What does it mean to say emotions and attitudes are conceptually distinct? Prejudice, like contempt, is best understood as a general evaluation rather than a specific emotion. This perspective departs from the view of prejudice as specific emotion adopted by several recent theories of prejudice (Cottrell & Neuberg Reference Cottrell and Neuberg2005; Cuddy et al. Reference Cuddy, Fiske and Glick2007; Fiske et al. Reference Fiske, Cuddy, Glick and Xu2002; Mackie et al. Reference Mackie, Devos and Smith2000). Prejudice, like contempt, is an intergroup sentiment. It is an affective representation of a social group's relational value – friend or foe?

A prejudice-as-sentiment view seamlessly adopts many of the features of the Attitude–Scenario–Emotion (ASE) model presented by Gervais & Fessler (G&F) in the target article. First, the attitudinal component of prejudice is an object-specific affective representation of a social group. It represents the perceiver's “feelings about” the group, and it can be “coldly” considered, as evidenced by the common use of self-report measures in prejudice research (e.g., Brigham Reference Brigham1993; McConahay Reference McConahay, Dovidio and Gaertner1986). People usually complete these measures in private, reporting their evaluation of a group when its members are not present.

Second, the affective representation of prejudice is enduring. Both explicit and implicit prejudice measures predict subsequent discrimination (e.g., Fazio et al. Reference Fazio, Jackson, Dunton and Williams1995; McConnell & Leibold Reference McConnell and Leibold2001), even when the measurement of attitude and behavior is separated by as much as one week (Dasgupta & Rivera Reference Dasgupta and Rivera2006). The same group may at times be seen as “lazy” and other times “dangerous,” but the underlying representation of affect is equally negative in both scenarios. The stable time course of the attitude should be taken as evidence of the generality of the representation (McDougall Reference McDougall1933).

Third, prejudice can moderate diverse emotions across scenarios. Prejudice-as-specific-emotion theorists are right to point out that different emotions are often experienced in relation to different groups (Cottrell & Neuberg Reference Cottrell and Neuberg2005; Mackie et al. Reference Mackie, Devos and Smith2000). African Americans may be stereotyped as dangerous and elicit fear, while gay men may be stereotyped as immoral and elicit disgust. But rather than claiming that groups eliciting fear and groups eliciting disgust represent different kinds of prejudice, a prejudice-as-sentiment model recognizes a common attitudinal core at the heart of all prejudices. (This allows for the sensible possibility that different contexts will lead to different emotions toward a group – surely we are excited rather than fearful when aggressive people join our team.)

G&F present two main hypotheses associated with the ASE model. The first is that there are many distinct sentiments, and the second is that sentiments are emotionally pluripotent. To garner support for the first hypothesis, G&F cite a central idea in evolutionary theory (e.g., Neuberg & Cottrell Reference Neuberg and Cottrell2008), noting there are many distinct threats and opportunities which determine a person's social-relational value. Therefore, they argue, sentiments are more complex than good/bad or like/dislike. Although this hypothesis seems appropriate for the interpersonal context, an extension to the intergroup context would mean there are many distinct prejudices. This cannot be true if prejudice is a general evaluation.

Indeed, it should be possible to represent a group's value along the single dimension of friend/foe. To be sure, evaluative judgments can be made along other dimensions as well (e.g., warmth, competence, status; Fiske et al. Reference Fiske, Cuddy, Glick and Xu2002), but these judgments come later. A prejudice-as-sentiment model easily accommodates this prediction. It conceives of prejudice as a general evaluation of the outgroup's value in relation to the ingroup. At the highest level, prejudice marks a group as friend or foe. This core affective representation then shapes the way the group is perceived (e.g., dangerous, trustworthy, warm), depending on the information available in a particular intergroup context (Bahns Reference Bahns2017).

The second hypothesis associated with the ASE model, the idea that sentiments are emotionally pluripotent, is fundamental to the argument that prejudice is a general evaluation. Emotional pluripotence means the same attitude can moderate a diverse range of emotions toward its object across situations. The hierarchical structure of a sentiment helps to account for the fact that multiple emotions may be elicited in the same intergroup context (see Cottrell & Neuberg Reference Cottrell and Neuberg2005); there need not be a one-to-one correspondence between emotions and target groups. For example, groups that threaten the physical safety of the ingroup may evoke both fear and anger, depending on the immediacy of the threat. The core affective representation of the group is the same – the group is a foe – but the emotions and behaviors directed toward the group may differ from one context to another.

One important implication of adopting a prejudice-as-sentiment model is that the particular stereotypes and emotions associated with a group can shift in different scenarios – they are not part of the enduring affective representation. Different stereotypes may be activated in different intergroup contexts, but each serves the same basic function of justifying the negative evaluation and appropriately directing behavior (Crandall & Eshleman Reference Crandall and Eshleman2003). In this way, stereotypes can be understood as justifications of prejudice rather than as forming the source of the prejudice itself (Crandall et al. Reference Crandall, Bahns, Warner and Schaller2011).