Bilinguals' and second language learners' knowledge of Japanese syllable structure

Abstract

The acquisition of second language phonology has been commanding researchers' attention in recent years. The aim of this thesis is to contribute to this area with a study on Japanese as a second language. The thesis explores both the development of phonological competence by post-puberty second language learners and the end state of pre-puberty bilingual acquisition. Reviewing the literature on the theoretical aspects of mora, syllables and syllable structure, we see that the mora is distinctive and plays vital role in Japanese phonology. We next look at the acquisition theories proposed in recent years, and adopt a Universal Grammar-based approach. Comparing first, bilingual and second language acquisition, three research hypotheses are presented: 1) the Mora Assignment Hypothesis, 2) L2 phonological Acquisition and Age Onset Hypothesis, and the 3) Quality and Quantity of Input Hypothesis. To test these hypotheses, a study was designed involving 24 bilingual children and adults, and 94 adult L2 learners of Japanese at varying levels of proficiency. The results provide evidence to support all three research hypotheses. First the data show that the both English-dominant bilinguals and second language learners at all levels deleted morae and all but the beginning second language learners added morae in oral and written production tasks, indicating non-native competence with respect to morae. In addition, learners attempt to preserve the overall mora count. Since English is not a mora sensitive language, the mora conservation exhibited here is from their Japanese. The learners, including English-dominant bilinguals, first become sensitive to the mora and only at a later stage assign segments to the correct mora slot. The difference in performance between English-dominant bilinguals and Japanese-dominant bilinguals was such that by the age of eight, those who had spent more years in Japan demonstrated native phonological competence, whereas the English-dominant bilinguals' performance pointed to non-native competence. With respect to the second and third hypotheses, results from the bilinguals indicate that in addition to age of onset, the amount of exposure to a second language must be taken into account as a factor influencing ultimate attainment. The study also reveals strong influence of literacy in both oral and written production of Japanese.