This sociolinguistic study of 45 speakers - interviewed between 1993 and 1997 in a tri-ethnic, Southern, rural county - shows cultural identity to be the most important social factor influencing language variation. Through participant-observation, variationist research was conducted on two sociolinguistic variables, past be and copula absence, for African Americans, European Americans, and Native Americans (Haliwa-Saponi) in Warren County, North Carolina. Through the quantitative and qualitative analysis of these sociolinguistic data, this dissertation explains the synchronic and diachronic effects of several important social factors, including ethnicity, sex, and cultural identity.
Cultural identity appears to be the most important social factor affecting a speaker's rate of stigmatized variants. Cultural identity cuts across all three ethnic groups so that Warren County has two distinct, identity-based speech communities rather than three ethnically-defined speech communities. The two identity-based speech communities are the local-identity speakers and the expanded-identity speakers.
Local-identity speakers are identified by their strong and exclusive ties to Warren County; expanded-identity speakers have both strong ties with Warren County and ties with areas outside Warren County. Local-identity speakers have high rates of stigmatized variants in both formal and casual contexts. Expanded-identity speakers have equally high rates of stigmatized variants in casual contexts, but in more formal contexts, they have significantly lower rates than their local-identity counterparts. The youngest expanded-identity speakers have few instances of stigmatized variants in any context, indicating that their elders' style switching has become language change for them.
This dissertation also makes several linguistic claims about past be and copula absence. First, Warren County has a tripartite division in the negative past be paradigm, with the variants wasn't and weren't and the innovative variant wont. Wont is the regularized form for all Warren County speakers and functions similarly to ain't. Second, copula absence in Warren County is greatly constrained by phonological factors. This phonologically-constrained copula absence supports the argument for different varieties of copula absence, each with possibly different origins and paths of development.
|
E-mail this message to a friend
