Colloquium: Holman Tse, "A merger and a split, but no chain shifting in Toronto Heritage Cantonese vowels"

Contact-induced vowel mergers and contact-induced vowel splits are not as well documented in the sociolinguistics literature as are vowel chain shifts. In this talk, I address this gap by presenting key findings from the largest variationist study ever conducted on the vowel system of Cantonese. These findings include evidence for a vowel merger and a vowel split but no evidence for chain shifting in Toronto Heritage Cantonese. I will argue that the vowel merger and the vowel split are contact-induced changes. The data comes from sociolinguistic interviews conducted as part of the Heritage Language Variation and Change in Toronto Project (Nagy 2011). The analysis includes 32 speakers. This includes 12 GEN 1 (first-generation, born in Toronto, immigrated to Canada as adults) and 12 GEN 2 (second-generation, grew up in Toronto) speakers as well as eight Homeland (lifelong Hong Kong residents) speakers. The vowel space for these 32 speakers was created based on midpoint F1 and F2 measurements of 11 monophthong categories and a grand total of 33,179 vowel tokens. Results show evidence for inter-generational change in two parts of the vowel system. The first change involves overlapping F1/F2 distributions of the vowels /y/ and /u/. The loss of acoustic distinction between these vowels appears to be due to English influence since English lacks a contrast between high round tense vowels. The second change involves fronting of the F2 of /ɛ/ (phonologically tense in Cantonese) in pre-nasal context. This appears to be influenced by a similar allophonic split in Toronto English involving the tensing of the pre-nasal vowel in BAN. Further analysis shows that the GEN 2 speakers who lead in these two changes are speakers who use the least amount of Cantonese in the recorded interview samples. This, along with the lack of the same changes observed among Homeland speakers, supports an argument for contact-induced change. These findings have important implications for the study of sound change since they complicate the traditional dichotomy between internally motivated and externally motivated change.