Certain spoken phrases, when removed from context and repeated, begin to sound as if they were sung. Prior work has uncovered several acoustic factors that determine whether a phrase sounds sung after repetition. However, the reason why repetition is necessary for song to be perceived in speech is unclear. One possibility is that by default pitch is not a salient attribute of speech in non-tonal languages, as spectral information is more vital for determining meaning. However, repetition may satiate lexical processing, increasing pitch salience. A second possibility is that it takes time to establish the precise pitch perception necessary for assigning each syllable a musical scale degree. Here we tested these hypotheses by asking participants to rate the musicality of spoken phrases and complex tones with matching pitch contours after each of eight repetitions. Although musicality ratings were overall higher for the tone stimuli, both the speech and complex tone stimuli increased in musicality to a similar degree with repetition. Thus, although the rapid spectral variation of speech may inhibit pitch salience, this inhibition does not decrease with repetition. Instead, repetition may be necessary for the perception of song in speech because the perception of exact pitch intervals takes time.

中文翻译：

---

重复将语音和声调刺激的音乐性提高到相似的程度

将某些口语短语从上下文中移开并重复播放后，它们听起来就像是在唱歌。先前的工作已经发现了几种声学因素，这些因素决定了重复后某个短语是否演唱。但是，不清楚为什么要在语音中感知歌曲需要重复。一种可能性是，默认情况下，音调不是非调性语言中语音的显着属性，因为频谱信息对于确定含义更为重要。但是，重复可以满足词法处理的需要，从而提高音调的显着性。第二种可能性是花费时间来建立为每个音节分配音阶程度所必需的精确音高感知。在这里，我们通过要求参与者在八个重复中的每个重复之后，对口语短语和复杂音调与匹配音高轮廓的音乐性进行评分，从而检验了这些假设。尽管总体上音调刺激的音乐性等级更高，但语音和复杂音调刺激的音乐性却随着重复而增加了相似的程度。因此，尽管语音的快速频谱变化可能会抑制音调突出，但是这种抑制不会随着重复而减小。取而代之的是，对于感知语音中的歌曲，重复可能是必要的，因为精确的音高间隔的感知需要时间。尽管语音的快速频谱变化可能会抑制音调显着性，但是这种抑制不会随着重复而减小。取而代之的是，对于感知语音中的歌曲，重复可能是必要的，因为精确的音高间隔的感知需要时间。尽管语音的快速频谱变化可能会抑制音调显着性，但是这种抑制不会随着重复而减小。取而代之的是，对于感知语音中的歌曲，重复可能是必要的，因为精确的音高间隔的感知需要时间。