UC Berkeley

This dissertation compares the phonetic and phonological features of adult non-speakers' productions of words in an endangered Native American language, Oregon Northern Paiute (also known, and hereafter referred to, as Numu), to productions by fluent speakers. The purpose of this comparison is two-fold. The first purpose is to examine the differences in pronunciation that non-speakers bring to the language, which point to possible directions of future language change in a language that is no longer being learned as a first language by children. Changes brought to the language by second language learners are likely to occur due to transfer effects from English and processes of regularization, but may also occur due to the intensification of socially salient language features, or hypercorrection (see Wolfram, 2002). For this reason, two groups of non-speakers were included in the study: English speaking members of the community where Numu is spoken (Warm Springs, Oregon) and English speakers from outside the community. It was hypothesized that the latter group would only exhibit transfer effects or regularization, while the Warm Springs group would also exhibit hypercorrection of what they perceive to be salient features of Numu. By comparing the productions of the two non-speaker groups, specific aspects of potential change are identified and classified as transfer, regularization, or hypercorrection.

The second purpose of the comparison between speaker and non-speaker productions is to ascertain specific differences in pronunciation that result in perceivably accented speech. This research goal is achieved by examining fluent speakers' reactions to non-speakers' productions. It was hypothesized that not all features unique to non-speaker produced speech would result in a perceivable accent. Learners who wish to improve their pronunciation from the perspective of the Numu community could then focus particularly on the features that do contribute to a noticeable accent.

This research makes contributions to our understanding of phonetic and phonological change in endangered language contexts, both from a second language acquisition perspective and a socio-phonetic perspective.