The Apparent‐Time Construct and stable variation: Final /z/ devoicing in northwestern Indiana¹

Abstract

As real-time language data becomes increasingly available for sociolinguistic research, a growing number of studies are benefitting from it in order to study language changes in progress, some of which even explicitly seek to scrutinize the APPARENT-TIME CONSTRUCT itself. Vanishingly few real-time studies, however, have focused specifically on stable sociolinguistic variables, leaving an important gap in our understanding of the Apparent-Time Construct's abilities to model real-time facts. In an effort to address this gap, the present study analyzes a presumably stable sociolinguistic variable -final /z/ devoicing -in extreme northwestern Indiana through real and apparent time. A series of VARBRUL analyses indicate that this variable is, indeed, stable throughout the 20 years of real time covered by the data and that its stability is successfully modeled in apparent time. Additionally, similarities in /z/ devoicing between this community and some other communities where it has also been studied are identified and discussed.

VARIATION IN CHICANO English: THE CASE OF FINAL (z) DEVOICING

“…results overall, final (z) devoicing is common in the speech of buena Vista young people. of 1,827 tokens analyzed, 711, or 38.9%, were devoiced, slightly higher than the rate of 34% that José (2010) found among european American speakers in indiana or the 31.3% that Doviak and hudsonedwards (1980) found among Mexican American children in New Mexico. individual rates of devoicing ranged from a high of 69.8% for speaker 008, a male oriented toward life in the Courts, to a low of 22.8% for speaker 022, a male with an outward orientation.…”

“…Some recent studies report that it is also affected by the morphological class of the fricative and the sex of the speaker (José 2010;Verhoeven, hirson, and basavaraj 2011). the phonological and prosodic environment following the fricative appears to exert the strongest influence both in english and cross-linguistically.…”

“…Smith (1997) also found that prosodic context affected rates of devoicing, with utterance final fricatives completely devoiced, word-final (but not prepausal) fricatives tending to be partially devoiced, and syllable-final word-medial fricatives fully or partially voiced for most speakers. José (2010) found morphological class to have an influence on (z) devoicing. the devoiced variant is favored for possessives and plurals and disfavored for verbal inflections (third-person singular) and other nonmorphemic tokens.…”

“…the phonological and prosodic environment following the fricative appears to exert the strongest influence both in english and cross-linguistically. Devoicing is favored when the frica- ( José 2010;Verhoeven, hirson, and basavaraj 2011) american speech 89. 4 (2014) 388 tive is adjacent to a voiceless consonant and in contexts where articulatory effort tends to be reduced.…”

“…Although the situation has improved greatly with the appearance of a number of survey articles and full-length studies (e.g., Santa Ana 1993Ana , 1996bayley 1994, 1999, 2013fought 1999Mendoza-Denton 1999bayley and Santa Ana 2004;Santa Ana and bayley 2004;Konopka and Pierrehumbert 2008;bayley and bonnici 2009;roeder 2010), many Che features have yet to be fully investigated. final (z) devoicing (e.g., boy[z] ü boy[s]) is one such feature.…”