B U L L E T I N
Annual Meeting Coverage
Domain Analysis: A Socio-Cognitive Orientation for Information Science Research
Birger Hjørland is professor in the department for information studies at the Royal School of Library and Information Science, Denmark; e-mail: firstname.lastname@example.org
Editor's note: This article is based on the author's talk given on October 20, 2003, at the ASIS&T 2003 Annual Meeting for the second Plenary Session: Humanizing Information Technology: New Theoretical Approaches in Play.
This article outlines my view on humanizing information technology as well as on the fundamentals of information science (IS) and on the deep connection between these two issues.
What is the meaning of "humanizing information technology" – our conference theme? Is it just making information technology (IT) "smart," "efficient," "reliable" and "user friendly"? Or are wider social consequences of computing involved, too? Is it just a question of getting information quickly, or is the quality of the information that the users gain also involved? I will argue in support of a wider conception of "humanizing information technology," especially as a concern that distinguishes IS from computer science.
While computer science is the science about IT, we in IS try to optimize people's access to information, knowledge and documents. This optimization is extremely dependent on IT, which explains the close relationship between the two fields. But an important goal for IS involves the quality of information and the social perspective related to information intermediating. That goal is to enable users to make informed choices about how they are informed.
Almost everyone agrees that IS badly needs good theories. There is a need to examine and discuss the basic approaches to IS. A field that cannot rationally confront its problems is in crisis. Therefore I find this session a fruitful initiative.
Domain analysis offers a theoretical perspective, which in my opinion is able to satisfy the need for a comprehensive theory of IS. Domain analysis is an approach that connects theory and practice, has a coherent view of all major concepts in IS and provides an identity for IS consistent with the history of the field. Domain analysis is able to unite different sub-disciplines such as bibliometrics, knowledge organization, information retrieval and information literacy. I hereby invite everyone to examine this claim and to participate in the discussion of how our field should develop.
Domain analysis states that the most fruitful horizon for information science is to study the knowledge domains or thought or discourse communities that are parts of society's division of labor. Marcia Bates' presentation at the 1987 ASIS Annual Meeting (Information: The Last Variable) is an early and clear formulation of this focus, which is central to the perspective of domain analysis.
In a previous article I suggested the following specific ways to study domains, which together define the specific competencies of information scientists:
While these 11 approaches may be used separately (and some of them are often used and taught in a "general" way), the application of more than one to the same domain may provide a deeper understanding of underlying dynamics. The approaches need to be tested on specific fields. It is a real mistake to believe that one is better off if one knows nothing about information systems in any particular domain.
If we take a given domain, say art, a domain study can map the different actors, institutions and communication processes in that domain, including the artists, the art reviewers, the museums, the art historians, the scholarly literature, the libraries and the databases as in the tenth approach above. Such a mapping depends on a view of what (good) art is. In a given society there are majority and minority views about (good) art, the reason an epistemological or critical study of the art domain is necessary. Different views of art may also be revealed through anthropological studies of art.
Ørom has pointed out in a recent article that art has been influenced by different paradigms, such as the "iconographic paradigm" and "the stylistic paradigm." These paradigms have influenced the way art literature is written and the way both art exhibitions and library classifications are designed. Based on such an analysis, Ørom has been able to discern a dominant view in, for example, the Library of Congress and Dewey decimal classifications. It should not be difficult to expand this analysis to all other information science areas including bilbiometreics, information retrieval and relevance assessments. One may, for example, study bibliometric patterns in scholarly art literature and the relative influence of different paradigms. If one were going to construct a guide to information sources about art, then both this epistemological study and the study of actors, institutions and other players would be a precondition if one wants to base such a guide on well-supported criteria. My basic argument is that any work of art, any text on art, must be considered in information system design and management, and domain analysis is a research program in IS that provides general methodological principles for such tasks.
Domain analysis does not, however, imply that information science should dissolve into separate studies of various domains. There are general methods and principles by which domains should be explored in IS (see, for instance, the UNISIST model presented in Fjordback Søndergaard, Andersen and Hjørland) as displayed at a poster session at this conference. IS should be organized both in relation to different domains and to different processes, levels, approaches, kinds of systems and other appropriate criteria. The comparative perspective – the examination of how knowledge domains differ on some points and are similar on other points – is important in order to construct a general information science that is not just an empty abstraction.
The Socio-Cognitive View
Even if domain analysis has the domain as its primary focus (and thus not the individual), it nevertheless also has a view on individual cognitive processes. This view is termed the socio-cognitive view and is related to both American pragmatism and to Russian historical-cultural psychology. Important names in these traditions are John Dewey and L. S. Vygotsky, respectively.
Yes, the socio-cognitive view turns the traditional cognitive program upside down. It emphasizes the internalization of culturally produced signs and symbols and the way cognitive processes are mediated by culturally, historically and socially constructed meanings. Less priority is given to hardware whether in brains or computers.
Domain analysis consequently does not conceive users in general but sees them as belonging to different cultures, social structures and domains of knowledge. Information producers, intermediaries and users are more or less connected in communities that share common languages, genres and other typified communication practices. There are different semantic distances between the agents.
Emergence of Information Processing
Domain analysis has a coherent view of information and information systems as Capurro and I demonstrate in our ARIST 2003 article on "The Concept of Information." Bateson's famous definition of information in Steps to an Ecology of the Mind is that information is "a difference which makes a difference" (p. 453).
Such differences appear as physical signals. Organisms or systems may be designed to react to specific physical signals. Information is only information in relation to a specific mechanism sensitive to a specific signal. Such mechanisms are information processing mechanisms.
Information processing systems probably first emerged in living organisms. As discussed in my recent paper "Principia Informatica," their biological evolution may be described in qualitative stages followed by a cultural evolution, which may also be described in qualitative stages. In human societies information processing mechanisms are developed socially in, for example, different professions. A stone in the field contains some information for the geologist, other information for the archeologist. Different professionals describe informative objects in different ways, and they organize their descriptions according to domain specific criteria.
Information is thus an emergent phenomenon, which is always relative to certain mechanisms, subjects or collective criteria in a community. Information cannot be identified or measured unless a certain mechanism is specified. In human societies such mechanisms are more or less domain specific.
Domain analysis and socio-cognitivism thus share with Sandstrom's socioecological approach (see article in this issue) a broad biological and anthropological view. We agree on the fruitful analogy of comparing information resources with resources like food as well as the hypothetical-deductive way of putting research questions, for example in relation to specialization versus generalization in the utilization of different kinds of resources. Domain analysis may, however, differ somewhat in its view on methodological individualism (for further discussion, see Bhargava). Domain analysis emphasizes the exploration of specific social environments, such as science and the humanities and their documents, genres and symbolic systems as detailed in the 11 ways to study domains listed above. Domain analysis puts greater emphasis on qualitative issues related to cultural-historical evolution as well as to domain specific documentation systems.
Is Domain Analysis Only a "Metatheoretical" View?
Does domain analysis lack empirical foundation? No, this question reflects (in my opinion) a wrong way to look at the relations between (meta)theory and empirical investigation. Any view in the behavioral, cognitive and social sciences is related to different philosophical assumptions that often have deep roots in the history of thinking.
Behaviorism was once an extremely productive empirical approach, but was nevertheless largely ousted by cognitivism, which was based on other premises. Traditions such as "the physical paradigm" or "the cognitive view" have not been better proven empirically than, for example, domain analysis or "social constructivism." Any such view tries (more or less successfully) to integrate all existing empirical and theoretical knowledge as their basis. They have, however, different criteria of what counts as relevant findings. Overall evaluations of the relative merits of such metatheories can only be done by examining and comparing their basic arguments, not by measuring their productivity, their impact or similar factors.
Much empirical support exists for the socio-cognitive view in psychology. In IS the domain-analytic view is empirically supported by, for example, my recent JASIS&T article, "Epistemology and the Socio-Cognitive Perspective in Information Science."
In addition, domain analysis is in accordance with some new interdisciplinary discussions concerning (scientific) realism versus antirealism. Questions about realism versus antirealism are really difficult. They are, however, extremely important to address for IS (see my forthcoming article in Library Trends – manuscript at www.db.dk/bh/Arguments%20for%20Philosophical. . .doc). I term my own position pragmatic realism, and I interpret Thomas Kuhn, among others, as being a pragmatic realist.
While Thomas Kuhn emphasizes how our ontologies are implied by our theories and paradigms, he nevertheless emphasizes that we cannot freely invent arbitrary structures: "nature cannot be forced into an arbitrary set of conceptual boxes." On the contrary " . . . the history of the developed sciences shows that nature will not indefinitely be confined in any set which scientists have constructed so far" (Kuhn, p. 263). The world provides resistance to our conceptualizations in the form of anomalies, that is, situations in which it becomes clear that something is wrong with the structures given to the world by our concepts. In this way Kuhn's view may be interpreted as (pragmatic) realist, although he is often being interpreted as anti-realist (for example by Niiniluoto).
Pragmatic realism is not opposed to the view that (scientific) knowledge is constructed as claimed by social constructivists. Clearly disciplines, theories, instruments, terminology, documents, information systems and other aspects of science are constructed entities. Domain analysis shares with social constructivism an interest in historical perspectives on the social conditions associated with knowledge production. It is opposed, however, to tendencies towards ontological antirealism in (some forms of) social constructivism.
This realistic view of IS implies that documents have given informative potentialities whether these are recognized by users or by the discourse community. Relevance is not just what users believe is relevant. Users may change their relevance criteria when they encounter new information. The implication is that criteria of indexing and retrieval of information cannot rely just on "user studies" but are primarily related to methodological and epistemological norms. IS should thus be open to different views and help users to identify them in information systems.
Are There Drawbacks to Domain Analysis?
Any scientist should of course consider the strong and weak side of every approach. Are there any weaknesses in domain analysis? Yes, I think there are drawbacks to domain analysis as an approach to IS compared to other approaches. Universal solutions and smart technical solutions have advantages compared to solutions that emphasize differences in cultures and domains. Such universal solutions have been developed by IT (mainly by computer science). Another drawback might be that domain analysis will draw IS more in the direction of the humanities and information studies compared to computer science and information science.
There may be less prestige and probably less money connected to more "soft" research. Domain analysis may, however, be based on a truer understanding of the nature of information services, the reason it may be superior in the long run. It can be harmful for disciplines if they do not cumulate real knowledge because short-term interests drive them. Domain analysis may also provide IS with a stronger identity compared to computer science. Perhaps it will also be possible to combine the "harder" part of IS with the domain analysis-approach.
From the point of view of domain analysis, where would I like IS to be in – say – 10 years? I share many of the same wishes as the rest of you. I want IS to prosper. I want information scientists to play important roles in the development and management of digital libraries as well as in all kinds of memory institutions and knowledge organization, in the further development of citation analysis and many other activities.
In order for this to happen, I think it is necessary to strengthen IS in many ways. We should not fool ourselves by disregarding the nature of knowledge, information and documents and make unfruitful abstractions on a problematic basis. I hope we can contribute to fine-tuning search algorithms and knowledge organizing systems to specific kinds of media, genres, domains and (sub)cultures. We should encourage the study of the rich flora of such entities as documents, domains, genres and communities from a specific information point of view. It would be a good thing if information scientists provided highly respected courses in information searching and knowledge organization in different occupations. We should contribute with quality textbooks and research journals in scientific, social scientific, humanistic and other major domains.
Information science should be better integrated with other fields studying domains including computational linguistics and LSP, compositional and writing studies, sociology and history of science, philosophy of science and conceptual history.
I hope that different approaches to information science will have fruitful dialogues and that the basic assumptions in different views will be better explicated than they are today. I also hope that different perspectives will enrich the field, that the connections between research and practice will be strengthened and that a growing feeling of the relevance of IS will arise in the library, documentary and information communities.
Conclusion: Humanizing Information Technology
Domain analysis' special contribution in humanizing IT is strongly connected to its foundation in a social understanding. Domain analysis shares this social emphasis with social constructivism. Domain analysis might contribute to making IT and information systems better adapted to different user groups and interests. Domain analysis may also contribute to making information systems more transparent by combining advanced and multidimensional semantic information with visualization technologies. The aim is not just to map semantic relations such as synonymy and homonymy in a general way, but also to display how such relations are related to different theories and citation patterns in the literature, thus allowing users better control in using IR.
For Further Reading
Bates, Marcia J. (1987): Information: the last variable. In Proceedings of the 50th Annual Meeting of the American Society for Information Science, 24 (pp.6-10). Medford, N.J.: Published by Knowledge Industry Publications for American Society for Information Science.
Bateson, G. (1972). Steps to an ecology of mind. New York: Ballantine Books.
Bhargava, R. (1998). Holism and individualism in history and social science. In Routledge Encyclopedia of Philosophy, Version 1.0, London: Routledge.
Capurro, R., & Hjørland, B. (2003). The concept of information. Annual Review of Information Science & Technology, 37, 343-411.
Fjordback Søndergaard, T., Andersen, J., & Hjørland, B. (2003). Documents and the communication of scientific and scholarly information: Revising and updating the UNISIST model. Journal of Documentation, 59, 278-320. www.db.dk/bh/UNISIST.pdf
Gärdenfors, P. (1999). Cognitive science: From computers to anthills as models of human thought. Human IT, 3(2), 9-36. www.hb.se/bhs/ith/2-99/pg.htm
Hjørland, B. (2002a), Epistemology and the socio-cognitive perspective in information science. Journal of the American Society for Information Science and Technology, 53(4), 257-270.
Hjørland, B. (2002b). Domain analysis in information science: Eleven approaches - traditional as well as innovative. Journal of Documentation, 58, 422-462. www.db.dk/bh/publikationer/Filer/JDOC_2002_Eleven_approaches.pdf
Hjørland, B. (2002c). Principia informatica. Foundational theory of information and principles of information services. In H. Bruce, R. Fidel, P. Ingwersen, & P. Vakkari (Eds.) Emerging frameworks and methods: Proceedings of the Fourth International Conference on Conceptions of Library and Information Science (CoLIS4), 109-121. Greenwood Village, CO: Libraries Unlimited.
Hjørland, B. (IN PRESS, 2003). Social and cultural awareness and responsibility in library, information and documentation studies. In B. Rayward, J. Hansson & V. Suominen (Eds.) Aware and responsible (pp 71-91). Lanham, MD: Scarecrow Press. (Papers presented at the Nordic and International Colloquium on Social and Cultural Awareness and Responsibility in Library, Information and Documentation Studies (SCARLID), 13-14 December 2001, Oulu, Finland).
Hjørland, B. (2004). Arguments for philosophical realism in library and information science. Paper accepted for Library Trends, issue on "Philosophy of Information," edited by Ken Herold, 52(3), Winter 2004. Manuscript available on www.db.dk/bh/Arguments%20for%20Philosophical. . . .doc.
Hjørland, B., & Hartel, J. (Eds.). (2003). Special issue of Knowledge Organization devoted to domain analysis. (Forthcoming, issue vol. 30, issue3/4, 2003).
Hjørland, B., & Sejer Christensen, F. (2002). Work tasks and socio-cognitive relevance: A specific example. Journal of the American Society for Information Science and Technology, 53, 960-965.
Ingwersen, P. (1992). Information retrieval interaction. London: Taylor Graham.
Kiel, F. C. (1989). Concepts, kinds, and cognitive development. Cambridge, MA: The MIT Press.
Kuhn, T. S. (1970). Reflections on my critics. In I. Lakatos & A. Musgrave (Eds.) Criticism and the growth of knowledge (pp.231-278). Cambridge: Cambridge University Press.
Niiniluoto, I. (1991). Realism, scientific. In Handbook of metaphysics and ontology (Vol.2, pp. 761-763). Munich: Philosophia.
Ørom, A. (Forthcoming). Knowledge organization in the domain of art studies – History, transition and conceptual changes. Knowledge Organization, 30 (3/4).
Sandstrom, P. E. (1994). An optimal foraging approach to information-seeking and use. Library Quarterly, 414-449. See also Scholars as subsistence foragers. Bulletin of the American Society for Information Science, 25(3) www.asis.org/Bulletin/Feb-99/sandstrom.html
Sandstrom, P. E. (2001). Scholarly communication as a socioecological system. Scientometrics, 51, 573-605
Wertsch, J. V. (1985). Vygotsky and the social formation of mind. Cambridge, MA: Harvard University Press.
Addendum: Is domain analysis only concerned with academic users and academic subjects?
After the presentation, a member of the audience asked whether domain analysis is only concerned with scientists and scholars or whether it is also concerned with the information needs of ordinary people?
In my reply I argued that domain analysis is an approach to information science that covers all kinds of information users. I referred to a forthcoming special issue of Knowledge Organization (Hjørland & Hartel, Eds.) devoted to domain analysis. Fields such as art and music are discussed. Classification systems, terminology systems and all other parts of information systems are influenced by their views of, for example, art and music. Ordinary users are influenced by different views, too, and the success of information systems such as public libraries depends on the institution's ability to cope with different "paradigms" in, for example, art and music. One can study, for example, hobbies as domains, and children's cognitive processes in different domains are an important area of study in modern child psychology (see, for example, Kiel). Thus any claim that domain analysis is only concerned with academic subjects is not true.
Copyright © 2004, American Society for Information Science and Technology