstone. The material was commonly used for rock hearths by the precontact inhabitants of Fort Bragg. Five pieces of arenite were collected during the investigation, totaling 119.67 gm. Float, fire-cracked rock and rock were identified. (2) Quartzite. A single piece of float composed of a dark gray quartzite resembling orthoquartzite from the Santee River Valley of South Carolina (Anderson 1982:120-124). The latter material derives from the Thanetian Black Mingo Formation and it is typically described as chert and chalcedony cemented quartz arenite or chert and chalcedony cemented sandstone. Large cobbles and chunks of lag gravel made of this material are found in terrace exposures along the Lower Santee and Black River valleys. The fact that the material was found in float suggests a local origin for the specimen from 31RH478. Lithic Artifacts Definitions of the various lithic artifact classes were presented in Chapter 4. Specific descriptions of the 31RH478 lithic collection are presented below. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 377 Chapter 8. 31RH478 Debitage sode. It is likely that similar proportions characterize directional core reduction from raw nodules. The UER sample is positioned within the lower end of this range, suggesting that the debitage was produced from either raw nodules or from minimally prepared directional cores with high percentages of cortex remaining. The Cortical percentages are lower for MMR, which suggests that the debitage from this raw material type was derived from a minimally prepared directional core or cores. The debitage from both raw material types exhibits river cobble cortex, indicating that original nodules were extracted from terrace contexts or from Triassic Basin deposits. The data are insufficient to present a detailed discussion of reduction strategies for the other raw material types, however, the small size of flakes and the absence of cortex in the Type I RT and USR (1) material would suggest that the recovered debitage was produced from tool maintenance activities. The debitage classification is adapted from formal (Bradley 1973; Frison and Bradley 1980; House and Wogaman 1978; Newcomer 1971), attribute (Moore 2002; Shott 1994), and mass (Ahler 1989; Moore 2002) analysis approaches. Debitage classes were devised to identify both reduction stages and reduction/production systems (ie. biface core reduction, directional core reduction, and flake blank production). This classification was supplemented with the recording of three attributes: (1) size class, (2) percent cortex, and (3) condition. Table 110 presents the complete inventory (all field stages) of debitage type by raw material type. Core Flakes are unusually abundant in the two main raw material types. Core flake comprise nearly 74 percent of the identifiable debitage types in the MMR sample and 79 percent by weight. Similar proportions are registered in the UER sample in which 66 percent of the identifiable debitage types consist of core flakes by count and 77 percent by weight. The high proportion of core flakes indicates that the debitage concentrations produced by these raw material types involved directional core reduction as the main activity. Cores Metric and attribute data on cores are provided in Appendix C. Two classes of cores were identified in the 31RH478 collection: (1) Core Fragments and (2) Flake Blanks. Each class is discussed below. (1) Core Fragments. Angular pieces of raw material containing portions of one or more flake scars suggestive of directional core reduction, but which lacked striking platforms, were classified as core fragments. Three such fragments were recognized in the collection, two of which were made of MMR while a third was composed of UER The speci- The distribution of cortex on dorsal faces and platforms of debitage provides a basis for further evaluating the nature of reduction strategies associated with the various raw material types (Table 111). Dibble et al. (2005:550) indicate that raw nodule reduction from bifaces should yield between about 6 and 21 percent flakes with cortex for any single epi- Table 110. Weight and Frequency Distributions of Debitage Type by Raw Material Type, 31RH478. WEIGHT* AND FREQUENCY DISTRIBUTIONS OF DEBITAGE TYPE BY RAW MATERIAL TYPE: 31RH478 LITHIC RAW MATERIAL TYPE BRF** Count BTF Weight Count FBR Weight CORE FLAKE Weight Count Weight 9 11.85 25 45.84 Count Weight FRAGMENT GRAND TOTAL Count Weight Count Weight 37 35.15 71 92.84 2 1.01 2 1.01 336 224.46 660 621.55 Uwharries Southern Rhyolite (1) 3 0.42 3 0.42 White Quartz 2 1.28 2 1.28 380 262.32 738 717.10 Mill Mountain Rhyolite Count CHUNK Type I Rhyolite Tuff Uwharries Eastern Rhyolite GRAND TOTAL * WEIGHT = GRAMS 1 1 0.40 0.40 21 21 4.22 4.22 87 96 85.56 97.41 214 239 303.61 349.45 1 1 3.30 3.30 ** BRF: Biface Resharpening Flake Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 378 Chapter 8. 31RH478 Table 111. Percent Cortex on Debitage and Percent Core Flakes by Raw Material Type, 31RH478. PERCENT CORTEX AND CORE FLAKES FOR DEBITAGE BY LITHIC RAW MATERIAL: 31RH478 PERCENT W/CORTEX PERCENT CORE FLAKES 71 4.2 73.53 2 0.0 - 660 6.5 66.25 Uwharries Southern Rhyolite (1) 3 0.0 - White Quartz TOTAL 2 0.0 - RAW MATERIAL TYPE TOTAL DEBITAGE Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite mens are all small and, although they were derived from directional cores, they exhibit very few of the features of said cores. A better idea of the physical structure of the directional cores can be seen from the characteristics of the debitage (Figure 126:A-F). Cobble cortex patches commonly occur on the dorsal faces of large flakes and the thick platforms on core flakes display near 90° striking angles. The flake scar structure of the original directional cores is clearly evident on the larger core flakes in the collection (i.e. Figure 126:B, C, G and J). (2) Flake Blanks. Flake blanks were removed from a core for the purpose of further modification or use as a tool (Bradley 1973:6). However, they do not exhibit obvious macroscopic evidence of use or modification. One specimen was tentatively identified as a flake blank (Figure 126:G). The specimen was placed in this category primarily because of its large size, but there are a number of large flakes in the collection that appear to have been nothing more than manufacturing debris. Bifaces Bifaces are derived from both large mass packages (i.e. unmodified and/or prepared cores) and flake blanks. Determination that a biface was derived from a flake blank was made when flake characteristics were observed on the item, such as remnant striking platforms, flake curvature, or bulbs of percussion 692 from flake detachment. Only a single, small lateral section from Type II Biface was identified in the collection. Type II Bifaces, in the main, represent intermediate stage preform rejects (Daniel (2002:51-54). They are roughly flaked and exhibit only large, conchoidal flake scars. No edge trimming is evident, but an ovate to lanceolate outline is generally recognizable and the perimeter of the edge margin is entirely formed. Facial retouch, however, may not be complete and features of the original blank are sometimes evident. Projectile Points Projectile points are bifacially flaked tools with retouched haft elements. In addition, projectile point fragments consist of specimens with the fine bifacial retouch characteristic of finished projectile points, but not necessarily exhibiting evidence of a haft element. The collection from 31RH478 consisted of two tip fragments, one from a large, thick projectile point of MMR (Figure 126:I) and other from a broad, but thin point made of Type I RT (Figure 126:H). Metric and descriptive data on the projectile points can be found in Appendix E. Utilized Flakes One non-retouched flake exhibited wear patterns on one or more edges, indicating their use as Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 379 Chapter 8. 31RH478 Figure 126. Lithic Artifacts, Site 31RH478. A: Flake Fragment (m176), B-E: Core Flakes (m207, m207, m80, and m86 respectively), F: Flake of Bifacial Retouch (m127), G: Flake Blank (a366), H-I: Projectile Point Tips/Upper Blade Fragments (a17 and a206 respectively), J: Utilized Flake (a153), K: Cobble Fragment (m363). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 380 Chapter 8. 31RH478 expedient tools (Figure 126:J). Metric and attribute data for utilized flakes are presented in Appendix G. The specimen was made of UER on an elongated core flake. It measured 55.64 mm long, 30.44 mm wide and was 7.80 mm thick. Miscellaneous Lithic Items A final lithic item consisted of a broken quartz cobble of the appropriate size to function as a hammer stone (Figure 126:K). However, battering wear patterns or any other type of wear was not observed on its surfaces. One end was broken off and it weighed 147.81 gm. Historic Artifacts Eighteen historic artifacts were recovered during the investigation. Most of these appear to have been deposited during military operations at Camp MacKall. The inventory consists of one sherd of amber bottle glass, three pieces of barb wire, two brick fragments, a clear mold made bottle glass fragment, seven green bottle glass sherds, a key wind can lid opener, a piece of window glass, a piece of wire and a penny with a date of 1950 which was minted in Philadelphia. All of this material was recovered between 0 and 20 cm bs. OCCUPATION PATTERNS Diagnostic markers of culture-chronological association were not recovered during the Phase I or Phase II investigations at 31RH478. The predominance of lithic artifacts and the absence of precontact ceramics indicate that the most if not all of the occupation is associated with Archaic phases. Precontact artifact density across the site was extremely low. The Stage I shovel test sample produced only three pieces of debitage, a projectile point tip, two ferrous quartzite rocks and 12 historic artifacts. Chipped stone debitage density in positive Stage I shovel tests was only 0.80 artifacts (SD=0.45). The mode was 1 and the range was 0 to 1 chipped stone artifact per positive shovel test. Based on positive and negative Stage I shovel tests excavated within the established boundaries of the site (n=21), the overall artifact density for 31RH478 is estimated to be 0.19 artifacts (SD=0.40). This is an extremely low figure, but further estimation reveals that the site contains a sizable artifact population nevertheless. Given a sampling fraction of .09 percent from the Stage I shovel test sample, it can be estimated that the site contains a mean of approximately 4,400 chipped stone artifacts, with a 1-standard deviation range of up to 9,300 artifacts. Due to the extremely low artifact densities at the site, the general approach taken to characterize the occupation history at the other sites in the package is only marginally applicable to 31RH478. The combined Stage I and II shovel test sample, which is biased toward locations identified as artifact bearing in the Stage I sample, contained only 43 chipped stone artifacts. However, the vertical patterns it produces provide a tentative model of the occupation history of the site (Table 112). MMR and UER are positioned between 10 and 50 cm bs. From what we have learned about the vertical distributions of culturalchronological occupations on sites in the Sandhills (see Cable 2010; Cable and Cantley 2005a, b, 2006), these profiles indicate ages ranging between Late and Early Archaic. The Type I RT profile is lower, concentrating between 40 and 70 cm bs. This vertical position is more narrowly associated with Early Archaic occupation. USR (1) occurs too infrequently to effectively evaluate its vertical position, but the single piece of debitage found in the Stage I and II sample of this raw material type is situated between 40 and 50 cm bs, which suggests an Early Archaic affiliation. A more detailed appreciation for precontact land-use patterns evidenced at the site can be derived from a consideration of occupation type distributions and associations. Currently, four occupation types based on chipped stone criteria have been recognized on sites in the Fort Bragg area (see Cable and Cantley 2005c, 2006): (1) Type I occupations consist of a debitage concentration and an associated tool cluster situated at a confined location along its periphery. It is inferred that the tool cluster was formed around a hearth and in several instances calcined bone fragments have been recovered from the deposit delimited by a Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 381 Chapter 8. 31RH478 Table 112. Vertical Distribution of Chipped Stone Artifacts, Stage I and II Shovel Tests, 31RH478. VERTICAL DISTRIBUTION OF CHIPPED STONE ARTIFACTS IN THE COMBINED STAGE I AND II SHOVEL TEST SAMPLE: 31RH478 RAW MATERIAL TYPE LEVEL 2 Mill Mountain Rhyolite 3 4 5 4 1 3 4 11 11 Type I Rhyolite Tuff Uwharries Eastern Rhyolite 1 3 Uwharries Southern Rhyolite (1) GRAND TOTAL 6 7 1 1 8 12 16 9 1 3 1 30 1 3 GRAND TOTAL 1 1 3 43 tool cluster. It is also believed that ephemeral shelters were situated adjacent to these tool clusters and opposite the concentrations. Debitage concentrations appear to represent lithic reduction loci established away from a shelter or sleeping area where tools were manufactured during the stay at the camp. These concentrations are generally composed of single lithic raw material types, and the associated tool clusters commonly consist of manufacturing rejects and discarded broken or worn-out tools made of the same raw material that comprises the associated debitage concentration. Although many Type I residences occur in isolation and appear to be the by-products from single nuclear or small extended family occupations, others appear to represent multi-family occupations that extend beyond the ability of a single shovel test to identify. Woodland period Type I occupations are characterized by an additional element, partial vessel sherd aggregations. These occur adjacent to debitage concentrations and they are hypothesized to occupy a portion of the living floor associated with a hut. The model would stipulate that tool clusters are situated between the debitage concentrations and sherd aggregations. (2) Type II residences are similar in spatial organization to Type I residences, but they exhibit much higher tool-to-debitage ratios and may commonly be comprised of diverse lithic raw materials. Lithic reduction on these sites consisted of tool maintenance and late stage shaping of tools from flake blanks or reduced biface cores. Consequently, debitage density is much lower than that of Type I debitage concentrations and the scatters consist primarily of late stage reduction debitage. Type II occupations are inferred to be the output from single or small, multiple household aggregations and are thought to represent the residue from High Technology Forager (HTF) residences (Spiess 1984; Todd 1983:231-233). The HTF model proposes a specialized forager adaptation combining high logistical and high residential mobility into a single settlement system characterized by highly curated technologies. Type II residences have thus far only been assigned to the Early Archaic period at Fort Bragg. The lithic raw material diversity characteristic of Type II residences suggests that these occupations represent the aggregation of social units that may have been seasonally dispersed, transporting a wide range of lithic raw materials from different sources. (3) Type III occupations are inferred to represent logistical camps and very short-term residences characterized by low-density debitage scatters and sporadic tool discard. (4) Type IV occupations represent extraction or processing loci that are not believed to have involved, necessarily, overnight stays. They are difficult, however, to distinguish from Type III occupations, even in situations of greater excavation exposure. Isolated sherd aggregations are hypothesized to represent one kind of Type IV element associated with Woodland occupations. It is difficult to distinguish effectively between these various occupation types at shovel test intervals of greater than 1.25 m, but a less pre- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 382 Chapter 8. 31RH478 cise classification can be used at greater shovel test intervals to aid in population projections. High-density debitage concentrations tend to indicate Type I residences, while low-density scatters tend to signal short-term residences, special purpose camps and locations (i.e. Type II, III and IV occupations). Highdensity debitage concentrations are temporarily stipulated to correspond to shovel test outcomes of five or more pieces of debitage of a specific raw material subtype. Eventually, it will be necessary to confirm this arbitrary threshold by examining shovel test counts from a large sample of known high-density debitage concentrations. Deposits containing less than five pieces of debitage are viewed as more likely to represent low-density debitage scatters associated with the other three site types. Since shovel tests can intersect the peripheral zones of high-density debitage concentrations where debitage densities are much lower, however, it is likely that Type I residential occupations will be under-represented in macrointerval shovel test data. Because it is difficult to determine whether a single occupation is represented in more than one shovel test, in the cases of multi-family residences or associated special activity zones on the periphery of a camp, the term “element” is preferred over the term “occupation” to refer to the occupational data in shovel tests. At a sampling interval of 5 m- or 10 m-intervals, we are assured that the presence of an element will not be repeated in adjacent shovel tests. Elements are identified by lithic raw material subtypes and/or by precontact sherd aggregations associated with one or more vessels. This methodology runs the risk of overestimating occupations because more than one raw material subtype may comprise a single occupation. The “element” concept recognizes this fact, focusing on depositional events and episodes rather than entire occupations. Elements can represent segments of contemporaneous multifamily occupations, separate occupations of various function, or contemporary members of multiple raw material reduction episodes. The current methodology for identifying elements in shovel tests consists of running crosstabs of lithic debitage raw material subtypes or subtype groups by shovel test number in a spreadsheet. Sherd aggregations are not used in these calculations because, in general, they are associated with debitage scatters and their inclusion would result in potentially double counting a good number of elements. Once the cross-tab is generated, the sum of each lithic raw material subtype in a shovel test is identified as an element. In some instances, the vertical separation between individual items of the same lithic subtype can be so great as to suggest the presence of distinct elements. Due to the vagaries of bioturbation and possible recovery errors in which higher positioned material can be included in lower levels through sidewall displacement, however, the lower item(s) is assigned to the higher positioned item(s) to define a single occupation element. In cases where large debitage concentrations of the same lithic raw material type are found at distinctively separated levels, though, each is counted as separate elements. Elements have horizontal dimensions and, as a result, sample units that are smaller than those dimensions have the benefit of increased sample efficiency (see Cable and Donaldson 1988; Rice 1987; Rice and Plog 1983). The typical element found at Fort Bragg has a diameter of about 3 m. Thus, a shovel test of .09 m2 samples only .09 percent of the area of a 10 m-square, but it samples 9.0 percent of the theoretical mean element space of 9.0 m2. Consequently, a 10 m-interval shovel test pattern actually samples about 9.0 percent of the available element space at a site, while a 5 m-interval shovel test pattern samples about 36.0 percent of the available element space. A mean estimate of the population of elements at shovel test intervals of 10 m and 5 m, then, requires a simple arithmetic calculation of counting the occupation elements identified in shovel tests and dividing this figure respectively by 9.0 percent and 36.0 percent. Confidence intervals can be constructed around these means, but they have not been calculated in the Fort Bragg testing projects. Unfortunately, the only unbiased and fullcoverage sample that can be worked with at 31RH478 is the Stage I sample. In low density situations like this, the chances for sampling error increase significantly, especially when sample unit density is low like it is with 10 m-interval shovel tests. Four elements were identified in the Stage I shovel test sample, two each of Type I RT and UER (Table 113). All were represented by a single artifact and hence were assigned to the low-density element grouping. No Type I elements were identified in the sample, but subsequent Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 383 Chapter 8. 31RH478 Table 113. Vertical Distribution of Element Types, 31RH478. VERTICAL DISTRIBUTION OF ELEMENT TYPES BY RAW MATERIAL TYPE: 31RH478 ELEMENT TYPE RAW MATERIAL TYPE TYPE I TYPE II/III/IV LEVEL OF OCCURRENCE 4 5 Type I Rhyolite Tuff 2 Uwharries Eastern Rhyolite 2 1 1 4 1 2 GRAND TOTAL 0 investigation documented the presence of at least two of these elements at the site. The Type I RT elements are located in the center of the site at grid coordinates N520/E500 and N530/E510, while the UER elements were situated on the northern and southern ends of the site at grid coordinates N540/E480 and N490/E510 (Figure 122). Following the procedure for projecting element populations employed at the other sites, a population mean of 44 elements is produced. SAMPLE BLOCK INVESTIGATIONS Only two sample blocks (Figure 124) were established at 31RH478. However, both accomplished intensive coverage within areas that were much more expansive than the commonly configured 6.25 m2 area encompassed by a nine shovel test set. Sample Block 1 was centered at N515/E500 and covered a rectangular shaped area measuring 7.5 x 6.25 m. Forty-five shovel tests were excavated in this sample block, generally at intervals of 1.25 m. Sample Block 2 was centered at N488/E508 and covered a rectangular shaped area measuring 11.25 x 8.75 m. Seventy-two shovel tests were excavated in this sample block, spaced again at intervals of 1.25 m. The results of shovel testing in each sample block are presented below. Sample Block 1 (SB 1) Debitage recovered from SB1 consisted almost entirely of UER and MMR lithic raw mate- 6 1 7 1 0 1 rial types (Table 114). The UER debitage profile may occupy a slightly higher vertical position than the one for MMR, but the former also has a more dispersed vertical distribution and significant quantities of it occur in the same position as the MMR sample. Sample sizes are too low to make a definitive statement about their vertical relationships, however, beyond the observation that both raw material types are primarily positioned between 20 and 50 cm bs. Also included in the sample was a flake of Type I RT found between 60 and 70 cm bs and two flakes of USR(1) recovered between 20 and 50 cm bs. Horizontal mapping of the raw material type distributions indicates that the MMR scatter is concentrated on the eastern side of the sample block (Figure 127). The low-density debitage scatter has maximum dimensions of 3.75 x 5.00 m, with the primary concentration occurring in an area measuring about 3.00 x 3.50 m. The UER distribution (Figure 128) identifies two separate concentrations within a larger extent of about 6 m in diameter. All three of the concentrations exhibit shovel test outcome characteristics consistent with Type III elements (see Table 115). Three measures related to these outcomes have been developed to describe element variability. These are: (1) number of positive shovel tests within which the element occurs, (2) number of shovel tests with greater than or equal to 5 pieces of debitage and (3) mean debitage frequency per positive test. Each of these clusters are characterized by low individual shovel test yields of debitage, almost all of which produced less than 5 pieces of debitage (Table 116). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 384 Chapter 8. 31RH478 Table 114. Vertical Distribution of Chipped Stone Artifacts in SB 1, 31RH478. VERTICAL DISTRIBUTION OF DEBITAGE AND STONE TOOLS BY RAW MATERIAL TYPE IN SB1: 31HK478 DEBITAGE RAW MATERIAL TYPE STONE TOOLS LEVEL 1 2 Mill Mountain Rhyolite 3 4 5 5 1 3 6 Type I Rhyolite Tuff Uwharries Eastern Rhyolite 1 6 Uwharries Southern Rhyolite (1) GRAND TOTAL 12 6 1 1 6 5 7 GRAND TOTAL 1 10 1 1 1 31 1 18 7 LEVEL 3 4 Core Frag. Core Frag. 2 9 3 44 31RH478, SB1: MMR Debitage Density Distribution (CI=1 Item) 522.5 0 N Scale 2.5 m 520.0 517.5 515.0 512.5 Core Fragments (2) 510.0 Positive ST Negative ST 495.0 497.5 500.0 502.5 505.0 507.5 Figure 127. Density Distribution of MMR Debitage, SB1, Site 31RH478. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 385 Chapter 8. 31RH478 31RH478, SB1: UER Debitage Density Distribution (CI=1 Item) 522.5 0 N Scale 2.5 m 520.0 517.5 515.0 512.5 510.0 Positive ST Negative ST 495.0 497.5 500.0 502.5 505.0 507.5 Figure 128. Density Distribution of UER Debitage, SB1, Site 31RH478. Table 115. Shovel Test Outcome Model for the Purpose of Identifying Element Types. INFERRED ELEMENT TYPE NO. POSITIVE STPS NO. STPS WITH ≥ 5 DEBITAGE MEAN DEBITAGE/ POSITIVE TEST TYPE I HIGH HIGH HIGH TYPE I OR TYPE III HIGH MODERATE MODERATE TYPE III MODERATE LOW LOW Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 386 Chapter 8. 31RH478 The low density of occupations across the site, suggests that these clusters may in fact represent elements within a single occupation. There is clear evidence of a void between the two UER clusters and the MMR cluster is offset to the east in such a way as to complete a hypothetical circle of elements around a central void. Concentric arrangements of activity spaces are common features of camp structure when multiple social units aggregate (see Binford 1983; Fischer and Strickland 1991; Yellen 1977). The MMR scatter is situated at a slightly lower vertical position that the UER clusters, which could disqualify its membership, but there is no difference between the UER clusters. The MMR scatter has an average depth of 37.70 cm bs, while the UER clusters are positioned at average depths of 31.00 and 32.50 cm bs. The possible contemporaneity of these elements, however, would only move closer to resolution through the recovery of matching diagnostics. All of the debitage scatters are composed primarily of late stage biface reduction debitage and none of it displays cortex. Average debitage weight is low. For the UER sample it was 0.29 gm (n=32, SD=0.30) and for MMR it was 0.71 gm (n=10, SD=1.56). The latter average was skewed by a single large FBR that weighed 5.14 gm. Without this outlier, the average weight for the MMR debitage (0.22 gm) is comparable to that of the UER sample. These reduction profiles are consistent with late stage reduction and maintenance activities, which is commonly the case for Type III elements. Sample Block 2 (SB 2) Debitage recovered from SB2 consisted almost entirely of UER (Table 117). Its vertical profile was centered between 30 and 50 cm bs, in what is normally considered the Middle and Early Archaic zone of upland Sandhills sites in the region (see Cable 2010; Cable and Cantley 2005a, b, 2006). Single flakes of USR (1) and MMR were recovered in this same zone. Only a small core fragment of UER accompanied the UER debitage in the shovel test sample. Horizontal mapping of the UER debitage resulted in the identification of three separate clusters resting in a larger area measuring approximately 7.5 x 9.0 m (Figure 129). The UER distribution may extend to the east of the mapped area, but the additional area does not appear to be significant judging by the distribution of negative shovel tests in the larger site grid. The clusters are relatively limited in extent, measuring 2.5 x 3.5 m (NW), 2 x 3 m (SW) and 2.5 x 2.5 m (NE). Again, the concentric arrangement of the clusters suggests that they could represent Table 116. Inferred Element Types in SB 1, 31RH478. 2 0 2.50 III Southern UER Cluster 13 1 2.08 III 5 0 2.00 III MMR Cluster INFERRED ELEMENT TYPE MEAN DEBITAGE/ + SHOVEL TEST Northern UER Cluster ELEMENTS NO. OF STs REPRESENTED NO. STs WITH ≥ 5 DEBITAGE INFERRED ELEMENT TYPES IDENTIFIED IN SB1: 31RH478 STONE TOOLS Core Fragments (n=2) Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 387 Chapter 8. 31RH478 Table 117. Vertical Distribution of Chipped Stone Artifacts in SB 2, 31RH478. VERTICAL DISTRIBUTION OF DEBITAGE AND STONE TOOLS BY RAW MATERIAL TYPE IN SB2: 31HK478 DEBITAGE RAW MATERIAL TYPE STONE TOOLS LEVEL 1 2 3 4 Mill Mountain Rhyolite 5 6 7 1 Uwharries Eastern Rhyolite 3 44 Uwharries Southern Rhyolite (1) 39 3 44 41 4 1 14 3 103 1 GRAND TOTAL LEVEL GRAND TOTAL Core Frag. 1 14 3 105 31RH478, SB2: UER Debitage Density Distribution (CI=1 Item) 495.0 NE Element Quartz Cobble FCR (n=2) 492.5 NW Element 490.0 Core Fragment 487.5 SW Element 485.0 N 482.5 Positive ST 0 Scale 2.5 m Negative ST 480.0 500.0 502.5 505.0 507.5 510.0 512.5 515.0 Figure 129. Density Distribution of UER Debitage, SB2, Site 31RH478. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 388 Chapter 8. 31RH478 a contemporaneous, multiple social unit camp (see Binford 1983; Fischer and Strickland 1991; Yellen 1977). Vertical debitage profiles for the three clusters are nearly identical, adding further support to the proposed contemporaneity (Table 118). Shovel test outcome data indicate that the elements appear to exhibit functional differences (Table 119). The NE and SW elements display outcome results typical of Type III elements, but the NW element is clearly within the range of a Type I element. Further variation can be seen in the debitage reduction profiles (Table 120). The SW cluster exhibits a profile consistent with a Type III element. Bifacial reduction types dominate, cortex is low and average flake weight is low, all characteristics that imply only maintenance and late stage reduction. By contrast, the NW and NE clusters display profiles in line with early stage reduction and tool manufacture. Core flakes dominate the identifiable debitage, the incidence of cortex is high and within the range of primary nodule reduction and mean flake weight is high. Table 118. Vertical Distribution of Debitage for the UER Elements, SB 2, 31RH478. VERTICAL DISTRIBUTION OF DEBITAGE FOR THE UER ELEMENTS IN SB2: 31RH478 LEVEL ELEMENT 3 4 NE 5 6 7 10 1 GRAND TOTAL 7 18 NW 1 29 19 6 2 57 SW 2 9 11 7 1 30 GRAND TOTAL 3 45 41 14 3 106 Table 119. Inferred Element Types in SB 2, 31RH478 NO. STs WITH ≥ 5 DEBITAGE MEAN DEBITAGE/ + SHOVEL TEST NE 6 1 3.00 III NW 12 3 4.75 I SW 12 1 2.50 III UER ELEMENTS INFERRED ELEMENT TYPE NO. OF STs REPRESENTED INFERRED ELEMENT TYPES IDENTIFIED IN SB 2: 31RH478 STONE TOOLS Core Fragment Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 389 Chapter 8. 31RH478 Table120. Debitage Reduction Profiles for Elements in SB 2, 31RH478. DEBITAGE REDUCTION PROFILES FOR ELEMENTS IN SB2: 31RH478 IDENTIFIABLE DEBITAGE (n) % CORE FLAKES NE 12 83.3 16.7 1.93 (2.60) NW 32 87.5 7.0 1.19 (1.84) SW 15 40.0 0.0 0.48 (0.71) UER ELEMENTS % CORTEX MEAN WEIGHT (SD) Variation of this type should be expected in multiple social unit camps because of the potential for the occupying group to vary by age and sex. This would be particularly true for extended family units where it is common ethnographically for different locations to be occupied by primary nuclear families and attached elderly households or specially comprised demographic cohorts of unmarried men, women or young adults (see Binford 1983; Hayden 1979; Yellen 1977). In this scenario, the Type I element might represent the primary nuclear family residence. This is because mature adults and children are generally members of larger households, while young adults and the elderly tend to be members of small households in hunter-gatherer societies (Howell 1979:325). In this regard it would be reasonable to expect that the greatest amount of debris accumulation would occur in stem nuclear family households. Further attention to such variation may one day provide a foundation for conducting anthropological studies of the demographic structure of archaeological campsites. TEST UNITS Test unit investigation was focused on further elaborating the patterns identified in SB2. Distributional patterns suggested that this area might contain the remains of a multiple household camp. One of the primary goals was to locate and sample potential hearth areas within the three clusters to derive a better picture of the activities that took place and to recover diagnostic artifacts that would simultaneously allow for assigning cultural chronological associations to the deposits and to test for their contemporaneity. Four 1-x-1-m test units were excavated within the confines of SB2 (Figure 130, Table 121). Initially, TU1 sampled the NW element, TU2 was placed in the NE element and TU3 was excavated in the SW element. Subsequently, TU4 was placed contiguous and to the east of TU1 in the NW element. Finally, four 0.5-x-0.5 m test units were excavated contiguous to the northern and northeastern side of TU4 in an attempt to search further for evidence of a hearth area. Because the test unit profiles are so similar due to close proximity and low levels of subsurface disturbance, they will be described together (Figures 131-133). Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corners of the units, all of which were excavated to a depth of 70 cm bd. The 0.5-x-05-m test units were terminated at 60 cm bd. Three soil zones were consistently present, conforming closely to the Wakulla series profile description (Evans 1999). The upper zone was 10 to 15 cm thick and appears to represent a plow zone (Ap). It was composed of dark gray (10YR 4/1) to light brownish gray (10YR6/2) sand. The darker coloration was due to leaching of prescribed burn charcoal. The E-Horizon extended from the base of the plow zone to between 40 and 50 cm bd. It consisted of brownish yellow (10YR 6/6) sand. Below the E-Horizon and extending beyond the base of the test units was a strong brown (7.5YR 5/6) Bt-Horizon, which was composed of argillic loamy coarse sand. As is typical of Wakulla series soils, the silt/clay content of the BtHorizon was low. Excavations conducted in each of the three clusters recovered UER debitage in the same vertical Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 390 Chapter 8. 31RH478 31RH478, SB2: UER Debitage Density Distribution (CI=1 Item) 495.0 Flake Blank Quartz Cobble FCR (n=2) Type II Biface frag. Utilized Flake Large Quartz Cobble 492.5 NW Element TU1 1 2 3 TU4 4 NE Element TU2 490.0 PPK Tip 487.5 Core Fragment Quartz Cobble FCR (n=7) SW Element TU3 485.0 N 482.5 Positive ST 0 Scale 2.5 m Negative ST 480.0 500.0 502.5 505.0 507.5 510.0 512.5 515.0 Figure 130. Locations of Test Units in SB2, Site 31RH478. position as was reported for shovel tests in SB2 (Table 122). The central tendency of the vertical profiles in the test units placed in the NE and SW elements is slightly higher than that of the NW element, but this is most likely a consequence of subtle variation in the elevation of the occupation surface rather than an indication of chronological differences. MMR debitage, which was not found in shovel tests, was present in low-densities in the units excavated in the NW and NE elements. A projectile point tip of MMR was recovered from TU1 (Figure 126:I) as well. Whether this material is associated with the UER clusters could not be determined, although it was found in the same vertical position, between 20 and 50 cm bd. Only three stone tools of UER were recovered, a utilized flake (Figure 126:J) and a small lateral section from a Type II Biface found respectively Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 391 Chapter 8. 31RH478 Table 121. Summary Data for Test Units, 31RH478. SUMMARY OF TEST UNITS: 31HK2521 SIZE NORTH EAST NO. LEVELS MAX. DEPTH (cm bd) VOLUME 3 (m ) ARTIFACT DEPTH (cm bd) TU1 1x1m 491.75 505.85 7 70 0.70 70 TU2 1x1m 491.25 511.25 7 70 0.70 70 TU3 1x1m 487.50 506.50 7 70 0.70 50 TU4 1x1m 491.75 506.85 7 70 0.70 60 EX1 1x1m 492.25 506.35 6 60 0.15 50 EX2 1x1m 492.25 506.85 6 60 0.15 60 EX3 1x1m 492.25 507.35 6 60 0.15 60 EX4 1x1m 491.75 507.35 6 60 0.15 60 TEST UNIT between 30 and 40 cm bd and 40-50 cm bd in the NW quadrant of TU2 (NE element) and a flake blank (Figure 126:G) positioned between 30 and 40 cm bd in Extension-4 off of TU4 in the NW element (Figure 130). These locations probably represent the edges of tool clusters, both of which are situated on the outside perimeter of the concentric ring forming the hypothetical multiple household camp. If these locations indeed represent hearth areas, it would then be likely that sleeping surfaces (i.e. huts or open areas) ringed the exterior of the larger debitage scatter in areas relatively free of debris. In this model, the debitage concentrations would have been situated to the sides of the sleeping surfaces. Additional evidence in support of this hypothetical layout was recovered. A large quartz cobble (Figure 126:K), perhaps a hammer stone, was recovered from Extension-2 of TU4, in the same general area as the flake blank in the NW element. It too was situated on the outside perimeter of the debitage concentration (Figure 130). In addition, a cluster of seven quartz cobble fire-cracked rock fragments (149.56 gm) was found in the NW quadrant of TU3 in the SW element. This location would have been situated on the exterior of the debitage concentration too and might evidence the former location of a hearth. Finally, two small quartz cobble fire-cracked rock fragments were also found in a shovel test on the northern perimeter of SB2 in the hypothesized location of the tool cluster on the north edge of the NW element. Although meager, this evidence is consistent with a multiple household camp in which the sleeping surfaces are flanked by debitage concentrations and ring a central communal area (Figure 134). The hearth and sleeping area for the NW element are depicted as larger to reflect its hypothesized stem nuclear family role in what might represent an extended family with dependent smaller households composed of narrow cohort groups. Debitage profiles generated for the elements from the shovel test sample indicated variation in the activity structure of the elements. The NW and NE element were characterized by profiles indicative of tool manufacture and primary reduction of nodules to directional cores, while the SW element displayed a profile in keeping with tool maintenance and only secondary reduction from directional cores or flake blanks. The results of the test unit excavations agree with this general reconstruction (Table 123). High percentages of core flakes and high incidences of cortex characterize the NE and NW elements, while much lower percentages of these two measures are found in the SW element. Moreover, mean weight in grams per piece of debitage is much higher in the NW and NE elements. Interestingly, the MMR profiles from the NW and NE elements display nearly identical profiles to the UER debitage in these elements, which might indicate that this material was also deposited during the time the multiple household camp was occupied. Reduction profile variation may not relate to anything more than random differences Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 392 Chapter 8. 31RH478 EAST WALL Root cm bd 0 Ap Lev 1 10 Lev 2 20 Lev 3 E 30 Lev 4 40 Lev 5 50 Lev 6 60 Bt Lev 7 70 TU1 N491.75/E504.85 TU4 N491.75/E505.85 0 N491.75/E506.85 2m Soil Strata Ap: Dark Gray (10YR 4/1) Sand E: Brownish Yellow (10YR 6/6) Sand Bt: Strong Brown (7.5 YR 5/6) Loamy Sand/Sand Figure 131. Profile Drawing, TU1 and TU4, West Wall, Site 31RH478. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 393 Chapter 8. 31RH478 NORTH WALL cm bd 0 Ap Tree Roots Lev 1 10 Lev 2 20 Lev 3 30 E Lev 4 40 Lev 5 50 Lev 6 60 Bt Lev 7 70 TU2 N491.25/E510.25 0 N491.25/E511.25 1m Soil Strata Ap: Dark Gray (10YR 4/1) Sand E: Brownish Yellow (10YR 6/6) Sand Bt: Strong Brown (7.5 YR 5/6) Loamy Sand/Sand Figure 132. Profile Drawing, TU2, North Wall, Site 31RH478. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 394 Chapter 8. 31RH478 SOUTH WALL cm bd 0 Ap Lev 1 10 Lev 2 20 Lev 3 30 E Lev 4 40 Lev 5 50 Lev 6 60 Bt Lev 7 70 TU3 N486.5/E506.5 0 N486.5/E505.5 1m Soil Strata Ap: Dark Gray (10YR 4/1) Sand E: Brownish Yellow (10YR 6/6) Sand Bt: Strong Brown (7.5 YR 5/6) Loamy Sand/Sand Figure 133. Profile Drawing, TU3, South Wall, Site 31RH478. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 395 Chapter 8. 31RH478 Table 122. Vertical Distribution of Chipped Stone Artifacts in Test Units, 31RH478. VERTICAL DISTRIBUTION OF DEBITAGE AND STONE TOOLS FROM TEST UNITS: 31RH478 NW ELEMENT TU1 RAW MATERIAL TYPE DEBITAGE 1 Mill Mountain Rhyolite Uwharries Eastern Rhyolite White Quartz GRAND TOTAL 2 1 4 3 16 56 LEVEL 4 5 78 5 72 83 Mill Mountain Rhyolite Uwharries Eastern Rhyolite White Quartz GRAND TOTAL 7 24 168 LEVEL 4 PPK Tip 5 192 DEBITAGE 1 1 1 2 3 1 21 1 21 LEVEL 4 4 94 1 99 EXTENSION UNITS RAW MATERIAL TYPE 32 6 FCR TU4 RAW MATERIAL TYPE 5 2 30 STONE TOOLS GRAND TOTAL 5 9 82 1 92 STONE TOOLS 6 1 4 7 GRAND TOTAL 2 5 Mill Mountain Rhyolite Uwharries Eastern Rhyolite White Quartz GRAND TOTAL STONE TOOLS 3 2 5 LEVEL 4 1 18 11 3 3 37 7 19 11 3 40 5 5 15 202 2 219 DEBITAGE 1 LEVEL 4 6 7 GRAND TOTAL LEVEL 4 5 Flake Blank Cobble NE ELEMENT TU2 RAW MATERIAL TYPE DEBITAGE 1 2 Mill Mountain Rhyolite Uwharries Eastern Rhyolite GRAND TOTAL 3 4 17 21 LEVEL 4 8 26 34 5 5 18 23 STONE TOOLS 6 1 1 7 1 1 GRAND TOTAL 18 62 80 LEVEL 4 5 Utilized Flake SW ELEMENT TU3 RAW MATERIAL TYPE Uwharries Eastern Rhyolite White Quartz GRAND TOTAL DEBITAGE 1 1 2 4 3 20 LEVEL 4 22 1 4 20 22 5 8 8 STONE TOOLS 6 7 GRAND TOTAL LEVEL 4 5 55 FCR (n=7) 55 in lithic raw material provisioning, but in situations where a multiple household occupation is suspected, these differences may inform directly on the demographic composition of the various households. A model has yet to be developed that links lithic reduction strategies to variation in social unit composition. However, various scenarios can be imagined. For instance, the NW element, because of the large mass of debitage associated with it, has been posited to represent a nuclear stem family. Mature males in the prime of their hunting prowess head nuclear stem families and replenishing hunting tool kits might require major tool manufacturing activities, thus resulting in the kind debitage profile and debitage density characterizing the NW element. The NE element is distinguished by a much lower mass accumulation of debitage, but with a similar tool manufacturing activity profile. This is the kind of pattern that might be expected to be manifest in minority social units containing a mature or young adult hunter without a fully expressed nuclear family. For instance, a young adult son belonging to the stem family, an attached mature male sibling or fictive sibling of the head of the stem family, or a young adult couple without children. A social unit without a hunter, such as Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 396 Chapter 8. 31RH478 Hypothetical Multiple Household Model for SB2, 31HK478 Sleeping Surface Sleeping Surface Hearth NW Element NE Element Hearth Debitage Concentration Debitage Concentration Communal Area Hearth Debitage Concentration Sleeping Surface 0 N Scale 2.5 m SW Element Figure 134. Hypothetical Model of Multi-Household Camp, SB2, Site 31RH478. Table 123. Debitage Reduction Profiles for Elements in SB 2 and Associated Test Units, 31RH478. ELEMENT RAW MATERIAL TYPE COUNT IDENTIFIABLE DEBITAGE TYPES PERCENT CORE FLAKES TOTAL DEBITAGE PERCENT DEBITAGE WITH CORTEX DEBITAGE REDUCTION PROFILES BY RAW MATERIAL TYPE FOR SB2 ELEMENTS: 31RH478 MEAN DEBITAGE WEIGHT (SD) NE UER 27 81.4 61 9.8 1.06 (1.53) NE MMR 10 70.0 18 5.6 2.60 (3.93) NW UER 197 71.1 407 7.4 0.97 (1.52) NW MMR 20 85.0 42 4.8 0.92 (1.41) SW UER 25 24.0 55 1.8 0.63 (0.80) Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 397 Chapter 8. 31RH478 an elderly couple or unmarried women, may have occupied the SW element since tool kit replenishing was not undertaken there. Obviously we have moved past our ability to test propositions such as this, but these are the kinds of considerations that will ultimately deliver a much better understanding of the dynamics of camp structure than we hold today. Complete recovery of deposits such as this will enhance our ability to entertain detailed hypotheses of social unit composition and function. The recovery of associated tool kits will, moreover, greatly enhance our ability to begin to discuss patterns that might be associated with differences in gender, age and social unit size. EVALUATION Site 31RH478 is recommended not eligible for inclusion on the National Register of Historic Places. Although it contains discrete deposits that have retained a high degree of structural and organizational integrity, similar deposits occur on a large number of sites in the Sandhills. The occupation history of the site was not well defined from the Phase II investigation. It is generally established, however, that the primary deposits are Archaic in age and appear to reside in an in tact Early to Middle Holocene surface. Subsequent investigations at nearby 31RH480 and 31RH491 demonstrate that this surface is widespread across sites in the Camp MacKall segment of the Drowning Creek valley. Moreover, much larger samples of Middle and Early Archaic occupations occur at those sites, offering much more comprehensive representations of hunter-gatherer settlement variability and chronological time depth. Consequently, the opportunity to address a much wider range of research questions is presented by the archaeological deposits from these sites. One of these sites, 31RH491, is recommended eligible for the NRHP in this report and the Early to Middle Holocene deposits at 31RH478 would not significantly enhance the information contained at this site beyond what has been learned form the Phase II investigation. Consequently, the data contained at 31RH478 is viewed as redundant and no further work is recommended at this site. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 398 Chapter 9. 31RH480 Site 31RH480 is located in Richmond County within the confines of Camp MacKall and just to the east of the airfield (Figure 135). The site is situated on a ridge finger overlooking a beaver pond that has backed-up in an unnamed tributary that feeds into Moss Gill Lake (Figure 136). The site can be reached by taking any number of unimproved trails and roads intersecting Glider Road just east of the airfield and traveling north about 0.6 km or about 0.35 miles. The SCS soil survey for Richmond County, NC (Evans 1999) classifies the landform as Ailey loamy sand, a well-drained soil typically found in the uplands of the Sandhills on both ridge tops and slopes. The site rests on the edge of a narrow ridge finger leading down to the channel of an unnamed tributary that now supports a beaver pond (Figure 137). The tributary eventually empties into Moss Gill Lake located just south of Glider Road. The confluence of the tributary and Big Muddy Creek is approximately 1.1 km south of the site. A trial follows the central spine of the ridge and dead-ends at the base of the ridge toe (Figure 138). The surface of the site is dotted with historic disturbances and small pits, most of which are attributable to military training exercises. Elevations within the site boundaries ranged between about 96.75 m on the north along the mid-slope of the ridge apex to about 93.25 m amsl near the margins of the tributary. Slope from north to south within the site limits is about 4 percent. Vegetation across the site consisted of a thinned stand of young long-leaf pines and an understory of turkey oak and wiregrass. Maximum site dimensions were established at 80 m x 150 m, covering an area of approximately 1.98 acres (8,000 m2). PREVIOUS RESEARCH Site 31RH480 was originally recorded by TRC Garrow Associates, Inc in the spring of 2001 during a Phase I survey of portions of Fort Bragg and the KK1 Tract at Camp MacKall (Ruggiero 2003:442-445). The site boundaries established at that time measured 60 x 120 m (Figure 139). Phase I site definition proceeded through the deployment of 15 m-interval shovel tests around a positive survey transect shovel test. Including site discovery transect shovel tests, 49 shovel tests were excavated in total, 10 of which were positive. Ruggiero (2003:442) described three soil zones exposed in shovel tests. The upper zone was 0 to 15 cm thick and consisted of very dark grayish brown (10YR 3/2) sand, while the next lower zone was composed of very pale brown (10YR 7/3) sand that extended to depths of at least 75 cm bs. A third zone of yellowish brown (10YR 5/8) or brownish yellow (10YR 6/8) sand was occasionally encountered during shovel testing and would appear to correspond to the Bt-Horizon of the Ailey sandy loam profile (Evans 1999). Twenty-five artifacts were recovered in total, including 16 pieces of rhyolite debitage and nine precontact sherds. These were found in the Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 399 Chapter 9. 31RH480 640000 638000 639000 640000 3880000 3879000 3879000 3880000 3881000 639000 3881000 638000 3878000 3877000 3877000 3878000 31RH480 Figure 135. Location of Site 31RH480 (Pine Bluff, NC 7.5’ USGS Quadrangle) Scale: 1:24,000. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 400 Chapter 9. 31RH480 637800 638000 638200 638400 3878200 3878400 17 th 31RH480 AB N Beaver DI V Ro ad 3878200 3878400 3878600 637600 3878600 637400 3878000 er Glid d Roa 3877800 3877800 3878000 nd Po MacKall Airfield 3877400 3877200 3877200 3877400 ke La 3877600 ill G 3877600 ss Mo 637400 637600 637800 638000 638200 638400 Figure 136. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31RH480 (Scale: 1 inch = 200m, Projection: UTM, NAD83). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 401 Chapter 9. 31RH480 Figure 137. View of Site 31RH480, Looking Southeast from the Central Trail at N510/E490 Toward the Beaver Pond. second soil zone, the E-Horizon, at depths of 10 to 80 cm bs. Seven of the sherds were found in a single shovel test and belonged to a single fabric impressed vessel tempered with crushed quartz. The other two sherds were sand/grog-tempered and exhibited indeterminate exterior surface treatments. Three separate artifact concentrations were identified within the site boundaries. Ruggiero (2003:445) concluded that the bulk of the occupation was associated with the Woodland period. Based on the apparent horizontal separation of Woodland components and the integrity of the subsurface deposits, the site was recommended potentially eligible for inclusion on the National Register of Historic Places. Further work here was considered likely to yield important information concerning Woodland period site function and spatial organization and possibly chronology. FIELDWORK OVERVIEW Phase II fieldwork was implemented in four stages, each stage building on the information generated from the previous stage. Stage I investigation consisted of a program of 10 m-interval shovel tests that were excavated to establish firm site boundaries. Site boundaries were determined by the documentation of two consecutive negative shovel tests in all directions on a 10-m grid, or in some cases when wetland was encountered. Once artifact distributions from the Stage I program were mapped and evaluated, the Stage II investigation involved the excavation of closer-interval shovel tests at 5 m-intervals in most of the artifact bearing areas of the site. Stage III investigations involved the deployment of closeinterval shovel tests of 1.25 m around the perimeter of targeted shovel tests yielding deposits deemed of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 402 Chapter 9. 31RH480 97.0 560 96.5 550 Beaver Pond 96.0 540 il Tra 530 95.5 520 510 95.0 Dirt R 500 oad 490 94.5 480 31RH480 470 94.0 460 450 93.5 440 430 420 410 N 400 0 40 m 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 138. Site 31RH480, Base Map. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 403 Chapter 9. 31RH480 E440 E410 E470 E500 E560 E530 E590 N590 N560 T22 T21 N530 Beaver Pond N500 T20 + T19 N470 N440 T18 d Roa N410 Positive STP Positive transect STP Negative STP Negative transect STP Site boundary + GPS Contour (approximate) * Surface find (if any) Tests shown are not to scale 0 Meters 45 Figure 208. Site Plan of 31RH480. Figure 139. Site 31RH480, Phase I Survey Sketch Map, TRC (from Ruggiero 2003). 443 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 404 Chapter 9. 31RH480 theoretical importance for evaluating the site. These close-interval shovel test locations were identified as sample blocks and generally consisted of eight shovel tests excavated around a targeted Stage I or Stage II shovel test. The sample blocks were also instrumental in guiding the precise placement of 1 x 1 m test units, which constituted the final stage of the investigation. Test units were placed in locations judged to provide optimal opportunities to reconstruct the occupation history of the site and to assess the integrity and scientific importance of the deposits. Stage I investigations were initiated by relocating the datum established by Ruggiero (2003) and laying out a 10-m interval grid with a total station. The grid coordinates (N500/E500) for the datum were maintained for the Phase II work. Relative topographic elevations at each 10 m-grid intersection were recorded with the total station and marked with a pin flag. The elevation values (in meters) illustrated on the base map (Figure 138) are approximated real elevations based on the correlation of GPS points with a geo-referenced copy of the USGS, Pine Bluff, NC 7.5-minute topographic quadrangle map. Cultural and landform features were also mapped and additional elevations were taken when appropriate to produce the topographic map. Shovel tests measured 30-x-30-cm and were square in shape. Excavation proceeded in 10 cm arbitrary levels measured from the surface. One hundred and two Stage I shovel tests were excavated at 10 m-intervals to define the limits of the site (Figure 140). Of these, 26 produced artifacts. Maximum site boundaries were established at 80 m x 150 m, covering an area of about 8,000 m2, or approximately 1.98 acres. Mean shovel test depth for the Stage I sample was 71.67 cm bs (SD=21.80) with a mode of 70 cm bs and a range of 10 to 110 cm bs. Generally, shovel tests were terminated after contact with substrate, which was variably deep across the site due to disturbances. The SCS soil survey for Richmond County (Evans 1999) maps the area of the site as Ailey loamy sand (8 to 15 percent slopes). The on-site soil profile corresponds well to the typical Ailey series pedon. The A-Horizon was very thin, extending between only about 5 and 10 cm bs. It consisted of grayish brown (10YR 5/2) to ligh grayish brown (10YR 6/2) loamy sand that does not appear to represent a plow zone. Dark gray (10YR 4/1) to very dark gray (10YR 3/1) coloration occurred sporadically in the A-Horizon as a consequence of prescribed burn leaching and stump burns. Below this, the E-Horizon generally extended to depths of 50 to 80 cm bs and to depths as great as 115 cm bs in several locations. The E-Horizon is composed of light brown (7.5YR 6/3) to pale brown (10YR 6/3) loamy sand with very low silt/clay content. Below this rests a strong brown (7.5YR 5/6) sandy loam Bt-Horizon substrate, with a relatively high silt/clay content. Masses of reddish to brownish oxidized iron patches occur throughout the Bt-Horizon matrix. The Stage II shovel test sample consisted of 5 m-interval shovel tests placed in the southern portion of the site to expand coverage in the highest artifact density locations and in the northern portion of the site to firm up the site boundary and to investigate these lower density areas (Figure 141). Ninety-six Stage II shovel tests were excavated across the site. Fifty-one of these were positive. Mean shovel test depth for the Stage II sample was 80.63 cm bs (SD=16.01) with a mode of 90 cm bs and a range of 45 to 115 cm bs. The Stage III shovel test sample was distributed to eight locations anchored by Stage I or II shovel tests containing data relevant to evaluating the occupational history and integrity of the site matrix (Figure 142). Sixty-four Stage III shovel tests were excavated at typically 1.25 m-intervals around targeted Stage I and II shovel tests. All but one of these produced artifacts. Mean shovel test depth for the Stage III sample was 93.59 cm bs (SD=8.97) with a mode of 100 cm bs and a range of 50 to 100 cm bs. The greater mean depth of shovel tests in both the Stage II and Stage III samples resulted from focusing on less disturbed areas of the site. Once the results of the Stage III shovel test sample were analyzed, specific locations yielding shovel test outcomes of interest to further evaluating the site were selected for test unit excavation. Stage III shovel testing around targeted shovel test outcomes provided the basis for precise placement of test units. Four 1 m-x-1-m test units were excavated in total and their locations are illustrated in Figure 143. All of the test units were terminated at 70 cm bd, 5 to 10 cm below the E-/Bt-Horizon contact. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 405 Chapter 9. 31RH480 97.0 560 96.5 550 Beaver Pond 96.0 540 il Tra 31RH480 530 95.5 520 510 95.0 Dirt R 500 oad 490 94.5 480 470 94.0 460 450 93.5 440 430 Positive ST Negative ST 420 410 N 400 0 40 m 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 140. Site 31RH480, Stage I Shovel Test Sample. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 406 Chapter 9. 31RH480 97.0 560 96.5 550 Beaver Pond 96.0 540 il Tra 530 31RH480 95.5 520 510 95.0 Dirt R 500 oad 490 94.5 480 470 94.0 460 450 93.5 440 430 Positive ST Negative ST 420 410 N 400 0 40 m 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 141. Site 31RH480, Stage I and II Shovel Test Samples. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 407 Chapter 9. 31RH480 97.0 560 96.5 550 Beaver Pond 96.0 540 il Tra 530 95.5 520 510 95.0 Dirt R 500 31RH480 oad 490 SB1 SB2 SB3 94.5 480 SB4 SB5 SB7 470 SB6 94.0 SB8 460 450 93.5 440 430 Positive ST Negative ST 420 410 N 400 0 40 m 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 142. Site 31RH480, Stage III Shovel Test Sample and Locations of Sample Blocks. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 408 Chapter 9. 31RH480 97.0 560 96.5 550 Beaver Pond 96.0 540 il Tra 530 95.5 520 510 TU1 95.0 Dirt R 500 31RH480 oad 490 TU2 94.5 480 TU3 TU4 470 94.0 460 450 93.5 440 430 Positive ST Negative ST 420 410 N 400 0 40 m 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 143. Site 31RH480, Test Unit Locations. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 409 Chapter 9. 31RH480 ARTIFACT INVENTORY The artifact inventory from Phase II operations at 31RH480 consisted of 1600 items including 1281 pieces of lithic debitage, 14 cores/flake blanks, 4 bifaces, a projectile point, 6 unifaces, 5 utilized flakes, 3 cobble tools, 82 pieces of biotite mica crystal, 135 precontact sherds, 63 pieces of rock/fire-cracked rock, a steatite spall and a piece of metasedimentary float. Appendix B contains the artifact database for the project, while appendices C through I provide additional data on lithic chipped and ground stone tools and precontact ceramics. Descriptions of the various artifact classes and lithic raw material types are presented below. Lithic Raw Material Types Lithic raw material identification proceeded on two different levels. At the macro-scale, individual specimens were typed in accordance with major hard rock geological terminology to achieve comparability with previous projects at Fort Bragg. At the micro-level, however, an attempt was made to further partition some of the hard rock types into “core groupings” to distinguish between specimens derived from single hard rock types that may have originated from different cores. This was done to facilitate spatial analysis in close-interval shovel test grids where overlapping of distinct deposits or complex multi-household occupations were suspected. Core grouping analysis was undertaken principally in metavolcanic types where subtle differences in patina coloration, core color, phenocryst size and density, and matrix texture was easily discerned and monitored. Descriptions of the various rock types and the associated core grouping sub-types recognized in the collection are presented below. Metavolcanic Types Metavolcanic types make up the majority of the chipped stone collection from the site (Table 124). As a group, these types represent 92.8 percent of the total chipped stone artifact inventory (i.e. debitage and chipped stone cores and tools). Descriptions of the five identified metavolcanic types and associated core group subtypes follow. The best reflection of actual representation, however, is supplied by the combined Stage I and II shovel test samples, which is relatively unbiased in terms of spatial coverage. In this sample, metavolcanic material makes up 80.4 percent of the chipped stone, an indication that the Stage III and IV investigations favored contexts that contained greater proportions of metavolcanic chipped stone. Descriptions of the five identified metavolcanic types and associated core group subtypes follow. The typological system varies somewhat from earlier reports in response to the recent stone quarry sourcing study conducted by Steponaitis et al. (2006). (1) Mill Mountain Rhyolite (MMR). MMR was not clearly differentiated in the sourcing study (Steponaitis et al. 2006), but it is a readily recognizable type in the assemblages from Fort Bragg. Mill Mountain is located between the Uwharries Southern and Uwharries Eastern source areas on the opposite bank of the Yadkin River from Table Top Mountain. It was not included in either of these source areas, however, because of the presence of distinctive clear, glassy quartz phenocrysts. Daniel and Butler (1996:18) describe this type as a medium gray, aphanitic, rhyolite porphyry with sparse, glassy phenocrysts of quartz generally less than 1 mm in diameter. Very fine-grained, disseminated grains of pyrite are also present. Thin sections indicate a microcrystalline matrix composed primarily of feldspar and quartz, with some biotite, chlorite, and disseminated pyrite. The specimens assignable to this type from the 31RH480 collection are characterized by a dark reddish gray to moderate brown coloration when not patinated. Patination is generally grayish orange to dark yellowish orange. Four subtypes were assigned to the type, based primarily on the completeness of the observation field and patination. Subtype R8 corresponds to specimens that exhibit sparse glassy quartz phenocrysts only. Subtypes R8Hhf, R8Hhm and R8Hhc are fine-, medium- and coarsegrained specimens respectively that exhibit sparse glassy quartz, white subhedral quartz phenocrysts and, rarely, feldpsar phenocrysts. Although these white quartz phenocrysts were not included in the type description, the subtypes containing them were included as a variant of Mill Mountain Rhyolite due to the co-occurrence with glassy quartz phenocrysts. Mill Mountain Rhyolite makes up 12.8 percent of the combined Stage I and II chipped stone sample. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 410 Chapter 9. 31RH480 Table 124. Lithic Raw Material Frequencies by Chipped Stone Tool Class, 31RH478. UTILIZED FLAKE UNIFACE PROJECTILE POINT CORE BIFACE RAW MATERIAL TYPE DEBITAGE LITHIC RAW MATERIAL FREQUENCIES BY CHIPPED STONE TOOL CLASS: 31RH480 GRAND TOTAL Metavolcanic Hydrated metavolcanic 4 4 Metavolcanic 1 1 Mill Mountain Rhyolite 82 Type I Rhyolite Tuff 32 2 32 Uwharries Eastern Rhyolite 126 Uwharries Southern Rhyolite (1) 162 1 Uwharries Southern Rhyolite (2) 784 3 Uwharries Western Rhyolite 84 1 1 3 1 128 4 2 170 2 1 793 1 4 5 Metasedimentary Black Metasedimentary 1 1 Gray/Tan Metasedimentary 3 3 1 1 Crystal Quartz 5 5 Gray Quartz 2 2 White Quartz 72 Yellow Quartz 1 Quartzite Quartz 7 79 1 Chert Chert, Dark Gray Purplish Tan Mottled Chert GRAND TOTAL 1 1 1 1281 1 4 14 1 6 5 1311 (2) Type I Rhyolite Tuff (Type I RT). An exact correlate of this type was not discernible in the sourcing study (Steponaitis et al. 2006). Benson (1999:30) describes it as fine-grained dark green to gray material that looks and feels very similar to Piedmont chert. He noted that plagioclase phenocrysts were sporadically present. Two subtypes were identified in the 31RH480 collection. Subtype R2 is only rarely identified in assemblages on Fort Bragg. Macroscopically it appears very similar to Subtype R3t, but its coloration can be described as a dusky blue green. The texture is microcrystalline and the edges are semi-translucent. Subtype R2A specimens exhibit a pale green patina, but broken edges reveal that the underlying material corresponds to R2. Darker green splotches and swirls on the faces of these specimens represent more resistant matrix. The matrix is aphantitic and microcrystalline in texture and is highly isotropic in fracture characteristics. The origin of Type I Rhyolite Tuff is not well understood and it is possible that it represents a very fine-grained metamudstone similar to some of those described for Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 411 Chapter 9. 31RH480 the Chatham Pittsboro source (see Moore and Irwin 2006:27-28; Stoddard 2006:57). In support of this is the occurrence of what appear to be bedding laminations in some of the patinated material. Moreover, the color of the patinated specimens is similar to that reported for the Chatham Pittsboro source. Type I Rhyolite Tuff makes up 3.4 percent of the combined Stage I and II chipped stone sample. (3) Uwharries Eastern Rhyolite (UER). Stoddard (2006:52) describes the specimens from this sourcing area as light to dark gray metadacite porphyry or crystal lithic tuff. All samples collected contained plagioclase feldspar and white quartz phenocrysts, ranging from less than 2 percent to 7 percent phenocryst density. Common metamorphic minerals within the matrices include biotite, stilpnomelane and calcite. Similar dacites and rhyolites are typical of the Uhwarries Southeastern source (Stoddard 2006:55-56) as well, but many of these specimens contain silica levels beyond the range typical of igneous rocks. Macroscopic criteria for distinguishing between lithic material from these two sources does not at present exist. Consequently, specimens from both source areas may be present in the analyzed lithic assemblages at Fort Bragg. Typical specimens from 31RH480 are characterized by dark gray (nonpatinated) and grayish green (patinated) matrices with splotches of silicate and variable percentages (2 to 4 percent) of white quartz and plagioclase feldspar phenocrysts. Some specimens exhibit purplish to brown background matrices. Quartz phenocrysts predominate over feldspar in the examples from the site. Type I Rhyolite, which has been distinguished separately in previous reports (Cable and Cantley 2005b, 2006, 2010), is also subsumed under this type. Two subtypes were identified in the collection: R8Ihm and R8Ihc, both of which contained white quartz and plagioclase feldspar phenocrysts in variable percentages. They are distinguished from each other by differences in grain. Respectively, the subtypes correspond to medium- and coarse-grained specimens. UER makes up 17.2 percent of the combined Stage I and II chipped stone sample. (4) Uwharries Southern Rhyolite (USR). This type includes sources found on Morrow Mountain and surrounding areas, most notably Table Top Mountain. Daniel and Butler (1996:10-13) describe this type as a dark gray, aphanitic, aphyric rhyolite, that commonly exhibits flow-banding. Weathering tends to bring out the flow lines visually, which generally alternate in a pattern of light and dark gray. Although the matrix is usually homogeneous, some specimens exhibit small spherulites of less than 1 mm in diameter (see Stoddard 2006:52). Thin sections indicate a microcrystalline intergrowth of feldspar and quartz, with minor biotite and chlorite elements. The individual minerals are difficult to distinguish and strings of dark minerals can occur along fracture planes. The specimens identified as USR in the 31RH480 assemblage are composed of two larger groupings, the R3m and the R3/R3t groups. These are respectively referred to as USR (2) and USR (1). The R3m Group corresponds to material that typically occurs on Morrow Mountain and includes a series of subtypes that describe variation in flow banding, texture and patination. Subtype R3m is non-patinated and it exhibits a dark gray fine-grained matrix devoid of phenocrysts. Subtypes R3mD, R3mDf and R3mDs are patinated and range in color from grayish green to greenish gray to moderate yellow or tan. R3mD represents patinated specimens of R3m, while R3mDf is made up of patinated specimens that exhibit flow banding. R3mDs exhibits blotches and linear arrays of silicate containing biotite and perhaps chlorite elements, respectively. It too, appears to represent a patinated form of R3m. The R3/R3t group is composed of an extremely fine grained or microcrystalline dark gray matrix that is semi-translucent when viewed on attenuated edges. Finely disseminated feldspar is commonly seen on exposed surfaces, but the small blotches do not form phenocrysts. R3, R3t, R3f are non-patinated. R3 is distinguished from R3t by a somewhat grainier matrix and greater opacity. R3f is the flow-banded version of R3. Subtypes R3p and R3pat are fully patinated. Both are tan in color, but R3p also exhibits finely disseminated biotite specks. R3pat appears to represent patinated specimens of R3t. USR (1) comprises 21.6 percent the Stage I and II chipped stone sample, while USR (2) makes up 24 percent. (5) Uwharries Western Rhyolite (UWR). This source area includes specimens from Wolf Den Mountain and Falls Dam (Stoddard 2006:52). This Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 412 Chapter 9. 31RH480 material is described as gray to black microcrystalline felsic volcanic rock that contains rounded plagioclase phenocrysts and glomerocrysts. Green biotite and pale green amphibole are dispersed locally in the matrix and spherulites are sometimes present. Three subtypes were identified in the collection. The main subtype, R3mp, has been patinated to a grayish green or greenish gray color. Non-patinated specimens may have been assigned to one of the other Uwharrie source areas in the analysis, but the distinctive glomerocyst structure suggests that this was not a major problem. Due to patination, most of the specimens in the collection were characterized by fine to medium texture. A single Subtype was identified, R3mp. UWR was not represented in the Stage I and II chipped stone sample. Only 5 specimens of this raw material type were recovered during the Phase II operation at 31RH480. (6) Indeterminate Metavolcanic. A small amount of the metavolcanic material identified in the analysis could not be further classified to type due to excessive hydration. These were simply classified as “indeterminate metavolcanic” or “hydrated metavolcanic.” Only 5 specimens were assigned to this category in the chipped stone inventory. Quartz Types Pure quartz forms as veins in igneous and metamorphic rock formations characterized by slow crystallization. This generally produces anhedral, milky structures (Spock 1962). In some cases, however, crystallization occurs earlier in the sequence, producing clear crystals with euhedral structure. Precontact groups in the project area utilized both vein and euhedral crystal quartz, but most of it appears to have come from river gravel deposits. A great deal of variation was evident in the vein quartz category ranging from mixed clear and cloudy exposures to opaque milky white. Since these characteristics appeared to be repeatedly present on single cores when debitage concentrations were analyzed, all of this variation was subsumed under the single white quartz subtype Q5. The category “crystal quartz” (CQ) was retained for instances where glass-like examples and crystal facets strongly suggested a euhedral crystal origin. A single piece of yellow stained, fractured quartz was recovered and it was identified as “yellow quartz.” Staining of this type is commonly observed in quartz cobbles. Highly isotrophic gray quartz was also identified in low quantities at the site (n=2). White quartz represented 16.7 percent of the combined Stage I and II chipped stone sample. Only 5 pieces of crystal quartz were found in the Phase II operations and none of these showed up in the Stage I and II sample. Gray quartz and yellow quartz made up 0.5 percent of the sample each. In addition to debitage, quartz cobbles and fragments of various types were recovered from the investigation. These included 30 pieces of firecracked rock, 3 cobble fragments and a hammer stone. Minority Types Several other rock types were recognized in the 31RH480 collection. Descriptions of each are presented below. (1) Arenite. This material has been previously identified as ferruginous sandstone, but thinsectioning of a piece of this material indicates that a more accurate characterization is arenite, or hematite cemented quartz sandstone (Appendix K). Three pieces of arenite were classified as fire-cracked rock. It occurs in a variety of shapes including flattened and irregular nodules. The exterior of the nodules consist of fine-grained ruddy brown precipitate while the centers consist of dense, reddish brown to purple, densley cemented sandstone. The material was commonly used for rock hearths by the precontact inhabitants of Fort Bragg. Twenty-three pieces of arenite were collected during the investigation, totaling 521.58 gm. Float, fire-cracked rock and rock were identified. (2) Biotite. Numerous dark, tabular crystals were recovered during the Phase II operation that appear to be biotite. Biotite is a sheet silicate in which iron, magnesium, aluminium, silicon, oxygen, and hydrogen form thin planar layers that are weakly bound together by potassium ions. Biotite is sometimes referred to as “iron mica” or “black mica” because it is extremely iron-rich and dark in coloration. In North Carolina biotite is occasionally found in large cleavable crystals, especially in pegmatite veins (Rickwood 1981). Whether these crystals were intentionally gathered by precontact groups and brought to 31RH480 for use is not clear, as cut marks or other manufacturing evidence was not observed Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 413 Chapter 9. 31RH480 on the fragments. Mica, however, was long used by precontact groups as a raw material to produce ornamental objects. (3) Chert. Two pieces of chert were identified in the collection. One specimen, a piece of debitage, was a purplish tan, fossiliferous chert with abundant disseminated mica. Although not a precise match, it resembles the characteristics of cherts found in the Thanetian Black Mingo Formation in the Santee River Valley of South Carolina (Anderson et al. 1982:127-128). The other specimen, a utilized flake, was a medium to dark gray chert of dull luster and abundant, finely disseminated mica. Numerous gray to dark gray chert sources have been identified in the Durham-Wadesboro Basin in the eastern Piedmont (Lautzenheizer 1996). (4) Gneiss. A single river cobble composed of a foliated, coarse-grained, micaeous igneous rock was identified in the collection. The raw material type is tentatively identified as gneiss. (5) Metasedimentary. Seven items of metasedimentary rock were identified in the collection. One core flake was made of a black material, while the other specimens were composed of a greenish gray, laminated material. The latter material was represented by two fragments from an abrader, three pieces of debitage and a piece of float. The relative coarseness of the grain suggests that the material can be further classified as metasiltstone or fine metasandstone. Numerous source areas in the Durham-Wadesboro Basin contain similar rock types, including the Chatham Pittsboro, Person County and Chatham Siler City (Stoddard 2006:57-63). (6) Quartzite. A single piece of debitage was composed of a pale red, highly micaceous medium to fine grained quartzite resembling orthoquartzite from the Santee River Valley of South Carolina (Anderson 1982:120-124). The latter material derives from the Thanetian Black Mingo Formation and it is typically described as chert and chalcedony cemented quartz arenite or chert and chalcedony cemented sandstone. Large cobbles and chunks of lag gravel made of this material are found in terrace exposures along the Lower Santee and Black River valleys. (7) Sandstone Conglomerate. Also abundantly present in the deposit were chunks of sandstone conglomerate containing a wide range of sand and pebble sizes. Thin-sectioning of a piece of this material determined that individual grains of sand were cemented with clays (wacke) and exhibited udulatory extinction indicating that it originated from metamorphic rock (Appendix K). Occasionally interfaces with arenite could be observed on the chunks, indicating that the two materials were derived from the same formations (e.g. Middendorf). The matrix of the conglomerate consistently exhibited a grayish orange to dusky yellow coloration. Chunks of arenite were sometimes viewed in the matrix as well. This material appears to have been used to make rock hearths. Four pieces of sandstone conglomerate were collected, weighing 139.35 gm. (8) Steatite. A single spall of a talc-schist material of medium light gray coloration may represent a fragment form a steatite bowl. Steatite vessel fragments are sporadically found on Fort Bragg (see McNutt and Gray 2007; Millis et al. 2010; Ruggiero 2003). The material is commonly found at locations in the Piedmont, although the nearest known quarries are in South Carolina and south central Virginia (see Millis et al. 2010:460-461; Truncer 2004). Lithic Artifacts Basic definitions of the various lithic artifact classes were presented in Chapter 4. Here, specific descriptions of the 31RH478 lithic collection are presented. Debitage The debitage classification is adapted from formal (Bradley 1973; Frison and Bradley 1980; House and Wogaman 1978; Newcomer 1971), attribute (Moore 2002; Shott 1994), and mass (Ahler 1989; Moore 2002) analysis approaches. Debitage classes were devised to identify both reduction stages and reduction/production systems (ie. biface core reduction, directional core reduction, and flake blank production). This classification was supplemented with the recording of three attributes: (1) size class, (2) percent cortex, and (3) condition. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 414 Chapter 9. 31RH480 Table 125 presents the complete inventory (all field stages) of debitage type by raw material type from 31RH480. Biface reduction types (FBRs and BTFs) comprise the majority of identifiable debitage, but directional core reduction as represented by core flakes is relatively high in some raw material types. Directional core debitage comprises 11 to 50 percent of the identifiable debitage types (i.e. core flakes, FBRs and BTFs) within the metavolcanic types and 60 percent of the white quartz. The distribution of cortex on dorsal faces and platforms of debitage provides a basis for further evaluating the nature of reduction strategies associated with the various raw material types (Table 126). Dibble et al. (2005:550) indicate that raw nodule reduction from bifaces should yield between about 6 and 21 percent flakes with cortex for any single episode. It is likely that similar proportions characterize directional core reduction from raw nodules. Type I RT, USR (1) and USR (2) exhibit percentages (6.3 to 17.9 percent) within the range of raw nodule reduction. Only USR (2), however, displays a correspondingly high core flake percentages. This suggests that the other two types are characterized by only prepared core and biface reduction strategies. The debitage profiles of the remaining types are also characteristic of prepared core and biface reduction strategies. Cores Metric and attribute data on cores are provided in Appendix C. Two classes of cores were iden- Table 125. Distribution of Debitage Type by Raw Material Type, 31RH480. DISTRIBUTION OF DEBITAGE TYPE BY RAW MATERIAL TYPE: 31RH480 RAW MATERIAL TYPE BTF FBR CORE FLAKE CHUNK FLAKE FRAGMENT GRAND TOTAL Metavolcanic Hydrated Metavolcanic Mill Mountain Rhyolite Type I Rhyolite Tuff 7 29 9 5 5 37 82 6 10 2 14 32 Uwharries Eastern Rhyolite 16 27 29 54 126 Uwharries Southern Rhyolite (1) 41 42 17 62 162 Uwharries Southern Rhyolite (2) 90 158 114 415 784 1 1 2 4 Uwharries Western Rhyolite 7 Metasedimentary Black Metasedimentary 1 Gray/Tan Metasedimentary 1 1 2 3 3 5 1 2 56 72 Quartz Crystal Quartz 1 Gray Quartz White Quartz 1 1 5 Yellow Quartz 1 9 1 1 1 Miscellaneous Quartzite 1 Purple Tan Mottled Chert GRAND TOTAL 1 1 165 271 1 185 9 651 1281 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 415 Chapter 9. 31RH480 Table 126. Percent Cortex on Debitage and Percent Core Flakes by Raw Material Type, 31RH480. PERCENT CORTEX AND CORE FLAKES BY RAW MATERIAL: 31RH480 RAW MATERIAL TYPE TOTAL DEBITAGE PERCENT W/CORTEX PERCENT CORE FLAKES Mill Mountain Rhyolite 82 1.2 20.0 Type I Rhyolite Tuff 32 6.3 11.1 Uwharries Eastern Rhyolite 126 2.4 40.3 Uwharries Southern Rhyolite (1) 162 17.9 17.0 Uwharries Southern Rhyolite (2) 784 13.3 31.5 4 0.0 50.0 72 1.4 60.0 Uwharries Western Rhyolite White Quartz TOTAL 1262 tified in the 31RH480 collection: (1) Core Fragments and (2) Flake Blanks. Each class is discussed below. (1) Core Fragments. Angular pieces of raw material containing portions of one or more flake scars suggestive of directional core reduction, but which lacked striking platforms, were classified as core fragments. Seven such fragments were recognized in the collection, one of which was made of USR (2) while the remaining specimens were composed of white quartz (Table 127). The specimens are all small and, although they were most probably derived from directional cores, they exhibit very few of the features of said cores. (2) Flake Blanks. Flake blanks were removed from a core for the purpose of further modification or use as a tool (Bradley 1973:6). They do not exhibit obvious macroscopic evidence of use or modification, but they may have been used for a brief activity. Seven flake blanks were identified in the collection (Table 127, Figure 144:A-D). Most of these were composed of metavolcanic materials, but one large white quartz core flake fragment was also included in this category. Both blade flakes (flakes approximately twice as long as they are wide) and more amorphous core flake flakes are represented in the collection. The dorsal surfaces of the flakes reveal the former facial structure of directional cores. Bifaces Four biface fragments were recovered during the investigation (Table 128). Bifaces are derived from both large mass packages (i.e. unmodified and/ or prepared cores) and flake blanks. Determination that a biface was derived from a flake blank was made when flake characteristics were observed on the item, such as remnant striking platforms, flake curvature, or bulbs of percussion from flake detachment. The bifaces in the collection do not have sufficient mass to suggest that they were derived from large biface cores. Instead, they all appear to have been produced from flake blanks of varying sizes. Type I and III bifaces, as adapted from Daniel (2002:51-54), are represented in the collection. Each type is described below. (1) Type I Bifaces. Daniel (2002:51) describes Type I Bifaces as irregularly shaped flake blanks and core masses with large conchoidal scars forming at least one section of a sinuous edge along the core edge. No evidence of secondary thinning Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 416 Chapter 9. 31RH480 Table 127. Summary Data for Cores, 31RH480. CORES: 31RH480 SPECIMEN # BAG CORE TYPE PORTION FLAKE TYPE LITHIC RAW MATERIAL TYPE MAXIMUM THICKNESS (mm) MAXIMUM LENGTH (mm) MAXIMUM WIDTH (mm) WEIGHT (gm) 271269a48 29 Core Fragment Fragment White Quartz 23.91 38.65 18.85 16.18 271269a87 51 Core Fragment Fragment White Quartz 16.67 26.06 16.07 7.73 271269a157 98 Core Fragment Fragment White Quartz 10.37 29.86 16.59 5.26 271269a269 174 Core Fragment Fragment White Quartz 12.47 32.78 21.58 6.9 271269a335 207 Core Fragment Fragment White Quartz 20.98 29.03 14.31 8.32 271269a612 351 Core Fragment Fragment White Quartz 12.31 23.30 13.26 6.66 271269a820 401 Core Fragment Fragment Uwharries Southern Rhyolite (2) 15.81 44.70 21.29 9.92 (51.80) (48.45) (48.79) 43.42 271269a3 2 Flake Blank Fragment Core Flake White Quartz 271269a75 48 Flake Blank Fragment Core Flake Mill Mountain Rhyolite 3.16 35.38 32.62 4.5 271269a98 56 Flake Blank Fragment Core Flake Uwharries Eastern Rhyolite 9.92 37.16 (57.20) 19.26 271269a305 191 Flake Blank Whole Core Flake Mill Mountain Rhyolite 5.66 30.49 43.87 8.22 271269a561 333 Flake Blank Whole Blade Flake Uwharries Southern Rhyolite (2) 5.56 51.77 21.02 6.79 271269a568 334 Flake Blank Whole Blade Flake Uwharries Southern Rhyolite (2) 4.94 35.78 21.97 3.37 271269a674 374 Flake Blank Whole Core Flake Uwharries Western Rhyolite 5.02 46.03 35.07 5.63 scars is observable and a relatively high thickness to width ratio obtains. Three Type I Biface fragments were recognized in the collection. Specimen a135 was made of USR (2) material and represents a lateral section of a thin biface made from a flake blank. Specimen a247 (Figure 144:E) represents an ovoid end fragment made of USR (2) material. Cortical remnants are still observable from the dorsal surface of the original flake blank. The cross-section is plano-convex and flake scars occur along the margins of both faces. Finally, Specimen a440 (Figure 144:F) was made of USR (1) material. The dorsal surface of the thin blank is covered almost entirely in cortex and bifacial shaping had not yet been completed along the margins when it snapped. All of the Type I Bifaces were discarded upon breakage during manufacture, as reflected by step and hinge fracture terminations across the blade faces. (2) Type III Bifaces. Type III Bifaces represent late stage preform rejects. The edges exhibit secondary thinning and shaping scars and facial retouch is globally distributed. Outline shapes are well defined and reflect ovate and lanceolate forms. Most of these specimens display step and hinge fractures that terminated production. Type III Bifaces tend to have the lowest ratios of thickness-to-width. Sometimes features of diagnostic projectile point haft elements are present, allowing correlation with specific culture-historic types. In these cases, specimens are differentiated and identified as “Preforms.” One Type III biface fragment (a789) was identified in the collection (Figure 144:G). It was made of USR (2) material and it represented a pointed tip of what appears to have been a projectile point preform in the final stages of manufacture. The cross-section is biconvex. The large size and thickness of the fragment strongly suggest that it is Archaic in affiliation. Projectile Points Projectile points are bifacially flaked tools with retouched haft elements. In addition, projectile point fragments consist of specimens with the fine bifacial retouch characteristic of finished projectile points, but not necessarily exhibiting evidence of a haft element. The collection from 31RH480 con- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 417 Chapter 9. 31RH480 Figure 144. Chipped Stone Tools, Site 31RH480. A-D: Flake Blanks (a3, a98, a561 and a674 respectively), E-F: Type I Biface Fragments (a247 and a440 respectively), G: Type III Biface Tip (a789), H: Palmer II Corner Notched (a679), I-J: Type IIa Side Scrapers (a554 and a374 respectively), K: Type IIb Side Scraper (a630), L: Core Scraper (a121), M-N: Gravers (a125 and a460 respectively). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 418 Chapter 9. 31RH480 Table 128. Summary Data for Bifaces, 31RH480. BIFACES: 31RH480 SPECIMEN # BAG TYPE PORTION 271269a135 83 Type I Biface Fragment Lateral Section 271269a247 164 Type I Biface Fragment End 271269a440 266 Type I Biface Fragment End 271269a789 399 Type III Biface Fragment Tip SHAPE Ovoid Ovate/Triangular LITHIC RAW MATERIAL TYPE Maximum Thickness (mm) Maximum Length (mm) Maximum Width (mm) Weight (gm) Uwharries Southern Rhyolite (2) 3.05 (26.43) (9.41) 0.86 Uwharries Southern Rhyolite (2) 11.47 (23.89) (40.43) 11.22 Uwharries Southern Rhyolite (1) 5.18 (21.01) (24.59) 3.64 Uwharries Southern Rhyolite (2) 9.47 (24.65) (20.71) 2.76 sisted of a single projectile point base/lower blade fragment. Metric and descriptive data on projectile points can be found in Appendix E. The projectile point fragment (a679) was typed as a Palmer II Corner Notched point (Figure 144:G). Coe (1964:67) described the Palmer Corner Notched type as a “small corner-notched blade with a straight, ground base and pronounced serrations.” Blade edges were generally straight, but were secondarily both incurved and excurvate. Bases were commonly straight and typically ground. Blade serrations were at times deep, averaging about 3 mm in width and 5 to 7 mm in length. The width of the shoulder barbs usually exceeded that of the base. Metric dimensions ranged between 28 and 60 mm in length, 15 and 25 mm in width, and 5 to 12 mm in thickness. At the Hardaway Site, Palmer Corner-Notched points were primarily associated with Level III, sandwiched between Hardaway Side-Notched and HardawayDalton styles in Level 4 and Kirk Serrated and Kirk Corner-Notched in Level II. As an outcome of the Haw River excavations, Cable (1982) further divided the Palmer type into three variants, Palmer I, II and III. The Palmer I style is smaller, typically exhibits ground bases and was found in a slightly lower vertical position at Haw River than the larger variants described as Palmer II and Palmer III. The latter is the largest of the three variants and is characterized by a low incidence of basal grinding Palmer II is stylistically intermediate and it may overlap the temporal distributions of the other variants. Others suggest that all of this variation is largely a consequence of life-stage variation rather than temporal differences (Daniel 1998; Sassa- man 2002). The time range of this style is generally accepted at 9500–8900 BP (Sassaman et al. 2002:10). The specimen from 31RH480 was made of USR (1) material. The base was moderately ground and had been thinned in the center during use and repair. The basal outline was straight. The blade margins, which were straight and recurved outward near the shoulders, exhibited worn down serrations. The upper blade had been snapped off in a step fracture, which may have led to its discard. The blade took on a “Christmas tree” appearance from episodes of resharpening and the margins were highly serrated but not beveled. The tang was wide and relatively long. By virtue of its intermediate size and ground base, it was typed as a Palmer II Corner Notched point. The specimen was recovered from a depth of 40 to 50 cm bd in TU3. Unifaces Unifaces consist of tools that are retouched along flake margins in a single direction (unifacial edge shaping). Retouch was applied to shape a working edge or bit and, sometimes, a butt element for hafting. Limited bifacial retouch or resharpening is sometimes present along haft elements or working edges. Metric and attribute data for unifaces are presented in Appendix F. Six items classified as unifaces were recognized in the collection (Table 129). These include 3 side scrapers, a core/scraper and 2 gravers. The specimens are described below under the rubric of the typological system developed by Coe (1964) and elaborated by Daniel (1998). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 419 BAG 331 228 356 73 75 280 SPECIMEN NO. 271269a554 271269a374 271269a630 271269a121 271269a125 271269a460 Graver Graver Core/Scraper Type IIb Side Scraper Type IIa Side Scraper Type IIa Side Scraper UNIFACE/RETOUCHED FLAKE TYPE PORTION Whole Whole Fragment Whole Fragment Whole Core Flake Core Flake Lenticular Lenticular Lenticular Directional Core/Flake BLANK TYPE Lenticular Lenticular Lenticular CROSS-SECTION Core Flake FBR Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) LITHIC RAW MATERIAL TYPE 5.71 3.13 (24.55) 12.12 (18.92) 22.73 RETOUCH LENGTH (mm) SCRAPERS AND RETOUCHED FLAKES: 31RH480 (36.00) RETOUCH LENGTH (mm), OTHER MARGIN MAXIMUM EDGE ANGLE 28° 44° 80° 54° 59° 67° 1 1 3 1 1 2 NUMBER OF RETOUCHED EDGES 420 4.26 1.44 20.15 4.32 2.71 4.50 MAXIMUM THICKNESS (mm) Table 129. Summary Data for Scrapers and Retouched Flakes, 31RH480. MAXIMUM LENGTH (mm) 25.22 11.95 (53.08) 17.61 (18.87) 28.49 MAXIMUM WIDTH (mm) 12.42 8.44 42.73 12.92 (10.96) 32.82 WEIGHT (gm) 0.88 0.11 38.75 0.62 0.5 4.5 Graver Spur and hafting Notches on Lateral Margins. Graver Spur and Spoke Shave Concavity Opposite Scraper Margin NOTES Chapter 9. 31RH480 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC Chapter 9. 31RH480 (1) Side Scrapers. Three side scraper specimens were identified in the 31RH480 collection. Daniel (1998:6683-84) defines side scrapers as unifaces containing “one or more unifacially retouched working edges that parallel the long axis of the flake blank.” Daniel identified four side scraper types and numerous subtypes in the Hardaway scraper collection. Side scrapers from 31RH480 are all placed in the Type II category. Building on Coe’s (1964:79) original description for the Type II Side Scraper, Daniel (1998:84-93) separated individual specimens further into two types based on the size of the flake blanks used to manufacture the tools. Type IIa refers to the classic description of a Type II Side Scraper. That is, scrapers made on large, irregular flakes whose bit edges were retouched without concern for regularizing the marginal outline. Daniel’s Type IIb was reserved for very small flakes, usually of less than 30 mm in length. Coe (1964:79) observed that Type II Side Scrapers were strongly associated with the Palmer and Hardaway levels (III and IV) at the Hardaway Site. Two Type IIa Side Scrapers were recognized in the 31RH480 collection. Specimen a554 (Figure 144:H) was made on a large FBR of flow-banded USR (2) material. Regularized and fine pressure flaked marginal retouch was carefully applied to a straight to slightly convex working edge. The pointed edge opposite the scraper margin was retouched along one margin to enhance the point into a graver spur. The concavity along this edge may also have functioned as a spoke shave. Most likely this tool was not hafted, as this action would have hindered the use of one or more edges. The other Type IIa specimen (a374) represents a small fragment of an edge exhibiting steep marginal retouch along a straight side (Figure 144:I). It was made of USR (2) material. Both scrapers were found in deep vertical positions in keeping with their probable Early Archaic association. The former specimen was found at a depth of 40 to 50 cm bd in TU2, while the latter was recovered in a Stage III shovel in SB1 at a depth of 50 to 60 cm bs. A single Type IIb Side Scrapers was identified in the collection (Figure 144:J). It was made on a small, whole core flake with cortex on the platform and part of the dorsal face. The scraper margin was formed on the side opposite the striking platform of the flake. The margin was only minimally retouched and the bit was convex in outline. It was also found in a deep vertical position, at a depth of 40 to 50 cm bd in TU4. (2) Core/Scraper. Daniel (1998:106) identified six scrapers in the Hardaway collection that he identified as Core/Scrapers. These items were similar to rough choppers recovered from the site, but were smaller and characterized by more careful and regular retouch. These items were plano-convex to trapezoidal in cross-section and appear to have been cores or core fragments that were modified into scrapers. The specimens displayed flat, oval surfaces that served as striking platforms for all flake removals and two of the specimens resembled quite closely the form of ployhedral cores. The specimen (a121) from 31RH480 shares these characteristics (Figure 144:K). Steep and undercutting marginal retouch was evident across two sides, but the bit end of the tool appears to have partially broken off by a step fracture. The two sides appear to have hafting notches and one of these notches is separated from the bit by a graver projection. The fact that the tool was hafted and one side of the bit supported a graver spur suggests that this specimen may actually represent a roughly finished Type III End Scraper. It was found in a Stage II shovel test at a depth of only 20 to 30 cm bs. (3) Gravers. Gravers represent flakes with short unifacially flaked projections (Daniel 1998:104; Goodyear 1974:55). It is generally held that gravers served as engravers or points for slotting bone and wood. Two small flakes in the collection display graver projections. Specimen a125 (Figure 144:L) was made on an extremely small core flake measuring only about 8.5 x 12 mm. Unifacial retouch was applied to one side of the projection to enhance the natural acute angle. It was made of USR (1) material and was found in a Stage II shovel test at a depth of only 0-10 cm bs. The other specimen (a460) was made on a slightly larger flake of USR (1) material (Figure 144:M). Again, unifacial retouch was applied to one side of the natural projection to enhance its sharpness. This specimen was found in a Stage III shovel test in SB8 at a depth of 20 to 30 cm bs. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 421 Chapter 9. 31RH480 Utilized Flakes Five un-retouched flakes exhibited wear patterns on one or more edges, indicating their use as expedient tools (Table 130). Metric and attribute data for utilized flakes are presented in Appendix G. All of these specimens exhibited edge attrition and bifacial nibbling scars indicative of cutting functions. Utilized edge angles fell between 28°and 58°. Most of the specimens appear to have been derived from block cores (i.e. core flakes and blade flakes), but one was made from a large FBR. Most of the utilized flakes were made on metavolcanic raw material, but one was composed of a dark gray chert. Cobble Tools Three items were classified as cobble tools. One was a quartz cobble hammer stone (a656) that exhibited battering and grinding wear on one observable end and the two side margins (Figure 145:A). The opposite end was broken off. It weighed 140.98 gm and it was found at a depth of 20 to 30 cm bd in TU3. The other two items (a193 and a194) are fragments from a single abrading tool (Figure 145:B). The original item was made from a greenish gray medium grained metasedimentary river nodule that appears to have been modified on only one flat surface. This surface was ground to a regularized contour and displayed numerous striations running roughly parallel to the long axis of the former implement. The combined weight of the two fragments was 10.95 gm. They were found at depths of 40 to 50 and 50 to 60 cm bs in a Stage I shovel test. Metric and attribute data for these items are presented in Appendix H. Minerals Eighty-two pieces of what appears to represent tabular biotite mica crystals were recovered during the Phase II operation (Figure 145:C). Biotite is a sheet silicate in which iron, magnesium, aluminium, silicon, oxygen, and hydrogen form thin planar layers that are weakly bound together by potassium ions. All of the material was collected form a single location at the site, in TUs 3 and 4 and the surrounding SB4 (Figures 142-143). The crystal fragments were primarily concentrated in an area measuring about 1.5 m in diameter at depths of 20 and 40 cm bs/bd. It is not clear how the crystals might have come to rest in this particular context, unless a precontact group brought them to the site. A piece of white quartz with biotite sheets attached to it was found within the concentration, indicating that the material originated in a quartz vein or in a piece of stream gravel detached from a vein at some time in the past. Evidence of Table 130. Summary Data for Utilized Flakes, 31RH480. LITHIC RAW MATERIAL TYPE MAXIMUM THICKNESS (mm) MAXIMUM LENGTH (mm) MAXIMUM WIDTH (mm) WEIGHT (gm) EDGE ANGLE 271269a37 23 Whole Cutting FBR 1 Uwharries Southern Rhyolite (1) 3.02 22.68 17.58 1.19 28° 271269a94 56 Whole Cutting Core Flake 2 Uwharries Eastern Rhyolite 8.42 68.48 29.82 15.4 46° 271269a479 290 Whole Cutting Core Flake 1 Uwharries Southern Rhyolite (1) 4.96 25.98 32.6 3.42 46° 271269a512 310 Whole Cutting Core Flake 2 Chert, Dark Gray 8.72 21.48 10.22 1.20 58° 271269a777 398 Fragment Cutting Blade Flake 2 Uwharries Southern Rhyolite (2) 2.41 36.57 11.54 0.95 29° SPECIMEN # BAG CONDITION UTILIZED FLAKE FUNCTION NO. UTILIZED EDGES UTILIZED FLAKES: 31RH480 BLANK TYPE Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 422 Chapter 9. 31RH480 Figure 145. Miscellaneous Stone Tools and Minerals, Site 31RH480. A: Hammer Stone (a656), B: Abrader Fragment (a193), C: Biotite Crystals (m660). cutting or other manufacturing modifications was not observed on the crystals. However, use of mica to manufacture ornaments and to provide power objects for shaman’s tool kits is well documented in the ethnographic and archaeological records throughout North America and the world. Precontact Ceramics One hundred thirty-five precontact ceramics were recovered from the Phase II operation at 31RH480. This is a large collection by Sandhills standards in the region, but 95 of the sherds belong to a single vessel. The remaining 40 sherds are divided between 26 other vessels. Although monitoring of temper types within series has proceeded in an ad hoc fashion in earlier investigations (see Cable and Cantley 2005a, b, 2006, Cable 2010), the enormity of the inventory of temper constituents and various compositional combinations seen in ceramic collections in the region makes it imperative that in the future a universal system be developed to accommodate cross-project comparability. Formerly, temper classes were assigned letter designations as they were historically recognized within each series, creating a confusing list of dif- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 423 Chapter 9. 31RH480 ferent letter designations for the same temper type within each series. Here, temper constituent designations are standardized into a single classificatory system. The letters “a” and “b” are reserved for the basic paste representatives diagnostic of each series as described in previous reports. The remaining letter designations note the addition of other temper constituents to these basic paste categories as described below: (c) Crushed Arenite (d) Granule to Pebble Sized Angular/Subangular Quartz Sand and/or Crushed Quartz (e) Crushed and Ground Indurated Granite (f) Medium to Granule Sized Angular/Subangular Feldspar Sand More than one of these temper constituents is sometimes present in a single sherd. In these cases, the constituent designations would include combinations of letter designations (e.g. “c/e” or “e/f”, etc.). During the course of the ceramic analysis of the DO5 site package it also became apparent that some of the pottery exhibited extremely hard pastes, much harder, in fact, than the typical paste previously defined for each series. A hardness scale consisting of three attribute states was developed to describe this variation. These consisted of: (1) friable, (2) compact and (3) hard. The attribute state of “friable” corresponds to sherds that are easily crushed on their edges with only minimal application of pressure with the thumb. A “compact” state was identified in instances where moderate pressure with the thumb was required to break the edge of a sherd. Finally the “hard” state was reserved for those sherds whose edges were not easily broken even with extreme thumb pressure. Generally, pottery achieves greater hardness when one or more of three conditions prevail (Shepard 1954:114). These are: (1) the pottery is made from low-fusing, dense-firing clay, (2) the pottery is fired at a high temperature and/or (3) the pottery is fired in an atmosphere promoting vitrification. It is common to observe in the local ceramic sequences of the Southeast, a progression from friable ceramics to much harder, more durable pottery in the later stages of cultural development. Clearly, Mississippian potters had developed technological strategies to produce extremely hard, sometimes vitrified pot- tery and it is quite possible that this advancement was meant to adapt to a more sedentary, agricultural life way in which durability and longevity of pots would have been advantageous. Similar adaptive technologies may have been incorporated into sedentary, agricultural societies peripheral to the Southeast Appalachian Mississippian culture as well, and for the same reasons. Currently, a sophisticated method for measuring hardness in “primitive” pottery does not exist due to its porosity and heterogeneity of paste composition. The “thumb” test, although somewhat imprecise, provides a relatively objective and empirical basis for identifying relative hardness in precontact pottery assemblages. For an extended discussion of hardness in precontact ceramics see Chapter 4. This method of analysis resulted in segregating extremely hard or vitrified pottery that would have traditionally been placed in the Cape Fear (sandtempered), Hanover (sand- and grog-tempered) and Yadkin (crushed rock- and granule sand-tempered) series. As a means of preserving data to test the “hardness trajectory” hypothesis, these hard examples were separated into new series variants distinguished by the Roman numeral suffix “III.” Thus, extremely hard variants of the series are respectively referred to here as Hanover III, Cape Fear III and Yadkin III. If the efficacy of the hypothesis is later confirmed, then it would be appropriate to assign new series names to the Cape Fear III and Yadkin III variants since these series are generally viewed as Middle Woodland in affiliation. New River, Hanover, Yadkin and Cape Fear series are represented in the Phase II collection (Table 131). Descriptions of each series are presented below. Metric and attribute data on precontact ceramics are contained in Appendix I. New River Series New River series ceramics are associated with the Early Woodland period in the region. The series was originally defined by Loftfield (1976) who characterized it as a coarse sand-tempered pottery of compact composition and gritty consistency characterized by cord marked, fabric impressed, thong simple stamped and plain exterior surface treatments. Correlated series of the Early Woodland period Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 424 Chapter 9. 31RH480 Table 131. Precontact Ceramic Sherd and Vessel Representation, 31RH480. CERAMIC SERIES SHERD AND VESSEL REPRESENTATION: 31RH480 CERAMIC SERIES SHERD COUNT VESSELS REPRESENTED New River I 3 2 New River II 7 4 Hanover II 99 5 Hanover III 15 7 Yadkin III 10 8 Cape Fear III 1 1 GRAND TOTAL 135 27 include Lenoir (Crawford 1966) and Deep Creek (Phelps 1983). The New River series in the Fort Bragg area is poorly understood. Herbert (2003:156) indicates that both net-impressed and parallel/overstamped cord marked types are recognized, while Cable and Cantley (2005a, b) have also recognized simple stamped, fabric impressed and plain surface treatments. The predominant classificatory criterion used to identify the series is an abundance of quartz sand temper as Loftfield (1976) prescribed. Herbert (2003:156) proposes that an arbitrary cut-off at about 15% sand temper density should be used to distinguish Cape Fear (< 15%) and New River (>15%) in the Fort Bragg area. Cable and Cantley (2005a, b) have recognized two major paste variants in Fort Bragg assemblages. These are variants Ia and IIa. Variant Ia consists of sherds dominated by abundant medium to very coarse subangular to rounded quartz sand grains in a gritty, friable paste matrix. Variant IIa contains a fine, compact, relatively harder paste matrix with abundant to moderately dense medium to very coarse quartz sand inclusions. Variant IIa may correlate with the traditionally defined Cape Fear series. Feldspar sand is sometimes found in New River series sherds, and it is designated as New River If or IIf. Ten New River series sherds were recognized in the collection from 31RH480. The New River Ia sample consists of three sherds, one fabric impressed specimen (V24) and two indeterminate exterior surface treatments (V12). The weave pattern of the fabric impressed sherds was not identifiable. The IIa paste variant, which occurred in three sherds, was friable and contained abundant medium to pebble sized, subangular to rounded quartz sand. The New River IIa sample totaled seven sherds, including six fabric impressed sherds and one indeterminate sherd (Figure 146:A-C). Four vessels are represented in this sample. Rigid warps and medium wefts characterized all of the recognizable weave patterns (n=5) on the fabric impressed sherds. Quartz sand temper inclusions were less coarse than those found in the New River Ia sherds. Most specimens were characterized by compact paste with abundant medium to very coarse sand. Two sherds also contained moderate amounts of granular subangular quartz sand and one specimen contained only moderate amounts of medium sand. Rim sherds were not present in the sample. Vessel wall thickness for the New River Ia sample ranged from 6.62 to 8.88 mm with a mean of 7.66 mm (SD=1.14 mm), while the New River Ia sample ranged between 5.09 and 7.88 mm with a mean of 6.38 mm (SD=0.89 mm). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 425 Chapter 9. 31RH480 Hanover Series South (1960:16-17; 1976: 28) defined the Hanover series from surface collections made in 1960 on sites located between Wilmington, NC and Myrtle Beach, SC. Pottery affiliated with this series is tempered with aplastic clay lumps, most of which consist of grog fragments from crushed sherds. A correlate on the north coastal plain of North Carolina is referred to as the Carteret series (Loftfield 1979:154157). Cord marked and fabric impressed exterior surface treatments were included in the original series description, but other grog-tempered types have been recognized subsequently. Herbert (2003:75) has identified minority proportions of check stamped, plain (smoothed), net impressed, simple stamped, punctate, and random straw bundle punctate surface treatments. The latter two types are traditionally subsumed under the Refuge series on the central South Carolina coast where grog/clay temper occurs as a major temper variant (see Cable 2002: 195-201). A great deal of variability in paste composition has been noted in Hanover assemblages, primarily related to differences in quartz sand contributions, clay versus grog distinctions, and the density and size of clay/grog particles (see Cable et al. 1998; Herbert 2003; 74-75). Herbert (2003:191-192) recognizes two broad variants that he believes are sequential. The earliest variant, Hanover I, is primarily sandtempered with minor amounts of finely crushed grog. Surface treatments attributed to Hanover I pottery consist of cord marking, fabric impressing and check stamping. He estimates an age range of AD 400-800 for Hanover I pottery. The later variant, Hanover II, is inferred to correlate with a Late Woodland temporal range (AD 800 through 1500). Hanover II pottery is characterized by pastes with only small amounts of sand and abundant grog particles. The dominant exterior surface treatment of the later variant is fabric impressing. Cable and Cantley (2005a, b) follow Herbert (2003) in describing Hanover series ceramics in the Fort Bragg area. However, in an attempt to more fully capture the temper and paste variation, these basic groupings have been expanded to include sub-variants. As research has progressed, the number of subvariants has increased to accommodate an expanding range of temper constituents and paste qualities. There are three major paste variants found in Hanover pottery at Fort Bragg based on relative hardness. Hanover I is characterized by a compact but friable paste, Hanover II exhibits a harder, more compact paste and Hanover III contains a hard to vitrified paste. Within each of these relative hardness groupings are two primary temper constituent designations. Designation “a” refers to pastes with sparse to moderate amounts of fine to medium quartz sand and abundant medium to large grog particles. Designation “b” includes pastes with moderate to abundant amounts of medium to coarse quartz sand and sparse to moderate densities of small to medium particles. Other temper constituents are sometimes added to these basic paste variants. These include crushed arenite (“c”), granule to pebble sized angular to subangular quartz sand and, rarely, crushed quartz (“d”), crushed and ground indurated granite (“e”) and medium to granule sized subangular feldspar sand (“f”). Crushed arenite and indurated granite are strongly associated with Hanover III paste. Only Hanover II and Hanover III subseries were identified in the 31RH480 collection (Table 131). The Hanover II sample consisted of 99 sherds from four vessels, 95 sherds of which belonged to a single vessel. The Hanover III sample was represented by 15 sherds belonging to seven vessels. Each subseries is described below. (1) Hanover II. The Hanover II sample from 31RH480 contained only fabric impressed sherds, which is starkly different than the Hanover II sample from the Quewhiffle site (31HK2521) where check stamped and cord marked surface treatments made up nearly 73 percent of the recognizable Hanover II exterior surface treatments. The sample from 31RH480, however, is very limited, made up of only five vessels of which only three could be assigned surface treatment types. Vessel 1 was a Hanover IIa Fabric Impressed open mouth jar represented by 95 sherds (Figure 146:D-I). A good portion of the vessel was recovered, producing a combined weight of 264.67 gm. Eight small to medium sized rim sherds from the vessel were recovered and their profiles (Figure 147) indicate a straight-sided open moth jar with a slightly recurved wall. The lip was flat and stamped with fabric impressions. Bi-conical mend holes or strap holes were identified on two large sherds pictured in Figure 146 (G-I). The weave pattern appears to Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 426 Chapter 9. 31RH480 Figure 146. New River II and Hanover II Series Ceramics, Site 31RH480. A-C: New River IIa Fabric Impressed (p148, p919 and p435 respectively), D-I: Hanover IIa Fabric Impressed (p650, p42-Exterior, p42-Interior, p645, p644-Exterior and p644-Interior respectively). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 427 Chapter 9. 31RH480 Hanover II Yadkin III Cape Fear III p13 p433 Direct Thinned p650 Simple Direct p42 Simple Direct Figure 147. Rim Profiles, Site 31RH480. consist of a rigid warp and a medium weft. The interior (Figure 146:I) was well smoothed and displayed a very regular contour. The dark coloration appears to have been made during the firing process. The interior of one of the rim sherds was not smoothed over, leaving an unfinished, thickened on a portion of the lip interior (Figure 146:F). Weave patterns on the other two fabric impressed vessels, each represented by a single sherd, could not be determined. Wall thickness for Vessel 1 ranged between 4.22 and 6.90 mm, one outlier sherd had a thickness of 9.48 mm, with a mean of 5.69 mm (SD=0.76). The relatively thin character of these sherds suggests that most of the sample came from the upper part of the jar near its mouth. The background paste of the vessel was very fine, consisting of abundant, well-sorted fine sand. Large pieces of sherd temper were abundantly and evenly distributed in the paste matrix. Wall thickness of the remaining four sherds ranged between 4.47 and 7.17 mm, with a mean of 6.30 mm (SD=1.24). This group consisted of three Hanover IIa and one Hanover IIb paste variants. The paste of the former variant was similar in structure to that of Vessel 1, while the latter had slightly coarser background sand and smaller, moderately dense grog temper particles. (2) Hanover III. Fifteen sherds assignable to the Hanover III subseries were identified in the analysis, representing fragments of seven different vessels (Table 131). Fabric impressed sherds made up 73.3 percent of the sample (n=11) of the sample, while the remaining sherds were cord marked (26.7 percent, n=4). Fabric impressing was present on 85.7 percent (n=6) of the identifiable vessels, while cord marking was represented on only 14.3 percent (n=1) of the vessel assemblage. Of the fabric impressed sherds (Figure 148:A-B) with identifiable weave patterns (n=9) all had rigid warps, while five displayed fine wefts and the remaining four sherds had medium wefts. The cord marked sherds all belonged to a single vessel (V2). They exhibited a carefully applied perpendicular cross stamp pattern. Cord widths averaged 1.75 mm in diameter (SD=0.30) and ranged between 1.38 to 2.04 mm. Paste variation in the sample was confined to two variants, IIIa and IIIb. The latter paste dominates, representing 86.7 percent of the assemblage. Variant IIb is characterized by moderate to abundant quartz sand of predominantly medium to coarse grain size and sparse to moderately dense grog of small to medium size. Variant IIa consists of sparse quartz sand of medium to coarse grain size and with abundant medium to large sized grog particles. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 428 Chapter 9. 31RH480 Figure 148. Hanover III and Yadkin III Series Ceramics, Site 31RH480. A-B: Hanover IIIb Fabric Impressed (p432 and p448 respectively), C: Hanover IIIa Cord Marked (p915), D-E: Hanover IIIb Cord Marked (p611 and p595 respectively), F: Yadkin IIId Fabric Impressed (p97), G-I: Yadkin IIId Cord Marked (p433, p434 and p122). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 429 Chapter 9. 31RH480 Wall thickness for the Hanover III sample ranged between 5.34 and 8.46 mm, with a mean of 7.02 mm (SD=0.79). Rim sherds were not recovered for the sample. A single Hanover IIb Cord Marked sherd displayed a small portion of a mend hole/handle hole on its edge margin. Interiors were variably finished. Some were carefully smoothed with even contours while others were irregular. Yadkin Series Coe (1964:30-32) defined Yadkin series pottery from his excavations at the Doershuck Site. Abundant, angular crushed quartz was the diagnostic trait of the series, which was characterized by three main surface treatments, fabric impressed, cord marked, and linear check stamped. Subsequent investigations at Town Creek (Coe 1995) also associated simple stamped exteriors with the series. In the North Carolina Sandhills, the definition of Yadkin has been expanded to include all types of crushed rock temper (Herbert 2003:58–59). Currently, examples of crushed granite, crushed arenite, crushed and ground sandstone/siltstone, granule sized pieces of crushed and ground indurated granite, subangular feldspar sand and granule/pebble subangular quartz sand have been identified in assemblages in the region. The results of the Quewhiffle site ceramic analysis suggests that Yadkin III and Hanover III are at least partially contemporaneous. The presence of large pieces of crushed and ground sandstone/siltstone conglomerate particles in sherd collections within Fort Bragg was first recognized during Phase II testing at 31MR326 and 31HT748 conducted by Palmetto Research (Cable and Cantley 2005b). Due to its texture and composition, it can be confused with grog and probably has been on occasion in the Fort Bragg area of the Sandhills. The material is light tan to grayish yellow when viewed in sherds and generally contains fine to medium grains of arenite that can become finely disseminated in the paste matrix. The two materials are found in the same geological deposits and large particles of each can co-occur in individual sherds. Significant quantities of the sandstone tempered sherds at 31MR326 contained grog temper. The Yadkin material from 31RH480 exhibited hard to vitrified pastes; consequently, all of it was assigned to the Yadkin III subseries. Ten Yakin III sherds were identified in the collection, representing fragments of eight separate vessels. Exterior surface treatments on individual sherds number five cord marked, two fabric impressed and one plain or smoothed. Relative exterior surface treatment proportions controlled by vessel data are: 50 percent cord marked, 33 percent fabric impressed and 17 percent plain/smoothed. Recognizable stamp patterns on cord marked sherds are evenly divided between parallel (Figure 148:G-H) and oblique cross-stamped (Figure 148:I). Cord impression range between 0.67 and 1.88 mm, averaging 1.42 mm (SD=0.55). Only one of the fabric impressed sherds (Figure 148:F) exhibited a recognizable weave pattern. It displayed a rigid warp and a medium weft. A single rim sherd (Yadkin IIId Cord Marked) was recovered in the assemblage (Figure 147 and Figure 148:G). Its form can be described as “direct thinned,” which consists of direct rims that have been thinned on the interior contour beginning at a distance of about 8 to 10 mm below the lip. This gives a slightly flaring appearance to the lip termination, but the vessel walls are straight-sided to very slightly recurved. In the Quewhiffle site (38HK2521) rim assemblage most of these rims are thick and appear to represent large open mouth jars. This appears to be the case with the example from 31RRH480 as well. This rim form predominates in both the Hanover III and Yadkin III samples from the Quewhiffle site. Three paste variants were recognized. Variant IIId was most numerous and consisted of granule sized crushed white quartz and/or granule to pebble sized subangular quartz sand in a hard to vitrified matrix containing medium to coarse sand. Variant IIIf contained medium to granule sized subangular quartz and feldspar sand. Finally, variant IIId/e contained abundant granule sized pieces of crushed white quartz and moderate amounts of granule sized pieces of crushed and ground indurated granite. Vessel wall thickness for the Yadkin III sample was 7.17 mm (SD=0.63 mm), with a range of 6.34 and 8.42 mm. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 430 Chapter 9. 31RH480 Cape Fear Series Stanley South (1976:18-20) first described the Cape Fear series, or ware-group as he described it, from surface collections made in 1960 on sites located between Wilmington, NC and Myrtle Beach, SC. He applied the term to virtually all of the sand-tempered cord marked, fabric impressed, and net impressed pottery within the collection. Many have suggested that the series was too broadly applied by South and other investigations in North Carolina indicate that such material should be separated into a sequence of series based on sand inclusion coarseness (see Anderson and Logan 1981:107-108; Trinkley 1981:11). Coastal North Carolina sequences contain at least two sand grain size modes that are used to distinguish sand-tempered series (Phelps 1983). The Deep Creek series consists of pastes with abundant, coarse sand and it is chronologically correlated with Deptford. Other series (i.e. Mount Pleasant and Cape Fear) contain finer grained pastes and they are roughly dated to a post-Deptford or late Middle to Late Woodland context (i.e. AD 500-900 or 1500 to 1100 BP). Herbert (2003:156) distinguishes the sandtempered New River and Cape Fear series by quartz sand density in the North Carolina Sandhills. Cape Fear series material is characterized by low to moderate densities of quartz sand amounting to less than 15% of the paste body. In addition, he places all perpendicular stamped cord marked exterior surfaces in the Cape Fear series, while all net-impressed is assigned arbitrarily to New River. Cable and Cantley (2005a, b) have distinguished two series variants for Cape Fear in the Fort Bragg area based on quartz sand density and paste characteristics. Cape Fear I was characterized by a friable, gritty paste and a moderate density of medium to coarse quartz sand temper. Cape Fear II exhibited a finer, more compact paste with sparse medium to coarse quartz sand temper. Only a single sherd of Cape Fear series was identified in the collection. Just as was true of the Quewhiffle site sample, the sherd possessed an extremely hard to vitrified paste. Consequently, it was assigned to a third Cape Fear subseries, Cape Fear III, which is assumed to belong to the Late Woodland period. The paste was composed of abundant fine to medium quartz sand with apparently few coarser grained inclusions. Cape Fear III series paste resembles most closely the paste varieties that characterize Mississippian series pottery as it is represented in the upper Pee Dee River Valley (Town Creek), the upper Wateree River Valley (Mulberry-Adamson) and the lower Santee River Valley (Jeremy-Tibwin). Whether there is a direct correlation with neighboring Mississippian cultures is an issue that cannot be resolved at present. The exterior surface treatment on the sherd, which was a small rim, was not identifiable, but it appears be either fabric impressed or cord marked. The rim can be described as simple direct (Figure 147) and its thinness suggests that it may represent a fragment of a hemispherical bowl. Vessel wall thickness at a location 10 mm below the lip is 5.27 mm. The lip is flat and undecorated. Concluding Remarks The ceramic assemblage recovered from the 31RH480 evidences Early, Middle and Late Woodland occupation. Vessel representation suggests that about 60 percent of the occupation occurred during the Late Woodland period. Hanover II and New River II series comprise about 33 percent of the vessel population between them and this may represent the Middle Woodland contribution to the site occupation history. The Early Woodland affiliated New River I series accounts for about 7 percent of the vessel sample. Historic Artifacts Four historic artifacts were recovered during the investigation. All of this material appears to be related to military training exercises. The artifact inventory consists of three bullet casings found in shovel tests and a seemed Coca-Cola can that was recovered from an historic trash pit intersected by TU4. OCCUPATION PATTERNS Diagnostic stone tool markers of culturechronological association were rarely recovered at 31RH480. Unlike most sites in the interior uplands of Fort Bragg, however, this site produced evidence of a moderately extensive Woodland period occu- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 431 Chapter 9. 31RH480 pation as reflected by precontact ceramics. Only a single diagnostic projectile point was recovered, a Palmer II Corner Notched point affiliated with the Early Archaic period. Several unifacial tools also suggest the presence of Early Archaic associations, but diagnostic evidence of Middle and Late Archaic occupation is lacking. In general, however, moderate to large precontact sites in the Sandhills that contain large quantities of lithic artifacts contain a wide range of Archaic occupations. Ceramic analysis identified the presence of Early Woodland (New River I), early Middle Woodland (New River II), late Middle Woodland (Hanover II) and early Late Woodland (Hanover III/Yadkin III) components. In addition, the Cape Fear series, which was present in small quantities, is interpreted as evidence of a temporary intrusion of transitional Mississippian seasonal occupations into the region. 31RH480 is characterized by moderately high artifact density. Mean artifact density in positive Stage I and II shovel tests (n=78) was 3.45 artifacts (SD = 3.34), indicating that two-thirds of the positive tests yielded densities between one and seven artifacts. Although the modal positive shovel test outcome across the sample was only one artifact, 54 percent of the positive tests contained three or more artifacts. The entire site area (including negative Stage I and II shovel tests) had a mean artifact density of 2.61 artifacts (SD = 3.26) per shovel test. These values are comparable to artifact densities at 31HK2510 and about half of that seen at 31HK2521. Metavolcanic debitage recovered during Stage I and II shovel testing was primarily concentrated on the nose of the ridge overlooking the confluence of the unnamed tributary and a smaller drainage on the southwestern border of the site (Figure 149). The densest clusters occurred in an area measuring 20 x 40 m between N510 and NN470 and E500 and E520. Quartz debitage densities were much lower and although its distribution was also centered on the ridge nose, the main concentration was located slightly south of the metavolcanic concentration (Figure 150). Vertical distributions by raw material type indicate that debitage is centered between 30 and 50 cm bs (Table 132). From what we have learned about the vertical distributions of cultural-chronological occupations on sites in the Sandhills (see Cable 2010; Cable and Cantley 2005a, b, 2006), this distri- bution should indicate that most of the lithic debris is associated with the Archaic period and predominantly with the Middle and Early Archaic. Ceramic vertical distributions indicate that the Woodland period occupation at the site is positioned in the upper 30 cm of deposit (Table 133). Sherds below 30 cm bs are “out of position” and were likely displaced by natural disturbances, precontact feature construction, modern land use activities or recovery error. Approximately one-quarter of the lithic debitage is situated in the upper 30 cm of deposit. Although it is not likely that all of this material is associated with Woodland period occupation, it is reasonable to suggest that as much as 10 percent of the lithic debris on the site is associated with Woodland occupation. Horizontal ceramic distributions (Figures 151-153) indicate that Woodland occupations are concentrated on the ridge toe in discrete, spatially limited clusters. These data include shovel tests from all stages to derive a relatively complete picture of Woodland occupation. Hanover II and Yadkin III series are the most extensively distributed and the limited size of the clusters strongly argues that they represent separate reoccupations of the site by single households. However, the three Hanover III clusters are concentrated at a single location defined by shovel test sampling in SBs 7 and 8. This tight clustering may indicate a multiple household occupation (Figure 153). One of the Yadkin III clusters was also found at this location, which may support the hypothesis that the Yadkin III and Hanover III series are, at least, partially contemporaneous. The New River series clusters also appear to represent reoccupations by single households, although each series is concentrated in a specific area of the site. This may suggest that campsite selection during reoccupation was conditioned by previous camp locations. A more detailed appreciation for precontact land-use patterns at the site can be derived from a consideration of occupation type distributions and associations. Currently, four occupation types based on chipped stone criteria have been recognized on sites in the Fort Bragg area (see Cable and Cantley 2005c, 2006): Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 432 Chapter 9. 31RH480 Metavolcanic Debitage (Contours=1 Item) 97.0 560 96.5 550 Beaver Pond 96.0 540 il Tra 31RH480 530 95.5 520 510 95.0 Dirt R 500 oad 490 94.5 480 470 94.0 460 450 93.5 440 430 Positive ST Negative ST 420 410 N 400 0 40 m 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 149. Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31RH480 (Contours = 1 piece of debitage). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 433 Chapter 9. 31RH480 Quartz Debitage (Contours=1 Item) 97.0 560 96.5 550 Beaver Pond 96.0 540 il Tra 31RH480 530 95.5 520 510 95.0 Dirt R 500 oad 490 94.5 480 470 94.0 460 450 93.5 440 430 Positive ST Negative ST 420 410 N 400 0 40 m 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 150. Density Distribution of Quartz Debitage, Stage I and II Shovel Test Sample, Site 31RH480 (Contours = 1 piece of debitage). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 434 Chapter 9. 31RH480 Table 132. Vertical Distribution of Chipped Stone Artifacts, Stage I and II Shovel Tests, 31RH480. VERTICAL DISTRIBUTION OF DEBITAGE BY RAW MATERIAL TYPE, STAGE I AND II SAMPLES: 31RH480 RAW MATERIAL TYPE Metavolcanic Hydrated Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Quartz Gray Quartz White Quartz Yellow Quartz Metasedimentary Black Metasedimentary Gray/Green Metasedimentary Minor Types Quartzite Purple Tan Mottled Chert GRAND TOTAL LEVEL 1 2 3 1 1 3 1 4 5 1 7 4 8 11 14 17 1 4 6 5 5 2 3 6 7 1 3 2 15 8 12 4 2 5 6 7 8 4 1 4 9 1 1 1 1 2 3 1 9 GRAND TOTAL 3 25 7 34 42 48 1 1 1 1 30 1 1 1 1 1 1 1 7 20 24 1 65 47 20 4 7 1 1 195 1 Table 133. Vertical Distribution of Precontact Ceramics, All Proveniences, 31RH480. VERTICAL DISTRIBUTION OF PRECONTACT CERAMICS, ALL SHOVEL TESTS: 31RH480 CERAMIC SERIES LEVEL 1 2 New River I New River II 2 Hanover II 3 2 1 3 1 6 5 Hanover III 5 4 Yadkin III 1 6 4 8 21 1 3 1 7 14 1 3 10 GRAND TOTAL 3 11 10 Cape Fear III GRAND TOTAL 5 4 1 1 3 46 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 435 Chapter 9. 31RH480 Contours=1 Sherd New River I 97.0 560 New River II 96.5 550 Beaver Pond 96.0 540 il Tra 31RH480 530 95.5 520 510 95.0 Dirt R 500 oad 490 94.5 480 470 94.0 460 450 93.5 440 430 Positive ST Negative ST 420 410 N 400 0 40 m 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 151. Density Distribution of New River Series Sherds, All Shovel Tests, Site 31RH480 (Contours = 1 sherd). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 436 Chapter 9. 31RH480 Contours=1 Sherd Hanover II 97.0 560 96.5 550 Beaver Pond 96.0 540 il Tra 31RH480 530 95.5 520 510 95.0 Dirt R 500 oad 490 94.5 480 470 94.0 460 450 93.5 440 430 Positive ST Negative ST 420 410 N 400 0 40 m 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 152. Density Distribution of Hanover II Series Sherds, All Shovel Tests, Site 31RH480 (Contours = 1 sherd. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 437 Chapter 9. 31RH480 Contours=1 Sherd Hanover III Yadkin III 97.0 560 96.5 Cape Fear III 550 Beaver Pond 96.0 540 il Tra 31RH480 530 95.5 520 510 95.0 Dirt R 500 oad 490 94.5 480 470 94.0 460 450 93.5 440 430 Positive ST Negative ST 420 410 N 400 0 40 m 390 400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550 560 Figure 153. Density Distribution of Hanover III, Yadkin III and Cape Fear III Series Sherds, All Shovel Tests, Site 31RH480 (Contours = 1 sherd). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 438 Chapter 9. 31RH480 (1) Type I occupations consist of a debitage concentration and an associated tool cluster situated at a confined location along its periphery. It is inferred that the tool cluster was formed around a hearth and in several instances calcined bone fragments have been recovered from the deposit delimited by a tool cluster. It is also believed that ephemeral shelters were situated adjacent to these tool clusters and opposite the concentrations. Debitage concentrations appear to represent lithic reduction loci established away from a shelter or sleeping area where tools were manufactured during the stay at the camp. These concentrations are generally composed of single lithic raw material types, and the associated tool clusters commonly consist of manufacturing rejects and discarded broken or worn-out tools made of the same raw material that comprises the associated debitage concentration. Although many Type I residences occur in isolation and appear to be the by-products from single nuclear or small extended family occupations, others appear to represent multi-family occupations that extend beyond the ability of a single shovel test to identify. Woodland period Type I occupations are characterized by an additional element, partial vessel sherd aggregations. These occur adjacent to debitage concentrations and they are hypothesized to occupy a portion of the living floor associated with a hut. The model would stipulate that tool clusters are situated between the debitage concentrations and sherd aggregations. (2) Type II residences are similar in spatial organization to Type I residences, but they exhibit much higher tool-to-debitage ratios and may commonly be comprised of diverse lithic raw materials. Lithic reduction on these sites consisted of tool maintenance and late stage shaping of tools from flake blanks or reduced biface cores. Consequently, debitage density is much lower than that of Type I debitage concentrations and the scatters consist primarily of late stage reduction debitage. Type II occupations are inferred to be the output from single or small, multiple household aggregations and are thought to represent the residue from High Technology Forager (HTF) residences (Spiess 1984; Todd 1983:231-233). The HTF model proposes a specialized forager adaptation combining high logistical and high residential mobility into a single settlement system characterized by highly curated technologies. Type II residences have thus far only been assigned to the Early Archaic period at Fort Bragg. The lithic raw material diversity characteristic of Type II residences suggests that these occupations represent the aggregation of social units that may have been seasonally dispersed, transporting a wide range of lithic raw materials from different sources. (3) Type III occupations are inferred to represent logistical camps and very short-term residences characterized by low-density debitage scatters and sporadic tool discard. (4) Type IV occupations represent extraction or processing loci that are not believed to have involved, necessarily, overnight stays. They are difficult, however, to distinguish from Type III occupations, even in situations of greater excavation exposure. Isolated sherd aggregations are hypothesized to represent one kind of Type IV element associated with Woodland occupations. It is difficult to distinguish effectively between these various occupation types at shovel test intervals of greater than 1.25 m, but a less precise classification can be used at greater shovel test intervals to aid in population projections. High-density debitage concentrations tend to indicate Type I residences, while low-density scatters tend to signal short-term residences, special purpose camps and locations (i.e. Type II, III and IV occupations). Highdensity debitage concentrations are temporarily stipulated to correspond to shovel test outcomes of five or more pieces of debitage of a specific raw material subtype. Eventually, it will be necessary to confirm this arbitrary threshold by examining shovel test counts from a large sample of known high-density debitage concentrations. Deposits containing less than five pieces of debitage are viewed as more likely to represent low-density debitage scatters associated with the other three site types. Since shovel tests can intersect the peripheral zones of high-density debitage concentrations where debitage densities are much lower, however, it is likely that Type I residential occupations will be under-represented in macrointerval shovel test data. Because it is difficult to determine whether a single occupation is represented in more than one shovel test, in the cases of multi-family residences or associated special activity zones on the periphery Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 439 Chapter 9. 31RH480 of a camp, the term “element” is preferred over the term “occupation” to refer to the occupational data in shovel tests. At a sampling interval of 5-m or 10-m intervals, we are assured that the presence of an element will only rarely be repeated in adjacent shovel tests. Elements are identified by lithic raw material subtypes and/or by precontact sherd aggregations associated with one or more vessels. This methodology runs the risk of overestimating occupations because more than one raw material subtype may comprise a single occupation. The “element” concept recognizes this fact, focusing on depositional events and episodes rather than entire occupations. Elements can represent segments of contemporaneous multi-family occupations, separate occupations of various function, or contemporary members of multiple raw material reduction episodes. The current methodology for identifying elements in shovel tests consists of running crosstabs of lithic debitage raw material subtypes or subtype groups by shovel test number in a spreadsheet. Sherd aggregations are not used in these calculations because, in general, they are associated with debitage scatters and their inclusion would result in potentially double counting a good number of elements. Once the cross-tab is generated, the sum of each lithic raw material subtype in a shovel test is identified as an element. In some instances, the vertical separation between individual items of the same lithic subtype can be so great as to suggest the presence of distinct elements. Due to the vagaries of bioturbation and possible recovery errors in which higher positioned material can be included in lower levels through sidewall displacement, however, the lower item(s) is assigned to the higher positioned item(s) to define a single occupation element. In cases where large debitage concentrations of the same lithic raw material type are found at distinctively separated levels, though, each is counted as separate elements. Elements have horizontal dimensions and, as a result, sample units that are smaller than those dimensions have the benefit of increased sample efficiency (see Cable and Donaldson 1988; Rice 1987; Rice and Plog 1983). The typical element found at Fort Bragg has a diameter of about 3 m. Thus, a shovel test of .09 m2 samples only .09 percent of the area of a 10 m-square, but it samples 9.0 percent of the theoretical mean element space of 9.0 m2. Con- sequently, a 10 m-interval shovel test pattern actually samples about 9.0 percent of the available element space at a site, while a 5 m-interval shovel test pattern samples about 36.0 percent of the available element space. A mean estimate of the population of elements at shovel test intervals of 10 m and 5 m, then, requires a simple arithmetic calculation of counting the occupation elements identified in shovel tests and dividing this figure respectively by 9.0 percent and 36.0 percent. Confidence intervals can be constructed around these means, but they have not been calculated in the Fort Bragg testing projects. Although Stage I shovel test samples have been relied upon in the past to produce population projections because of their even and complete coverage, a concerted effort was made in the DO5 package to extend 5 m-interval shovel tests across the entire site area so that advantage could be taken of greater sample density. Unfortunately, the Stage II sample only partly covered the site area of 31RH480. Consequently, only the Stage I shovel tests could be confidently used to make population projections. Following the procedures outlined above, the method succeeded in identifying one Type I element and 31 indeterminate (Type II, III and IV) elements in the Stage I sample (Table 134). Elements are relatively evenly distributed between UER, USR (1), USR (2) and white quartz, which range between 18.2 and 27.3 percent of the sample each. MMR and Type I RT elements respectively represent 6.1 and 9.1 percent of the sample. Based on these sample results it can be further estimated that the site contains 367 elements, of which 11 represent Type I elements. Only two chipped stone tools were recovered in the Stage I sample, a white quartz flake blank and a utilized flake made of USR (1) material. Only the flake blank was found in isolation from debitage and it was included in the element count. The proportional representation of Type I residential elements (3.0 percent) is quite low, suggesting that the predominant precontact land use history of the site was short-term camps of single households or specially comprised task groups of less than a weeks duration. A large segment of the Fort Bragg site assemblage has been demonstrated to contain as much as 20 to 30 percent Type I residences (see Cable 2010). Expanded over nearly the full range of human occupation (about 10,000 years), the population esti- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 440 Chapter 9. 31RH480 Table 134. Recognized Elements for the Stage I and II Shovel Test Sample, 31RH480. IDENTIFIED ELEMENTS FROM THE STAGE I SHOVEL TEST SAMPLE: 31RH480 RAW MATERIAL TYPES TYPE I TYPE II, III, IV GRAND TOTAL PERCENTAGE OF ELEMENTS Mill Mountain Rhyolite 2 2 6.3 Type I Rhyolite Tuff 3 3 9.4 Uwharries Eastern Rhyolite 6 6 18.8 Uwharries Southern Rhyolite (1) 7 7 21.9 Uwharries Southern Rhyolite (2) 6 6 18.8 White Quartz 1 7 8 25.0 GRAND TOTAL 1 31 32 100.0 mate indicates that precontact groups, on average, visited the site a maximum of once every 27.2 years. The occupation, however, was probably not evenly distributed through time. The site was no doubt used intensively by specific culture-historic groups, probably for strings of continuous years punctuated by abandonments in response to long-term land-use patterns dependent upon resource depletion cycles. Due to occupation concentrating on the ridge nose at 31RH480, a fairly high incidence of element superimposition would be expected. However, mean superimposition in the Stage I sample was only 1.60 (SD=0.68) elements per shovel test. This is lower than the incidence of superimposition at 31HK2502 (2.24) and 31HK2521 (2.07), which were located on even more restricted land forms. In fact, the 31RH480 value is in line with 31HK2510 (1.62), which was situated on a more generic terrace edge with extensive stretches of flat topography. In the case of 31RH480, it would appear that the low incidence of superimposition is due to the dominance of short-term occupations by single households and specially comprised task groups. Not included in the population estimate was the Woodland period ceramic collection. Only one sherd was recovered from the Stage I sample, indicating that the ceramic deposit at the site is relatively light. Concentrating the Stage II shovel test sample on the ridge nose produced a much larger sample of precontact ceramics. In the combined Stage I and II shovel test sample, 15 individual vessels were recovered. Because ceramics appear to be almost entirely contained on the ridge nose, this expanded sample represents the best data set with which to evaluate Woodland occupation at the site. The standard approach to this problem has been to calculate vessel populations using the aggregated model of vessel deposit organization. We have seen in previous analyses of the ceramic assemblages from sites in the DO5 package, however, that the ceramic deposit appears to be much more disorganized and displaced than the aggregated ceramic model of vessel population estimate would assume. That model stipulates that we can expect most sherds from single vessels to be contained within an area of about 2 m in diameter. Under this assumption the Stage I and II shovel tests would be expected to achieve 16 percent coverage for sherd aggregate space. In the case of 31RH480, the aggregated model would yield a total vessel population estimate of 94. There is good reason to believe, however, that this method of calculation underestimates the actual vessel population, as was discussed in depth in Chapter 5. It was also concluded from that discussion that it is unlikely that many full pot concen- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 441 Chapter 9. 31RH480 trations survived on these sites due to recycling of large sherds and off-site transport of the larger vessel fragments. An idea of the magnitude of the sherd population at the site can be supplied by calculating a straight area mean. The sample of 117 Stage I and II shovel tests specifically located on the ridge nose sums to an area of 10.53 m2. The ride nose covers an area of approximately 2,900 m2, which yields an area sample of 0.36 percent. Twenty-three sherds, weighing 101.97 gm, were recovered from the sample, resulting in a population mean estimate of 6,389 sherds and a total sherd weight of 28.3 kg. If whole pots were entirely represented in the assemblage, it is doubtful that this total would be sufficient to cover the weight of more than 21 typical open-mouth jars from the region (see Herbert and Irwin 2003:2). Given an average weight of 3 pounds (1,377 gm) for a typical pot, if there were 94 whole pots present on the site, a mean population weight of 129 kg would be required. Consequently, only about 22 percent of the expected vessel weight relative to the estimate of the vessel population is present on the site. Similar results were obtained at 31HK2502, 31HK2510 and 31HK2521, where only about 10, 18 and 30 percent, respectively, of the expected sherd weight was found in the sample. This suggests that the intensity of use recycling is great on Sandhills sites and that large sherds from broken pots were probably recycled for other uses (see David and Hennig 1972:21-21) and stored in caches or transported upon leaving camps, while smaller sherds were probably left at the point of breakage. If so, sherds from the same pot might be spread across multiple sites and a high degree of displacement and disorganization should be evident in the spatial distribution of ceramics on sites with short-term occupation records. Other methods of calculating vessel populations at 31HK2521 resulted in values on the order of 4 to 20 times larger than those produced by the aggregated model. The Stage I and II shovel test sample succeeded in identifying New River I, New River II, Hanover II, Hanover III, Yadkin III and Cape Fear III series ceramics. Using the proportions of individual vessels associated with each of these sub-series (Table 135) it can be inferred that 6.7 percent of the Woodland occupation is affiliated with the Early Woodland period (New River I), 13.3 percent with the early Middle Woodland (New River II), 33.3 percent with the late Middle Woodland period (Hanover II) and 40 percent with the early Late Woodland (Hanover III and Yadkin III). Cape Fear III, which is postulated to represent an Early Mississippian intrusion into the area, comprised 6.7 percent of the vessel sample. Numerous studies have documented vertical patterning in the relative depth of deposits in the unconsolidated sandy soils of the Coastal Plain (Cable and Cantley 2005 b, 2006; Michie 1990). The actual depth patterns vary according to the character of the Table 135. Representation of Vessels by Sub-Series, Stage I and II Shovel Tests, 31RH480. REPRESENTATION OF VESSELS BY SERIES, STAGE I AND II SAMPLE: 31RH480 CERAMIC SERIES PERCENT REPRESENTATION New River I 1 6.7 New River II 2 13.3 Hanover II 5 33.3 Hanover III 1 6.7 Yadkin III 5 33.3 Cape Fear III 1 6.7 15 100.0 GRAND TOTAL VESSELS (n) Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 442 Chapter 9. 31RH480 deposit. Those deposits with higher B-horizons tend to have compressed sequences, while those in deep C-horizon and E-horizon sediments generally display more discreet and vertically expansive sequencing. Michie’s (1990) C-horizon model developed in the Waccamaw Neck area on the central coast of South Carolina begins with Mississippian and Woodland materials positioned in the upper 30 cm of sediment. This zone is underlain by ceramic Late Archaic occupations situated at 28 to 35 cm bs, Middle Archaic horizons at 35 to 55 cm bs, and Early Archaic corner-notched and side-notched components at 55 to 62 cm bs. This model is generally applicable to sites on Fort Bragg. However, the deposits at Fort Bragg appear to be somewhat more compressed, judging by the vertical positions of samples of diagnostic artifacts. At Fort Bragg three vertical groupings have been developed for deep Sandhills sites: (1) Woodland/Late Archaic/Middle Archaic between 15 and 40 cm bs, (2) Middle Archaic/Early Archaic between 35 and 50 cm bs, and (3) Early Archaic below 50 cm bs. The divisions are broadly constructed and realistically anticipate a certain degree of vertical mixing and overlap between the various periods. is non-diagnostic lithic chipped stone debitage. The vertical positions of the various elements recognized in the Stage I shovel test sample allow a rough picture of the representation of occupation periods at the site (Table 136). Most elements were situated in a single level, but some, especially the higher density elements, spanned several levels. In the latter cases, the central tendency of the vertical distribution of a particular element was used to estimate its position. Vertical data from the shovel test sample was not precisely congruent with the model ranges discussed above. However, the following adaptation was developed to correlate culture-historic association with level data: (1) 0-20 cm bs, Woodland, (2) 20-40 cm bs, Woodland/Late Archaic/Middle Archaic, (3) 40-50 cm bs, Middle Archaic/Early Archaic and (4) 50 to 80 cm bs, Early Archaic. Approximately 65.7 percent of the occupation elements are positioned below 40 cm bs. These depths are inferred to represent primarily Middle and Early Archaic occupation. Based on the recovery of a Palmer II Corner Notched point and several scrapers and a lack of recovery of Middle Archaic diagnostics, it is likely that most of the elements situated below 40 cm bs are affiliated with Early Archaic phases. About 34.4 percent of the elements are situated between 20 and 40 cm bs and are inferred to represent Middle and Late Archaic and Woodland components. The actual representation of This vertical model can be of great utility in reconstructing the occupation history of sites like 31RH480 where the predominant artifact class Table 136. Inferred Culture-Historic Associations of Recognized Lithic Elements, 31RH480. INFERRED ASSOCIATIONS OF RECOGNIZED LITHIC ELEMENTS, STAGE I SAMPLE: 31RH480 Woodland 1 20-40 1 Middle & Early Archaic 40-50 1 2 Early Archaic & Paleoindian 50-80 2 3 WHITE QUARTZ USR (2) USR (1) UER TYPE I RT 0-20 Woodland/Late & Middle Archaic GRAND TOTAL MMR INFERRED ASSOCIATION VERTICAL RANGE (cm bs) LITHIC ELEMENTS GRAND TOTAL PERCENT 1 2 6.3 3 1 3 1 9 28.1 3 4 3 5 18 56.3 1 3 9.4 6 7 6 8 32 100.0 1 1 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 443 Chapter 9. 31RH480 Woodland occupation is difficult to determine, but the earlier estimate of 10 percent may be close, although perhaps slightly underestimated. The single Type I element identified in the sample was composed of white quartz and was positioned between 40 and 50 cm bs, suggesting a Middle or Early Archaic association. (Table 138) suggest that the USR (1) element is vertically consistent with a Woodland occupation. The other raw material types are characterized by small sample sizes and erratic vertical distributions, which limits their value to make chronological inferences. The remaining sample blocks contain low frequencies of ceramics and exhibit the characteristics of disorganized deposits. SAMPLE BLOCK INVESTIGATIONS The high degree of ceramic deposit disorganization on sites in the DO5 package strongly indicates that methods for projecting vessel population emphasizing aggregated vessel completeness lead to significant underestimation of vessel populations. An alternative method, discussed earlier in Chapter 7, uses the area of the sample block as the basis for an areal sample fraction of vessels. Each sample block covers an area of 6.25 m2. The ceramics recovered during the investigation of 31RH480 are concentrated on the ridge nose, which covers an area of about 2,400 m2. The locations of sample blocks were not selected randomly, but they achieve a relatively even coverage of the ridge nose. Consequently, an argument can be made that together they yield a representative sample of the ridge nose area. Each of the sample blocks would constitute a 0.26 percent sample fraction of the ridge nose and together they achieve a 2.08 percent sample fraction. Seventeen vessels were found in the eight sample blocks. This would yield a mean vessel population projection of 817, which is over 8 times the mean estimate of 94 generated from the aggregated vessel model. Moreover, since it is reasonable to assume that not all vessels within each sample block were recovered (the shovel tests usually cover only about 13 percent of the block area), this projection should be viewed as a minimum vessel estimate. This may seem like an outrageously high estimate for a site where ceramics are not particularly dense, but over a span of about 1,000 years, the minimum span for the Woodland occupation based on ceramic types, this would yield a mean vessel deposition rate of only 0.817 vessels per year. Eight shovel test outcomes from the Stage I and II sample were targeted for close-interval shovel testing to further elaborate the character of the occupations extant at the site. In general, a series of eight additional shovel tests was excavated at 1.25 m-intervals surrounding the target shovel test. These are referred to as sample blocks (SBs) and their locations are illustrated in Figure 142. Investigation focused on the high-density area on the ridge nose. Tables 137 and 138 display the vertical distributions of debitage by raw material type for each of the sample blocks. Six of the sample blocks yielded vertical profiles centered between 20 and 60 cm bs in the Archaic zone of the deposit. Of this group, three (SB1, SB3 and SB7) were sufficiently deep (30 to 60 cm bs) to infer nearly pure Early Archaic occupation surfaces. The other two sample blocks (SB4 and SB8) primarily contained debitage in the upper 30 cm of deposit, which suggested a predominance of Woodland and/or later Archaic occupation. Ceramics were recovered from all but one of the sample blocks (Table 139). SB4 and SB8, as would be predicted by the high vertical debitage profiles, yielded the greatest sherd frequencies. The nine sherds in SB4 belonged to a single Hanover IIa Fabric Impressed vessel and were positioned between 10 and 40 cm bs. Reference to Table 137 indicates that the clearest vertical association with debitage would be with the white quartz element. The other raw material types in SB4 appear to be positioned lower in the matrix and probably represent Archaic elements. The SB8 ceramic deposit is disorganized, consisting of 12 sherds representing three ceramic series and six vessels. Most of the ceramics are positioned between 0 and 30 cm bs, with two-thirds occurring between 0 and 20 cm bs. Debitage vertical distributions in SB8 All of the sample blocks contain a wide variety of lithic raw material types. Much of this heterogeneity can be attributed to occupation superimposition, but some of it may result from site functional variability. Type II residences characteristically contain heterogeneous lithic raw material profiles and the high density of scraper forms in the stone tool inven- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 444 Chapter 9. 31RH480 Table 137. Vertical Distribution of Debitage, Sample Blocks 1—4, 31RH480. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 1-4:31RH480 SB1 RAW MATERIAL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL LEVEL 1 2 3 4 5 6 1 1 1 1 1 1 5 1 1 7 1 6 2 21 1 4 1 6 15 24 6 7 5 6 7 8-10 1 2 7 4 1 7 1 2 3 4 1 1 1 1 3 5 1 2 5 8 SB3 1 2 3 3 6 1 8-10 3 4 5 6 7 1 4 1 3 1 2 3 4 1 1 3 1 1 1 6 1 12 2 3 4 2 1 1 1 SB4 metavolcanic Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 3 4 2 GRAND TOTAL 2 1 2 6 14 25 LEVEL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL RAW MATERIAL GRAND TOTAL 5 1 17 4 41 1 69 LEVEL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) GRAND TOTAL RAW MATERIAL 8-10 1 1 SB2 RAW MATERIAL 7 4 1 11 5 1 12 2 1 5 6 7 8-10 1 GRAND TOTAL 12 7 3 2 16 2 42 LEVEL 1 1 3 4 3 2 9 2 1 4 tory suggests that many of the occupations at the site may represent Early Archaic Type II residences. The sample block frames generally cover an area of only about 2.5 m square, which is smaller than the average element size. Moreover, it would be uncommon for the sample block to expose nearly whole elements. However, a method was devised during the DO4 investigation (Cable 2010) to identify and distinguish 1 2 3 8 1 15 GRAND TOTAL 2 2 2 2 6 12 6 32 Type I and III elements that are mostly contained within the extent of a sample block. Three measures related to shovel test outcomes were developed to describe this variability (Table 140). These are: (1) number of positive shovel tests within which the element occurs, (2) number of shovel tests with greater than or equal to 5 pieces of debitage and (3) mean debitage frequency per positive test. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 445 Chapter 9. 31RH480 Table 138. Vertical Distribution of Debitage, Sample Blocks 5—8, 31RH480. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 5-8:31RH480 SB5 RAW MATERIAL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL LEVEL 1 2 3 4 5 6 7 1 10 2 1 6 1 8 1 6 3 1 1 3 3 17 1 1 4 1 1 1 7 8-10 1 4 1 3 1 Metasedimentary, gray/green Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 20 13 1 2 3 4 5 6 1 1 1 2 5 1 1 1 1 1 1 1 3 1 2 3 5 2 5 7 13 3 11 4 4 4 5 6 7 8-10 2 1 3 1 2 10 16 7 8-10 1 2 2 1 1 3 GRAND TOTAL 2 7 9 3 8 1 15 45 LEVEL 1 2 3 1 1 1 1 5 8 SB8 Crystal Quartz Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) GRAND TOTAL GRAND TOTAL 27 6 4 14 8 4 63 LEVEL Crystal Quartz Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL RAW MATERIAL 2 6 SB7 RAW MATERIAL 1 1 SB6 RAW MATERIAL 8-10 1 10 2 13 GRAND TOTAL 1 15 4 3 16 39 LEVEL 1 2 3 4 5 2 10 3 13 1 2 7 10 6 2 1 4 1 7 1 1 2 3 2 GRAND TOTAL 4 1 3 22 6 36 Since Type I residential elements are characterized by high-density debitage concentrations, they are expected to yield high outcome values for each of these measures. Type III residences, by contrast, are defined by the presence of low-density debitage scatters of about the same size as Type I high density scatters. Consequently, they should be expected to measure high or moderate on the number of positive shovel tests within which they are represented, but low on the number of shovel tests yielding five or more pieces of debitage of a specific raw material type and low on the mean frequency of debitage per positive shovel test. Given the structure of the sample blocks, there are generally nine outcomes (shovel tests) to measure. Some elements present measures that are intermediate between the two extremes and these can only be classified as Type I/III. Type II residential elements would be expected to reflect similar patterns to Type III residences, but they should contain multiple raw material types and a greater density of tools. Clearly, the identification of Type II residences in sample blocks is subjective and it would require addi- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 446 Chapter 9. 31RH480 Table 139. Vertical Distribution of Ceramic Types in Sample Blocks, 31RH480. VERTICAL DISTRIBUTION OF CERAMIC TYPES IN SAMPLE BLOCKS: 31RH480 CERAMIC TYPE LEVELS 1 2 3 4 5 6 7 Vessel # 8 SB2 New River Ia indet. 2 V12 SB3 Hanover IIa indet. 1 V23 SB4 Hanover IIa Fabric Impressed 4 2 3 V1 SB5 Hanover IIIb Fabric Impressed New River IIa Fabric Impressed 2 1 V17 1 V14, V18 SB6 Hanover IIIb Fabric Impressed 1 V27 Yadkin IIId indet. 1 V26 SB7 Hanover IIIa Fabric Impressed 2 V8 Hanover IIIb Fabric Impressed 1 Yadkin IIId Plain 1 1 V10 V9 SB8 Hanover IIIb Fabric Impressed 3 1 1 V3, V6 New River II a Fabric Impressed 1 New River IIa indet. dec. 1 V7 1 V4 Yadkin IIId Cord Marked 1 Yadkin IIId Fabric Impressed 1 1 1 V5 V19 Table 140. Shovel Test Outcome Model for the Purpose of Identifying Element Types. INFERRED ELEMENT TYPE NO. POSITIVE STPS NO. STPS WITH ≥ 5 DEBITAGE MEAN DEBITAGE/ POSITIVE TEST TYPE I HIGH HIGH HIGH TYPE I OR TYPE III HIGH MODERATE MODERATE TYPE III MODERATE LOW LOW Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 447 Chapter 9. 31RH480 tional confirmation from expanded sample blocks or test unit excavation to firmly identify such elements. shovel test representation (1 to 7 shovel tests), low frequencies of shovel tests containing greater than 5 pieces of debitage (0 to 1 shovel test) and debitage frequency means in positive shovel tests of less than or equal to 3.0. Stone tools were scarce. Type I elements were slightly more numerous than was predicted by the Stage I and II sample, but anticipated by an acknowledgement of the limitations of the method. Type I elements identified in the sample blocks represented two of the 44 identified elements (4.6 percent). Table 141 summarizes the results of applying the shovel test model to the elements defined by lithic raw material type within each sample block. As established by the Stage I and II element study, Type III elements predominate in the sample blocks, comprising 86.4 percent (n=38) of the sample. These elements are characterized by variably moderate to high Table 141. Inferred Element Types by Sample Blocks, 31RH480. MEAN DEBITAGE/ + SHOVEL TEST 2 0 2.50 III Type I Rhyolite Tuff 1 0 1.00 III x EA Uwharries Eastern Rhyolite 6 1 2.83 III x MA/EA Uwharries Southern Rhyolite (1) 2 0 2.00 III x EA Uwharries Southern Rhyolite (2) 5 3 8.20 I x EA White Quartz 1 0 1.00 III x MA/EA Mill Mountain Rhyolite 2 0 1.00 III Type I Rhyolite Tuff 1 0 1.00 III x EA ASSOCIATION NO. STs WITH ≥ 5 DEBITAGE Mill Mountain Rhyolite ELEMENTS INFERRED ELEMENT TYPE NO. OF STs REPRESENTED TYPE II CANDIDATE INFERRED ELEMENT TYPES IDENTIFIED IN SAMPLE BLOCKS: 31RH480 STONE TOOLS SB1 Archaic Type IIa Side Scraper (L6) SB2 Archaic Uwharries Eastern Rhyolite 2 0 1.00 III Uwharries Southern Rhyolite (1) 3 0 2.00 III x EA LA/MA Uwharries Southern Rhyolite (2) 7 0 2.00 III x MA/EA Mill Mountain Rhyolite 4 0 3.00 III x EA Type I Rhyolite Tuff 2 1 3.5 III x EA Uwharries Eastern Rhyolite 3 0 1.00 III x EA Uwharries Southern Rhyolite (1) 2 0 1.00 III x EA Uwharries Southern Rhyolite (2) 7 1 2.29 III x EA White Quartz 2 0 1.00 III x EA Mill Mountain Rhyolite 2 0 1.00 III W/LA/MA Type I Rhyolite Tuff 2 0 1.00 III W/LA/MA Uwharries Eastern Rhyolite 2 0 1.00 III LA/MA Uwharries Southern Rhyolite (1) 3 0 2.00 III MA/EA Uwharries Southern Rhyolite (2) 7 0 1.71 III MA/EA White Quartz 4 0 1.50 III W/LA/MA Type I Biface frag. (L7) SB3 SB4 Gray/Black Chert Core frag. (L4) Utilized Flake (L2) Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 448 Chapter 9. 31RH480 Table 141. Inferred Element Types by Sample Blocks, 31RH480 (Continued). NO. STs WITH ≥ 5 DEBITAGE MEAN DEBITAGE/ + SHOVEL TEST INFERRED ELEMENT TYPE Mill Mountain Rhyolite 7 2 3.86 I/III x MA/EA Type I Rhyolite Tuff 3 0 1.75 III x MA/EA ELEMENTS ASSOCIATION NO. OF STs REPRESENTED TYPE II CANDIDATE INFERRED ELEMENT TYPES IDENTIFIED IN SAMPLE BLOCKS: 31RH480 STONE TOOLS SB5 Uwharries Eastern Rhyolite 3 0 1.33 III Uwharries Southern Rhyolite (1) 5 1 2.33 III Archaic Uwharries Southern Rhyolite (2) 4 0 2.00 III x MA/EA White Quartz 3 0 1.33 III x MA/EA Metasedimentary, gray-green 1 0 2.00 III x EA Mill Mountain Rhyolite 4 0 1.75 III Uwharries Eastern Rhyolite 5 1 1.80 III x EA Uwharries Southern Rhyolite (1) 2 0 1.50 III Uwharries Southern Rhyolite (2) 5 0 1.60 III x MA/EA Uwharries Western Rhyolite 1 0 1.00 III x EA White Quartz 6 1 3.00 I/III x MA/EA LA/MA Core frag. (L3) SB6 Archaic Flake Blank (L3) Indet. Core frag. (L2) SB7 Uwharries Eastern Rhyolite 4 1 3.75 I/III x EA Uwharries Southern Rhyolite (1) 3 0 1.33 III x MA/EA Uwharries Southern Rhyolite (2) 2 0 1.50 III x MA/EA White Quartz 4 1 4.00 I/III x MA/EA Crystal Quartz 1 0 1.00 III x EA Crystal Quartz 1 0 4.00 I/III Type I Rhyolite Tuff 1 0 1.00 III W Indet. Flake Blank (L3), Utilized Flake (L3) W/LA/MA Type I Biface (L5), Graver (L3) Utilized Flake (L3) SB8 Uwharries Eastern Rhyolite 2 0 1.50 III Uwharries Southern Rhyolite (1) 6 2 3.67 I Uwharries Southern Rhyolite (2) 3 0 2.00 III W/LA/MA Indet. These elements possessed high shovel test representation (5 to 6 shovel tests), greater numbers of high debitage density shovel test outcomes (2 to 3 shovel tests) and high mean debitage densities, which ranged between 3.67 and 8.20 pieces of debitage. Approximately 9.1 percent (n=4) of the elements exhibited intermediate characteristics and could not be further differentiated. Cultural-chronological associations were inferred from depth profiles and from diagnostic stone tools that were matched to the raw material composition of the debitage concentrations. Only one marginally diagnostic stone tool was recovered, a Type IIa Side Scraper recovered at a depth of 50 to 60 cm bs. It was matched to a USR (2) debitage concentration in SB1 classified as a Type I element (Table 141). The vertical profile of the debitage concentration was consistent with the tool and an Early Archaic assignment (Table 137). It was positioned between 30 and 80 cm bs and peaked between 50 and 60 cm bs. Other than this one instance, all elements were assigned culture-chronological associations based on the vertical profiles of the debitage. Sixteen Early Archaic, 12 Middle Archaic/Early Archaic, three Late Archaic/Middle Archaic, seven Woodland/Late Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 449 Chapter 9. 31RH480 Archaic/Middle Archaic, one Woodland, four generalized Archaic and three indeterminate elements were identified in the sample blocks. Early Archaic and Middle/Early Archaic elements comprise together about 60.9 percent of the identified elements, which is consistent with earlier projections from the Stage I and II shovel test sample. Candidates for Type II elements were recognized in each of the six blocks containing deep vertical debitage profiles (see Table 141). In these instances, groups of Type III elements could hypothetically belong to single Type II residences. Since such elements have only been positively associated with the Early Archaic period, only elements of this inferred affiliation were considered for this procedure. If groups of Type III Early Archaic elements belong to single occupations, the representation of Early Archaic occupations would be significantly diminished in the sample block inventory. Type II residences are characterized by low-density scatters of debitage composed of multiple raw material types and relatively high densities of stone tools of heterogeneous raw material composition. It is suggested that these elements represent multiple family units that came together for short durations to conduct activities such as communal hunts that resulted in abnormally high tool discard rates related to processing animals. Each of the identified sample blocks meets these criteria. Type II elements are rare and not well understood in the region. Similar magnitudes of Type II candidates of Early Archaic age were identified at 31HK2502 and 31HK2510. In several instances, debitage concentrations identified as Type I or Type I/III elements were included in the status of Type II candidates because the debitage reduction profiles are in line with expectations for Type II residences. Late stage reduction types (i.e. BTFs and FBRs) dominate and overall debitage weight is low. For instance, the Type I element of USR (2) in SB1 consists of 85 percent late stage reduction types and the total collected debitage weight is only 12.43 gm. Average weight per piece of debitage is only 0.30 gm and the projection of total debitage weight across the sample block is only about 96 gm, which is equal to the mass of one to two spent directional cores. Type I elements are typically composed of high proportions of early stage debitage and exhibit much greater mass concentrations of debitage. TEST UNITS Four 1-x-1-m test units were excavated at 31RH480 (Table 142). Each unit was quartered and excavated in arbitrary 10 cm levels. The units were generally terminated below the E-/Bt-horizon contact. The precise location of each test unit was determined after close-interval shovel testing was deployed around targeted shovel tests. Two of the units (TU1 and TU2) were excavated as isolated 1-x1-m units, while TU3 and TU4 were placed contiguous to one another forming a 1-x-2-m block (Figure 143). The results of the test unit excavations are presented below. Table 142. Summary Data for Test Units, 31RH480. SUMMARY OF TEST UNITS: 31RH480 SIZE NORTH EAST NO. LEVELS MAX. DEPTH (cm bd) VOLUME 3 (m ) ARTIFACT DEPTH (cm bd) TU1 1x1m 506.00 510.25 7 70 0.70 70 TU2 1x1m 499.75 498.60 7 70 0.70 70 TU3 1x1m 480.50 505.60 7 70 0.70 60 TU4 1x1m 480.50 506.60 7 70 0.70 70 TEST UNIT Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 450 Chapter 9. 31RH480 Test Unit 1 (SB 1) TU1 was placed at grid coordinates N506/ E510.25 (NE corner) in the northwest quadrant of SB1 (Figures 142 and 143). Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of the unit (Figure 154). Three natural soil layers were identified in the profile. The A-Horizon was comprised of light brownish gray (10YR 6/2) loamy sand that was partially stained a very dark grayish brown (10YR 3/2) from leached charcoal derived from prescribed burn episodes. A semi-hemispherical pit filled with recent charcoal was intrusive into the A-Horizon in the southeast corner of the unit that appears to represent a fire pit, probably from military training exercises at the site. The E-Horizon extended below the base of the unit and was divided into two zones of indistinct transition. The upper zone consisted of very pale brown (10YR 6/4) loamy sand and is identified as an E1-Horizon. Between 30 and 40 cm bd, it gave way to a slightly darker, light yellowish brown (10YR 6/4) loamy sand E2-Horizon. Although artifact-density was still high, the unit was temporarily terminated at 50 cm bd due to heavy rains that raised the water table. At the end of the project, excavation was resumed and carried down to 70 cm bd before the water table prevented further downward progress. The Bt-Horizon was never reached. The south wall profile in Figure 154 includes only the upper five levels, as a dry face could not be maintained during the final stages of the excavation. SB 1 was established around a Stage II shovel test excavated at N505/E510 that produced a high-density metavolcanic debitage concentration positioned primarily at 40 to 60 cm bs. Subsequent expansion around the anchor test revealed a complex Early Archaic zone containing a range of metavolcanic elements (see Table 141), the most prominent of which was a USR (2) Type I element associated with a matching Type IIa Side Scraper fragment (Figure 144:J). The USR (2) profile was situated between 40 and 80 cm bs, in a clear Early Archaic vertical position. The vertical distributions of the other, less frequent raw material types appeared to be somewhat higher in the matrix, but within vertical positions indicative of Early or Middle Archaic occupation. Excavation of TU1 confirmed the general picture of the area supplied by the SB1 shovel test sample (Table 143). The USR (2) debitage concentration dominated the collection, consisting of 567 pieces of debitage. It was centered between 40 and 60 cm bd and three matching tools, a Type III Biface tip (Figure 144:G), a utilized flake and a core fragment were found in this same vertical position. Lowdensity USR (1) and Type I RT debitage scatters were also found in this vertical position. MMR and UER scatters were positioned at higher levels. The former displayed a wide vertical profile indicating potential disturbance or multiple elements. The UER debitage scatter was also characterized by wide vertical dispersion, but a clear central tendency could be identified between 10 and 30 cm bd, suggesting a Woodland/ Late Archaic/Middle Archaic age. The USR (2) debitage concentration exhibits a classic raw nodule reduction profile (Table 144). Core flakes are abundant, the incidence of cortex is high and mass output is high. A little more than 0.3 kg of the material was recovered from TU1 and the surrounding SB1 shovel test area. This constitutes a 28.96 percent areal sample of the 6.25 m2 square of SB1, which suggests that as much as 1.05 kg of USR (2) debitage is contained within this area. These characteristics are clearly within the range of a Type I residence. The tool-to-debitage ratio is also consistent with the low values seen in Type I residences. By mass, this ratio is 1:138. The Type I RT and USR (1) profiles, which are situated in the same general vertical position of the USR (2) Type I element, reflect low-level flake blank production and tool maintenance from directional cores. It is unclear whether they belong to the USR (2) occupation or whether they represent separate Type III occupations of Early Archaic groups. The UER and MMR profiles are similar, also appearing to represent low-level reduction from directional cores and manufacturing of tools from flake blanks. Investigations in TU1 and SB1 demonstrate that intact Early Archaic deposits below 40 cm bs are present at 31RH480. In this case, superimposition of occupations on the Early Archaic surface appears to be probable, but raw material profiles supply a basis for analytically differentiating elements and occupations. The primary element was a Type I USR (2) debitage concentration and matching chipped stone tools centered at a depth of 40 to 60 cm bs/bd. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 451 Chapter 9. 31RH480 cm bd SOUTH WALL 0 A Lev 1 10 A MILITARY FIRE PIT Lev 2 20 E1 Lev 3 Root 30 Lev 4 E2 40 Lev 5 50 TU1 N505/E510.25 0 N505/E509.25 1m Soil Strata A: Light Brownish Gray (10YR 6/2) Loamy Sand E1: Very Pale Brown (10YR 7/4) Loamy Sand E2: Light Yellowish Brown (10YR 6/4) Loamy Sand MILITARY FIRE PIT: Charcoal and Loamy Sand Figure 154. Profile Drawing, TU1, South Wall, 31RH480. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 452 Chapter 9. 31RH480 Table 143. Stone Tools and Debitage by Level for TU 1, 31RH480. VERTICAL DISTRIBUTION OF DEBITAGE AND STONE TOOLS, TU1: 31RH480 RAW MATERIAL TYPE LEVEL 1 2 3 Mill Mountain Rhyolite 5 Type I Rhyolite Tuff 4 5 1 6 1 1 1 Uwharries Eastern Rhyolite 1 20 7 2 7 29 White Quartz 3 53 3 27 43 9 1 2 36 5 19 1 25 138 265 73 567 1 58 STONE TOOLS 11 2 2 GRAND TOTAL 3 8 Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) 7 GRAND TOTAL Type III Biface Tip (L5), Utilized Flake (L5), Core frag. (L6) 3 147 294 79 651 Table 144. Debitage Reduction Profiles by Raw Material Type, TU 1 and SB 1, 31RH480. DEBITAGE REDUCTION PROFILES BY RAW MATERIAL TYPE, TU1 and SB1: 31RH480 RAW MATERIAL TYPE % CORE FLAKES IDENTIFIABLE DEBITAGE %CORTEX TOTAL DEBITAGE TOTAL WEIGHT AVERAGE DEBITAGE WEIGHT Mill Mountain Rhyolite 20.0 10 6.3 16 4.98 0.3 Type I Rhyolite Tuff 28.6 7 0.0 10 3.77 0.4 Uwharries Eastern Rhyolite 38.9 36 1.9 53 20.14 0.4 Uwharries Southern Rhyolite (1) 12.5 16 10.3 29 7.04 0.2 Uwharries Southern Rhyolite (2) 34.6 269 15.5 608 304.93 0.5 100.0 2 0.0 4 2.4 0.6 720 343.26 0.5 White Quartz GRAND TOTAL 340 Test Unit 2 (SB 2) TU2 was placed at grid coordinates N499.75/ E498.60 (NE corner) in the southwest quadrant of SB2 (Figures 142 and 143). Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of the unit (Figure 155). Three natural soil layers were identified in the profile. The A-Horizon was comprised of light brownish gray (10YR 6/2) loamy sand that was partially stained a very dark grayish brown (10YR 3/2) from leached charcoal derived from prescribed burn episodes. It extended only about 5 to 7 cm below the surface. The E-Horizon extended to the base of the unit, where it contacted the Bt-Horizon. Color zones could not be discerned in the E-Horizon, which consisted of brownish yellow (10YR 6/6) loamy sand. The Bt-Horizon was contacted across much of the base of the unit at 70 cm bd. It was composed of compact, strong brown (7.5YR 5/6) sandy loam. SB 2 was established around a Stage II shovel test excavated at N500/E499 that produced metavolcanic debitage at depths of 50 to 70 cm bs. It Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 453 Chapter 9. 31RH480 WEST WALL cm bd 0 A Lev 1 10 SHOVEL TEST Roots Lev 2 20 Lev 3 30 Lev 4 40 E Lev 5 50 Lev 6 60 Bt Lev 7 70 TU2 N498.75/E497.60 0 N499.75/E497.60 1m Soil Strata A: Light Brownish Gray (10YR 6/2) Loamy Sand E: Brownish Yellow (10YR 6/6) Loamy Sand Bt: Strong Brown (7.5 YR 5/6) Sandy Loam Figure 155. Profile Drawing, TU2, West Wall, 31RH480. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 454 Chapter 9. 31RH480 was assumed that the area contained a discrete Early Archaic occupation. Subsequent expansion around the anchor test revealed a low-density deposit that appeared to have vertically separated occupations (see Table 141). Analysis of the raw material types, however, suggested that a certain degree of vertical displacement might have occurred in matrix as USR (1) and USR (2) material was found in both high and low vertical positions. Only a single stone tool was recovered, a Type I Biface fragment (Figure 144:E) of USR (2) material recovered between 60 and 70 cm bs. porting an Early Archaic age for the USR (2) element. Along with the debitage, three matching stone tools/ cores were recovered. These included a Type IIa Side Scraper (Figure 144:I) between 40 and 50 cm bd and two flake blanks (Figure 144:C), respectively, from 40 and 50 cm bd and 50 and 60 cm bd. Unlike the USR (2) concentration in SB1, the SB2 concentration exhibits a debitage reduction profile more in keeping with a Type II or III element. Core flake proportions are relatively low, the incidence of cortex is low and mass output is low to moderate (Table 146). Only .045 kg of USR (2) debitage was recovered in TU2 and SB2, while a little more than 0.3 kg of the material was recovered from TU1 and the surrounding SB1. A projected USR (2) weight estimate for the entire area of SB1 was 1.05 kg for USR (2) debitage, while the same projection for SB2 would be 0.155 gm. The tool-to-debitage ratio was 1:1.73 compared to 1:138 for the SB1 USR (2) Type I element. This low ratio is most in keeping with a Type II element as Type III elements usually have higher ratios like those for Type I elements. Although the debitage profiles from shovel tests indicated possible disturbance, the recovery of an ovoid biface fragment at such a low depth merited further investigation to more broadly define the record of Early Archaic occupation at the site. TU2 was placed at a location that appeared to represent the center of the USR (2) debitage concentration. Excavation of TU2 resulted in the recovery of this material, almost to the exclusion of other types (Table 145). The vertical profile of the debitage was broad, ranging between 20 and 60 cm bs, but the central tendency was positioned between 40 and 60 cm bd, sup- Table 145. Stone Tools and Debitage by Level for TU 2, 31RH480. VERTICAL DISTRIBUTION OF DEBITAGE AND STONE TOOLS, TU2: 31RH480 RAW MATERIAL TYPE LEVELS 1 2 Uwharries Southern Rhyolite (1) 3 4 1 1 5 6 7 GRAND TOTAL STONE TOOLS 2 Uwharries Southern Rhyolite (2) 1 13 12 19 26 3 74 GRAND TOTAL 1 14 13 19 26 3 76 Type IIa Side Scraper (L5), Flake Blank (L5), Flake Blank (L6) Table 146. Debitage Reduction Profiles by Raw Material Type, TU 2 and SB 2, 31RH480. DEBITAGE REDUCTION PROFILES BY RAW MATERIAL TYPE, TU2 and SB2: 31RH480 RAW MATERIAL TYPE Uwharries Southern Rhyolite (2) % CORE FLAKES 24.4 IDENTIFIABLE DEBITAGE 45 %CORTEX 4.8 TOTAL DEBITAGE 88 TOTAL WEIGHT (gm) AVERAGE DEBITAGE WEIGHT (gm) 44.84 0.51 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 455 Chapter 9. 31RH480 Investigation in TU2/SB2 again demonstrates that intact Early Archaic deposits below 40 cm bs are present at 31RH480. Here, however, only minimal superimposition and mixing with other occupations on the Early Archaic surface was documented. The primary element was what appears to be a Type II USR (2) debitage concentration with matching chipped stone tools positioned at depths of 40 to 60 cm bs/bd based on the vertical distributions of stone tools and, secondarily, on the debitage vertical profile. The latter appears to have experienced a degree of vertical displacement upward. Test Units 3 and 4 (SB 4) TUs 3 and 4 were placed contiguous to one another at grid coordinates N480.50/E505.5 (NE corner) and N505.5/E506.5 (NE corner), respectively, in the east central portion of SB 4 (Figures 142 and 143). Excavation proceeded in arbitrary 10 cm levels reckoned from the NE corner of TU3 (Figure 156). Three natural soil layers were identified in the profile. The A-Horizon was comprised of light brownish gray (10YR 6/2) loamy sand that was partially stained a very dark grayish brown (10YR 3/2) from leached charcoal derived from prescribed burn episodes. It extended only about 5 to 10 cm below the surface. An historic trash pit containing a seamed Coca-Coal can at its base intruded through the A-Horizon on the southern side of TU4. A thin, humic zone of charcoal covered the very top of the pit, which was slightly depressed at the surface. This zone was probably produced during prescribed burning. The pit extended to the very top of the Bt-Horizon at a depth of about 55 cm bd. The E-Horizon was divided into two zones of indistinct transition. The upper zone consisted of light brown (7.5YR 6/4) loamy sand and is identified as an E1-Horizon. This layer gave way to an E2-Horiozn between 40 and 45 cm bd, which was composed of pink (7.5YR 7/4) loamy sand. The BtHorizon was composed of compact, strong brown (7.5YR 5/6) sandy loam. Its contact with the overlying E2-Horiozn was wavy and occurred between 55 and 60 cm bd. SB 4 was established around a Stage II shovel test excavated at N480/E505 that produced six sherds from a single Hanover IIa Fabric Impressed vessel (V1) positioned between 10 and 30 cm bs. In addition to the sherds were a large biotite crystal fragment, a white quartz core fragment and two pieces of metavolcanic debitage positioned between 10 and 40 cm bs. SB4 was established to further explore this aggregated sherd cluster and possible associated features and lithic elements. Shovel testing produced three more sherds belonging to V1, a light concentration of debitage, seven more biotite crystal fragments, a dark gray chert utilized flake and seven pieces of quartz cobble fire-cracked rock and arenite and sandstone conglomerate rock (Tables 137 and 139). The low-density debitage scatters were all classified as Type III elements and, owing to their relatively high vertical positions, were inferred to represent Woodland/Late Archaic/Middle Archaic aged occupations (Table 140). Excavation of TUs 3 and 4 succeeded in the recovery of a large portion of the Hanover IIa Fabric Impressed vessel (V1) in the upper 20 cm of deposit and a portion of an Early Archaic element of USR (1) material positioned primarily between 20 and 50 cm bd (Table 147). Sherds from the vessel (Figure 146:D-I) were concentrated in the northern half of TU3 and the aggregation extended north and west of this unit (Figure 157). The limits of the aggregation are enclosed by the northern and western shovel tests in SB4, which failed to produce additional sherds of this vessel. Based on data from SB4 and the test units, the aggregate’s maximum diameter is estimated to extend not more than about 2.5 m and the main concentration appears to be confined to an area measuring about 1 x 1.5 m. Total weight for V1 from sherds recovered in the test units and the sample block shovel tests was 262.31 gm, which equals only about 19 percent of the idealized vessel weight of 1,377 gm for an open mouth jar as suggested by Herbert and Irwin (2003:2). More of the vessel remains in the ground, but based on the size projection of the aggregate and the coverage achieved in the sample block, it is unlikely that more than 50 percent of the idealized vessel weight is contained at this location. The fact that most of the sherds recovered represent thinner fragments from the area of the orifice, though, might indicate that heavier sherds associated with the base still remain to be recovered and that a greater portion of the vessel is present than is projected. Alternatively, it might suggest that the aggregation represents a vessel fragment only, consisting of a large area of one side of the original vessel. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 456 Chapter 9. 31RH480 SOUTH WALL H cm bd 0 A Lev 1 10 Lev 2 MILITARY TRASH PIT 20 Root E1 Lev 3 30 Lev 4 Coca-Coal Can 40 Lev 5 50 E2 Lev 6 60 Bt Lev 7 70 TU4 N479.5/E506.6 TU3 N479.5/E505.6 N479.5/E504.6 0 2m Soil Strata H: A: E1: E2: Bt: Black (10YR 2/1) Loamy Sand Humus/ Prescribed Burn Charcoal Light Brownish Gray (10YR 6/2) Loamy Sand Light Brown (7.5YR 6/4) Loamy Sand Pink (7.5 YR 7/4) Loamy Sand Strong Brown (7.5 YR 5/6) Sandy Loam MILITARY TRASH PIT: Jumbled A- and E-Horizon Loamy Sands Figure 156. Profile Drawing, TU3 and TU4, South Wall, 31RH480. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 457 Chapter 9. 31RH480 Table 147. Vertical Artifact Distributions, TU 3 and TU 4, 31RH480. VERTICAL ARTIFACT DISTRIBUTIONS, TU3 AND TU4: 31RH480 TU3 DEBITAGE RAW MATERIAL TYPE LEVEL 1 2 Mill Mountain Rhyolite 3 4 5 6 7 2 Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) 1 STONE TOOLS 2 1 2 5 4 6 1 2 Uwharries Western Rhyolite GRAND TOTAL 1 4 1 17 3 White Quartz Palmer II Corner Notched (L5) Flake Blank (L4) Hammer Stone (L3) GRAND TOTAL 1 8 7 8 1 1 26 TU4 DEBITAGE RAW MATERIAL TYPE LEVEL 1 2 3 4 Uwharries Eastern Rhyolite 1 1 1 Uwharries Southern Rhyolite (1) 5 2 10 Uwharries Southern Rhyolite (2) 1 2 White Quartz 3 2 7 15 GRAND TOTAL 6 5 6 7 GRAND TOTAL STONE TOOLS 3 7 5 29 Type IIb Side Scraper (L5) 3 5 7 5 Core frag. (L4) 40 TU3 ARTIFACT TYPE LEVEL 1 2 Biotite Fragment Hanover IIa Fabric Impressed Hanover IIa indet. GRAND TOTAL 52 7 7 1 59 8 3 4 27 27 5 9 6 7 1 GRAND TOTAL 64 2 61 8 29 27 9 1 133 TU4 ARTIFACT TYPE LEVEL 1 2 3 Biotite Fragment Hanover IIa Fabric Impressed 5 6 1 10 Hanover IIa indet. 4 1 10 GRAND TOTAL 15 1 1 Hanover IIIb Cord Marked 7 1 1 Hanover IIIa Cord Marked GRAND TOTAL 4 2 1 7 1 1 3 2 1 21 The USR (1) debitage concentration was matched to a Palmer II Corner-Notched point (Figure 144:H) and a Type IIb Side Scraper (Figure 144:K), both of which were found between 40 and 50 cm bd, well within the expected vertical position for intact Early Archaic deposits in the region. The debitage has a slightly more variable vertical distribution, but 71 percent of it occurs below 30 cm bd in the test units. The intrusive trash pit in the south one-quarter of TU4 may have contributed to the dispersed vertical profile of the USR (1) debitage, but there is very little difference between the profiles from the two test units (Table 147). USR (1) debitage was concentrated in the northeast corner of SB4 and TU4 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 458 Chapter 9. 31RH480 482.0 Distribution of Vessel 1, Hanover IIa Fabric Impressed (Contours = 5 gm, Starting with 10 gm) SB4 481.5 481.0 TU3 North 480.5 TU4 480.0 479.5 479.0 478.5 478.0 503.0 503.5 504.0 504.5 505.0 505.5 506.0 506.5 507.0 East Scale N 1 meter Figure 157. Density Distribution of Vessel 1 Sherds, SB4, 31RH480. (Figure 158) and the matching tools were recovered on the western side of it, in it what likely represents a former hearth area. Although found slightly higher, between 30 and 40 cm bd, a cobble hammer stone and a flake blank of UWR material may belong to this tool cluster as well. The debitage reduction profile for the USR (1) material is in keeping with a Type II or III element. Core flake proportions are relatively low, the incidence of cortex is only moderate and mass output is low (Table 148). Only 0.14 kg of USR (1) debitage was recovered in the TUs and shovel tests and since coverage in SB4 equaled about 45 percent of the area it is projected that only about 31.5 gm of debitage is present in the entire element. This weight is less than that of a typical exhausted core, suggesting activity related to tool manufacturing was limited to flake blank production and subsequent thinning rather than primary core reduction. The tool-to-debitage ratio was 1:1.91, a result that might suggest that the concentration represents a Type II element. Given that USR (1) element may represent the anchor of a Type II residence, the hammer stone (Figure 145:A), the UWR flake blank (Figure 144:D) and a small amount of UWR and USR (2) debitage found in this general vicinity may also represent elements of the occupation. Biotite crystal fragments (Figure 145:C) were concentrated in the NW and NE quadrants of TU3 and appear overlap the distribution of the V1 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 459 Chapter 9. 31RH480 Hypothesized Type II Element, Palmer II Phase 482.0 SB4 481.5 481.0 TU3 North 480.5 TU4 480.0 479.5 479.0 478.5 478.0 503.0 503.5 504.0 504.5 505.0 505.5 506.0 506.5 507.0 East Palmer II Corner-Notched, USR (1) Type IIb Side Scraper, USR (1) Scale 1 meter Flake Blank, UWR Hammer Stone, Quartz Cobble N USR (1) Debitage (Contours= 1 item, starting at 5) Figure 158. Hypothesized Palmer II Type II Element, SB4, 31RH480. sherd aggregate (Figure 159). The material is primarily confined to an area of 1.5 m in diameter, but it may extend a little farther north of the sample block. Biotite crystals are primarily distributed between 20 and 40 cm bd, which is clearly above the Palmer II surface, but somewhat lower than the sherd aggregate. It is possible that the biotite was cached in a pit originating in the Woodland zone of the site. If so, the concentration might be associated with the sherd aggregate and belong to a sleeping surface or hut interior. The apparent absence of a Woodland debitage scatter in the sample block supports this inference, as such features are predicted to be segregated from sherd aggregates, which are hypothesized to be associated with sleeping surfaces. Investigation in SB4 succeeded in partially recovering a relatively complete sherd aggre- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 460 Chapter 9. 31RH480 Table 148. Debitage Reduction Profiles for USR (1) , TUs 3 and 4 and SB 4, 31RH480. DEBITAGE REDUCTION PROFILES FOR USR (1) MATERIAL, TUs 3 & 4 and SB4: 31RH480 % CORE FLAKES RAW MATERIAL TYPE Uwharries Southern Rhyolite (2) 0.08 IDENTIFIABLE DEBITAGE %CORTEX 36 TOTAL DEBITAGE 13.50 TOTAL WEIGHT (gm) AVERAGE DEBITAGE WEIGHT (gm) 14.15 0.27 52 Distribution of Biotite Crystal Fragments (Contours = 5 gm, Starting with 10 gm) 482.0 SB4 481.5 481.0 TU3 North 480.5 TU4 480.0 479.5 479.0 478.5 478.0 503.0 503.5 504.0 504.5 505.0 505.5 506.0 506.5 507.0 East Scale N 1 meter Figure 159. Distribution of Biotite Crystal Fragments, SB4, 31RH480. gate composed of portions of a Hanover IIa Fabric Impressed open mouth jar which was superimposed over an intact Early Archaic Type II or III occupation. A cluster of biotite crystals coterminous with the sherd aggregate may represent cached material associated with the Woodland period occupation. The sherd aggregate and the biotite concentration may define a Woodland period sleeping surface or hut area, as the area appears devoid of other Woodland period debris. The Early Archaic element is affiliated Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 461 Chapter 9. 31RH480 with the Palmer II phase and may very well represent a Type II camp. In spite of superimposition, the Early Archaic surface was intact. EVALUATION Site 31RH480 is situated on a ridge slope paralleling an unnamed tributary of Big Muddy Creek. The surface has been extensively damaged by logging and military training activities. Much of the Woodland period occupation has been disturbed by these historic land uses, although spotty intact Woodland deposits were documented. These remnants are not considered sufficiently extensive to yield significant data on the Woodland period that could not be more efficiently and effectively recovered at other sites. The site also contains an extensive Middle to Early Holocene occupation surface situated between 30 and 60 cm bs that has escaped significant historic impacts. Test unit excavations succeeded in isolating Early Archaic occupations in three separate areas. These cultural deposits were intact, but exhibited variable degrees of mixing from reoccupation. Even in these instances, however, individual occupations could be analytically segregated through detailed lithic raw material type identifications. Consequently, Archaic camp structure and spatial organization are accessible at this site and it follows that this surface could yield important information concerning Archaic settlement patterns and regional adaptation. However, a nearby site, 31RH491, in the Camp MacKall segment of the Drowning Creek stream valley that is recommended eligible for inclusion on the NRHP in this report, contains an identical record of occupation, but has the added advantage of offering a much larger and more comprehensive sample of Early and Middle Archaic components. Therefore, the opportunity to address a much wider range of research questions is presented by the more extensive record at 31RH491. In this light, the Early and Middle Holocene deposits at 31RH480 would not significantly enhance the information contained at 31RH491 beyond what has been learned from the Phase II investigation. Thus, the data contained at 31RH480 is viewed as redundant and the site is not recommended eligible for inclusion on the National Register of Historic Places for this reason. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 462 Chapter 10. 31RH491 Site 31RH491 is located in Richmond County, NC, within the confines of Camp MacKall (Figure 160). The site is situated on a ridge terrace overlooking the broad floodplain of Drowning Creek (Figure 161), the channel of which is situated approximately 0.5 km east of the site (Figure 162). The confluence of and Drowning Creek and Big Muddy Creek is approximately 2 km distant and southeast of the site. To reach the site travel southwest on General Ridgeway Road from the bridge at Drowning Creek and turn right onto an unnamed gravel road located approximately 0.43 miles from the bridge. Travel northwest along the gravel road for 0.35 miles. A trail intersects the gravel road on the west at this points and leads to a wildlife plot at the north end of the site. The SCS soil survey for Richmond County, NC (Evans 1999) classifies the landform as Pelion loamy sand, a moderately well drained soil typically found in the uplands of the Sandhills on both ridge tops and slopes. The rolling landform on which the site rests is topographically complex due to both historic land-use modifications and natural contour variation (Figure 163). The northern end of the site is relatively flat and supports a fallow wildlife plot (Figure 164). This area is separated by the crest of the ridge to the south by a swale sloping off to the southwest. A portion of the ridge apex has been bulldozed away on the west side of the gravel road (Figure 165), which has undoubtedly resulted in the removal of some of the original archeological deposits. Other historic disturbances include two earthen gun implacement constructions measuring between 6 and 8 m in diameter each (Figure 166). Elevations within the site boundaries range from about 82.0 amsl along the wetland/floodplain interface to about 84.5 amsl on the apex of the ridge. Vegetation across the site consisted of a thinned stand of young long-leaf pines and an understory of turkey oak and wiregrass (Figure 167). Maximum site dimensions were established at 165 m x 315 m, covering an area of approximately 5.38 acres (21,600 m2). PREVIOUS RESEARCH Site 31RH491 was originally identified by TRC Garrow Associates, Inc in the fall/winter of 2002/2003 during a Phase I survey of portions of Fort Bragg and the KK1 Tract at Camp MacKall (Grunden and Ruggiero 2006:482-486). The site boundaries established at that time measured 90 x 180 m, covering an estimated area of 18,255 m2 (Figure 168). Phase I site definition proceeded through the deployment of 15 m-interval shovel tests around four positive survey transect shovel tests. Including site discovery transect shovel tests, 96 shovel tests were excavated in total, 30 of which were positive. A 0.5 m-x-0.5 m test unit was excavated near the survey grid coordinates of N485/E485. Grunden and Ruggiero (2006:482) described three soil zones exposed in shovel tests. The upper zone was 0 to 10 cm thick and consisted of very dark gray (2.5Y 3/1) loamy sand. Below this layer was a Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 463 Chapter 10. 31RH491 640000 3880000 3879000 3879000 3880000 3881000 639000 3881000 638000 3878000 3877000 3877000 3878000 31RH491 638000 639000 640000 Figure 160. Location of Site 31RH491 (Pine Bluff, NC 7.5’ USGS Quadrangle) Scale: 1:24,000. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 464 Chapter 10. 31RH491 639800 640000 Dr ow 640200 640400 3879400 639600 3879400 639400 3879200 ee Cr 3879200 g nin 3879000 3879000 k 3878800 3878800 31RH491 n 3878400 3878200 11th A 3878200 3878400 Ge lR a er ew g id 3878600 3878600 d R ay 3878000 iv Rd 3878000 BN D 639400 639600 639800 640000 640200 640400 Figure 161. Portion of Fort Bragg Aerial Photograph Series (2006), Vicinity of Site 31RH491 (Scale: 1 inch = 200m, Projection: UTM, NAD83). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 465 Chapter 10. 31RH491 Figure 162. View of Site, Drowning Creek Looking North from the Bridge on General Ridgeway Road. variably thick light yellowish brown (2.5Y 6/4) sand extending to depths of as much as 95 cm bs. Although not specifically described, it is presumed that the third layer would have been an argillic Bt-Horizon. The test unit contacted the water the water table before the Bt-Horizon was reached. Subsurface and surface contexts yielded 334 precontact artifacts, including seven sherds, 322 pieces of debitage, a cobble fragment, a biface, a projectile point tip, a corner notched projectile point fragment and a Kirk Corner Notched point. Most of the chipped stone artifacts were made of rhyolite, including the projectile points and biface. Precontact sherds consisted of fabric impressed and cord marked surface treatments and a variety of tempers were noted, including crushed quartz, crushed rock, clay/grog and sand. A sherd of Bristol slip stoneware and a piece of brown container glass were recovered as well. Two artifact concentrations were recognized from shovel test density distributions. One was located in the area of the wildlife plot and the other was situated on the ridge apex to the south and on the other side of the swale. Artifacts were found throughout the E-Horizon matrix. The test unit produced abundant lithic debitage between 40 and 80 cm bs. Based on the presence of extensive subsurface deposits in apparent undisturbed contexts, the site was recommended potentially eligible for inclusion on the National Register of Historic Places. Further work here was considered likely to yield important information concerning precontact settlement patterns and adaptive strategies. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 466 Chapter 10. 31RH491 560 MN 540 l .0 520 83 na Ca 0 10 20 meters Gun Implacement 500 82 .5 Food Plot 480 .0 82 460 ra G d oa lR ve 31RH491 440 83.5 .5 81 420 83 .5 84.5 400 .0 84 83.5 380 360 .5 84 340 84.0 320 83.0 300 280 260 240 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 163. Site 31RH491, Base Map. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 467 Chapter 10. 31RH491 Figure 164. View of Site 31RH491 Looking West from N510/E510 to the Wildlife Plot. FIELDWORK OVERVIEW Phase II fieldwork was implemented in four stages, each stage building on the information generated from the previous stage. Stage I investigation consisted of a program of 10 m-interval shovel tests that were excavated to establish firm site boundaries. Site boundaries were determined by the documentation of two consecutive negative shovel tests in all directions on a 10-m grid, or in some cases when wetland was encountered. Once artifact distributions from the Stage I program were mapped and evaluated, the Stage II investigation involved the excavation of closer-interval shovel tests at 5 m-intervals in most of the artifact bearing areas of the site. Stage III investigations involved the deployment of closeinterval shovel tests of 1.25 m around the perimeter of targeted shovel tests yielding deposits deemed of theoretical importance for evaluating the site. These close-interval shovel test locations were identified as sample blocks and generally consisted of eight shovel tests excavated around a targeted Stage I or Stage II shovel test. The sample blocks were also instrumental in guiding the precise placement of 1-x-1 m test units, which constituted the final stage of the investigation. Test units were placed in locations judged to provide optimal opportunities to reconstruct the occupation history of the site and to assess the integrity and scientific importance of the deposits. Stage I investigations were initiated by relocating the datum established by Grunden and Ruggiero (2006) and laying out a 10-m interval grid with a total station. The grid coordinates (N500/E500) for the datum were maintained for the Phase II work. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 468 Chapter 10. 31RH491 Figure 165. View of Site 31RH491 Looking South from N460/E540 along Grade Cut Paralleling the Gravel Road. Relative topographic elevations at each 10 m-grid intersection were recorded with the total station and marked with a pin flag. The elevation values (in meters) illustrated on the base map (Figure 163) are approximated real elevations based on the correlation of GPS points with a geo-referenced copy of the USGS, Pine Bluff, NC 7.5-minute topographic quadrangle map. Cultural and landform features were also mapped and additional elevations were taken when appropriate to produce the topographic map. Shovel tests measured 30-x-30-cm and were square in shape. Excavation proceeded in 10 cm arbitrary levels measured from the surface. Two hundred and eighty-five Stage I shovel tests were excavated at 10 m-intervals to define the limits of the site (Figure 169). Of these, 101 produced artifacts. Maximum site boundaries were established at 165 m x 315 m, covering an area of approximately 5.38 acres (21,600 m2). Mean shovel test depth for the Stage I sample was 56.74 cm bs (SD=25.91) with a mode of 50 cm bs and a range of 10 to 105 cm bs. Generally, shovel tests were terminated after contact with substrate, which was variably deep across the site due to disturbances. The SCS soil survey for Richmond County (Evans 1999) maps the area of the site as Pelion loamy sand (0 to 2 percent slopes). The on-site soil profile corresponds well to the typical pedon of the soil type, but the E-Horizon is generally thicker in undisturbed locations. The top layer represents a 20 cm thick plow zone (Ap-Horizon) composed of dark gray (10YR 4/1) to very dark gray (10YR 3/1) loamy sand. Below this, the E-Horizon generally extended to depths of 10 to 100 cm bs. It consisted of pale Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 469 Chapter 10. 31RH491 Figure 166. View of Site 31RH491 Looking Southeast from N370/E570 toward Gun Implacement. brown (10YR 6/3) to light brown (10YR 6/4) loamy sand. A strong brown (7.5YR 5/6) sandy clay loam Bt-Horizon underlay the E-Horizon. The boundary between the two layers was smooth and the transition consisted of a sandy loam. The Stage I sample provided a basis for defining a Stage II sample frame within which to deploy a 5 m-interval shovel test grid. This guaranteed that the artifact bearing locations of the site would receive systematic sampling at a relatively intensive sample density, so that a detailed picture of the occupation history of the site could be generated. Three hundred and ninety-three Stage II shovel tests were excavated at 5 m-intervals within the sample frame (Figure 170). Two hundred and forty-seven of these were positive. Mean shovel test depth for the Stage II sample was 66.96 cm bs (SD=24.82) with a mode of 90 cm bs and a range of 10 to 105 cm bs. The Stage III shovel test sample was distributed to eight locations anchored by Stage I or II shovel tests containing data relevant to evaluating the occupational history and integrity of the site matrix (Figure 171). One hundred nineteen Stage III shovel tests were excavated at typically 1.25 m-intervals around targeted Stage I and II shovel tests. Of these, 107 were positive. Mean shovel test depth for the Stage III sample was 80.33 cm bs (SD=22.56) with a mode of 100 cm bs and a range of 30 to 110 cm bs. The greater mean depth of shovel tests in both the Stage II and Stage III samples resulted from focusing on less disturbed areas of the site. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 470 Chapter 10. 31RH491 Figure 167. View of Site 31RH491 Looking South from N520/E530 Down the Gravel Road. Once the results of the Stage III shovel test sample were analyzed, specific locations yielding shovel test outcomes of interest to further evaluating the site were selected for test unit excavation. Stage III shovel testing around targeted shovel test outcomes provided the basis for precise placement of test units. Five 1-m-x-1-m test units were excavated in total and their locations are illustrated in Figure 172. All test units were terminated once the Bt-Horizon was reached, which ranged between 30 and 70 cm bd. ARTIFACT INVENTORY The artifact inventory from Phase II operations at 31RH491 consisted of 4,513 items including 3,872 pieces of lithic debitage, 47 cores/flake blanks, 21 bifaces, 28 projectile points, 60 unifaces, 29 utilized flakes, 8 cobble tools, 8 pieces of float, 82 precontact sherds, 221 pieces of rock/fire-cracked rock, 5 faunal bone fragments, 2 ethnobotanical charcoal fragments and 130 historic artifacts. Appendix B contains the artifact database for the project, while appendices C through I provide additional data on lithic stone tools and precontact ceramics. Descriptions of the various artifact classes and lithic raw material types are presented below. Lithic Raw Material Types Lithic raw material identification proceeded on two different levels. At the macro-scale, indi- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 471 - 575 - 560 - 545 - 530 - 515 - 500 - 485 - 470 - 455 - 440 - 425 - 410 Chapter 10. 31RH491 - 530 Gr - 515 51-3 lR ave 51-2 oad TR 51 - 500 51-1 - 485 Wildlife food field TR 52 13-1 52-2 TR 13 - 470 52-3 52-4 S sh pu - 455 es pil TR 53 TR 12 S 53-4 53-3 - 440 12-1 - 425 modern brick TR 54 - 410 54-3 54-2 - 395 gun implacement - 380 - 365 South Wall Centimeters 0 10 2.5Y 3/1 Very dark gray Sand - 350 20 - 335 30 2.5Y 6/4 Light yellowish brown Sand Water Table S GPS datum/Positive ST Negative transect ST Positive ST Positive transect ST Negative ST Site boundary Surface positive (if any) Contour (approximate) 50 cm x 50 cm test unit Wetland (if any) N Scale in Meters 0 15 30 45 Shovel test pits shown are not to scale Figure 255. Site Plan of 31RH491. Figure 168. Site 31RH491, Phase I Survey Sketch Map, TRC (from Grunden and Ruggiero 2006). 483 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 472 Chapter 10. 31RH491 560 MN 540 l .0 520 83 na Ca 0 10 20 meters Gun Implacement 500 .5 82 Food Plot Positive ST 480 Negative ST .0 82 460 lR ve ra G d oa 31RH491 440 83.5 .5 81 420 83 .5 84.5 400 .0 84 83.5 380 360 .5 84 340 84.0 320 83.0 300 280 260 240 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 169. Site 31RH491, Stage I Shovel Test Sample. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 473 Chapter 10. 31RH491 560 MN 540 l .0 520 83 na Ca 0 10 20 meters Gun Implacement 500 82 .5 Food Plot 480 Positive ST Negative ST .0 82 460 d oa lR ve ra G 440 83.5 .5 81 31RH491 420 83 .5 84.5 400 .0 84 83.5 380 360 .5 84 340 84.0 320 83.0 300 280 260 240 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 170. Site 31RH491, Stage I and II Shovel Test Samples. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 474 Chapter 10. 31RH491 560 MN 540 l .0 520 83 na Ca SB3 SB15 0 20 meters SB2 SB14 500 Gun Implacement SB13 82 SB11 .5 SB12 Food Plot 480 10 Positive ST Negative ST .0 82 SB10 SB9 460 ve ra G d oa 440 lR 31RH491 83.5 81 .5 SB1 420 SB7 83 .5 SB8 84.5 SB6 400 .0 84 83.5 SB5 380 SB4 360 .5 84 340 84.0 320 83.0 300 280 260 240 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 171. Site 31RH491, Stage III Shovel Tests and Sample Block Locations. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 475 Chapter 10. 31RH491 560 MN 540 l .0 520 83 na Ca 0 10 20 meters Gun Implacement 500 .5 82 480 TU5 Food Plot TU1 .0 82 460 lR ve ra G d oa 31RH491 440 83.5 81 .5 TU2 420 83 .5 TU3 84.5 400 .0 84 83.5 380 TU4 360 .5 84 340 84.0 320 83.0 300 280 260 240 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 172. Site 31RH491, Test Unit Locations. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 476 Chapter 10. 31RH491 vidual specimens were typed in accordance with major hard rock geological terminology to achieve comparability with previous projects at Fort Bragg. At the micro-level, however, an attempt was made to further partition some of the hard rock types into “core groupings” to distinguish between specimens derived from single hard rock types that may have originated from different cores. This was done to facilitate spatial analysis in close-interval shovel test grids where overlapping of distinct deposits or complex multi-household occupations were suspected. Core grouping analysis was undertaken principally in metavolcanic types where subtle differences in patina coloration, core color, phenocryst size and density, and matrix texture was easily discerned and monitored. Descriptions of the various rock types and the associated core grouping sub-types recognized in the collection are presented below. Metavolcanic Types Metavolcanic types make up the majority of the chipped stone collection from the site (Table 149). As a group, these types represent 91.3 percent of the total chipped stone artifact inventory (i.e. debitage and chipped stone cores and tools). Descriptions of the five identified metavolcanic types and associated core group subtypes follow. The best reflection of actual representation, however, is supplied by the combined Stage I and II shovel test samples, which is relatively unbiased in terms of spatial coverage. In this sample, metavolcanic material makes up 86.0 percent of the chipped stone, an indication that the Stage III and IV investigations slightly favored contexts that contained greater proportions of metavolcanic chipped stone. Descriptions of the five identified metavolcanic types and associated core group subtypes follow. The typological system varies somewhat from earlier reports in response to the recent stone quarry sourcing study for the Carolina Slate Belt conducted by Steponaitis et al. (2006). (1) Mill Mountain Rhyolite (MMR). MMR was not clearly differentiated in the sourcing study (Steponaitis et al. 2006), but it is a readily recognizable type in the assemblages from Fort Bragg. Mill Mountain is located between the Uwharries Southern and Uwharries Eastern source areas on the opposite bank of the Yadkin River from Table Top Mountain. It was not included in either of these source areas, however, because of the presence of distinctive clear, glassy quartz phenocrysts. Daniel and Butler (1996:18) describe this type as a medium gray, aphanitic, rhyolite porphyry with sparse, glassy phenocrysts of quartz generally less than 1 mm in diameter. Very fine-grained, disseminated grains of pyrite are also present. Thin sections indicate a microcrystalline matrix composed primarily of feldspar and quartz, with some biotite, chlorite, and disseminated pyrite. The specimens assignable to this type from the 31RH480 collection are characterized by a dark reddish gray to moderate brown coloration when not patinated. Patination is generally grayish orange to dark yellowish orange. Four subtypes were assigned to the type, based primarily on the completeness of the observation field and patination. Subtype R8 corresponds to fine-grained specimens that exhibit sparse glassy quartz phenocrysts only. Subtype R8m represents medium-grained examples of Subtype R8. Subtypes R8Hhf, R8Hhm and R8Hhc are fine-, medium- and coarse-grained specimens respectively that exhibit sparse glassy quartz, white subhedral quartz phenocrysts and, rarely, feldpsar phenocrysts. Although these white quartz phenocrysts were not included in the type description, the subtypes containing them were included as a variant of Mill Mountain Rhyolite due to the co-occurrence with glassy quartz phenocrysts. Mill Mountain Rhyolite makes up 5.7 percent of the combined Stage I and II chipped stone sample. (2) Type I Rhyolite Tuff (Type I RT). An exact correlate of this type was not discernible in the sourcing study (Steponaitis et al. 2006). Benson (1999:30) describes it as fine-grained dark green to gray material that looks and feels very similar to Piedmont chert. He noted that plagioclase phenocrysts were sporadically present. Two subtypes were identified in the 31RH480 collection. Subtype R2 is only rarely identified in assemblages on Fort Bragg because the material patinates quickly with exposure tosunlight. Macroscopically it appears very similar to Subtype R3t, but its coloration can be described as a dusky blue green. The texture is microcrystalline and the edges are semi-translucent. Subtype R2A specimens exhibit a pale green patina, but broken edges reveal that the underlying material corresponds to R2. Darker green splotches and swirls on the faces of these specimens represent more resis- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 477 Chapter 10. 31RH491 Table 149. Lithic Raw Material Frequencies by Chipped Stone Tool Class, 31RH491. PROJECTILE POINT UNIFACE UTILIZED FLAKE 24 Mill Mountain Rhyolite 158 3 7 3 1 172 Type I Rhyolite Tuff 253 2 4 7 6 272 Uwharries Eastern Rhyolite 343 2 6 2 5 3 361 Uwharries Southern Rhyolite (1) 350 8 9 3 19 6 395 Uwharries Southern Rhyolite (2) 2249 4 11 10 23 12 2309 166 3 BIFACE Hydrated metavolcanic RAW MATERIAL TYPE DEBITAGE CORE LITHIC RAW MATERIAL FREQUENCIES BY CHIPPED STONE TOOL CLASS: 31RH491 GRAND TOTAL Metavolcanic Uwharries Western Rhyolite 24 2 171 Metasedimentary Metasedimentary, Black 4 Metasedimentary, Green 16 Schist 18 4 2 1 2 19 20 Quartz Crystal Quartz 14 White Quartz 273 White Quartz, Fractured 2 3 3 1 10 1 18 1 287 1 4 Chert Fossiliferous Coastal Plain Chert GRAND TOTAL 1 1 3872 21 47 28 60 29 4057 tant matrix. The matrix is aphantitic and microcrystalline in texture and is highly isotropic in fracture characteristics. Subtype R2Af corresponds to the description for R2A, but it also exhibits fine laminations. R2Am and R2Ams represent medium grained specimens of R2 material. The latter also included splotches and swirls of more resistant material. The origin of Type I Rhyolite Tuff is not well understood and it is possible that it represents a very fine-grained metamudstone similar to some of those described for the Chatham Pittsboro source (see Moore and Irwin 2006:27-28; Stoddard 2006:57). In support of this is the occurrence of what appear to be bedding laminations in some of the patinated material. Moreover, the color of the patinated specimens is similar to that reported for the Chatham Pittsboro source. Type I Rhyolite Tuff makes up 4.6 percent of the combined Stage I and II chipped stone sample. (3) Uwharries Eastern Rhyolite (UER). Stoddard (2006:52) describes the specimens from this sourcing area as light to dark gray metadacite porphyry or crystal lithic tuff. All samples collected contained plagioclase feldspar and white quartz phenocrysts, ranging from less than 2 percent to 7 percent phenocryst density. Common metamorphic minerals within the matrices include biotite, stilpnomelane and calcite. Similar dacites and rhyolites are typical of the Uhwarries Southeastern source (Stoddard 2006:55-56) as well, but many of these specimens Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 478 Chapter 10. 31RH491 contain silica levels beyond the range typical of igneous rocks. Macroscopic criteria for distinguishing between lithic material from these two sources does not at present exist. Consequently, specimens from both source areas may be present in the analyzed lithic assemblages at Fort Bragg. Typical specimens from 31RH491 are characterized by dark gray (nonpatinated) and grayish green (patinated) matrices with splotches of silicate and variable percentages (2 to 4 percent) of white quartz and plagioclase feldspar phenocrysts. Some specimens exhibit purplish to brown background matrices. Quartz phenocrysts predominate over feldspar in the examples from the site. Type I Rhyolite, which has been distinguished separately in previous reports (Cable and Cantley 2005b, 2006, 2010), is also subsumed under this type. Four subtypes were identified in the collection: R8Ihf, R8Ihm, R8Ihms, R8Ihc. All of these subtypes were dominated by white quartz phenocrysts. They are primarily distinguished from each other by differences in grain. Respectively, the subtypes correspond to fine-, medium-, medium- and coarse-grained specimens. Subtype R8Ims also contains veins and embayments of silicate. UER makes up 13.7 percent of the combined Stage I and II chipped stone sample. texture and patination. Subtype R3m is non-patinated and it exhibits a dark gray fine-grained matrix devoid of phenocrysts. Subtypes R3mD, R3mDf and R3mDs are patinated and range in color from grayish green to greenish gray to moderate yellow or tan. R3mD represents patinated specimens of R3m, while R3mDf is made up of patinated specimens that exhibit flow banding. R3mDs exhibits blotches and linear arrays of silicate containing biotite and perhaps chlorite elements, respectively. It too, appears to be a patinated form of R3m. The R3/R3t group is composed of an extremely fine grained or microcrystalline dark gray matrix that is semi-translucent when viewed on attenuated edges. Finely disseminated feldspar is commonly seen on exposed surfaces, but the small blotches do not form phenocrysts. R3, R3t, R3f are non-patinated. R3 is distinguished from R3t by a somewhat grainier matrix and greater opacity. R3tf and R3f are the flow-banded versions of R3t and R3, respectively. Subtypes R3p and R3pat are fully patinated. Both are tan in color, but R3p also exhibits finely disseminated biotite specks. R3pat appears to represent patinated specimens of R3t. USR (1) comprises 13.6 percent the Stage I and II chipped stone sample, while USR (2) makes up 42.8 percent. (4) Uwharries Southern Rhyolite (USR). This type includes sources found on Morrow Mountain and surrounding areas, most notably Table Top Mountain. Daniel and Butler (1996:10-13) describe this type as a dark gray, aphanitic, aphyric rhyolite, that commonly exhibits flow-banding. Weathering tends to bring out the flow lines visually, which generally alternate in a pattern of light and dark gray. Although the matrix is usually homogeneous, some specimens exhibit small spherulites of less than 1 mm in diameter (see Stoddard 2006:52). Thin sections indicate a microcrystalline intergrowth of feldspar and quartz, with minor biotite and chlorite elements. The individual minerals are difficult to distinguish and strings of dark minerals can occur along fracture planes. (5) Uwharries Western Rhyolite (UWR). This source area includes specimens from Wolf Den Mountain and Falls Dam (Stoddard 2006:52). This material is described as gray to black microcrystalline felsic volcanic rock that contains rounded plagioclase phenocrysts and glomerocrysts. Green biotite and pale green amphibole are dispersed locally in the matrix and spherulites are sometimes present. Three subtypes were identified in the collection. The main subtype, R3mp, has been patinated to a grayish green or greenish gray color. Non-patinated specimens may have been assigned to one of the other Uwharrie source areas in the analysis, but the distinctive glomerocyst structure suggests that this was not a major problem. Due to patination, most of the specimens in the collection were characterized by fine to medium texture. A single Subtype was identified, UWR comprises 5.2 percent the Stage I and II chipped stone sample. The specimens identified as USR in the 31RH491 assemblage are composed of two larger groupings, the R3m and the R3/R3t groups. These are respectively referred to as USR (2) and USR (1). The R3m Group corresponds to material that typically occurs on Morrow Mountain and includes a series of subtypes that describe variation in flow banding, (6) Indeterminate Metavolcanic. A small amount of the metavolcanic material identified in the analysis could not be further classified to type due to excessive hydration. These were simply classified as “indeterminate metavolcanic” or “hydrated metavol- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 479 Chapter 10. 31RH491 canic.” Twenty-four specimens were assigned to this category in the chipped stone inventory. Quartz Types Pure quartz forms as veins in igneous and metamorphic rock formations characterized by slow crystallization. This generally produces anhedral, milky structures (Spock 1962). In some cases, however, crystallization occurs earlier in the sequence, producing clear crystals with euhedral structure. Precontact groups in the project area utilized both vein and euhedral crystal quartz, but most of it appears to have come from river gravel deposits. A great deal of variation was evident in the vein quartz category ranging from mixed clear and cloudy exposures to opaque milky white. Since these characteristics appeared to be repeatedly present on single cores when debitage concentrations were analyzed, all of this variation was subsumed under the single Subtype Q5 or white quartz. The category “crystal quartz” (CQ) was retained for instances where glass-like examples and crystal facets strongly suggested a euhedral crystal origin. A single piece of reddish brown stained, fractured quartz was recovered and it was identified as fractured “rose quartz.” Staining of this type is commonly observed in quartz cobbles. Examples of fractured white quartz were also identified in the collection and were distinguished as Subtype “Q5 Fract.” White quartz represented 11.0 percent of the combined Stage I and II chipped stone sample, while crystal quartz made up 0.7 percent of this sample. Fractured white quartz contributed 0.2 percent to the Stage I and II sample. In addition to debitage, quartz cobbles and fragments of various types were recovered from the investigation. These included 57 pieces of white quartz and 22 pieces of fractured white quartz that represented fire-cracked rock fragments. Most of these were derived from exploded river cobbles. Approximately 58 percent of the quartz fire-cracked rock sample exhibited cobble cortex. Six items of quartz were also classified as cobble tools. These consisted of two hammer stones (a1399 and a2581), a hammer stone spall (a2620) and three pebbles (a356, a2650 and a3105). Minority Types Several other rock types were recognized in the 31RH491 collection. Descriptions of each are presented below. (1) Arenite. This material has been previously identified as ferruginous sandstone, but thinsectioning of a piece of this material indicates that a more accurate characterization is arenite, or hematite cemented quartz sandstone (Appendix K). Three pieces of arenite were classified as fire-cracked rock. It occurs in a variety of shapes including flattened and irregular nodules. The exterior of the nodules consist of fine-grained ruddy brown precipitate while the centers consist of dense, reddish brown to purple, densley cemented sandstone. The material was commonly used for rock hearths by the precontact inhabitants of Fort Bragg. One hundred and one pieces of arenite were collected during the investigation, totaling 2.284 kg. In addition, one flat piece of arenite was used to make a grooved abrader (a1653) on a smooth precipitate surface. (2) Chert. A single flake fragment (m2173) was composed of a semi-lustrous pale whitish yellow, fossiliferous chert. It resembles the characteristics of cherts found in the Thanetian Black Mingo Formation in the Santee River Valley of South Carolina or in the Allendale formations along the lower Savannah River (Anderson et al. 1982:127-128). Numerous gray to dark gray chert sources have been identified in the Durham-Wadesboro Basin in the eastern Piedmont (Lautzenheizer 1996), but little sourcing work has been conducted on yellowish fossiliferous cherts on the North Carolina Coastal Plain. (3) Granite. A single rock fragment of granite was recovered from what appears to have been a precontact context, at a depth of 30 to 40 cm bs in a Stage II shovel tests at N420/E545. This specimen (m1245) is distinguished from the various indurated granite samples by its crisp, unweathered condition. Local sources for granite are unknown. If the specimen is precontact in association, its presence at the site is likely attributed to importation from the Piedmont or to its inclusion in stream transported gravels from Triassic Basin sediments. The specimen weighed 16.67 gm. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 480 Chapter 10. 31RH491 (4) Grano-diorite. A hammer stone (a314) was made of a rounded stream cobble composed of a mafic, fine-grained, dense igneous rock. The material is tentatively identified as grano-diorite. Grano-diorite is gradational between granite and diorite and is formed when plagioclase replaces orthoclase during magma cooling (Spock 1962:58). (5) Indurated Granite. Six rock fragments of fine-grained granite referred to here as “indurated granite” were recovered at 31RH491 A sample of this material recovered from 31HK2502 was thin-sectioned and identified as a fine-grained granite composed of subhedral plagioclase feldspar, subhedral quartz magnetite, subhedral muscovite and subhedral zircon (Appendix K). Some quartz occurred “as intergrowths within alkali feldspar crystals (this texture is referred to as myrmeckitic texture and is often observed in rocks that have been altered by fluids during metamorphism).” Outcrops of this material are unknown in the region, but Seramur (personal communication, August, 2011) notes that its distinctive structure would make determining its precise source probable. This material was extensively used as temper in portions of the Hanover III and Yadkin III series precontact ceramics recovered from the DO5 package sites. The outer surfaces of the specimens were rounded, suggesting that they were transported from their source by errosional and/or alluvial agents. It is possible that this material occurs as a minor element in stream gravels, perhaps deriving from Triassic Basin sediments. On the other hand, it may have been intentionally imported by potters to use as tempering material for pottery manufacture. If the latter is true, it would provide another line of evidence to suggest that pottery manufacture was undertaken locally. The six specimens from 31RH491 weighed about 204 gm, combined. (6) Metasedimentary. Twenty-two items of metasedimentary material were recovered from the investigation. These included four pieces of black metasedimentary debitage, 18 pieces of green metasedimentary debitage and one utilized flake (a2407) of green metasedimentary material. The relative coarseness of the grain suggests that the material can be further classified as metasiltstone or fine metasandstone. Numerous source areas in the Durham-Wadesboro Basin contain similar rock types, including the Chatham Pittsboro, Person County and Chatham Siler City (Stoddard 2006:57-63). (7) Petrified Wood. One piece of talc-like rock (m1056) was recovered that appeared to exhibit wood-grain contouring, suggesting that it might represent petrified wood. Petrified wood has been reported from Uwharrie National Forest in nearby Montgomery County. Petrified wood might be expected to occur in locations across the Triassic Basin, in the same formations that produce the abundant sandstone conglomerates and arenite and sandstones widespread throughout the Sandhills (Rogers 2006). (8) Sandstone Conglomerate. Also abundantly present in the deposit were chunks of sandstone conglomerate containing a wide range of sand and pebble sizes. Thin-sectioning of a piece of this material determined that individual grains of sand were cemented with clays (wacke) and exhibited udulatory extinction indicating that it originated from metamorphic rock (Appendix K). Occasionally interfaces with arenite could be observed on the chunks, indicating that the two materials were derived from the same formations (e.g. Middendorf). The matrix of the conglomerate consistently exhibited a grayish orange to dusky yellow coloration. Chunks of arenite were sometimes viewed in the matrix as well. This material appears to have been used to make rock hearths. Thirty-two pieces of sandstone conglomerate were collected, weighing about 0.90 kg. (9) Schist. Eighteen pieces of chunky debitage flake fragments and two bifacially flaked tool fragments (a1301 and a2360) were made of a soft, laminar, highly micaceous material identified as schist. The material was yellowish brown to gray in color with dark gray and reddish brown streaks. It did not have a “soapy” feel, although it was clearly of hardened talc-like consistency. This might suggest that material may be more properly identified as a pyrophyllite. The specimen identified as petrified wood might also represent a fragment of this material. (10) Float. Eight small water worn nodules of lithic material were recovered that are worth mentioning because they were either brought to the site by precontact groups or were transported and deposited during flood events. These included three pieces of schist, a green metasedimentary tabular nodule, a fragment of talc/steatite, a piece of hydrated metavol- Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 481 Chapter 10. 31RH491 Table 150. Distribution of Debitage Type by Raw Material Type, 31RH491. DISTRIBUTION OF DEBITAGE TYPE BY RAW MATERIAL TYPE: 31RH491 RAW MATERIAL TYPE BTF FBR CHUNK/ CORTICAL CHUNK CORE FLAKE CORE REJUVINATION FLAKE FLAKE FRAGMENT GRAND TOTAL Metavolcanic Hydrated Metavolcanic 1 1 22 24 65 64 158 73 108 253 Mill Mountain Rhyolite 16 13 Type I Rhyolite Tuff 43 28 Uwharries Eastern Rhyolite 24 18 166 135 343 Uwharries Southern Rhyolite (1) 89 46 68 1 146 350 Uwharries Southern Rhyolite (2) 212 174 872 1 982 2249 9 18 65 166 Uwharries Western Rhyolite 1 8 74 Metasedimentary Black Metasedimentary 2 2 4 Green Metasedimentary 3 13 16 2 10 14 63 191 273 3 3 Quartz Crystal Quartz 2 White Quartz 14 3 2 White Quartz, Fractured Miscellaneous Fossiliferous Chert Schist GRAND TOTAL 409 301 11 1389 2 1 1 18 18 1760 3872 Table 151. Percent Cortex on Debitage and Percent Core Flakes by Raw Material Type, 31RH491. PERCENT CORTEX AND CORE FLAKES BY RAW MATERIAL: 31RH491 RAW MATERIAL TYPE TOTAL DEBITAGE PERCENT W/CORTEX TOTAL IDENTIFIABLE DEBITAGE PERCENT CORE FLAKES MEAN CORE FLAKE WEIGHT (gm) Mill Mountain Rhyolite 158 7.50 94 69.15 0.99 Type I Rhyolite Tuff 253 8.30 144 50.69 0.39 Uwharries Eastern Rhyolite 343 4.96 208 79.81 1.03 Uwharries Southern Rhyolite (1) 350 10.40 204 33.82 0.58 Uwharries Southern Rhyolite (2) 2249 12.54 1260 69.29 0.72 Uwharries Western Rhyolite 166 9.21 101 73.27 1.06 White Quartz 273 5.13 80 78.75 1.76 TOTAL 3792 2091 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 482 Chapter 10. 31RH491 canic material and two small nodules of URS (2) material. If the metavolcanic “float” were brought to the site by natural forces or as unintentional inclusions within rock hearth bundles, it would suggest that Uwharries metavolcanic material could be present in stream terrace formations in the area or in exposures of Triassic Basin sediments crosscut by streams. Rogers (2006) discusses the distribution of Carolina Slate Belt elements and the Triassic Basin in locations near Fort Bragg. Lithic Artifacts Basic definitions of the various lithic artifact classes were presented in Chapter 4. Here, specific descriptions of the 31RH478 lithic collection are presented. Debitage The debitage classification is adapted from formal (Bradley 1973; Frison and Bradley 1980; House and Wogaman 1978; Newcomer 1971), attribute (Moore 2002; Shott 1994), and mass (Ahler 1989; Moore 2002) analysis approaches. Debitage classes were devised to identify both reduction stages and reduction/production systems (ie. biface core reduction, directional core reduction, and flake blank production). This classification was supplemented with the recording of three attributes: (1) size class, (2) percent cortex, and (3) condition. Table 150 presents the complete inventory (all field stages) of debitage type by raw material type from 31RH491. Directional core reduction types (Core Flakes and Core Rejuvination Flakes) comprise the majority of identifiable debitage. Biface reduction types (BTFs and FBRs) comprise only about 34 percent of the identifiable debitage types. The distribution of cortex on dorsal faces and platforms of debitage provides a basis for further evaluating the nature of reduction strategies associated with the various raw material types (Table 151). Dibble et al. (2005:550) indicate that raw nodule reduction from bifaces should yield between about 6 and 21 percent flakes with cortex for any single episode. It is likely that similar proportions characterize directional core reduction from raw nodules. MMR, Type I RT, USR (1), USR (2) and UWR raw material types display profiles consistent with raw nodule reduction, while UER and white quartz exhibit profiles more in keeping with prepared directional core reduction. Bifacial thinning occurs in all of the raw material types, but this activity is generally overwhelmed by directional core reduction. Mean core flake weight values further suggest that Type I RT and USR (1) are characterized by small raw nodules or prepared directional cores, while the other raw material types have larger core flakes more indicative of larger nodule and prepared core size. Cores Metric and attribute data on cores are provided in Appendix C. Three classes of cores were identified in the 31RH491 collection: (1) Directional Cores, (2) Core Fragments and (3) Flake Blanks. Each class is discussed below and summary data on cores are presented in Table 152. (1) Directional Cores. Directional, or block, cores consist of relatively large masses of material that exhibit flake scars running in one or more directions from flat striking platforms rather than from biface edges. The platform angle approximates 90° on early stage cores, but specimens nearing exhaustion sometimes exhibit more acute angles. Directional cores can be further divided into those with scars emanating from only one platform (unidirectional) and cores with multiple platforms (multidirectional). Generally, multidirectional cores represent the final stage of core use as new platforms are formed to extend their lives. In addition, directional cores are sometimes split in two in the later life stages to create thick flake blanks, either by intention or because of a flake production error. These are called “split cores.” Two directional core fragments were recognized in the collection (Figure 173:A-B), respectively made of USR (2) and USR (1). Two split cores were also identified, one made of crystal quartz in which the faceted cortex of the crystal is still present on the dorsal face (Figure 173:C) and the other of white quartz (Figure 173:D). The morphology of the original directional cores is preserved on the dorsal faces of both specimens. (2) Core Fragments. Angular pieces of raw material containing portions of one or more flake Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 483 Chapter 10. 31RH491 Table 152. Summary Data for Cores, 31RH491. CORES: 31RH491 SPECIMEN # PORTION FLAKE TYPE LITHIC RAW MATERIAL TYPE MAXIMUM THICKNESS (mm) MAXIMUM LENGTH (mm) MAXIMUM WIDTH (mm) WEIGHT (gm) BAG CORE TYPE 271270a3188 805 Directional Core Frag. Fragment Uwharries Southern Rhyolite (1) 26.55 34.94 23.00 22.79 271270a2707 1254 Directional Core Frag. Fragment Uwharries Southern Rhyolite (2) 49.92 (32.65) (19.86) 53.23 271270a2625 1238 Split Core Whole Crystal Quartz 35.67 30.53 17.16 20.56 271270a2089 1031 Split Core Fragment White Quartz 22.10 42.81 20.95 26.45 16.29 271270a64 32 Core Fragment Fragment Mill Mountain Rhyolite 17.81 17.95 38.1) 271270a126 67 Core Fragment Fragment Uwharries Eastern Rhyolite 19.09 9.44 40.42 6.12 271270a353 189 Core Fragment Fragment White Quartz 20.22 16.12 10.63 3.28 271270a436 234.1 Core Fragment Fragment Mill Mountain Rhyolite 30.99 20.51 15.04 10.23 271270a681 328 Core Fragment Fragment Type I Rhyolite Tuff 24.82 24.65 10.45 4.6 271270a826 398 Core Fragment Fragment White Quartz 18.85 24.94 17.06 9.28 271270a840 404 Core Fragment Fragment Metasedimentary, green 34.79 26.94 26.98 11.62 271270a1006 493 Core Fragment Fragment Uwharries Southern Rhyolite (1) 10.70 21.49 17.49 4.13 271270a1024 502 Core Fragment Fragment White Quartz 13.71 24.44 19.08 6.83 271270a1316 647 Core Fragment Fragment Uwharries Eastern Rhyolite 39.85 39.30 17.25 33.17 271270a1339 664 Core Fragment Fragment White Quartz 13.84 15.77 15.53 3.94 271270a1393 698 Core Fragment Fragment Uwharries Eastern Rhyolite 17.21 21.32 11.54 3.95 271270a1427 725 Core Fragment Fragment Uwharries Eastern Rhyolite 25.06 22.66 13.25 6.01 271270a1645 852 Core Fragment Fragment Crystal Quartz 16.20 20.64 12.76 3.96 271270a2713 1255 Core Fragment Fragment White Quartz (40.89) (22.90) (20.93) 22.51 271270a2738 1257 Core Fragment Fragment Uwharries Eastern Rhyolite (30.47) (31.72) (19.45) 21.11 271270a14 8 Flake Blank Fragment Core Flake Metasedimentary, green 13.13 57.56 (45.49) 31.88 271270a110 57 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (2) 11.06 48.51 27.47 14.18 271270a241 128 Flake Blank Whole Core Flake Uwharries Eastern Rhyolite 6.66 42.63 22.73 6.83 271270a677 327 Flake Blank Whole Core Flake White Quartz 19.84 42.50 39.32 35.49 271270a717 346 Flake Blank Fragment Core Flake White Quartz 8.33 35.66 29.81 8.62 271270a760 364 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (1) 3.84 30.21 16.87 2.22 271270a1188 598 Flake Blank Whole Blade Flake Uwharries Southern Rhyolite (1) 2.28 27.97 12.09 0.75 271270a1290 634 Flake Blank Whole Blade Flake Uwharries Southern Rhyolite (2) 3.41 32.86 16.81 1.47 271270a1459 741 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (1) 8.35 28.69 29.64 8.3 271270a1512 769 Flake Blank Fragment Core Flake Crystal Quartz 3.38 (20.90) (20.19) 1.44 271270a1605 839 Flake Blank Whole Bladelet Uwharries Southern Rhyolite (2) 271270a1680 867 Flake Blank Whole Core Flake White Quartz 271270a1865 909 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (2) 4.71 35.06 35.12 5.5 271270a1903 926 Flake Blank Whole Core Flake Mill Mountain Rhyolite 8.40 56.98 26.16 9.81 271270a2100 1036 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (2) 3.17 31.30 18.37 2.29 271270a2149 1057 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (1) 4.68 30.62 27.94 3.97 271270a2256 1110 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (2) 2.41 32.84 18.08 1.65 271270a2300 1126 Flake Blank Fragment Blade Flake White Quartz 8.16 (23.39) (26.13) 3.88 271270a2319 1136 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (2) 10.62 56.13 17.39 15.12 271270a2349 1151 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (1) 5.42 27.55 25.85 3.03 271270a2370 1161 Flake Blank Fragment Uwharries Southern Rhyolite (1) 8.92 (27.05) (37.03) 7.17 271270a2379 1166 Flake Blank Fragment Blade Flake Uwharries Southern Rhyolite (2) 3.07 (35.29) 18.90 2.06 271270a2422 1178 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (2) 5.50 41.08 28.19 4.13 271270a2458 1192 Flake Blank Whole FBR Type I Rhyolite Tuff 2.18 22.21 23.85 1.49 271270a2501 1216 Flake Blank Whole Core Flake White Quartz 7.84 28.25 30.76 6.85 271270a2554 1232 Flake Blank Fragment Core Flake Uwharries Southern Rhyolite (2) 4.88 (27.09) 22.79 3.2 271270a2592 1236 Flake Blank Whole Core Flake Uwharries Southern Rhyolite (2) 5.79 37.59 24.81 3.99 1.92 22.34 8.47 0.42 14.17 44.25 42.19 24.77 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 484 Chapter 10. 31RH491 Figure 173. Cores and Type I Bifaces, Site 31RH491. A-B: Directional Core Fragments ( a2707 and a3188 respectively), C-D: Split Core (a2625 and a2089 respectively), E-I: Type I Bifaces (a723, a1146, a1149-Ventral Face, a1149-Dorsal Face and a763 respectively). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 485 486 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC BAG 348 367 576 577 369 304 476 478 699 771 917 1240 1257 697 194 258 1128 456 509 642 1156 SPECIMEN # 271270a723 271270a763 271270a1146 271270a1149 271270a770 271270a630 271270a970 271270a979 271270a1394 271270a1526 271270a1890 271270a2652 271270a2752 271270a1387 271270a363 271270a499 271270a2304 271270a929 271270a1037 271270a1301 271270a2360 Worked Schist Fragment Worked Schist Hardaway Blade Chipped Stone Axe Biface Fragment Type III Biface Fragment Type III Biface Fragment Type II Biface/Preform Type II Biface Fragment Type II Biface Fragment Type II Biface Fragment Type II Biface Fragment Type II Biface Fragment Type II Biface Fragment Type II Biface Fragment Type II Biface Fragment Type I Biface Fragment Type I Biface Type I Biface Type I Biface Type I Biface TYPE Lateral Section End Whole End Missing Lateral Section Ovoid Ovoid Paleo-indian/ Early Archaic Schist Schist Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Uwharries Eastern Rhyolite Large CornerNotched Point? Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) White Quartz Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Uwharries Western Rhyolite Uwharries Western Rhyolite Uwharries Southern Rhyolite (1) Uwharries Western Rhyolite LITHIC RAW MATERIAL TYPE Blade/Shoulder Section Morrow Mt. Stemmed? ASSOCIATION Uwharries Southern Rhyolite (1) Lanceolate Ovoid Ovoid Lanceolate Ovate Ovoid Ovoid Ovoid SHAPE Lateral Section Whole End End Lateral Section Fragment END End End End End Missing Whole Whole End Missing Whole PORTION BIFACES: 31RH491 Table 153. Summary Data for Bifaces, 31RH491. (10.37) 9.55 7.54 15.77 7.2 5.53 6.31 32.84 (8.18) 9.61 (7.49) (9.93) (13.77) 4.89 5.05 6.12 10.62 13.88 22.03 5.61 14.02 Maximum Thickness (mm) (72.36) (55.40) 36.46 (52.19) (15.61) (30.52) (16.73) 107.13 (26.21) (26.09) (33.69) (28.01) (27.08) (13.27) (14.25) (9.44) (32.23) 66.93 64.62 29.63 45.93 Maximum Length (mm) (27.60) (66.67) 31.15 47.37 (11.10) (34.76) 21.42 43.93 (14.98) (24.65) (17.52) (25.72) (24.28) 17.00 (26.98) 16.31 17.21 40.45 50.52 (21.70) 40.66 Maximum Width (mm) 26.06 54.79 10.29 50.61 1.1 6 2.55 105.19 1.99 7.97 3.55 6.29 7.86 1.36 1.87 0.85 7.19 45.84 58.82 4.78 30.58 Weight (gm) Chapter 10. 31RH491 Chapter 10. 31RH491 scars suggestive of directional core reduction, but which lacked striking platforms, were classified as core fragments. Sixteen such fragments were recognized in the collection, six of which were composed of quartz and one was made of green, laminated metasedimentary material. The remaining specimens were made of various metavolcanic raw material types, including five UER, two MMR, one Type I RT and one USR (1) (Table 152). The specimens are all small and, although they were most probably derived from directional cores. Sections of the platforms from the original directional cores were observed on 11 of the specimens. (3) Flake Blanks. Flake blanks were removed from a core for the purpose of further modification or use as a tool (Bradley 1973:6). They do not exhibit obvious macroscopic evidence of use or modification, but they may have been used for a brief activity. Twenty-seven flake blanks were identified in the collection (Table 152). Most of these were composed of metavolcanic materials, but one large white quartz core flake fragment was also included in this category. Blade flakes (flakes approximately twice as long as they are wide), core flake flakes, FBRs and bladelets are represented in the collection. Some of the flake blanks were small. Five of the whole specimens weighed less than 2 gm, but most of the collection consisted of larger core flakes weighing more than5 gm each. The smaller specimens owe their membership in this category to the fact that they manifest the characteristics of a utilized flake, although evidence of wear could not be identified. The dorsal surfaces of most of the flake blanks reveal the former facial structure of directional cores. Bifaces Twenty-one biface or biface fragments were recovered during the investigation (Table 153). Bifaces are derived from both large mass packages (i.e. unmodified and/or prepared cores) and flake blanks. Determination that a biface was derived from a flake blank was made when flake characteristics were observed on the item, such as remnant striking platforms, flake curvature, or bulbs of percussion from flake detachment. Most of the bifaces in the collection do not have sufficient mass to suggest that they were derived from large biface cores. Instead, most appear to have been produced from flake blanks of varying sizes. Each biface category as adapted from Daniel (2002:51-54) is described below. (1) Type I Bifaces. Daniel (2002:51) describes Type I Bifaces as irregularly shaped flake blanks and core masses with large conchoidal scars forming at least one section of a sinuous edge along the core edge. No evidence of secondary thinning scars is observable and a relatively high thickness to width ratio obtains. Five Type I Biface fragments were recognized in the collection (Figure 173:E-I). All of these appear to have been derived from flake blanks of various sizes. The platform and dorsal surface morphology of large flake blanks is clearly displayed on the reverse sides of two ovoid/ovate Type I Bifaces (Figure 173:F-H). These specimens (a1146 and a1149) were recovered together between 0 and 20 cm bs in a Stage II shovel test excavated at N405/E535. These bifaces were composed of UWR material and their co-association suggests that they may have been cached for future use. Another Type I Biface (a723) of UWR (Figure 173:E) was recovered high in the site matrix as well. It was found between 0 and 10 cm bs in a Stage I shovel test excavated at N490/E495. The other two specimens were small and made of USR (1) material (Figure 173:I). These specimens (a763 and a770) were also recovered from a single shovel test (N465/E465) between 20 and 50 cm bs. USR (1) raw material tends to be found at lower depths across the site and is strongly associated with Early Archaic occupations. These items were manufacturing rejects and appear to have been deposited as a consequence of manufacturing rather than caching behavior. (2) Type II Bifaces. These bifaces, in the main, represent intermediate stage preform rejects. They are roughly flaked and exhibit only large, conchoidal flake scars. No edge trimming is evident, but an ovate to lanceolate outline is generally recognizable and the perimeter of the edge margin is entirely formed. Facial retouch, however, may not be complete and features of the original blank are sometimes evident. Nine Type II Biface fragments were identified in the collection. Most of these represent small lateral and end sections, but one large specimen (a1387) was whole and appears to affect the outline of a Morrow Mountain Stemmed point (Figure 174:A). It was made of UER material, which is consistent Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 487 Chapter 10. 31RH491 Figure 174. Type II and III Bifaces and Biface Tools, Site 31RH491. A: Type II Biface/Preform (a1387, Morrow Mountain Stemmed Preform?), B: Type III Biface Fragment (a499, Palmer II/III Preform Reject?), C: Type III Biface/Hardaway Blade (a1037), D-E: Chipped Stone Axe (a929, D: Dorsal face, E: Ventral Face), F: Worked Schist (a1301). with a Middle Archaic association as porphyries are much more frequently identified in Middle and Late Archaic point collections than they are in Early Archaic ones. The specimen was recovered from a Stage II shovel test (N363/E598) at a depth of 10 to 20 cm bs. Only large conchoidal flake scars occur along the biface margins and the edges are sharp, displaying no evidence of utilization. The item was apparently abandoned due to a large mass of material on one face that could not be successfully removed as indicated by numerous step fractures on one side of the face. Whether it was cached for possible future use is undetermined, but its presence suggests that the manufacturers were not operating under a preconceived notion of extreme raw material scarcity. (3) Type III Bifaces. Type III Bifaces represent late stage preform rejects. The edges exhibit Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 488 Chapter 10. 31RH491 secondary thinning and shaping scars and facial retouch is globally distributed. Outline shapes are well defined and reflect ovate and lanceolate forms. Most of these specimens display step and hinge fractures that terminated production. Type III Bifaces tend to have the lowest ratios of thickness-to-width. Sometimes features of diagnostic projectile point haft elements are present, allowing correlation with specific culture-historic types. In these cases, specimens are differentiated and identified as “Preforms.” Two Type III Bifaces were identified in the collection (Table 153). One was a small lateral section, but the other was a larger specimen (a499) that appears to consist of blade and shoulder section of a corner notched point preform (Figure 174:B). Secondary flaking and use wear were not observed along the blade margin, suggesting that the item represented a preform reject rather than a broken finished tool. The specimen was found at a depth of 40 to 50 cm bs in a Stage II shovel test (N410/E505) that formed the anchor for SB7. This depth range is within the Early Archaic horizon at the site and is consistent with the expected vertical position for Palmer Corner Notched points, of which the specimen appears to represent a preform reject. The broad width of the tang further suggests a Palmer II or III association. (4) Hardaway Blade. A whole, ovoid Type III Biface (a1037) that exhibits the characteristics of a Hardaway Blade (Coe 1964:64-65) was recovered in the collection (Figure 174:C). The basal outline was slightly convex and sharply differentiated from the blade margins by pointed “shoulders.” Long flute-like thinning flake channels are evident on one face, but these flakes originated from the presumptive top of the blade rather than the base. The Hardaway Blade was recovered in the lowest levels of the Hardaway Site in primary association with the Hardaway Dalton point style. Coe’s illustrated examples range from deeply concave bases with flaring ears similar in form to the Quad point to less elaborate ovate to lanceolate forms with slightly concave to straight bases. Similar Quad-like, ovate bifaces with flared ears were recovered in the lower levels at the Haw River Site as well (Claggett and Cable 1982). Moreover, excavations at 31HT690 on Fort Bragg have produced a number of well made ovate and squat lanceolate Type III Bifaces at the base of the Early Archaic block (Cable 2008:147-155). These specimens exhibited both concave and convex bases, suggesting that basal concavities were formed during the final stages of manufacture, which may have been aided by basal fluting as a thinning strategy. Two specimens from 31HK2521 appear to also fit into this category and one of these (a231) exhibited a basal thinning flute on one face (Figure 80:F). The specimen from 31RH491 was found at a vertical position consistent with an Early Archaic/Late Paleoindian context, between 40 and 50 cm bs, in a Stage II shovel test (N515/E530) that formed the anchor for SB2. (5) Chipped Stone Axe. A tool manufactured on a large “quarry blade” or flake blank is identified as a chipped stone axe (Figure 174:D-E, Table 153). The specimen (a929) is largely a unifacial tool, but it is included in the biface tool category because it was apparently hafted and the margins of the remaining end were bifacially flaked. Opposing notches formed by unifacial percussion flaking occur approximately one-third of the way along the inferred longitudinal axis of the tool from the remaining end. The other end was separated by a clean step fracture. The fact that the hafting notches are positioned closer to the remaining end suggests that it originally served as the butt end of the tool. The bifacial margin, however, exhibits edge dulling indicative of use. This may have occurred once the original bit was broken off, allowing the tool to remain hafted. It was found in a Stage II shovel test (N515/E495) at a depth of 20 to 30 cm bs, suggesting a Late or Middle Archaic association. The specimen was manufactured from USR (2) material and is similar in technology to the Guilford Axe described by Coe (1964:113). The notches are not quite as wide as the illustrated examples and the projected expansion of the missing end does not seem as extreme. Its dimensions are on the low end of the reported metrics for the Guilford Axe. (6) Worked Schist. Two talc-schist specimens (a1301 and a2360) exhibited edges that were roughly flaked by percussion producing bifacial edges. Large pieces of roughly flaked, laminated raw material like this has been used as knives to cut and saw semi-resistant plant material with fleshy and fibrous matrices in certain archaeological cultures. Large flat phyllite knives were used to cut and harvest agave, for instance, in southern Arizona during precontact times (Fish and Fish 1994). The larger specimen (a1301) is an end fragment from an original piece that is projected to have been as much as 100 to Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 489 848 1297 130 1231 952 456 722 1218 1233 981 1201 473 470 870 239 469 735 756 857 882 961 1013 1014 1257 271270a1637 271270a3111 271270a244 271270a2544 271270a1941 271270a930 271270a1443 271270a2503 271270a2563 271270a1998 271270a2484 271270a957 271270a947 271270a1690 271270a452 271270a945 271270a1451 271270a1484 271270a1661 271270a1755 271270a1967 271270a2054 271270a2059 271270a2739 1240 271270a2651 40 271270a76 696 318 271270a659 271270a1386 BAG SPECIMEN No. PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment PPK Fragment Yadkin Large Triang., Var. B Yadkin Large Triang., Var. B Savannah River St. Preform Savannah River St. Savannah River St. Guilford Lanceolate Contracting St. Base Morrow Mountain I St. Morrow Mountain II St. Kirk/Stanley St. Base Square Stem Kirk/Stanley St. Base Waller Knife Waller Knife Corner Notched, indet. Palmer I Corner Notch. Palmer I Corner Notch. Taylor Side Notched TYPE Upper Blade Midsection Tip Midsection Tip Upper Blade Tip Lateral Section Midsection Midsection Tip Tip Missing Base Base Base Tip Missing Tip Missing Base Whole Base Base Base Base Base Whole Midsection Whole Tip Missing Base PORTION Type I RT MMR USR (2) USR (2) MMR UER Type I RT USR (2) MMR MMR MMR MMR USR (2) USR (2) USR (2) Type I RT MMR USR (2) UER USR (2) USR (2) USR (1) Type I RT White Qtz. USR (1) USR (2) USR (1) Cryst. Qtz. LITHIC RAW MATERIA L TYPE WEIGHT (gm) 3.75 4.55 0.56 0.49 5.61 0.89 0.82 4.15 0.86 1.25 4.25 0.79 15.09 10.9 20.05 13.42 2.51 2.85 3.22 5.46 2.74 2.48 0.62 3.43 12.8 2.92 3.82 1.42 MAXIMUM THICKNESS (mm) (6.76) (8.55) (4.84) (5.46) 6.85 (4.41) (4.69) (7.02) (4.57) 7.86 6.18 (4.25) 9.60 (9.15) 12.52 10.56 (6.24) 6.94 5.52 (7.86) (7.24) (6.64) (2.68) 6.00 3.34 5.28 6.44 (6.70) MAXIMUM LENGTH (mm) (31.43) (27.21) (10.22) (14.83) (36.19) (22.12) (22.34) (23.50) (15.60) (17.93) (30.27) (14.28) (44.67) (37.58) (65.47) (57.10) (19.85) 30.62 22.78 (25.79) (22.52) (17.48) (14.41) 28.49 (26.78) 31.10 (27.26) (11.23) 24.55 18.50 15.18 13.89 6.76 6.42 17.50 17.34 13.72 17.57 16.80 18.01 BASAL WIDTH (mm) PROJECTILE POINTS: 31RH491 TANG WIDTH (mm) 15.71 16.19 17.23 10.56 11.64 18.50 17.40 12.71 13.83 13.29 11.20 15.32 33.21 34.09 28.93 20.68 17.63 (27.16) 19.32 24.01 (18.79) (23.64) SHOULDER WIDTH (mm) 490 17.71 9.86 15.50 9.56 9.57 9.47 10.31 6.09 6.26 8.62 8.64 7.76 TANG LENGTH (mm) Table 154. Summary Data for Projectile Points, 31RH491. BLADE LENGTH (mm) 21.06 22.23 22.48 BASAL GRINDING Present Present Present Present Present Present Present BLADE SERRATION Present Present Present EDGE ANGLE 52° 61° 51° 55° 54° 55° 58° 45° 48° 66° 56° 56° 48° 64° 58° 50° 54° 58° 39° 49° 32° 49° 64° Concave Concave Concave/Fishtail Straight Straight Rounded Rounded Concave/Bifurcated Straight Concave/Bifurcated Straight Straight/Irregular Straight Straight Straight BASE SHAPE Chapter 10. 31RH491 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC Chapter 10. 31RH491 Figure 175. Projectile Points, Site 31RH491. A: Taylor Side Notched (a659), B-C: Palmer I Corner Notched (a76 and a1386 respectively), D-E: Corner Notched, indeterminate (a2651, Ventral and Dorsal Faces respectively), F-I: Waller Knives (a1637, Dorsal and Ventral Face respectively and a3111, Dorsal and Ventral Faces respectively), J-L: Kirk/Stanly Stemmed Bases? (a244, a1941 and a2544 respectively), M: Morrow Mountain II Stemmed (a930), N: Morrow Mountain I Stemmed (a1443), O: Base Fragment, indeterminate contracting stem (a2503), P: Guilford Lanceolate, Straight Base Variant (a2563), Q-R: Savannah River Stemmed (a1998 and a2484 respectively), S: Savannah River Stemmed Preform Reject (a957), T-U: Yadkin Large Triangular, Variant B (a1690 and a947 respectively). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 491 Chapter 10. 31RH491 120 mm in length (Figure 174:F). It was recovered between 20 and 30 cm bs in a Stage II shovel test (N385/E581) anchoring SB5. The smaller fragment (a2360) was found in a Stage III shovel test in SB3, at a depth of 20 to 30 cm bs as well. Projectile Points Projectile points are bifacially flaked tools with retouched haft elements. In addition, projectile point fragments consist of specimens with the fine bifacial retouch characteristic of finished projectile points, but not necessarily exhibiting evidence of a haft element. The collection from 31RH491 consisted of 18 diagnostic or partly diagnostic projectile points and 10 fragments (Table 154). Metric and descriptive data on the projectile points can be found in Appendix E. Descriptions of specific point types are presented below in chronological order from earliest to latest. (1) Taylor Side Notched. A single, crystal quartz base (a659) recovered from the 31RH491 investigation was identified as a Taylor Side Notched point (Figure 175:A). Although this type is more frequently found in South Carolina, its presence in North Carolina is gaining increasing recognition. Michie (1966:123-124; 1996:253-254) defined the type as an outcome of his excavation at the Taylor Site in central South Carolina. Morphological characteristics of the type include basal grinding, squared ears, deep side notches and alternate blade beveling. Taylor points have not been recovered in stratigraphic contexts, nor are there associated radiocarbon dates available. However, it is generally accepted that the morphological characteristics of the type place it in the very early range of the Early Archaic period, in a time horizon correlated with Bolen (Bullen 1975) and Big Sandy (DeJarnette et al. 1962:48; Lewis and Kneberg 1961:38) points. Michie suggests a chronological position of around 10,000 BP for the Taylor Side Notched point. The morphological characteristics of the point type place it between the Hardaway Side Notched and Small Dalton cluster on the early end and Palmer Corner Notched on the late end. Daniel (1998:53) recognized one specimen from the Hardaway site that resembles the type. The blade of the specimen was not beveled, as is common for the South Carolina sample. The specimen from 31RH491 exhibited a squared ear and U-shaped side notch on the complete side of the base and a slightly concave, ground base. The morphology of the blade was not observable. In these characteristics, the specimen also resembles the Bollen point type illustrated by Daniel (1998:54, Figure 4.3:B). Specimen a659 was manufactured from Western Uwharries Rhyolite. It was recovered at a depth of only 20 to 30 cm bs in a Stage II shovel test (N475/E470) that anchored SB9. This depth is “out of position” for Early Archaic points and suggests displacement due to bioturbation or other disturbance process. This particular location was situated within the wildlife plot on the north end of the site. (2) Palmer I Corner Notched. Coe (1964:67) described the Palmer Corner Notched type as a “small corner-notched blade with a straight, ground base and pronounced serrations.” Blade edges were generally straight, but were secondarily both incurved and excurvate. Bases were commonly straight and typically ground. Serrations were at times deep, averaging about 3 mm in width and 5 to 7 mm in length. The width of the shoulder barbs usually exceeded that of the base. Metric dimensions ranged between 28 and 60 mm in length, 15 and 25 mm in width, and 5 to 12 mm in thickness. At the Hardaway Site, Palmer Corner-Notched points were primarily associated with Level III, sandwiched between Hardaway SideNotched and Hardaway-Dalton styles in Level 4 and Kirk Serrated and Kirk Corner-Notched in Level II. As an outcome of the Haw River excavations, Cable (1982) further divided the Palmer type into three variants, Palmer I, II and III. The Palmer I style is smaller, typically exhibits ground bases and was found in a slightly lower vertical position at Haw River than the larger variants described as Palmer II and Palmer III. The latter is the largest of the three variants and is characterized by a low incidence of basal grinding. It would correspond closely to the ideal type description for the Kirk Corner Notched type (Coe 1964:69-70). Palmer II is stylistically intermediate and it may overlap the temporal distributions of the other variants. It is larger than Palmer I specimens, but is characterized by a high incidence of basal grinding. Others suggest that all of this variation is largely a consequence of life-stage variation rather than temporal differences (Daniel 1998; Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 492 Chapter 10. 31RH491 Sassaman 2002). Although this may be the case for Palmer II and III, this explanation does not take into account the much narrower tang widths of Palmer I specimens. The time range of this style is generally accepted at 9500–8900 BP (Sassaman et al. 2002:10). Two projectile point specimens from 31RH491 were identified as Palmer I Corner Notched. Specimen a76 (Figure 175:B) was recovered between 30 and 40 cm bs in a Stage I shovel test (N490/E490) excavated in SB12. It was manufactured from USR (1) material. The distal ears of the shoulder barbs and the upper blade were missing. Shallow serrations have been worn down through use on both blade edges. Blade outlines were slightly excurvate to straight and the base was heavily ground with a straight outline. Specimen a1368 (Figure 175:C) was found at a depth of 40 to 50 cm bs in a Stage II shovel test (N355/E395). It was manufactured from USR (2) material and was essentially whole. The distal end of one shoulder ear was missing. The blade margins were excurvate in outline and one was serrated. The other margin is more heavily resharpened, which may have removed evidence of serrations. Resharpening appears to have resulted in the formation of an appiculated, or awl-like, tip. The base was straight and heavily ground. Another corner notched projectile point midsection appears to represent a Palmer point, as well, but the base is missing and consequently it cannot be completely typed (Figure 175:D-E). It was composed of USR (1) material and was found at a depth of 40 to 50 cm bd in TU1. The specimen (a2651) is carefully pressure flaked on one face in a parallel pattern, while the reverse side is only very cursorily flaked, leaving a flat ventral face and patches of cortex. As a consequence, the blade cross-section is plano-convex like those of many scraper forms. However, the edge that was produced is delicate and intended for cutting functions as is attested to by the extremely acute edge angle of the blade, which averaged about 32°. Most projectile points with biconvex cross-sections have much higher edge angles. The two Palmer I Corner Notched points display average edge angles of 49° to 64° and are clearly made to withstand much more rigorous cutting applications. The blade margin outlines range from slightly excurvate to recurved on the specimen and its tang width is consistent with a Palmer I or Palmer II style. (3) Waller Knife. Two specimens in the collection correspond to the morphological characteristics of a notched, unifacially flaked stone tool known as the Waller Knife (Figure 175:F-I). It was first recognized in assemblages from Florida (Bullen and Bielman 1973; Purdy 1981:31-32; Waller 1971:173174), but has since been identified in a number of Early Archaic contexts along the Southeast Gulf and Atlantic coasts (Anderson and Schuldenrein 1985:292; Geiger and Brown 1983; Goodyear et al. 1979). Daniel (1998:103-104) also identified a Waller Knife in his reanalysis of the Hardaway Site lithic assemblage. This particular specimen displayed wide U-shaped notches that would not technically represent side notches, but rather more closely resemble corner notches. Daniel observes that the delicate unifacial edge would not have been appropriate for heavy scraper functions, but would have been more closely aligned with cutting soft materials. The two specimens from 31RH491 were made on small flakes and were shaped exclusively by unifacial retouch. Specimen a1637 was made of white quartz and exhibits broad corner notching and other characteristics resembling the Palmer Corner Notched type (Figure 175:F-G). The blade margins are excurvate and display very fine serrations, while the base is not ground. An inverse bulb of percussion emanates from the base which also appears to exhibit evidence of platform collapse, suggesting that the specimen may have once been a traditional Palmer Corner Notched point that was subsequently split by bipolar percussion perfectly down its medial axis. The tool was found in a Stage III shovel test in SB10 at a depth of 40 to 50 cm bs, clearly in an Early Archaic vertical position. The other specimen (a3111) was made on a thin flake of Type I RT and displays a clear, shallow side notch (Figure 175:H-I). The observable blade portion is excurvate in outline and the straight base is lightly ground. This specimen was recovered between 20 and 30 cm bd in TU3 and appears to be somewhat out of position based on its clear Early Archaic morphology. Just as was true of the corner notched point discussed above with the plano-convex cross-section, these specimens exhibit acute edge angles (Table 154) ranging between 32° and 49°. This tool type may very well crosscut Early Archaic point styles and ultimately be shown to be a little appreciated feature of side notched and corner notched point assemblages throughout the Southeast. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 493 Chapter 10. 31RH491 (4) Kirk/Stanly Stemmed Bases. Three square stem fragments with moderately concave bases appear too represent fragments of Stanly or Kirk Stemmed point styles (Figure 175:J-L). Kirk Stemmed and Serrated (Coe 1964:70) and Stanly Stemmed (Coe 1964:35) are characterized by square to slightly expanding stems with either straight or concave basal outlines. Although Savannah River Stemmed bases also exhibit these qualities, the specimens identified here appear to be more finely flaked in the Kirk/Stanly tradition. Moreover, the basal margins of each of these specimens display light to heavy grinding. Coe specifically observes that grinding is absent on Stanly Stemmed, but fails to discuss this feature for Kirk Stemmed and Serrated. All of these specimens were recovered in vertical contexts between 10 and 30 cm bs/bd. (5) Morrow Mountain I Stemmed. A single Morrow Mountain I Stemmed point (a1443) was identified in the collection from 31RH491 (Figure 175:N). The Morrow Mountain I Stemmed type was originally defined by Coe (1964:37-43) from his excavations at the Doershuk Site. The type is similar in morphology to the Morrow Mountain II Stemmed point. Both styles were characterized by contracting stems, but two differences were cited. The blade is proportionately longer and narrower and a noticeable angle occurs at the shoulder/stem juncture in the Morrow Mountain II Stemmed type. The shoulder of the Morrow Mountain I style, by contrast, curved into the stem without a break in the marginal contour. There is some evidence to suggest that Morrow Mountain II occurred at a slightly higher stratigraphic/vertical position than Type I in the Doershuk Site excavations, which led Coe (1964:43) to tentatively posit that the former was “transitional and survived to a later date.” This type is similar in length to the Morrow Mountain II Stemmed type (30–80 mm), but is noticeably broader at the shoulder (18–30 mm). Specimen a1443 was whole and occupied a position on the short end of the range of length measurements reported for the type. Its maximum length was 30.62 mm. Slight shifts in the orientation of the upper stem margins are observable near the juncture with the blade, but the shift was not deemed sufficiently distinct to merit assignment of the point to the Morrow Mountain II style. It was composed of USR (2) material. The blade margins were very slightly excurvate. The specimen was found in a Stage II shovel test anchoring SB15 at a depth of 30 to 40 cm bs. (5) Morrow Mountain II Stemmed. One specimen of this style was recovered from 31RH491 (Figure 175:M). The Morrow Mountain II Stemmed type was originally defined by Coe (1964:37-43) from his excavations at the Doershuk Site. It is similar in morphology to the Morrow Mountain I Stemmed point. Both styles were characterized by contracting stems, but two differences were cited. The blade is proportionately longer and narrower and a noticeable angle occurs at the shoulder/stem juncture in the Morrow Mountain II Stemmed type. There is some suggestive evidence that Morrow Mountain II occurred at a slightly higher stratigraphic/vertical position than Type I in the Doershuk Site excavations, which led Coe (1964:43) to tentatively posit that the former was “transitional and survived to a later date.” This type is similar in length to the Morrow Mountain I Stemmed type (30–80 mm), but is noticeably narrower at the shoulder (18–30 mm). The specimen (a930) from 31RH491 was a base/shoulder fragment composed of UER material. One shoulder ear was also missing. The blade margins are incurvate in outline. It was recovered at a depth of 20 to 30 cm bs in a Stage II shovel test (N515/E495). A contracting stemmed base fragment (a2503) made of MMR may also represent a Morrow Mountain II Stemmed point (Figure 175:O). A single shoulder adheres close to the stem juncture, which is a characteristic more consistent with a Morrow Mountain II Stemmed point rather than the Savannah River Stemmed style. It was recovered between 0 and 10 cm bd in TU2. (6) Guilford Lanceolate, Round Base Variant. A single Guilford Lanceolate, Straight Base point (a2563) was recovered from 31RH491 (Figure 175:P). Coe (1964:43) described the Guilford Lanceolate type as a “long, slender, but thick blade with straight, rounded, or concave base.” These are large points that ranged between 50 and 120 mm in Coe’s sample, while widths ranged between 20 and 35 mm. Guilford Lanceolate points were found sandwiched between the earlier Morrow Mountain Stemmed and the later Savannah River Stemmed types in Zone VI at the Doerschuk Site (see Coe 1964:34). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 494 Chapter 10. 31RH491 An important factor of variation noted in Coe’s (1964:43) broadly defined Guilford Lanceolate type is the morphology of the basal contour. Concave, straight, and rounded bases are noted to exist in the Doershuk Site assemblage and it is clear from the overall physical resemblance and close stratigraphic associations that these variants are relatively contemporaneous. However, this particular variation may evidence a very fine-grained chronological sequence and, consequently, it is advisable to draw typological distinctions between the various basal contours (Cable and Cantley 2005). Adoption of this terminology will facilitate communication and preserve potentially important chronological variation in the Guilford Complex. The specimen from 31RH491 is made from microcrystalline Type I RT. A small portion of the tip was broken off prior to discard. Percussion flaking was crudely applied and secondary thinning flakes were absent. The blade margins and faces above the “shoulder,” the point in the center of the blade where the marginal contour narrows in both the proximal and distal directions, exhibited extensive polish dulling from use. It was recovered between 20 and 30 cm bd in TU1. (7) Savannah River Stemmed. William Claflin (1931) defined the type from his excavations at the Late Archaic Stallings Island shell midden on the Savannah River. Coe (1964:44-45) described the Doershuk Site specimens as large, heavy points with broad blades and stems. Blade edges were commonly convex in outline and bases ranged from straight to concave in shape. In South Carolina, the type occurs in both Stallings and pre-pottery deposits of the Late Archaic period. The type appears to pre-date pottery in North Carolina. Two Savannah River Stemmed points and a Savannah River preform were identified in the 31RH491 collection. Specimen a1998 (Figure 175:Q) was made of USR (2) material and was nearly whole. The tip was missing. The base was relatively straight in outline and much of the platform of the original flake blank was not retouched. Blade margins were straight to slightly excurvate and percussion flaking was crude. It was found in a Stage III shovel test in SB11 at a depth of 20 to 30 cm bs. Specimen a2484 was a basal fragment containing approximately half of the original blade, which had been snapped off through step fracturing (Figure 175:R). The stem was slightly expanding and the base was strongly concave, resulting in a classic “fish-tail” stem. The specimen was also made of USR (2) material. Blade outlines were straight and secondary flaking along the margins was somewhat carefully applied. It was found in a Stage III shovel test in SB5 at a depth of 0 to 10 cm bs. The preform (a957) was also made of USR (2) material (Figure 175:S) and it was recovered from a Stage I (N520/E530) that anchored SB3 at a depth of 30 to 40 cm bs. Only broad conchoidal flake scars were evident and the blade margins were parallel, which is frequently found in examples of Savannah River points. The upper blade had broken off in a hinge fracture and stem was not yet fully formed. (8) Yadkin Large Triangular. Coe (1964:4549) defined the Yadkin Large Triangular type from excavations conducted at the Doerschuk Site. The type was described as “a large, symmetrical, and well made triangular point.” Bases were usually concave and blade margins ranged from excurvate to straight to incurvate. Coe (1964:45) observed that all specimens were finished by careful pressure flaking. A great deal of morphological variation is contained within the Yadkin Large Triangular type description. Although this variation may simply reflect life cycle stages within a single constellation of typological traits, it is just as possible that it represents a chronological sequence of design style changes. If the stratigraphic and spatial dimensions of this variation are not monitored through typological recognition, this hypothesis will remain inadequately addressed. For this reason, Cable (2010:416) has proposed that various shape configurations of the Yadkin type be segregated into variants, as he also proposed for the Badin type. Four variants are identified in the plate of Yadkin points in the Doerschuk Site report (Coe 1964: Figure 42): (1) Variant A: triangular blade w/excurvate blade margins and concave base, (2) Variant B: triangular blade with straight blade margins and concave base, (3) Variant C: triangular blade with incurvate blade margins and concave base, and (4) Variant D: Lanceolate blade with excurvate blade margins and concave bases with side notched ears. Two specimens were typed as Yadkin Large Triangular points from the 31RH491 collection, both Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 495 Chapter 10. 31RH491 of which are further classified as Variant B. Specimen a1690 (Figure 175:T) was made of MMR and was nearly whole with the exception of a missing tip. The blade margins were carefully pressure flaked, producing almost a serrated affect on one side. Blade margins were relatively straight in outline and the base was moderately concave. It was found at a depth of 30 to 40 cm bs in a Stage III shovel test in SB11, somewhat lower than would be expected. The other specimen, a947, consisted of a basal ear from a concave based Large Yadkin Triangular point also meeting the definition of Variant B (Figure 175:U). It was found in a Stage I shovel test (N420/E460) anchoring SB1 at a depth of 10 to 20 cm bs. It was also made of MMR material. Unifaces Unifaces consist of tools that are retouched along flake margins in a single direction (unifacial edge shaping). Retouch was applied to shape a working edge or bit and, sometimes, a butt element for hafting. Limited bifacial retouch or resharpening is sometimes present along haft elements or working edges. Metric and attribute data for unifaces are presented in Appendix F. Sixty items classified as unifaces were recognized in the collection. These include 11 end scrapers, 25 side scrapers, 5 pointed scrapers, 3 denticulates, 2 gravers, 4 retouched flakes, 2 spoke shavess and 8 scraper fragments. The scraper fragments consist of very small sections of unifacially retouched edges, most probably representing fragments from thin side scrapers. The specimens are described below under the rubric of the typological system developed by Coe (1964) and elaborated by Daniel (1998). (1) End Scrapers. Eleven end scraper specimens were identified in the 31RH491 collection (Table 155). Daniel (1998:66) defines end scrapers as unifaces with “steep, regular unifacial retouch that forms a relatively narrow convex bit or working edge.” He further notes that the bit is commonly positioned in a transverse orientation to the long axis of the tool. Daniel expanded Coe’s (1964) original typological breakdown of end scrapers to include more expediently fashioned types as well. Most of the specimens found at 31RH491 correspond to expedient types in which design constraints on form were minimal. (a) Type I End Scrapers. Two Type I End Scrapers were identified in the collection. These are distinctive unifacial tools commonly found in both Paleoindian and Early Archaic tool kits in North America (Collins 2005). Coe (1964:76) observed that the Type I End Scrapers from the Hardaway site varied little from the descriptions of similar tools found at the Lindenmeier, Quad, Williamson, Shoop and Bull Brook Paleoindian sites. He recognized two forms in the Hardaway collection that Daniel (1998:66-68) later formalized into two subtypes. Type Ia included triangular to trapezoidal shaped end scrapers with well-executed unifacial retouch on lateral and distal edges. Type Ib End Scrapers are distinguished by additional facial retouch across the dorsal surface of the flake blank that produced a smooth and rounded contour resulting in a ‘tear-drop” plan view. Specimen a45 was assigned to the Type Ia subtype because of a lack of extensive cross-face retouch on the dorsal surface (Figure 176:A). It was made of USR (1) material and it was recovered at a depth of 20 to 30 cm bs in a Stage I shovel test (N90/ E480) anchoring SB11. Opposing notches for hafting the implement were created on the lateral margins immediately below the widest extent of the bit. Retouch occurred along each lateral margin, including some bifacial retouch that appears to have been done to strengthen the haft element. One side of the bit had also been modified through bifacial retouch on the ventral surface, the flake scar ridges of which exhibited dull polish from subsequent scraper use. The other side of the bit had been resharpened just prior to discard and displayed only minimal use polish. Resharpening on the dorsal face had created a very steep bit edge angle that undercut the general 90° work surface plane. Clearly, the tool life cycle was nearing exhaustion at the time it was discarded. Specimen a903 was classified as a Type Ib End Scraper (Figure 176:B). This subtype represents the classic “tear-drop” form with regularized retouch across the dorsal surface of the flake blank. Most of the bit had been removed by a transverse step fracture. The original plan of the tool, however, appears to have approximated a squat, lanceolate outline. Hafting notches are not evident, but it appears that accommodation to hafting was made by narrowing of the lateral margin near the butt end. The butt of the tool was snapped off leaving a slight hinge fracture, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 496 22 438 1167 1261 1289 1236 877 203 950 1010 157 271270a903 271270a2387 271270a2808 271270a3099 271270a2590 271270a1728 271270a389 271270a1939 271270a2048 271270a281 BAG 271270a45 SPECIMEN NO. End Scraper Fragment Type V End Scraper Type V End Scraper Type V End Scraper Type IV End Scraper Type III End Scraper Type IIa End Scraper Type IIa End Scraper Type IIa End Scraper Type Ib End Scraper Type Ia End Scraper UNIFACE/RETOUCHED FLAKE TYPE PORTION Bit End Fragment Whole Fragment Fragment Whole Fragment Fragment Fragment Fragment Whole BLANK TYPE Core Flake Blade Flake Core Flake Core Flake Core Flake Core Flake CROSS-SECTION Plano-Convex Lenticular Lenticular Lenticular Trapezoidal Plano-Convex Trapezoidal Plano-Convex Trapezoidal Plano-Convex Trapezoidal Uwharries Southern Rhyolite (2) Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Mill Mountain Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Type I Rhyolite Tuff Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) LITHIC RAW MATERIAL TYPE 14.87 13.87 10.55 20.90 27.07 (14.25) (16.32) (15.25) 8.01 (15.68) 26.07 OTHER BIT RETOUCH LENGTH (mm) (24.00)/ (14.30) (27.33)/ (17.86) (22.57)/ (25.12) 25.95 (19.66)/ (17.2) 32.08/36.39 END SCRAPERS: 31RH491 Table 155. Summary Data for End Scrapers, 31RH491. BIT RETOUCH LENGTH (mm) MINIMUM EDGE ANGLE 71° 58° 71° 68° 90° 62° 66° 53* 84° 54° 65° MAXIMUM EDGE ANGLE 82° NUMBER OF RETOUCHED EDGES (1) 3 1 (3) 1 3 (1) (2) (3) 3 3 MAXIMUM THICKNESS (mm) (5.27) (4.11) 2.24 6.64 6.08 5.58 (4.32) (12.46) 7.10 (4.41) 6.43 MAXIMUM LENGTH (mm) (10.82) (22.37) 30.63 (30.74) 34.20 36.22 (23.84) (57.56) (26.57) (25.86) 42.38 MAXIMUM WIDTH (mm) (20.88) (27.86) 12.56 (30.22) (35.32) 24.90 (16.55) (14.59) 11.52 (20.40) 29.30 WEIGHT (gm) 0.91 2.46 1.14 7.26 7.07 7.3 1.89 8.8 2.09 2.67 7.43 Graver Spur Graver Spur Minimal Retouch on Bite, Lateral Margins Bifacially Retouched into Haft Notches One Lateral Edge Used for Cutting, Graver Spur on Right Bit Edge Graver Spur on Right Dorsal Margin of Bit Alternate Edge Retouch, Hafting Notch Hafting notches NOTES Chapter 10. 31RH491 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 497 Chapter 10. 31RH491 Figure 176. End and Side Scrapers, Site 31RH491. A: Type Ia End Scraper (a45), B: Type Ib End Scraper (a903), C-E: Type IIa End Scrapers (a2387, a3099 and a2808 respectively), F: Type III End Scraper (a2590), G: Type IV End Scraper (a1728), H-J: Type V End Scrapers (a389-Ventral Face, a389-Dorsal Face and a2048 respectively), K-L: Type I Side Scrapers (a1902 and a2213 respectively), M-Q: Type IIa Side Scrapers (a1440, a2309, a821, a2591 and a2364 respectively), R: Type IIb Side Scraper (a257), S: Type IV Side Scraper (a2165). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 498 Chapter 10. 31RH491 probably in the same event that collapsed the bit. The low angle of the remaining bit edge (54°) suggests that the tool was broken at an early stage in its projected life cycle. The specimen was made of USR (2) material and was found at a depth of 40 to 50 cm bs in a Stage II shovel test (N455/E485). (b) Type II End Scrapers. Three Type II End Scrapers were identified during the investigation at the site. Coe (1964:76) defined the Type II End Scraper from his excavations at the Hardaway Site. The type consisted of flakes of variable shapes and sizes that were retouched only along the narrowest margins to form scraping edges. He recognized two varieties that Daniel (1998:72) later formalized. Type IIa End Scrapers were made on large, thick irregular flakes that were casually shaped at one end to produce a scraper edge. Cutting edges were always irregular in marginal outline and sharp, suggesting to Coe that they were used to work hard material such as wood or bone. They were primarily confined to Level II at the Hardaway Site, which was primarily associated with Kirk Stemmed and Serrated and Stanly Stemmed points. Type IIb End Scrapers were made on thin, narrow, prismatic flakes. The scraping edge was formed along the end opposite the striking platform. This variety was not numerous, but it had the same stratigraphic association as the Type IIa variety. Daniel (1998:72-73) expanded the type definitions to also include specimens with lateral margin retouch. All of the Type II End Scrapers identified in the 31RH491 collection were classified as Type IIa. Specimen a2387 (Figure 176:C) was made of Type I RT and was recovered between 40 and 50 cm bs in a Stage III shovel test in SB3. Both lateral margins were “backed” with steep unifacial retouch to form a haft element. The bit, however, was retouched on only one side and this consisted of minimal shaping over the ventral surface of the flake only. Light use polish was observable on both sides of the bit and the non-retouched side also exhibited tiny nibbling scars along the dorsal face. Specimen a3099 was manufactured from USR (2) material (Figure 176:D). The natural contours of the flake approximated a trapezoidal shaped plan typical of hafted end scrapers and hence the lateral margins appear not to have been retouched. One margin, however, had been broken off of the tool and the new edge displayed dulling and nibble scars indicative of utilization. Whether this utilization occurred during the use life of the tool as an end scraper or as a consequence of recycling is not clear. The bit edge, however, was formed by classic steep edge retouch and one end of the bit supports a short graver spur projection. The dorsal side of the flake actually formed a flatter scraper edge and retouch forming the bit originated from it rather than from the ventral side of the flake blank. The dorsal surface was primarily covered in cortex. The narrow butt end, however, exhibits retouch that emanates from the ventral surface and the edge displays dulling from either use as a secondary scraper edge or wear from hafting. This specimen was recovered at a depth of 40 to 50 cm bd in TU5. Finally, Specimen a2808 was also found at a depth of 40 to 50 cm bd, in TU4 (Figure 176:E). It consists of a lateral section of a large scraper that included most of one lateral margin and a portion of the bit. A substantial graver spur projected from the end of the bit edge on the fragment. The lateral margin was retouched from the bit juncture to a point about half way down the side where the contour narrowed. The non-retouched portion of the margin exhibited use dulling, probably resulting from wear associated with the haft attachment at this location or from intentional dulling to accommodate the haft. The specimen was composed of flow banded, USR (2) material. (c) Type III End Scrapers. A Type III End Scraper (a2590) made of USR (1) material was recovered at a depth of 30 to 40 cm bd in TU1 (Figure 176:F). Coe (1964:76) defined this type as a rough and larger version of the finely made Type I End Scraper, or “tear-drop” end scraper. Daniel (1998:75) notes that Type III End Scrapers at Hardaway exhibit the widest median bit value (42.4 mm) of all end scrapers, suggesting to him that this tool needed more lateral and dorsal retouch to form a functional haft element. The bit width the specimen from 31RH491 achieved a maximum dimension of 25.06 mm, which is at the very low end of the range for the type. The haft element, or butt end, of the specimen was broken off in a hinge fracture termination, suggesting that it was discarded upon breakage. The lateral margins were casually formed by both unifacial and bifacial retouch and the edges displayed variable degrees of Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 499 Chapter 10. 31RH491 dulling from haft wear or from intentional dulling the edges to accommodate the haft. The bit end was only minimally retouched as the natural contour of the termination was taken advantage of to serve as a scraper edge. Light use polish was spread across the entire length of the bit edge and extended approximately 3 mm up its face. The tool was apparently discarded prior to an attempt to sharpen the bit edge with steep marginal retouch. Coe (1964:76) observed that all of the edges on Type III End Scrapers were irregular and sharp, which may set this example off from the Hardaway collection. Daniel suggests that edge dulling and rounding on end scrapers may evidence the scraping of, in particular, dry hides. Other functions proffered for scrapers not exhibiting edge rounding, were scraping of resistant material. It is highly unlikely, however, that scraping hard surfaces would maintain a sharp edge. Since some examples within each of the other end scraper types also failed to exhibit edge dulling, it is likely that much of this variation is explained by the state of resharpening at the time of discard rather than specific functional differences. (d) Type IV End Scrapers. A single Type IV End Scraper (a1728) made of MMR was recovered from investigations at 31RH491 (Figure 176:G). Daniel (1998:75-76) extracted this type from the Hardaway scraper collection to segregate a group of informal end scrapers that he believed were hand-held during use. Type IV End Scrapers are unshaped with the exception of the bit end and they exhibit large bit widths. He notes, however, that approximately one half of the specimens displayed restricted retouch on edge margins near the bit. Daniel further observes that blank selection for Type IV End Scrapers was quite variable, suggesting that few design restrictions were placed on their manufacture. Moreover, bit placement was, most often, lateral to the striking platform of the blank. The overall morphology suggests to Daniel that the tool was manufactured on an ad-hoc basis to be used in an immediate and specific task. In other words, this type was considered to represent an expedient end scraper form. The bit on the specimen from 31RH491 appears to have been manufactured on the platform of a core flake blank, although the features of the platform have been completely removed by steep marginal retouch with the exception of a bulb of percus- sion on the ventral face of the blank. Light use polish is distributed across the bit edge and extends across the entire bit face. The edge angle approximates 90° and the final stages of retouch resulted in undercutting the bit face. Retouch was not observed on the lateral edges of the tool, but these margins were partly broken off. Although the tool may have been hafted, it fits comfortably in the hand and could have been used in this manner without a haft. It was recovered from a Stage III shovel test excavated in SB14 at a depth of 40 to 50 cm bs. (e) Type V End Scrapers. Type V End Scrapers are similar to Type IV End Scrapers in that they appear to have been hand-held rather than hafted and they were made on block-derived flakes with variable cross-sections (Daniel 1998:76-77). They are distinguished, however, by possessing relatively narrow bits that are placed at the corners of flake blanks, generally at the distal-lateral junction of the flake blank. As a result these scrapers have very narrow bit widths, but typically exhibit wide tool widths. In addition, these scrapers have well executed, straight to incurvate retouched lateral margins that expand away from the bit. Three specimens of variable sizes and shapes were classified as Type V End Scrapers in the 31RH491 collection. Specimen a389 was made of USR (1) material (Figure 176:H-I). It was recovered between 20 and 30 cm bs in a Stage II shovel test (N480/E495). The tool was snapped in half in a step fracture. The remaining portion exhibits careful and well-executed parallel pressure flaked retouch along both lateral margins, emanating from the ventral surface of the flake. Light edge dulling occurs along both edges. The bit edge was made on the reverse side of the blank, on its ventral surface. As is specified by the type definition, the bit is cantered to one corner of the flake blank, at the distal-lateral juncture. A projection at one end of the bit edge may represent a graver spur. Moderate use wear polish occurs along the entire circumference of the bit edge and across the ventral face of the bit. It is unlikely that such care would have been taken to form the lateral edges of the tool if the intention was simply to form a haft element. It is likely that both lateral margins were used as scraper edges, which suggests that the tool may have been unhafted. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 500 Chapter 10. 31RH491 Specimen a2048 was manufactured from a thin flake of Type I RT (non-patinated) and was recovered from a Stage III shovel test excavated in SB9 at a depth of 40 to 50 cm bs (Figure 176:J). The tool was snapped in half in a step fracture and the end scraper bit was cantered to one side of the distallateral juncture. Both lateral margins are retouched and the retouch on one of the margins is placed on the opposite side of the blank from the bit and the other lateral margin. A delicate graver spur projects from one end of the bit edge. All of the retouched edges exhibit use polish and its status as a handheld tool appears firm. Specimen a1939 was made on a small blade flake of USR (1) material. It was recovered at a depth of only 0 to 10 cm bs in a Stage III shovel tests excavated on SB6. A delicate and narrow bit edge was formed on one distal-lateral juncture with fine unifacial retouch. Use polish was observed on the entire length of the bit edge and along the associated lateral margin. Approximately one-third of the bit was broken off. The absence of accommodations for hafting suggests that this tool was also handheld. (2) Side Scrapers. Twenty-five side scrapers and eight indeterminate scraper fragments that most probably represent side scrapers were identified in the collection (Table 156). Daniel (1998:6683-84) defines side scrapers as unifaces containing “one or more unifacially retouched working edges that parallel the long axis of the flake blank.” Daniel identified four side scraper types and numerous subtypes in the Hardaway scraper collection. Representatives of all side scraper types were identified in the 31RH491 collection. Although individual specimens were discussed at length in the end scraper discussion, the side scraper collection is too large to merit this treatment. In general, it is assumed that scrapers are primarily associated with Early Archaic/Paleoindian tool kits and this was generally born out in the end scraper descriptions presented above. Most of the items were found between 40 and 50 cm bs/bd. Items “out of position” and occurring higher in the deposit matrix are predominantly the result of disturbance factors or variable rates of deposition and erosion. Table 157 presents the vertical distributions of all unifaces and retouched flakes in the collection from 31RH491. End scrapers show display the lowest overall vertical position, with 73 percent occupying positions below 30 cm bs. The other uniface types, including side scrapers, show a somewhat higher position, with only 53 percent occurring below 30 cm bs. This suggests two things. First, other uniface forms occur more frequently in later phases and, second, a good deal of disturbance occurs across the site. Descriptions of the various side scraper types are presented below. (a) Type I Side Scrapers. Coe (1964:77) defined this scraper type as being made from large, wedge-shaped flakes produced from directional cores. One of the long sides of the flake blank was subjected to steep percussion retouch, while one or both ends of the blank also received retouch shaping. Daniel (1998:84) further adds that the bit edges were generally convex to straight in shape and cross-sections appeared wide and flat (lenticular). Type I Side Scrapers were equally distributed between Levels II, III and IV at the Hardaway Site, in association with Kirk and Palmer Corner Notched points, Hardaway Side Notched points and Hardaway Blades. Two Type I Side Scrapers were identified in the collection (Figure 176:K-L). Both were made on thick, prismatic core flakes and retouch was confined to one lateral margin (b) Type II Side Scrapers. Building on Coe’s (1964:79) original description for the Type II Side Scraper, Daniel (1998:84-93) separated individual specimens further into two types based on the size of the flake blanks used to manufacture the tools. Type IIa refers to the classic description of a Type II Side Scraper. That is, scrapers made on large, irregular flakes whose bit edges were retouched without concern for regularizing the marginal outline. Daniel’s Type IIb was reserved for very small flakes, usually of less than 30 mm in length. The generally thin edges on these scrapers suggest that they were used for more delicate uses and on softer materials than Type I Side Scrapers. Coe (1964:79) observed that Type II Side Scrapers were strongly associated with the Palmer and Hardaway levels (III and IV) at the Hardaway Site. Eleven Type IIa Side Scrapers were recognized in the 31RH491 collection (Figure 176:M-Q). They were generally manufactured on large, thin core flakes from directional cores. Angles on retouched Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 501 944 1122 1177 1282 197 1063 1256 188 194 243 295 375 470 686 271270a1931 271270a2295 271270a2415 271270a3034 271270a379 271270a2165 271270a2734 271270a349 271270a362 271270a461 271270a593 271270a786 271270a949 271270a1370 271270a2591 888 1236 271270a2364 271270a1778 1159 271270a2309 877 1131 271270a1984 835 970 271270a1440 271270a1725 731 271270a1340 271270a1600 665 271270a1016 349 499 271270a821 271270a725 396 271270a382 139 199 271270a345 271270a257 187 271270a283 88 158 271270a2213 271270a173 925 1092 271270a1902 BAG SPECIMEN NO. Side Scraper Fragment Side Scraper Fragment Side Scraper Fragment Side Scraper Fragment Side Scraper Fragment Side Scraper Fragment Side Scraper Fragment Scraper Fragment Type IV Side Scraper Type III Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIb Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type IIa Side Scraper Type I Side Scraper Type I Side Scraper SIDE SCRAPER TYPE Whole PORTION Fragment Fragment Fragment Fragment Fragment Fragment Fragment Fragment Fragment Fragment Fragment Fragment Whole Fragment Fragment Fragment Whole Fragment Fragment Whole Fragment Whole Whole Fragment Fragment Fragment Fragment Fragment Fragment Fragment Fragment Fragment Core Flake BLANK TYPE Core Flake Core Flake FBR Core Flake Core Flake Core Flake Core Flake Core Flake Core Flake Core Flake Reused Biface frag. Core Flake Core Flake Trapezoidal CROSS-SECTION Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Triangular Trapezoidal Lenticular Trapezoidal Trapezoidal Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Mill Mountain Rhyolite Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite LITHIC RAW MATERIAL TYPE (18.99) (13.77) (12.06) (10.78) (17.28) (6.30) (12.57) (7.35) (12.28) (12.84) (13.77) (10.62) 11.38 (8.81) (15.34) (6.28) 12.84 (13.39) (12.10) 12.32 (22.61) 16.42 17.74 (14.39) (17.55) (14.95) (11.10) (31.20) 17.04 (14.93) (25.68) (27.41) 33.53 BIT RETOUCH LENGTH (mm) OTHER BIT RETOUCH LENGTH (mm) (16.77) SIDE SCRAPERS: 31RH491 87° MINIMUM EDGE ANGLE 58° 37° 78° 63° 51° 49° 61° 69° 58° 49° 62° 68° 65° 65° 69° 61° 65° 36° 59° 58° 75° 77° 54° 72° 66° 70° 59° 63° 72° 59° 59° 76° (1) 1 (1) (1) (1) (1) 1 (1) 2 (1) (1) (1) 1 1 1 1 1 (1) 1 1 1 1 1 1 1 1 (1) (1) (1) (1) 1 (1) 1 NUMBER OF RETOUCHED EDGES MAXIMUM EDGE ANGLE Table 156. Summary Data for Side Scrapers, 31RH491. (4.57) (3.14) (4.60) 3.66 3.85 2.08 3.42 (3.09) 4.48 2.83 (4.53) (2.91) 2.87 (2.33) 2.37 2.13 2.98 2.16 3.30 4.36 13.11 7.09 4.77 7.42 8.37 3.52 (3.34) (3.60) 5.38 6.40 6.90 (11.88) 9.41 MAXIMUM THICKNESS (mm) 502 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 26.22 MAXIMUM LENGTH (mm) (24.52) (13.48) (12.03) (10.11) (18.63) (7.23) (16.02) (12.34) (17.16) (12.39) (24.37) (20.40) 15.03 (14.04) (15.32) (15.21) 17.86 (16.70) (24.34) 19.94 (23.29) 27.03 20.93 (18.00) (34.38) (23.41) (20.95) (38.55) (31.64) (14.98) (27.43) (30.92) 36.75 MAXIMUM WIDTH (mm) (11.91) (11.35) (11.13) (13.72) (14.40) (11.07) 12.97 (11.07) (2.60) (17.73) (13.64) (15.36) 12.50 (9.49) (14.34) (17.08) 13.45 (14.89) (18.89) 9.46 (38.47) 23.54 22.54 (15.16) (21.37) (16.52) (24.17) (25.05) (26.49) (31.58) (22.25) (25.71) 9.43 WEIGHT (gm) 1.48 0.36 0.51 0.5 1.39 0.22 0.89 0.36 2.49 0.63 1.44 0.79 0.47 0.32 0.5 0.65 0.48 0.79 2.06 0.71 7.43 4.67 2.55 2.18 6.17 2.01 1.57 3.03 5.03 2.73 3.94 10.02 Graver-like Projection at One End One Edge bifacially retouched on margin Non-retouched Spoke Shave on Reverse Side of Working Edge Utilized cutting edge opposite retouched edge NOTES Chapter 10. 31RH491 Chapter 10. 31RH491 edge margins ranged between 54° and 77° (Table 156). None of the specimens exhibited evidence of possessing more than one retouched edge. The working edges were delicately retouched by pressure flaking and bit margins exhibited light dulling. A wide range of vertical positions was reflected in the collection. Only a little more than one half were situated below 30 cm bs (Table 157). Ten Type IIb Side Scrapers were recovered (Figure 176:R). These specimens were fashioned from small, thin flakes that were predominantly derived from directional cores. Bit margins were delicately retouched with pressure flaking and working edge angles ranged between 36° and 75° (Table 156). Each specimen exhibited only one retouched lateral edge and bit margins were lightly dulled. Their vertical positions were widely situated in the site matrix and comparable in overall pattern to Type IIa Side Scrapers (Table 157). (c) Type III Side Scrapers. Type III Side Scrapers were made on thin and narrow flakes (Coe 1964:79). Retouch occurred along the edges parallel to the long axis of the flakes and generally involved reshaping of only the working edges. No accommodations for hafting were present. This scraper type was found predominantly in Levels II and III at Hardaway, levels associated with Palmer through Morrow Mountain occupation (Coe 1964:73). Very few of Table157. Vertical Distribution of Scrapers and Retouched Flakes, 31RH491. VERTICAL DISTRIBUTION OF UNIFACES: 31RH491 LEVELS UNIFACE TYPE 1 2 3 4 5 6 GRAND TOTAL END SCRAPERS Type Ia End Scraper 1 1 Type Ib End Scraper Type IIa End Scraper Type III End Scraper 1 3 3 1 Type IV End Scraper Type V End Scraper 1 1 1 End Scraper Fragment 1 1 1 1 3 1 1 2 2 SIDE SCRAPERS Type I Side Scraper Type IIa Side Scraper 1 Type IIb Side Scraper Type III Side Scraper 2 2 5 1 1 1 4 2 11 2 1 1 Type IV Side Scraper 1 Scraper Fragment 1 1 Side Scraper Fragment POINTED SCRAPERS 10 1 2 1 4 7 1 3 1 5 OTHER UNIFACES Denticulate 1 Graver Retouched Flake 1 Spokeshave 1 GRAND TOTAL 4 1 1 3 1 1 2 2 1 4 1 8 14 2 22 10 2 60 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 503 Chapter 10. 31RH491 these forms were found at Doerschuk, however, which suggests that their primary association may be with Early Archaic phases. Daniel (1998:93) informs that the Type III Side Scrapers from the Hardaway collection were made on bladelike flakes, which, in this report, are referred to as blade flakes. One specimen (a379) meeting these criteria was identified in the 31RH491 collection. It was made of UER material and it consisted of the broken upper one-quarter of a blade flake with delicate unifacial retouch observable on only one margin. The bit displayed use polish along the retouched edge segment. It was found in the upper 10 cm of deposit. (d) Type IV Side Scrapers. Daniel (1998:95) recognized a previously undefined side scraper type from the Hardaway collection that he referred to as a Type IV Side Scraper. This scraper was very similar in size and bit morphology to the Type I Side Scraper. However, in these specimens the side opposite the scraper edge is bifacially retouched. Daniel surmises that the bifacial retouch was applied to prepare the scraper for hafting. The fact that the bifacial edge was oriented parallel to the bit edge suggested to Daniel that the scraper was hafted in such a way as to allow the haft to be held by both hands in a draw-knife fashion. A single specimen (a2165) meeting these criteria was identified in the collection from 31RH491 (Figure 176:S). It was made from a core flake of USR (2) material that had been snapped in a step fracture, probably during use. Retouch on the unifacial edge was not steep and the resulting working edge was relatively acute (i.e. 58°). The bifacial edge was only marginally retouched and the edge was also acute. (3) Pointed Scrapers. This scraper type was infrequently found in the Hardaway Site assemblage. Coe (1964:79) described it as a tool manufactured from a thick flake in which steep unifacial retouch was applied to two sides of the dorsal surface to form a point at one end. The vertical distribution of the pointed scrapers at Hardaway was indistinguishable from Type I end scrapers and suggests, again, an Early Archaic association. Daniel (1998:102) recognizes two subtypes within the Hardaway collection. Type Ia corresponded to the original description provided by Coe, while Type Ib specimens were made on thinner flake blanks. Five Pointed Scrapers were identified in the 31RH491 collection (Table 158). Three of these (a2028, a2359 and a2723) were made on large, thick flakes and were identified as examples of Type Ia. Specimen a2723 (Figure 177:A) displayed marginal retouch on two converging lateral edges. The tip had broken off, but use polish ran across the broken edge indicating that the tool had been used after the breakage event. Both lateral margins were also dulled and displayed use polish. Specimen s2028 (Figure 177:B) exhibited retouch along two lateral margins converging in a point, but retouch on the two margins was placed on alternate faces of the flake. Light to moderate use polish and dulling was observed on both retouched edges. Specimen a2359 (Figure 177:C) was manufactured on a large white quartz core flake. Retouch along one of the converging lateral margins was extremely steep, while the other margin had collapsed from use. Again, edge dulling and polishing was evident along both margins. The remaining two specimens were made on small thin flakes and were therefore classified as Type Ib Pointed Scrapers. Specimen a2456 was made on an extremely thin flake and only the point and small sections of the converging retouched lateral edges were represented on the fragment. Specimen a2026 (Figure 177:D) was whole and consisted of converging, retouched lateral margins in which retouch continued around the point of the tool. Again, edge polish and dulling occurred along the entire circumference of retouch. In addition, a small spoke shaves concavity was present at the end of one of the lateral margins. The concavity exhibited use nibbling and edge polishing in the center. (4) Other Unifacial Tools. In addition to the more conventional scraper types, four other marginally retouched tool types were recognized in the 31RH491 collection. These consisted of denticulates, gravers, retouched flakes and spoke shavess. Each of these is described below. Summary data on the various specimens are presented in Table 158. (a) Denticulates. Three denticulates were identified in the 31RH491 collection (Figure 177:EF). Denticulates represent flakes that display a series of regularly spaced notches/projections along a working edge creating a saw-like appearance (Daniel 1998:103). The morphology of the working edge suggests that the tool functioned as a knife or saw for Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 504 1258 271270a2755 1278 271270a2966 1014 867 271270a2245 271270a2060 1287 271270a3091 1282 634 271270a1291 271270a3037 602 271270a1203 381 1191 271270a2456 271270a797 995 271270a2026 346 1256 271270a2723 271270a716 1155 271270a2359 52 996 271270a2028 271270a98 BAG SPECIMEN NO. Spoke Shave Spoke Shave Retouched Flake Retouched Flake Retouched Flake Retouched Flake Graver Graver Denticulate Denticulate Denticulate Type Ib Pointed Scraper Type Ib Pointed Scraper Type Ia Pointed Scraper Type Ia Pointed Scraper Type Ia Pointed Scraper UNIFACE TYPE PORTION Fragment Whole Fragment Fragment Fragment Fragment Whole Whole Fragment Fragment Fragment Fragment Fragment Tip Missing Whole Whole BLANK TYPE Core Flake FBR Core Flake Core Flake Core Flake Core Flake CROSS-SECTION Lenticular Lenticular Lenticular Trapezoidal Trapezoidal Lenticular Lenticular Lenticular Lenticular Lenticular Lenticular Triangular Lenticular Lenticular Uwharries Southern Rhyolite (2) Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Uwharries Southern Rhyolite (2) Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) White Quartz Uwharries Western Rhyolite LITHIC RAW MATERIAL TYPE BIT RETOUCH LENGTH (mm) (18.59) 17.59 (6.01) (10.22) (10.82) 15.48 4.27 (8.16) (25.17) (28.54) (14.31) (18.19) (27.43) 44.04 18.95 (18.53) 4.40 (12.12) (11.57) 15.61 (11.07) (7.24) 14.58 OTHER BIT RETOUCH LENGTH (mm) OTHER UNIFACES AND RETOUCHED FLAKES: 31RH491 78° 64° 44* 33° 44° 49° 38° 53° 62° 66° 55° 60° 77° 80° 66° MINIMUM EDGE ANGLE MAXIMUM EDGE ANGLE 73° 65° 40° 84° NUMBER OF RETOUCHED EDGES 2 1 (1) (1) (1) 1 2 (2) 2 1 (2) 2 2 2 2 MAXIMUM THICKNESS (mm) 4.98 4.74 (6.90) (1.42) 5.50 (7.43) 4.67 (3.07) 2.37 4.71 (1.18) 4.62 8.59 7.51 6.36 (24.51) 30.25 (5.24) (10.68) (13.67) (27.32) 24.02 (12.02) (25.62) (26.04) (11.46) (20.67) (45.24) 44.95 26.15 MAXIMUM LENGTH (mm) TABLE 158. Summary Data for Other Scrapers and Retouched Flakes, 31RH491. MAXIMUM WIDTH (mm) (20.34) 17.23 (28.00) (5.96) (13.39) (17.12) 21.77 (10.27) (15.74) (32.01) (14.87) (15.67) (33.01) 26.42 39.25 WEIGHT (gm) 2.74 2.1 0.93 0.08 1.31 0.38 2.64 2.43 0.32 1.15 4.06 0.2 1.49 14.95 9.14 7.94 Scraper Edge Opposite Spoke Shave Marginal Bifacial Retouch Marginal Bifacial Retouch Marginal Bifacial Retouch Utilized Edge Opposits (Cutting) Spokeshave Concavity Point Broken Off and then Repaired NOTES Chapter 10. 31RH491 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 505 Chapter 10. 31RH491 Figure 177. Pointed Scrapers and Other Retouched Tools, Site 31RH491. A-C: Type Ia Pointed Scrapers (a2723, a2028 and a2359 respectively), D: Type Ib Pointed Scraper (a2026), E-F: Denticulates (a1203 and a1291 respectively), G-H: Gravers (a1681 and a2966 respectively), I-J: Spoke Shaves (a2060 and 2755 respectively). processing materials of soft to moderate hardness. Two of the denitculates were found in low positions in the site matrix, between 40 and 60 cm bs, suggesting a primary association with the Early Archaic period (Table 157). (b) Gravers. Two gravers were recognized in the collection (Figure 177:G-H). Gravers represent flakes with short unifacially flaked projections (Daniel 1998:104; Goodyear 1974:55). It is generally held that gravers served as engravers or points for slotting bone and wood. Graver projections or spurs also occur on hafted end scrapers, but gravers occurring on otherwise unmodified flakes were probably handheld. The flake gravers were also found at relatively deep vertical positions, between 30 and 50 cm bs, also suggesting that they are primarily associated with Early Archaic occupation (Table 157). (c) Retouched Flakes. Four flakes with acute edges created by marginal bifacial retouch were identified in the collection (Table 158). The edge angles on these specimens ranged between 33° and 44°, suggesting that they were used as expedient handheld knives. The vertical positions of these tools ranged between 20 and 50 cm bs, indicating a primary Archaic association. (d) Spoke Shaves. Two spoke shaves were identified in the collection. Typically, unifacial retouch is applied to a natural concavity on a flake margin to produce a strong working edge from which to scrape or shave narrow convex surfaces such as spear or arrow shafts. Goodyear (1974:50-53) identified both hafted and unhafted spoke shaves at the Brand Site, and Daniel (1998:104-105) recognized two hafted specimens in the Hardaway collection. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 506 Chapter 10. 31RH491 Goodyear believes that most of the spoke shaves he recovered at the Brand Site were associated with the Dalton occupation. The spoke shaves recovered at 31RH491 are composite tools. Specimen a2060 includes a spoke shave concavity and a graver projection (Figure 177:I). The spoke shave concavity consists of a natural rc that was enhanced and strengthened by unifacial retouch. Edge polish and dulling were observable along the entire arc of the working edge. The graver projection displays minimal retouch along one edge of an otherwise natural point. Again, use polish and dulling occurred along both marginal edges and the tip. Although the concavity on the specimen might represent a hafting notch, the margin on the reverse side is unmodified and very thin. Had the tool been hafted in the manner entertained, the reverse side would have exhibited evidence of haft modification. Specimen a2755 displayed a broad spoke shave concavity and a retouched scraper edge on the reverse side (Figure 177:J). Again, use modification in the form of edge dulling and polish was present on the edges of each working edge or bit. Clear vertical patterning is not exhibited in the distribution of spoke shaves. Utilized Flakes Twenty-six non-retouched flakes exhibited wear patterns on one or more edges, indicating their use as expedient tools (Table 159). Metric and attribute data for utilized flakes are presented in Appendix G. All of these specimens exhibited edge attrition and bifacial nibbling scars indicative of cutting functions. Utilized edge angles fell between 21°and 63°. Most of the specimens appear to have been derived from block cores (i.e. core flakes, blade flakes and bladelet flakes), but three were made on large FBRs. All of the recognized utilized flakes were made on metavolcanic raw material, but it is likely that white quartz is under-represented due to the difficulty experienced in identifying wear on this material. Cobble Tools Seven items were classified as cobble tools. These consisted of three hammer stones, a spall from a hammer stone and three pebbles. Specimen a1399 was a quartz cobble hammer stone weighing 221.00 gm. Battering abrasion occurred on both ends of the cobble and along restricted locations on the lateral edges (Figure 178:A). It was collected from the surface of the site. Specimen a2581 was also a quartz cobble hammer stone (Figure 178:B) that was found in TU1 at a depth of 20 to 30 cm bd. Battering was present on both ends of the cobble. It weighed 114.72 gm. Specimen a314 (Figure 178:C) was a hammer stone fragment made of a dense meta-igneous or grano-diorite cobble. Battering occurs on the one observable end of the tool. It weighed 107.91 gm and was found at a depth of 20 to 30 cm bs in a Stage II shovel test (N500/E485). Specimen a2620 represented a spall from a quartz cobble hammer stone (Figure 178:D). Battering was observed on one side of the dorsal surface of the spall. It weighed 24.92 gm and was recovered between 30 and 40 cm bd in TU1. Finally, three pebbles were recovered (Figure 178:E-F) which displayed no evidence of use. These items appear too large to have been brought to the site by some natural process, so it is assumed that they were transported and deposited at the site by human agency. The pebbles were found at variable depths ranging from 0 to 90 cm bs. Intended uses for pebbles would include detailed retouch hammer stones and ceramic polishing stones. Grooved Abrader A flat fragment of arenite (a1653) exhibited a rounded, ground groove on its smooth face (Figure 178:H). The reverse face is unmodified and rough, while the surface containing the groove is smooth. One edge also appears to have been smoothed by grinding into a regularized curved outline. The rounded groove measured 8.91 mm in width and was less than 3.0 mm in depth. Goodyear (1974:69-71) reports similar grooved abraders from the Dalton Deposits at the Brand Site. These Brand site abraders were made of locally available sandstone and ferruginous sandstone and could be segregated into two groups. One group consisted of relatively narrow grooves ranging between 3 mm and 8 mm in width, while the other was composed of specimens of 10 mm to 15 mm in width. Goodyear argued that the size and shape of this narrower group suggested that they were used to manufacture thin cylindrical objects such as bone points, pins and perforating tools. He further argued that the wider group of abraders was likely used in the Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 507 Chapter 10. 31RH491 TABLE 159. Summary Data for Utilized Flakes, 31RH491. 39° 14.51 54° Type I Rhyolite Tuff 2.04 32.36 14.74 0.91 22° Type I Rhyolite Tuff 4.61 28.47 17.48 1.39 34° Core Flake Uwharries Southern Rhyolite (2) 5.13 (46.23) (24.41) 4.52 30° Cutting Core Flake Uwharries Southern Rhyolite (2) Cutting Core Flake Uwharries Eastern Rhyolite Fragment Cutting Bladelet Flake Whole Cutting Core Flake 271270a1235 Whole Cutting Core Flake Uwharries Southern Rhyolite (2) 4.54 271270a1341 Fragment Cutting FBR Uwharries Southern Rhyolite (2) 3.29 840 271270a1608 Fragment Cutting Uwharries Southern Rhyolite (1) (1.50) 271270a2039 1003 271270a2039 Whole Cutting Core Flake Uwharries Southern Rhyolite (1) 8.49 271270a2245 1104 271270a2245 Whole Cutting Core Flake Uwharries Southern Rhyolite (1) 4.13 271270a2276 1115 271270a2276 Fragment Cutting Uwharries Southern Rhyolite (2) 2.62 271270a2338 1146 271270a2338 Whole Cutting Uwharries Southern Rhyolite (2) 7.06 28.81 271270a2407 1172 271270a2407 Fragment Cutting Metasedimentary, green 5.72 39.38 24.71 5.31 21* 271270a2558 1232 271270a2558 Whole Cutting Blade Flake Uwharries Eastern Rhyolite 5.78 35.19 13.76 2.15 33° 271270a2724 1256 271270a2724 Whole Cutting Core Flake Uwharries Eastern Rhyolite 13.91 82.16 22.90 23.31 60° 271270a2745 1257 271270a2745 Whole Cutting Core Flake Uwharries Southern Rhyolite (2) 3.29 33.50 29.10 3.49 36° 271270a2772 1258 271270a2772 Whole Cutting Blade Flake Uwharries Southern Rhyolite (2) 5.96 40.94 21.50 3.74 28° 271270a2781 1259 271270a2781 Whole Cutting Core Flake Uwharries Southern Rhyolite (1) 5.24 21.83 18.19 1.28 26° 271270a3136 1302 271270a3136 Whole Cutting Bladelet Flake Type I Rhyolite Tuff 1.31 26.97 4.21 0.48 28° 271270a3147 1304 271270a3147 Fragment Cutting Blade Flake Type I Rhyolite Tuff 3.44 (17.74) (19.16) 1.01 30° 271270a3148 1304 271270a3148 Whole Cutting Core Flake Type I Rhyolite Tuff 2.04 11.68 9.31 0.21 22° 271270a3161 1307 271270a3161 Fragment Cutting Core Flake Type I Rhyolite Tuff 5.96 (25.31) 24.55 2.99 35° LITHIC RAW MATERIAL TYPE 271270a78 41 271270a78 Fragment Cutting Bladelet Flake Uwharries Southern Rhyolite (1) 271270a149 75 271270a149 Whole Cutting Core Flake Mill Mountain Rhyolite 271270a200 104 271270a200 Whole Cutting FBR 271270a201 104 271270a201 Whole Cutting FBR 271270a329 179 271270a329 Fragment Cutting 271270a559 285 271270a559 Whole 271270a850 408 271270a850 Whole 271270a981 480 271270a981 271270a1064 523 271270a1064 271270a1235 616 271270a1341 666 271270a1608 Core Flake EDGE ANGLE 0.18 40.29 BLANK TYPE WEIGHT (gm) (6.74) 33.30 CONDITION MAXIMUM WIDTH (mm) MAXIMUM LENGTH (mm) (15.59) BAG NO. UTILIZED EDGES 1.89 13.15 SPECIMEN # UTILIZED FLAKE FUNCTION MAXIMUM THICKNESS (mm) UTILIZED FLAKES: 31RH491 3.81 45.60 24.76 4.56 28° 10.11 36.32 43.97 12.43 63° Uwharries Southern Rhyolite (2) 3.16 (29.68) 11.50 1.07 38° Uwharries Southern Rhyolite (2) 5.51 30.80 36.04 4.44 46° 27.78 14.11 1.47 57° (20.91) (29.17) 2.76 42° (8.67) (13.43) 0.21 39° 50.33 23.73 6.4 37° 22.44 19.52 1.72 28° (20.68) 18.09) 1.04 24° 28.85 3.57 45° manufacture of Dalton Adzes based on the contents of the Hawkins Cache (see Morse 1971:18). A grooved abrader was also recovered at 31HK2502, at a depth of 20 to 30 cm bs. The specimen from 31RH491 was found between 30 and 40 cm bs in a Stage III shovel test excavated in SB10. The extremely regular, perfectly concave cross-section of Specimen a1653 indicates that it was used to grind objects of similar shape, such as bone sections or wooden spear shafts. Precontact Ceramics Eighty-two precontact ceramics were recovered from the Phase II operation at 31RH491. Seven subseries were identified, including New River I and II, Hanover I, II and III, Yadkin III and Cape Fear III (Tables 160 and 161). The large number of vessels represented (N=59) is partly a function of recovery strategy, which primarily consisted of widely spaced shovel tests at 10 and 5 m-intervals. However, as is true of other sites in the testing package, the ceramic deposit at the site is highly disorganized and fragmentary. Metric and attribute data on precontact ceramics are contained in Appendix I. Although monitoring of temper types within series has proceeded in an ad hoc fashion in earlier investigations (see Cable and Cantley 2005a, b, 2006, Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 508 Chapter 10. 31RH491 Figure 178. Miscellaneous Stone Tools, Site 31RH491. A-C: Hammer Stones (a1399, a2581 and a314 respectively), D: Hammer Stone Spall (a2620), E-G: Pebbles (a2650, a356 and a3105 respectively), H: Grooved Abrader (a1653). Cable 2010), the enormity of the inventory of temper constituents and various compositional combinations seen in ceramic collections in the region makes it imperative that in the future a universal system be developed to accommodate cross-project comparability. Formerly, temper classes were assigned letter designations as they were historically recognized within each series, creating a confusing list of different letter designations for the same temper type within each series. Here, temper constituent designations are standardized into a single classificatory system. The letters “a” and “b” are reserved for the Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 509 Chapter 10. 31RH491 with only minimal application of pressure with the thumb. A “compact” state was identified in instances where moderate pressure with the thumb was required to break the edge of a sherd. Finally the “hard” state was reserved for those sherds whose edges were not easily broken even with extreme thumb pressure. basic paste representatives diagnostic of each series as described in previous reports. The remaining letter designations note the addition of other temper constituents to these basic paste categories as described below: (c) Crushed Arenite (d) Granule to Pebble Sized Angular/Subangular Quartz Sand and/or Crushed Quartz (e) Crushed and Ground Indurated Granite (f) Medium to Granule Sized Angular/Subangular Feldspar Sand Generally, pottery achieves greater hardness when one or more of three conditions prevail (Shepard 1954:114). These are: (1) the pottery is made from low-fusing, dense-firing clay, (2) the pottery is fired at a high temperature and/or (3) the pottery is fired in an atmosphere promoting vitrification. It is common to observe in the local ceramic sequences of the Southeast, a progression from friable ceramics to much harder, more durable pottery in the later stages of cultural development. Clearly, Mississippian potters had developed technological strategies to produce extremely hard, sometimes vitrified pottery and it is quite possible that this advancement was meant to adapt to a more sedentary, agricultural life way in which durability and longevity of pots would have been advantageous. Similar adaptive technologies may have been incorporated into sedentary, agricultural societies peripheral to the Southeast Appalachian Mississippian culture as well, and for the same reasons. Currently, a sophisticated method for More than one of these temper constituents is sometimes present in a single sherd. In these cases, the constituent designations would include combinations of letter designations (e.g. “c/e” or “e/f”, etc.). During the course of the ceramic analysis of the DO5 site package it also became apparent that some of the pottery exhibited extremely hard pastes, much harder, in fact, than the typical paste previously defined for each series. A hardness scale consisting of three attribute states was developed to describe this variation. These consisted of: (1) friable, (2) compact and (3) hard. The attribute state of “friable” corresponds to sherds that are easily crushed on their edges Table 160. Precontact Ceramic Sherd and Vessel Representation, 31RH491. CERAMIC SERIES SHERD AND VESSEL REPRESENTATION: 31RH480 CERAMIC SERIES SHERD COUNT VESSELS REPRESENTED New River I 13 11 New River II 5 4 Hanover I 7 7 Hanover II 30 20 Hanover III 10 6 Yadkin III 8 7 Cape Fear III 4 4 residual sherds 5 GRAND TOTAL 82 59 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 510 Chapter 10. 31RH491 Table 161. Sherd Counts and Weights for Ceramic Types, 31RH491. SHERD COUNTS AND WEIGHTS FOR CERAMIC TYPES: 31RH491 CERAMIC TYPE SHERD COUNT WEIGHT (gm) New River Series New River Ia Plain 2 15.07 New River If Fabric Impressed 2 6.35 New River If indet. 3 7.26 New River If Plain 6 29.78 New River IIa Fabric Impressed 3 7.69 New River IIf Fabric Impressed 1 1.37 New River IIf Plain 1 3.50 Hanover Ia Check Stamped 1 5.11 Hanover Ia indet. 2 5.71 Hanover Ib Cord Marked 1 4.38 Hanover Ib indet. 1 1.56 Hanover Ib indet. dec. 2 3.94 Hanover IIa Check Stamped 9 31.25 Hanover IIa Fabric Impressed 7 12.44 Hanover IIa indet. 2 4.40 Hanover IIa indet. dec. 1 3.12 Hanover IIb Check Stamped 1 6.12 Hanover IIb Fabric Impressed 1 3.09 Hanover IIb indet. 1 0.85 Hanover IId Cord Marked 2 4.07 Hanover IId Fabric Impressed 1 0.88 Hanover IIe Fabric Impressed 5 7.23 Hanover IIIa indet. 1 1.20 Hanover IIIc/e Fabric Impressed 1 1.23 Hanover IIIc/e indet. dec. 1 3.65 Hanover IIId indet. dec. 1 2.90 Hanover IIIe Cord Marked 5 23.23 Hanover IIIe Fabric Impressed 1 6.72 Yadkin IIIc/e indet. 1 1.20 Yadkin IIId indet. dec. 1 5.43 Yadkin IIId Plain 2 21.77 Yadkin IIId/e Plain 1 4.38 Yadkin IIIe Cord Marked 1 1.77 Yadkin IIIe Fabric Impressed 2 3.70 Cape Fear III Cord Marked 1 1.51 Cape Fear III indet. 2 3.98 Cape Fear III indet. dec. 1 0.94 Residual Sherd 5 2.70 GRAND TOTAL 82 251.48 Hanover Series Yadkin Series Cape Fear Series Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 511 Chapter 10. 31RH491 measuring hardness in “primitive” pottery does not exist due to its porosity and heterogeneity of paste composition. The “thumb” test, although somewhat imprecise, provides a relatively objective and empirical basis for identifying relative hardness in precontact pottery assemblages. For an extended discussion of hardness in precontact ceramics see Chapter 4. This method of analysis resulted in segregating extremely hard or vitrified pottery that would have traditionally been placed in the Cape Fear (sandtempered), Hanover (sand- and grog-tempered) and Yadkin (crushed rock- and granule sand-tempered) series. As a means of preserving data to test the “hardness trajectory” hypothesis, these hard examples were separated into new series variants distinguished by the Roman numeral suffix “III.” Thus, extremely hard variants of the series are respectively referred to here as Hanover III, Cape Fear III and Yadkin III. If the efficacy of the hypothesis is later confirmed, then it would be appropriate to assign new series names to the Cape Fear III and Yadkin III variants since these series are generally viewed as Middle Woodland in affiliation. New River Series New River series ceramics are associated with the Early Woodland period in the region. The series was originally defined by Loftfield (1976) who characterized it as a coarse sand-tempered pottery of compact composition and gritty consistency characterized by cord marked, fabric impressed, thong simple stamped and plain exterior surface treatments. Correlated series of the Early Woodland period include Lenoir (Crawford 1966) and Deep Creek (Phelps 1983). The New River series in the Fort Bragg area is poorly understood. Herbert (2003:156) indicates that both net-impressed and parallel/overstamped cord marked types are recognized, while Cable and Cantley (2005a, b) have also recognized simple stamped, fabric impressed and plain surface treatments. The predominant classificatory criterion used to identify the series is an abundance of quartz sand temper as Loftfield (1976) prescribed. Herbert (2003:156) proposes that an arbitrary cut-off at about 15% sand temper density should be used to distin- guish Cape Fear (< 15%) and New River (>15%) in the Fort Bragg area. Cable and Cantley (2005a, b) have recognized two major paste variants in Fort Bragg assemblages. These are variants Ia and IIa. Variant Ia consists of sherds dominated by abundant medium to very coarse subangular to rounded quartz sand grains in a gritty, friable paste matrix. Variant IIa contains a fine, compact, relatively harder paste matrix with abundant to moderately dense medium to very coarse quartz sand inclusions. Variant IIa may correlate with the traditionally defined Cape Fear series. Feldspar sand is sometimes found in New River series sherds, and it is designated as New River If or IIf. Eighteen New River series sherds were recognized in the collection from 31RH491 (Table 161). The New River I sample consisted of 13 sherds representative of paste variants Ia (n=2) and If (n=11). Both pastes are friable and gritty, but they are differentiated by the mix of clastic inclusions/temper. The former is composed of abundant medium to coarse quartz sand with additions of granule to pebble size angular and subangular opaque white quartz sand. The latter contains, in addition, medium to coarse “feldspar” sand of subangular shape. This material is only tentatively identified as feldspar. It consists of opaque, grayish orange to moderate yellow, highly micaceous rock that contains feldspar and appears to represent metamorphosed material. Most of the New River I sample is comprised of plain exterior surface treatments (Figure 179:A-E). Eighty percent of the sherds with recognizable surface treatments are plain, while fabric impressed makes up the remaining portion of this sample ((Figure 179:F-G). Fabric impressed weave patterns were identifiable on only one of the two sherds. It consisted of a rigid warp and a medium weft. Vessel wall thickness for the New River I sample ranged from 7.02 to 10.78 mm with a mean of 9.26 mm (SD=1.23 mm). Rims were not recovered in the sample. The New River II sample (n=5) consists of the same two temper constituent variants, IIa and IIf. The paste, however, is more compact and background sand is somewhat finer and less abundant. The New River IIa sample is dominated by fabric impressed surface treatments, comprising 80 percent of the recognizable surface treatments (Figure 179:I-J). Only one plain sherd was identified (Figure 179:H). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 512 Chapter 10. 31RH491 Weave patterns on the fabric impressed sherds were too faint or restricted in view to identify. Vessel wall thickness for the New River II sample ranged from 6.06 to 10.69 mm with a mean of 7.37 mm (SD=1.91 mm). Rims were not recovered in the sample. Hanover Series South (1960:16-17; 1976: 28) defined the Hanover series from surface collections made in 1960 on sites located between Wilmington, NC and Myrtle Beach, SC. Pottery affiliated with this series is tempered with aplastic clay lumps, most of which consist of grog fragments from crushed sherds. A correlate on the north coastal plain of North Carolina is referred to as the Carteret series (Loftfield 1979:154157). Cord marked and fabric impressed exterior surface treatments were included in the original series description, but other grog-tempered types have been recognized subsequently. Herbert (2003:75) has identified minority proportions of check stamped, plain (smoothed), net impressed, simple stamped, punctate, and random straw bundle punctate surface treatments. The latter two types are traditionally subsumed under the Refuge series on the central South Carolina coast where grog/clay temper occurs as a major temper variant (see Cable 2002: 195-201). A great deal of variability in paste composition has been noted in Hanover assemblages, primarily related to differences in quartz sand contributions, clay versus grog distinctions, and the density and size of clay/grog particles (see Cable et al. 1998; Herbert 2003; 74-75). Herbert (2003:191-192) recognizes two variants that he believes are sequential. The earliest variant, Hanover I, is sand-tempered with minor amounts of finely crushed grog. Surface treatments attributed to Hanover I pottery consist of cord marking, fabric impressing and check stamping. He estimates an age range of AD 400-800 for Hanover I pottery. The later variant, Hanover II, is inferred to correlate with a Late Woodland temporal range (AD 800 through 1500). Hanover II pottery is characterized by pastes with only small amounts of sand and abundant grog particles. The dominant exterior surface treatment of the later variant is fabric impressing. Cable and Cantley (2005a, b) have followed Herbert (2003) in describing Hanover series ceram- ics in the Fort Bragg area. However, in an attempt to more fully capture the temper and paste variation, these basic groupings have been expanded to include sub-variants. As research has progressed, the number of sub-variants has increased to accommodate an ever-expanding range of temper constituents and paste qualities. There are three major paste variants found in Hanover pottery at Fort Bragg based on relative hardness. Hanover I is characterized by a friable paste, Hanover II exhibits a compact paste and Hanover III contains a hard to vitrified paste. Within each of these relative hardness groupings are two primary temper constituent designations. Designation “a” refers to pastes with sparse to moderate amounts of fine to medium quartz sand and abundant medium to large grog particles. Designation “b” includes pastes with moderate to abundant amounts of medium to coarse quartz sand and sparse to moderate densities of medium to large grog particles. Other temper constituents are sometimes added to these basic paste variants. These include crushed arenite (“c”), granule to pebble sized angular to subangular quartz sand and, rarely, crushed quartz (“d”), crushed and ground indurated granite (“e”) and medium to granule sized subangular feldspar sand (“f”). Crushed arenite and indurated granite almost exclusively occur in Hanover III paste. Indurated granite temper was originally identified as sandstone conglomerate because of similar texture characteristics, but thin-sectioning of rock samples indicated that most of the temper in this category corresponds to a fine grained, indurated granite (see Appendix K). Some of the temper in this category may actually represent sandstone conglomerate, but this was not differentiated in the ceramic analysis due to the late submittal of samples. All three subseries were identified in the 31RH491 collection. The Hanover I sample was composed of only 7 sherds from 7 different vessels, the Hanover II sample consisted of 30 sherds represented by 20 different vessels and the Hanover III sample was made up of 10 sherds from 6 vessels. Each subseries is described below. (1) Hanover I. The Hanover I sample from 31RH491 was limited and consisted of only two sherds with identifiable surface treatments, a Hanover Ib Cord Marked (Figure 179:K) and a Hanover Ia Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 513 Chapter 10. 31RH491 Figure 179. New River and Hanover I and II Series Ceramics, Site 31RH491. A: New River Ia Plain (p1469), B-E: New River If Plain (p892, p2025, p1073 and p2530 respectively), F-G: New River If Fabric Impressed (p2893 and p1900 respectively), H: New River IIf Plain (p534), I: New River IIa Fabric Impressed (p740), J: New River IIf Fabric Impressed (p2308), K: Hanover Ib Cord Marked (p1028), L: Hanover Ia Check Stamped (p1697), M-P: Hanover IIa Check Stamped (p148, p934-Exterior, p934-Interior and p1074 respectively), Hanover IIb Check Stamped (p246), R-S: Hanover IIa (p1110 and p114 respectively), T-U: Hanover IIa Fabric Impressed (p529 and p724 respectively), V: Hanover IIe Fabric Impressed (p1798), W: Hanover IId Cord Marked (p988). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 514 Chapter 10. 31RH491 Check stamped (Figure 179:L). The remaining 5 sherds in the subseries had indeterminate and eroded exterior surfaces. The cord marked sherd displayed a parallel application pattern and cord impressions were about 1.13 mm in diameter. Lands on the check stamped sherd are present on only one side of a transverse and, suggesting that the sherd might represent a linear check stamped motif. The view of the motif is not sufficient to determine this, however, as this vacant field might correspond to the edge of carved paddle motif. The paste is nearly identical to that of New River II. Subangular quartz sand is abundant and ranges from medium to very coarse in grain size. Mean vessel wall thickness for the Hanover I sample was 7.74 mm (SD=1.20 mm). (2) Hanover II. The Hanover II sample from 31RH491 was dominated by check stamped and fabric impressed sherds. Of the sherds with recognizable exterior surface treatments (n=26), fabric impressed made up 54 percent of the sample, check stamped comprised 38 percent and the remaining 8 percent was represented by cord marked treatments. Calculating percentages by vessel representation versus sherd counts indicates that check stamping is slightly more popular. Of the 19 vessels with identifiable surface treatments fabric impressed makes up 53 percent, check stamping occurred on 42 percent and cord marking was found on only 5 percent. These percentages are comparable to those for the Quewhiffle site (31HK2521), although cord marking was more prevalent there. Four paste variants were recognized in the sample. Variant IIa comprises nearly two-thirds of the sherds (n=19), while variants IIb (n=3), IId (n=3) and IIe (n=5) range from 10 to 17 percent of the sample each. Variant IIe contains crushed indurated granite, which is primarily a feature of the Hanover III subseries. Check stamped impressions are generally faint and the lands are narrow (Figure 179:M-S). Application is sloppy and over-stamping is common. The fabric impressed sherds consistently exhibited medium wefts and rigid warps on those with identifiable impressions (Figure 179:T-V). Only one cord marked sherd displayed a recognizable stamp pattern (Figure 179:W). It was parallel in application and cord impressions averaged 0.64 mm in diameter. Five rim sherds from two vessels were present in the sample (Figure 179:N-O and V). Both vessels had direct orientations and appear to represent open mouth jars (Figure 180). In both cases the rims were thinned on both the exterior and interior. The check stamped rim exhibited check stamped field on its interior as well (Figure 179:O). Mean wall thickness for the sample was 6.99 mm (SD=1.22). (3) Hanover III. Ten sherds assignable to the Hanover III subseries were identified in the analysis, representing fragments of six different vessels (Table 160). Seven sherds exhibited recognizable exterior surface treatments. Of these 5 were cord marked and 2 were fabric impressed. The two treatments were proportionally equal in the vessel sample, each represented by two vessels. One fabric impressed sherd displayed a rigid warp and a medium weft (Figure 181:A). Recognizable cord marked sherds exhibited both oblique/perpendicular (Figure 181:B-C, E) and parallel (Figure 181:D) applications, but sherds from the same vessel displayed both patterns. Most of the sample was comprised of paste variants IIIc/e (crushed arenite and indurated granite) and IIIe (crushed indurated granite) (Table 161). These two variants made up 80 percent of the sample. Paste variants IIIa and IIId were represented by one sherd each. Mean wall thickness for the Hanover III sample was 7.84 mm (SD=1.13). Rim sherds were not present in the sample. Yadkin Series Coe (1964:30-32) defined Yadkin series pottery from his excavations at the Doershuck Site. Abundant, angular crushed quartz was the diagnostic trait of the series, which was characterized by three main surface treatments, fabric impressed, cord marked, and linear check stamped. Subsequent investigations at Town Creek (Coe 1995) also associated simple stamped exteriors with the series. In the North Carolina Sandhills, the definition of Yadkin has been expanded to include all types of crushed rock temper (Herbert 2003:58–59). Currently, examples of crushed granite, crushed quartzite, crushed and ground sandstone/siltstone conglomerate, subangular feldspar sand and granule/pebble subangular quartz sand have been identified in assemblages in the region. The results of the Quewhiffle site (31HK2521) Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 515 Chapter 10. 31RH491 A. Direct Thinned, Hanover IIa Check Stamped (p934) B. Direct Thinned, Hanover IIe Fabric Impressed (p1798) Figure 180. Rim Profiles, Site 31RH491. Figure 181. Hanover III, Yadkin III and Cape Fear III Series Ceramics, Site 31RH491. A: Hanover IIIe Fabric Impressed (p1486), B-E: Hanover IIIe Cord Marked (p2872, p42, p2883 and p2924 respectively), F-G: Yadkin IIId Plain (p1, F: Exterior and G: Interior, Floated), H: Yadkin IIId/e Plain (p1675), I: Yadkin IIIe Fabric Impressed (p50), J: Cape Fear III Cord Marked (p1769), K: Cape Fear III indeterminate (p1207). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 516 Chapter 10. 31RH491 ceramic analysis suggest that Yadkin III and Hanover III are at least partially contemporaneous. The presence of large pieces of crushed and ground sandstone/siltstone conglomerate particles in sherd collections within Fort Bragg was first recognized during Phase II testing at 31MR326 and 31HT748 conducted by Palmetto Research (Cable and Cantley 2005b). Due to its texture and composition, it can be confused with grog and probably has been on occasion in the Fort Bragg area of the Sandhills. The material is light tan to grayish yellow when viewed in sherds and generally contains fine to medium grains of arenite that can become finely disseminated in the paste matrix. The two materials are found in the same geological deposits and large particles of each can co-occur in individual sherds. Significant quantities of the sandstone tempered sherds at 31MR326 contained grog temper. The Yadkin material from 31RH491 exhibited hard pastes; consequently, all of it was assigned to the Yadkin III subseries. The Exterior surface treatments on individual sherds consisted of two fabric impressed (Figure 181:I), one cord marked and three plain or smoothed (Figure 181:F-H). Relative exterior surface treatment proportions controlled by vessel data are: 60 percent plain, 20 percent fabric impressed and 20 percent cord marked. One large plain sherd (p1) displayed a floated surface on its interior (Figure 181:G). Mean vessel wall thickness for the Yadkin III sample was 7.67 mm (SD=2.37 mm). Rim sherds were not present in the sample. Cape Fear Series Stanley South (1976:18-20) first described the Cape Fear series, or ware-group as he described it, from surface collections made in 1960 on sites located between Wilmington, NC and Myrtle Beach, SC. He applied the term to virtually all of the sand-tempered cord marked, fabric impressed, and net impressed pottery within the collection. Many have suggested that the series was too broadly applied by South and other investigations in North Carolina indicate that such material should be separated into a sequence of series based on sand inclusion coarseness (see Anderson and Logan 1981:107-108; Trinkley 1981:11). Coastal North Carolina sequences contain at least two sand grain size modes that are used to distinguish sand-tempered series (Phelps 1983). The Deep Creek series consists of pastes with abundant, coarse sand and it is chronologically correlated with Deptford. Other series (i.e. Mount Pleasant and Cape Fear) contain finer grained pastes and they are roughly dated to a post-Deptford or late Middle to Late Woodland context (i.e. AD 500-900 or 1500 to 1100 BP). Herbert (2003:156) distinguishes the sandtempered New River and Cape Fear series by quartz sand density in the North Carolina Sandhills. Cape Fear series material is characterized by low to moderate densities of quartz sand amounting to less than 15% of the paste body. In addition, he places all perpendicular stamped cord marked exterior surfaces in the Cape Fear series, while all net-impressed is assigned arbitrarily to New River. Cable and Cantley (2005a, b) have distinguished two series variants for Cape Fear in the Fort Bragg area based on quartz sand density and paste characteristics. Cape Fear I was characterized by a friable, gritty paste and a moderate density of medium to coarse quartz sand temper. Cape Fear II exhibited a finer, more compact paste with sparse medium to coarse quartz sand temper. Four Cape Fear series sherds were identified in the collection. Just as was true of the Quewhiffle site sample, the sherds possessed an extremely hard to vitrified paste. Consequently, they were assigned to a third Cape Fear subseries, Cape Fear III, which is assumed to belong to the Late Woodland period. The paste was composed of abundant fine to medium quartz sand with apparently few coarser grained inclusions and cores were generally dark. Cape Fear III series paste resembles most closely the paste varieties that characterize Mississippian series pottery as it is represented in the upper Pee Dee River Valley (Town Creek), the upper Wateree River Valley (Mulberry-Adamson) and the lower Santee River Valley (Jeremy-Tibwin). Whether there is a direct correlation with neighboring Mississippian cultures is an issue that cannot be resolved at present. Three of the four sherds displayed indeterminate exterior surface treatments, while one was cord marked (Figure 181:J). A single, small rim sherd was present in the collection (Figure 181:K). Insufficient area below the lip was available to confidently determine the orientation of the sherd. The lip was subrounded and the rim may have been slightly flared. The meager Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 517 Chapter 10. 31RH491 sample produced a mean vessel thickness of 8.08 mm (SD=0.65). Concluding Remarks The ceramic assemblage recovered from the 31RH491 evidences Early, Middle and Late Woodland occupation. Vessel representation suggests that about 25 percent of the occupation was associated with New River components, 46 percent was made up of Hanover I and II components and about 29 percent of the components were affiliated with the Late Woodland period (i.e. Hanover III, Yadkin III and Cape Fear III). Historic Artifacts One hundred and thirty historic artifacts were recovered during the Phase II operation (Table 162). Most of this material was affiliated with the military occupation of Camp MacKall, much of it dating to the World War II time period. A fragment of a clay ball pipe bowl, however, probably dates to the nineteenth century. During the Phase I survey a piece of Bristol slipware was recovered that would probably belong to this same occupation. The surveyors (Grunden and Ruggiero 2006:482) did not describe the provenance of the slipware, other than to mention that it was found on the surface of the site. The pipe bowl fragment was found in a Stage II shovel tests excavated at grid coordinates N340/E605 at a depth of 10 to 20 cm bs on the rise at the extreme south end of the site. OCCUPATION PATTERNS Diagnostic stone tool markers of culturechronological association were abundantly recovered at 31RH491. The recovery of a Taylor Side Notched point, two Palmer I Corner Notched points, a Palmer Corner Notched point of indeterminate subtype and a wide range of scraper and uniface types documented extensive evidence of Early Archaic occupation. The scraper collection is one of the largest from any site in the North Carolina Sandhills. In particular, the 11 end scrapers in the collection exhibit very low depth positions consistent with Early Archaic affiliations. Vertical distributions of side scrapers exhibit more variable depths, but a significant proportion of them occur in the Early Archaic horizon of the site matrix. Middle Archaic occupation is also extensively represented in the diagnostic stone tool assemblage. Three partly diagnostic square stems with concave bases, all of which display basal grinding, are most likely representative of Kirk Stemmed/Serrated and Stanly Stemmed phases. In addition, a Morrow Mountain I Stemmed point, a Morrow Mountain II Stemmed point and a Guilford Lanceolate, Straight Base variant identify later Middle Archaic occupation. Finally, two Savannah River Stemmed points and a preform evidence Late Archaic occupation. Ceramic analysis identified the presence of Early Woodland (New River I), early Middle Woodland (New River II), late Middle Woodland (Hanover I and II) and early Late Woodland (Hanover III/Yadkin III) components. In addition, the Cape Fear series, which was present in small quantities, was present in small quantities and it may represent temporary, seasonal intrusions by transitional Mississippian groups. In addition, two Yadkin Large Triangular, variant B points were recovered, which are probably associated with the Middle Woodland ceramic phases. 31RH491 is characterized by high artifact density. Mean artifact density in positive Stage I and II shovel tests (n=342) was 5.03 artifacts (SD = 6.46), indicating that two-thirds of the positive tests yielded densities between one and 12 artifacts. Although the modal positive shovel test outcome across the sample was only one artifact, 57 percent of the positive tests contained three or more artifacts. The entire site area (including negative Stage I and II shovel tests) had a mean artifact density of 2.54 artifacts (SD = 5.23) per shovel test. These values are comparable to artifact densities at 31HK2510 and 38RH480 and about half of that seen at 31HK2521. Metavolcanic debitage recovered during Stage I and II shovel testing was primarily concentrated on the two rises, in the area of the wildlife plot and on the south end of the site (Figure 182). The contour algorithm suggests that the southern concentration was truncated by road grading near the gravel road and most probably extended 20 to 30 m farther to the east at one time. Quartz debitage densities were much lower and the main area of concentration was Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 518 Chapter 10. 31RH491 Table 162. Historic Artifact Inventory, 31RH491. HISTORIC ARTIFACTS: 31RH491 HISTORIC ARTIFACT TYPE ARTIFACT COUNT Brick Fragment 1 Formica Laminate 2 Window Glass 8 Cut Nail 1 Square Twisted Nail 5 Wire Nail 2 Brown Glass Bottle w/ Metal Screw Cap 1 Clear Bottle Glass 8 Clear Bottle Glass, Embossed 6 Clear Container Glass Clear Container Glass Base 1 Clear Container Glass Screen Top Lid 1 Milk Bottle, Clear Glass 1 Milk Glass 6 Plastic Salt/Pepper Shaker Screen Cap Lid 1 Soda Bottle, Clear Glass 1 Whiteware 3 Whiteware Plate 3 Road Gravel 4 UID Flat Metal 9 Washer 1 Foil Button 1 India Ink Container w/ Cap 1 Clay Ball Pipe Bowl Fragment 2 Armament Clip 4 Lead Bullet 1 M14/16 Casing 4 M14/M16 Casing 7 M60 Bullet Belt Link 1 GRAND TOTAL 44 located in the wildlife plot (Figure 183). Vertical distributions by raw material type indicate that debitage is centered between 10 and 50 cm bs, with a central tendency situated between 20 and 40 cm bs (Table 163). From what we have learned about the vertical distributions of cultural-chronological occupations on sites in the Sandhills (see Cable 2010; Cable and Cantley 2005a, b, 2006), this distribution should indicate that most of the lithic debris is associated with the Archaic period, which is consistent with the wide 130 range of Archaic projectile point styles and scraper types. Ceramic vertical distributions indicate that the Woodland period occupation at the site is positioned in the upper 20 cm of deposit (Table 164). These data include all shovel tests from the site sample. Sherds below 30 cm bs can be confidently assumed to be “out of position” and were likely displaced by natural disturbances, precontact feature construction, modern Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 519 Chapter 10. 31RH491 land use activities or recovery error. Approximately one-quarter of the lithic debitage is situated in the upper 20 cm of deposit. Although it is not likely that all of this material is associated with Woodland period occupation, it is reasonable to suggest that as much as 10 percent of the lithic debris on the site is associated with Woodland occupation. Horizontal ceramic distributions (Figures 184-186) indicate that Woodland occupations are concentrated in the northern end of the site in the area of the wildlife plot and on the slope leading down to the Drowning Creek floodplain on the east side of the gravel road. Data used to generate these maps included Stage III shovel tests as well, so that a broader picture of ceramic distributions could be viewed. There is a tendency for “chronologically adjacent” ceramic series to exhibit a degree of horizontal overlap. Whether these distributions reflect punctuated reoccupation intervals (sub-generational) by lineally related social groups or simply unrelated reoccupation by small social groups over larger time spans is not clear. This is an issue that should be addressed for all cultural periods when addressing the occupation history of sites, but it will require data recovery efforts to effectively investigate these patterns. If these distributions reflect punctuated, serial reoccupation, it would stand to reason that campsite selection during reoccupation would have been conditioned by previous camp locations. A more detailed appreciation for precontact land-use patterns at the site can be derived from a consideration of occupation type distributions and associations. Currently, four occupation types have been recognized on sites in the Fort Bragg area (see Cable and Cantley 2005c, 2006): (1) Type I occupations consist of a debitage concentration and an associated tool cluster situated at a confined location along its periphery. It is inferred that the tool cluster was formed around a hearth and in several instances calcined bone fragments have been recovered from the deposit delimited by a tool cluster. It is also believed that ephemeral shelters were situated adjacent to these tool clusters and opposite the concentrations. Debitage concentrations appear to represent lithic reduction loci established away from a shelter or sleeping area where tools were manufactured during the stay at the camp. These concentrations are generally composed of single lithic raw material types, and the associated tool clusters commonly consist of manufacturing rejects and discarded broken or worn-out tools made of the same raw material that comprises the associated debitage concentration. Although many Type I residences occur in isolation and appear to be the by-products from single nuclear or small extended family occupations, others appear to represent multifamily occupations that extend beyond the ability of a single shovel test to identify. Woodland period Type I occupations are characterized by an additional element, sherd aggregations generally composed of partial vessels. These occur adjacent to debitage concentrations and they are hypothesized to occupy a portion of the living floor associated with a hut. The model would stipulate that tool clusters are situated between the debitage concentrations and sherd aggregations. (2) Type II residences are similar in spatial organization to Type I residences, but they exhibit much higher tool-to-debitage ratios and may commonly be comprised of diverse lithic raw material types. Lithic reduction on these sites consisted of tool maintenance and late stage shaping of tools from flake blanks or reduced biface cores. Consequently, debitage density is much lower than that of Type I debitage concentrations and the scatters consist primarily of late stage reduction debitage. Type II occupations are inferred to be the output from single or small, multiple household aggregations and are thought to represent the residue from High Technology Forager (HTF) residences (Spiess 1984; Todd 1983:231-233). The HTF model proposes a specialized forager adaptation combining high logistical and high residential mobility into a single settlement system characterized by highly curated technologies. Type II residences have thus far only been assigned to the Early Archaic period at Fort Bragg. The lithic raw material diversity characteristic of Type II residences suggests that these occupations represent the aggregation of social units that may have been seasonally dispersed, transporting a wide range of lithic raw materials from different sources. (3) Type III occupations are inferred to represent logistical camps and very short-term residences characterized by low-density debitage scatters and sporadic tool discard. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 520 Chapter 10. 31RH491 560 MN 540 l .0 520 83 na Ca 0 10 20 meters Gun Implacement 500 .5 82 Positive ST Food Plot 480 Negative ST .0 82 460 lR ve ra G d oa 31RH491 440 83.5 .5 81 420 83 .5 84.5 400 .0 84 83.5 380 360 .5 84 340 84.0 320 83.0 300 280 260 240 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 182. Density Distribution of Metavolcanic Debitage, Stage I and II Shovel Test Sample, Site 31RH491 (Contours = 1 piece of debitage). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 521 Chapter 10. 31RH491 560 MN 540 l .0 520 83 na Ca 0 10 20 meters Gun Implacement 500 .5 82 480 Food Plot Positive ST Negative ST .0 82 460 d oa lR ve ra G 31RH491 440 83.5 .5 81 420 83 .5 84.5 400 .0 84 83.5 380 360 .5 84 340 84.0 320 83.0 300 280 260 240 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 183. Density Distribution of Quartz Debitage, Stage I and II Shovel Test Sample, Site 31RH491 (Contours = 1 piece of debitage). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 522 Chapter 10. 31RH491 Table 163. Vertical Distribution of Debitage by Raw Material Type, Stage I and II Shovel Tests, 31RH491. VERTICAL DISTRIBUTION OF DEBITAGE BY RAW MATERIAL TYPE, STAGE I AND II SAMPLES: 31RH491 LEVEL RAW MATERIAL TYPE 1 2 3 4 5 6 4 1 11 17 20 24 14 7 8 9 10 GRAND TOTAL Metavolcanic Hydrated Metavolcanic Mill Mountain Rhyolite Type I Rhyolite Tuff 5 3 2 14 20 13 14 5 35 50 59 43 21 8 Uwharries Southern Rhyolite (1) 6 34 68 43 31 11 1 5 Uwharries Southern Rhyolite (2) 48 114 177 162 123 42 11 5 7 19 17 23 12 4 1 Uwharries Eastern Rhyolite Uwharries Western Rhyolite 2 89 1 71 216 199 2 1 685 83 Metasedimentary Black 1 2 Green 3 1 7 3 2 13 Quartz Crystal Quartz 1 2 1 1 2 2 White Quartz 17 26 47 40 32 10 White Quartz, Fractured 9 3 1 176 3 3 9 7 16 427 366 Other Schist GRAND TOTAL 128 277 252 82 21 12 2 1 1568 Table 164. Vertical Distribution of Precontact Ceramics, All Shovel Tests, 31RH491. VERTICAL DISTRIBUTION OF PRECONTACT CERAMICS, ALL SHOVEL TESTS: 31RH491 CERAMIC SUBSERIES LEVEL 1 2 3 New River I 3 4 2 New River II 1 1 1 Hanover I 2 3 Hanover II 11 10 Hanover III 3 2 Yadkin III 4 1 Cape Fear III 3 1 GRAND TOTAL 27 22 4 5 1 6 1 1 GRAND TOTAL 11 4 5 3 2 3 27 1 6 1 8 4 8 6 1 1 65 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 523 Chapter 10. 31RH491 560 MN 540 l .0 520 83 na Ca 0 10 20 meters Gun Implacement 500 .5 82 Food Plot Positive ST 480 Negative ST .0 82 Contours=1 Sherd New River I 460 G lR ve ra New River II d oa 31RH491 440 83.5 .5 81 420 83 .5 84.5 400 .0 84 83.5 380 360 .5 84 340 84.0 320 83.0 300 280 260 240 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 184. Density Distribution of New River Series Sherds, All Shovel Tests, Site 31RH491 (Contours = 1 sherd). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 524 Chapter 10. 31RH491 560 MN 540 l .0 520 83 na Ca 0 10 20 meters Gun Implacement 500 .5 82 Food Plot Positive ST 480 Negative ST .0 82 Contours=1 Sherd Hanover I 460 lR ve ra G Hanover II d oa 31RH491 440 83.5 .5 81 420 83 .5 84.5 400 .0 84 83.5 380 360 .5 84 340 84.0 320 83.0 300 280 260 240 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 185. Density Distribution of Hanover I and II Series Sherds, All Shovel Tests, Site 31RH491 (Contours = 1 sherd). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 525 Chapter 10. 31RH491 560 MN 540 l .0 520 83 na Ca 0 10 20 meters Gun Implacement 500 .5 82 Food Plot Positive ST 480 Negative ST .0 82 Contours=1 Sherd Hanover III 460 lR ve ra G Yadkin III Cape Fear III d oa 31RH491 440 83.5 .5 81 420 83 .5 84.5 400 .0 84 83.5 380 360 .5 84 340 84.0 320 83.0 300 280 260 240 420 440 460 480 500 520 540 560 580 600 620 640 660 Figure 186. Density Distribution of Hanover III, Yadkin III and Cape Fear Series Sherds, All Shovel Tests, Site 31RH491 (Contours = 1 sherd). Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 526 Chapter 10. 31RH491 (4) Type IV occupations represent extraction or processing loci that are not believed to have involved, necessarily, overnight stays. They are difficult, however, to distinguish from Type III occupations, even in situations of greater excavation exposure. Isolated sherd aggregations are hypothesized to represent one kind of Type IV element associated with Woodland occupations. It is difficult to distinguish effectively between these various occupation types at shovel test intervals of greater than 1.25 m, but a less precise classification can be used at greater shovel test intervals to aid in population projections. High-density debitage concentrations tend to indicate Type I residences, while low-density scatters tend to signal short-term residences, special purpose camps and locations (i.e. Type II, III and IV occupations). Highdensity debitage concentrations are temporarily stipulated to correspond to shovel test outcomes of five or more pieces of debitage of a specific raw material subtype. Eventually, it will be necessary to confirm this arbitrary threshold by examining shovel test counts from a large sample of known high-density debitage concentrations. Deposits containing less than five pieces of debitage are viewed as more likely to represent low-density debitage scatters associated with the other three site types. Since shovel tests can intersect the peripheral zones of high-density debitage concentrations where debitage densities are much lower, however, it is likely that Type I residential occupations will be under-represented in macrointerval shovel test data. Because it is difficult to determine whether a single occupation is represented in more than one shovel test, in the cases of multi-family residences or associated special activity zones on the periphery of a camp, the term “element” is preferred over the term “occupation” to refer to the occupational data in shovel tests. At a sampling interval of 5-m or 10-m intervals, we are assured that the presence of an element will only rarely be repeated in adjacent shovel tests. Elements are identified by lithic raw material subtypes and/or by precontact sherd aggregations associated with one or more vessels. This methodology runs the risk of overestimating occupations because more than one raw material subtype may comprise a single occupation. The “element” concept recognizes this fact, focusing on depositional events and episodes rather than entire occupations. Elements can represent segments of contemporaneous multi-family occupations, separate occupations of various function, or contemporary members of multiple raw material reduction episodes. The current methodology for identifying elements in shovel tests consists of running crosstabs of lithic debitage raw material subtypes or subtype groups by shovel test number in a spreadsheet. Sherd aggregations are not used in these calculations because, in general, they are associated with debitage scatters and their inclusion would result in potentially double counting a good number of elements. Once the cross-tab is generated, the sum of each lithic raw material subtype in a shovel test is identified as an element. In some instances, the vertical separation between individual items of the same lithic subtype can be so great as to suggest the presence of distinct elements. Due to the vagaries of bioturbation and possible recovery errors in which higher positioned material can be included in lower levels through sidewall displacement, however, the lower item(s) is assigned to the higher positioned item(s) to define a single occupation element. In cases where large debitage concentrations of the same lithic raw material type are found at distinctively separated levels, though, each is counted as separate elements. Elements have horizontal dimensions and, as a result, sample units that are smaller than those dimensions have the benefit of increased sample efficiency (see Cable and Donaldson 1988; Rice 1987; Rice and Plog 1983). The typical element found at Fort Bragg has a diameter of about 3 m. Thus, a shovel test of .09 m2 samples only .09 percent of the area of a 10 m-square, but it samples 9.0 percent of the theoretical mean element space of 9.0 m2. Consequently, a 10 m-interval shovel test pattern actually samples about 9.0 percent of the available element space at a site, while a 5 m-interval shovel test pattern samples about 36.0 percent of the available element space. A mean estimate of the population of elements at shovel test intervals of 10 m and 5 m, then, requires a simple arithmetic calculation of counting the occupation elements identified in shovel tests and dividing this figure respectively by 9.0 percent and 36.0 percent. Confidence intervals can be constructed around these means, but they have not been calculated in the Fort Bragg testing projects. Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 527 Chapter 10. 31RH491 Although Stage I shovel test samples have been relied upon in the past to produce population projections because of their even and complete coverage, a concerted effort was made in the DO5 package to extend 5 m-interval shovel tests across the entire site area so that advantage could be taken of greater sample density. Following the procedures outlined above, the method for 5 m-interval patterns succeeded in identifying 41 Type I elements and 751 indeterminate (Type II, III and IV) elements (Table 165). As would be anticipated by the raw material type frequencies, USR (2) is most frequently identified in the census, comprising nearly 29 percent of the identified elements. Of numerical secondary importance are MMR, UER, USR (1) and white quartz, which range between 9.5 and 18.1 percent of the sample. Based on these sample results it can be further estimated that the site contains a mean of 2,200 elements, of which 114 represent Type I elements. A relatively large number of stone tools (n=42) were identified in isolation from matching debitage in shovel tests. The proportional representation of Type I residential elements (5.5 percent) is relatively low, suggesting that the predominant precontact land use history of the site was short-term camps of single households or specially comprised task groups of less than a weeks duration. A large segment of the Fort Bragg site assemblage has been demonstrated to contain as much as 20 to 30 percent Type I residences (see Cable 2010). Expanded over nearly the full range of human occupation (about 10,000 years), the population estimate indicates that precontact groups, on average, visited the site a maximum of once every 4.5 years. The occupation, however, was probably not evenly distributed through time. The site was no doubt used intensively by specific culture-historic groups, probably for strings of continuous years punctuated by abandonments in response to long-term land-use patterns dependent upon resource depletion cycles. Given the large number of occupations, it was anticipated that a high incidence of superimposi- Table 165. Recognized Elements for the Stage I and II Shovel Test Sample, 31RH491. IDENTIFIED ELEMENTS FROM THE STAGE I & II SHOVEL TEST SAMPLE: 31RH491 RAW MATERIAL TYPES TYPE I Crystal Quartz GRAND TOTAL PERCENTAGE OF ELEMENTS 11 11 1.4 Metasedimentary, Black 2 2 0.3 Metasedimentary, Green 12 12 1.5 MMR 1 74 75 9.5 Type I RT 2 36 38 4.8 UER 8 106 114 14.4 USR (1) 5 138 143 18.1 USR (2) 22 206 228 28.8 UWR 1 48 49 6.2 White Quartz 2 115 117 14.8 3 3 0.4 751 792 100.0 Fractured White Quartz GRAND TOTAL TYPE II, III, IV 41 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 528 Chapter 10. 31RH491 tion of elements would be documented at 31RH491. It was relatively high, averaging 2.16 (SD=1.28) elements per shovel test, which is comparable to superimposition at 31HK2502 (2.24) and 31HK2521 (2.07). Much of this is due to occupation concentrating of the ridge top at the southern end of the site and the low-relief rise at the north end. Superimposition was much lower at 31HK2510 (1.62) and 31RH480 (1.60). Not included in the population estimate was the Woodland period ceramic collection. It is assumed that an unspecified proportion of debitage elements represent Woodland occupation and that including ceramics in the calculations would result in double counting elements. However, to arrive at some perspective on the magnitude of Woodland occupation at the site, an independent estimate of total vessels can be projected. In the combined Stage I and II shovel test sample, 41 individual vessels were recovered. The standard approach to this problem has been to calculate vessel populations using the aggregated model of vessel deposit organization. We have seen in previous analyses of the ceramic assemblages from sites in the DO5 package, however, that ceramic deposits appear to be much more disorganized and displaced than the aggregated ceramic model of vessel population estimate would assume. That model stipulates that we can expect most sherds from single vessels to be contained within an area of about 2 m in diameter. Under this assumption the Stage I and II shovel tests would be expected to achieve 16 percent coverage for sherd aggregate space. In the case of 31RH491, the aggregated model would yield a total vessel population estimate of 256. There is good reason to believe, however, that this method of calculation underestimates the actual vessel population, as was discussed in depth in Chapter 5. It was also concluded from that discussion that it is unlikely that many full pot concentrations survived on these sites due to recycling of large sherds and off-site transport of the larger vessel fragments. An idea of the magnitude of the sherd population at the site can be supplied by calculating a straight area mean. Since nearly all of the ceramics are concentrated on the flat ridge at the north end of the site, we will limit this evaluation to the area of greatest sherd concentration, which covers approximately 6,000 m2. Shovel tests in this area numbered 287, resulting in a straight areal coverage of 25.83 m2. This yields an area sample of 0.43 percent. Forty-seven sherds, weighing 157.59 gm, were recovered from the area of greatest sherd concentration, resulting in a population mean estimate of 10,930 sherds and a total sherd weight of 36.56 kg. If whole pots were entirely represented in the assemblage, it is doubtful that this total would be sufficient to cover the weight of more than 27 typical open-mouth jars from the region (see Herbert and Irwin 2003:2). Given an average weight of 3 pounds (1,377 gm) for a typical pot, if there were 256 whole pots present on the site, a mean population weight of 352 kg would be required. Consequently, only about 10.4 percent of the expected vessel weight relative to the estimate of the vessel population is present on the site. Similar results were obtained at 31HK2502, 31HK2510, 31HK2521 and 31RH480, where only about 10, 18, 30 and 22 percent, respectively, of the expected sherd weight was found in the sample. This suggests that the intensity of use recycling is great on Sandhills sites and that large sherds from broken pots were probably recycled for other uses (see David and Hennig 1972:21-21) and stored in caches or transported upon leaving camps, while smaller sherds were probably left at the point of breakage. If so, sherds from the same pot might be scattered at distances from one another on sites and also spread across multiple sites. Another factor that would add to this displacement would be reoccupation on top of old campsites. Other methods of calculating vessel populations at 31HK2521 resulted in values on the order of 4 to 20 times larger than those produced by the aggregated model. That is, the aggregated model under-estimates the number of vessels represented at sites because ceramic deposits are disorganized and contain primarily only portions of individual vessels. The Stage I and II shovel test sample succeeded in identifying New River I, New River II, Hanover II, Hanover III, Yadkin III and Cape Fear III series ceramics. Using the proportions of individual vessels associated with each of these sub-series (Table 166) it can be inferred that 17.1 percent of the Woodland occupation is affiliated with the Early Woodland period (New River I), 4.9 percent with the early Middle Woodland (New River II), 46.3 percent with the middle and late Middle Woodland period (Hanover I and Hanover II) and 24.4 percent with the early Late Woodland (Hanover III and Yadkin Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 529 Chapter 10. 31RH491 Table 166. Representation of Ceramic Vessels by Sub-series, Stage I and II Shovel Tests, 31RH491. REPRESENTATION OF VESSELS BY SUBSERIES, STAGE I AND II SAMPLE: 31RH491 CERAMIC SUBSERIES VESSELS (n) PERCENT REPRESENTATION New River I 7 17.1 New River II 2 4.9 Hanover I 3 7.3 Hanover II 16 39.0 Hanover III 5 12.2 Yadkin III 5 12.2 Cape Fear III 3 7.3 41 100.0 GRAND TOTAL III). Cape Fear III, which is postulated to represent an Early Mississippian intrusion into the area, comprised 7.3 percent of the vessel sample. Numerous studies have documented vertical patterning in the relative depth of deposits in the unconsolidated sandy soils of the Coastal Plain (Cable and Cantley 2005 b, 2006; Michie 1990). The actual depth patterns vary according to the character of the deposit. Those deposits with higher B-horizons tend to have compressed sequences, while those in deep C-horizon and E-horizon sediments generally display more discreet and vertically expansive sequencing. Michie’s (1990) C-horizon model developed in the Waccamaw Neck area on the central coast of South Carolina begins with Mississippian and Woodland materials positioned in the upper 30 cm of sediment. This zone is underlain by ceramic Late Archaic occupations situated at 28 to 35 cm bs, Middle Archaic horizons at 35 to 55 cm bs, and Early Archaic corner-notched and side-notched components at 55 to 62 cm bs. This model is generally applicable to sites on Fort Bragg. However, the deposits at Fort Bragg appear to be somewhat more compressed, judging by the vertical positions of samples of diagnostic artifacts. At Fort Bragg three vertical groupings have been developed for deep Sandhills sites: (1) Woodland/Late Archaic/Middle Archaic between 15 and 40 cm bs, (2) Middle Archaic/Early Archaic between 35 and 50 cm bs, and (3) Early Archaic below 50 cm bs. The divisions are broadly constructed and realis- tically anticipate a certain degree of vertical mixing and overlap between the various periods. This model can be of great utility in reconstructing the occupation history of sites like 31RH491 where the predominant artifact class is non-diagnostic lithic chipped stone debitage. Vertical positions of the various elements recognized in the Stage I and II shovel test sample allow a rough picture of the representation of occupation periods at the site (Table 167). Most elements were situated in a single level, but some, especially the higher density elements, spanned several levels. In these latter cases, the central tendency of the vertical distribution of a particular element was used to estimate its position. Vertical data from the shovel test sample were not precisely congruent with the model ranges discussed above. However, the following adaptation was developed to correlate culture-historic association with level data: (1) 0-20 cm bs, Woodland, (2) 20-40 cm bs, Woodland/Late Archaic/Middle Archaic, (3) 40-50 cm bs, Middle Archaic/Early Archaic and (4) 50 to 80 cm bs, Early Archaic. Approximately 22.2 percent of the occupation elements are positioned below 40 cm bs. These depths are inferred to represent Middle and Early Archaic occupation, but it is likely that Early Archaic elements predominate at this depth range based on stone tool distributions. About 51.4 percent of the elements are situated between 20 and 40 cm bs and are inferred to represent Middle Archaic, Late Archaic and Woodland components. The actual Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 530 Chapter 10. 31RH491 Table 167. Inferred Culture-Historic Associations of Recognized Lithic Elements, 31RH491. INFERRED ASSOCIATIONS OF RECOGNIZED LITHIC ELEMENTS, STAGE I & II SHOVEL TEST SAMPLE: 31RH491 MMR TYPE I RT UER USR (1) USR (2) UWR WHITE QUARTZ FRACTURED QUARTZ CRYSTAL QUARTZ METASEDIMENTARY, BLACK VERTICAL DEPTH RANGE (cm bs) METASEDIMENTARY, BLACK LITHIC ELEMENTS 0-20 3 1 1 24 7 41 28 51 20 33 Woodland/Late & Middle Archaic 20-40 4 1 9 38 20 58 77 116 24 57 Middle & Early Archaic 40-50 2 2 10 9 9 25 39 4 21 121 Early Archaic & Paleoindian 50-90 2 3 2 6 13 22 1 6 55 75 38 114 143 228 49 117 INFERRED ASSOCIATION Woodland GRAND TOTAL 11 2 12 GRAND TOTAL 209 3 3 407 792 representation of Woodland occupation is difficult to determine, but the earlier estimate of 10 percent may be close. This would indicate that the remaining percentage of 14.7 percent is probably associated with Middle and Late Archaic. About 78 percent of the Type I elements are positioned below 20 cm bs and most of these are likely Archaic in affiliation (Table 168). The remaining Type I elements in the upper 20 cm of deposit, which is a plow zone, are no doubt a mixture of Woodland and later Archaic components. SAMPLE BLOCK INVESTIGATIONS Fifteen shovel test outcomes from the Stage I and II sample were targeted for close-interval shovel testing to further elaborate the character of the occupations extant at the site. In general, a series of eight additional shovel tests was excavated at 1.25 m-intervals surrounding the target shovel test. These are referred to as sample blocks (SBs) and their locations Table 168. Inferred Culture-Historic Associations of Recognized Lithic Elements, 31RH491. VERTICAL DISTRIBUTION OF INFERRED TYPE I ELEMENTS: 31RH491 RAW MATERIAL TYPE LEVEL 1 2 3 4 GRAND TOTAL 1 1 1 2 1 8 Uwharries Eastern Rhyolite 4 3 Uwharries Southern Rhyolite (1) 1 2 2 Uwharries Southern Rhyolite (2) 3 5 8 5 5 1 1 Uwharries Western Rhyolite 6 1 Mill Mountain Rhyolite Type I Rhyolite Tuff 5 White Quartz 1 1 GRAND TOTAL 9 12 22 1 2 11 8 1 41 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 531 Chapter 10. 31RH491 are illustrated in Figure 171. Sample block locations coincided with the areas of the two main artifact concentrations at the site, at the north end in the wildlife plot and east of the gravel road and on the southern ridge top. Table 169 displays the vertical distributions of debitage by raw material type for each of the sample blocks. Five of the sample blocks contained profiles predominantly confined to the upper 40 cm of deposit (SBs 1, 7, 9, 11 and 14). The remaining SBs reflected slightly lower distributions, with debitage concentrated between 30 and 60 cm bs. Due to the variable deposit depth across the site and loci of disturbance, the vertical model was of only of marginal utility, however. Ceramics were not targeted in the sample block investigation because of their extremely fragmented and eroded condition. Plowing contributed to this condition, but recycling, as has been repeatedly discussed in this report, was also instrumental is the high degree of disorganization and displacement of the ceramic deposit. However, small quantities of sherds were recovered from 10 of the sample blocks (Table 170). The high degree of ceramic deposit disorganization demonstrated to be extant on sites in the DO5 package strongly indicates that methods for projecting vessel population emphasizing aggregated vessel completeness lead to significant underestimation. An alternative method, discussed earlier in Chapter 7, uses the area of the sample block as the basis for an areal sample fraction of vessels. Each sample block covers an area of 6.25 m2. The locations of sample blocks were not selected randomly, but they achieve a relatively even coverage of the area of primary ceramic concentration at the north end of the site. Consequently, an argument can be made that the nine SBs in this area yield a representative sampling of this 6,000-m2 frame, which contains most of the ceramic age occupation at the site. Each of the sample blocks would constitute a 0.26 percent sample fraction of the frame and together they achieve a 2.34 percent sample fraction. Fifteen vessels were found in the nine sample blocks established at the north end of the site. This would yield a mean vessel population projection of 641, which is about 2.5 times the mean estimate of 256 generated from the aggregated vessel model discussed earlier. Moreover, since it is reasonable to assume that not all vessels within each sample block were recovered (the shovel tests usually cover only about 13 percent of the block area), this projection should be viewed as a minimum vessel estimate. In addition, a small amount of the ceramic deposit occurs in the southern part of the site, but was not accounted for in this estimate. This may seem like an outrageously high estimate for a site where ceramics are not particularly dense, but over a span of about 1,000 years, the minimum span for the Woodland occupation based on ceramic types, this would yield a mean vessel deposition rate of only 0.641 vessels per year. All of the sample blocks contain a wide variety of lithic raw material types. Much of this heterogeneity can be attributed to occupation superimposition, but some of it may result from site functional variability. Type II residences characteristically contain heterogeneous lithic raw material profiles and the high density of scraper forms in the stone tool inventory suggests that a significant number of the occupations at the site may represent Early Archaic Type II residences. The sample block frames generally cover an area of only about 2.5 m square (6.25 m2), which is smaller than the average element size. Moreover, it would be uncommon for the sample block to expose nearly whole elements. However, a method was devised during the DO4 investigation (Cable 2010) to identify and distinguish Type I and III elements that are significantly contained within the extent of a sample block. Three measures related to shovel test outcomes were generated to describe this variability (Table 171). These are: (1) number of positive shovel tests within which the element occurs, (2) number of shovel tests with greater than or equal to 5 pieces of debitage and (3) mean debitage frequency per positive test. Since Type I residential elements are characterized by high-density debitage concentrations, they are expected to yield high outcome values for each of these measures. Type III residences, by contrast, are defined by the presence of low-density debitage scatters of about the same size as Type I high density scatters. Consequently, they should be expected to measure high or moderate on the number of positive shovel tests within which they are represented, but low on the number of shovel tests yielding five or more pieces of debitage of a specific raw material type and low on the mean frequency of debitage per positive Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 532 Chapter 10. 31RH491 Table 169. Vertical Distribution of Debitage by Sample Block, 31RH491. VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 1-5:31RH491 SB1 LEVEL RAW MATERIAL 1 2 Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 5 6 2 2 2 15 3 1 1 4 SB2 RAW MATERIAL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL LEVEL 1 1 1 2 2 3 2 2 1 1 3 1 2 1 1 6 10 10 4 Metasedimentary, Green Mill Mountain Rhyolite Schist Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 8 9 Schist Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 2 1 1 1 3 1 1 1 5 2 5 11 3 2 1 1 1 5 3 20 4 1 38 4 Hydrated Metavolcanic Mill Mountain Rhyolite Schist Type I Rhyolite Tuff Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 7 8-10 3 1 1 5 5 1 4 1 13 7 1 5 6 7 1 7 3 1 8 3 1 1 5 6 7 8-10 8-10 1 3 1 4 16 1 26 GRAND TOTAL 2 9 3 8 27 2 1 52 1 GRAND TOTAL 2 3 1 5 8 11 55 7 1 93 LEVEL 1 2 3 4 3 4 4 1 41 1 61 35 7 4 46 1 63 2 38 8 4 5 6 7 1 5 1 9 14 2 1 21 1 SB5 RAW MATERIAL 6 LEVEL SB4 RAW MATERIAL 5 1 SB3 RAW MATERIAL GRAND TOTAL 7 6 3 2 1 19 4 GRAND TOTAL 1 12 2 167 3 4 189 LEVEL 1 2 3 1 2 8 1 3 3 3 15 4 1 4 4 2 2 1 14 3 2 1 6 1 1 GRAND TOTAL 1 4 12 1 9 10 2 39 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 533 Chapter 10. 31RH491 Table 169. Vertical Distribution of Debitage by Sample Block, 31RH491 (Continued). VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 6-10:31RH491 SB6 RAW MATERIAL LEVEL 1 Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 2 3 4 5 6 1 1 2 1 7 7 5 7 5 5 6 1 2 2 3 4 1 6 7 5 6 7 6 7 2 3 1 1 9 SB7 LEVEL RAW MATERIAL 1 Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL 1 2 1 4 2 3 4 Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite GRAND TOTAL 5 11 3 9 16 1 13 8 1 1 11 Crystal Quartz Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL 1 5 5 2 3 4 2 1 1 1 1 1 2 6 2 1 2 3 1 9 3 Hydrated Metavolcanic Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL GRAND TOTAL 2 1 13 36 2 3 57 GRAND TOTAL 3 2 10 6 3 24 1 1 LEVEL 1 2 1 1 1 5 2 1 1 1 4 1 8 3 4 1 2 3 1 7 11 5 2 1 6 8-10 1 1 1 2 12 SB10 RAW MATERIAL 8 LEVEL SB9 RAW MATERIAL 7 1 SB8 RAW MATERIAL GRAND TOTAL 22 2 1 1 26 7 1 GRAND TOTAL 1 6 1 6 3 19 3 39 LEVEL 1 2 3 1 2 1 3 1 4 2 7 1 1 5 16 4 2 3 1 6 2 11 3 9 37 5 6 2 2 3 9 4 4 22 1 1 7 1 12 7 8-10 GRAND TOTAL 4 8 1 14 8 36 9 13 93 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 534 Chapter 10. 31RH491 Table 169. Vertical Distribution of Debitage by Sample Block, 31RH491 (Continued). VERTICAL DISTRIBUTION OF DEBITAGE, SAMPLE BLOCKS 11-15:31RH491 SB11 RAW MATERIAL Hydrated Metavolcanic Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) White Quartz GRAND TOTAL LEVEL 1 2 3 6 2 9 1 2 1 10 3 17 1 1 1 1 1 7 11 23 1 2 3 4 1 4 5 6 6 1 3 1 5 3 15 6 3 12 SB12 RAW MATERIAL Fossiliferous Chert Metasedimentary, Black Mill Mountain Rhyolite Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL Crystal Quartz Hydrated Metavolcanic Mill Mountain Rhyolite Type I Rhyolite Tuff Uwharries Eastern Rhyolite Uwharries Southern Rhyolite (1) Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite White Quartz GRAND TOTAL Mill Mountain Rhyolite Uwharries Southern Rhyolite (2) Uwharries Western Rhyolite GRAND TOTAL 6 1 1 5 6 7 1 4 2 3 8 2 4 6 7 8-10 1 1 2 3 6 7 1 4 3 8 1 1 2 5 1 4 7 20 16 1 2 3 4 GRAND TOTAL 1 1 12 5 6 35 23 83 1 1 3 6 1 2 14 GRAND TOTAL 1 1 3 9 14 34 3 8 73 LEVEL 1 1 3 1 1 2 1 1 7 5 8-9 2 2 3 3 4 12 1 7 3 9 3 24 6 4 1 1 12 2 1 2 6 5 6 7 4 11 1 3 4 16 4 4 5 6 2 2 GRAND TOTAL 3 1 3 19 13 19 7 65 LEVEL 1 2 1 1 3 4 1 1 50 1 53 1 3 72 3 2 82 1 1 4 65 2 2 75 1 1 3 9 31 2 4 51 SB15 RAW MATERIAL 4 1 1 SB14 RAW MATERIAL 8-9 LEVEL SB13 RAW MATERIAL 7 8-9 GRAND TOTAL 1 2 1 2 5 22 236 8 8 285 LEVEL 1 2 3 4 1 2 1 1 1 2 GRAND TOTAL 1 2 1 4 Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 535 Chapter 10. 31RH491 Table 170. Vertical Distribution of Ceramic Types in Sample Blocks, 31RH491. VERTICAL DISTRIBUTION OF CERAMIC TYPES IN SAMPLE BLOCKS: 31RH491 CERAMIC TYPE LEVELS 1 2 3 4 5 6 7 Vessel # 8 SB2 New River If Plain 1 V59 New River IIf Fabric Impressed 1 V47 SB3 New River IIa Fabric Impressed 1 V48 SB4 Hanover Ia indet. 1 V46 SB9 New River If Plain 1 V45 SB10 Yadkin IIId/e Plain 1 V38 1 V56 SB11 Hanover IIIe Cord Marked SB12 New River If Plain 1 V57 SB13 Hanover IIa Check Stamped 1 V58 SB14 Cape Fear III Cord Marked 1 Hanover Ia Check Stamped V41 1 Hanover IIa Check Stamped 1 Hanover IIe Fabric Impressed 1 Yadkin IIIe Cord Marked 1 V39 V43 4 V40 V42 SB15 New River If Fabric Impressed GRAND TOTAL 1 10 6 1 1 V44 1 Table 171. Shovel Test Outcome Model for the Purpose of Identifying Element Types. INFERRED ELEMENT TYPE NO. POSITIVE STPS NO. STPS WITH ≥ 5 DEBITAGE MEAN DEBITAGE/ POSITIVE TEST TYPE I HIGH HIGH HIGH TYPE I OR TYPE III HIGH MODERATE MODERATE TYPE III MODERATE LOW LOW Phase II Archaeological Investigations of Six Precontact Sites (C5890020435-D5095060228), Fort Bragg and Camp Mackall, NC 536 Chapter 10. 31RH491 shovel test. Given the structure of the sample blocks, there are generally nine outcomes (shovel tests) to measure. Some elements present measures that are intermediate between the two extremes and these can only be classified as Type I/III. Type II residential elements would be expected to reflect similar patterns to Type III residences, but they should contain multiple raw material types and a greater density of tools. Clearly, the identification of Type II residences in sample blocks is subjective and it would require additional confirmation from expanded sample blocks or test unit excavation to firmly identify such elements. Table 172 summarizes the results of applying the shovel test model to the elements defined by lithic raw material type within each sample block. As established by the Stage I and II population projections of elements, Type III elements predominate in the sample blocks, comprising 80.1 percent (n=55) of the identifiable elements. These elements are characterized by variably moderate to high shovel test representation (1 to 7 shovel tests), low frequencies of shovel tests containing greater than 5 pieces of debitage (0 to 1 shovel test) and debitage frequency means in positive shovel tests of less than or equal to 3.0. Stone tools were scarce. Type I elements were slightly more numerous than was predicted by the Stage I and II sample, but anticipated by an acknowledgement of the limitations of the method. Type I elements identified in the sample blocks represented ten of the 68 identified elements (14.7 percent). These elements possessed high shovel test representation (5 to 6 shovel tests), greater numbers of high debitage density shovel test outcomes (2 to 3 shovel tests) and high mean debitage densities, which ranged between 3.67 and 8.20 pieces of debitage. Approximately 4.4 percent (n=3) of the elements exhibited intermediate characteristics and could not be further differentiated. Cultural-chronological associations were inferred from depth profiles and from diagnostic stone tools that were matched to the raw material composition of the debitage concentrations. Due to relatively high reoccupation rates, localized deflation and historic disturbances the general vertical model was violated in a number of instances when diagnostic artifacts were compared to the vertical profiles of matching debitage distributions. In general, Archaic components of all ages occupy the zone between 20 and 50 cm bs, with only marginal vertical separation when viewed as a group. Forty-one Early Archaic, 17 Middle Archaic/Early Archaic, 4 Middle Archaic, 13 Late Archaic/Middle Archaic, 3 Late Archaic, 2 Woodland/Late Archaic/Middle Archaic, 4 Woodland/Late Archaic and 3 Woodland elements were identified in the sample block investigation. Together, Early and Middle Archaic elements comprise about 71.3 percent of the identified elements, which suggests that most of the lithic elements situated below 30 cm bs are Early and Middle Archaic in association. Candidates for Type II elements were recognized in each of the eight sample blocks (see Table 172). In these instances, groups of Type III elements could hypothetically belong to single Type II residences. Since such elements have only been positively associated with the Early Archaic period, only elements of this inferred affiliation were considered for this procedure. If groups of Type III Early Archaic elements belong to single occupations, the representation of Early Archaic occupations would be significantly diminished in the sample block inventory. Type II residences are characterized by low-density scatters of debitage composed of multiple raw material types and relatively high densities of stone tools of heterogeneous raw ma

RELATED PAPERS

RELATED TOPICS

Log In

Phase II Archaeological Testing and Evaluation of Six Precontact Sites in Hoke and Richmond Counties, Fort Bragg and Camp Mackall, NC.

Phase II Archaeological Testing and Evaluation of Six Precontact Sites in Hoke and Richmond Counties, Fort Bragg and Camp Mackall, NC.

Phase II Archaeological Testing and Evaluation of Six Precontact Sites in Hoke and Richmond Counties, Fort Bragg and Camp Mackall, NC.

Phase II Archaeological Testing and Evaluation of Six Precontact Sites in Hoke and Richmond Counties, Fort Bragg and Camp Mackall, NC.

Phase II Archaeological Testing and Evaluation of Six Precontact Sites in Hoke and Richmond Counties, Fort Bragg and Camp Mackall, NC.

Related Papers

RELATED PAPERS

RELATED TOPICS