Abstract

Michael Dummett, Neil Tennant, and Crispin Wright all appeal to a priori restrictions on a philosophical explanation of linguistic competence to criticize classical logic and semantics. They also use these restrictions to defend verificationsim. In the negative part of my project I uncover interesting structural analogies between the Dummettian arguments for logical revision to show that they all require the truth of a stronger, and less plausible, form of verificationism than even the logical positivists were willing to countenance. This result, I argue, genuinely is a modus tollens from which we should conclude that the explanatory demands Dummett and others place on the theory of meaning are mistaken. I then draw from the linguistic, lexicographic, and psychological literature to undermine the Dummettian view that an explanation of competence should recursively correlate dispositions necessary and sufficient for grasp of meaning with the aspects of meanings generated by a compositional semantics. Thus is eliminated the Dummettian impetus to identify such dispositions with the ability to recognize verifications. ;In the positive discussion I: characterize and defend a conception of tacit knowledge which renders the postulation of the mental reality of classical semantics both explanatory and plausible, show that many traditional philosophical questions concerning the epistemology of language can and should be recast as questions concerning how the theory of sentence meaning interacts with the theory of word meaning , and defend a roughly Davidsonian alternative to Dummett's theory of grasp of meaning.