The adaptationist analysis of revenge and forgiveness as offered by McCullough et al. provides a very welcome overarching theoretical framework to better understand these concepts. As the authors note, empirical research on revenge and forgiveness has been rather scattered, and mainly driven by mini-theories. The authors have done an impressive job to integrate a host of previously isolated findings to support their evolutionary approach to forgiveness and revenge. Their analysis leads to several interesting testable predictions about when people will be more, or less, likely to take revenge, and when they are more likely to forgive a transgressor.

The target article concludes with proposing a computational model that helps humans to decide whether to take revenge or forgive an offender, or essentially, which response offers the most fitness benefits. Ultimately this decision depends on perceived relationship value (as indexed by psychological constructs such as closeness and commitment) and perceived future exploitation risk, and their interaction. As cited by the authors, research with both human and nonhuman subjects has revealed strong support for the relationship value prediction (e.g., Finkel et al. Reference Finkel, Rusbult, Kumashiro and Hannon2002; Karremans & Aarts Reference Karremans and Aarts2007; Watts Reference Watts2006). Recently, in line with their evolutionary argument, we have demonstrated that the positive association between interpersonal closeness and forgiveness is robust across several different (both independent and interdependent) cultures – albeit with some cross-cultural variation regarding the strength of the association (see Karremans et al. Reference Karremans, Regalia, Paleari, Finham, Cui, Takada, Ohbuchi, Terzino, Cross and Uskul2011).

However, although the theorized computational system offers a very useful tool for understanding when and why humans forgive or take revenge, less attention is paid to the how of revenge and forgiveness. Recent studies have provided important insights into the basic processes that lead to forgiveness, demonstrating that executive functioning – in particular the cognitive ability to control and inhibit impulsive responses (as assessed with Stroop-like measures) – is an important facilitator of forgiveness (e.g., Pronk et al. Reference Pronk, Karremans, Overbeek, Vermuist and Wigboldus2010; Wilkowski et al. Reference Wilkowski, Robinson and Troop-Gordon2010; cf. Finkel & Campbell Reference Finkel and Campbell2001). Whereas the initial impulsive response to a transgression often is to retaliate, individual differences in executive control are positively associated with the ability to inhibit such retaliatory and negative affective responses, and instead to respond in a forgiving manner (Pronk et al. Reference Pronk, Karremans, Overbeek, Vermuist and Wigboldus2010).

Importantly, it appears that individuals low in executive control have difficulty forgiving an offending relationship partner even when the partner is someone they feel close to – or, to use McCullough et al.'s terminology, even when relationship value is high. In a recent series of studies in primary schools, we have found initial evidence that 11- and 12-year old children are more likely to forgive their friends than non-friends, but crucially, that this “relationship value” effect was more strongly pronounced among children high in executive control (Van der Wal et al. Reference Van der Wal, Karremans and Cillessen2012). In fact, although based on the relationship value hypothesis we should have expected stronger forgiveness when the transgressor is a friend rather than non-friend across the range, children low in executive control basically did not show this effect. Similar effects were found in a study with late adolescents, revealing that closeness only predicted forgiveness among participants high in executive control. These findings suggest that high relationship value generally leads to the recruitment of executive control in order to down-regulate negative emotions toward the offender, unless the individual lacks such executive control resources. Thus, relationship value is not always the best possible predictor of forgiveness – at least not for everyone, or under all circumstances (e.g., when executive control resources are temporarily depleted).

Admittedly, this literature has so far not looked at how executive control may be related to perceived exploitation risk. A possible prediction based on the authors' proposed model is that low executive control individuals may have more difficulty in estimating exploitation risk, which may prevent them from forgiving valuable relationship partners. Or, alternatively, low executive control might disrupt the entire computational process, failing to successfully integrate relationship value and exploitation risk information in order to decide whether or not to forgive.

The fact that low executive control hinders forgiveness, even in the face of high relationship value (and possibly, low exploitation risk), raises intriguing and complex questions. For example, and following the authors' adaptationist logic, do individuals low in executive control – which is strongly genetically determined (see Friedman et al. Reference Friedman, Miyake, Young, DeFries, Corley and Hewitt2008) – adopt alternative strategies to minimize the fitness costs of their relative inability to forgive valuable relationship partners? Or, as with the example of sex differences provided in the target article, does executive control modify the costs and benefits of forgiving valuable others, such that the lack of forgiveness may be less detrimental to the valuable relationships of individuals low versus high in executive control? Do low executive control individuals in some way compensate for the loss of fitness benefits from their relative struggle to forgive? Although very speculative, perhaps low executive control individuals – or more broadly, any individual with a lower forgiveness propensity for whatever reason – may seek out relationship partners who possess particularly well developed conflict-resolution skills.

To conclude, whereas in the target article the authors have built an evolutionary theoretical account of forgiveness and revenge by integrating largely dispersed research findings, in turn this account inspires many novel and specific questions. However, although relationship value and exploitation risk are the factors that help to explain when and why revenge or forgiveness are the most adaptive and thus most likely responses, they may be limited in addressing how revenge and forgiveness actually occurs. Yet, knowledge about the basic processes that describe how forgiveness occurs also informs us about when forgiveness or revenge is the most likely response. Hence, we believe that recent findings on executive control and forgiveness – and more generally, any past and future findings on the role of the proximate cognitive and neural mechanism involved in forgiveness and revenge – can help to further inform and refine the authors' theoretical approach.