汉语普通话母语者的节奏组块感知偏好

个体对声音节奏的感知具有明显的偏好，强度上表现为扬抑组块偏好，时长上表现为抑扬组块偏好，这被称作抑扬-扬抑规律，但这一感知规律是否具有跨语言的普遍性却存在争议。本研究以汉语普通话母语者为被试，考察他们对非语音音调序列的节奏组块感知情况，结果发现强度和时长参数下被试对扬抑形式的判断比例均显著高于随机水平，说明汉语普通话母语者在两个参数上的节奏感知中均表现出明显的扬抑偏好，在强度上支持节奏感知偏好具有跨语言普遍性的观点，在时长上不支持节奏感知偏好受到母语短语韵律特征影响的观点。

The Study of the Rhythmic Perception on Chinese Mandarin

Two perceptual grouping principles, iambic/trochaic law (ITL), have been found over a century ago by Bolton (1984) and Woodrow (1909). Hayes (1995) proposed this law describes humans’ tendency to form trochaic groups over sequences varying in intensity (i.e., the loudest sounds mark group beginning), and iambic groups over sequences varying in duration (i.e., the longest sounds mark group ending). This perceptual grouping has traditionally been thought to be governed by innate, universal principles. However, recent work has found that language or music background may modulate the use of ITL, meaning that these perceptual grouping preferences may instead be a consequence of language exposure. Researches has investigated how adults and infants speaking Italian, Turkish, Japanese, English or Spanish group sequences of syllables, tones or musical instrument sounds. Compared with these languages, Mandarin has its own unique features, such as rhyme, tone and so on. In this study, we explore Mandarin speakers' perceptual grouping preference of non-linguistic sounds by using a grouping task. Each sound sequence was constituted with basic and altering sound in turn, which would last 10s. To assure that the sequences were fully ambiguous in terms of an iambic or a trochaic organization, the beginning and the end of each intensity and duration sequence faded in and out 1.5s. In addition, each sequence was played forward (i.e. starting with basic sound) or backward (i.e. starting with the altering sound). In the intensity condition, changing tone was built with either 60dB or 65dB. In the duration condition, changing tone was built with either 250ms-long or 200ms-long sine wave. The acoustics parameters of base tone were set 150ms in duration, 55dB in intensity and 256Hz in F0 successively. In addition, six filler sequence were also inserted with each condition. Hence in each condition, participants would finish 30 trails. Twenty-nine participants completed the intensity condition, and twenty-six participants finished the duration condition. After each sound, they were asked to indicate whether they have heard a trochaic (i.e., strong-weak) or an iambic (i.e., weak-strong) rhythmic pattern. In the intensity condition, participants showed a soft trochaic grouping bias (M = 0.549, p = 0.091,d = 0.33), as predicted by the ITL. In the duration condition, participants exhibited a trochaic preference (M = 0.566, p = 0.019,d = 0.52), which is different from ITL. These experiments provide new evidence for the proposal that intensity information draws upon a universal mechanism in ITL across languages, yet ITL in duration is shaped by language experience. Especially, light tone may be one of the factors for Mandarin native speakers. Further studies should investigate the effects of Mandarin tone or even dialectal tone on the perception of rhythmic grouping.