Abstract: This article explains and motivates an account of one way in which we might have substantive a priori knowledge in one important class of domains: domains in which the central concepts are response-dependent. The central example will be our knowledge of the connection between something's being harmful and the fact that it is irrational for us to fail to be averse to that thing. The idea is that although the relevant responses (basic aversion in the case of harm, and (...) a kind of interpretive failure in the case of irrationality) are produced by independent psychological mechanisms, they have distal causes that turn out to be related in ways that—once language enters the picture—yield epistemically accessible necessary connections between the referents of their corresponding terms. (shrink)

The author rejects Leon Galis's claim ( Inquiry, Vol. II, No. 2) that in 'Of Words and Tools' ( Inquiry, Vol. 10, No. 2) he attacks a form of the 'use' theory of meaning that no one has held. Galis's other claim, that the author criticizes a needlessly weak form of the theory, is found to be justified, but the author argues that his procedure was adequate, and parallel to that oi Galis's own reconstruction of the 'use' theory in terms (...) of 'goal-directed action'. Difficulties in this reconstruction are pointed out, and some meta-semantic issues about theories of meaning raised. (shrink)