I advocate a relativist semantics for epistemic modal claims such as “the treasure might be under the palm tree”, according to which such utterances determine a truth value relative to something finer-grained than just a world (or a pair). Others have argued for relativist semantics in other areas. Anyone who is inclined to relativise truth to more than just worlds and times faces a problem about assertion. It’s easy to be puzzled about just what purpose would be served by assertions of this kind, and how to understand what we’d be up to in our use of sentences like “the treasure might be under the palm tree”, if they have such peculiar truth conditions.
In what follows I will first present an example of the kind of case that motivates relativism about epistemic modals. (I’ll be talking about ‘might’, but nothing much hangs on this choice of examples. In fact, the intuitions that I’m appealing to are probably stronger for ‘probably’. So if you think I might be wrong about ‘might’, you’ll probably be happier to go along if you think about the parallel argument for ‘probably’ instead.) I’ll then sketch a relativist theory in a bit of detail. I’ll then show why there is a problem, given such a theory, about the role of epistemic modals in assertion and communication, and set out to solve it. Solving this problem will be helpful in several ways: not only does it eliminate an apparently forceful objection to relativism, but the account of the role of such claims in assertion and communication helps to make clear just what the relativist position is, exactly, and why it’s interesting.