Bayesian inference in dialogue
Abstract
A word is referentially ambiguous if it has several potential referents. Observing how listeners make choices among thosereferents can reveal their hidden beliefs and preferences, as well as reflect their reasoning strategies. We asked subjectsto observe how one of the objects is chosen following a possibly ambiguous utterance and to infer which preferences thelistener may have had in mind when choosing that particular object. In order to adjust this interaction to a dialogue-likesetting, we extended the traditional one-shot reference game to a round of 4-trial games. Moreover, we modeled theprocess within the Rational Speech Act framework, implementing iterative inference over multiple trials, where posteriorsfrom previous trials carry over to the next trial as priors. The model predicts human inference behavior better than abaseline uniform model, as well as better than a non-iterative model. The results imply that, in principle, humans areable to compute Bayesian-like inferences in dialogue, learning about the beliefs and preferences of others in a cumulativemanner.