Context-dependency and comparative adjectives

Abstract

might be true when said of a professional chess player in a context where the subject is being compared to all adult chess players, casual and non casual, but false in a context where the same player is being compared to other chess professionals. A natural way to account for the context dependency of (1) is to treat 'impressive' in (1) on the model of simple indexicals. On this account, the context-dependency of (1) is not so much the result of hidden semantic or syntactic structure but rather due to the fact that both 'He' and 'impressive' can take on different semantic values in different contexts depending on who is being talked about and on what standard of impressiveness is being deployed. On a simple version of this story, and ignoring intensional issues, 'impressive' picks out in context a function from objects to truth values. In a high chess standards context, for example, the function associated with 'impressive' delivers truth only for individuals who are good at chess by comparison with professional chess players. On a slightly less simple approach (and again ignoring intensional issues), 'impressive' denotes a function from sets to sets. In a high standards context, the function picked out by 'impressive' takes as argument the extension of 'chess player' and delivers as value the set of people that are impressive by professional standards. Call such approaches 'indexical approaches'. (I shall not be examining their relative merits here.1,2) In a series of thoughtful and important papers, Jason Stanley has advo cated an alternative semantic approach to comparative adjectives, one that attempts to assimilate the context-dependency of interest in (1) to the kind