Committee: John Singler (chair), Renee Blake, Gregory Guy, Laurel MacKenzie, Robin Dodsworth (North Carolina State University)
Throughout the history of the field of variationist sociolinguistics, studies have typically focused on urban spaces. Such studies tend not to investigate patterns of language use in suburbs, and often assume that a city and its suburbs share an urban dialect. This is rather surprising, as over 50% of Americans live in suburbs. The relative absence of suburban language studies from the literature thus poses an issue for the field in two ways: if we do not understand patterns of language variation and change in suburbs, we do not understand patterns of variation and change in the United States. At the same time, the development, social structure, and ideologies surrounding suburbs are in many ways peculiarly American. Much of the work in variationist sociolinguistics is American-centric, and without an understanding of language variation and change in suburbs, we cannot properly contextualize variationist theory and findings in crosscultural contexts. This dissertation represents an attempt to contribute to such an understanding.
Suburbs have a complex spatiality and relation to their central city. They lie some distance away from the urban core, and metropolitan areas are often highly fragmented, yet suburbs have close economic and cultural ties to the central city. To study them necessitates attention to their history and development, as well as attitudes surrounding them. As such, this project uses a case study of Greater St. Louis to explore the interaction of the spatiality of suburbs with linguistic variation and change from spatio-temporal, social, and moral geographic perspectives. The study uses the apparent time construct to compare 52 speakers who grew up in South St. Louis, which is an urban environment, Kirkwood/Webster Groves, which are older commuter suburbs with long-standing ties to the city, and St. Charles County, which is a set of newer suburbs that have developed rapidly since 1970. Drawing on sociolinguistic interviews and a questionnaire, the study examines patterns of variation and change in the three field sites vi with respect to four questions: 1. What is the trajectory of language change over the course of the development of a suburb, and how does it relate to language change in the central city? 2. Is the metropolitan area a single speech community? 3. Is there a difference in language use between suburbanites who maintain frequent contact with the central city and those whose lives are more peripheral to it? 4. Is urban/suburban identity reflected in language use? To answer these questions, the study relies on three linguistic features that distinguish St. Louis English from the surrounding Midland: the raised and fronted TRAP vowel associated with the Northern Cities Shift (NCS), the fronted LOT and lowered THOUGHT that do not participate in the low back vowels, and the START/NORTH merger traditionally associated with St. Louis.
Some linguistic changes, like the fronted LOT and lowered THOUGHT of the NCS, diffused to St. Charles County while the area was still rural. Other features like a raised TRAP, however, entered St. Charles County rather abruptly during the period of suburbanization. This change appears to be an example of relocation diffusion (Johnson 2007). However, the findings suggest that while abrupt relocation diffusion is a possible outcome of suburbanization, the more robust outcome is that the trajectory of language change in the developing suburb abruptly shifts to match that of the rest of the metropolitan area. Instead of finding that innovations diffuse throughout the metropolitan area, we find that the entire metropolitan area participates in sound change together.
The city and its suburbs share a great deal of linguistic features and norms. There are some key differences, however: suburbs are leading the shift to a nasal TRAP system and backed LOT, and LOT-fronting is conditioned differently. St. Charles County and Kirkwood/Webster Groves have a comparatively backed DOLLAR and DOLL, while South St. Louis only has a backed DOLL. Furthermore, linguistic norms are not fully shared throughout the region, with location- vii based differences in how the low back vowels and TRAP system are targeted in careful speech contexts. This suggests that the fragmented nature of metropolitan areas plays a role in linguistic variation in the short term, but the close ties that make a metropolitan area metropolitan play a greater role in the long term.
Urban/suburban contact has a bidirectional effect on linguistic production. Speakers from St. Charles County who commute to St. Louis County have a higher pre-oral TRAP, and the youngest commuters have a lower NORTH. Both of these features are more STL-like by virtue of being closer to the NCS TRAP system and traditional START-NORTH merger. At the same time, speakers raised in South St. Louis who moved to the suburbs as adults have a higher NORTH and backer LOT than those who remained. Both of these features are less STL-like by virtue of being further from NCS LOT-fronting and the traditional START-NORTH merger. This suggests that a metropolitan dialect may be a continuum in which suburbanites with little urban contact are clearly different from urbanites with little suburban contact, but those in contact with the other group fall somewhere in between. Urban/suburban identity has a strong effect on linguistic production and patterns of change. Over time, speakers in St. Charles County came to view themselves as living in the St. Louis area, and there is a corresponding rise in urban orientation in apparent time. This effect seems to be linked to the location’s abrupt adoption of metropolitan features and norms and shift toward the metropolitan trajectory of language change. Identity also influences engagement in contact situations; speakers from St. Charles County with long commutes evaluate suburbs more negatively than those who do not commute, while speakers from South St. Louis who moved to the suburbs evaluate the city more negatively than those who remain.
These research findings point to the beginnings of a theory of suburban language variation and change in which place-based identity plays a key role. This identity is deeply connected to migration patterns, and the interaction of migration and identity may determine the outcome of language change during suburbanization. Place-based identity is scalar and multivalent, which enables speakers to engage in complex linguistic practice. This requires us to view locality as scalar between hyper-local municipalities and the local metropolis. A model of the speech community is offered that is similarly scalar in TimeSpace (Wallerstein 1998), from the hyper-local and synchronic to higher levels of locality and diachrony. Individual suburbs can therefore constitute a speech community synchronically, and in this sense may differ from the rest of the metropolitan area with respect to linguistic features and/or norms. The metropolitan area as a whole constitutes a speech community from a diachronic perspective, as over time the close economic and cultural ties lead to a shared set of linguistic features and norms.