McCullough et al. analyze revenge and reconciliation in terms of computation of risk and the welfare tradeoff ratio (WTR). Their functional analysis does not clarify how these mechanisms are implemented, but suggests that the same computational psychological mechanisms are always involved. This leaves out a significant component of a complete account.
Risk computation cannot explain why most of us are willing to forgo benefits, including fitness gains, to avoid the experience of betrayal. People will make different decisions on whether to take a risk, depending on whether the potential loss will be inflicted by a person or by some probabilistic machine with the same probability (for “betrayal aversion,” see Bohnet & Zeckhauser Reference Bohnet and Zeckhauser2004; Koehler & Gershoff Reference Koehler and Gershoff2003). The reason for this is that a perceived risk of betrayal triggers emotions which, in turn, influence decision making. Moreover, attitudes to trust and the impact of fear of betrayal are culture-dependent, indicating something more than a case by case cost–benefit analysis. In some cultures, betrayal is so aversive that essentially no risks are taken, and people will forgo significant potential gains attainable with substantial probability, in order to not incur the risk of betrayal (Bohnet et al. Reference Bohnet, Herrmann and Zeckhauser2010).
McCullough et al. analyze the different costs and benefits involved in exacting revenge from close as opposed to distant associates. Our work (Joskowicz-Jabloner & Leiser, forthcoming) demonstrates that trust violations evoke different patterns of emotions, depending on the closeness of the perpetrator. These emotions in turn predict avenues for reconciliation, as well as, crucially, the ineffectiveness of certain remedies to promote reconciliation. We established the existence of (at least) two distinct betrayal-domains that differ in the pattern of emotional response evoked, in the actions that relieve negative emotions and the influence of values held by the victim on the emotional response. In the social norms domain, betrayal evokes predominantly anger-related emotions that can be alleviated effectively, whereas in the personal domain more profound negative emotions are elicited by betrayal and fewer actions can relieve them.
For example, according to the analysis of McCullough et al., offers of compensation or hyper-compensation should be effective. In our study, participants did perceive monetary (hyper-) compensation as highly effective when the perpetrator was a stranger, but as an inappropriate remedy when the perpetrator was a friend. In fact, such offers to a friend compound the injury and make forgiveness more difficult – a situation described as a “taboo trade-off.” A taboo trade-off (Fiske & Tetlock Reference Fiske and Tetlock1997) is an incommensurable comparison between two values, in which one value is desecrated by being weighed against another, for example to attach “a monetary value to one's friendship” (p. 256). Because taboo trade-offs themselves trigger negative emotional, behavioral and cognitive responses, offering compensation for a betrayal in the personal domain is likely to be perceived by the betrayed as an additional act of betrayal. In terms of appraisal theory, the victim of personal betrayal is forced to undertake an ongoing assessment of the situation (Ellsworth & Scherer Reference Ellsworth, Scherer, Davidson, Goldsmith and Scherer2003). The betrayal victim's reappraisal of the relationship is first based on the betrayal incident and then by the offer of monetary compensation, which implies that the perpetrator seeks or sees an instrumental relation to the victim.
The reason for this finding is quite obvious: In the case of friends and close allies, the computations about WTR, past investment, and future expectations are mediated by the feelings or emotion of personal trust, that are absent with respect to strangers. Moreover, emotions such as personal trust or honor are of use beyond computation of WTR, on a cumulative or case by case basis, and relate to broader psychological structures, such as values. The value hierarchies espoused by people modulate the functioning of such emotions. McCullough et al. would of course not dispute this, but the consequences need to be spelled out. Let us take an example from our work: We found that the experience of trust-betrayal is influenced by the importance of the values “Power” (indicating a concern for self-interest) and “Benevolence” (expressing the concern for the welfare of others) (Schwartz Reference Schwartz, Tamayo and Porto2005) for the betrayal victim. Power amplifies the emotional response and also enhances the effectiveness of several relief actions, whereas benevolence may ease negative emotions, in particular emotions of anger and regret, but it strengthens the efficiency only of apology. Similarly, Stouten et al. (Reference Stouten, De Cremer and Van Dijk2005; Reference Stouten, De Cremer and Van Dijk2006) found that in the face of norm violators, “prosocials” (the benevolence-oriented; see De Cremer & Van Lange Reference De Cremer and Van Lange2001) are more concerned with fairness and conflict resolution, whereas “proselfs” (the power-oriented) focus on efficiency.
In view of these considerations, a complete account of revenge and forgiveness should go beyond WTR and incorporate several central constructs from social psychology. To explain specific behavioral patterns, it is necessary to identify the specific emotions that are used to tally and compute the relevant risks and benefits in each case. These emotions evolved for a variety of broad functions, out of which WTR computation is but one. They are differentially evoked and involved according to the circumstances, and related to yet broader psychological structures, such as values and values hierarchies.
McCullough et al. analyze revenge and reconciliation in terms of computation of risk and the welfare tradeoff ratio (WTR). Their functional analysis does not clarify how these mechanisms are implemented, but suggests that the same computational psychological mechanisms are always involved. This leaves out a significant component of a complete account.
Risk computation cannot explain why most of us are willing to forgo benefits, including fitness gains, to avoid the experience of betrayal. People will make different decisions on whether to take a risk, depending on whether the potential loss will be inflicted by a person or by some probabilistic machine with the same probability (for “betrayal aversion,” see Bohnet & Zeckhauser Reference Bohnet and Zeckhauser2004; Koehler & Gershoff Reference Koehler and Gershoff2003). The reason for this is that a perceived risk of betrayal triggers emotions which, in turn, influence decision making. Moreover, attitudes to trust and the impact of fear of betrayal are culture-dependent, indicating something more than a case by case cost–benefit analysis. In some cultures, betrayal is so aversive that essentially no risks are taken, and people will forgo significant potential gains attainable with substantial probability, in order to not incur the risk of betrayal (Bohnet et al. Reference Bohnet, Herrmann and Zeckhauser2010).
McCullough et al. analyze the different costs and benefits involved in exacting revenge from close as opposed to distant associates. Our work (Joskowicz-Jabloner & Leiser, forthcoming) demonstrates that trust violations evoke different patterns of emotions, depending on the closeness of the perpetrator. These emotions in turn predict avenues for reconciliation, as well as, crucially, the ineffectiveness of certain remedies to promote reconciliation. We established the existence of (at least) two distinct betrayal-domains that differ in the pattern of emotional response evoked, in the actions that relieve negative emotions and the influence of values held by the victim on the emotional response. In the social norms domain, betrayal evokes predominantly anger-related emotions that can be alleviated effectively, whereas in the personal domain more profound negative emotions are elicited by betrayal and fewer actions can relieve them.
For example, according to the analysis of McCullough et al., offers of compensation or hyper-compensation should be effective. In our study, participants did perceive monetary (hyper-) compensation as highly effective when the perpetrator was a stranger, but as an inappropriate remedy when the perpetrator was a friend. In fact, such offers to a friend compound the injury and make forgiveness more difficult – a situation described as a “taboo trade-off.” A taboo trade-off (Fiske & Tetlock Reference Fiske and Tetlock1997) is an incommensurable comparison between two values, in which one value is desecrated by being weighed against another, for example to attach “a monetary value to one's friendship” (p. 256). Because taboo trade-offs themselves trigger negative emotional, behavioral and cognitive responses, offering compensation for a betrayal in the personal domain is likely to be perceived by the betrayed as an additional act of betrayal. In terms of appraisal theory, the victim of personal betrayal is forced to undertake an ongoing assessment of the situation (Ellsworth & Scherer Reference Ellsworth, Scherer, Davidson, Goldsmith and Scherer2003). The betrayal victim's reappraisal of the relationship is first based on the betrayal incident and then by the offer of monetary compensation, which implies that the perpetrator seeks or sees an instrumental relation to the victim.
The reason for this finding is quite obvious: In the case of friends and close allies, the computations about WTR, past investment, and future expectations are mediated by the feelings or emotion of personal trust, that are absent with respect to strangers. Moreover, emotions such as personal trust or honor are of use beyond computation of WTR, on a cumulative or case by case basis, and relate to broader psychological structures, such as values. The value hierarchies espoused by people modulate the functioning of such emotions. McCullough et al. would of course not dispute this, but the consequences need to be spelled out. Let us take an example from our work: We found that the experience of trust-betrayal is influenced by the importance of the values “Power” (indicating a concern for self-interest) and “Benevolence” (expressing the concern for the welfare of others) (Schwartz Reference Schwartz, Tamayo and Porto2005) for the betrayal victim. Power amplifies the emotional response and also enhances the effectiveness of several relief actions, whereas benevolence may ease negative emotions, in particular emotions of anger and regret, but it strengthens the efficiency only of apology. Similarly, Stouten et al. (Reference Stouten, De Cremer and Van Dijk2005; Reference Stouten, De Cremer and Van Dijk2006) found that in the face of norm violators, “prosocials” (the benevolence-oriented; see De Cremer & Van Lange Reference De Cremer and Van Lange2001) are more concerned with fairness and conflict resolution, whereas “proselfs” (the power-oriented) focus on efficiency.
In view of these considerations, a complete account of revenge and forgiveness should go beyond WTR and incorporate several central constructs from social psychology. To explain specific behavioral patterns, it is necessary to identify the specific emotions that are used to tally and compute the relevant risks and benefits in each case. These emotions evolved for a variety of broad functions, out of which WTR computation is but one. They are differentially evoked and involved according to the circumstances, and related to yet broader psychological structures, such as values and values hierarchies.