母语和非母语者对北京话相似和相异调拱声调的范畴感知——基于MMN的研究

Categorical Perception of Similar-contour and Dissimilar-contour Tones in Beijing Mandarin of Native and Non-native Speakers: An MMN Study

本文采用非注意的失匹配负波(MMN)实验范式,考察母语者和非母语者(重庆话)对北京话相异调拱(阴平-去声)和相似调拱(阳平-上声)的感知模式。在两组被试中,阴平-去声序列的跨范畴偏差刺激均诱发了MMN,范畴内偏差刺激均没有诱发MMN;不同母语背景的MMN波幅没有显著差异,但重庆话母语者的潜伏期长于北京话母语者。阳平-上声序列的范畴内偏差刺激和跨范畴偏差刺激均没有诱发MMN。本文的结果再次证明了前注意阶段的语音加工既包括声学加工,也包括音系加工。本文结果与前人结果的不一致之处在于,在某些条件下,跨音系范畴的偏差刺激没有诱发MMN,且本研究中所有范畴内的偏差刺激都没有诱发MMN。这些分歧可能表明,前注意阶段的语音加工主要是声学层面的,音系加工受到了语音声学变异程度的制约。

Previous studies have not reached a consistent conclusion regarding whether the perception of tones in a given Chinese dialect is categorical or continuous. Some tones have acoustical perceptual boundaries, while others in the same dialect do not. We hypothesize that there is no clear categorical boundary between any two tones with similar pitch contours, while there is a categorical boundary between tones with dissimilar pitch contours. Since the task of discrimination in behavioral experiments may introduce bias to the results, this paper tests this hypothesis by combining the results of two category perception experiments(behavioral experiments) and two electrophysiological experiments. The subjects were native speakers of Beijing Mandarin(BJS) and native speakers of Chongqing Mandarin(CQS). The stimuli included a Tone 1-Tone 4(T1-T4, dissimilar-contour tones) continuum, within which the stimuli varied along the ending pitch, and a Tone 2-Tone 3(T2-T3, similar-contour tones) continuum, within which the stimuli varied along the turning position of the pitch contour. Electrophysiological experiments were conducted according to the paradigm of inattentive mismatch negativity(MMN) to investigate whether phonological information is processed in the preattention stage. In the behavioral experiments, the perceptions of the T1-T4 continuum in Beijing Mandarin of the two groups of subjects showed categorical characteristics, and there was a significant difference in the category boundary between the two groups, while the perceptions of the T2-T3 continuum of the two groups showed no categorical tendency. In the electrophysiological experiments, standard and deviant stimuli were designed based on the results of the behavior perceptual experiments. The within-category stimulus of the T1/T4 sequence did not elicit MMN for either group. The cross-category deviant stimulus of the T1/T4 sequence elicited MMN for both groups;the MMN amplitudes of the two groups were not significantly different, while the MMN latency of the CQS group was longer than that of the BJS group. Neither the within-or cross-category deviant stimuli of the T2/T3 sequence elicited MMN for either group. The results of the behavioral and electrophysiological experiments all support our hypothesis. The results of the MMN experiments in this study are consistent with those of previous studies;that is, phonetic processing in the preattention stage is carried out at both the acoustic and phonological levels. The inconsistencies between the results of this study and those of previous studies are that in this study, under certain conditions, cross-category deviant stimuli did not elicit MMN, and no within-category deviant stimuli elicited MMN. These inconsistencies may indicate that categorical processing during the preattention stage is mainly shaped by acoustic variation and that categorization in phonological processing is constrained by the degree of acoustic variation.