Comments on Overfitting and Heuristics

In the second half of Overfitting and Heuristics in Philosophy, Timothy Williamson defends intensionalism, the view that necessarily equivalent propositions are identical, against objections in the area of the ‘problem of logical omniscience’. His defense relies on proposed psychological heuristics for making and interpreting attitude ascriptions and on the notion of having an attitude towards a proposition under a guise. These comments raise three questions about this apparatus. First, Williamson says almost nothing systematic about guises, even though they bear the weight of his account of cognitive achievement in mathematics and metaphysics. Second, he specifies no relationship between having an attitude under a guise and having it simpliciter, with the result that his theory makes no predictions about the truth of even ordinary attitude reports. Third, his objections to contextualism assume a connection between semantics and psychological processing that his own view appears to deny. In a final section, I argue that Williamson’s own commitments about guises and communication supply the materials for a contextualist semantics of attitude reports, that his reasons for rejecting contextualism do not withstand scrutiny, and that there are good reasons, given his other commitments, to endorse this form of contextualism.