Science as instrumentation. The case for psychiatric rating scales

The aim of this article is to test the model analysis conceived by Terry Shinn on the autonomy and unity of science. For him, the differentiation of sciences can be explained in a large part by the diffusion of generic instruments created by research-technologists moving in interstitial arenas between higher education, industry, statistics institutes or the military. We have applied this analysis to research on depression by making the hypothesis that psychiatric rating scales could have played a similar role in the development of this scientific field. To that purpose, we proceeded to a lexicographic study of keywords mentioned in articles listed by the PsycINFO© data base on this subject between 1950 and 2000. In order to realize an associated words analysis, we constructed a co-occurrence matrix and used clustering analysis based on a grouping index; that is, the equivalency index. We obtained significant aggregates of keywords associated with significant periods, or major moments, of the development of research on depression. This periodization confirmed the structural role played by psychiatric rating scales in the development of this scientific field, and led us to discuss and to extend some elements of the model initiated by Shinn.