Abstract

According to a prominent account of epistemic possibility endorsed by John Hawthorne and Jason Stanley (“H-S Account”), a proposition q is epistemically possible for a subject just in case what the subject knows doesn’t obviously entail not-q. I argue that H-S Account is false by its own lights by first showing that H-S Account entails a different account of epistemic possibility—q is epistemically possible for a subject just in case not-q is not obvious to that subject (“Obvious Account”)—and then showing that H-S Account is false on the basis of Obvious Account. Obvious Account is good news for fallibilists. H-S Account is in tension with fallibilism, which requires that fallibilist adherents of H-S Account do extra work to relieve the tension. Obvious Account, however, does not require any of this work; it is straightforwardly compatible with fallibilism. Obvious Account also has implications for the truth of concessive knowledge attributions (CKAs)—statements of the form: ‘I know p, but possibly q’, where q obviously entails not-p. Obvious Account allows some CKAs to be true, whereas H-S Account does not.