Abstract

If two successive theories are semantically incommensurable, we have no way to make a complete comparison of their contents. If so, we have no way to verify that the highly confirmed content of the successor is greater than that of its predecessor, and we cannot verify that scientific knowledge has accumulated across the theory change. Thus, incommensurability creates a problem for the justification of the standard cumulative conception of scientific progress. ;To resolve this problem, I distinguish irresolvable strong in commensurability from weak incommensurability, which is resolvable. I argue that Kuhn's arguments, insofar as they are sound, support only the latter. Cumulative progress is therefore not only possible, but in principle justifiable. Nonetheless, I support most of Kuhn's claims about the incommensurability of successive paradigms. ;My argument for weak incommensurability depends on an interpretation of scientific theories which makes the way a theory is understood an integral part of the theory. Both syntactic and semantic approaches to theories fail to deal with the incommensurability problem because they ignore this pragmatic aspect. I offer a context-dependent semantics based on contemporary pragmatics which can both represent the incommensurability problem and show how it can be resolved