| Recent work on postcolonial, non-native varieties of English has begun to distinguish these varieties from generic second language acquisition in both structural and social terms. However, quantitative evidence of this divergence is still scarce. This dissertation examines both aspects of stable non-native English dialects, drawing on Indian English data collected through sociolinguistic interviews in an urban immigrant neighborhood in Northern California.; A central finding of the study is that emergent dialect features can in fact be quantitatively distinguished from second language learning stages, and their divergence indicates an interaction of language transfer and discourse universals. I use implicational scaling to model the 'cline of bilingualism' and show that certain non-native features (article use) show more stable patterns of use than others (copula use, agreement, past tense marking). The emergent article system is not identical to the first language system. When a form exists in the substrate language, it often influences the second language. However, when a gap occurs in the substrate, universal discourse principles intervene. The present speakers use more overt articles with newer information and less so with increasing givenness in discourse, and more with bare NPs than with modified or quantified NPs: this results in an economical, disambiguating function of articles. Universal discourse knowledge appears to play a particularly important role due to the absence of native input.; A secondary finding of the study is that proficiency (based on amount of English education and functional use of English) can only account for part of the observable variation among speakers: the full range requires a consideration of speakers' ideologies and dialect awareness. Speakers' choice of certain American or Indian phonological variants does not correspond to their level of proficiency, but rather to specific ideological stances they develop through contact with other communities. Evidence of conscious non-native linguistic ideologies can also be seen in many speakers' meta-linguistic commentaries about 'grammar', 'accent', and 'Americanization'.; The methodologies and findings presented here allow the study of such varieties to move beyond a dichotomy between second language acquisition and native variation and to develop an integrated account of language change in stable bilingual situations. |
