The Interaction of Cognitive Structures and Cooperation Behavior in Scientific Communities

by Peter Mutschke
Social Science Information Center (IZ)
Bonn, Germany


Scientific communities have an underlying cognitive structure that changes over time. The changes reflect the popularity of specific themes or research topics. Researchers (e.g. Heintz 1993) have argued that changes in the popularity of themes are related to characteristics of the communication context, particularly the social structure of the community. Those characteristics include the social positions of actors as well as their roles. Subsequently, the relationship between cognitive and social structures in scientific communities has been investigated in several scientometric studies.

A study conducted by Renner (Renner 1995, Mutschke & Renner 1995) looks into the question how characteristics of the communication structure of a scientific field influences the development of themes: Why do certain subjects rise to mainstream themes of a research field (and others not), and which constellations are the basis for such processes? Renner examined the careers of topics in different social science research fields with the help of a cognitive mapping procedure called Leximappe (Callon et al. 1983, see also Grivel et al. 1995). The surprising result from this study is that over time cyclic process forms can be observed concerning the career of themes: After increasing centrality (which indicates increasing interest in the topic) a rising density of the topic follows. Finally the theme promotes to a mainstream theme of the research field (characterized by high density and high centrality). Diminishing interest corresponds then with decreasing centrality but with still high density indicating that the theme moves (in tendency) into the area of "ivory tower" research (Renner 1995, Mutschke & Renner 1995, Renner 1997).

In a further investigation Renner (1997) examined whether there is a connection between the social coherence of a scientific community and its innovativeness. By the example of the career of innovative topics in sociostructurally different research fields, it could be proven that innovativeness (indicated by the community's consensus concerning the relevance of a new idea) arises rather not in the center, but in the periphery of a research field. A comparison of the "attraction curves" of the cluster 'violence against children' in the period from 1980 to 1992 turned out that within the context of the sociostructurally fragmented research field 'violence' the topic gained quite rapidly in attractiveness (high centrality, high density), whereas in the sociostructurally relatively coherent 'sociology of family' the same topic appeared, in contrast, much later and had only a much flatter career cycle. This contradicts the generally assumed thesis that new ideas are accepted more easily in socially coherent groups than in networks whose members have only weak ties to each other (Coleman 1990, Rogers 1977, Valente 1995), and, moreover, are first taken over by persons who are particularly well integrated into the network (opinion leader model).

As a reason for this phenomenon the sociostructurally caused topic orientation of the actors is seen (Renner 1997): Highly integrated actors are more involved in research-field-internal consensus finding processes concerning the specific contents of the main topics of the research field. Therefore those scientists tend rather towards thematic conformity: There is an internal orientation of central actors. In contrary, scientists who have peripheral positions in researcher networks are more involved in external discussions, i.e. other research fields. This situation predestines them to the early transfer of new ideas from one research field to another. This is considered as an external orientation of marginal actors. Therefore, in contrast to the opinion leader model, a coherent system of mutual influence globally seems to hinder the acceptance of innovative ideas.

The interaction processes which cause this phenomenon are obviously dependent on sociostructural factors, e.g. the position of the actors within the communication context. In a study of Mutschke and Renner (1995) using the tools Leximappe (topic development) and Akcess (development of networking) (Mutschke 1994) a strong statistical interaction between the centrality of persons and the centrality of themes could actually be proven. The study could show, for the research field 'youth and violence', that the inclination to select attractive themes (topics of high centrality and density) seems to rise with membership and central social status in large networks.

These studies indicate that the choice of subjects and cooperation partners seems to be strongly influenced by the social coherence of a research field and the degree of integration of the involved actors into cooperation networks. However, although the relationship between cognitive and social structures has been a central question in these studies, it has been difficult to determine how these two aspects are related. On the background of the foundations described above particularly the question arises whether the phenomenon of cyclic topic careers corresponds with increasing social networking of the involved actors. Therefore, the paper addresses two central issues: (1) Is there a linear relationship between the centrality/density of themes and the social status, especially the centrality of actors in researcher networks? (2) Who are the innovative actors who push the carrier of new innovative themes and, accompanied by this, possibly cause the evolution of cooperation networks in scientific communities? For this purpose between four actor categories (single fighters, free riders, social climbers, experts) are distinguished according to their centrality in cooperation networks.
Due to these issues the interaction of cognitive structures and social networks is investigated for two social science research fields (sociology of youth, and women) over a ten-year-period (1989-1998). The datasets are obtained from the German social science database SOLIS. For the analysis the tools Leximappe (cognitive mapping) and Akcess (social networking) are used.

The paper will demonstrate that that all actors, independent from their social status in a scientific community, generally tend to be engaged in central themes. However, the rate of involvement in central themes does not correspond in a linear manner with increasing centrality of the actors in cooperation networks: Apart from the single fighters, social climbers correlate much stronger with central themes than free riders and experts. This confirms Renner's hypothesis that new innovative topics are rather not pushed by very central actors. However, Renner's reversal conclusion that innovations are rather introduced by actors who are peripheral in networks could not be confirmed: free riders correlate clearly much weaker to the centrality of themes than social climbers.

This indicates that innovations are rather to be expected from actors who are neither located in the (structural) periphery of a research field nor in its (structural) center, but in its structural midfield where the best chances to rise are to be suspected. Those actors have a strong tendency to push ahead new innovative ideas and to establish them as mainstream topics of the society. This characteristic "elevate" them to interesting cooperation partners for the central actors (experts) who put particular emphasis on the consolidation of the mainstream. This hypotheses corresponds with the cooperation behavior of experts: The more central a theme, the more experts cooperate with social climbers, i.e. actors who are (already) engaged in themes with increasing centrality.

The observed orientation patterns concerning topic and partner selection in social science research fields suggest that its social climbers play an important role regarding the evolution of cognitive and social structures. Both their innovativeness as well as their social relevance seems to give the development of networks and themes in scientific communities the strongest impetus.


Callon, M.; J.-P. Courtial, W.A. Turner, and S. Bauin (1983): From translation to problematic networks: an introduction to co-word analysis, Social Science Information, 22, 191-235.
Coleman, J.S. (1990): Foundations of Social Theory. Harvard University Press, Boston.
Grivel, L., P. Mutschke, and X. Polanco (1995): Thematic Mapping on Bibliographic Databases by Cluster Analysis: A Description of the SDOC Environment with SOLIS, Knowledge Organisation, 22(2), 70-77.
Heintz, B. (1993). Wissenschaft im Kontext: Neuere Entwicklungstendenzen in der Wissenschaftssoziologie. Kölner Zeitschrift für Soziologie und Sozialpsychologie, 45 (3), 528-552.
Mutschke, P. (1994): Processing Scientific Networks in Bibliographic Databases, in: H.H. Bock et al. (Eds.): Information Systems and Data Analysis. Prospects-Foundations-Applications. Proceedings 17th Annual Conference of the GfKl 1993. Springer-Verlag, Heidelberg-Berlin, 127-133.
Mutschke, P., and I. Renner (1995): Akteure und Themen im Gewaltdiskurs. Eine Strukturanalyse der Forschungslandschaft, in: E. Mochmann et al. (Eds.): Gewalt in Deutschland. Soziale Befunde und Deutungslinien. Oldenbourg Verlag, München, 147-192.
Renner, I. (1995): Gewalt in der Gesellschaft: zur sozialwissenschaftlichen Konstruktion des Phänomens. Informationszentrum Sozialwissenschaften. Jahresbericht 1994. Bonn.
Renner, I. (1997): Soziale Kohärenz und Innovativität. Struktureffekte zur Akzeptanz neuer Themen in sozialwissenschaftlichen Forschungsfeldern, Kölner Zeitschrift für Soziologie und Sozialpsychologie, 49(1), 74-97.
Rogers, E.M. (1977): Network Analysis of the Diffusion of Innovations, in: D. Lerner et al. (Eds.): Communication Research - A Half Century Appraisal. EastWest Center Honolulu, 137-164.
Valente, Th.W. (1995): Network Models of the Diffusion of Innovations. Cresskill/New Jersey: Hampton Press.

Back to Papers ListBack to Papers List