In the target article, McCullough et al. do an admirable job of arguing that revenge is not a disease, and instead may be an adaption to prevent exploitation. This approach is long overdue in many social sciences, as it moves away from pejorative preconceptions about behaviours we don't like. As with any adaptive explanation for behaviour, there is a high risk of the authors' argument being misunderstood to claim that all instances of revenge should be adaptive. Such misunderstandings regularly occur with other evolutionary explanations of human social behaviour. As such, the authors' argument requires extension to examine when revenge and forgiveness will overstretch (or understretch) the bounds of utility, and why.
A complete explanation of revenge and forgiveness will include errors of absence as well as errors of excess. While there are popularized cases of ridiculous revenge, we often overlook the excessive “lack of revenge” or excessive forgiveness. These are the things that fill books like Chicken Soup for the Soul. They are potentially equally maladaptive, but we don't see them as “errors” because we “like” this behaviour (see also Wakefield Reference Wakefield1992). There is likely an optimal level of revenge and forgiveness for any situation. Too little revenge is an insufficient deterrent, but too much revenge invites further retaliation (Barclay Reference Barclay2008). Too little forgiveness prevents reparation of a relationship, but too much forgiveness invites future exploitation (Axelrod Reference Axelrod1984). Finding the optimal level of revenge involves “brinksmanship” (Daly & Wilson Reference Daly and Wilson1988), a difficult game when people have imperfect information about the world or about others' past and future intentions (e.g., Todd Reference Todd, Gigerenzer and Selten2001). Because of such constraints, no evolved psychological mechanism is expected to produce optimal results in every single instance, but is expected to be adaptive on average (Haselton & Buss Reference Haselton and Buss2000; Nesse Reference Nesse2005; Barclay Reference Barclay and Roberts2011). The following are some causes of excessive or insufficient revenge or forgiveness, and when combined in one individual, they could result in markedly abnormal behaviour indeed.
Misperceptions of costs and benefits
To produce adaptive levels of revenge and forgiveness, our brains must use environmental cues of the costs and benefits. Assessing these costs and benefits is no small task, as one must assess one's strength relative to a transgressor (and possibly an audience), audience presence and characteristics, one's need, the risk, the relationship value, and a host of other factors. Naturally, there is error associated with assessing any of these (Barclay Reference Barclay2008). Sometimes these errors will balance out, but when they don't, they will compound to produce more revenge or forgiveness than is optimal. Statistically speaking, these misperceptions alone will cause deviations from optimality that are normally distributed about the optimum, with most individuals near optimality but with some individuals displaying excessively high or low revenge or forgiveness.
Misperceptions of others' actions
The optimal level of revenge and forgiveness likely depends on others' intentions, both in terms of the reasons for their past actions and their intended future behaviour. Assessing such intentions is a mind-reading game, and is also prone to error. “Rules of thumb” based on past experience will only sometimes be right, and will sometimes overestimate hostility. Based on this, we might predict that people who are better at reading others' intentions will produce more optimal levels of revenge and forgiveness.
Minimizing costly errors
Different errors have different costs, and natural selection has presumably designed our emotions so that we avoid committing more costly or more frequent errors (Haselton & Buss Reference Haselton and Buss2000; Nesse Reference Nesse2005). If being too vengeful is more costly than being insufficiently vengeful, then our revenge systems should be biased towards producing less revenge than is “needed,” and vice versa. A similar argument holds for forgiveness. Which is more costly, excessive or insufficient revenge (or forgiveness)? This is probably an empirical question. In fact, the answer may vary in different social environments depending on the frequency and importance of exploitation (bias towards excessive retaliation) and long-term cooperation (bias towards excessive forgiveness). By focusing on the costs and frequencies of these different errors, we can predict when we will observe excessive vengeance or excessive forgiveness.
Genetic or development noise variation
The target article outlines a number of tasks performed by our revenge and forgiveness systems, each of which involves many steps. As with any complex trait, each of these sub-tasks will be affected by multiple genes and environmental influences. Because these influences can combine in different combinations, it will cause natural variation about an optimum for each sub-task, resulting in some individuals in the tail ends of excessive revenge.
Pathologies within the revenge systems
Although the target article suggests that revenge is not a “disease,” it does leave open the possibility of genuine diseases within our evolved revenge systems. Some individuals might indeed have something “broken” in the brain areas responsible for assessing costs, benefits, and intentions, or for producing an appropriate level of revenge. For example, if a (subconscious) assessment of costs tends to inhibit revenge, then anything that damages the brain's inhibitory systems will prevent this inhibition and will result in excessive revenge. Also, if someone is insensitive to costs or punishment in general (e.g., psychopaths), then there will be nothing to lower their vengefulness down to optimal levels. In other words, the capacity for revenge is not pathological, nor is the acting on that capacity, but there can be pathologies associated with expressing that capacity. It is these pathologies that probably produce the types of revenge and forgiveness that make newspaper headlines (Barclay Reference Barclay2008).
These are but some of the potential causes of abnormal levels of revenge and forgiveness. Some will result in normally distributed variation in revenge and forgiveness, whereas others will cause systematic biases towards excess (e.g., pathologies, error management). Altogether, they show how not every instance of revenge will be beneficial, nor will every instance of forgiveness. Thus, we can extend the framework that McCullough et al. provide to make predictions about “abnormal” levels of revenge and forgiveness.
In the target article, McCullough et al. do an admirable job of arguing that revenge is not a disease, and instead may be an adaption to prevent exploitation. This approach is long overdue in many social sciences, as it moves away from pejorative preconceptions about behaviours we don't like. As with any adaptive explanation for behaviour, there is a high risk of the authors' argument being misunderstood to claim that all instances of revenge should be adaptive. Such misunderstandings regularly occur with other evolutionary explanations of human social behaviour. As such, the authors' argument requires extension to examine when revenge and forgiveness will overstretch (or understretch) the bounds of utility, and why.
A complete explanation of revenge and forgiveness will include errors of absence as well as errors of excess. While there are popularized cases of ridiculous revenge, we often overlook the excessive “lack of revenge” or excessive forgiveness. These are the things that fill books like Chicken Soup for the Soul. They are potentially equally maladaptive, but we don't see them as “errors” because we “like” this behaviour (see also Wakefield Reference Wakefield1992). There is likely an optimal level of revenge and forgiveness for any situation. Too little revenge is an insufficient deterrent, but too much revenge invites further retaliation (Barclay Reference Barclay2008). Too little forgiveness prevents reparation of a relationship, but too much forgiveness invites future exploitation (Axelrod Reference Axelrod1984). Finding the optimal level of revenge involves “brinksmanship” (Daly & Wilson Reference Daly and Wilson1988), a difficult game when people have imperfect information about the world or about others' past and future intentions (e.g., Todd Reference Todd, Gigerenzer and Selten2001). Because of such constraints, no evolved psychological mechanism is expected to produce optimal results in every single instance, but is expected to be adaptive on average (Haselton & Buss Reference Haselton and Buss2000; Nesse Reference Nesse2005; Barclay Reference Barclay and Roberts2011). The following are some causes of excessive or insufficient revenge or forgiveness, and when combined in one individual, they could result in markedly abnormal behaviour indeed.
Misperceptions of costs and benefits
To produce adaptive levels of revenge and forgiveness, our brains must use environmental cues of the costs and benefits. Assessing these costs and benefits is no small task, as one must assess one's strength relative to a transgressor (and possibly an audience), audience presence and characteristics, one's need, the risk, the relationship value, and a host of other factors. Naturally, there is error associated with assessing any of these (Barclay Reference Barclay2008). Sometimes these errors will balance out, but when they don't, they will compound to produce more revenge or forgiveness than is optimal. Statistically speaking, these misperceptions alone will cause deviations from optimality that are normally distributed about the optimum, with most individuals near optimality but with some individuals displaying excessively high or low revenge or forgiveness.
Misperceptions of others' actions
The optimal level of revenge and forgiveness likely depends on others' intentions, both in terms of the reasons for their past actions and their intended future behaviour. Assessing such intentions is a mind-reading game, and is also prone to error. “Rules of thumb” based on past experience will only sometimes be right, and will sometimes overestimate hostility. Based on this, we might predict that people who are better at reading others' intentions will produce more optimal levels of revenge and forgiveness.
Minimizing costly errors
Different errors have different costs, and natural selection has presumably designed our emotions so that we avoid committing more costly or more frequent errors (Haselton & Buss Reference Haselton and Buss2000; Nesse Reference Nesse2005). If being too vengeful is more costly than being insufficiently vengeful, then our revenge systems should be biased towards producing less revenge than is “needed,” and vice versa. A similar argument holds for forgiveness. Which is more costly, excessive or insufficient revenge (or forgiveness)? This is probably an empirical question. In fact, the answer may vary in different social environments depending on the frequency and importance of exploitation (bias towards excessive retaliation) and long-term cooperation (bias towards excessive forgiveness). By focusing on the costs and frequencies of these different errors, we can predict when we will observe excessive vengeance or excessive forgiveness.
Genetic or development noise variation
The target article outlines a number of tasks performed by our revenge and forgiveness systems, each of which involves many steps. As with any complex trait, each of these sub-tasks will be affected by multiple genes and environmental influences. Because these influences can combine in different combinations, it will cause natural variation about an optimum for each sub-task, resulting in some individuals in the tail ends of excessive revenge.
Pathologies within the revenge systems
Although the target article suggests that revenge is not a “disease,” it does leave open the possibility of genuine diseases within our evolved revenge systems. Some individuals might indeed have something “broken” in the brain areas responsible for assessing costs, benefits, and intentions, or for producing an appropriate level of revenge. For example, if a (subconscious) assessment of costs tends to inhibit revenge, then anything that damages the brain's inhibitory systems will prevent this inhibition and will result in excessive revenge. Also, if someone is insensitive to costs or punishment in general (e.g., psychopaths), then there will be nothing to lower their vengefulness down to optimal levels. In other words, the capacity for revenge is not pathological, nor is the acting on that capacity, but there can be pathologies associated with expressing that capacity. It is these pathologies that probably produce the types of revenge and forgiveness that make newspaper headlines (Barclay Reference Barclay2008).
These are but some of the potential causes of abnormal levels of revenge and forgiveness. Some will result in normally distributed variation in revenge and forgiveness, whereas others will cause systematic biases towards excess (e.g., pathologies, error management). Altogether, they show how not every instance of revenge will be beneficial, nor will every instance of forgiveness. Thus, we can extend the framework that McCullough et al. provide to make predictions about “abnormal” levels of revenge and forgiveness.