Perceived accents and comprehensibility of native and nonnative English

Abstract

This study investigates the relation between perceived accentedness and comprehensibility of utterances spoken by native and nonnative English speakers. Native and nonnative English listeners heard six types of stimuli: single words, 2–10-word utterances from native and nonnative English speakers. Listeners classified the stimuli as either native or nonnative English speech and transcribed some utterances. Overall, results from the classification and transcription tasks indicate that perceived native-ness or nonnative-ness does not directly correspond to good or poor comprehensibility, as measured by transcription scores. Some nonnative utterances were classified as native speech by both native and nonnative listeners, although the transcription accuracy was not necessarily high. In other cases, native speech was correctly identified as native; however, dialectal pronunciation variation seemed to contribute to low transcription accuracy in some cases. On the other hand, although the transcriptions were accurate, native speech was sometimes classified as nonnative. In other cases, nonnative utterances were correctly identified as nonnative, and the speech was transcribed accurately. These results suggest that familiarity with certain dialects or accents influences listeners’ perception and comprehension for both native and nonnative speech, and that perceived nonnative-ness may not be the main cause for nonnative speech to be less comprehensible.