The hypotheses generated in McCullough et al.'s article can be tested in multiple ways and contexts. Having worked in the interesting context of the new South Africa (Allan et al. Reference Allan, Allan M, Kaminer and Stein2006; Kaminer et al. Reference Kaminer, Stein, Mbanga and Zungu-Dirwayi2001; Stein et al. Reference Stein, Seedat, Kaminer, Moomal, Herman, Sonnega and Williams2008), we read the target article from the vantage of whether observations and data about revenge and forgiveness in this setting might be relevant. Insofar as South Africa underwent a rapid transformation from an apartheid system characterized by racial discrimination to a democratic dispensation characterized by universal human rights, it may provide a unique laboratory for investigating certain aspects of the psychology of revenge and forgiveness. We note McCullough et al.'s warning that their model is not intended to apply to groups, but groups are of course comprised of individuals.
A first point to note is that in South Africa, with the advent of the new democratic dispensation in 1994, a decision was made that instead of retributive justice (i.e., punishment), there would be reparative justice (as embodied by a “Truth and Reconciliation Commission”) (Asmal et al. Reference Asmal, Asmal and Roberts S1996; Stein Reference Stein1998). One key rationale was that it simply wasn't possible to ensure retributive justice, given that the institutions responsible for implementing justice after 1994 had not yet democratized. It is notable that McCullough et al. posit from basic evolutionary principles that punishment is less effective when the costs of punishment are high. This was certainly the case in post-democratic South Africa.
Nevertheless, there were many who objected to the Truth and Reconciliation Commission (TRC) over precisely this issue. In particular, families of those who had been victims of gross human rights violations were vehemently opposed to the idea that perpetrators would not be formally prosecuted and punished. Notably, McCullough et al. provide a model which predicts that revenge is more likely when a kin group member is involved. Thus, revenge may be well be sweeter for some than for others (de Quervain et al. Reference de Quervain, Fischbacher, Treyer, Schellhammer, Schnyder, Buck and Fehr2004; Stein & Kaminer Reference Stein and Kaminer2006). Although the courts are not always seen as active in the South African setting, in the particular instance of the TRC the judicial system was clearly active (albeit administering reparative rather than retributive justice), predicting, per McCullough et al., less need for retaliatory feelings in the majority of observers (Stein et al. Reference Stein, Seedat, Kaminer, Moomal, Herman, Sonnega and Williams2008).
Indeed, a second point is that forgiveness levels appear to be moderately high in all sectors of the South African population in the aftermath of the TRC (Stein et al. Reference Stein, Seedat, Kaminer, Moomal, Herman, Sonnega and Williams2008). This may well be consistent with McCullough et al.'s model, which predicts that shared interests, similar values, and many opportunities for mutually beneficial interactions are good candidates for forgiveness. Certainly, despite clear racial heterogeneity in South Africa, there are many values that are held in common by the population, including religious values, and despite considerable geographical separation of races, there are many opportunities for interaction (e.g. in domestic and commercial settings). Relationships are in many ways “valuable,” and therefore considerable conciliatory behaviour could perhaps have been predicted.
Nevertheless, there have been many in South Africa, including participants in the TRC, with low levels of forgiveness for past violations (Stein et al. Reference Stein, Seedat, Kaminer, Moomal, Herman, Sonnega and Williams2008). Again, there are a number of potential explanatory variables. McCullought et al. note that apology is important, and the South African data confirm this (Allan et al. Reference Allan, Allan M, Kaminer and Stein2006). McCullough et al. cite meta-analytic findings that women score higher on self-reports of tendencies to forgive, but the South African data suggest that men are more likely to forgive in the South African context (Kaminer et al. Reference Kaminer, Stein, Mbanga and Zungu-Dirwayi2001). It may be that sex differences in forgiveness pertain to forgiveness of different kinds of violations; indeed, in the setting of the TRC women were more likely than men to report violations to family members (Allan et al. Reference Allan, Allan M, Kaminer and Stein2006). Speculatively, in the South African setting, forgiving may have higher benefits for males, who are perhaps more involved in commercial interactions with one another than are females. Similarly, in nonhuman primates, it may be hypothesized that reconciliation after contests is particularly adaptive for males who subsequently need to cooperate in key ways (de Waal Reference de Waal2000).
A third point we want to make is that even within groups of individuals with apparently similar interests (from an adaptive perspective), there is considerable variation (consider, for example, Nelson Mandela and Desmond Tutu's emphasis on reconciliation at a time when many of their comrades and colleagues were less concerned with this issue). Indeed, a comprehensive account of the full range of revenge and forgiveness in settings such as the new South Africa would seem to require a model that goes beyond a purely computational approach that weights costs and benefits, to also include an account of how values are embodied in cognitive-affective systems. Thus, for example, a detailed understanding of the moral metaphors that individuals live by, and of underlying motivational structures, appears relevant to a full understanding of the complex range of revenge and forgiveness behaviours seen in real-life settings (Boulding Reference Boulding1969; Johnson Reference Johnson1993).
Along these lines, we would emphasize that our observations and data from South Africa are tangential to the question of whether revenge and forgiveness behaviour is mediated by specially evolved brain “modules.” Although tackling the question of whether any specially evolved modules exist is not the main focus of our commentary, we wish to note that, in our view, the brain-mind has evolved in a considerably “messier” way than might be suggested by some views of neat modules proposed by evolutionary psychology, that evolved cognitive-affective systems instead have components with indistinct boundaries and distributed functions, and that behaviour is best understood to emerge from an interaction between relatively few ancient special-purpose circuits and more recent general purpose mechanisms (Nesse & Stein 2012; Panksepp & Panksepp 2000).
In summary, then, based on observations from one country that arguably conducted a nation-wide social experiment on revenge and forgiveness, we would conclude that these are partially consistent with the hypotheses generated by McCullough and colleagues. At the same time, we note that data were not specifically collected with this thesis in mind, and therefore do not provide a very direct test of the hypotheses. The South African findings on gender are of particular interest insofar as they apparently conflict with other published findings, and yet might be explicable on the basis of particular circumstances in South Africa, and with the underlying explanation that revenge and forgiveness involve adaptive mechanisms, and therefore will be triggered in different ways in different environments. Finally, we suggest that the rich range of revenge and forgiveness phenomena in real-life settings is likely to require explanatory concepts other than specialized “modules” and their computational outputs.
The hypotheses generated in McCullough et al.'s article can be tested in multiple ways and contexts. Having worked in the interesting context of the new South Africa (Allan et al. Reference Allan, Allan M, Kaminer and Stein2006; Kaminer et al. Reference Kaminer, Stein, Mbanga and Zungu-Dirwayi2001; Stein et al. Reference Stein, Seedat, Kaminer, Moomal, Herman, Sonnega and Williams2008), we read the target article from the vantage of whether observations and data about revenge and forgiveness in this setting might be relevant. Insofar as South Africa underwent a rapid transformation from an apartheid system characterized by racial discrimination to a democratic dispensation characterized by universal human rights, it may provide a unique laboratory for investigating certain aspects of the psychology of revenge and forgiveness. We note McCullough et al.'s warning that their model is not intended to apply to groups, but groups are of course comprised of individuals.
A first point to note is that in South Africa, with the advent of the new democratic dispensation in 1994, a decision was made that instead of retributive justice (i.e., punishment), there would be reparative justice (as embodied by a “Truth and Reconciliation Commission”) (Asmal et al. Reference Asmal, Asmal and Roberts S1996; Stein Reference Stein1998). One key rationale was that it simply wasn't possible to ensure retributive justice, given that the institutions responsible for implementing justice after 1994 had not yet democratized. It is notable that McCullough et al. posit from basic evolutionary principles that punishment is less effective when the costs of punishment are high. This was certainly the case in post-democratic South Africa.
Nevertheless, there were many who objected to the Truth and Reconciliation Commission (TRC) over precisely this issue. In particular, families of those who had been victims of gross human rights violations were vehemently opposed to the idea that perpetrators would not be formally prosecuted and punished. Notably, McCullough et al. provide a model which predicts that revenge is more likely when a kin group member is involved. Thus, revenge may be well be sweeter for some than for others (de Quervain et al. Reference de Quervain, Fischbacher, Treyer, Schellhammer, Schnyder, Buck and Fehr2004; Stein & Kaminer Reference Stein and Kaminer2006). Although the courts are not always seen as active in the South African setting, in the particular instance of the TRC the judicial system was clearly active (albeit administering reparative rather than retributive justice), predicting, per McCullough et al., less need for retaliatory feelings in the majority of observers (Stein et al. Reference Stein, Seedat, Kaminer, Moomal, Herman, Sonnega and Williams2008).
Indeed, a second point is that forgiveness levels appear to be moderately high in all sectors of the South African population in the aftermath of the TRC (Stein et al. Reference Stein, Seedat, Kaminer, Moomal, Herman, Sonnega and Williams2008). This may well be consistent with McCullough et al.'s model, which predicts that shared interests, similar values, and many opportunities for mutually beneficial interactions are good candidates for forgiveness. Certainly, despite clear racial heterogeneity in South Africa, there are many values that are held in common by the population, including religious values, and despite considerable geographical separation of races, there are many opportunities for interaction (e.g. in domestic and commercial settings). Relationships are in many ways “valuable,” and therefore considerable conciliatory behaviour could perhaps have been predicted.
Nevertheless, there have been many in South Africa, including participants in the TRC, with low levels of forgiveness for past violations (Stein et al. Reference Stein, Seedat, Kaminer, Moomal, Herman, Sonnega and Williams2008). Again, there are a number of potential explanatory variables. McCullought et al. note that apology is important, and the South African data confirm this (Allan et al. Reference Allan, Allan M, Kaminer and Stein2006). McCullough et al. cite meta-analytic findings that women score higher on self-reports of tendencies to forgive, but the South African data suggest that men are more likely to forgive in the South African context (Kaminer et al. Reference Kaminer, Stein, Mbanga and Zungu-Dirwayi2001). It may be that sex differences in forgiveness pertain to forgiveness of different kinds of violations; indeed, in the setting of the TRC women were more likely than men to report violations to family members (Allan et al. Reference Allan, Allan M, Kaminer and Stein2006). Speculatively, in the South African setting, forgiving may have higher benefits for males, who are perhaps more involved in commercial interactions with one another than are females. Similarly, in nonhuman primates, it may be hypothesized that reconciliation after contests is particularly adaptive for males who subsequently need to cooperate in key ways (de Waal Reference de Waal2000).
A third point we want to make is that even within groups of individuals with apparently similar interests (from an adaptive perspective), there is considerable variation (consider, for example, Nelson Mandela and Desmond Tutu's emphasis on reconciliation at a time when many of their comrades and colleagues were less concerned with this issue). Indeed, a comprehensive account of the full range of revenge and forgiveness in settings such as the new South Africa would seem to require a model that goes beyond a purely computational approach that weights costs and benefits, to also include an account of how values are embodied in cognitive-affective systems. Thus, for example, a detailed understanding of the moral metaphors that individuals live by, and of underlying motivational structures, appears relevant to a full understanding of the complex range of revenge and forgiveness behaviours seen in real-life settings (Boulding Reference Boulding1969; Johnson Reference Johnson1993).
Along these lines, we would emphasize that our observations and data from South Africa are tangential to the question of whether revenge and forgiveness behaviour is mediated by specially evolved brain “modules.” Although tackling the question of whether any specially evolved modules exist is not the main focus of our commentary, we wish to note that, in our view, the brain-mind has evolved in a considerably “messier” way than might be suggested by some views of neat modules proposed by evolutionary psychology, that evolved cognitive-affective systems instead have components with indistinct boundaries and distributed functions, and that behaviour is best understood to emerge from an interaction between relatively few ancient special-purpose circuits and more recent general purpose mechanisms (Nesse & Stein 2012; Panksepp & Panksepp 2000).
In summary, then, based on observations from one country that arguably conducted a nation-wide social experiment on revenge and forgiveness, we would conclude that these are partially consistent with the hypotheses generated by McCullough and colleagues. At the same time, we note that data were not specifically collected with this thesis in mind, and therefore do not provide a very direct test of the hypotheses. The South African findings on gender are of particular interest insofar as they apparently conflict with other published findings, and yet might be explicable on the basis of particular circumstances in South Africa, and with the underlying explanation that revenge and forgiveness involve adaptive mechanisms, and therefore will be triggered in different ways in different environments. Finally, we suggest that the rich range of revenge and forgiveness phenomena in real-life settings is likely to require explanatory concepts other than specialized “modules” and their computational outputs.