Contempt as non-prototypical emotion

G&F find support for eight different and coherent features of contempt from the literature, but still conclude that contempt cannot be considered a (basic) emotion. We do not see this literature on contempt as perplexing or inconsistent. Our review took as its basis a prototype view (Russell Reference Russell1991b) in which non-prototypical emotions lack one or more key traits. For example, we argued that contempt may lack hot physiology and a clear term in English, but its appraisals, action tendencies, and expressions are distinctive. Contempt thus can be seen as a non-prototypical emotion (similar to other non-prototypical emotions like guilt or worry). This more parsimonious account shows that the literature does not lead inevitably to the model proposed by G&F.

Contempt as a sentiment, not an emotion

In our review, we argue that contempt can be both emotion and sentiment, based on Frijda et al.'s (Reference Frijda, Mesquita, Sonnemans, Van Goozen and Strongman1991) definition of sentiment as an enduring object–emotion association, that is, a disposition to respond emotionally to a certain object. G&F assume that contempt can only be seen as an enduring sentiment, organizing more momentary related emotions such as anger and disgust around “a common attitudinal core” (sect. 3.3, para. 3). However, many other emotions can become enduring object-linked emotional sentiments, such as anger (e.g., Giner-Sorolla Reference Giner-Sorolla2012; Halperin & Gross Reference Halperin and Gross2011), disgust (Ortony et al. Reference Ortony, Clore and Collins1990; Olatunji et al. Reference Olatunji, Forsyth and Cherian2007), envy, or admiration (Fiske et al. Reference Fiske, Cuddy, Glick and Xu2002; Harris & Fiske Reference Harris and Fiske2007). In other words, any emotion can become a sentiment when it becomes temporally extended, characterized by enduring changes in one's relation with others. There is no evidence that one of these is a “master sentiment” leading the other mere emotions along.

According to G&F, because contempt is a sentiment, it cannot also be an emotion, elicited by a single action of a person. We think that this assumption makes a category error: yes, contempt represents negative traits in others, but need not itself arise from an enduring attitude; one action can suffice. For example, a son who has a very good relationship with his parents hears that they have decided to divorce because his father betrayed his mother by sleeping with another woman, and he feels contempt towards his father. No pre-existing negative attitude is required to feel contempt for his father. On the contrary, the experience of contempt may lead to an attitude change towards his father, and he may not have the same relationship with his father ever again. This story, as reported by a respondent in one of our studies (Fischer & Roseman Reference Fischer and Roseman2007), clearly illustrates that contempt can be an immediate response.

Although it is true that various studies have shown that contempt is more likely to endure than anger, this is plausibly because its eliciting factors are more likely to endure (a single act versus a character), and not a reason to deny contempt's manifestation as an emotion. In fact, it is hard to see how evidence that contempt has distinctive expressions (e.g., the unilateral lip curl [Matsumoto & Ekman Reference Matsumoto and Ekman2004], or clucking or tutting noises [Hawk et al. Reference Hawk, van Kleef, Fischer and van der Schalk2009]) fits the view of contempt as a recruiter of diverse, context-sensitive feelings and expressions. What transitory emotion is recruited, if not contempt itself, when the contempt sentiment leads to such displays?

The Attitude–Scenario–Emotion model of sentiments

While Frijda and colleagues' definition of sentiment is easy to operationalize – does an object elicit the same emotion at different times or when presented out of context – G&F propose an Attitude–Scenario–Emotion model of sentiments. This model's components have many hard-to-verify, hard-to-falsify or overlapping features (see G&F's Table 2 in the target article), for example, “diverse motives, behaviors, and expressions across scenarios” (said of sentiments in their Table 2, but also potentially true of emotions or attitudes). In fact, the authors assert that sentiments are “higher-level functional networks of attitudes and emotion; each sentiment is an attitude state and the various emotions disposed by that representation” (sect. 4.2, para. 6). The interrelatedness and fuzzy boundaries of three core concepts of this model stand in sharp contrast with the authors' aim to disentangle folk concepts and the neurobiological basis of affect systems.