Training for learning Mandarin tones: A comparison of production and perceptual training

Abstract

Mandarin Chinese lexical tones pose difficulties for non-native speakers whose first languages contrast or do not contrast lexical tones. In this study, both tone language and non-tone language speaking learners of Mandarin Chinese were trained for three weeks to identify the four Mandarin lexical tones. One group took the production training with both visual and audio feedback using Kay Sona Speech II software. The target tones produced by native Mandarin speakers were played back through a pair of headphones and the pitch contours of the target tones were displayed on the computer screen on the top window to be compared with the trainees productions which appear in real time in the bottom window. Another group of participants took the perceptual training only with four-way forced choice identification tasks with immediate feedback. The same training tokens were used in both training modes. Pretest and post test data in perception and production were collected from both groups and were compared for effectiveness of training procedures.