| Summary: | This PhD thesis contributes new empirical knowledge to the question of what paths bilingual acquisition of sound structure can take in early simultaneous bilinguals. The issues of language differentiation and interaction are considered in their relationship to language input, crosslinguistic structure and longitudinal effects. Two Russian-Scottish English subjects aged between 3;4 and 4;5 were recorded longitudinally. Russian was spoken in their families, and Scottish English in the community (Edinburgh, UK). The family environments were similar, but one subject had received substantially more input in Russian than the other one. We addressed the detail of their production of prominent syllable-nuclear vowels in Scottish English and in Russian with regard to their vowel quality, duration and vocal effort. Language differentiation and interaction patterns were derived by accounting for the language mode, and by statistical comparison of the crosslinguistic structures to the speech of monolingual peers (n=7) and adults (n=14). Subjects' bilingual results revealed both substantial language differentiation and systematic language interaction patterns. The extent of language differentiation and directionality of interaction depended on the amount of language exposure. Its directionality did not necessarily depend on the markedness of the crosslinguistic structures, and could be bi-directional for the same properties. Longitudinally, language differentiation increased, while interaction reduced. The amount of reduction depended on both language input and the structural complexity of the languages with segmental tense/lax contract and complex postvocalic vowel duration conditioning showing more persistent language interaction effects. The results confirmed the importance of language input. We showed that in bilingual phonological development language interaction should be considered as a normal but non-obligatory process. Besides, some structurally complex processes potentially explainable by 'markedness' (applied to isolated segments) could rather be explained by lexical and phonotactic factors.
|
|---|
