Key points are not available for this paper at this time.
Lewis (1976, 1986b) concentrates on material conditionals, arguing that their semantic values are propositional but that their pragmatic values are given by conditional subjective probabilities. Thus, from the start, the general theses (1) were linked to close linguistic analysis involving both semantic and pragmatic considerations. One comes away from Lewis’s papers with the sense that they could form the cornerstone for a successful (probabilistic) formal pragmatic theory. The present paper investigates the pragmatics of evidential sentences in the general Lewisian terms (1). Now, evidential morphemes are rich and varied in their contributions, they might not form a natural class cross-linguistically (Matthewson et al. 2007:3, Speas 2007), and analyzing even a single morpheme generally proves complex (Davis et al. 2007, Fasola 2007). Thus, we do not, in this short paper, venture a comprehensive analysis. Rather, we focus on the pragmatic strategy that evidentials facilitate. Section 2 discusses the pragmatics of evidentiality. Section 3 carves out a role for probabilities in pragmatic theory, and Section 4 describes our analysis. We close the paper with a discussion of the connections between evidentiality and modality (Section 5) and a brief look at various promising extensions of these ideas (Section 6).
Davis et al. (Wed,) studied this question.