Lee, C.-Y., Tao, L., & Bond, Z. S. (2010). Identification of acoustically modified Mandarin tones by nonnative listeners. Language and Speech, 53(2), 217–243.
Presentation: Hsiao-chien
Summary: Sally
As a follow-up work of Lee et al. (2008), this study investigated the relationship between the amount of acoustic information and the effectiveness of tonal identification in Mandarin among nonnative listeners. Four kinds of acoustically modified syllables, intact, silent-center, center-only, and onset-only, were generated from monosyllabic Mandarin words preceded with a carrier phrase. The tone of the phrase offset either matched or mismatched the tone of the onset of the target words in F0 height. Forty native English speakers in their 1st, 2nd, or 3rd year of Mandarin learning attended all three experiments. In Experiment 1, the acoustically modified syllables were presented with their original carrier phrases. In Experiment 2, these syllables were presented in isolation. In Experiment 3, they were excised from their original carrier phrase and cross-spliced with another one.
Similar results were obtained throughout the experiments: Unlike the native listeners, who were able to automatically compensate for the coarticulation effect (Lee et al., 2008), the effect of the preceding tone was significant in this study. Test syllables with F0 height of the offset of the carrier phrase matching that of the onset of the target syllable were responded with higher accuracy. For acoustic modification, the accuracy of tone identification was positively correlated with the amount of acoustic information (intact ≈ center-only > silent-center > onset-only). In terms of tones, for both accuracy and RT, Tone 4 was consistently the easiest whereas Tone 2 the most difficult for the nonnative listeners. A possible account for the former is that Tone 4 is the only tone that is similar in its acoustic properties to an English contour (Broselow et al., 1987); the latter may be explained using the tonal contour—Tone 2 in fact is not a linear rise, but remains flat or even drop slightly before making the rise. Furthermore, among the silent-center syllables, the Tone 1-Tone 4 confusion was found less troublesome for the 3rd-year learners, whose experience is believed to facilitate the integration of the onset and offset information of a syllable. Finally, consistent with Gandour’s (1983) finding that English listeners are more sensitive to F0 height than F0 direction, the bias toward Tone 3 in the Tone 2-Tone 3 confusion among native Mandarin listeners (Lee et al., 2008) was not found in this study.
