UNISIST Model and Knowledge Domains

UNISIST Model and Knowledge Domains 

Birger Hjørland 

Trine Fjordback Søndergaard 

Jack Andersen 

Royal School of Library and Information Science, Copenhagen, Denmark 

INTRODUCTION 

Scientific and scholarly communication is an important 

area within information science (IS). The social 

system of communication consists of knowledge 

producers, intermediaries, and users. These groups of 

people (or actors) are different kinds of professionals. 

The social system also comprises institutes such as 

research institutes, publishers, and libraries. The actors 

and institutions perform information services such as 

writing, publishing, storing, and retrieving documents 

and information. The actors communicate in both 

formal and informal ways and produce different kinds 

of documents such as journal articles, books, book 

reviews, proceedings, bibliographies, catalogs, dictionaries, 

handbooks, encyclopedias, and review articles. 

It is of obvious interest for IS to develop adequate 

models of this system of scholarly communication. It 

is of course a dynamic system, and new information 

technologies are one of the causes underlying the 

dynamics. Many different attempts have been made 

to model this system. The best one is probably that 

proposed by UNISIST in 1971. The United Nations 

Information System in Science and Technology 

(UNISIST) was a program developed by the United 

Nations Educational, Scientific and Cultural Organization 

(UNESCO) to improve scientific and technological 

communication. The UNISIST program has been 

terminated, but the model has been updated and modified 

and is an important analytical tool in IS. Whereas 

the original model only considered scientific and technological 

communication, the modified version also 

considers social sciences and humanities and is 

regarded an important analytical model in domain 

analysis. 

OVERALL STRUCTURE AND ITS ELEMENTS a 

The original UNISIST model of scientific communication 

was proposed in 1971 in a report. [1, p.26] Fig. 1 is a 

a The description of this model is mainly based on Fjordback Søndergaard, 

Andersen, and Hjørland. [3] Readers are referred to this paper 

for more details. 

reproduction of this original model. This is but one of 

many communication models. In spite of its age, newer 

models have not been able to replace it. The model 

describes information communication between knowledge 

producer and knowledge user, as a socio-technical 

system consisting of diverse organizational and documentary 

units, each contributing to the division of 

labor in scholarly communication. A mapping of these 

agents, their information services, and document 

production should be an essential task of library and 

IS research. 

This model was originally proposed as a generalized 

model of the information structures within science and 

technology but may be generalized to other fields as 

well. 

The point of departure in the model is the knowledge 

producers. According to the model, there are 

three main categories of information distribution channels 

available for communicating research: informal 

and formal communication channels, and tabular 

channels. 

Informal communication may be in oral or written 

form and takes place when the producer and the user 

know each other and exchange information via channels 

such as personal correspondence (i.e., letters), 

manuscripts and preprints, private exchange of bibliographical 

references, etc. It might also occur semiformally 

as in professional conferences, meetings, or 

lectures. 

Two means of formal ways of communicating 

research are depicted in the UNISIST figure: published 

documents and unpublished documents. Published 

documents (books and journals) go from the knowledge 

producer through publishers or editors, from 

whom they reach the users, perhaps through abstracting 

and indexing services, and libraries and information 

centers. Bibliographic control of the documents 

contributes in ensuring the visibility and publicity of 

documents. A published document has an increased 

amount of visibility compared to unpublished ones. 

Unpublished documents consist, among other 

things, of theses, and research and technical reports 

distributed in limited copies by, e.g., government agencies. 

These documents may reach the users through 

clearinghouses and information centers. Hence, they 

Encyclopedia of Library and Information Science DOI: 10.1081/E-ELIS-120024989 

Copyright # 2005 by Taylor & Francis. All rights reserved. 1

2 UNISIST Model and Knowledge Domains 

Fig. 1 The flow of scientific and technical information. (From Ref. [1] ; p. 26.) 

are not subject to the same selection, production, and 

distribution procedures of publishers and editors that 

published documents go through. The notion of gray 

literature is somewhere in between published and 

unpublished documents. A private letter is unpublished, 

while a report registered by a clearinghouse is 

semipublished or gray literature. Ph.D. dissertations, 

published on demand at the University Microfilm 

International (UMI) and indexed in Dissertation 

Abstracts International, are another example of 

semipublications. 

The third category of communicating research is 

through tabular channels. This consists of numerical 

data. The original model chooses to regard this an 

independent channel, although it recognizes that 

much tabular information is distributed in normal

UNISIST Model and Knowledge Domains 3 

publications (and also in semipublications or in unpublished 

documents). b 

So far, the primary sources of scientific and technical 

information have been described in terms of their 

selection, production, and distribution functions. 

Primary literature is the point of departure in the production 

of scientific and scholarly knowledge and for 

the communicative division of labor of the literatures 

involved. The task of this literature is to produce 

and present original knowledge claims; this is why primary 

literature places emphasis on methodological 

descriptions. 

In the UNISIST model, we observe three levels of 

information source services between the knowledge 

producers and users: Besides the primary sources, there 

are also secondary and tertiary sources=services. c This 

is the basic taxonomy of information sources in the 

model (see the Appendix for the full model including 

the categories added in the revised version). The 

secondary information services register and describe 

primary documents for the purpose of retrieval and 

documentation. Secondary literature, such as subject 

bibliographies; citation indexes; library catalogs; and 

databases, analyses, describes, and registers primary 

literature d mainly in these bibliographical instruments. 

The central working processes of the secondary sources 

are analysis, storage, and dissemination. The model 

shows abstracting and indexing services, libraries, 

information centers, clearinghouses, and data centers, 

which are considered secondary information services, 

each with particular functions to perform. Tertiary 

literature is, for example, professional handbooks 

and encyclopedias and literature reviews, which consolidate, 

collect, and synthesize the primary literature. 

Abstracting and Indexing Services 

The UNISIST model distinguishes between two kinds 

of abstracting and indexing services. The first one 

consists of printed secondary journals prepared and 

distributed by scientific associations that operate on a 

b 

The UNISIST model privileged ‘‘tabular data’’ that presumably 

meant socio-economic numeric data series, tables of constants, observational 

data in numeric form, and the like; but it was an awkward 

phrase even then. That view (or at least wording) is obsolete in the 

sense that the key characteristic now is ‘‘digital’’ not ‘‘tabular’’ 

and includes vast databases of images and other resources not 

usually considered ‘‘tabular data.’’ 

c 

The terms primary (or secondary or tertiary) information sources 

may be used interchangeably with primary (or secondary or tertiary) 

information services: The actors or institutions that produce the 

sources represent the services. Also, the term literature(s) might be 

used interchangeably with information sources. 

d 

Although emphasis is on primary sources, secondary sources may 

also include tertiary sources. 

profit basis. An example of this kind is Chemical 

Abstracts or the Citation Indexes produced by the 

Institute for Scientific Information (ISI); that is the 

content of these does not represent collection of a 

physical location such as a library. The other service 

consists of catalogs and index files compiled by the 

staff of libraries or information centers as a mean of 

accessing their own collections. Also, guides and referral 

services belong to the second service produced by 

libraries and information centers. Referral services 

provide an indication of sources (persons, institutions, 

publications, etc.), from which scientific information 

may be obtained on a given subject and are mechanisms 

for switching users to such sources. 

Information Centers 

The concept of an information center is ambiguous, 

and the model’s relation to similar concepts such as 

libraries or documentation centers is not clarified. 

According to the model, the information center combines 

some of the functions of secondary journals 

and specialized libraries, to which are added specific 

duties such as the selective dissemination of information, 

or the preparation of state-of-art monographs, 

trend reports, reviews, etc. for the benefit of a specialized 

field or well-defined and more restricted user 

groups. "The role of such [information] centers is 

sometimes spoken of as that of ‘repackaging’ the information 

provided by abstracting and indexing services, 

according to the requirements of specific users; they 

operate then as ‘tertiary’ services, with a synthesis 

function added to those of indexing and classification: 

reorganization, quality control, compression, synthesis, 

evaluation etc.’’ [1, p. 30] Typical examples of tertiary 

documents are as previously mentioned literature 

reviews and syntheses. 

Because information centers perform both secondary 

and tertiary functions, they actually differ from 

libraries. It may be stated that whereas libraries are a 

kind of information center, the latter is not a kind of 

library. Libraries do not usually produce reviews, 

syntheses, or other kinds of tertiary documents. 

Moreover, information centers are normally not in 

the possession of a physical collection of documents 

and are not primarily concerned with giving access to 

these collections, as are libraries. When the UNISIST 

(1971) model was proposed, information centers were 

often conceived of as centers that provided bibliographic 

references or information from documents 

‘‘information’’ but not documents. See also Ref. [2] ). 

It has often been seen as an ideal means to present pure 

information, whereas documents have been viewed as 

obsolete medium of information. Centers that delivered 

bibliographical records (usually for a fee) from


electronic databases grew rapidly developed from 

about 1965 until 1990 and were usually part of 

libraries, particularly in the scientific and technological 

fields. Such departments were important, for example, 

those in the Danish Technological Library and the 

Main Medical Library in Copenhagen (and correspondingly, 

for example, centers in Stockholm as well 

as in other cities in the world). They were usually 

referred to as Documentation Departments. With the 

introduction of CD-ROM databases and ‘‘end-user’’ 

oriented search interfaces around 1990, these departments 

have by and large disappeared or have been 

downsized. 

The UNISIST model operates only for a basic 

concept of information center. In the following, we 

use information center as an umbrella term for 

libraries, documentation centers, and other similar 

activities concerning the collection, dissemination, 

storage, retrieval, and organization of documents 

(or knowledge). 

Clearinghouses 

Clearinghouses are defined by the UNISIST as 

‘‘ ...institutions entrusted with the procurement and 

dissemination of special categories of documents, such 

as technical reports, dissertation theses, thesauri, 

etc.’’ [1, p. 147] According to the UNISIST report, the 

modes of analysis, storage, and dissemination here 

are the same as those of libraries or information centers. 

However, what differentiates clearinghouses from 

libraries or information centers is the attention exclusively 

paid to unpublished documents. e 

Data Centers 

The UNISIST model conceives of data centers as being 

different from the abovementioned secondary services 

because ‘‘ ...they [data centers] deal exclusively with 

the raw material of science, parallel or even prior to 

publication ...,’’ [1, p. 30] and because functions (catalog 

ing, abstracting, indexing, synthesizing, etc.) of the 

other kinds of secondary services ‘‘ ...are normally 

defined with respect to written documents.’’ 

[1, p. 31] 

That is, the UNISIST model conceives of data centers 

as being concerned with ‘‘raw’’ data and nonwritten 

documents such as quantified surveys. 

Fjordback Søndergaard, Andersen, and Hjørland [3] 

in their revision decided, however, to consider data 

e When broadening the UNISIST model to other fields such as the 

humanities, there is an obvious need for comparing the functions 

of clearinghouses and archives. 

centers as part of the other units in the model as data 

that are published through formal communication 

channels such as publishers. This implies that we will 

exclude data centers as an independent form of information 

service and information source. This decision 

to not regard data centers as a specific communication 

channel is also motivated by the facts that other kinds 

of messages such as computer programs, pictures, and 

sounds are not represented by separate channels. 

Special Bibliographies, Translations, etc. 

As seen in the UNISIST model, organizations for 

information and documentation may produce special 

bibliographies, whether current or those of earlier 

times. Translation services may be commercial, or as 

in government centers they may be special translation 

journals (e.g., ‘‘cover to cover’’ translation of Russian 

scientific journals to English). They may also be bibliographies 

covering translations (such as Index translationum. 

Répertoire International des Traductions= 

International Bibliography of Translations—published 

by UNESCO). 

Review, Syntheses, etc. 

Here, reviews should not be confused with book 

reviews. Reviews are syntheses of the primary literature, 

e.g., in the form of handbooks, review articles, 

scientific and professional encyclopedias, and the like. 

(This is unlike works like general encyclopedias that 

are not primarily written for subject specialists nor 

are a part of the scientific literature described in the 

original UNISIST model. They may be seen as a 

special kind of literature, popular literature.) One 

important kind of tertiary literature is the series titled 

‘‘Annual Reviews of X,’’ e.g., the Annual Review of 

Information Science and Technology. Many of such 

annual reviews have high impact factors in their 

respective fields. 

Users 

The final unit in the model is the users. Users of scientific 

and technical information are in most cases also 

identical to the producers, or they may also be practitioners 

such as physicians. The different roles of users 

determine their information needs and make specific 

constraints on their use of information. Researchers 

are generally intensive users of informal and primary 

sources within a narrow specialty, whereas practitioners 

are generally users of a broader range of 

sources and may rely more on tertiary sources.


THEORETICAL REVISION AND 

TECHNOLOGICAL UPDATION 

OF THE MODEL 

There are two main reasons why Fjordback Søndergaard, 

Andersen, and Hjørland [3] found it necessary 

to update and revise the original model: 

1. The need to consider differences between different 

disciplines and domains and to generalize 

the model and its document typology (or the system 

of kinds of document) from science and 

technology to other domains. 

2. The growing impact of the Internet on scientific 

and scholarly communication. 

These considerations have been incorporated in the 

modified model shown in Figs. 2 and 3 and are introduced 

in the rest of this article. In Fig. 2, the Internet 

services have been displayed as a separate entity, 

whereas print and electronic media are integrated as 

seen in Fig. 3 

Concern 1: Considering Different 

Domains 

Fig. 2 has a punctured ellipse symbolizing a scientific 

discipline or a knowledge domain. The ellipse can symbolize, 

for example, the biological, the medical, or the 

legal discourse community. Such domains are typically 

overlapping, open structures, as is the case in the biomedical 

domain, where the borders of biology and 

medicine are difficult to identify. The ellipse is (more 

or less) punctured in order to illustrate this open nature 

of domains. The knowledge producers, the users, 

and the intermediaries are thus all considered members 

of a domain (or discourse community). Different services 

and institutions may, however, be shared between 

different domains. This is especially the case with interdisciplinary 

institutions and tools such as, for example, 

the Science Citation Index and national libraries. A 

given domain may, in other words, exercise more or 

less control with regard to its own informational infrastructure, 

and the institutions (e.g., publishers and 

libraries) may or may not have adequate subject 

knowledge and engagements to fulfill the functions in 

optimal ways. 

A new feature in the revised model is the arrows 

symbolizing the import and export of knowledge into 

the domain and out of the domain. The revised model 

recognizes that the information producers also use 

information sources. They may get information from 

sources in their own domain, other domains, direct 

observation of natural information sources, or from 

the general public. Some domains get most of their 

knowledge from their own domain. This is the case 

with mathematics, economics, and psychology. Other 

domains such as agriculture get most of their input 

from other domains, e.g., chemistry. Different epistemologies 

in a given domain will emphasize different 

knowledge sources. Social constructivism is, for example, 

an epistemological position that claims that 

researchers’ direct observations of nature are mediated 

by information sources of a social nature, which is a 

contrast to an empiricist or positivist epistemology. 

Export may be directed toward other disciplines or 

the general public. In engineering, physical products 

are, e.g., speakers or cars, the normal products 

exported from the knowledge production in the 

domain. In this domain, researchers seem to prefer 

making commercial products rather than writing 

research papers. Such products may however, be documented 

in patents. In science, publication of articles in 

journals is the norm, whereas journals compete with 

books in humanities. The articles may be printed in 

journals of the researcher’s own discipline, in those 

related to other disciplines, or in general scientific 

journals. The category of popular literature in the 

document typology is especially designed to serve 

dissemination of knowledge to the public domain or 

to beginners (See Appendix, ‘‘Popular literature’’). 

Domains vary much in their exports of knowledge to 

the general public. Humanities typically have a much 

closer relationship to the mass media and the general 

public compared to scientific publications. 

The domain itself reflects the division of labor in 

society (e.g., the division between people working with 

health problems in the health domain and those working 

with legal problems in the legal domain). Inside 

each domain, there is a more or less developed internal 

division of labor, e.g., between theoretical and empirical 

researchers, assistants, administrators, computer 

specialists, librarians=information specialists, translators, 

publishers, practitioners, etc. Often, the practitioners 

are the end users of the knowledge produced 

by the researchers. This can be, for example, the doctor 

curing a patient by applying new research results. The 

most important thing to realize is that each domain has 

a unique structure that should be described empirically 

and explained theoretically. A central point in the 

domain analytic approach to IS is the claim that tools, 

languages for special purposes (LSP), concepts, meaning, 

information structures, information needs, and 

relevance criteria are shaped in discourse communities. 

Different scientific, scholarly, or professional domains 

have unique structures of documents, which reflect an 

adaptation to the special needs in the domain. Some 

examples of unique kinds of documents are: 

Almanacs in astronomy 

Patents in engineering


Fig. 2 The revised UNISIST model modified for the domain analytic approach. # Emerald. (From Ref. [3] ; p. 303.)


Fig. 3 The revised UNISIST model integrating printed and Internet resources and modified according to the domain analytic 

approach. (From Ref. [4] ; p. 96.) (View this art in color at www.dekker.com.) 

Maps and atlases in geography 

Pedigrees and genealogical trees in genealogy 

Codes, bodies of law in law 

Sheets of music in music 

Tests in psychology 

Such specific elements are, however, not contained 

in general versions of the UNISIST model; one must 

but await the development of domain-specific models. 

Some scholarly fields have special information services 

providing prepared materials for research.


While all documents may serve research, some do this 

in a special way in the form of providing access to reliable 

source documents. This is especially the case in 

historical research. Archives and museums are important 

institutions missing in the original UNISIST 

model. They are very important in humanities and 

may also be important in some scientific fields. The 

study of archives and museums may thus be seen as 

parts of IS. Often, archives and other institutions 

reproduce and publish important historical documents 

in order to make such unique documents much more 

visible and available to scholars. This is a unique kind 

of work and should not be overlooked. In the Appendix, 

we have added ‘‘Source Literature’’ as a new category 

beyond the primary, secondary, tertiary forms 

known from the original UNISIST model. 

Among the other changes in the revised model that 

can be mentioned are integration of libraries and information 

centers into one box and addition of book 

reviews. The fundamental structure of the original 

model remains, i.e., the differentiation between primary, 

secondary, and tertiary sources and services. 

As mentioned earlier, we have added source literature 

as a new category. The figure (and the Appendix) 

shows the addition of thesauri and dictionaries also 

as a new category. Finally, the classification ‘‘Incidental 

Information’’ in the Appendix is a primary one for 

other sources and services that do not fit into other 

categories but nevertheless are important in scholarly 

communication. 

National and regional substructures 

The original UNISIST model was conceived of as one 

universal international structure. In reality, however, 

there exist more or less independent and elaborated 

national or regional information substructures in most 

domains. If we take psychology as an example, there 

exist regional journals of psychology, as well as 

internationally published journals from American 

and Germany. The German Information System for 

Psychology is the most elaborate information system 

outside the U.S.A. in this domain. In this national system, 

we find a complete system of primary, secondary, 

and tertiary information services. The primary information 

system in German psychology consists of journals 

covering all major subfields such as experimental 

psychology, social psychology, clinical psychology, 

developmental psychology, etc. proceedings of conferences, 

books in all fields of psychology, and scholarly 

treatises. It has under its fold about 150 publishers 

and producers of tests. The secondary information 

system contains a comprehensive bibliographical database, 

PSYNDEX, with abstracts and indexing of 

the German literature in both English and German. 

Also, many kinds of dictionaries are being developed. 

The tertiary information system contains comprehensive 

handbooks, encyclopedias, review journals, etc. 

The German system also contains specialized information 

systems developed to communicate psychological 

knowledge to students and to the general public (e.g., 

textbooks and the magazine Psychologie Heute). 

Today, there is increasing political pressure toward 

internationalization. Such a pressure may motivate 

German psychologists (among others) to publish in 

American (or international) journals, and it may leave 

the impression that other national or regional substructures 

are obsolete and inappropriate in scientific 

communication. This issue has been intensively 

debated and we shall not go deeper into this issue here. f 

As long as regional systems exist, they should be taken 

seriously by information scientists and should be 

reflected in our modeling and bibliometric studies of 

domains and the coverage of databases. The concept 

of regional subsystems is related to philosophical and 

sociological question about the nature of different 

traditions in science and scholarship. 

Paradigmatic differences 

In addition to disciplinary and regional=geographical 

differences, each domain will have—more or less 

noticeably—variations in its information system that 

are due to paradigmatic differences between the actors 

in the field. In psychology, for example, there is almost 

a complete information system for the psychoanalytic 

approach to psychology, consisting of primary journals, 

specific organizations, indexes and abstracts 

journals, encyclopedias and terminological works, etc. 

This system is not just relatively independent of the 

‘‘general’’ psychological system; it also has attributes 

that reflect the special nature of that field. Thus, psychoanalytic 

literature is more related to the humanities 

compared to the literature of behaviorism, cognitivism, 

and neuroscience. This is reflected in the tendency to 

organize the literature on authorship basis (Sigmund 

Freud’s works, C.G. Jung’s works, Melanie Klein’s 

works, etc). It is also reflected by the relative dominance 

of books compared to articles and by the rhetorical 

structure of the texts. 

Although such paradigmatic tendencies may be 

weak, it is our opinion that they always tend to exist 

to some extent. One could say that in a given (sub) 

f Michael Brittain [5] wrote ‘‘The claims of the social sciences to be of 

universal interest, value, and use are challenged. Citation data show 

that there is not a free flow of information across language and 

national boundaries,’’ and he considers the implication for information 

services. This is still an important issue.


culture or domain, there will almost always exist 

different ‘‘paradigms’’ for how practices should be 

changed and how the discipline should be defined 

and further developed. Any change in practices implies 

a need to change the documents, the symbolic systems, 

and the concepts that support the existing practice. The 

given concepts and documents will always serve certain 

policies and practices better than other concepts and 

documents. This is the case whether or not people 

are aware of this relationship. Domains are dynamic, 

and an important factor is the theoretical and epistemological 

development. When paradigms change, the 

whole model of the domain (as reflected in the UNI- 

SIST model) has to change too. 

Different concepts, documents, ways of cooperation, 

etc. are simply better suited for certain 

‘‘paradigms’’ than for other paradigms. This is why 

there is always a more or less latent tendency to 

develop separate information systems for separate 

views in any field. Only some disciplines (e.g., psychology 

and economics) have, however, specific journals 

specially devoted to different views. The epistemological 

issues may, however, turn out to be the most important 

dynamics underlying any information system. 

Concern 2: Internet-Based Communication 

As mentioned previously, Internet services have been 

displayed as a separate entity at the right in Fig. 2. 

The dynamics in scientific and scholarly communication 

caused by the technological changes from print 

to electronic media are of course very important. As 

stated by Smith, [6] the flow of computer mediated 

communication may cause a modification of the traditional 

categorization of documentary units. Some of 

the units may broaden or even modify their categorization 

in document type (primary, secondary, and 

tertiary literature). Also, concepts such as documents 

and publications may have to be redefined. It is, however, 

with some modifications still possible to classify 

information sources according to the UNISIST model. 

How the growing impact of Internet based communication 

channels has changed the flow of scientific communication 

since the creation of the UNISIST model 

has been subject to research. Among others, Kling 

et al., [7–8] Hurd, [9–10] Meadows [11] and Russel [12] have 

proposed or discussed models for communication in the 

digital age. However, none of the suggested new models 

can replace the UNISIST model. Most descriptions of 

the documentary and organizational units on the Internet 

emerge from a classification of hardware or software technicalities, 

rather than that by functional communication 

parameters as used in the UNISIST model. 

As early as 1978, Lancaster suggested that ‘‘the 

distinction between formal and informal communi- 

cation will be much less clear in an all-electronic 

environment.’’ [13, p. 113–114] Though this is true for 

many of the new hybrid forms, most people would 

agree that a distinction between formal and informal 

is possible in everyday Internet based communication. 

Informal communication 

Consistent with contemporary use of the informal 

communication channels, the most significant informal 

documentary units on the Internet are: 

1. Email 

2. List servers, which are discussion groups or 

interest groups that distribute messages via 

mailing lists. Electronic conferences or newsletters 

are usually listserv mediated. 

3. Usenet News, which is a collective term for 

thousands of newsgroups or discussion groups. 

Usenet News is managed centrally without the 

use of email in contrary to List servers. The 

messages or articles are most often cumulated 

and archived at least for a while. In most cases, 

this group includes bulletin boards, which are 

now rarely used on their own but rather as a 

feature among others in newsgroups. Thus, 

Lancaster [13, p.130] defined bulletin boards as a 

‘‘public space to permit messages to be entered 

and made accessible without restriction to all 

users of the system.’’ 

4. Electronic meeting or webcam conferencing. 

Each of these informal communication channels on 

the Internet may be located by the user through either 

(1) various search engines, including meta search tools 

such as Metacrawler that allows one to access several 

search engines from one place, or (2) diverse types of 

virtual libraries. However, informal communication 

channels often become known to the users in the 

course of serendipity, general browsing, or interpersonal 

contacts. 

Generally speaking, the Internet mediates a less 

selective number of informal communication channels 

than the preceding nonelectronic ones. Information once 

available only through the professional grapevine is now 

found on personal or institutional web pages, [14, p. 274] this 

is why the Internet is believed to have a positive effect on 

the development of invisible colleges in the otherwise 

stratified scientific community. Harnad [15] has argued 

that when (informal) manuscripts and feedback are 

exchanged through the network, scholarship can progress 

at a speed more similar to that of natural thought 

and speech. Because of the ease of using the informal 

Internet based communication channels, the path from 

the producer to the user and vice versa is more freely 

and quickly accessed and less troublesome.


Formal communication 

The increasing number of computer literate scholars 

through the 1990s, among other things, brought about 

the transition of the Internet from a predominantly 

informal communication channel to a significant 

formal communication channel. 

The most significant formal documentary units on 

the Internet are: 

1. E-journals and online journals: The true concept 

of the ‘‘electronic journal’’ or simply 

e-journal, as opposed to the ‘‘online journal,’’ 

requires that the material be produced and 

stored only in electronic form. [16] Online journals 

on the other hand are electronic spin-offs 

from paper journals. 

2. Preprints: Though some redaction has often 

occurred, preprints are documents distributed 

before the actual publication and perhaps 

before the peer-reviewing process is completed. 

Preprints are often considered a part of gray 

literature, but in recent years the emergence of 

preprint databases on the Internet has offered 

the means to gain access to this document type. 

However, not all disciplines have preprint databases; 

therefore depending on the knowledge 

domain is considered, preprints may or may 

not be considered gray. g 

3. Gray or unpublished literature such as theses, 

reports, etc. mostly found on scientific and 

research organizational servers. 

Real, operational e-journals are essentially a 

phenomenon of the late 1990s. The first peer-reviewed 

electronic, full-text e-journal including graphics was 

Online Journal of Clinical Trials, which began publication 

only in 1992. [17] Research completed in 1996 

showed that some e-journals actually have a high 

impact factor. However, as Harter points out, few articles 

are published. ‘‘Indeed, e-journals cannot have a 

major impact on the advancement of knowledge until 

they publish many more articles annually than they 

do [in 1996], while maintaining the apparent high 

quality of their articles.’’ 

[18, p.155] 

Weller [19] indicates that the peer-review process of 

e-journals is generally similar to the traditional process 

found in paper-based journals. However, new models 

of editorial peer review have been suggested, and some 

are being experimented with, which either alter or 

g The widely adopted practice of authors making copies of their 

papers available on their personal=institutional web sites should also 

be mentioned. In some domains, notably computer science, this has 

become the norm. It may be seen as a modern form of offprint. 

eliminate the traditional model of peer review. It is 

most important that any new model maintains the 

integrity of science and scholarly communication and 

yet implements the new emerging electronic environment 

and the need for decreasing turn-around time. 

As an example, Weller mentions that in the electronic 

environment, there is a need to re-examine the anonymity 

of reviewers. 

Regarding the development of preprint databases, 

the Internet as a new media has played a vital role. 

Due to the scholar community’s discontent with publishing 

delays and distribution problems with paper 

journals, Paul Ginsparg created Los Alamos ePrint 

archive in 1991. Arthur Smith who sees journals merely 

as an overlay on preprint databases describes the 

powerful position of ePrint archives. ‘‘The tension concerning 

responsibility for public distribution and communication 

of new work has been resolved in favor of 

the electronic preprint databases. Traditional journals 

still have some role in communication, providing archival 

material and interlinking, but they no longer form 

the primary communication medium at either the formal 

or the public levels.’’ [6] As can be seen, the Internet 

has changed the perception and use of this document 

type at least in some science domains where rapid 

dissemination is required. However, further research 

is needed to identify and explain differences between 

domains on this matter. 

If the development continues in line with Arthur 

Smith’s ideas for the future, it could be argued that 

journals (electronic or paper-based) should be positioned 

as a secondary source instead of its present location 

among the primary sources. Smith [6] argues that 

the main purpose of the typical journal will be storage, 

and as a sign of formal confirmation the preprints will 

adopt the traditional journals’ communicative role. 

Whether or not this will come about depends on various 

conditions. The traditional position of journals 

is therefore maintained in the domain general model 

of Internet based scientific and technical information. 

Gray literature 

The Fourth International Conference on Grey Literature 

defined gray literature as ‘‘That which is produced 

on all levels of government, academics, business and 

industry in print and electronic formats, but which is 

not controlled by commercial publishers.’’ [20] Gray 

literature on the Internet is, if possible to locate, very 

easily accessed compared with non-Internet based 

alternatives. The Internet has created an opportunity 

to make gray literature publicly available without the 

expenses of traditional publication. 

Several organizations, associations, and information 

systems such as the European Association for 

Grey Literature in Europe (EAGLE), Information


for Grey Literature in Europe (SIGLE), and the 

British Library Document Supply Centre (DSC) are 

making special efforts to raise awareness of and provide 

access to gray literature such as reports, theses, 

translations, noncommercial conferences, and official 

(government) material. Several bibliographies devoted 

to gray literature can be found on the Internet. The 

problems concerning gray literature are therefore being 

addressed on international level but may still demand 

the searcher to be rather persistent. 

Formal communication channels 

The formal documentary units (in particular journal 

articles) may be directly accessed if their address is 

known. Otherwise, the units on the Internet may reach 

the user through, for example: 

1. Preprint databases. h 

2. Bibliographic or full text databases. i 

3. Scientific and research organizations servers. 

4. Publisher web sites. 

5. Virtual libraries. 

6. Search engines and meta search tools. 

As with the role played by preprints, Smith [6] also 

anticipates changes in the function of preprint databases. 

Preprint databases are expected to become 

responsible for public distribution and communication 

of new works. This means that the user needs to have 

great domain specific knowledge or expertise since the 

content of these databases is not necessarily finalized 

by the peer-review process. For this reason, Smith 

presumes review papers to take on a more prominent 

role in providing guidance to the literature for those 

not familiar enough with the domain to deal with the 

nonreviewed preprint literature. 

The majority of bibliographic or full-text databases 

are available via the Internet (given a password), but 

many databases have yet to make value addition to 

their services further to this media. The terms electronic 

libraries (e-libraries), digital libraries, and virtual 

libraries are often used rather inconsistently and 

h In the literature, these databases are not referred to as clearinghouses 

as found in the original UNISIST model. On the Internet, 

the term Clearinghouse seems to denote some kind of annotated 

directory or resource guide (See for example the Argus Clearinghouse 

at http:==www.clearinghouse.net=mission.html for further 

information). Preprint databases primarily flourish within the science 

domains such as Los Alamos ePrint, LANL preprint database, and 

SPIRES. 

i Representing both commercial (First Search, DIALOG, STN, Lexis- 

Nexis) and noncommercial databases available on the Internet 

(OPACs also termed electronic libraries or e-libraries). 

several different definitions are seen in the literature 

and on the Internet. 

The preservation of paper based scientific communication 

is a part of the secondary organizational units in 

the original UNISIST model (e.g., libraries’ copyright 

deposits). Several archival initiatives are seen on the 

Internet. The Internet archive (http:==www.archive. 

org) is as an example of the preservation of former 

versions of web sites. Some countries (e.g., Denmark 

since 1998) have copyright deposits for some web 

documents, but coverage is selective and probably 

unsatisfactory. Also more subject specific archives 

exist. 

Publishers’ web sites are increasingly providing 

access to publications in addition to more traditional 

information such as subscription prizes, contributor 

instructions, and review policies. The specific searching 

and browsing facilities on these sites vary. The users’ 

opportunities to view, print, or request documents 

differ as do their requirement for payment. The publishers’ 

traditional role connected to selection, production, 

and distribution of the primary sources is 

increasingly supplemented with new roles concerned 

with storage and dissemination. Some publishers are 

beginning to utilize cross-referencing or reference linking 

as a browsing option. CrossRef is a collaborative 

reference linking service that functions as a sort of digital 

switchboard. It holds no full text content, but rather 

effects linkages through Digital Object Identifiers 

(DOI) that are tagged to article metadata supplied by 

the participating publishers. The result is a linking 

system through which a researcher can click on a 

reference in a journal and access the cited article. 

These facilities, however, are only available to the 

users who subscribe to the various publications. A 

researcher clicking on a CrossRef link will be automatically 

connected to a page on the publisher’s web site 

showing a full bibliographical citation of the article, 

and, in most cases, the abstract as well. Subscribers 

are generally authenticated for full text access, and 

nonsubscribed users presented with other options for 

access (such as subscription, document delivery, or 

pay-per-view). Researchers in library environments 

may find that CrossRef links redirect to local holdings. 

This development suggests that the publishers are 

taking up the traditional secondary sources such as 

indexing and abstracting services, libraries, and information 

centers. If cross-referencing between the diverse 

publishers becomes standard, a new and potentially 

powerful information-searching tool may rise. 

Aids such as domain specific dictionaries, glossaries, 

taxonomies, and thesauri of various quality and coverage 

can be found on the Internet, mostly for free or as 

value-added service connected to fee based databases, 

virtual libraries, or clearinghouses. These secondary 

sources are both effectively and efficiently utilized on


the Internet compared to traditional paper versions. 

However, the various sources must be individually 

assessed with regard to quality. 

The documentary units of the Internet can also be 

reached by search engines or diverse meta search tools. 

Though very often helpful, these retrieval algorithms 

or search engines typically suffer from a lack of semantic 

control (e.g., with synonyms, homonyms, broader 

and narrower terms, etc. as known in traditional 

metadata systems). Although they almost seem like 

miracles, there are still problems that they cannot 

tackle and for which further research and competing 

alternatives are necessary. 

CONCLUSIONS 

The UNISIST model is a fruitful model of scientific 

communication that help conceptualize IS in a perspective 

that is of great heuristic value and also fruitful 

for further empirical investigations. The model is also 

useful for practical information work, e.g., the 

construction of information guides. 

Today we do not even have a description or model 

of the communicative system according to the 

UNISIST model for even one single discipline based 

on empirical studies. There is a big need to study different 

domains according to the model (thus also contribute 

to the further development of the model). 

We also need to consider some basic issues in the 

model further. For example, to what degree do the 

primary, secondary, tertiary, source producing and 

intermediating level (text books and mass media) function 

as relatively independent systems? Do they have 

relatively independent groups of professionals? Do 

they have specific guidelines and norms? Do they have 

specific channels for publication (output)? Do they 

have specific educational programs and information 

input channels? What internal and external factors 

determine the structure of scientific communication 

systems? 

Each point in the model as well as a large number of 

relations are also in need of more research. It is our 

hope that this model may stimulate further interest in 

scholarly communication and in documents and thus 

provide library and IS a much-needed theoretical 

inspiration. 

APPENDIX: TYPOLOGY OF DOCUMENTS 

I. Primary Literature 

(Primary literature is the researchers’ and knowledge 

producers’ primary medium for claiming original 

findings, theoretical analysis, empirical data, etc.) 

Monographs 1 (as long as they communicate original 

findings) 

Journal articles and articles in edited books (as long 

as they communicate original findings) 

Critical analyses 

Conference presentations 

Gray literature including dissertations, treatises, 

master theses, 

Reports, various kinds of official publications and 

governmental publications 

Patents 

Standards 

Ia. Source Literature 

(Source literature is either literature produced in order 

to supply researchers with information (e.g., translation 

journals) or information produced to other purposes 

than research, but used as information by 

researchers (e.g., music and fiction). Primary literature 

(and anything else) serves of course as information 

sources, while source literature is negatively defined 

as not being primary, secondary, tertiary, accidental, 

or popular literature) 

Facsimiles 

Transcriptions 

Source editions, scientific editions, and standard 

editions. [Model: The works of Søren 

Kierkegaard ...]. 

Laws, court findings 

Music 

Data archives 

Statistical documents, tabular documents 1 (reporting 

original data) 

Translations (only qua translations; the translated 

work is, for example, primary literature) 

Product information=‘‘trade literature’’ 

(Not applicable: Sourcebooks) 

II. Secondary Literature =Bibliographical 

Literature 

(Literature that registers, describes, and organizes the 

primary literature as well as the other categories 

(including the secondary literature itself). Secondary 

information systems are the core focus of the library, 

documentation, and IS profession. Bibliography is a 

discipline that studies this area). 

Subject bibliographies and bibliographical 

databases 

Abstract journals 

Indexes 

Citation indexes


Current Contents 

Bibliographical guides, metabibliographies 

Bio-bibliographies=author-encyclopedias (including 

auto-bibliographies on personal web pages) 

Source inventories 

Catalogs 

IIa. Dictionaries and Thesauri 

(Dictionaries are the focus of the linguistic subdiscipline 

lexicography. Thesauri are a kind of dictionaries 

that has mostly been studied and developed in relation 

to bibliographical databases) 

Historical=etymological dictionaries 

Translation dictionaries 

Conceptual dictionaries and thesauri 

III. Tertiary Literature/Review 

Literature/‘‘Outlines’’ 

(Literature summarizing and synthesizing knowledge 

in the primary literature) 

Handbooks 

(Textbooks) 

Monographs 2 (synthesizing existing literature 

without providing new, independent knowledge) 

Review articles (do) 

Scientific encyclopedias (general encyclopedias are 

normally considered as popular literature) 

Short, indicative reviews 

Chronological surveys 

Data handbooks, tabular documents 2 (synthesizing 

original statistical sources) 

IV. Incidental Information 

(Information about tools (including computers and 

software), developments in the job market, in the 

discipline=domain, etc. as long as such information 

cannot be seen as part of the domains’ regular knowledge 

production.) 

Biographical documents 

Directories 

Conference calendars 

Lists of archives 

Directory to grants, scholarships, etc. 

Yearbooks (annual reports) 

Newsletters 

Personal homepages 

V. Popular Literature 

(Export of knowledge produced in a domain to the 

general public, other domains, or students.) 

Textbooks 

Magazines 

Newspapers (e.g., science journalism) 

Popular books (including general encyclopedias) 

Faction, science fiction 

Mass media, multimedia presentations, etc. 

REFERENCES 

1. UNISIST. Study Report on the Feasibility of a 

World Science Information System, By the United 

Nations Educational, Scientific and Cultural 

Organization and the International Council of 

Scientific Unions, Paris, UNESCO, 1971. 

2. Hjørland, B. Documents, memory institutions, 

and information science. J. Doc. 2000, 56 (1), 

27–41. 

3. Fjordback Søndergaard, T.; Andersen, J.; 

Hjørland, B. Documents and the communication 

of scientific and scholarly information. Revising 

and updating the UNISIST model. J. Doc. 2003, 

59 (3), 278–320. Available at: http:==www.db.dk= 

bh=UNISIST.pdf. 

4. Hjørland, B. Fundamentals of knowledge organization. 

Knowl. Org. 2003, 30 (2), 87–111. 

5. Brittain, J.M. Internationality of the social 

sciences: implications for information transfer. J. 

Am. Soc. Inf. Sci. 1984, 35 (1), 11–18. 

6. Smith, A. The journal as an overlay on preprint 

databases. Talk prepared for ALPSP, 9 April 

1999. Available at:http:==ridge.aps.org= 

APSMITH=ALPSP=talk.html (accessed July 24, 

2004). 

7. Kling, R.; Callahan, E. Electronic journals, the 

internet, and scholarly communication. Annu. 

Rev. Inf. Sci. Technol. 2003, 37, 127–177. 

8. Kling, R.; McKim, G.; King, A. A bit more to it: 

scholarly communication forums as socio-technical 

interaction networks. J. Am. Soc. Inf. Sci. 

Technol. 2003, 54 (1), 47–67. 

9. Hurd, J.M. The transformation of scientific communication: 

a model for 2020. J. Am. Soc. Inf. 

Sci. 2000, 51 (14), 1279–1283. 

10. Hurd, J.M. Models of scientific communication 

system. In From Print to Electronic: The Transformation 

of Scientific Communication; 

Crawford, S.Y., Hurd, J.M., Weller, A.C.; 

Medford: ASIS, 1996. 

11. Meadows, A.J. Communicating Research; 

Academic Press, 1998.


12. Russel, J.M. Scientific Communication at the 

Beginning of the Twenty-First Century, ISSJ 

168, UNESCO, 2001; 271–282. 

13. Lancaster, F.W. Indexing and Abstracting in 

Theory and Practice, 2nd Ed.; Library Association: 

London, 1998. 

14. Russel, J.M. Scientific Communication at the 

Beginning of the Twenty-first Century. UNESCO, 

Paris, 2001; 271–282. 

15. Harnad, S. Post-Gutenberg galaxy: the fourth 

revolution in the means of production of knowledge. 

Public Access Comput. Syst. Rev. 1991, 2 (1), 

25–38. 

16. Piternick, A.B. Attempts to find alternatives to 

the scientific journal: a brief review. J. Acad. 

Librariansh. 1989, 15 (5), 260–266. 

17. Keyhani, A. The online journal of current clinical 

trials: an innovation in electronic journal publishing. 

Database 1993, 16 (1), 14–23. 

18. Harter, S.P. Scholarly communication and 

electronic journals: an impact study. J. Am. Soc. 

Inf. Sci. 1998, 49 (6), 507–516. 

19. Weller, A.C. Editorial peer review for electronic 

journals: current issues and emerging models. J. 

Am. Soc. Inf. Sci. 2000, 51 (14), 1328–1333. 

20. Grey Literature Network Service. GL 1999 

Conference Program. Fourth International 

Conference on Grey Literature: New Frontiers 

in Grey Literature, GreyNet, Grey Literature 

Network Service, Washington, DC, 1999, Oct 

4–5. Available at: http:==www.nyam.org=library= 

greywhat.shtml (accessed July 24, 2004).

UNISIST Model and Knowledge Domains

Create successful ePaper yourself

Delete template?

Save as template?