Abstract: From the moment we are born, humans begin to interact with the world through emotions, which play a central role in early communication. Music is a particularly powerful medium for expressing and understanding emotions, and prior research has shown that even young children possess an innate ability to recognize the emotions conveyed through musical pieces. However, to fully grasp the development of this emotional understanding, it is crucial to investigate even younger populations. While the majority of studies on musical emotion perception have focused on children older than three years, only a limited number of studies have explored infants’ responses to emotional music, and the findings have been somewhat inconsistent. Our study aims to contribute to this growing area of research by examining whether infants can distinguish between different emotions in music. To do this, we measured changes in their pupil size while they listened to various types of emotionally expressive music, a method that offers objective insight into emotional arousal and cognitive engagement. In particular, pupil dilation is known to reflect heightened emotional or cognitive responses, often increasing in response to negative emotional stimuli compared to positive ones.
The study included 34 infants aged 5 to 9 months (mean age = 6.84 ± 1.57 months, with 13 females), all of whom had normal hearing. Two additional infants were unable to complete the experiment due to crying or falling asleep during the sessions. Each infant was exposed to a piece of happy music and a piece of sad music, presented in a counterbalanced sequence to control for order effects, while viewing a simple animated video designed for infants. We recorded the changes in pupil diameter using a Tobii Spectrum 600 eye-tracker. To validate the emotional content of the musical stimuli, we asked 83 adults between the ages of 16 to 50 to rate the music used in the study. Of the adult evaluators, 82% rated the happy music as happy or relatively happy, while 89% rated the sad music as sad or relatively sad. This helped ensure that the emotional content of the music was perceived as intended by adult listeners.
A 2×2 ANOVA examined the between-subjects variable of music order (happy first/sad first) and the within-subjects variable of music emotion type (happy/sad). There were no significant main effects for either order (F(1, 32)= .004, p > .05; BF_incl= .264) or music emotion (F(1, 32) = .187, p > .05; BF_incl= .320), but their interaction was significant (F(1, 32)=8.968, p < .01, η²= .219; BF_incl=12.865). For the group that heard happy music first, pupil dilation was smaller during sad music (paired-samples t-test, p < .05). For the group that heard sad music first, pupil dilation was smaller during happy music (paired-samples t-test, p = .062). Thus, pupil dilation was not directly linked to specific emotional content, and when the emotional type of music changed, participants' pupil dilation decreased.
The findings suggest that music that conveys distinct emotions to adults did not elicit significant changes in pupil size in infants. This finding implies that infants at this age may be unable to differentiate between happy and sad emotions in music. Interestingly, the reduction in pupil dilation following a change in music emotion suggests that infants may not yet be highly sensitive to emotional shifts in music, especially if they have not had sufficient time to become familiar with the musical stimuli.
These findings advance our understanding of how infants under one year old perceive emotional content in music. While some previous research suggests that infants can distinguish emotional music, it remains unclear whether they are responding to the acoustic features of the music or genuinely perceiving the emotions being conveyed. More definitive evidence of emotional perception seems to emerge between 12 and 18 months of age. Future research should focus on how infants within this age range process emotional information and social cues in music, possibly exploring whether music influences infants’ emotions and whether they use music to anticipate the emotions of others. This line of research could further elucidate the role of emotional prosody in the development of human emotional perception and social cognition.

Key words: music emotion, infant development, pupil dilation

摘要： 人类是否天生具有感知音乐情绪的能力?给5~9个月的婴儿播放快乐和悲伤音乐,比较其瞳孔直径变化量,探究婴儿的情感唤起程度是否有差异（以往研究表明婴儿在观看消极情绪刺激时情感唤起程度更高）。结果显示,婴儿瞳孔直径变化量在音乐情绪发生变化时显著降低,但整体上在听快乐和悲伤音乐时没有显著差异。该结果表明,5~9个月的婴儿的瞳孔反应尚未与音乐所表达的情绪建立明确的联系,说明婴儿感知音乐情绪的能力可能仍在发展之中。该研究提出了一种研究婴儿音乐情绪感知的新方法,并为人类情绪感知能力的早期发展规律提供了重要证据。

关键词: 音乐情绪, 婴儿研究, 瞳孔直径变化