Specific Language Impairment (SLI) revisited : evidence from a psycholinguistic investigation of grammatical gender abilities in Brazilian Portuguese-speaking children

This thesis is about Specific Language Impairment (SLI) in children. Its aim is twofold: 1. To provide a theoretical analysis of the field of SLI and discuss its controversies from a novel angle and 2. to present the outcome of a behavioural study which evaluated abilities related to gender agreemen...

Full description

Bibliographic Details
Main Author: Silveira, M.
Published: University College London (University of London) 2011
Subjects:
410
Online Access:http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.565214
Description
Summary:This thesis is about Specific Language Impairment (SLI) in children. Its aim is twofold: 1. To provide a theoretical analysis of the field of SLI and discuss its controversies from a novel angle and 2. to present the outcome of a behavioural study which evaluated abilities related to gender agreement in Brazilian Portuguese-speaking children with language impairment. In the first part, I develop a critique of the field, focusing on the multiplicity of conceptions of the term language embraced by different disciplines that study the disorder. A critical review of how the field of SLI has developed in recent decades reveals that the conceptual fluctuation in the use of the term language has, in many ways, impeded progress in the field. I claim that the only way SLI could be a valid category is if studies focus on basic language skills which, under typical conditions, are acquired spontaneously, without any formal instruction. In Part II, I report an experimental study carried out on the basis of the approach to SLI advocated in Part I. I present a series of experiments that explore the processing of grammatical gender agreement in the Determiner Phrase (DP), in a range of lexical, morphophonological and morphosyntactic conditions in Brazilian Portuguese. Participants were six children with language impairment and 60 typically developing children, including equal numbers from middle class and working class backgrounds. Results showed that gender agreement was very robust for the two groups of typically developing children but problematic for children with SLI, particularly when adjective agreement and gender assignment to novel nouns were involved. The pattern of errors observed and the theoretical discussion throughout Part II suggest that the processing of determiner/noun agreement is a different phenomenon from the processing of noun/adjective agreement, which is vulnerable in children with SLI. In addition, their difficulties with novel nouns suggest that they may require more exposure to input than typically developing children to acquire the gender of nouns.