A Corpus-Based Study On Schwa/A/Production In Chinese EFL Learners’ Read Speech

The goal of this research is to study the acoustic features of schwa produced by Chinese EFL learners in their read speech.The reasons of these features will also be discussed.The data of the present study was extracted from the recordings of ESCCL(English Speech Corpus of Chinese Learners)and the target words are to,that,of,at,does,handsome,and somebody,all of whose weak forms are realized as schwa.This study randomly chose the recordings of 32 participants from 8 dialectal areas in China,and 6 American college students’ recordings of the same reading materials were considered as referents.After annotating all the recordings by using Praat,F1(the first formant),F2(the second formant),FO(fundamental frequency),intensity and duration of schwa in the target words were extracted and examined to see how Chinese EFL learners differ with native speakers in these five parameters and how Chinese EFL learners from different dialectal areas differ in schwa production.The major findings can be summarized as follows.As for the first research question,Chinese EFL learners have difficulty in schwa production.Specifically,the vowel quality(F1,F2)of certain Chinese learners’production of schwa showed significant difference compared with that by NSs.Chinese learners tended to produce the strong forms of the vowels where schwa should have been adopted,or to use the Chinese phoneme[(?)]to replace schwa.As to the schwa produced by Chinese learners,the duration was significantly longer,FO was saliently higher,but intensity did not show significant difference.Moreover,zero schwa occurred at a high frequency in the production of NSs;only a few of Chinese participants managed to produce it.As to the strong forms produced by Chinese learners,the vowel quality(F1,F2)of certain target sounds showed significant difference compared with the production by NSs;Chinese learners did not show significant difference in duration or intensity but had a saliently higher F0.The mispronunciation of Chinese students could be a result of L1 transfer or lack of phonetic knowledge of stress.As for the second research question,acoustic analysis of schwa based on dialectal distributions showed that all the eight dialectal groups of Chinese EFL learners had some acoustic features in common:they all showed a tendency to replace schwa with the Chinese phoneme[(?)]and produced schwa with a significantly longer duration.Dialectal backgrounds may contribute to the mispronunciation of schwa in vowel quality,FO and intensity.MIN produced the vowel quality of schwa closest to that by NSs of all dialectal groups.HAKKA showed a saliently higher FO than NSs,JIN and HAKKA showed a significantly greater intensity,and GAN also showed a saliently higher intensity than NSs.This study has both theoretical and pedagogical implications.Theoretically,it enriches the understanding of vowel reduction and provides important theoretical evidence for second language phonological acquisition.Pedagogically,the present study reveals the problems in the schwa production by Chinese EFL learners and provides empirical supports to English pronunciation teaching.