ABSTRACT
The goal of this study is to shed light on Mexican listeners’ social evaluations of [v] in the voices of three Spanish-speaking groups in the United States: late immigrants from Mexico, heritage speakers of Mexican descent, and learners of Spanish as a second language. To this end, 75 Mexican listeners participated in a matched-guise test in which 12 speakers (4 from each of the aforementioned groups) gave directions. One baseline clause was used for each speaker, with cases of <v> spliced once to contain [v] and once to contain the standard [b] or [β], and listeners evaluated each voice along a matrix of social properties. A statistical analysis finds that speaker gender significantly conditions evaluations: [v] decreases evaluations of intelligence and good Spanish for all male speakers but increases evaluations of intelligence and confidence for women in the late immigrant group. I contend that [v] is perceived as a hyperarticulation strategy that indexes precision, literacy, and education, enhancing positive perceptions of femininity, but the variant is evaluated negatively where hyperarticulated speech is not considered the gendered norm—that is, in men’s speech. Finally, this study proposes that monolingual listeners may hold biased linguistic expectations for heritage speakers, potentially resulting in linguistic discrimination.
