Abstract: Processing the music rhythm and tempo is based on time perception, which contains the processing of time duration, time interval, rhythmic pattern and grouping. In the studies of time duration, we have found an interesting phenomenon called Time Order Error (Woodrow, 1951; Allan & Gibbon, 1994; Schab & Crowder, 1988), which said people have some biases on distinguishing the duration when it is less than 500ms or more than 700ms, and when the time duration is between 500ms and 700ms, there is no perceptible difference. Later, some researchers found a similar phenomenon in the studies of time interval, when the time interval is less than 500ms, people will estimate accelerated tempo changes more accurately, when the time interval is more than 700ms, people will estimate slowdown tempo changes more accurately. When the time interval is between 500ms and 700ms, various investigations indicated different views on this issue. This work was aimed to investigate whether people have perceptual bias when the time interval is 600ms (tempo 100bpm). 600ms is said to be the most natural interval in music (Fraisse, 1982; Krumhansl, 2000), but few experiments are based on this interval. In addition, there is no relevant study on the issue whether meter types (such as duple and triple meter) will affect people’s judgment on the beat ahead or behind. So this paper also attempted to explore the impact of tempo changes by certain factors such as change variety, change direction, meter type as well as different music learning background. This experiment chose two drum beats as downbeat and upbeat, to perform duple and triple meter sequences respectively. In each trail there were three standard bars before the variable beat, the variable beat will put ahead or delay for 15%, 10%, 8%, 5 %, 2% or arrive on time. Participants who were music and non-music college students were divided into two groups, according to whether he or she had more than 10 years learning for music. The two groups were asked to listen to each trail carefully and make the choice whether the last beat came early, delayed, or on time. There were 96 trails for both duple and triple meter. The experiment have shown that people are more sensitive when they feel the advance beat than the delayed beat, and this superiority showed especially obvious in triple meter. Furthermore, music students performed better than non-music students in both accuracy and reaction time, especially obvious in triple meter. It can be inferred that people distinguish accurately speed-up tempo changes compared to slowdown tempo changes, and professional music training experiences may help on improving people’s ability of judging tempo changes, and this effect acts more obviously in triple meter.

Key words: Music Cognition, Meter, Tempo, TOE, Duple, Triple

摘要： 辨别节拍速度变化是音乐认知能力的重要组成部分。本文通过控制节拍速度的变化量(15%、10%、8%、5%、2%)、变化方向(提前与滞后)、节拍类型(2拍子和3拍子)3个变量，比较了音乐专业大学生和非音乐专业大学生对于节拍速度变化的感知能力。研究结果显示：人们对提前的变化比滞后的变化更敏感，并且这种知觉优势在3拍子条件下更为明显。同时音乐组被试对节拍速度变化的辨别能力显著高于非音乐组，并且这种优势在3拍子中更明显。

关键词: 音乐认知 节拍 速度 时序误差 2拍子 3拍子