Abstract

In the society of English as a lingua franca, non-native listeners have been more frequently exposed to situations where they need to evaluate non-native talkers’ L2 speech. This sheds light on bringing up several empirical questions of whether their accentedness ratings agree with those of native listeners, whether their L2 proficiency affects the extent to which the agreement occurs, and finally whether their L1 backgrounds contribute to the judgment of foreign accent. The current study attempts to answer those questions, investigating both native and non-native Chinese and Korean listeners’ perceptions of foreign accent for Chinese-accented English L2 speech. Additionally, segmental and suprasegmental acoustic measures were examined in association with accentedness ratings to find out a best predictor in the judgment of foreign accent. Results showed that non-native listeners did not agree with native listeners, except for the case that high proficiency listeners judged high proficiency talkers’ speech. This suggests that non-native listeners rated L2 speech based on their own criteria which had been established from the phonological and phonetic features existing in their interlanguages. They tended to rate L2 speech as less accented than native listeners because they had more deviant sound features in the interlanguage and created more tolerable and extended accentedness standards. In addition, we found the effects of non-native listeners’ proficiency on accentedness ratings but no effects of L1 background. Finally, suprasegmental variables were constantly correlated with the ratings but the number of segmental variables facilitating the perception of foreign accent decreased as listeners’ proficiency decreased. The suprasegmental features seemed to be easily accessible to both native and non-native listeners in comparison with segmental variables, because pitch range and speech rate were a holistic and global characteristic of speech while the processes of deletion, insertion and substitution occurred locally in a tiny portion along the temporal dimension.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call