Explaining the performance register: stylistic shift as realignment strategy

Natalie Schilling-Estes

Sun. 9-10:40 B

Because sociolinguists typically focus on utterances which minimize the attention paid to speech, they have tended to dismiss "performance speech," in which speakers deliberately highlight what they perceive to be salient features of their own or another dialect. However, performance speech is quite pervasive in a number of communicative contexts, and in some communities it even plays a central role in daily life (e.g. Bauman 1974). Thus, the performance register cannot be neglected in any thorough investigation of the patterning of inter- and intraspeaker language variation in its social setting.

In this presentation, I conduct a detailed analysis of performance and non-performative speech in a series of extended interviews conducted with a middle-aged male resident of Ocracoke, North Carolina. Ocracoke is an island community which existed in relative isolation from mainland dialect areas for two and a half centuries as it developed its unique language variety. Over the past several decades, islanders have come into increasingly frequent contact with tourists and new residents, and their dialect is receding as a result. With this influx of outsiders, islanders have become accustomed to solicitations for samples of their "quaint" speech. In response, the community's prominent performers have developed stock phrases which highlight island features, including the highly salient raised/backed /ay/ diphthong (that is, [[[radical]]/^{^}] which has come to characterize the Ocracoke, or "hoi toider," dialect. My investigation centers on the acoustic analysis of the nucleus of the /ay/ diphthong as it occurs in the performance phrases my informant utters and as it occurs in several non-performative contexts.

The current analysis reveals several important insights about the nature of the performance register and language variation. First, performance speech can be shown to display quite regular patterning rather than the irregularity traditionally ascribed to "subordinate dialect shift" (e.g. Labov 1972, Baugh 1992). Further, the regular patterning of performance speech provides naturalistically based insight into speaker perception of language features, since, in order to successfully perform their own or another dialect, speakers must decide which features are "important" or "noticeable"--that is, which dialect features are most salient. Finally, the incorporation of performance speech into language variation study allows us to rethink generally accepted, yet often incomplete, notions of register and style, particularly Labov's (1972) "attention to speech" model and Bell's (1984) "audience design" model. The current study indicates that the framing of conversational interactions and participant alignments within these frames (e.g. Tannen and Wallat 1993, Goffman 1981) may play a far more central role in intraspeaker variation than either attention to speech or the composition of the literal audience. In the sociolinguistic interview, these alignments may encompass not only immediate participants but also linguists who will listen to the interview upon its completion. This "deferred" audience has been largely overlooked even in audience-design based studies of style shift. However, speaker awareness of this audience may well account for a wide range of register and style shifts, including shifts from naturalistic conversation to the exaggerated vernacular that characterizes the performance register.