Crucial to Peter Achinstein's philosophy of science, as presented in Concepts of Science, is the concept of semantic relevance. First, the concept of semantic relevance is central to his analysis of definition and it is presupposed both in his analysis of the concepts of theory and model and in his critique of alternative analyses. Secondly, Achinstein's way of doing philosophy of science rests heavily on his analysis of semantic relevance. Philosophical analysis for Achinstein seems primarily to consist in specifying the use of certain terms. As we shall see, use and semantic relevance are, according to Achinstein, closely connected.

Yet, despite the centrality of this concept in Achinstein's philosophy, I will argue that the concept has serious philosophical problems and that the use Achinstein makes of the concept is dubious. First, I will argue that the notion of semantic relevance inherits some of the traditional problems of analyticity as well as having problems of its own. In particular, I will show that given Achinstein's construal of semantic relevance it is extremely difficult if not impossible to determine the semantic relevance of a term. Secondly, I will maintain that given these problems, Achinstein is not in any way justified in using the notion of semantic relevance in his criticism of other philosophers' views and in the construction of his own.