Ethnicity and Phonetic Variation in a San Francisco Neighborhood.

Abstract

This dissertation advances research in sociolinguistics by analyzing phonetic variation in a majority Asian American community in San Francisco, California. As one of the first community studies focusing on Asian Americans in an urban US context, this work speaks to ongoing discussions about speaker ethnicity, phonetic variation, and regional sound change. The analyses shows Asian Americans sharing the same changes in apparent time as their European American cohorts. In some cases, the correlation with speaker age is even more robust within the Asian American speaker sample than within the European American sample. I argue that this pattern is particularly likely given the social history of San Francisco and the settlement history of the particular community. The community of study is the Sunset District, a large residential neighborhood in western San Francisco. The area has undergone a relatively rapid demographic shift since the 1970s. Today its population is approximately 52% Asian American and 48% European American. Generational differences among neighborhood residents are prominent, in terms of how residents characterize and relate to their community. These changing prestige values map onto the emergence of competing linguistic markets. The phonetic analysis draws on these ethnographic insights in an effort to explore the social meaning of particular variants and the motivations behind participation in local sound change. The variables analyzed are two well-known features of sound change in US English: the merger of the low back vowel classes, as in LOT and THOUGHT, and the fronting of the nuclei of the mid- and high back vowels, as in GOOSE and GOAT. The results show that residents of San Francisco's Sunset District are moving towards the low back merger in apparent time. This development, shown through correlations between production measures and speaker age, is in occurring in parallel with other regions of the West. Specific to San Francisco, however, is that some speakers still maintain this phonemic distinction, regardless of age. While ethnic variation does not predict vocalic variation, Asian Americans show change in apparent time towards low back merger, while the correlation among European Americans is not significant. Furthermore, while speaker sex class does not predict vocalic variation, women exhibit change in apparent time towards more merger, while the correlation among men is not significant. Other trend correlations among speaker subsets suggest that some speakers may be orienting towards a broader and newer regional pattern of merger, while other speakers may be orienting to a more local and older linguistic market that grants prestige to the low back distinction. The analysis of mid- and high back vowel fronting shows that Sunset residents are moving toward more fronted productions for both vowel classes, with significant correlations between fronting measures and speaker age across the speaker sample. The pattern for the GOAT vowel is similar to low back merger: while ethnic variation does not predict vocalic variation, Asian Americans show change in apparent time towards GOAT-fronting, while the correlation among European Americans is not significant. In contrast, the fronting of the GOOSE vowel does not appear to vary according to speaker ethnicity or within ethnic subsamples. Instead, while speaker sex class does not predict variation in GOOSE, in the environment following an anterior coronal, women show change in apparent time towards more fronted variants, while the correlation among men is not significant. The sociolinguistics literature contains relatively little work on phonetic variation in the English of Asian Americans, and there is an increasing interest in research exploring the complex interactions between ethnic and regional identities. This dissertation speaks to these gaps, and further argues that Asian American ethnicities are integral to San Franciscan identities and ideologies of place.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 116 citations exist upstream.