Relativism, Vagueness and What is Said

In a series of recent papers, John MacFarlane (2003, 2005a, 2005b) has formulated a version of truth-relativism, and argued for its application in some cases—future contingents, knowledge attributions and epistemic modals among them. Mark Richard (2004) also defends a version of relativism, which he applies to vagueness-inducing features of the semantics of gradable adjectives; Richard thinks that it provides a model to properly articulate contextualist intuitions about knowledge attributions. On MacFarlane’s characterization, and on Richard’s as I will be interpreting it, truth-relativist claims posit a distinctive kind of context-dependence, to be distinguished from the more familiar one that has been closely studied since Kaplan’s (1989) work in the 1970s. Radical truth-relativists posit the dependence of the evaluation of an assertion as true or otherwise on aspects of the context of the evaluation itself—in contrast with the context of the assertion. The main motivation for truth-relativism is to make sense of alleged cases of ‘‘faultless disagreement’’. These are cases of ‘‘substantive disagreement’’ (Richard, op. cit., 219) between two thinkers, ‘‘in the sense that one asserts something that is inconsistent with what the other asserts’’ (ibid.: 218), while, nonetheless, neither of them is at fault, with respect to the fundamental norms to evaluate acts of assertion such as those they are engaged in, and in some unexpected way: it cannot just be that they are talking about different subject-matters, without realizing it, so that in fact there is no real inconsistency; or that both assertions are equally well justified, given the evidence available to the asserters, although at