Juslin & Västfjäll (J&V) acknowledge in their concluding remarks that scientists will generally find their proposal uncontroversial. I concur. As a neuropsychologist, I find the proposed framework sound and clear. One provocative claim, however, is that the field has made little progress due to the lack of control for mechanisms underlying musical emotions. I do not share this view. For example, emotional responses to dissonance have been studied for a long time in musical acoustics, and the studies have yielded consistent data in both developmental psychology and neuroscience. In what follows, I examine to what extent the six proposed mechanisms by J&V are needed to understand emotional responses to dissonance. I also highlight what remains to be understood to develop a theory of emotional responses to dissonance, in particular, and to unpleasant sounds, in general.
Perception of dissonance is a striking and distinct experience in music listening. To experience it, it suffices to play together two adjacent keys on a keyboard – forming a minor second in musical terminology – or to imagine the initial tuning of an orchestra. Both sound complexes are dissonant, and usually judged unpleasant by ordinary listeners. In contrast, striking together two keys that lie 12 keys (i.e., an octave) apart or hearing the first chord of Beethoven's Fifth Symphony produce sound experiences that are usually judged pleasing or consonant. Ordinary listeners, including infants, easily distinguish consonant from dissonant pitch combinations and consider the former more pleasant than the latter (Hannon & Trainor Reference Hannon and Trainor2007; Plomp & Levelt Reference Plomp and Levelt1965). Even musicians who sometimes claim that dissonance can be highly pleasant, exhibit enhanced electrodermal activity in response to dissonant music as compared to the consonant versions of the same pieces (Khalfa & Peretz Reference Khalfa and Peretz2004).
The mechanism underlying responses to dissonance arises from the roughness created at the level of the basilar membrane in the inner ear. The overlap in vibration patterns compromises the resolution of pitches of different frequency on the basilar membrane, leading to beating and the perception of roughness (von Helmholtz Reference von Helmholtz and Ellis1954). Hence, in the absence of detailed analysis of the music, a sudden dissonant event may elicit a rapid, reflex-like reaction in the neural pathway. This mechanism is referred to as a brain stem reflex by J&V (sect. 3.1.1). Neuropsychologists would rather qualify this type of fast emotional responses as a subcortical reflex, to acknowledge the possibility that the response may arise from a number of different subcortical structures, not just the brain stem.
There is ample evidence from both neuroimaging and brain lesion studies that subcortical structures are involved in emotional responses to dissonance. Neuroimaging studies have identified the parahippocampal gyrus (Blood et al. Reference Blood, Zatorre, Bermudez and Evans1999; Koelsch et al. Reference Koelsch, Fritz, von Cramon, Müller and Friederici2006) and the amygdala (Ball et al. Reference Ball, Rahm, Eickhoff, Schulze-Bonhage, Speck and Mutschler2007; Koelsch et al. Reference Koelsch, Fritz, von Cramon, Müller and Friederici2006) as key brain structures. Lesion data have confirmed the critical involvement of the parahippocampal cortex rather than the amygdala (Gosselin et al. Reference Gosselin, Samson, Adolphs, Noulhiane, Roy, Hasboun, Baulac and Peretz2006). Thus, and contrary to J&V's claims, dissonance has been controlled for in these studies and the available data converge nicely on the involvement of one critical brain structure. More importantly, the data do not suggest that different mechanisms are involved in the emotional responses to dissonance or that the underlying mechanisms have been neglected.
Such reflex-like responses to dissonance may be innate or result from evaluative conditioning (J&V's mechanism 2 [sect. 3.1.2]). Indeed, there are early and multiple opportunities for consonance calibration at all levels along the auditory pathways. Consonance is ubiquitous in the auditory environment. Most natural sounds, and speech in particular, are composed of consonant harmonic intervals. Thus, central neural networks may be preferentially attuned to consonant intervals by a process of generalization because of their prevalence or biological significance in the environment, and not simply because of hard-wired constraints of the peripheral hearing system. Here again, the data are consistent and informative. In three different laboratories, infants have been shown to prefer consonance over dissonance (Masataka Reference Masataka2006; Trainor & Heinmiller Reference Trainor and Heinmiller1998; Zentner & Kagan Reference Zentner and Kagan1996). More importantly, this preference for consonance is not dependent on prenatal or early postnatal experience. Hearing newborns from deaf parents prefer consonance over dissonance (Masataka Reference Masataka2006). Thus, the mechanism underlying the unpleasant sounding of dissonance appears to be innate. Here, too, there is no need to consider additional mechanisms underlying dissonance to resolve inconsistencies.
What is needed, however, is an account for the fact that preference for consonance (or avoidance of dissonance) appears specific to humans. Although many species discriminate dissonance from consonance, nonhuman primates (tamarins) do not prefer consonant over dissonant chords (McDermott & Hauser Reference McDermott and Hauser2004). Rather, tamarins prefer species-specific feeding chirps over species-specific distress calls. The latter category of sounds, suggestive of danger, is qualified as dissonant by J&V (sect. 3.1.1). Indeed, the increased tension of the vocal chords in distress calls introduces many subharmonics in the vocal sounds that are otherwise harmonic signals. These distortions in screams sound unpleasant to both humans and animals and would trigger an avoidance response in the listeners (who will try to stop the screams). Thus, there would be an evolutionary advantage to have such dissonant screams in the vocal repertoire, especially in the youngsters (Fitch et al. Reference Fitch, Neubauer and Herzel2002). Nonetheless, the research with tamarins suggests that dissonance in music and dissonance in vocal sounds might be different phenomena. What is needed is a definition of the acoustical correlates of dissonance so that we can measure it in non-musical sounds (J. McDermott, personal communication). In other words, we need to learn more about the acoustical properties of emotional vocal sounds in order to understand better the neurobiological mechanism that accounts for the avoidance of dissonance.
The need to examine perceptual determinants of emotional responses within and across domains should not be minimized. They provide grounds to generalize a theory. For example, within the musical domain, one would think that dissonance is the major perceptual determinant of threat (consider the soundtrack of Hitchcock's Psycho [1960]). Yet, current research points to the recruitment of a different neural structure in response to dissonance (i.e., the parahippocampal cortex) compared to musical threat (i.e., the amygdala; Gosselin et al. Reference Gosselin, Peretz, Noulhiane, Hasboun, Beckett, Baulac and Samson2005; Reference Gosselin, Samson, Adolphs, Noulhiane, Roy, Hasboun, Baulac and Peretz2006; Reference Gosselin, Peretz, Johnsen and Adolphs2007). More research is needed to identify the perceptual determinants of scary music and to distinguish the neural correlates of perception from those of emotion. Similarly, defining perceptual determinants of dissonance in vocal communication would enable us to examine the relevance of emotional contagion, visual imagery, and episodic memory (J&V's mechanisms 3, 4, and 5) in driving emotional responses to dissonance. If dissonance is to music what screams are to vocal communication, one can easily conceive how these broad categories of mechanisms might come into play. For the moment, consideration of these mechanisms is premature. More work is needed to specify the acoustical determinants of unpleasantness in particular, and the perceptual triggers of emotions in general.
