McCullough et al. posit that revenge is an evolved cognitive mechanism meant to reduce the likelihood that a victim will incur future harm. Specifically, they discuss the ways in which this rational system (a deliberative calculation of the Welfare Tradeoff Ratio) operates in individuals and modern cultures. However, although they include affect in their model as the “proximate motivating system” (sect. 3.1.3, para. 10), their model assumes that the role of affect in decision-making is superseded by the cognitive mechanism responsible for calculating costs and benefits (see sect. 3.1.3, para. 11). We review social psychological research suggesting that revenge is far less deliberative than the authors argue and results from more emotionally-driven reactions and biased perceptions. In particular, such research has shown that emotions are the primary drivers of decisions to seek revenge, and that individuals are insensitive to deterrence motives.

Individuals frequently respond to injustice with anger, outrage, and a heightened motivation to retaliate against the perpetrator (Darley & Pittman Reference Darley and Pittman2003; Carlsmith & Darley Reference Carlsmith, Darley and Zanna2008). Inasmuch as deliberative cognitions play a role in the processes leading to revenge, they are likely to be influenced by the decision-maker's emotions (Schwarz Reference Schwarz2000; Schwarz & Clore Reference Schwarz and Clore1983).

Some theorists have argued that anger is the primary emotion involved in responses to injustice, leading to an increased tendency to support the punishment of offenders (Darley & Pittman Reference Darley and Pittman2003; J. S. Lerner et al. Reference Lerner, Goldberg and Tetlock1998). Anger not only increases tendencies toward punitive action but also reduces cognitive processing, leading to heuristic-based judgments with stronger attributions of blame and desires for retribution (Goldberg et al. Reference Goldberg, Lerner and Tetlock1999). Moreover, anger induced by events unrelated to a transgression can increase the desire for punishment, suggesting that even incidental emotions shape justice judgments and responses. In these instances, “hot” emotional states override deliberate cognition and decision-making (Loewenstein Reference Loewenstein1996). Avengers who are victims of the transgression grant themselves a special claim to justice (Zitek et al. Reference Zitek, Jordan, Monin and Leach2010), and are particularly likely to experience such emotional states.

Consistent with research on the tension between affect and deliberation (Zajonc Reference Zajonc1968), research on moral intuitions suggests that individuals express moral disapproval and outrage primarily as a result of emotional reactions, and only later develop and express cognitive justifications (Haidt Reference Haidt2001). More recent work has suggested that even such apparently ideological variables as political attitudes (Graham et al. Reference Graham, Haidt and Nosek2009) and voting behavior (Inbar et al. Reference Inbar, Pizarro, Iyer and Haidt2012) are influenced by individuals' propensity to experience disgust. Similarly, research from outside the moral domain has established that when asked to elaborate on their decisions, individuals are quite inaccurate when it comes to describing the reasons for their actions (Nisbett & Wilson Reference Nisbett and Wilson1977). In sum, any cognitive justification that individuals claim motivates revenge (e.g., deterrence) may be post-hoc and less predictive of their decision than their emotional reactions.

Taken together, these findings run counter to the idea that revenge is based on a rational economic calculus meant to reduce the future likelihood of repeated wrongdoing. Although individuals may believe that they seek revenge for primarily utilitarian reasons, this research suggests that they may actually be relatively insensitive to probabilities of deterrence and are instead driven to punish by more emotion-based motives (Carlsmith et al. Reference Carlsmith, Darley and Robinson2002). Indeed, some research has supported this contention specifically for judgments of blame and punishment. Such research has found that individuals are more likely to blame and punish offenders whose intentions evoke moral outrage than those who perpetrate the same harmful acts for other reasons (Cushman Reference Cushman2008); this runs contrary to the deterrence perspective, according to which people should be insensitive to motives so long as the probabilities of future wrongdoing are comparable. For instance, individuals attribute more blame and causation to a transgressor (who, e.g., injured another driver in an accident) if the reason for the outcome was attributed to an undesirable behavior (e.g., speeding home to hide a vial of cocaine) versus a desirable one (e.g., speeding home to hide an anniversary present) (Alicke Reference Alicke1992). In other research suggesting that punishers are not driven by cognitively based deterrence calculations, individuals confronted with incontrovertible evidence that capital punishment is an ineffective deterrent nonetheless maintained strong support for capital punishment (see Carlsmith et al. Reference Carlsmith, Darley and Robinson2002; Ellsworth & Ross Reference Ellsworth and Ross1983; Vidmar & Miller Reference Vidmar and Miller1980). Thus, individuals seem to be motivated to punish based primarily on the degree of their outrage, not the potential incapacitation of the offender (Darley et al. Reference Darley, Carlsmith and Robinson2000; Kahneman et al. Reference Kahneman, Schkade and Sunstein1998).

Certainly, any decision-making process is likely to contain both cognitive and affective processes (Schwarz Reference Schwarz2000). Indeed, there are contexts in which revenge may be a more deliberative “cold” process (Bies & Tripp Reference Bies, Tripp, Kramer and Tyler1996), and we do not claim that revenge is always and only emotionally driven. However, we do maintain that victims' affective reactions to wrongdoing or injustice, such as anger, outrage, and resentment, are likely to be stronger predictors of their revenge responses than are their cognitive calculations of future benefits and costs.

In light of the research reviewed above, we believe McCullough et al.'s analysis does not capture the full extent to which emotions serve as predictive causes of revenge and punishment. While the authors remind us that evolved mechanisms may not manifest themselves in identical ways for which they were selected, they do provide examples of present-day behavior (e.g., recent data from economic games) as support for their evolved mechanism. In contrast, the social psychological research on emotions, moral decision-making, and punishment which we have reviewed paints a picture of revenge as more of a hot, reflexive reaction than a cold, judicious decision.