Northern Arizona: Sound Change and Dialect Contact

Previous work on (uw), or GOOSE, and (ow), or GOAT in Northern Arizona English (Hall-Lew 2004, 2005) found that young townspeople and old ranchers both front GOOSE. This suggests two distinct processes of GOOSE-fronting (see also Koops 2010), with Northern Arizona framed as an area of dialect contact between the Southern Shift (Labov 1994) and the California Vowel Shift (Eckert 2004). The present paper builds on these findings to look at another California Vowel Shift feature, the short-a ‘nasal split’ (Eckert 2008), in Arizona. California English shows fronting and raising of pre-nasal (ae), or BAN, and backing and lowering of (ae), or TRAP, elsewhere. We consider these variables in data from Northern Arizonans from 2002-2003. The demographic representation is, unfortunately, not balanced (Table 1). Two subsets of the data were analyzed statistically: gender is tested within the town group only and the town/ranch contrast within the men only. We then qualitatively compare the statistical model for the men to data from five women, the one female rancher and four townies of her approximate age. The analysis is based on normalized midpoint F1 and F2 Hz values (Fabricius et al. 2009; Kendall & Thomas 2009-2014). We ran eight mixed-effect models including following phonological environment (PLACE, MANNER) and YEAR-OF-BIRTH. Four on the town subset included GENDER as a predictor, and four on the men subset included the TOWN/RANCH contrast. Table 2 summarizes the six best-fit models. None of the factors accounted for variation in BAN for the men subset. PLACE was never a significant predictor, nor were any interaction effects. MANNER was not tested for BAN and was always significant for TRAP. TRAP lowering and BAN fronting show apparent-time correlations. Women are leading in TRAP lowering. Women also favor a backer TRAP and a higher BAN than men, although neither variable is changing in apparent time the way they are in California (Eckert 2008). Although not a precise fit, evidence of phonetic divergence between BAN and TRAP still suggests that urban Northern Arizona in 2002 was participating in the short-a nasal split. Only TRAP F2 correlates with TOWN/RANCH, with rancher men producing a fronter vowel than town men. Qualitatively, this is also true for the age-matched subset of women. TRAP fronting is part of the Southern Vowel Shift, and the Southern TRAP vowel is fronter than the Californian TRAP vowel. We take this as potential evidence that rural Northern Arizona in 2002 was better described as having a Southern Vowel System than a Californian one, thus making the region a site of dialect contact between two major US English varieties.