Sociological aspects of
information science

T.D. Wilson
Postgraduate School of Librarianship and Information Science
University of Sheffield, UK

Information science

'Information science' is defined here as that set of practices and related disciplinary studies which is concerned with the generation, transmission, organization, storage, retrieval and use of information* together with studies of the user of information. In brief it is concerned with the subject areas outlined in Figure I.

It will be clear that 'information science' is not a unitary discipline revolving about a single set of unique theoretical laws' and principles. Rather, it is a practical science in the same way that 'medical science' and 'agricultural science' are practical fields to which specific disciplines contribute research methods and findings.

This view of information science is derived from the belief that the phenomena open to investigation are qualitatively different at different places in the field described by Figure 1. Thus, at one point, for example, the retrieval of chemical data from computer files, we are concerned with finding ways of expressing in machine-readable and machine-manipulable form the content of chemical formulae; at another, say the study of citations to scientific literature, we are concerned with the statistical structure of the occurrence of a special type of linguistic string in bodies of text; while in a third, for example, the study of how people go about searching for information, we are interested in human behaviour. In each case different theories, different methods, and different research purposes prevail.

Sociology

Sociology might be called a 'hybrid' discipline as the following quotation suggests:

'Broadly it may be said that sociology has had a fourfold origin in political philosophy, the philosophy of history, biological theories of evolution, and the movements for social and political reform which found it necessary to undertake surveys of social conditions.' (Ginsberg, 1947)

The name itself is a nineteenth century invention of Auguste Comte and precisely what it connotes is a topic of some dispute related to the theoretical schools of the protagonists. In fact writers on sociology, sociological methodology, and the philosophy of the social sciences either assume that the reader has his own definition of sociology or give reasons for failing to define the subject matter in very precise terms. However, it is possible to identify some of the interests of sociology.

Sociology is interested in the way man creates and uses social institutions: institutions such as family, work, education, organization; it is interested in structures that arise out of social life such as class and bureaucracy; it is interested in the processes whereby these institutions develop and change over time and in cross-cultural comparisons; it is interested in the norms and values which define other kinds of order which 'society' defines as deviant; it is interested in how these norms and values become embodied in law, religion and other value-structures; it is interested in how individuals define their everyday social world for the purpose of acting within it; it is interested, in other words, in the whole variety of the life of man in society. It is this very universality of interests that makes definition difficult and which leads to the problem of identifying points of contact between information science and sociology.

Sociological theory

The problem of elucidating the possible sociological analysis of information phenomena is compounded by the variety of 'schools' of sociology that exist at the present time. The two basic perspectives of social system and social action (Dawe, 1970) have spawned a variety of 'sociologies', some of which are represented in Figure 2.

The division of schools is directly related to a philosophical argument over the nature of social reality and the consequent methodological dispute over how this reality may be understood. Crudely stated, the social systems perspective takes social phenomena as having existence similar to that of phenomena in the natural world and, therefore, capable of investigation in a manner modelled on the methods of natural science. This approach is labelled 'positivist' (a term of abuse among some of its opponents). These opponents are found under the umbrella of the social action perspective which, again stated crudely, holds that social phenomena are qualitatively different from natural phenomena in that they are socially constructed by men and women who are the actors in the everyday social world. In other words, social phenomena cannot be understood without reference to the meaning that their acts have for those who produce them.

In terms of what is considered to be suitable for investigation the two schools naturally differ - the system perspective concentrates upon institutions and structures, the action perspective concentrates upon processes. Conflict-sociology may contradict this but, typically, the roots of conflict are assumed to exist in structural aspects of society.

These two basic orientations give rise to a number of different sub-schools, the most prolific in this sense being the action perspective where the influence of phenomenology and existentialism has been strongly felt. Some of these sub-schools are identified in Figure 2.

It is not intended to define and discuss these sub-schools at length. They are presented simply to make the point that there cannot be a single sociological view of communication behaviour - what one chooses to investigate and how one chooses to investigate it will depend upon one's position vis-a-vis the fundamental philosophical issue. It should also be recognized that, such is the virulence of the debate, virtually no sociologist would accept Figure 2 as a representation of the field!

Communication behaviour

For a sociologist to be interested in the concept of 'information' it seems clear that there must be some link between that concept (however defined) and social behaviour. For this author that link is found in the fact of communication: uncommunicated information, or communication outside of any social context (such as computer-to-computer or satellite to earth receiving-station) is not a social phenomenon and, therefore, is outside the province of the sociologist. This clearly brings to mind the possibility of a demarcation dispute between those who would wish to call investigations in this area 'information science', those who would claim it for 'communication science' and those who want to call it 'the sociology of communication'. At the organizational level it is also the province of 'organization theory' and of 'management science'. One can ignore these disputes and simply say that there is an area of behaviour concerned with the communication of information and that this behaviour has a social context making it a suitable subject for sociological study.

One question remains as to what this behaviour is: Figure I may aid discussion at this point. As a diagram it is more or less self-explanatory and much like many others used by different authors. Information (some prefer 'knowledge') is produced within a social context of academic or other organizational settings and the process is subject to the influence of persons other than the creator of information. Sociological studies in this area range from all-embracing theories in the sociology of knowledge to small-scale studies of collaboration in the writing of scholarly papers.

Clearly, this area merges into 'primary dissemination': how information is made available to the world at large or to specialized sub-groups. This is a very large field for investigation which ranges from the study of mass communications in general to studies of the refereeing activities of scholarly journals, stopping off at the sociology of professions and the sociology of sociology on the way. In other words these areas of sociology provide the theoretical ideas on how the 'population of messages' comes about.

The recipient of information, his information-seeking behaviour (which involves social acts), his information-exchange activities and his use of information are also areas for sociological research. A.1 the back of these activities is the troublesome notion of information need: if one accepts that knowledge is a social product then it seems logical to assume that information need is also a social product. All of this last set of problem areas may be considered at several different levels from the aspect of society at large, through the study of communication processes and information needs in professional groups, formal organizations, and informal associations to the investigation of personal information needs and their social formation.

An earlier article (Wilson, 1977) set out a schema of the factors that may be held to influence information needs and information-seeking behaviour:

"Finding out about information needs involves asking:

Does this person or group need information? (Influencing factor - social role).

Does he know he needs information? (Influencing factor - problem recognition abilities).

What kind of information does he need? (Influencing factors - level of performance of role, nature of specific problem, environment).

These questions are difficult to answer because they imply that the person who needs information may not have defined that need and so may find it difficult to think of information in the same terms as the researcher. Whether one pursues questions of this kind depends upon:

• how well the target group is defined, and
• whether it is really necessary to have answers to such questions.

In other words, an interest can be claimed in only those who have defined information needs for themselves. Attention can then be given to information-seeking behaviour.

Finding out about information-seeking behaviour involves asking:

What does he do about his need? (Influencing factors - environment, urgency of problem, benefit of taking action, recognition of information sources, expectation of success).

How does he select information resources? (influencing factors - recognition of information sources, expectation of success).

How does he carry out a search for information? (Influencing factors - knowledge of information resource, ability to formulate search strategy, subjective factors)."

In a very crude way this presentation may be said to offer the outlines of a sociological theory for these areas. Each suggested factor with social connotations can be expanded into sets of propositions and derived hypotheses which could then be tested, for example:

Proposition I
Social role is a determinant of perceived needs for information.

Proposition 2
Roles which require the performer to generate new information or synthesise existing information are more likely to result in perceived information needs for their performers than other roles.

Hypothesis A: Scientists are more likely to regard the use of published documents as central to their role than are engineers.
Hypothesis B: Engineers are more likely to regard the use of published documents as central to their role than are managers.
Hypothesis C: Managers are more likely to regard the use of published documents as central to their role than are those they manage.

Proposition 3
The level of social system at which a social role is performed will be a determinant of perceived needs for information.

- and so on.

To elaborate a theory of this kind in a convincing manner is beyond the scope of this paper but it is clearly possible. The fact that such a theory has not been evolved seems to be due to the fact that so-called 'information needs studies ' have often turned out to be studies of document use. This points to the need for this general area to be carefully defined in terms of information need, information-seeking behaviour (which may include information exchange) and information use. Certainly the form that hypotheses take will depend very much upon which of these sub-areas one is investigating.

The third general area is that of information systems and their sub-systems for secondary dissemination. In principle this entire area can be regarded as a sub-field in the sociology of organizations (or mass communications, depending upon the ature of the system) but special consideration may need to be given to social aspects of the use of these systems to provide a bridge to the study of the information-seeker.

Social research methods

At this point I think it is necessary to deal with a point of potential confusion, that is, the distinction between social research methods and sociological research. There have been many studies in the information field which have employed social research methods; that is, methods such as questionnaire surveys, interviewing and observation. These methods are described as social research methods to signify their common usage across the social sciences - consequently sociological research is only one of the fields in which they are employed. When they have been employed in the field of users' needs or in systems evaluation they have often been misused and only very rarely has there been any attempt to embed them in a framework of sociological explanation. Sociological research proper may or may not employ social research methods but it must employ sociological concepts in explanations of communication behaviour.

Looking for work of this latter kind is like looking for a needle in a haystack, largely because of an overly positivistic approach to the problems before they have been adequately identified and defined. The phenomena to be investigated are assumed to be unproblematic in their existence and, hence, capable of investigation at a distance through the use of self-completed, mail questionnaires to collect data capable of computer analysis. The results have been uniformly disappointing both for those who were looking for increased understanding of the phenomena and for those who were seeking practical guidance on the design and management of information systems.

One conclusion, therefore, is that when social research methods are employed (whether for basic research or for evaluation) more care must be taken. This conclusion has implications for those funding research: properly conducted survey research with preliminary observational and unstructured interview studies of the subjects in their social setting, together with careful design of research instruments is time-consuming and expensive: any proposal which claims that it can be done cheaply and quickly should be regarded with suspicion.

Some examples

Examples of work which may be said to have been carried out within a framework of sociological explanation are scarce, as I have said: however, some work shows the beginnings of such a framework. If we accept a broad definition of information science that includes the whole spectrum of communication behaviour and associated sub-disciplines within management and organization theory, then a number of interesting examples can be found. A small number of cases to illustrate different aspects of the field as defined by Figure I follows:

Generation and production of knowledge

It would be tempting to review some of the work in the field of the sociology of knowledge under this heading and, certainly, that field is of significance to the whole of information science. However, the more obvious examples come from the sociology of science, or perhaps what might be better termed the "sociology of disciplinary areas" since it now covers more than science. The work of writers like Hagstrom (1965) Storer (1966) and Crane (1972) is well known in this respect but perhaps a less well-known paper might appropriately be reviewed here. It is a paper by Vaughan and Reynolds (1968) entitled 'The sociology of symbolic interactionism'. Their aim was to discover the relationship between adherence to a particular theoretical standpoint, ie whether or not symbolic interactionism was suited to handling the topic of social change, and certain sociological variables such as the academic background of the theorists, their relationships as students and teachers, and teaching in the same institution.

The respondents divided into two categories on the question: those in category A "offered a relatively unequivocal argument that symbolic interaction theory is capable of handling social change", while those in category B held a less comprehensive view of the theory. Analysis showed that the two groups were strikingly different in that 12 of the 15 in category A received their Ph.D degrees from the University of Chicago and, of these 12, 10 were students there when the other two were teachers. Finally, 10 of the 15 were members of the same faculty as at least one other member of the group. The other group was much less closely bonded in all respects. Since the theoretical viewpoint of the members of category A is the less orthodox, Vauehan and Reynolds argue that:

'...a cohesive, multi-bonded network of relationships is a necessary condition for the development of a social theory that challenges the theoretical status quo. '

and that:

'... some recent trends are possibly working against the kinds of relationship patterns that are most conducive to the development of new forms of social thought.'

Thus, this particular sociological analysis of the production of knowledge points to a quite significant finding for higher education and research.

Primary dissemination

The scientific journal has been one of the main channels for the dissemination of scientific knowledge for the past 500 years and, therefore, the control exercised over the appearance of papers in these journals is a matter of some importance. Diana Crane (1967) compared the acceptance of papers by the American Sociological Review and Sociometry which evaluate articles anonymously, with acceptance by the American Economic Review which does not preserve anonymity. She concluded that:

'... the distribution of characteristics such as academic affiliation, doctoral origin and professional age of contributors ... is similar to the distribution of these same characteristics among journal editors.'

and that the evaluation of articles is affected, to some degree, by non-scientific factors, the strongest of which is the sharing of common viewpoints based upon common training.

This work clearly points to the difficulty that authors, whose academic background is outside that of the 'gatekeeper', may have in achieving publication for their papers, and suggests the consequence that useful work may fail to find an audience.

Information-seeking, information exchange, and information use

These areas can be grouped together partly for convenience and partly because, in common with the other areas, there is a paucity of sociologically-oriented work.

In this area it is probably T.J. Allen's (1966) work which is best known. His study of communication in research groups revealed the significance of educational level (possession of the Ph.D degree) as a predictor of who communicates with whom. It seems likely, however, that this is partly the result of cultural factors since Pruthi and Nagpaul (1977) in a similar investigation in India found that the type of work performed and organizational hierarchy were of more significance.

Work by Wilson and Streatfield (1978) also draws attention to the significance of the organizational hierarchy in communication. For four groups of staff in social services departments (directorate, advisors, line-managers, social workers) observation of a total of 5,839 communication events of 22 members of staff showed that communication with line subordinates was the most frequently occurring category for three of the groups: communication with line superiors also accounting for significant proportions. Communication with non-line groups and with peers was less frequent.

The interview phase of this project also revealed differentially-perceived needs for certain types of information (eg legal, training, client records, statistical data, etc) on the part of these different levels.

Clearly, work of this kind is very closely allied to other studies in the area of organizational communication - a sub-field of organization theory which seems to be the most fruitful area for information researchers to look for concepts and theories. In view of the fact that Allen's work is so well known it is rather surprising that organizational communication in general is not paid more attention, but apart from, say. Alien and McAlpine et al (1972), citation to writers in that area seems to be very rare.

Organizational communications is a very extensive field but the following may serve to indicate the kinds of concerns there are within the field and the kind of results obtained. The first is by two well-known researchers, O'Reilly and Roberts (1977), and is entitled 'Task group structure, communication, and effectiveness in three organizations'. Briefly, its theoretical position is:

"... to view organizations as information-processing systems. To survive, a system must accurately sense its relevant environments, process information to make decisions, and coordinate and control its subunits and members. From this perspective, information and the capability to receive, process, and transmit it become crucial organizational functions."

They then show how the social structure (e.g., the social cohesiveness of the groups, the number of status levels in the group, and the number of occupational specialties) of work groups in three organizations affect perceptions of information accuracy and communications openness and how these are related to the effectiveness of the work groups. For information practitioners looking for justifications for their activities work of this kind seems to be very practically relevant.

The second piece of work is by Sarata (1976), which, starting from a completely different point, comes even closer to the concerns of the information practitioner. Sarata looked at seven job design variables in relation to staff in U.S. Child Care agencies including the opportunity for learning (clearly information-related) and 'information', defined in terms of the existence of institutionalized procedures for keeping staff informed on new practices in their own and other agencies. Sarata's research showed that the variable that best predicted the level of job satisfaction was the opportunity for learning, including institutionalized procedures for reviewing professional literature. 'Information' was significantly correlated with job satisfaction in general and particularly with positive attitudes concerning the field and the agency, and with the ability to use skills. Again, findings like these are directly relevant to those who wish to justify information systems on the basis of their contribution to the overall effectiveness of the parent organization.

Conclusion

What can we conclude from this brief, impressionistic survey? Three points appear to be worth stressing: firstly, there can be little doubt that, while some parts of the 'information science' field can draw upon natural science theories and methods, it is also necessary if the claims that information science is a social science (see for example Roberts, 1976) are to be pursued, to identify those parts of the field to which the social sciences can contribute. Some of the work reported above clearly shows that the links can be made, indeed it can be seen that sociological theory can contribute in different ways to almost any of the sub-fields identified in Figure I, and that particularly relevant contributions can be made to the study of the generation, dissemination and use of information. In addition Swift and his colleagues (1979) have suggested that retrieval systems for the social sciences will be less effective if these fail to take account of the theoretical orientation of research in these fields. Secondly, although the subject has been touched upon only lightly in this paper, there can be little doubt that investigators with a natural science background have tended to use social research methods in a very naive manner.

The problem seems to lie in two places: firstly, the scientist (whose education rarely includes adequate treatment of the philosophy of his discipline) fails to take sufficient cognisance of the complexities involved in investigating human behaviour; and secondly, he assumes that social research methods are easy methods to employ, that asking questions, or devising questionnaires are simple, common-sense activities requiring a minimum of effort. Fortunately, now that an increasing number of well-designed studies are being undertaken (for example, the Baltimore study by Warner et al., 1973), this attitude is disappearing. The lesson needs to be stressed, however, that anyone who steps outside his (or her) own discipline should have the intellectual honesty to admit ignorance and seek to improve his knowledge before conducting surveys. Finally, the fact that the information scientist can recognize the relevance of sociological research in areas such as the sociology of science and organizational communication should alert us to the possibility that there is a good deal of work going on in these and other areas which may usefully contribute theories, methods, and findings to information science. Some of these areas are particularly relevant to the information practitioner who is looking for ways of identifying the benefits of his services. Attention to work that points in this direction from the fields of organizational communication, mass communication, and opinion research, could enable the information researcher to demonstrate very clearly that his work is not ivory-tower-bound but has direct relevance to practice. Equally, it may be the case that sociologists could learn something from information science research: at the very least a dialogue is called for.

Note: Whether 'information' is a process or a 'stuff' does not seem to be relevant to a general discussion of this kind, but it is recognized that the way in which information is defined will affect the research process.

Acknowledgements I am grateful to my colleague Norman Roberts for commenting upon an earlier version of this paper (presented to a seminar on information research supported by the British Library Research and Development Department)

References

• Allen, T. J. (1966) Managing the flow of scientific and technological information. Cambridge, Mass.: M.I.T. (Ph D thesis).
• Crane, D. (1967) 'The gatekeepers of science: some factors affecting the selection of articles for scientific journals.' American Sociologist, 2, 195-201.
• Crane, D. (1972) Invisible colleges. Chicago: University of Chicago Press.
• Dawe, A. (1970) 'The two sociologies.' British Journal of Sociology, 21, 207-218.
• Ginsberg, M. (1947) Reason and unreason in society: essays in sociology and social philosophy. London: Longmans, Green and Co. for the London School of Economics and Political Science. (cited in Bottomore, T. B. Sociology. London: Unwin, 1966).
• Hagstrom, W. 0. (1965) The scientific community. New York: Basic Books.
• McAlpine, A., Whitley, R. D., and Frost, P. A. (1972) The flow and use of scientific information in university research. Manchester: Manchester Business School, R.& D. Unit.
• O'Reilly, C. and Roberts, K. (1977) 'Task group structure, communication and effectiveness in three organizations.' Journal of Applied Psychology, 62, 674-681.
• Pruthi, S. and Nagpaul, P. S. (1977) 'Communication patterns in small R.& D. projects.' Journal of Scientific and Industrial Research, 36, 505-510.
• Roberts, N. (1976) 'Social considerations towards a definition of information science.' Journal of Documentation, 32, 249-257.
• Sarata, B. P. V. (1976) 'Improving staff satisfaction through job design.' Administration in Mental Health, 4, 60-65.
• Stover, N. (1966) The social system of science. New York: Holt, Rinehart and Winston.
• Swift, D. F., Winn, V. A., and Bramer, D. A. (1979) 'A sociological approach to the design of information systems.' Journal of the American Society for Information Science, 30, 215-223.
• Vaughan, T. R. and Reynolds, L. T. (1968) 'The sociology of symbolic interactionism.' American Sociologist, 3, 208-214.
• Warner, E. S., Murray, A. D., and Palmour, V. E. (1973) Information needs of urban residents. Baltimore, Md: Regional Planning Council.
• Wilson, T. D. (1977) 'The investigation of information needs as a basis for training users in information use. ' International Forum on Information and Documentation, 2, 25-29.
• Wilson, T. D. and Streatfield, D. R. (1978) 'Information needs in local authority social services departments: an interim report on Project INISS.' Journal of Documentation, 33, 277-293.