Comments on MacFarlane∗

The natural answer to this question, MacFarlane takes it, is “yes”. I don’t propose to harp on the question of whether MacFarlane has the data right. Let us just assume, for the sake of argument, that he does. Let us further assume that his interpretation of the data is correct—i.e., that these judgments are assessments of the the whole clause and not simply of the prejacent. Granting all this—maybe a lot—we need a semantics for epistemic modals that will make sense of the judgments in this case, and in relevantly similar cases. MacFarlane argues that contextualism about epistemic modals cannot make sense of the judgments. His central worry is that it can only get the truth-value judgments of speakers right by making the truth-conditions of epistemic modal claims outrageously strong—too strong to be assertable in cases where they are, in fact, assertable. We might call it the contextualist’s dilemma: either our semantics systematically fails to capture the truth-value judgments that people actually make, or it captures these judgments but turns users of epistemic modal sentences into irrational asserters. Let us suppose, again for the sake of argument, that this diagnosis of the situation is correct. How, then, does MacFarlane propose to escape the contextualist’s dilemma and make sense of the data? ∗Comments on John MacFarlane’s ‘Epistemic Modals are Assessment-Sensitive’ (manuscript, August 23rd, 2006), for the 2006 University of Michigan Workshop in Philosophy and Linguistics. Unless otherwise noted, page references are to this paper.