Related Experiment Videos
Training American listeners to perceive Mandarin tones.
Y Wang1, M M Spence, A Jongman
1Cornell Phonetics Laboratory, Cornell University, Ithaca, New York 14853, USA. yw36@cornell.edu
The Journal of the Acoustical Society of America
|January 1, 2000
Summary
Auditory training significantly improved Mandarin tone identification for American learners. This training enhanced perception of non-native suprasegmental features and showed lasting retention.
Area of Science:
- Second Language Acquisition
- Phonetics and Phonology
- Auditory Perception
Background:
- Auditory training is effective for identifying non-native speech sounds (segmental distinctions).
- Its applicability to non-native suprasegmental contrasts, specifically Mandarin tones, remains less explored.
- Understanding suprasegmental perception is crucial for advanced second language (L2) proficiency.
Purpose of the Study:
- To investigate the effectiveness of auditory training for acquiring non-native suprasegmental contrasts (Mandarin tones).
- To determine if training gains generalize to new auditory stimuli and talkers.
- To assess the long-term retention of Mandarin tone identification skills.
Main Methods:
- Employed the high-variability paradigm for auditory training.
- Eight American learners of Mandarin participated in eight training sessions over two weeks.
- Trained participants to identify the four Mandarin tones in natural words produced by native speakers.
Main Results:
- Significant average increase of 21% in identification accuracy from pretest to post-test.
- Training gains generalized to new stimuli (18% increase) and new talkers/stimuli (25% increase).
- A six-month retention test showed sustained improvement (average 21% increase from pretest).
Conclusions:
- Auditory training is effective for improving the perception of non-native suprasegmental contrasts like Mandarin tones.
- Perceptual modification of suprasegmental features is achievable through targeted training.
- L2 acquisition processes at segmental and suprasegmental levels may share analogous learning mechanisms.