Speech development and disorder in bilingual children

Abstract

Speech-language pathologists have no clear guidelines on how to assess, diagnose or treat bilingual children with speech disorders. This thesis addresses this issue. The phonological development of 91 Cantonese-English and Punjabi-English bilingual children is described. Two Cantonese-English bilingual children's phonological development over the year they were first exposed to English is also presented. The bilingual children's phonological systems were clearly differentiated. The bilingual children's speech also included many phonological processes that would be considered atypical for a monolingual child. The use of these processes is argued to be characteristic of normal bilingual development. The longitudinal data showed that the atypical error patterns were transient and directly related to the introduction of the second language. Some `atypical' error patterns could be plausibly explained by referring to the nature of the two phonological systems. Other atypical processes could be explained by language-specific differences in normal developmental or adult variation patterns. This thesis argues that the differences evident in the bilingual children's phonological patterns are due to `hypothesis testing' resulting in underspecified realisation rules. There was no indication that bilingual children process phonological input and output differently to monolingual children. However, they differentiate the cognitive-linguistic information they abstract from the two languages, and they use separate phonological realisation rules for each language. This thesis argues that bilingual children use the same phonological processing mechanism for both languages, however they are able to filter each language through the appropriate language-specific phonological information. Case studies of 21 children with disordered speech and treatment case studies of 2 children are also presented. The disordered speech data supports current psycholinguistic models of speech processing the hypothesised levels of breakdown fit with the error profiles evident. The bilingual children with speech disorder validate Dodd's (1995) classification system: four different types of disorder were evident. The results of the two treatment case studies presented suggest that unless intervention targets the underlying deficit the effect of intervention will be language-specific. The investigation into bilingual children with disordered speech indicates that speechlanguage pathologists need to assess both languages of a bilingual child to determine the language-specific patterns and the type of disorder and that it is important to compare bilingual children to their bilingual normally developing peers, not to monolingual developmental data.