One approach to mapping the thematic structure of science is to utilize citation relations between publications in order to create publication clusters which are then classified into thematic groups. In this case, every citation relation in the publications’ citation network is treated equally and contributes in the same way to the signal which is taken to indicate thematic relatedness.
We have known for a long time, however, that citations serve all kinds of functions, and that we can expect different citations to express different degrees of thematic relatedness between the publications. Thus, this “lumping” of citations will create “aggregate” clusters which will be difficult to interpret.
In order to examine this problem, we present a categorization of citations which captures different dimensions of publications’ topical relatedness, such as the research methods used, the empirical objects studied, the theoretical resources built upon, the research questions pursued, or the external motivation for and relevance of the work. By distinguishing the ‘epistemic functions’ that citations have when relating a new knowledge claim to the existing literature, science mapping could improve the interpretability of science maps and inform new approaches for creating a more differentiated picture of thematic groupings of science. An important question that this kind of thinking raises is what kinds of citations are most relevant for identifying specific thematic structures such as, for example, research specialties, that borrow knowledge from other fields in order to produce new knowledge claims in their own domain.
This presentation shows results of an empirical case study where, first, the core and the boundary of the specialty of invasion biology have been determined. Then, selected core publications were manually analyzed for the epistemic functions that underlie the in-text citations. And, finally, the relatedness of the cited sources with the specialty was analyzed. Results indicate that (i) even among similar types of invasion biological studies (e.g. empirical studies), the dominant epistemic functions of citations vary remarkably; (ii) even core invasion biology publications “borrow” knowledge extensively from outside the field; (iii) publications belonging to the core of a specialty might not be detectable solely using the citation signal.
Theresa Velden (The German Centre for Higher Education Research and Science Studies - DZHW)
Matthias Held (Technical University Berlin)