Epistemic contextualism defended

Epistemic contextualists think that the extension of the expression ‘knows’ (and its cognates) depends on and varies with the context of utterance. In the last 15 years or so this view has faced intense criticism. This paper focuses on two sorts of objections. The first are what I call the ‘linguistic objections’, which purport to show that the best available linguistic evidence suggests that ‘knows’ is not context-sensitive. The second is what I call the ‘disagreement problem’, which concerns the behaviour of ‘knows’ in disagreement reports. These may not be the only objections to epistemic contextualism, but they are probably the most influential. I argue that the best current epistemic contextualist response to the linguistic objection is incomplete, and I show how it can be supplemented to deal with the full range of linguistic objections. I also develop a new solution to the disagreement problem. The upshot is that neither sort of objection gives us any reason to reject epistemic contextualism. This conclusion is, in a sense, negative no new arguments for epistemic contextualism are advanced but it’s a vital step towards rehabilitating the view. 0. Introductory Remarks Epistemic contextualists henceforth, ‘contextualists’ think that the extension of the expression ‘knows’ (and its cognates) depends on and varies with the context of utterance. It used to be thought that contextualism offered us both an attractive solution to the sceptical problem (see DeRose 1995) and a nice explanation of linguistic usage, in particular the observation that the appropriateness of ascribing ‘knowledge’ depends on the context (see the ‘bank cases’ in DeRose 1992). These motivations are separate, but they are mutually reinforcing. The solution to the sceptical problem gains independent plausibility from the explanation of linguistic usage, and the explanation of linguistic usage gains philosophical interest from the application to scepticism. I say ‘used to be thought’ because the solution to scepticism has been criticised for merely sidestepping the problem (see Kornblith 2000 and Sosa 2000) and the explanation of linguistic usage has been attacked on all sides, with some questioning the data (see