The influence of phonetic features on the perception of accented speech

Abstract

When learning a second language, speakers often produce certain phonetic features of that language differently than native speakers, which results in accented speech. Between Spanish and English, there are a number of phonetic differences which may be produced differently by Spanish speaking learners of English. I focus on two specific features of Spanish-English accented speech. The first is the lack of aspiration in pre-vocalic voiceless consonants, which results in English listeners perceiving the voiced counterparts. The second is the devoicing of the post-vocalic voiced consonants, which results in English listeners perceiving the voiceless counterparts. To test whether these two features have a strong effect on the intelligibility of accented speech, participants were asked to perform a forced-choice word recognition task in which they were presented with an auditory token, spoken by a Spanish-English bilingual, and asked to identify which of two visually presented words the spoken word was. The auditory token differed on pre- and post-vocalic voicing. I predicted that English listeners’ accuracy would be worse when they listened to the pre-vocalic, voiceless tokens compared to the pre-vocalic, voiced tokens and that the opposite would occur in the post-vocalic position. As predicted, results indicated that the subjects’ accuracy was better in the pre-vocalic voiced condition and the post-vocalic voicing condition. This implies that these two dimensions might have a direct effect on English listeners’ perception of an accent, which suggests that at some level, individual segmental qualities can result in accent perception, independent of the presence of suprasegmental features, or even other phonetic differences.