The cognitive approach to
information-seeking behaviour and
information use

T. D. WILSON
Department of Information Studies
University of Sheffield, Sheffield S10 2TN, UK

The cognitive approach to 'information behaviour' centres upon the idea of meaning. Meaning is involved not only in all aspects of information generation, transfer and use, but also in the way people define themselves, their lives and their actions.
The cognitive approach, therefore, draws attention to the need for a bridge between the meanings of everyday life and the information that may have relevance for everyday life. In this sense, of course, 'everyday life' is different for every person—for some it may involve research as an everyday activity, for others the practice of a profession, for others, involvement in business and commerce.
The aim of this paper will be to explore these concepts in the context of research into information behaviour in the everyday world of work, where individuals may be constrained to a greater or lesser degree in their ability to define the content, direction and function of their work and where the diversity of roles may prevent the emergence of coherent groups capable of expressing clear needs for information support.
De Mey's presentation of the 'cognitive paradigm' will be examined and parallels to elements of such a 'paradigm' will be sought in the world of work.
Throughout, the paper will draw upon the author's present and past research.

The 'cognitive approach' is associated with other terms which draw attention to its relevance to information science. The term 'image' in the sense used by Boulding (1961) or the idea of 'world view' or 'frame' as used by De Mey (1982) are examples. They draw attention to the fact that the essence of the cognitive approach is the idea of human perception, cognition, and structures of knowledge. Of course, the approach is also relevant to ideas of machine 'intelligence' and the problems of building computers (or, more properly, programs) that learn, but this paper is about human beings.

The concepts 'understanding' and 'meaning' must lie behind any of the terms used to describe the cognitive approach. For example (although neither of the above terms appears in his index) in examining how scientific knowledge is generated and structured, De Mey is concerned with how 'shared understandings' arise in the course of development of scientific specialisms and how 'meaning' is attributed to data by reference to the context within which the data are produced, the discipline to which they refer, the paradigmatic methods adopted, and so forth. Data generated by non-paradigmatic means from outside the discipline by persons who are not recognized members of the group may be rejected by the insiders as having no meaning for the work upon which they are engaged.

The Velikovsky case (de Grazia et al., 1963) is an example: Velikovsky is an 'outsider' both to scientists and to historians and, therefore, his work in both areas is rejected by them as having no 'relevance' to their work. The fact that Velikovsky may well be mistaken in all of his ideas is irrelevant: the point of interest is that unless a researcher shares his 'ways of knowing' with others in a way that gives them legitimacy his work will be disregarded or rejected.

Following Kulin, De Mey uses the concept 'paradigm' as his world model of science and draws attention to four components of the scientific paradigm:

1. Its use of symbolic generalizations.
2. Its underlying metaphysical beliefs.
3. Its values.
4. Its examplars (examples of best practice).

An extensive account is not necessary here but a brief example for each element may be instructive:

Schutz discusses something similar to the Kuhnian idea of a paradigm under the heading 'systems of relevance and typification'. He is worth quoting at length:

A system of relevances and typifications as it exists at any historical moment is itself a part of the social heritage and as such is handed down in the educational process to the members of the in-group. It has various important functions:

1. It determines which facts or events have to be treated substantially,...for the purpose of solving in a typical manner typical problems that emerge or might emerge in situations typified as being equal…
2. It transforms unique individual actions of unique human beings into typical functions of typical social roles, originating in typical motives aimed at bringing about typical ends...
3. It functions as both a scheme of interpretation and as a scheme of orientation for each member of the in-group and constitutes therewith a universe of discourse among them…
4. The chance of success of human interaction… is enhanced if the scheme of typification is standardized, and the system of pertinent relevances institutionalized…
5. The socially approved system of typifications and relevances is the common field within which the private typifications and relevance structures of the individual members of the group originate…

(Schutz, 1970:120-122)

The scientist, working in the context of this system of relevances and typifications (or 'image') approaches the ideal type of 'the expert':

The expert's knowledge is restricted to a limited field but therein it is clear and distinct. His opinions are based upon warranted assertions; his judgements are not mere guesswork or loose suppositions (Schutz 1946:464).

In information science, those who need access to expert knowledge have been given a good deal of attention over the years to that extent that the study of information-seeking behaviour can be said to be the study of scientists' information-seeking behaviour. It can be argued that the attention which scientists have been given over the years has resulted in a dis-service to information science, particularly so far as the behavioural aspects of information science are concerned. The conduct of science is a highly patterned affair with its own peculiar norms of what is acceptable behaviour and the role of information-seeking and information use in those norms is rather clearly defined.

The scientist deals with relatively well-structured problems for which lines of approach (examplars) to a solution may well exist and many others may have followed those lines for years previously. Examining that previous work, searching for missed associations, following through approaches that were formerly believed to be blind alleys, picking up promising hints overlooked by previous explorers, is part of the essential business of being a scientist. Paying recognition to previous work through citation, thereby recording 'use' of a document is also part of the rules-of-the-game, or in Schutz's words, part of the scientist's system of relevances and typifications. As a consequence, the identification of potentially useful information, the design of information systems, and the delivery of information services is a relatively straightforward matter for the information professional not least because his system of relevances, as it relates to documents and document use, overlaps with that of the scientist—a similar degree of significance is attached to documents, to information searching, and to information transfer.

The 'scientific paradigm' has also had another effect upon the behavioural side of information science: until the Baltimore investigation adopted the idea of 'problems' in the study of citizens' information needs (Warner et al., 1973) and until Project INISS (Wilson and Streatfield, 1977) adopted an alternative approach for the study of social work staff, the norm in investigating 'information needs' was to investigate 'document use'. The self-completed questionnaire ruled virtually supreme as a data collection instrument, in spite of its very severe disadvantages, and the increase in our understanding of the processes of information-seeking and information use and their relevance to the work of scientists was virtually zero since such studies told us little more than what we already knew.

However, science is only one area of human life and work, and rather asmall area at that. The other kinds of work in which people engage are too numerous to mention but each is distinct from all others and very few share the same characteristics as science. To treat all potential users of information as though they were scientists and to regard all contexts of work as the same as the context within which science is carried on is, to say the least, 'unscientific'. If the concepts of 'image', 'world view', and 'frame of reference' mean anything at all they mean that the way the reality of these concepts in any one individual affect that person's behaviour will vary according to how they are constituted, formed, and changed. The aim of a cognitive approach to information-seeking behaviour and information use, therefore, is to discover how the images and frames of reference of people in other areas of work than science (or rather as well as in science, because the task has not yet been carried out there) relate to the availability of information, how the choice of information is determined by the image, and how information may change the image, or otherwise affect a frame of reference.

Here again we can turn to Schutz for insight. He identifies two other 'ideal types' in relation to the social distribution of knowledge: the 'well-informed citizen' and the 'man on the street'. The man on the street relies upon 'recipe knowledge' for dealing with typical situations to achieve typical results. His treatment of situations may be as much affected by emotion and passion as by knowledge. The well-informed citizen, on the other hand:

…neither is, nor aims at being, possessed of expert knowledge; on the other hand, he does not acquiesce in the fundamental vagueness of mere recipe knowledge or in the irrationality of his unclarified passions and sentiments. To be well informed means to him to arrive at reasonably founded opinions… (Schutz, 1946: 468).

Schutz's trio of ideal types suggests that different persons, in different occupations may possess different world views and make different demands upon sources of knowledge as a consequence. For example, some occupations may require no more than 'recipe knowledge' for their effective performance; others, falling short of a need for 'expert' knowledge, may demand more in the nature of 'reasoned opinion' and, hence, a greater need for access to sources of information.

Of course, the 'images' with which we as information scientists are most familiar, through research, are the images of research scientists. Berger et al. (1974), however, offer a view of the 'cognitive style' of the bureaucrat, an occupation which is widely followed in most countries of the world and which, as yet, has had little in the way of information science research applied to it.

Berger and his colleagues begin by considering the view of bureaucracy which is held by its clients and note that:

A key notion in the individual's knowledge of the bureaucratic system is that of competence, each jurisdiction and each agency within it is competent only for its assigned sphere…
Another general notion about bureaucracy is that of proper procedure. Bureaucracy is assumed to operate within rational rules and sequences. These are known or in principle knowable… This implies the possibility of improper procedure and of avenues of redress. Indeed, very often the laws that set up a particular bureaucratic agency provide such avenues explicitly…,
There is finally a general notion of anonymity. Bureaucratic competences, procedures, rights and duties are not attached to concrete individuals but to holders and clients of bureaucratic offices...
(Berger et al., 1974: 46-48).

These elements of the clients' consciousness of bureaucracy are a consequence of a number of factors: in part, the client's past dealings with bureaucratic agencies; in part, generally understood concepts of bureaucracy within a particular society; and in part, characteristics of the agencies expressed through documentation or practices of dealing with clients. Clearly, the clients' consciousness is not necessarily the same for all clients of all bureaucratic agencies in all places. For example, in some countries, a characteristic of clients' consciousness would be that bureaucracies are slow, or corrupt.

Berger et al. also identify elements of the bureaucratic consciousness:

An overriding element is orderliness. Every bureaucracy must produce a system of categories into which everything within a certain jurisdiction can fit and in terms of which everything can be handled… this orderliness is based on a taxonomic propensity… Phenomena are classified rather than analysed or synthesized… the bureaucrat is typically satisfied once everything has been put into its proper box…
Bureaucracy presupposes general and autonomous organizability. In principle everything is organizable in bureaucratic terms…
There is a general assumption of predictability. It is assumed that bureaucracy will operate with certain regular procedures. These procedures are known and can therefore be predicted…
There is a general expectation of justice. It is expected that everyone in the relevant category… will receive equal treatment… What emerges from the combination of the aforementioned elements of cognitive style might be described as moralized anonymity… The bureaucratic system as a whole is deemed to have moral obligations towards its anonymous clientele (Berger etal., 1974: 50-53).

The differences between this view of the 'bureaucratic consciousness' and that of the scientist do not need to be laboured: it is clear that bureaucrat and scientist inhabit different spheres of life, perform different kinds of work, and will make different kinds of assumptions about information and access to it as a consequence. The question is: 'Do we have any evidence that this is indeed the case?'

The answer is, a little. I do not propose to give yet another paper on Project INISS and associated work – there are enough sources on that already. However, I do want to draw out some of the conclusions in a way that may illustrate the consequences for information services of a greater awareness of the users' images or consciousness.

First, the 'expert' knowledge claimed by social services staff is, in general, expert knowledge about the organization and its clients: that is, it tends towards the 'commonsense' end of the spectrum of specialized knowledge rather than towards the 'esoteric' end. This is clear from Table 1 below:

Secondly, it is pretty evident that, as bureaucracies are often established on a legal basis and, as Berger and his colleagues note, follow 'procedure', legal and procedural information may play a significant part in the work of employees of bureaucracies. Social services departments in England and Wales are established in law, have functions allocated to them by laws, and carry out those functions according to predefined procedures (whenever possible). The effect of this on the nature of the documents that pass across their desks is clear from Table 2.

Secondly, because of its concern for orderliness, and because of the general principle that everything is capable of being organized, the bureaucracy is much concerned with its own operations. Indeed, in looking at a public service bureaucracy, we found that the biggest category of communications related to the administration of the organization and communications relating to the clients came only second (Tables 3 and 4).

Both of the preceding points are supported by additional data on the frequency of perceived need for information of different kinds collected in interviews (Table 5).

The data suggest the 'commonsense' nature of information need (for example, the frequent need for 'directory' information and the weekly need for 'news' compared with the much less frequent need for 'research' information). They also suggest the concern with orderliness and the operation of the bureaucracy in the need for legal', 'procedural', 'client records', compared with the need for less organization-centred documentation such as 'central government statistics', 'research', and 'training' information.

Thirdly, the idea of 'moralized anonymity' is probably nowhere more strong than in social services departments. In running courses on the applications of information technology in social services departments we have found that one of the overriding concerns is that, if client records are converted to machine-readable form, security should be absolute. Great difficulty has been experienced in these departments in attempting to introduce computerized systems when fears about security have not been allayed.

Additional work carried out in subsequent years in other departments of local government, such as education departments, housing departments, and, most recently (Wilson and Masser, 1983) in planning departments could be analysed in these terms – if one had time – and one begins to see possibilities for the conduct of research into information-seeking behaviour and information use under the cognitive paradigm which may prove beneficial for the further development of the discipline and of information systems and services.

BERGER, P. L., BERGER, B. and KELLNER, F. (1974). The homeless mind: modernization and consciousness. Harmondsworth: Penguin.

BOULDING, K. E. (1961). The image. Ann Arbor, Michigan: University of Michigan Press.

DE GRAZIA, A., JUERGENS, R. E. and STECCHINI, L. C. (1963). The politics of science and Dr. Velikovsky. New York: American Behavioral Scientist.

DEMEY.M. (1982). The cognitive paradigm. Dordrecht: Reidel.

SCHUTZ, A. (1946). The well-informed citizen: an essay on the social distribution of knowledge. Social Research, 13, 463-478. (Also reprinted in Collected Papers, Vol. II, 1964,120-134)

SCHUTZ, A. (1970). On phenomenology and social relations: selected writings, edited and with an introduction by Helmut R. Wagner. Chicago: University of Chicago Press.

WARNER, E.S., MURRAY, A. D. and PALMOUR, V. E. (1973). Information needs of urban residents. Washington, D.C.: US Department of Health, Education and Welfare.

WILSON, T. D. and MASSER, I. M. (1983). Information management in local authority planning departments. Paper presented at the 5th International Research Forum on Informa-tion Science, Heidelberg, September.

WILSON, T. D. and STREATFIELD, D. R. (1977). Information needs in local authority social services departments: an interim report on Project INISS. Journal of Documentation, 33, 277-293.