Ethnicity, L1 interference, and sound change in New York City

The linguistic and social landscape of New York City is more diverse today than it has ever been, with an estimated 800 languages being spoken, and the majority of the population speaking a language other than English at home. This ethno-linguistic heterogeneity has the potential to foster complex ethnic identities as New Yorkers of various backgrounds interact with one another. Affiliation with one group over another may also have linguistic consequences on an individual's adoption of New York dialect features, and may ultimately alter the dialect itself.;This dissertation investigates linguistic differences among "white" New Yorkers by focusing on a group that has cultivated ethnic pride and ethnic neighborhoods despite a long history of immigration, i.e. the Polish diaspora. The analytic focus rests on regional variables of New York City English. At the linguistic level, I ask whether the adoption of sociolinguistic variation in the L2 (English) can be facilitated or hindered by L1 (Polish) phonology. At the social level, I investigate how the adoption or non-adoption of regional features reflects speakers' multivalent identities (Mendoza-Denton 2002), such as "Polish", "American", "New Yorker", or "male/female".;The data for this dissertation are drawn from recordings of first and second generation bilingual Polish New Yorkers and a control group of non-Polish New Yorkers who identify as "white". The data represent a range of tasks, such as sociolinguistic interviews, reading tasks, and a Polish involvement survey. I examine several sociolinguistic variables considered to be long-standing features of New York City English: TH-stopping ([d]is for this and [t]in for thin), /[open o]/-raising, and the short- a split (Babbitt 1896, Frank 1948, Hubbell 1950, Labov 1966). The latter two features have been reported to be abandoned in contemporary New York City in favor of low /[open o]/ and the nasal split, respectively (Becker 2010).;The results point to substantial linguistic differences between groups of "white" New Yorkers (i.e. Polish vs. non-Polish), as well as within the Polish group. I argue that these differences are accounted for by social factors that are relevant to the construction of ethnic identity, such as the strength of transnational ties. Polish New Yorkers who maintain strong ties with Poland tend to exhibit high rates of a variable linked to Polish ethnicity, TH-stopping (realized with Polish-like voicing), and avoid a stigmatized variable, raised /[open o]/, much like other white New Yorkers (Becker 2010). By contrast, weak transnational ties correlate with low rates of TH-stopping and the presence of raised /[open o]/. The two groups of Polish New Yorkers are consistent in their adoption of a New York City feature that is devoid of ethnic associations, i.e. the nasal split for /ae/. The analyses highlight that members of this "white" ethnic group have developed complex ethnic identifications, and that these identities, such as "Polish New Yorker", are linguistically expressed by jointly employing two sources of variation: ethnolinguistic variables and regional change.