The authors describe two emotions that they characterize as “mixed,” being moved and suspense, as points in case for their second component of Embracing. We have studied being moved ourselves, and focus our comments on it. We feel that the evidence collected by ourselves and others does not fit with the characteristics of being moved that the authors propose, which challenges the second pillar of the Embracing factor of the model. So, what are those characteristics?

We found three relevant empirical claims in the model. The first is that being moved is a mixed emotion, supposedly meaning that it is defined as having concurrently both positive and negative affective components (cf. Deonna Reference Deonna, Tappolet, Teroni and Ziv2011). The second is that its relation to a corresponding negative emotion, sadness, is one of metonymy. The third is that sadness and being moved cluster in emotion space. We present evidence against all three claims.

The claim that being moved is a mixed emotion appears to be based on evidence gathered in studies that ask about emotional experiences retrospectively, often much later. Such data are useful for many purposes, but cannot clarify whether positive affect and negative affect were experienced concurrently or in succession. In a recent study, we concurrently and continuously assessed reports of being moved or touched, happiness, and sadness (among other variables [Schubert et al. Reference Schubert, Zickfeld, Seibt and Fiske2016]). Across each of six different video clips, the time series of feeling moved or touched cross-correlated consistently and strongly with that of happiness. However, the cross-correlation of feeling moved or touched with sadness varied highly between clips and was very small overall. To the extent that happiness and sadness can be used as proxies for affect, these data are not in line with the idea that positive affect and negative affect consistently co-occur in being moved. (For a visualization of the data, see http://www.kamamutalab.org/visual/visual.html.)

The second claim assumes a relation of metonymy between sadness and being moved. As we understand it, this means that sadness is referred to as being moved and thereby acquires some of the positivity inherent in the mixed emotion being moved. However, the precise implication of the linguistic concept metonymy remains unclear from the theory presented. People do not always label their emotional experiences while experiencing them, and any assumption based on labeling would only be valid for a subset of experiences. Of course, there are transfer effects from one emotion to the next, as well as contrast effects. These are described in the first component of the Embracing factor, the dynamic unfolding of the story. In the case of being moved, well-orchestrated videos evoking being moved usually evoke some negative emotion first, often sadness, but sometimes anxiety or fear, which set the viewer up for then being moved. Here, being moved “profits from” sadness, and not the other way around, because being moved follows sadness rather than preceding it (see also Schubert et al. Reference Schubert, Zickfeld, Seibt and Fiske2016). This profit is not due to metonymy, but rather the law of comparative feeling (Frijda Reference Frijda1988).

The third assumption is that being moved and sadness (related to social bonding and attachment) are highly associated and cluster closely together. According to our reading, cluster analyses on German vernacular emotion terms suggest that sadness does not cluster with being moved, but instead with other negative emotions. Being moved instead clusters closest with compassion, but also with nostalgia, admiration, tenderness, gratitude, and other positive emotion terms (Kinateder Reference Kinateder2011; Menninghaus et al. Reference Menninghaus, Wagner, Hanich, Wassiliwizky, Kuehnast and Jacobsen2015b; Schmidt-Atzert & Ströhm Reference Schmidt-Atzert and Ströhm1983). The close relationship of being moved and compassion is corroborated by a robust correlation between the trait of empathic concern and feeling moved by a large variety of stimuli (Zickfeld et al. Reference Zickfeld, Schubert, Seibt and Fiske2017).

Taken together, this implies that being moved is not a mixed emotion, but rather a positive one that enhances the enjoyment of art that elicits it. It does so by contrasting with negative emotions, especially sadness over the loss or suspension of social bonds and, thus, the process described in the first component of the Embracing factor.

The question remains why people feel moved and why being moved contrasts with sadness. According to our kama muta theory (Seibt et al. Reference Seibt, Schubert, Zickfeld and Fiske2017), the relation of being moved and social sadness is one of polar opposites: Whereas being moved is elicited by the sudden intensification of communal sharing relations, social sadness is about their attenuation or loss. Communal sharing is a basic building block of human social relations (Fiske Reference Fiske1992). This hypothesis can account for the data presented in the target article. To get back to one of Menninghaus et al.'s examples: At funerals, people often oscillate between contemplating their loss versus their feelings of connection and. Thus. between grief and being moved. It is fully possible to feel both at once, or in quick succession.

One reason for our divergent perspective may be our belief that scientific conceptualizations of emotions should be distinguished from vernacular emotion terms and folk concepts. The use of the vernacular terms being moved, being touched, and their counterparts in other languages does not perfectly correspond to the emotional state caused by the intensification of communal sharing. Sometimes these terms are used for other states, and states that fit this appraisal are not labeled with them. Furthermore, neither does sadness match perfectly the state caused by the loss of a communal relation. We believe one cannot build theories on specific emotions by equating vernacular emotion labels with emotions themselves. Likewise, linguistic concepts such as metonymy may be applied to emotion labels, but not to emotions pars pro toto.