Fallibilists Can Deal with Concessive Knowledge Attributions

Fallibilism is typically taken to face a problem from the apparent infelicity of concessive knowledge attributions (hereafter, CKAs). CKAs are of the form: “S knows that p, but it’s possible that q”, where q obviously entails not-p. CKAs sound to the ears of many philosophers as contradictory or infelicitous. CKAs, though, appear to be statements that fallibilists are committed to being true if fallibilism is true, since if S fallibly knows that p, then she cannot properly rule out some possibility in which not-p. Do fallibilists, then, have some way of explaining the seeming infelicity of CKAs that does not impugn the truth of fallibilism? Fallibilists think so. In this connection, there are two well-known responses to the problem in the philosophical literature: Patrick Rysiew’s pragmatic strategy and Jason Stanley’s semantic strategy. While both strategies have real virtues, there are aspects of each strategy that face objections. In this paper, I outline those objections and develop some remedies for them. This paper will show that the challenge posed by CKAs is not a grave problem for fallibilism at all. In particular, I argue that either the semantic strategy succeeds or the pragmatic strategy succeeds. So, if CKAs really are statements that fallibilists are committed to being true if fallibilism is true, then the pragmatic strategy succeeds, but if the pragmatic strategy fails because CKAs are obviously false or are not statements that fallibilists are committed to being true if fallibilism is true, then the semantic strategy succeeds. Thus, the problem for fallibilism posed by CKAs is a problem that fallibilists are in a position to resolve.