In this paper, I show that, given seemingly plausible assumptions about the epistemic‘might’ and conditionals, we cannot explain why in some circumstances it is appropriate to utterconditional ‘might’-sentences, like “If Angelica has crumbs in her pocket, then she might be thethief” and not the corresponding simple ones, like “Angelica might be the thief.” So, one of ourassumptions must be incorrect. I argue that the root of the problem is an umbrella thesis aboutthe pragmatics of ‘might’-communication – one that says that the communicative impact of anutterance of a ‘might’-sentence is the performance of a consistency check on the information ofthe context. I conclude that we must reject this thesis. And I close the paper by sketching analternative view about what assertive uses of ‘might’-sentences typically do – one which avoidsthe problem. Such uses typically present a possibility as a serious option in reasoning anddeliberation.


Philosophy | Philosophy of Mind