中国科学院心理研究所机构知识库

摘要：乐句边界知觉对音乐理解至关重要。已有研究发现，边界前长音或休止符能够诱发终止正漂移（Closure Positive Shift,CPS）成分。CPS被认为是乐句边界加工的脑电指标，其时间窗口大概在400-700 ms之间，分布在中央顶部。长音和休止符属于音乐的声学要素。的确，这些声学要素可以暗示乐句的终止。在西方调性音乐中，和声进行（即和弦连接）对乐句的终止具有关键性的作用。比如，属和弦-主和弦进行形成了正格终止，该终止式意味着乐句的结束；属和弦-下中音和弦的进行则不会形成乐句的终止。在音乐作品中，即便句尾不是长音或休止符，但如果出现了属和弦-主和弦进行，乐句必然终止。遗憾的是，尚未有研究探讨听者能否知觉由和声终止产生的乐句边界的问题。考虑到音乐训练可能对乐句边界知觉产生影响效应，本研究通过ERP手段考察音乐家和非音乐家对和声终止产生的乐句边界的知觉。我们通过操纵和声进行，由此形成2种条件：属和弦-主和弦进行（正格终止）构成有边界条件；属和弦-下中音和弦进行则构成无边界条件。本实验采用内隐任务，12名音乐家与12名非音乐家在听完每条音乐刺激后，判断音乐是否包含非钢琴音色。脑电结果显示，与无边界条件相比，音乐家在有边界条件下诱发了CPS成分，而非音乐家诱发了一个持续负波。这些结果表明，听者对和声终止产生的乐句边界知觉受到音乐训练经验的影响。

Abstract：The processing of phrase boundaries in music is critical for musical understanding. It has been reported that the processing of musical phrase boundaries elicits a music Closure Positive Shift (CPS) between 400 and 700 ms. This positivity has a centroparietal maximum. Previous studies have demonstrated that the music CPS is present when the boundary is indicated by a pause and alengthening of preboundary tone. Indeed, both the pauses and lengthening of preboundary tones are acoustic cues, which can imply the ends of phrases. In Western tonal music, however, harmonic progressions such as cadences can also mark phrase boundaries. For example, V–I progression forms an authentic cadence andusually indicates a closure of phrase, whereas V–VI progression is not a cadence and would not lead to a closure of phrase. In the real musical context, musical phrase boundaries are usually marked only by harmonic cadences without acoustic cues. However, no research has examined the neural responses to phrase boundariesmarked by harmonic cadences. Given that musical expertise may affect the processing of phrase boundaries, the present study used the ERP technique to investigate the processing of phrase boundaries marked by harmonic cadences in musicians and nonmusicians. We manipulated harmonic progression. This resulted in two conditions: the phrased condition (V–I progression) and the unphrased condition (V–VI progression). Twelve musicians and 12 nonmusicians were asked to decide whether the stimulus was played with a piano timbre. The results showed that musicians processed the phrased condition by eliciting a music CPS in the time window of 450~600 ms, while nonmusicians exhibited a sustained negativity. These findings suggest that music training modulates the processing of phrase boundaries marked by harmonic cadences.