CUASI NOMÁS INGLÉS: PROSODY AT THE CROSSROADS OF SPANISH AND ENGLISH IN 20TH CENTURY NEW MEXICO

This dissertation investigates prosodic change in the long-term language contact setting of Traditional New Mexican Spanish (NMS). NMS prosody is perceptually distinct from other contemporary varieties of Spanish (Hills 1906, Bowen 1952, Lipski 2011), yet the features which make it unique have not been acoustically examined. This study hypothesizes that bilingualism with English has affected NMS prosody and analyzes three features which are known to differ between Spanish and English and therefore provide a quantitative point of comparison: pitch peak alignment, pitch variability, and rhythmic timing. These variables have been demonstrated to be susceptible to transfer in contact situations, including Spanish-English settings. This study asks whether NMS prosody has remained relatively stable over the 20th century or whether patterns have changed in a more English-like direction. This study also asks whether socio-demographic and language use factors correlate with change in Spanish prosodic patterns. This study answers these questions using naturalistic speech from 60 speakers in the New Mexico Colorado Spanish Survey (NMCOSS) (Bills & Vigil 2008). Participants in the NMCOSS have birth dates that span approximately 100 years (1897-1978), precisely the period that saw the change from a Hispanic majority to a Hispanic minority in New Mexico concomitant with an increase in English-Spanish societal bilingualism. Results of the mixed-effects regression analyses indicate that there is no evidence for change in apparent-time, demonstrating that the features considered in this study have not significantly changed in a more English-like direction throughout the 20th century. In particular, speakers retain the delayed peak alignment pattern typical of pre-nuclear broad focus declarative pitch accents in monolingual varieties of Spanish. Regarding rhythmic timing, measured using the PVI formula (Low & Grabe 1995), findings suggest that rhythmic timing in New Mexican Spanish has remained relatively stable across the three generations in this sample (mean PVI = .36), and there are no clear indications of influence from English. Regarding pitch variability, also measured using the PVI formula, younger speakers were expected to exhibit more pitch variability due to influence from English, which is reported to have a wider pitch range than Spanish (Kelm 1995, Enzinna 2015, Navarro Tomás 1944, Stockwell & Bowen1965, Cruttenden 1986). While Age was found to significantly correlate with pitch PVI, it was not in the expected direction; it was found that as the age of the speaker decreases, the less likely they are to exhibit more pitch variability. In sum, the analysis of prosodic features in this study suggests NMS has remained relatively impervious to English influence. This study provides a picture of the situation of language contact between Spanish and English in New Mexico, and how this speech community has been resistant to change from English, suggesting that social factors should be considered along with linguistic factors when discussing prosodic change in contact situations. This study adds to the literature on prosodic variation in Spanish, including Spanish in the United States. Thus, this dissertation study contributes to the fields of sociophonetics, language contact, and bilingualism