Speaker: Keny Chatain (MIT)
Title: Exploring a new recipe for implicature calculation
Time: Wednesday, September 16th, 1pm – 2pm
Abstract: This is the very preliminary give-me-feedback stage of a work on implicature calculation. My goal is to tackle puzzling data on distributive implicatures in the scope and restrictor of quantifiers (due to Filipe Kobayashi) and try to validate a parallel between distributive inferences and ignorance inferences (cf Meyer, 2014). The main idea is to minimally modify Anvari’s Logical Integrity principle (and with it, the theory of local contexts) and use it to compute regular scalar implicatures, rather than the Maximize Presupposition and Magri cases for which this generalization was originally designed.
The resulting system shows promising results: it captures the problematic data points that motivated it, it also captures a number of the standard cases (Chierchia’s problem, multiple disjunctions) and, with minimal cheating, the more tricky cases of free choice and some under some (Bassi, Del Pinal and Sauerland, 2019). There are some missed predictions (e.g. no account of Hurford disjunctions in sight) and some unanswered conceptual questions (is this a pragmatic or semantic approach? is it a descriptive generalization or a genuine explanation?). Hopefully, you can help me sort this out and figure out whether this is a dead end or an idea worth pursuing!