Use-conditional meaning in Rational Speech Act models

Abstract

Motivated by shortcomings in the modeling of repeated utterances in Burnett’s Rational Speech Act (RSA) model of social meaning, we propose an alternative analysis in terms of use-conditional meaning, which we implement in an extended RSA framework. We show that it can not only model the production of a sequence of utterances throughout a discourse, but also captures the similarities between social and expressive meaning.