A subsegmental approach to coda /s/ weakening in Dominican Spanish

Abstract

Analyses of the variable weakening of coda /s/ in Spanish have largely relied on a categorical conception of variability. While categorical descriptions have proven to be useful heuristics, they have nevertheless tended to obscure important facts about phonological variation and the factors that probabilistically condition it. A growing body of research on sociophonetic variation offers evidence that correlations between linguistic forms and social factors can be manifested in fine-grained subsegmental aspects of speech. The current study conducts a subsegmental analysis of coda /s/ in the speech of five Dominican women. Instrumental analysis of 625 tokens: (1) demonstrates that a strictly segmental description groups together tokens that are significantly different from one another acoustically; (2) offers an instrumentally-based, alternative description of coda /s/ tokens in terms of two continuous, acoustic measures, duration and center of gravity; and (3) argues that a subsegmental approach is better at describing the variation present in speech, and also better exploiting the explanatory power of several linguistic factors known to condition the phenomenon. As a result, the present analysis offers a more accurate picture of the Spanish of Dominicans in New York, and provides a model for greater descriptive and explanatory adequacy in the study of Spanish in general.