The Weekly Newsletter of MIT Linguistics

LingLunch 11/12 - Daniel Asherov, Danny Fox (MIT), and Roni Katzir (Tel Aviv University)

Speaker: Daniel Asherov, Danny Fox (MIT), and Roni Katzir (Tel Aviv University)
Title: A problem for iterated rationality approaches to scalar implicatures
Time: Thursday, November 12th, 12:30pm – 2pm

Abstract: This talk focuses on two seemingly equivalent approaches to scalar implicatures, namely the grammatical approach (Fox 2007, Chierchia, Fox, and Spector 2012) and the more recent iterative models of rationality (IRMs; Franke 2009, 2011, Frank and Goodman 2012, Goodman and Stuhlmüller 2013). Recent work has demonstrated the success of IRMs in capturing scalar implicatures derived by listeners in reference games, in which listeners are given a word and a set of objects and are asked to choose which object the word refers to. In many typical cases, the grammatical approach and IRMs make identical predictions. We identify two scenarios where the two approaches make different predictions and report results from an experiment testing these predictions. We conclude that participants’ judgments accord with the predictions of the grammatical approach, but not with those of IRMs.

To help in identifying differing predictions, we will use the simplified IRMs discussed in Fox and Katzir 2020 (https://ling.auf.net/lingbuzz/005519). But we will also explain why the same holds for the actual IRMs discussed in the literature (e.g. Rational Speech Act in Frank and Goodman 2012).