Speech style and authenticity: Quantitative evidence for the performance of identity

The question of what constitutes an authentic speaker, particularly with regard to African American Vernacular English (AAVE), has been the subject of some debate in sociolinguistics (Butters, 1984; Labov, 1980; Sweetland, 2002) and arises anew in the case of white hip-hop affiliated youth (WHHs) who converge toward AAVE in their speech. This paper takes a quantitative approach to this question by examining how speech style alters the relationship between the frequencies of a variable in different linguistic environments. Guy (1991b) showed that the exponential relationship in English among rates of coronal stop deletion (CSD) in several morphological categories is systematically distorted by constraints on the surface-level phonology. Because stylistic variation appears to operate at this level, such distortion provides an internal measure of a speaker's stylistic shifting away from their neutral vernacular usage. Data on CSD deletion from WHHs who style shift toward AAVE show this kind of distortion when compared with the speech of AAVE speakers. This data provide strong internal evidence in support of the idea that some WHHs are performing a speech style that diverges from their unmarked style.