The Composition of Meanings

Abstract

The title of this paper is intended to suggest a pair of related probe lems: first, the issue of how the meanings of sentences are built out of the meanings of their constituent words; and second, the issue of what sort of stuff meaning is, how it is created, which of an expression's underlying properties give it the particular meaning it has. My focus here is going to be on the connection between these two issues-on the question, roughly speaking, of what meanings must be like in order to be compositional. Specifically: what constraint is placed on an account of the underlying nature of meaning properties by acknowledging that the meanings of complex expressions are typically engendered by the meanings of their parts? My aim will be to answer this question, first, by giving a general, deflationary account of what it is to understand a sentence, second, by showing how this account yields an exceedingly simple explanation of how such understanding arises; and third, by concluding that the compositionality of meaning imposes no constraint at all on how the meaning properties of words are constituted. Let me start with an example. Presumably our understanding of the sentence 'dogs bark' arises somehow from our understanding of its components and our appreciation of how they are combined. That is to say, 'dogs bark' somehow gets its meaning (or, at least, one of its meanings) from the meanings of the two words 'dog' and 'bark', from the meaning of the generalization schema 'ns v', and from the fact that the sentence results from placing those words