Williamson on Knowledge and Evidence

It is difficult to reconcile Williamson's rejection of 'decompositional' analyses with his positive proposals for an 'account' of knowledge. After arguing that epistemologists should remain interested in sufficiency conditions, this chapter challenges the sufficiency of Williamson's safety-based account. Examination of cases suggests that the safety-based account is probably inferior to two of its reliabilist 'cousins': the relevant-alternatives approach and the reliable-process approach. The remainder of the chapter offers reasons to doubt Williamson's 'evidence equals knowledge' thesis. Interpreting evidence as noninferential propositional justifiedness is at least as promising as interpreting evidence as knowledge. The case of the diffident doxastic agent - whose reluctance to believe is compatible with an abundance of evidence - helps support this claim.