Past Event

Title: Disjunction, Acquisition, and the Theory of Implicature

Speakers: Andrea Astle (Carleton Psychology), Dr. Deepthi Kamawar (Carleton Institute of Cognitive Science/Psychology), and Dr. Raj Singh (Carleton Institute of Cognitive Science) (joint work with Dr. Danny Fox (MIT Linguistics) and Dr. Ken Wexler (MIT Brain and Cognitive Sciences/Linguistics))

Abstract: In so-called “truth-value judgment tasks” (Crain and Thornton 1999), adults often judge “A or B” to be false while children typically judge it to be true when A and B are both true. Do children assign a different semantics to disjunction than adults, or they do they assign the same semantics but differ along some other cognitive capacity, e.g., pragmatics or processing?

A recent consensus has emerged that: (i) adults and children assign or the semantics of inclusive disjunction, and (ii) adults do, but children do not, compute an additional pragmatic inference, a so-called scalar implicature, from which it follows that A and B is false. More generally, there is a consensus that children assign the same semantics to logical operators as adults, but differ from adults in that they find it difficult, due to processing limitations, to compute pragmatic inferences such as scalar implicatures  (e.g., Chierchia et al. 2001, Gualmini et al. 2001, Noveck 2001, Reinhart 2006, Barner et al. 2011, Stiller et al. 2011).

We report on the results of a recent study where we found that the consensus has generalized from the wrong data. Specifically, we found evidence that children treat or as ambiguous between three readings: (1) inclusive disjunction, (2) exclusive disjunction, and (3) conjunction. The conjunctive reading in (3) was missed in the previous literature on child acquisition of scalar implicature (though cf. Paris 1973, Braine and Rumain 1981), and cannot be explained by assumptions (i) and (ii) above. We argue that while the results are puzzling from the perspective of pragmatic theories of scalar implicature (e.g., Horn 1972, Sauerland 2004, Geurts 2010), they receive a natural interpretation under the recent idea that implicatures are not pragmatic, but are computed as ambiguities within the grammar (e.g., Chierchia 2004, Fox 2007, Chierchia et al. 2008).