From translations to problematic networks: An introduction to co-word analysis

Empirical studies of the sociology of science have called into question the distinction between internal & external influences on the development of science. An alternative basis for the distinction is sought through the concept of translation. The function of the language of scientific writings is not only to convey information, but also to convince the reader that interest in its subject is appropriate. The analysis by John Law of a specific scientific article ("Enrolement et contre-enrolement: les luttes pour la publication d'un article scientifique," see abstract, this section) illustrates this process. The definition of a problem involves linking together words that represent previously existing networks of interest into a common reality that unites these networks. The coword analysis methodology generates models of these networks, which are illustrated for a sample of 172 articles indexed by 1 of 3 key words in a datafile. The identification of problematic networks through this method also identifies key actors in the formation of networks. A concept of internal/external can exist for any specific network without being defined for the sciences as a whole. 1 Table, 3 Figures, 69 References. W. H. Stoddard.