Self Adaptive Phonetic Training for Mandarin Nasal Codas

Abstract

This study proposed a new perceptual training paradigm called Self Adaptive Phonetic Training (SAPT) for Japanese listeners’ perception of nasal codas in Mandarin, which dynamically provided a customized training plan for each participant. 17 participants were divided into two groups: the SAPT group and the High Variability Phonetic Training (HVPT) group. They were required to conduct 12 training sessions in a self-controlled environment with the training software. The results of a perception and production assessment showed that participants could benefit from both training paradigms and the effect size of SAPT was a little larger than HVPT. The training effect was transferred to new stimuli and new talkers in SAPT and HVPT. However, it did not transfer to new phonetic environments in both training paradigms. Moreover, perceptual training turned out to be effective in the production domain in both paradigms. These findings suggest that, in general, SAPT is an effective training paradigm for improving listeners’ perceptual and production ability to identify nasal codas in Mandarin. Notably, it reduces the time spent on unnecessary items and raises the efficiency of training. It sheds new light on constructing the perceptual training paradigm and may have implications for phonetic training in other second language learning.