Quantification and perspective in relativist semantics
Peter Lasersohn, UIUC
A common feature of recent relativist semantic analyses is that they treat sentences as expressing contents which are true or false not only relative to a possible world, but also to some other parameter such as an individual, perspective, epistemic state, or standard of taste. The use of such parameters raises interesting questions for the relation between syntax and semantics — questions which parallel, to some extent, the issues pertaining to the debate about “unarticulated constituents”. If a definite truth value can be assigned to a sentence content only if some particular value for one of these parameters is supplied, must that value be the denotation of some (perhaps hidden) syntactic element of the sentence? Or is it reasonable to treat such parameters purely as indices, with no syntactic correlates? Advocates and critics of relativism have both taken a wide variety of positions on such issues, from claims that a proper use of hidden arguments rather than indices supports a more conventional contextualist semantic theory over a relativist one; to claims that such theories are intertranslatable, and therefore equivalent; to claims that both non-standard indices and relativistically interpreted hidden arguments are necessary; to claims that these non-standard parameters function purely as indices, with no syntactic representation. In this paper I review the arguments for these various positions, attempt a clarification of the issues, and present semantic and syntactic evidence in favor of a relativist semantic analysis in which non-standard parameters are not syntactically represented. A relatively novel feature of this analysis is the use of “index binding” operators, which allow simultaneous modal and objectual quantification.
Quantification and perspective in relativist semantics