Child language, Teaching and Therapy. Volume29(1). pp 77-89.

In recent years, there has been growing recognition of a need for a general, non-language-specific assessment tool that could be used to evaluate general speech and language abilities in children, especially to assist in identifying atypical development in bilingual children who speak a language unfamiliar to the assessor. It has been suggested that a non-word repetition task (NWRT) may be a suitable candidate to fill this role, as it does not rely on knowledge of particular words for performance, and it may be possible to devise non-words that are not specific to any given language. The current study reports performance on a Welsh non-word repetition task by typically developing WelshEnglish bilingual children with varying levels of exposure to Welsh in the home (Only Welsh at Home, Only English at Home, or Welsh and English at Home). The focus of the study was on repetition of initial consonants and consonant clusters in novel words. Both quantitative and qualitative differences were found across groups, according to level of exposure to Welsh, on sounds unique to Welsh, but not on sounds shared by Welsh and English. The data suggest that level of knowledge of the language tested has an important impact on childrens performance on non-word repetition and that the use of the NWRT as a universal speech and language assessment tool should be adopted with caution.