Modelling atypicality inferences in pragmatic reasoning

Anonymous

Keywords: world knowledge, experimental pragmatics, Bayesian modeling, noisy channel

TL;DR: Atypicality inferences can be modelled using an RSA model which reasons about background knowledge and a noisy channel.

Abstract:

Empirical studies have demonstrated that when comprehenders are faced with informationally redundant utterances, they may make pragmatic inferences to accommodate the informationally redundant utterance (Kravtchenko and Demberg, 2015; Kravtchenko, 2021).
Consider for instance the second utterance in "John went shopping. He paid the cashier." As paying the cashier is easily inferable in the context of shopping, the utterance is redundant, and has been shown to raise an atypicality implicature, namely that John doesn't usually pay the cashier. We name these inferences triggered by a redundant utterance a habituality inference. Previous work has also shown that the strength of these inferences depends on prominence of the redundant utterance -- if it is stressed prosodically, marked with an exclamation mark, or introduced with a discourse marker such as ``Oh yeah'', habituality inferences are stronger (Kravtchenko, 2021; Ryzhova and Demberg, 2020).

The goal of the present paper is to propose a computational model that can capture both the habituality inference and the effect of prominence. Using the rational speech act model, we show that habituality inferences can be captured by introducing joint reasoning about the habituality of events, similar to Degen et al., 2015; Goodman and Frank (2016). However, we find that joint reasoning models principally cannot account for the effect of differences in utterance prominence. This is because prominence markers do not contribute to the truth-conditional meaning. We then proceed to demonstrate that leveraging models which have previously been used to model low-level acoustic perception can successfully account for the empirically observed patterns of utterance prominence.