Socio-Phonetic Cues Differentiating African American and European American Voices

North Carolina State University, Raleigh NC

Investigators

Abstract

With National Science Foundation support, Dr. Erik Thomas will conduct two years of linguistic research on the phonetic cues that differentiate African American and European American English. The project has two components, one examining differences in the production of African American and European American speech and the other examining differences in their perception - i.e., how listeners distinguish the two. Past research on African American English has focused largely on grammatical variants (such as copula deletion: "He happy"; and habitual BE: "He be working"), and consonantal variants (such as R-lessness: HEAH for HERE; and consonant cluster simplification: PAS for PAST). Research on vocalic and intonational variation in African American English has been scarce, and research on variation in rhythm and voice quality even rarer. This project will examine these lesser-known variables. The production component will include acoustic analyses of recordings of African Americans and European Americans from the same communities in North Carolina, examining vowel, intonational, rhythm, and voice quality variables. The perception component will include a series of speech identification experiments in which the effects on ethnic identification of the variables investigated in the production component are compared. It is important to study variables involving vowels, intonation, rhythm, and voice quality because they are the factors that give speakers accents and thus are critical for African American and European American identity. These variables are likely to appear in running speech before the better-studied grammatical variables and usually before the consonantal variables. Unlike those variables, variables involving vowels, intonation, rhythm, and voice quality seem to persist in standard African American speech. African Americans can use them to project their ethnic identity even in situations in which they have to avoid the more stigmatized grammatical and consonantal variants. In situations such as telephone conversations and radio broadcasts, listeners can distinguish a speaker's ethnicity using these variables. This ability is involved in both racial profiling and discrimination. More information about these variables, including their distribution within the two ethnic groups and how they are used by listeners, can provide insights as to how racial profiling occurs. This information could be useful in legal cases.