A Co-Citation Model of a Scientific Specialty: A Longitudinal Study of Collagen Research

The phenomenon of specialization in science is receiving increasing attention as it becomes clear that the 'specialty' is the principal mode of social and cognitive organization in modern science. Recently, there have been some attempts to formulate theories of specialty growth and change.' Before sufficient evidence can accumulate to confirm or refute these theories, it seems likely that systematic and consistent methodological frameworks must be developed. The specialty of collagen research is presented here as the vehicle for exploring the possibilities of one such methodology. Collagen research was selected because its recent history illustrates what is possibly an important type of specialty change -- namely, that of rapid shift in research focus, or what might be loosely termed 'revolutionary' change. In this paper I outline a general method for the study of specialty change, based upon citation data. The point of departure is a system which clusters highly cited documents, using their frequency of cocitation as a measure of association.2 This system uses the magnetic tape version of the Science Citation Index (SCI) as its data base. The clusters obtained from the SCI have been found to correspond to narrow subject matter groupings and it has been suggested that these reflect the social and cognitive structures of research specialties. The collagen specialty was selected from roughly 1,000 clusters identified by applying this algorithm to the SCI.3