Processing of English-accented Spanish voice onset time by Spanish speakers with low English experience

Abstract

Listen

Translate article

Previous research (Llanos & Francis, 2014) shows that the processing of foreign accented speech sounds can be affected by listeners’ familiarity with the language that causes the accent. Highly familiar listeners treat foreign accented sounds as foreign sounds while less familiar listeners treat them natively. The present study tests the hypothesis that less familiar listeners may nevertheless be able to apply foreign categorization patterns to accented words by recalibrating phonetic expectations according to acoustic information provided by immediate phonetic context. Two groups of Spanish native speakers with little English experience will identify tokens drawn from a digitally edited VOT continuum ranging from baso "glass" (-60 ms VOT) to paso "step" (60 ms VOT). Tokens are embedded in a series of Spanish words beginning with /b/ and /p/ to provide phonetic context. In the English-accented condition, context words are digitally modified to exhibit English-like VOT values for /b/ (10 ms) and /p/ (60 ms). In the Spanish condition, these tokens are edited to exhibit prototypical Spanish /b/ (-90 ms) and /p/ (10 ms) VOT values. If listeners can accommodate foreign accented sounds according to expectations provided by immediate phonetic context, then listeners’ VOT boundary in the English-accented condition should be significantly higher than in the Spanish condition.