Emmanuel Chemla (2017). “Heuristic Local Contexts”. Ms. LSCP.

Abstract: Local contexts provide a conceptually motivated system of presupposition projection (Schlenker, 2010). The system works in two steps: first, compute the local context in which the presuppositional phrase occurs and, second, ensure that this local context entails the presupposition triggered. The system is conceptually and technically elegant, but the actual calculations necessary to obtain projected presuppositions are complicated (by some intuitive measure). We first describe a theorem that helps compute the predicted presupposition in one step (see also Schlenker, 2009). We then propose that, at least in some occasions, speakers may follow heuristics and approximate what the projected presupposition is. We describe such heuristics. They provide a computationally simpler way to derive projected presuppositions, which match the prediction of the plain system in numerous cases (certainly for all propositional triggers). From the perspective of cognitive plausibility, a simpler computational system may a priori be seen as valuable, but it is hard to evaluate such a claim properly: humans surely perform some complex computational tasks with ease, language use itself is one such task. Hence, the system will be tested not through complexity measures, but through its divergent predictions. The Heuristic Local Context system accounts for a critical empirical difficulty for the ideal Local Context theory (and many of its alternatives), namely the variability in presupposition projection from quantified sentences.

Keywords: presupposition, projection, local contexts, reasoning, heuristics, quantification