ABSTRACT
In this paper I first present Kripke's thesis regarding the possibility of acquiring a priori knowledge of contingent truths and Keith Donnellan's criticism of this thesis. Second, I explore a distinction that Donnellan makes between (a) knowing that a sentence expresses a truth and (b) knowing what truth this sentence expresses. I argue that this distinction is not relevant only in the context of his criticism of the contingent a priori, but also to our practices with proper names in general. I try to show that knowing the meaning of proper names cannot be reduced to our linguistic competence with them, but that it depends on an acquaintance relation to their bearers. If this is true, then the thesis of the contingent a priori, as formulated by Kripke, cannot be correct.
Keywords:
Contingent a priori; Reference; Proper names; Belief de re