Reanalysis and Hypercorrection Among Extreme /s/ Reducers

Western Nicaragua is an immensely understudied region, and it also represents one of the most advanced coda /s/-weakening dialects of Spanish. Coda /s/ is reduced nearly categorically before a following consonant, vowel, or pause, e.g. cesta ‘basket’ becomes [sehta], más ajo ‘more garlic’ becomes [mah aho], and misas ‘masses’ becomes [misa], respectively. These reductions result in a “breathy Spanish” with rates of reduction similar to extreme Caribbean varieties (Lipski 1994: 291). Given the nearly absolute weakening, this work investigates the present status of coda /s/ in the dialect through an exploration of (i) diachronic data to determine how [s] production has changed over time, (ii) synchronic comparisons with other /s/-reducing dialects, and (iii) [s] hypercorrections. I conclude that /s/-weakening has advanced over the past thirty years as rates of coda sibilance decrease and rates of deletion rise; that [s] in Nicaragua is not a linguistically conditioned, local variant due to its deviant behavior; and that [s] hypercorrections do occasionally emerge in formal tasks, suggesting a loosening of the association between underlying coda /s/ and surface sibilance. Based on these conclusions, I argue that sibilance serves as a social strategy to index education, power, and precision on a global scale, while linguistically, many Nicaraguan speakers are operating with underlying coda /h/ instead of /s/, which helps to account for the innovative behavior of the glottal stop. Not only does this work document a highly understudied language variety, it also elucidates the complex linguistic and social motivations for selecting a particular variant in a radical dialect. This working paper is available in University of Pennsylvania Working Papers in Linguistics: http://repository.upenn.edu/pwpl/