Shared Content as Speaker Meaning

Cappelen and Lepore (2005; 2006; 2007) have recently emphasised the significance of a minimal notion of perfectly shared content for pragmatic theories. This paper argues for a similar notion, but assumes that a satisfactory defence cannot be achieved along the lines of the existing debate between Minimalism and Contextualism (e.g. Carston 2002, Recanati 2004). Rather, it is necessary to consistently distinguish two functional domains: the subjective processing domain and the interpersonal domain of communication, each with its own kind of utterance meaning. I will argue that it is the mutually recognised content of the speaker's overt commitment that should be identified as ‘speaker meaning’. Diverging from the (post-) Gricean tradition, it is conventionally restricted (minimal), but genuinely pragmatic (speaker-dependent). Functional considerations show that it is, moreover, unnecessary to include further elements in ‘speaker meaning’. The distinction between two notions of utterance meaning with very different characteristics allows us to integrate the assumption of perfect sharing, which explains people's trust in communication, and the subjectivity of the hearer's inferences into a coherent and powerful model.