Resolving Quantity- and Informativeness-Implicature in Indefinite Reference

Abstract

A central challenge for all theories of conversational implicature (Grice, 1957, 1975) is characterizing the fundamental tension between Quantity (Q) implicature, in which utterance meaning is refined through exclusion of the meanings of alternative utterances, and Informativeness (I) implicature, in which utterance meaning is refined by strengthening to the prototypical case (Atlas & Levinson, 1981; Levinson, 2000). Here we report a large-scale experimental investigation of Q-I resolution in cases of semantically underspecified indefinite reference. We found strong support for five predictions, strengthening the case for recent rational speaker models of conversational implicature (Frank & Goodman, 2012; Degen, Franke, & J¨ ager, 2013): interpretational preferences were affected by (i) subjective prior probabilities (Informativeness), (ii) the polarity and (iii) the magnitude of utterance cost differentials (Quantity), (iv) the felicity conditions of indefinite NPs in English, and (v) the ‘relatability’ of X and Y.