‘A logical theory may be tested by its capacity for dealing with puzzles, and it is a wholesome plan, in thinking about logic, to stock the mind with as many puzzles as possible, since these serve much the same purpose as is served by experiments in physical science.’ This paper is an attempt to follow Russell’s advice by using a puzzle about the contingent a priori to test and explore certain theories of reference and modality. No one could claim that (...) the puzzle is of any great philosophical importance by itself, but to understand it, one has to get clear about certain aspects of the theory of reference; and to solve it, one has to think a little more deeply than one is perhaps accustomed about what it means to say that a statement is contingent or necessary. (shrink)

I defend a version of Kripke's claim that the metaphysically necessary and the knowable a priori are independent. On my version, there are two independent families of modal notions, metaphysical and epistemic, neither stronger than the other. Metaphysical possibility is constrained by the laws of nature. Logical validity, I suggest, is best understood in terms of epistemic necessity.