This dissertation is a sociophonetic analysis of the English spoken by Hmong Americans living in the "Twin Cities" of Minneapolis and St. Paul, Minnesota. The Twin Cities has the largest urban population of Hmong Americans in the United States. Through studies of production and perception of vowels involved in sound changes, I investigate whether Hmong Americans|a relatively new ethnic group in the United States have established any elements of an ethnic dialect of English that communicates an identity that is uniquely Hmong American.
Sound changes are particularly fruitful objects of sociophonetic study as they provide a spectrum of potential indexical variables for speakers exposed to those sound changes. I examine Hmong Americans' participation in three sound changes: the Northern Cities Shift, the low back merger, and fronting of the high back vowel (/u/ or goose). Their degrees of participation in those sound changes are compared to age-matched European Americans from the same area.
It was expected that the inferred tight-knit nature of Hmong Americans' social networks would cause a slower uptake of current regional and supra-regional sound changes versus the comparatively looser networks of many European Americans in the Twin Cities. Furthermore, the target population should presumably experience some influence in their English from the Hmong language. Crucially for this study, the Hmong language has phonemic nasal vowels whereas English does not. This L2 influence of phonemic nasal vowels was hypothesized to emerge in Hmong Americans' English as less nasalization overall, and to decrease the likelihood that they will engage in the Northern Cities Shift.
The results of the production study show that European American speakers seem to be participating in one supra-regional sound change, the fronting of the goose vowel, to a greater extent than in the past, and to a greater extent than Hmong Americans. Two other sound changes, the Northern Cities Shift (a regional change) and the low back merger (a supra-regional change), show inconclusive evidence of adoption by either EA speakers or HA speakers. The perception study, which was conducted with a new set of participants, aimed to uncover whether phonetic differences between Hmong Americans' and European Americans' vowel pronunciations are actually detectable by others. Words recorded during fieldwork were rated on a visual analog scale by listeners on several different dimensions of speakers' social characteristics, including ethnicity. It was found that although certain expected phonetic differences were not used to make judgments of speakers' ethnicities, other phonetic differences, some expected and some not, did indeed predict listeners' judgments of speaker ethnicity. Listeners seemed to use either formant values or vowel nasalization (or sometimes both) to judge speaker ethnicity, depending on vowel class, listener ethnicity, and listener birthplace.
Taken together, the results of the two studies provide evidence that Hmong Americans' vowel pronunciations are not simply Hmong-influenced imitations of vowels as spoken by European Americans, and that listeners, especially other Hmong American listeners, can use these complex yet systematic phonetic patterns to make accurate decisions about speakers' ethnicities.