Development of Cantonese as a heritage language in children: Experiments in phonology and morphosyntax

This dissertation examines the development of Cantonese in young heritage speakers (age of testing 3;10–12;3) in New York City. These heritage speakers were raised speaking predominantly Cantonese at home, but were exposed to English (the majority language) from a young age. The research investigates whether they were acquiring Cantonese in the same way as Cantonese-English peers in Hong Kong, where Cantonese is the majority language, and which factors contributed to stronger abilities in the heritage language. The two groups of participants were compared in terms of tone discrimination, phono- logical production, and classifier production. The results showed that the heritage speakers discriminated between Cantonese tones less accurately, and that they spoke with lower native-likeness and lower comprehensibility. They were comparable to the participants in Hong Kong in producing the classifier structure, but were less able to use the appropriate classifier form. In the heritage speakers, age of testing predicted tone discrimination, and age of arrival and the amount of Cantonese input and output predicted classifier selection, but no other tested factors showed significant effects on the measures of Cantonese ability. These findings indicate that young heritage speakers do not acquire Cantonese in the same way as majority language speakers, although they can still be undergoing development. The data presented in this dissertation provides a comparison of two groups of bilingual Cantonese speakers, and illustrates the individual differences among heritage speakers.