The objective of the target article is to present an inventory of mechanisms that can cause musical emotions. The authors have succeeded in covering a broad range of mechanisms that are important in the causation of musical emotions. A first point we would like to make is that the inventory can be further improved in terms of both exhaustiveness and structure. Mechanisms that are overlooked are those underlying the mere exposure effect (Zajonc Reference Zajonc1968) and those underlying emotions that constitute a reaction to, rather than an imitation of, the emotions expressed in music. With regard to structure, we suggest that the organization of the inventory would benefit from distinguishing between two types of musical emotions (cf. Bell Reference Bell1914; Frijda Reference Frijda1986; Gabriel & Crickmore Reference Gabriel and Crickmore1977; Langer 1942; Payne Reference Payne1980; Swanwick Reference Swanwick1975), which we define below.

A first type is that of musical emotions that have as their object the music itself (i.e., music considered on the reality level). This type of musical emotions involves an appreciation of the music; the music is liked or disliked. The presence/absence of liked/disliked music can lead to positive emotions such as happiness, admiration, and relief. The absence/presence of liked/disliked music can lead to negative emotions such as anger or sadness, depending on the presence of other conditions such as coping potential and agency. A second type of musical emotions has as its object the content of the music, that which is represented or invoked by the music (i.e., music considered on the symbolic level). Music can represent emotions such as sadness, anger, fear, and happiness, or it can represent emotion-eliciting events such as hurricanes and carnivals. Musical emotions of the second type can be similar to the emotion expressed in the music, or they can constitute a reaction to the emotion or event invoked by the music.

Distinguishing between both types of musical emotions is important because it helps disentangle mechanisms responsible for the causation of musical emotions. For Type 1 emotions, we should consider mechanisms involved in liking/disliking of musical stimuli and those involved in evaluating the presence/absence of liked/disliked stimuli. Liking/disliking can be accounted for by mere exposure (music is liked more after multiple hearings), evaluative conditioning (music is liked better/less when paired with positive/negative events), and appraisal (music is compared with aesthetic goals or standards). Appraisal seems also suited for the further evaluation of the presence/absence of liked/disliked stimuli, as well as additional information about agency (e.g., who stopped the music?) and coping options (e.g., can we start the music again?).

The causation of Type 2 emotions consists of two steps. In a first step, the listener perceives an emotion or an emotion-eliciting event expressed in the music. For example, a piece of music can be perceived as expressing sadness, or it may evoke in the listener the image of a hurricane. In a second step, the emotion or emotion-eliciting event represented in the music evokes in the listener a similar emotion or an emotion that constitutes a response to it. For example, a sad piece of music can elicit sadness, and a musical fragment that expresses anger or that evokes the image of a hurricane can elicit fear. The two steps in the elicitation of Type 2 emotions rest on different mechanisms. In the first step, perceiving the emotional content of a piece of music can be based on similarities between features of the music and features of emotions. For example, the slow tempo of sad music is similar to the slow tempo of sad speech and behavior. These similarities can be picked up with episodic memory and imagery. In the second step, contagion accounts for emotions that are similar to the emotion expressed, and appraisal accounts for emotions that constitute a response to the emotion or emotion-eliciting event expressed in the music.

According to our analysis, both types of musical emotions seem to recruit appraisal at some point. This brings us to our next point. The authors of the target article have argued that appraisal is not crucially involved in the causation of musical emotions. We argue that this argument rests on a narrow conception of appraisal. If appraisal is understood according to current theoretical consensus, there is no reason to minimize the role of appraisal. Appraisal theories assume that emotions are caused by a process of appraisal in which the stimulus is evaluated on a number of variables, such as goal relevance, goal conduciveness, coping potential, and agency. Values on these variables form a pattern of appraisal, and this pattern determines the quality of the ensuing emotion. For example, fear occurs when a stimulus is appraised as goal relevant, goal inconducive, and difficult to cope with. Anger occurs when a stimulus is appraised as goal relevant, goal inconducive, and caused by another person. According to modern appraisal theories (Clore & Ortony Reference Clore, Ortony, Nadel, Lane and Ahern2000; Leventhal & Scherer Reference Leventhal and Scherer1987; Smith & Kirby Reference Smith, Kirby, Scherer, Schorr and Johnstone2001), appraisal can take the shape of (a) a comparison operation, (b) the activation or reinstatement of previous appraisal outcomes, or (c) activation of innate associations between stimuli and emotions. In addition, appraisal can be automatic (i.e., occur under suboptimal conditions) or nonautomatic (i.e., occur only under optimal conditions).

Following our conceptualization, appraisal is involved in both types of musical emotions discussed earlier. Moreover, it should be considered playing a role in some of the mechanisms discussed in the target article. For example, brain stem reflexes, to the extent that they signal inconduciveness with the goal of safety, can be considered an example of appraisal of the form described in (c) above. Another example is Meyer's (Reference Meyer1956) musical expectancy mechanism. We argue that the violation/confirmation of musical expectations does not directly lead to positive/negative emotions. Otherwise, after multiple hearings when a piece of music becomes entirely predictable, it would lose its emotion-eliciting power. We argue instead that the violation/confirmation of musical expectations is similar to the violation/confirmation of goals in real life (cf. the first step in Type 2 emotions) and that appraisal of it can lead to emotions (cf. the second step in Type 2 emotions).