Taking into account an evolutionary viewpoint, we hypothesize that music could hide a universal and adaptive code determining preferences. We consider the possible selective pressure that might have shaped, at least in part, our emotional appreciation of sound and music, and sketch a comparison between parameters of some naturalistic sounds and music.

With this commentary, I raise two issues relevant to the theoretical framework from a developmental perspective. First, the infants' emotional responses are induced by the music as well as by the multimodal information they perceive in interaction with their mothers, and these responses change with time. Second, contrary to what is suggested in the target article, musical expectancy is already experienced by young infants.

The mechanisms proposed in the target article are quite general and do not address variables specific for music. I argue that reactions to music include motivational mechanisms related to functions of music. To further the field, as the authors envision, consideration of internal mechanisms must be paired with specific hypotheses that include musical and musically relevant variables.

Juslin & Västfjäll (J&V) have suggested a promising theoretical framework for understanding musical emotions. However, the way they classify the hypothetical underlying psychological mechanisms suffers from certain weaknesses, both in principle and when it comes to details. It is proposed that the authors consider incorporating ideas from a recent dissertation that has advanced another multimechanism theory of musical emotions.

The target article inventories mechanisms underlying musical emotions. We argue that the inventory misses important mechanisms and that its structure would benefit from the distinction between two types of musical emotions. We also argue that the authors' claim that appraisal does not play a crucial role in the causation of musical emotions rests on a narrow conception of appraisal.

Are the neural systems involved in recognizing affective prosody in language also used for emotion recognition in instrumental music? One way to test this idea is to study musical affect perception in patients with receptive affective aprosodia (RAA). Music perception in RAA is totally unexplored and could provide a powerful way to test the idea that we perceive music as a kind of emotional voice.

Current research on emotional responses to dissonance has yielded consistent data in both developmental psychology and neuroscience. What seems to be lacking is a definition of what might constitute dissonance in non-musical domains. Thus, contrary to Juslin & Västfjäll's (J&V) proposal for the need to distinguish between six broad mechanisms, I argue that future research should rather focus on perceptual determinants of each basic emotion.

The emotional expression of laughter in opera scores and performance was evaluated by converting notation to temporal data and contrasting it with the conversational laughter it emulates. The potency of scored and sung laughter was assayed by its ability to trigger contagion in audiences.

Juslin & Västfjäll (J&V) think that all emotions aroused by music have the music itself as their “intentional object.” Some of the mechanisms they discuss almost certainly involve both cognitive appraisals and intentional objects. But some of the mechanisms are non-cognitive: they involve neither cognitive appraisals nor intentional objects. Partly for this reason they may produce moods rather than emotions proper.

We wonder about tying the universal appeal of music to emotion as defined by psychologists. Music is more generally about feelings, and many of these, such as moods and pleasures, are central to the enjoyment of music and fall outside the domain of emotion. The critical component of musical feelings is affective intensity, resulting from syntactically generated implications and their outcomes.

A basic aspect of emotional responding to music involves the liking for specific pieces. Juslin & Västfjäll (J&V) fail to acknowledge that simple exposure plays a fundamental role in this regard. Listeners like what they have heard but not what they have heard too often. Exposure represents an additional mechanism, ignored by the authors, that helps to explain emotional responses to music.

We disagree with Juslin & Västfjäll's (J&V's) thesis that music-evoked emotions are indistinguishable from other emotions in both their nature and underlying mechanisms and that music just induces some emotions more frequently than others. Empirical evidence suggests that frequency differences reflect the specific nature of music-evoked emotions: aesthetic and reactive rather than utilitarian and proactive. Additional mechanisms and determinants are suggested as predictors of emotions triggered by music.

We present evidence from neuroimaging and brain lesion studies that emotional contagion may not be a mechanism underlying musical emotions. Our brains distinguish voice from non-voice sounds early in processing, and dedicate more resources to such processing. We argue that super-expressive voice theory currently cannot account for evidence of the dissociation in processing musical emotion and voice prosody.

We propose that the six mechanisms identified by Juslin & Västfjäll (J&V) fall into two categories: signal detection and amplification. Signal detection mechanisms are unmediated and induce emotion by directly detecting emotive signals in music. Amplifiers act in conjunction with signal detection mechanisms. We also draw attention to theoretical and empirical challenges associated with the proposed mechanisms.

Instead of the discrete emotions approach adopted by Juslin & Västfjäll (J&V), the present perspective considers musical signals as functioning primarily to influence listeners in ways that are favorable to the signaler. Viewing music through the lens of social-emotional regulation fits with typical uses of music in everyday contexts and with the cross-cultural use of music for infant affect regulation.

There is certainly a need for a framework to guide the study of the physiological mechanisms underlying the experience of music and the emotions that music evokes. However, this framework should be organised hierarchically, with musical anticipation as its fundamental mechanism.

Most commentators have agreed with our thesis, that musical emotions cannot be studied without regard to underlying mechanisms. However, some commentators have expressed concerns that are addressed in this response. Others have suggested directions for future research. Topics discussed in our response include terminology, elaborations on particular mechanisms, possible additional mechanisms, ways of distinguishing among emotions and mechanisms, the prevalence of musical emotions, the relationship between perceived and felt emotions, developmental issues, and evolutionary perspectives. We end our response with a plea for researchers to reach beyond the traditional disputes in the field to pave the way for more theory-driven studies that can facilitate a deeper understanding of musical emotions.