Formal semantics and intentional states

Abstract

My aim in this note is to address the question of how a context ofutterance can figure within a formal, specifically truth-conditional, seman-tic theory. In particular, I want to explore whether a formal semantictheory could, or should, take the intentional states of a speaker to berelevant in determining the literal meaning of an uttered sentence. Theanswer I’m going to suggest, contrary to the position of many contempo-rary formal theorists, is negative. The structure of this note is then asfollows: first, I’ll very briefly sketch three distinct forms of semantic theory.One, ‘strong formal semantics’, will be seen to be immediately problem-atic, leaving us with two other options: use-based theories and what I’llterm ‘moderate formal semantics’. If we opt for the latter position, thequestion arises of what kinds of appeals to a context of utterance arelegitimate given a formal outlook. I’ll suggest that this question arises intwo distinct ways and explore the moderate formal semanticist’s positionin regard to both. However, the conclusion I will reach is that what ischaracteristic of formal semantics is that it makes only the most minimalsemantic concessions to context.To begin with, if we are thinking about formal approaches to linguisticmeaning, it seems that we might adopt what I’ll call ‘strong formalsemantics’. On this kind of account the right point at which to offersemantic analyses is that of formally described sentence-types. Thus literalmeaning is held to be entirely context-invariant – a sentence, individuatedin terms of its syntax, possesses the very same meaning no matter when,where, or by whom it is produced. Of course, what is communicated bythe utterance of a given sentence may change, but the thought is that, asfar as