IFLANET home - International Federation of Library Associations and InstitutionsAnnual ConferenceSearchContacts

63rd IFLA General Conference - Conference Programme and Proceedings - August 31- September 5, 1997

INTERNATIONAL LIS RESEARCH: A comparison of national trends

Maxine Rochester and Pertti Vakkari



The study of national trends in library and information science (LIS) research started with articles by Järvelin and Vakkari (1990; 1993) on trends in international LIS research. They described the trends in LIS research by comparing distribution of topics, approaches and methods in the years 1965, 1975 and 1985. The source material contained articles published in core journals of LIS. This study led to a research project that compared LIS research in Scandinavian countries (Aarek & al 1993; Vakkari & al 1993; Vakkari 1996).

The Section of Library Theory and Research of IFLA initiated a series of national studies based on the research by Järvelin and Vakkari. These were inspired by Cano and Rey’s (1993) presentation on trends in Spanish LIS research at the Section’s Open Forum Meeting in Barcelona 1993. Their paper was based on the classification schemes by Järvelin and Vakkari. The Section for Library Theory and Research considered these classification schemes as validated tools by the study of Cano and Rey for comparing research in LIS in various contries, although they also raised justified criticism towards the schemes. At the Open Forum Meetings of our Section we have heard reports on LIS research from Turkey (Yontar 1995), China (Chen 1996) and will hear one from the UK in Copenhagen (Layzell Ward 1997). Small project funding from IFLA assisted a study for Australia by Rochester (1995) and one just completed for Turkey by Yontar and Yalvac (1997). We will use these in our comparisons.

A small project grant to Rochester and Vakkari for 1997 allows a comparison of international and national trends in LIS research; this paper is a preliminary report. Our aim is to present a summary of findings from the studies mentioned above. We will compare most popular topics, subtopics and methods in the national LIS studies and relate them to the international trends reflected in the study by Järvelin and Vakkari. The countries include Finland, Spain, Turkey, Australia, China and the UK.

It is interesting to analyze how the national characteristics of LIS differ from the international trends in LIS research. The comparison reveals the peaks and valleys of the national LIS landscapes, and their relation to the international trends. This analysis gives us a descriptive account of the situation. If we are interested in the differences in national features of research, we have to seek the explanation from the social and cultural differences of those countries. These factors have an impact on the formation of the national innovation system in a country. An innovation system includes systems for higher education and research. The characteristics of the innovation system for its part determines the formation of LIS research. Both economic and cultural features of a country affect LIS research through its innovation system.

Social and cognitive institutionalization of research fields

Richard Whitley’s (1974) differentiation between social and cognitive institutionalization of research fields gives us a tool for analyzing differences in LIS research between countries. The social aspects concern in particular differences in the structuring of education and research and the degree of organization within the scientific community. The cognitive aspects refer to how central concepts and theories within the discipline are defined, as well as to main research areas, basic problems, methods and relevant solutions. Whitley’s basic idea is to examine the association between intellectual products and the way in which they are produced.

The basic difference between scientific structures is the degree of their institutionalization. Institutionalization refers to the patterning of actions and meanings. The degree of coherence and organization of actions and perceptions, and the extent to which ideas are articulated and adhered to constitute the degree of institutionalization. A field exhibits a high degree of institutionalization when the researchers share a common attitude in terms of its aims, methods, and explanation ideals. The more consensus there is about the central ways of conceptualizing the field, its basic problems and methods, relevant solutions and results, the more cognitively institutionalized it is.

Social institutionalization refers to the creation and maintenance of formal structures, which demarcate members of a cognitive structure. These arrangements function as a basis for the social identity and as an organizing principle for the activities. Social institutionalization is made up of two dimensions. On the one hand it refers to the degree of internal organization and boundary-definition, and on the other hand to the degree of integration in the social structures in terms of legitimization and resource allocation. (Whitley 1974, p. 72) The first dimension includes the foundation of research associations and formal communication channels like scientific conferences and journals. The latter dimension refers to the degree of integration of the field into university departments and teaching curricula. University departments, chairs, teachers’ and research posts, and doctoral programs are the hallmarks of a mature social structure of a speciality. One could suppose that a cognitive structure implied by a high degree of social institutionalization of LIS differs from the cognitive structure produced by a lesser institutionalized social structure.

The compared studies do not include data about the social structures of LIS research due to the fact that the source article did not include this problem formulation. Thus, the differences in the research between the analyzed countries can be explained only by giving civilized guesses. However it is possible to illustrate how the social dimension of research shapes its cognitive output by using Scandinavian countries as a case. A comparative study of LIS research in Scandinavia has used Whitley’s frame as its point of departure (Aarek & al. 1993; Vakkari 1996). We will complete our analysis of trends in LIS research by this case, and one for Australia.

Construction of library and information science and its subfields: How to squeeze reality in a mould

In order to be able to study the trends of LIS research one has to differentiate research from non research, e.g. from professional writings, and to demarcate research in LIS from other research. After that a content analysis scheme for the subfields of LIS and other features of research can be developed.

Järvelin and Vakkari (1990) excluded professional publications from the analysis. If the knowledge base consisted of other than research results and metatheoretical statements, we would be dealing with pseudoscience. Their definition of research, which was adopted from Peritz (1981), was expressed in quite general terms: "Research is an inquiry, where the goal is to elicit, through a systematic method, some new facts, concepts or ideas". Today they would use some additional criteria: a sound frame of reference, exact problem formulation, and connection to earlier research. However, Järvelin and Vakkari still believe that this definiton was strict enough for helping to differentiate research in a quite unequivocal sense.

When constructing a classification scheme of LIS one has to have some kind of understanding about the scope of LIS and its major subfields. How should the discipline be demarcated from other fields? Which themes and problems belong to the domain of LIS, and which do not? What are its central subfields? There is no pre-existing entity that can be called LIS, in the same way we are able to name familiar objects like an orange, and divide it into segments. Thus, the solutions, and definition of the domain is always to some extent normative. It is a construction guided by some metatheoretical presuppositions and directives.

In the first article Järvelin and Vakkari defined LIS ostensively by referring to the core journals of LIS. What is published in these journals indicates the domain of LIS. In connection with the Scandinavian study (Vakkari & al.1993; Vakkari 1996) a definition was developed: "We conceive of LIS as a discipline that views information processes from an information seeking perspective. This does not mean that the research exclusively focuses on information seeking, but this perspective essentially structures the discipline. The objective of the investigation is the information seeking of individuals and groups, the factors that generate this activity, as well as various arrangements and conditions that support the information seeking and provide access to information (e.g. LIS units)".

A ground plan for the topics of LIS was outlined from this perspective. However, the definition did not contain clearcut and detailed building bricks for constructing the branches of LIS. The original classification scheme was designed partly on the basis of the contents of the articles forming the data, and partly on the basis of relevant earlier classifications and the theoretical knowledge by the authors. Although the final version was a result of the interaction of the data and the theoretical understanding of the discipline, the most crucial was the latter one. One would claim that it is impossible to create a theoretical construct solely on the bases of the data. Single units of the data as such would not tell one what kind of classes one should form. In order to be able to infer classes from the data, to conceptualize it, one has to have some theoretical ideas in mind. All our observations are theory laden. It depends on our way of seeing, on our frame, whether the bottle is half empty or half full.

The classification scheme for topics of LIS by Järvelin and Vakkari (1990) was a construction of LIS and its subfields (Appendix 1). It reflects the understanding of the field as it was in the middle of the 1980s. One can consider how well this drawing of the map resembles the landscape of LIS about ten years later. At the time of its creation the topic scheme of LIS was the most comprehensive and systematic attempt to divide our field of research into subfields. That it has been used for further studies is an indication of that. It left, however, room for improvements. It was criticized - justifiably - by some colleagues. Cano & Rey (1993) suggested that it should include more elements of library and information systems input and also take into account the social dimension in more detail. Despite criticism, no one has presented a more valid version of it. An old truth is that it is always more difficult to be a system builder than its critic.

Also the classification of research methods (Appendix 2) by Järvelin and Vakkari is open to critical remarks. Some classes are not always so easy to differentiate from each other. The problem is the lack of the necessary detailed operational definitions of classes which would help sort the problematic cases into the right position in the scheme. The same problem holds for some cases in the scheme for the topics. Cultural differences especially might affect the understanding of the content of similar expressions. Although a common noun is used, persons from different backgrounds might refer to a different set of entities, e.g. interpretations of some classes both in the topic and method schemes are different in the Chinese study of LIS research (Cheng 1996).

Although the scheme provided a sound base for the analysis of national LIS research output, it is evident that it reflects the cognitive tradition of LIS in the western industrialized world. It might be more difficult to apply it to the research output of other countries.

Data from the participing countries

A summary of the data from the studies is presented in Table 1.

Tables are not available. Please contact author.

The time periods of the studies vary to some extent. International trends are from the years 1965, 1975 and 1985. The date from the UK is from the years 1965, 1975, 1985 and 1995. The Turkish data is a cumulation of ten years intervals 1952-1964, 1965-74, 1975-1984, and 1985-94. The rest of the studies include data from a shorter time period. Spain includes a cumulation from the 10 year period 198?-199?, and Australia also a ten year cumulation for the period 1985-1994. Data from Finland is also cumulations of three periods, 1965-74, 1975-84, and 1985-89. The Chinese results are based on data from the years 1985 and 1995. The time periodization of the national data sets is comparable. Also the international data representing the trends with intervals of ten years is comparable with the national data.

How well the data represents the research output of LIS in the studied countries depends upon the selection of the sources whence it has been acquired. The Finnish data included all the research publications, both articles and monographs. The data from the other countries consisted of articles from the core research journals of the field. However, there are not pure research journals in all the studied countries. Thus some of the journals were professional ones that included also research articles. It is difficult to assess how well the selected journals represent the total research output of each country. However it is plausible to suppose that the core journals are publishing the most important research results and thus reflect the main trends of research in the countries involved. Still one can doubt that the publication patterns in the subdomains of LIS differ so that e.g. humanistically oriented scholars publish their results in the form of monographs rather than as journal articles. One can claim, however, that humanists also publish articles in general and when preparing a monograph in particular. Thus the data is representative also in this respect.

Topics for research

When we consider the findings of the three most popular topics for research from an analysis of the international and the national studies of LIS research, as given in Table 2, we find some broad similarities. The classification of topics devised by Jarvelin and Vakkari is given in Appendix 1. The cautions already expressed about the application of the classification schemes should be kept in mind.

In the international journal literature information storage and retrieval was the most popular topic and library and information services the second most popular for the years 1965, 1975 and 1985. The analysis of the UK literature revealed that library and information services were the most popular topic in each of the years 1965, 1975, 1985 and 1995. Information storage and retrieval was the third most popular UK topic in 1965, and second most popular in the years 1975, 1985 and 1995. Also information seeking, the third most popular topic internationally at 8% in 1965 was the second most popular at 25% for the UK. Information seeking was still popular in the UK in 1985 at 22% and in 1995 at 16%. Thus the UK and the international journal literature showed similarities for topics to be investigated over an extended period of time.

For Turkey library and information services were investigated by 43% of the researchers in the early period, the same percentage in the period 1965 to 1974 and 60% in the period 1976 to 1984 and still 59% in the period 1985 to 1994. As for the international articles and the UK, information storage and retrieval was popular for Turkish researchers: 14% in the period to 1964, 50% in the next ten year period, 9% and 11% in the next ten year periods to 1994. Library history accounted for 43% of the research topic articles in the early period, but did not feature again as a popular topic. For Finland we also find library and information services a popular choice: 36% researched it in 1975, 21% in 1985 and 19% for the period 1985-1989. Information storage and retrieval was also popular: 20% in period to1985 and 28%, the most popular, in 1985-1989. Information seeking also featured for Finnish researchers: 32% in 1975, 25% in 1985 and 11% in the final period. Other topics were investigated by 11% in 1975 for Finland.

In the Chinese literature library and information services were the second most popular topic in the two periods: 25% in 1985 and 20% in 1995. Other popular topics were unique to China: principles of library and information science at 26% in 1985 and 28% in 1995; and related disciplines at 19% in 1985 and the information industry at 15% in 1995. For the Australian and Spanish library literature only a recent ten year period was analyzed. Library and information services was the most popular in Australia at 40%, and popular in Spain at 19%. In Spain information storage and retrieval also attracted 19% of researchers. Scientific communication, also at 19%, was unique for Spain, and library history at 14% for Australia was third most popular, as it had been for Turkey in the early period.

It seems that strong interest in LIS services has been typical of the research in Australia, Turkey and the UK. The literature in these countries contains relatively more publications on this topic than the international journals. Information seeking has been a more popular topic in Australia, Finland and the UK than in the international literature. The internationally most popular topic, information storage and retrieval, has been a well researched area only in Finland and the UK. LIS research in these two countries seems to reflect international trends more closely than research in the other participating countries.

We should also consider the most popular subtopics investigated (Table 3). The data is more sparse here and there seem to have been difficulties in applying the classification scheme. The popular subtopics for the topic information storage and retrieval were classification and indexing, information retrieval and cataloguing. The most popular subtopic in the international literature was classification and indexing: 22% in 1965 and 14% in 1975. It was most popular also in Turkey in the period to 1975. In China it attracted 8% in 1985 and in 1995. Cataloguing attracted 14% in Turkey to the period 1975. Information retrieval was the subtopic for 8% of international articles in 1975, and 13% in 1985. For Finland information retrieval attracted 9% of research authors.

For the popular topic library and information service activities we find the subtopics collections, administration and automation popular in the international articles. Administration and collections were also popular in Australia. Circulation, collections, administration, automation and several interconnected activities were also examined by researchers in Turkey.

The topic information seeking had some popular subtopics: use of library and information services attracted 10% in Australia, information seeking behaviour attracted 7% in Finland and use/users of information attracted 6% in Turkey in the period to 1984 and 8% in the last period.

Reseach methods

When we consider the three most popular research methods used we again find some methods are universally popular (Table 4). In the international literature the conceptual research method remained the most popular over time: 29% in 1965 and 1975, 23% in 1985. Also the survey method was often used: by 23% in 1965, 20% in 1975 and 23% in 1985. Historical method was used by 11% in 1965, and replaced in popularity by system design in 1975 and 1985, at 15%.

Findings for articles over three periods are available also for the UK literature. Here there were changes: the discussion method was very popular with 73% of research article authors in 1965, fell to 34% in 1975, and no longer featured as a popular method in 1985 and 1995. The survey was the second most popular in the first two periods, and in 1985 and 1995 was most popular at 20% and 29% respectively. The literature review featured as third most popular in the four test years. The conceptual method featured as the second most popular method in 1985 and 1995. Case or action research appeared as equal third most popular in 1995. Survey and conceptual methods were the only ones also popular in the international literature. For Turkey in the early period the conceptual method was also popular at 29%, as also was historical method: 29% in the early period and 17% in 1975-84. The survey method was also frequently used in each of the four time periods examined. The literature review took over as the most popular method for the three periods after 1965: 57%, 37% and 47% respectively.

For Finland for the three time periods we find the same three methods used as in the international studies: survey, historical and conceptual, but in different orders of frequency of use. Historical was the method most used in China for both periods examined, 25% and 18% respectively. Also conceptual and historical method appeared in both time periods. For Australia the survey was used by 44%, with historical and discussion method used by 14% and 10% respectively. For Spain we do not have detailed data, but empirical research strategies were employed in 33% of articles, conceptual/mathematical in 7% and descriptive and discursive methods in 36%.

Thus overall we see the survey method was popular internationally and nationally, as was the historical method and the conceptual. We need to ponder these findings and think about the popularity and changes in popularity of various research methods over time. We should also ask why some research methods, such as experimental or qualitative, widely used in other discipline areas, are so little used in LIS.

The Scandinavian case

In the following we examine the connection between social institutionalization and cognitive development of LIS in Scandinavia during the period 1965-1989. The case is based on the article by Vakkari (1996). First we analyze the development of the social structures within LIS and the research conditions that have been created in these countries. Then we examine the cognitive development of the research in relation to the social structures.

Social structures

Library and information science was and is integrated in the Nordic institutions of higher learning in various ways. One extreme is the situation in Finland, where LIS has been a part of the university structure since 1971. Elsewhere LIS has been linked to various separate professional colleges, although changes have been seen, starting in the late 1980s, especially in Sweden. In Finland, the discipline has established itself as a part of the traditional university structure. Separate disciplinary departments with professors and research positions of their own have been established. This has guaranteed a continuity of the research. The establishment of doctoral programs has also strengthened continuity. It offers opportunities to educate researchers who concentrate on LIS.

In the other Scandinavian countries LIS has been placed in separate professional schools. During 1965-89 the schools in Denmark, Norway, and Sweden did not have any professorships nor any research positions. Neither did they have any doctoral programs.

The social institutionalization was during the period of investigation most developed in Finland. In spite of scarce resources social institutionalization in Finland meets all the conditions that Whitley's (1974, p. 72) definition of a high degree of social institutionalization requires: a department, a professor, research positions, and programs of research training as well as a scientific association and a scientific journal that function as the basis for communication.

Up until 1989 no funds had been allocated for a professor level position in LIS in Sweden, but a few occasional research positions in connection with other disciplinary departments had been financed. There were no doctoral programs, and the researchers had to get their research training in other disciplinary programs. However, research opportunities increased with financing that was targeted exclusively to LIS. In Sweden, a scientific journal in the field is being published. In Denmark, the social organization of the research is limited to funding that The Royal School of Librarianship has allocated for research, the publishing of a journal and the library history year-book. Research positions or research training did not exist. Norway lacked most of the social structures of LIS. At the library school in Oslo there is a body called BRODD (School's consultancy and applied research unit). It can be viewed as a social structure that integrates the professional goals of the research and the profession.

Because of the low degree of social institutionalization in Denmark, Norway, and Sweden, the researchers in these countries had limited structural opportunities to associate with the social organization of LIS. The preconditions, everything from research training to communication channels, were inadequate. This is why the profession offers for researchers a noteworthy alternative for the creation of an identity. The low interest in the field's research themes on the part of other disciplines contributes to marginalizing those with research training and strengthens their connection with the profession. Other disciplines don't succeed in offering attractive enough social structures. In Sweden, however, there were a few academic departments that have been able to attract library researchers. They offered a research environment with opportunities for communication and identification.

When the degree of social institutionalization is low, a strong professional organization with significant financial resources can regulate the cognitive direction of the research by favouring certain subject areas and problems. A good example is Folkebibliotekens Rådighetssumma (The Public Libraries' Financial Research Aid) in Denmark, which led to Danish library research being directed toward public library problems. Another example is the problems concerning automated library systems, which in Norway and Sweden in the 1970s and 1980s led to a concentration on library research in these areas.

Social institutionalization, definition of LIS and other cognitive features of research

The degree of social institutionalization affects the way the discipline and the cognitive features of research are seen. In Denmark, Norway and Sweden the social structure of LIS was undeveloped. When the structural opportunities for research are almost completely lacking, the profession offers the only obvious road for creating them. The researchers had usually a professional background, thus, it was only natural to identify themselves with the ideals of the profession. The implication of this identification was the sharing of the ideals concerning the nature of the discipline and research. The discipline was linked to the problems of library and information service organizations. The internal distinctions within the profession were generalized to apply to the discipline as a whole. The functional differences were projected onto science as well. This has not been an uncommon feature in discussion about the nature of LIS outside Scandinavia (Biggs 1991; Cronin 1995; Ford 1990). The strong links with the professional structure caused also a stronger emphasis on applied research. The weak social institutionalization of LIS implied in research output a more system-oriented understanding of LIS, concentration on library-related problems, and seeking for solutions with have immediate applicational value. These features were reflected by the choice of library organizations, and topics which are related to them, as a research object in studies more often in these countries than in Finland.

In Finland, the integration of LIS within the university organization meant that the researchers started to identify with the norms of the research community. This was the only way to obtain legitimization of the discipline within that community. At the same time it caused a distancing from the norms of the profession.

The primary goal for the academics is to educate competent researchers and do good research. The primary contribution is the internal development of the discipline, the creation of new theories and concepts, and the improvement of methodology. The practical applications that are of importance to the profession are of secondary importance to the discipline (Bunge 1982; Giere 1988; Vakkari 1989). Basic research therefore has become a priority.

The integration of the discipline into a university community meant that the character and the definitions of the dicipline had to be problematized. The research community disassociated itself from a definition based on the library organizations and replaced it with a view that is based on a broader theoretical foundation.The idea of facilitating access to information was the integrator of the dicipline. This also affected the choice of research themes and problem formulations in other areas than the library-oriented problems. The library organizations were viewed as special cases.

It looks as if the social organization of LIS in the Scandinavian countries was associated with what kind of research was being pursued. The researchers' identification with either the research or the professional community functions as a mediating factor. Strong social institutionalization creates identification with the norms of the research community, which leads to an appreciation of basic research and theoretically broader problem formulations. Weak social institutionalization leads to a professional orientation, which is followed by a compliance with the research ideals of the professional community. It appears that the differences in identification in general leads to diverging views on the research object of LIS. The research community rewards non-system-oriented studies, where the LIS organizations are viewed as a component in the information gathering process. Professionally oriented research concentrates on library organizations. The consequence is a difference in the cognitive structures within LIS. The research community also considers themes, perspectives, problems and solutions other than those that focus on the LIS organizations as important for research, while professionally oriented research focuses on themes that are important for the LIS organizations.

The results suggest that the degree of social and cognitive institutionalization is not clearly linearly dependent in the Scandinavian countries. However, the general direction is that a well developed social structure implies a non-system-oriented cognitive structure. The Finnish research exhibited more of these features than the others. In Norway, where the social institutionalization was the weakest, the research most clearly exhibited a professional cognitive orientation. When the cognitive features were concerned, Denmark and Sweden were often placed between Norway and Finland. The differences were not always systematic, but often supported the hypothesis.

The Australian case

A preliminary analysis of the social structures and cognitive development of LIS in Australia during the period from the mid 1980s to the mid 1990s follows.

Social structures

In the mid 1980s in Australia there existed only two library schools in universities; they both had professors of librarianship. Other library schools were in higher education institutions known as colleges of advanced education or institutes of technology. These institutions emphasized teaching rather than research, but some began to offer research Masters degrees in the 1980s. In the late 1980s there was a major reorganization of higher education in Australia, with mergers and upgrades of these college institutions to university status. Material on social structures is based on a book by Rochester (1997).

So all eleven library schools/departments are now located in universities, which are enhancing their research cultures. Faculty are beginning to identify with the research community. Those teachers without doctoral qualifications are mostly acquiring them. There are now five full professors in the LIS area in Australia but few research positions. The discipline areas represented among the schools/departments are traditional LIS, teacher librarianship, archives and records management. As Australia has a population of only 18 million people and there are eleven schools/departments, there are some small departments. The number of schools can be partly attributed to the immense distances between cities in Australia. Courses are three year general librarianship and four year teacher/librarianship undergraduate programs, and one year graduate diploma and longer taught Masters programs. Several courses are available in distance education mode. Also available are research Masters and doctoral programs and the number of students taking research degrees is increasing. However as most students study part time progress is slow. The schools/departments are associated with larger academic groupings: computing, business and communications are favoured, and one is associated with an education school. Now Australian LIS is beginning to meet the conditions that Whitley(1974, p.72) laid down for a high degree of social institutionalization.

Cognitive features

Because of the social situation of education for librarianship in Australia, research has mainly been of an applied nature. As shown in the content analysis of the articles in the two main Australian librarianship journals for the period 1985 to 1994 (Rochester 1995) only 24% of articles could be classified as research ones, with most research being professionally oriented. Of these reearch articles 40% were concerned with research on library and information service activities, 20% with information seeking and 14% with library history. The main research strategies used were the survey method by 44% of the researchers and historical method by 14%.

The characteristics of the authors of the research articles for the 10 year period of the journals were then examined (Rochester 1997) to see whether they wrote alone, or collaborated with others, etc. There were 76% single authored papers; it was concluded that "most LIS researchers in Australia seem to be working in lonely isolation." There are few large research grants available, so there are few research collaborations. The research methods used are those that can be employed in non funded research.


The comparison has shown a remarkable variation of emphases and trends in research in the countries examined. Each has its own research profile, which does not follow very closely the international trends. Despite the differences there are similarities. A strong interest in LIS services was typical of the research in Australia, Turkey and the UK. Research trends in Finland and the UK reflected most closely the research profile in international core journals.

We can seek explanation for national differencies in the cultural differences of these countries. However, we were able to show only for the Scandinavian and Australian cases how the social institutionalization of LIS has an effect on research output. A developed social structure of the discipline has a positive impact on the quality and quantity of research output.