Acoustic analysis of American listeners’ Mandarin tone productions before and after perceptual training

Abstract

Training American listeners to perceive Mandarin tones has been shown to be effective, with the trainees’ identification improving by 21% [Y. Wang, A. Jongman, and J. Sereno, J. Acoust. Soc. Am. 106, 3649–3658 (1999)]. The effect on production as a result of perceptual training was first assessed by native Mandarin listeners’ evaluation of the trainees’ tone productions, showing that post-training productions were 18% more accurately identified than pre-training productions [Y. Wang, M. Spence, A. Jongman, and J. Sereno, J. Acoust. Soc. Am. 105, 1095 (1999)]. A detailed acoustic analysis has now been performed to further examine the nature of the production improvement. Trainees’ productions include 80 Mandarin words recorded both before and after perceptual training. For comparison, these words were also produced by native Mandarin speakers. For each production, a pitch track was derived. Results show that post-training tone contours approximate native norms to a greater degree than pre-training tone contours, indicating an improvement in production following perceptual training. Furthermore, pitch height and pitch contour are not mastered in parallel, with the former being more resistant to improvement than the latter. The results are discussed in terms of the production–perception relationship and native language influence.