Events

Title: Discovering the factivity of know.

Author: Rachel Dudley

Abstract: Think and know both express beliefs, but differ in “factivity”: (i) think can report false beliefs; (ii) know’s complement is presupposed to be true. How do children figure out that know is factive but think isn’t? We use corpus methods to examine input with the verbs and determine which distributional cues might signal factivity. We find that direct cues to factivity are sparse: (i) think is rarely used in contexts where the complement is false; (ii) know is rarely used in contexts where its complement is presupposed. However, we find that think and know differ greatly in how speakers use them in conversation: (iii) know is used to ask or answer questions, whereas think is used to make weak assertions. This suggests that noticing the goals of speakers who use the verbs might provide a less noisy signal than observing what speakers presuppose in using the verbs.


Title: Syntactic bootstrapping with minimal morphosyntactic cues: Learning Mandarin Chinese attitude verb meanings.

Author: Nick Huang

Abstract: The meanings of belief and desire verbs (e.g. “to think” and “to want”) is particularly difficult to acquire as these meanings lack reliable physical correlates. One hypothesis is that a child learns about the differences between these two classes of verbs via syntactic bootstrapping, i.e. using observed morphosyntactic cues, and exploiting principled links that relate these cues to their meaning (e.g. Gleitman 1990). In Romance and Germanic, these two classes of verbs are each associated with clausal complements with distinct tense/mood morphology (Bolinger 1968, White et al. 2016, a.o.), facilitating syntactic bootstrapping.

In a language with relatively little overt verbal morphology, such as Mandarin Chinese, it is less clear whether syntactic bootstrapping is a viable strategy for learning the meaning differences between attitude verbs. In this poster presentation, we argue that even though Mandarin does not have overt tense/mood morphology, (i) there are nonetheless syntactic properties that distinguish the clausal complements of belief verbs from those of desire verbs, such as the presence of an overt subject, modal auxiliaries, and negation (e.g. Huang 1982), and (ii) these properties are distributed sufficiently differently across these two verb classes in the input to make syntactic bootstrapping possible. We present initial findings from an ongoing CHILDES corpus study about the distribution of these properties in Mandarin child-directed speech.


Title: Pragmatic control of rationale clauses.

Author: Jeffrey Green

Abstract: Both grammatical and non-grammatical constraints can play a role in determining (co)reference. It has been claimed that PRO in non-finite rationale clauses (RatCs: Rita interviewed Harry [in order PRO to feel better]) receives its referent through syntactic control, since the syntactic position of the controller is important; the subject of the matrix clause, but not the object, can control PRO. I argue in favor of the alternative view, that control of PRO in a rationale clause is pragmatic; its referent is resolved to the party responsible for the fact expressed by the matrix/target clause. Grammatical theories of control are unable to account for many rationale clause phenomena without reference to non-grammatical constraints such as those presented here. This provides evidence that there can be strong discourse constraints on reference that are sensitive to grammatical structure.


Title: In what sense is might an epistemic modal?

Author: Quinn Harr

Abstract: When someone makes a modal claim, we can explain what makes that claim true. We can do this with both epistemic modal claims and non-epistemic ones. However, the claims made with one apparently epistemic modal—might—do not support the sort of epistemic explanations that other epistemic modal claims do. I consider possible reasons able to be offered by existing truth-conditional accounts of might for why this might be so but find none of them convincing. Such accounts take might to somehow have information states encoded into its semantics, but I propose to replace these with sets of circumstances and to explain the apparent epistemicity of might in purely pragmatic terms. In so doing, I argue, we can account for the explanations that bare might claims do and do not support while also accounting for data that motivated a truth-conditional account of might in the first place.