CLASP
The Centre for Linguistic Theory and Studies in Probability

Informativity in cooperative communication

Abstract In studying meaning in communication, a question arises as to which meanings are favored by interlocutors. While a range of candidate meanings may be possible and even plausible, how do speakers select which meanings to communicate and how do listeners make guesses as to the most probable meaning when trying to recover what a speaker intends or when anticipating what a speaker will say next?

In this talk, I compare two hypotheses for ranking candidate meanings that a speaker might contribute to a discourse. Under one account, listeners’ guesses simply reflect the probability that different meanings hold true: Speakers are taken to generate sentences that describe the world they see and listeners come to expect sentences about the typical situations speakers find themselves in. A second account combines this component with a component capturing the likelihood that a speaker, knowing some meaning to be true, would select that meaning as one worth conveying to a listener in an utterance. I present a series of psycholinguistic studies measuring listeners’ awareness of speakers’ production likelihoods. For example, although bananas are prototypically yellow, speakers rarely mention this yellowness in their utterances. In an eye-tracking study measuring anticipatory looking, listeners who hear a speaker use an ambiguous color adjective are found to anticipate subsequent mention of an object for which that color is less typical in the real world. Similarly, in a study on comprehenders’ guesses of what a speaker will say next, participants are shown to disprefer upcoming material that describes situation-typical outcomes. Further studies target properties of the speaker and show that the more aware comprehenders are of the speaker as an intentional knowledgeable communicator, the more informative they expect the speaker’s contribution to be and the more inferences they draw from the speaker’s content selection. This work raises questions about the ways in which learners (humans, machines) acquire information about the world if their linguistic input favors the description of real-world-atypical content.

The findings highlight the importance of establishing not only which meanings are possible and how they are derived, but also which meanings are probable as likely contributions to coherent discourse, despite – or perhaps as a result of – denoting non-typical situations.