Structural representation and sentence processing in SLI

Cuny Graduate Sch And Univ Ctr, New York NY

Investigators

Linked publications & trials

Abstract

[unreadable] DESCRIPTION (provided by applicant): The proposed research will examine the relation between errors in learning morphosyntactic rules (such as subject-verb agreement) and errors in correctly resolving the permissible referents of reflexive pronouns, such as "himself," in a sentence. Children with SLI make errors in both these areas of grammar but, so far, research has not succeeded in explaining why or in establishing a relation between these apparently distinct types of linguistic impairments. The hypothesis pursued is that SLI children, during real-time sentence comprehension, do not structure sentences in the same way as typically developing children and that this directly affects how their rule systems apply to the sentences they hear or produce. In particular, they may have the correct knowledge about the rule system, but since their processing system does not provide the type of representations to which these rules can apply correctly, they will be impeded in both how they learn rules and in their application of the rules once acquired. In order to test this hypothesis about an underlying deficit in structuring words into syntactic representations, three experiments are proposed. They are directed at examining SLI children's processing of grammatical processes that are essentially structure dependent. The project examines relations between reflexive pronouns like "himself' and the noun that they refer to in a sentence, as well as relations between relativized nouns and the syntactic positions they are anaphorically related to via syntactic movement. A truth-value judgment task will examine which antecedents SLI children assign to reflexives and test the hypothesis that they will select antecedents in a way that is not informed by structure. A cross-modal interference task experiment examines how the reference resolution process is carded out and tests the prediction that SLI children's resolution of reflexives will pattern with resolution of discourse-referring pronouns, measured in terms of processing load. The third experiment will use cross-modal picture priming to examine the time course of establishing anaphoric relations. In particular, the hypothesis will be tested that the lack of correct structure will cause the SLI parser to not immediately and quickly enough calculate how to connect a relativized noun (like zebra in a sentence like "The zebra that the hippo kissed on the nose ran far away") to the verb it is related to. These three experiments all address various consequences of the hypothesized structuring deficit in SLI children. If the experimental predictions are borne out for SLI children, new light will be shed on the relation between structure assignment and language acquisition in general, as well as the relation between morphosyntactic acquisition and syntactic structure processing. The potential findings can be used to inform clinical strategies for SLI. [unreadable] [unreadable]