A Study Of Training Effect On Th Perception Of The Voicing Contrast In French Plosives

This study aims to explore whether perceptual training could have some effect on the perception of the voicing contrast in French plosives by Chinese speakers. Specifically, it is aimed at answering the following research questions: (1) What effect can perceptual training have on the perception of the voicing contrast in French plosives by Chinese speakers? (2) What errors do the subjects make in each task type before and after training? Is there any pattern? (3) Can the training effect be transferred to the other places of articulation? If yes, what would be the transfer and in what task type?Thirty-one French-major freshmen from a natural class participated in this experiment. The whole experiment consisted of two one-hour sessions in two mornings, and was conducted in a normal classroom, where only multimedia equipment was availed. The subjects did not wear headphones. The subjects went through pretest and training session one on the first morning, and training session two and posttest on the second morning. Both pretest and posttest included identification and discrimination tasks about three pairs of plosives: /p/ and /b/, /t/ and /d/, and /k/ and /g/. During the two training sessions, the subjects were only trained with /p/ and /b/. The subjects' response sheets in pretest and posttest were collected and marked carefully. The errors that the subjects made in pre-test and post-tests were also checked precisely. All the data were analyzed with SPSS software.Result of data analysis show that on the whole, improvement has been made in the perception of the voicing contrast in French plosives after perceptual training and that the training effect could be transferred to other places of articulation to different extents. In the identification task, the training effect in the trained place of articulation could transfer to the other untrained places of articulation. In the discrimination task, there is some training effect for the trained pair pa-ba after training, but the training effect has not transferred to the other places of articulation within such a short period of time.In terms of error patterns, the study found that before training, the subjects tended to perceive voiceless plosives as voiced ones and made more errors in voiceless plosives. After training, however, voiceless and voiced plosives not significantly different in errors. The subjects made significant improvement in identifying voiceless plosives, but did not make significant improvement in voiced plosives. Before training, pa2, ta2, and ka2 are significantly hard to identify than pal, tal and kal. Improvement has been made with training, so pa2, ta2, and ka2 are not significantly hard to identify than pal, tal and kal after training. For ba, da, and ga, tokens one and two are almost equal in perception either before or after training. It has also been found that on the whole, in the discrimination task of the pretest, for bilabial and alveolar plosives, the pairs that are separated further along the VOT continuum are easier to discriminate than less distinct pairs. For velar plosives, the pair that is separated further along the VOT continuum is as easy to discriminate as the less distinct pair. The pattern in the discrimination task of the posttest is similar.This study is significant in different aspects. It not only demonstrates the plasticity in adult perceptual mechanisms to a more intense degree than previous studies, but also has strong implications for teaching and learning foreign language phonology.