Vague Utterances in Context

Take an utterance of (1):

(1) John is tall.

(1) is both context-sensitive and vague. What counts as tall depends on what kinds of things we are comparing it to, and what the standards of tallness that are useful in our circumstances are. Both of these dimensions of variation exhibit vagueness. But, we can perfectly well communicate with (1) even if we don’t know how tall exactly is tall or how tall John is. We argue for a theory according to which the contextual resolution of words like “tall” is determined by the mechanisms of discourse coherence that specify the linguistic relations the utterance containing it bears to the prior discourse and the real world situation it is embedded in. Our account allows us to capture how (1) can be used to make a useful distinction among the relevant class of individuals, as well as how it can be used to refine our understanding of what the standard for tallness is (cf. Barker 2002), while at the same time assigning a precise truth-conditional content to it. Moreover, we also capture that (1) exhibits vagueness: we argue that vagueness of (1) results from the fact that the agents typically have only incomplete information about the standards set by the mechanisms of discourse coherence.