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