Safety, Luck, and Gettierization

Sandy Goldberg gave an excellent talk yesterday. I’m looking forward to seeing some version of this coming out as a fully-fledged paper. Sandy is of course referring to one of the most famous papers in post-War philosophy.

Abstract. According to the widely-endorsed safety condition on knowledge, S knows that p only if: not easily would S believe that p were p false. It is natural to think that the safety condition on knowledge – or perhaps some version of it, relativized to grounds or methods – can be used to diagnose failures of knowledge that obtain owing to epistemic luck or gettierization. In this paper I argue against this natural view: there can be cases of justified, true, safely-formed belief that nevertheless suffer from a form of epistemic luck (and so fail to be knowledge). As a result, I argue, we would do well to reconsider the anti-luck condition on knowledge. In particular, we should think of the satisfaction of the modal component of this condition as something that must be explained by the subject’s epistemic access to the fact that makes her belief true. When we do so in the context of perceptual cases, as I do here, we see that the causal condition on perceptual knowledge, which everyone will acknowledge, has a non-negligible role to play as part of the anti-luck condition on such knowledge.

Here are some photos of the wonderfully irrepressible Sandy in action. Also, check out his latest book Relying on Others, a work I’ve mentioned before.