Abstract
In this paper, I re-examine Dummett's arguments against realism and, most notably, those against truth conditional semantics. Dummett claims that a (realist) truth conditional meaning theory will invariably encounter limitations when accounting for the meanings of the statements of the so-called "disputed class", and so must be rejected in favour of a theory of meaning couched in terms of proof or verification. The first part of this paper seeks to faithfully reconstruct Dummett's position on meaning. The subsequent parts are critical of this position. In particular, I attempt to expose the shortcomings of an antirealist theory of meaning on the one hand, and then show how truth conditional semantics (be they realist or not) emerge relatively unscathed from Dummett's criticism on the other.