Speaker's meaning: An essay in the philosophy of language.

摘要

The topic of this dissertation is how to understand the notion of speaker's meaning, and the line I pursue is Gricean. My thesis is that speaker's meaning consists of a specific type of nested audience-directed intentions, which I call M-intentions. I first introduce the Gricean distinction between natural and non-natural meaning, where speaker's meaning belongs to the latter category. Natural meaning is factive and agent-independent, while non-natural meaning is non-factive and agent-dependent. Building on this distinction, I argue that in order to understand what communication amounts to, we ought to distinguish between communication as indication, communication as influencing and communication as overt communication. The first coincides with natural meaning, the last with non-natural meaning (and is thus closely connected to speaker's meaning), while communication as influencing is an intermediary type of communication in-between the two. These categories make better sense of the empirical data on communicative interaction. In particular, they give us a handle on how to understand children's communicative output at various developmental stages, and autistic speakers. Also, building on the distinction between natural and non-natural meaning, I reintroduce the pragmatic phenomenon of double bind, and show how it should be understood as a clash between natural and non-natural meaning. In relation to the discussion of double bind, I suggest that humans have an innate trust in certain natural meaning signifiers and that this trust facilitates, and is a requirement for, humans to become trusting in non-natural meaning systems like natural language. The latter is a prerequisite for M-intentions to succeed. To M-intend is to intend to produce some particular response in one's audience, to intend the audience to recognize that intention, and to intend that that recognition is part of the reason for why the audience responds as intended. I address a variety of counterexamples and critiques to this way of understanding speaker's meaning. The M-intentions, I argue, are not only something speakers have when they address others, they also constitute a norm of speaker's meaning in the sense that an audience will assume that speakers M-intend when they address others.