This dissertation presents a study of African American and Hispanic speakers in East Austin, Texas, an area that has been historically segregated and is now experiencing rapid gentrification. The current literature is lacking research on minority participation in sound change. This dissertation contributes to redressing the deficit in this area by describing data from sociolinguistic interviews with African American and Hispanic speakers; the following questions are considered: How do social variables influence a linguistic system? How do the social pressures of gentrification influence linguistic variables? To what extent do minority speakers participate in sound changes present in the majority group's variety of English? Specifically, it considers GOOSE and GOAT fronting and the relative positions of LOT and THOUGHT. It has previously been suggested that only Anglo speakers participate in changes affecting these vowels. With regard to GOOSE and GOAT, in many varieties of Southern English, the back vowels become fronted in Anglo speech while they remain backed in African American and Hispanic speech (e.g., Thomas, 2001a). However, the findings discussed here show that African American and Hispanic speakers may front these vowels to achieve particular stylistic purposes. With regard to LOT and THOUGHT, these vowel classes have merged, or are undergoing merger, in many regions of the United States (Labov, Ash, & Boberg, 2006). In Texas, Anglo and Hispanic speakers exhibit the LOT~THOUGHT merger while African American speakers reportedly do not and in general resist the merger (Bailey, Wikle, & Sand, 1991; Bernstein, 1993; Labov et al., 2006; Thomas, 2001a). The findings presented here suggest a trend for LOT~THOUGHT moving towards merger among African American speakers in Central Texas. This is unexpected among African American speakers in Texas and in the U.S. at large. The findings presented illustrate the importance of ethnically diverse samples in describing speech. In Texas where there is a large population of Hispanic residents, we cannot claim to have a thorough knowledge of the regional variety of English without investigating minority speakers. The analysis presented here is a step towards describing a more diverse data set of regional American English. / text