The recognition of accented and unaccented English words by native speakers of Spanish and English

Abstract

This study examined effects of foreign accent and lexical factors (word frequency and neighborhood density) on the recognition of English words in noise. Two groups of native Spanish (NS) adults differing in overall degree of foreign accent (FA) in English (weaker versus stronger FAs) participated as well as a native English (NE) group. Participants identified words that had been spoken by a NE and a NS speaker. It was hypothesized that: (1) the stronger FA group would have less nativelike phonological representations than the weaker FA group, and so would benefit more from hearing Spanish‐accented English words; (2) words in dense lexical neighbors would require finer discrimination of English sounds, necessitating more nativelike phonological representations. The results showed that the weaker FA group recognized as many low‐neighborhood‐density English words as the NE group. For high‐neighborhood‐density words, the weaker FA group recognized fewer unaccented English words than the NE group, but more accented English words. The stronger FA group recognized as many accented words as the other groups, but fewer unaccented words. The differences between the weaker and stronger FA groups were interpreted as reflecting a change in their phonological representations for English words. [Work supported by NIH.]