Bootstrapping attitudes

Abstract

This paper explores two classic problems at the semantics-pragmatics interface from a learner's perspective. First, the meaning that speakers convey often goes beyond the literal meaning of the sentences they utter. Second, not all content encoded in utterances has equal standing: some is foregrounded, some backgrounded. Yet a sentence does not formally distinguish what a speaker asserts from what she presupposes or merely implicates. For this reason, the child acquiring a language has a daunting task. She must both extract the literal meaning from the overall message, and separate the background assumptions that are linguistically required from those that are incidental. This paper discusses the ways in which the syntax might guide the child with this daunting task, through a few case studies on children's acquisition of attitude verbs.

Children's attitude problems: Bootstrapping verb meaning from syntax and pragmatics

“…In this section, we develop the hypothesis that the way children acquire the semantics of attitude verbs is by exploiting correlations between the syntactic shape and pragmatic function of attitude reports (Hacquard, ). We first discuss the semantics of attitude verbs and their subcategorization behavior in different languages, and then turn to their pragmatic uses.…”

Three-Year-Olds' Understanding of Desire Reports Is Robust to Conflict

“…Alternatively, children's relative difficulty with think could be due to a P ragmatic A symmetry in the kinds of pragmatic enrichments that these verbs trigger (Hacquard, 2014 ; Harrigan, 2015 ; Hacquard and Lidz, to appear). Think sentences can be used to make indirect assertions.…”

Semantic Information and the Syntax of Propositional Attitude Verbs

“…On this view, learners begin with access to a set of unvalued syntactic features—for example, [+/− subjunctive ], [+/− tense ]—that a particular abstract structure—in this case, main clause —will instantiate, along with a rule that tells them which semantic property verbs that embed clauses with features similar to that structure instantiate—in this case, representational ← main clause . They must then identify what the actual feature valuation for the abstract structure is in order to figure out how to use the rule (Hacquard, ; Hacquard & Lidz, unpublished data).…”

“…First, though this distinction is often discussed as though it were mutually exclusive, some verbs appear to fall into both categories, and suggestively, show up in both frames. For instance, hope p involves both a desire that p come about and the belief that p is possible (Anand & Hacquard, 2013;Hacquard, 2014;Portner, 1992;Scheffler, 2009, but see also Portner & Rubinstein, 2013), and it occurs in both finite 1 and nonfinite 1 syntactic contexts.…”