Moral Origins

by Christopher Boehm

  To those models we must add kin selection, for nepotism accounts for the strong generosity that obtains within families. We must also keep in mind George Williams’s thoughts about extrafamilial generosity’s being similar in important ways to generosity that takes place within families. Indeed, in evolutionary terms nepotism may have served as a preadaptation for the evolutionary development of altruism.

  The Hadza study hints very strongly at altruism’s being a desirable feature in marital choice, but in any event people in these egalitarian bands pay close attention to whether others are sympathetic, generous with food, hardworking, or trustworthy—versus mean-spirited, stingy, lazy, or overly cunning.16 They often are choosing their associates accordingly,17 and this would apply to band membership, to safety nets within bands, to marriage arrangements, to routine nonmarital subsistence cooperation between families within the band, and to cooperation with people from other bands as a means of establishing long-distance safety nets or trading opportunities.18 Unusual generosity—or a noteworthy lack thereof—can be an important factor in all of these relations.

  Having a good, generous character is important, but it’s far from being all important. There are a variety of purely structural constraints in choosing these same types of partners, which will include ongoing kin ties and present or past in-law relationships. But even if an individual is most likely to join a band where close kin or in-laws are living, he or she also is more likely to try to choose the relatives who are more generous or more productive, in line with character traits the Aché and !Kung were favoring, rather than relatives who are unproductive, very stingy, or both. From the standpoint of gene selection, an altruist is more likely to be more successful in cooperation with kin and nonkin alike. And when two altruists choose each other, both will be profiting in fitness terms.

  From a theoretical perspective, it has been generosity based in feelings of sympathy for others and their needs that has concerned us here as a rather hard-to-study aspect of hunter-gatherer cooperation. The basis for such generosity can be sustained at the genome level if certain criteria are fulfilled, even though by definition there will be losses of fitness to be accounted for whenever generous behavior fulfills our definition of being altruistic. One criterion is that somehow the altruism must be compensated by the action of at least one of the five mechanisms discussed in Chapter 3. These include two possible piggybacking models, one-shot mutualism and long-term reciprocal altruism models, and the selection-by-reputation model. For several of these models, and also for the group selection model, free-rider suppression also has to be effective.

  For LPA foragers, the compensation criterion is being fulfilled, largely, I believe, in the form of reputational benefits. Thus, in combination with free-rider suppression my candidate as a main contributing mechanism to explain the limited but well-exemplified and socially significant altruism of hunter-gatherers has been selection by reputation. There’s a reason for giving these two social selection mechanisms such priority. This lies in the fact that reputation-based social selection is likely to gain special power precisely because the selection process is based on choice. This makes it similar to Darwinian sexual selection, but with respect to its efficacy among hunter-gatherers, selection by reputation might actually be going sexual selection one better.

  INTERACTIVE EFFECTS

  It would be useful if someone adept at evolutionary mathematics could model all the mechanisms we have discussed in combination, weighting each model for its probable contribution to the maintenance of altruistic genes in human gene pools over the past 45,000 years and more. However, the working hypothesis I’m offering here is that the selection-by-reputation model is likely to be especially powerful because not only is this type of social selection driven by well-focused, consistent preferences, but it also involves a two-way choice process. When the Hadza or the !Kung follow their preferences in choosing partners to mate with, or for that matter just partners to work with, one major effect of their choice behavior will be that generous and fair co-operators will be in a position to successfully choose others who are similarly altruistic and that on both sides fitness will profit.19

  If the sympathetic altruists (along with hard workers and trustworthy people) are tending to pair up with others who are similarly socially desirable like themselves,20 this leaves their less desirable counterparts tending to pair up with one another—and suffering the disadvantages of less effective cooperation. Obviously, the advantage goes to the pairs of worthier individuals who are reaping the fruit of superior collaboration—and at the same time are unlikely to suffer from free riding.

  This is significantly different from the pattern of having drab peahens who on a one-way basis choose resplendent males whose superior genes bring females more reproductive success. It’s as though both male and female fowl were developing attractive tail feathers to advertise their fitness—and the males were choosing the females with the best tail feathers at the same time that they themselves were being chosen on the same criterion. Were this the case, the power of sexual selection, which is widely acknowledged to be remarkably strong, would be stronger still because both parties would be simultaneously displaying superior fitness—and choosing it.21

  When geneticist Ronald Fisher became fascinated with Darwin’s treatment of sexual selection, and with the fact that such choice-driven selection can result in “exaggerated” traits, he thought that costly peacock tail feathers might be explained in terms of “runaway selection,” which results from the escalating interaction of astute female preferences with male signs of genetic superiority.22 When both the traits being chosen and the choosers are involved on both sides of the equation, I believe this could further intensify such effects. Mathematical modeling would be useful in testing this hypothesis, but as a cultural anthropologist obviously that is not my forté.

  Richard D. Alexander differentiates between “sexual selection” and “reciprocity selection” (I’ve been using the more descriptive term “selection by reputation” for the latter, for the sake of clarity for a general audience) as he considers this two-way element in the choice behavior that is inherent in selection by reputation:

  Sexual selection is a distinctive kind of runaway selection because joint production of offspring by the interacting pair causes the process to accelerate. . . . The defining feature of runaway selection is not acceleration, however, but the tendency of the process to go . . . much further beyond adaptiveness . . . than is ordinarily the case in the myriad compromises among the conflicting adaptive traits that create and maintain the unified organism.

  This aspect of runaway selection may hold for reciprocity selection, in which, unlike in sexual selection, both parties can carry tendencies not only to choose extremes but also to display extremes. In social selection . . . an individual can play both roles, of chooser and chosen.23

  Of course, one thing that Alexander means when he says that traits selected by reciprocity selection can go “significantly beyond adaptiveness” is that costly altruistic traits might be supported by this type of selection, as well as any other traits that appear otherwise to be maladaptive for individuals but are positively involved with reputational choices that are mutual. This may be the best one hypothesis so far to help explain human altruism—but only, I believe, if it is combined with the very effective kind of free-rider suppression I’ve been emphasizing throughout this book.

  WHY FREE RIDERS HAVEN’T SIMPLY VANISHED

  Trivers was correct when he suggested that moralistically aggressive groups could come down hard on detectable cheaters. However, I’ve suggested that with earlier humans, and still earlier with Ancestral Pan, it was selfish bullies who were taking the main free rides by competitive actions that favored their own genes at the same time that the genes of their less selfish and less powerful victims were being disadvantaged.

  In Chapter 4 we saw that in fifty LPA societies these bullies appear to have been executed far more often than thieves or cheater
s, or, most likely, than sexual offenders, which means that their selfish aggressions were seriously problematic for their groups—and also that these powerful free riders were paying a steep price in terms of loss of fitness. We also saw that when earlier humans were intent on being egalitarian, physically punitive types of social selection probably would have been acting far more powerfully than today because initially there was no conscience to help restrain the bullies.

  Thus, the earlier stages of free-rider suppression were likely to have been harsh indeed, and we may reasonably assume that the more selfishly driven alpha types, those prone to take risks, were bringing severe and quite frequent punishment on themselves until a conscience evolved to aid them in controlling themselves and thereby in avoiding such dire consequences. As bullying (or cheating) free-rider types increasingly became able to restrain themselves from actions that would bring on punishment, they would have suffered fewer and fewer reproductive disadvantages, so with the existence of more effective self-control there’s no reason to believe that their genes should have gone out of business entirely or have even come close to doing so.

  Keep in mind that even though LPA bands are so highly egalitarian, this does not mean there’s an absence of competition. Males compete for hunting prowess, and males (and females) compete for mates. Indeed, egalitarianism itself is based on competition between a few stronger individuals and the subordinates who unite to oppose them. Thus, if a person can channel his or her competitive tendencies in directions that are socially acceptable, and at the same time curb their expression when they will make for fitness-reducing punishment, selfishly competitive tendencies can be quite useful to fitness.

  Refraining strategically from aggressive behavior is well exemplified by Inuttiaq, for his efficient, self-conscious evolutionary conscience put him intuitively in touch with his personal social dilemma, and it enabled him to strategically inhibit many of his aggressive responses rather than expressing them in ways that would make his fellow Utku seriously apprehensive—and perhaps prone to dire action.

  Not all men prone to despotism manage to curb their despotic tendencies so efficiently. In fact, when Scandinavian explorers first contacted one Eskimo group in Greenland, they noted that a very dominant shaman had killed serially and was being treated with great respect by his group.24 They left before the dominator was killed, so that was not witnessed, but we may be quite confident that somehow he was disposed of. We’ve vicariously witnessed a similar !Kung execution in the Kalahari, and other Inuit groups do away with such men.

  In Chapter 7 Table IV showed us that such executions are quite widespread. But whether would-be despots successfully inhibit themselves or not, their actual potency as free riders who can take serious long-term advantage of altruists and others is very limited. Either like Inuttiaq they’re afraid to take the free ride, or like /Twi the aggressive !Kung Bushman they’ll be killed for taking their aggressive free ride too actively.

  This leads to an important theoretical point that requires some further emphasis. George Williams’s mathematical portrayal of free riders and altruists assumed that free riders were designed (by evolutionary process) to exploit altruists and thereby disadvantage their genes.25 As a result, altruistic genes could never reach fixation in the gene pool concerned. And if new altruistic genes were to appear as mutants, free-rider mutants would soon appear to drive them out of business.

  When we bring in the conscience as a highly sophisticated means of channeling behavioral tendencies so that they are expressed efficiently in terms of fitness, this scenario changes radically. Over time, human individuals with strong free-riding tendencies—but who exercised really efficient self-control—would not have lost fitness because these predatory tendencies were so well inhibited. And if they expressed their aggression in socially acceptable ways, this in fact would have aided their fitness.

  That’s why I believe that both free-riding genes and altruistic genes could have remained well represented and coexisting in the same gene pool. Genes that made for bullying free riding could have been useful because they were providing a useful competitive drive, whereas genes that made for altruism could have been useful because altruism was being compensated by reputational benefits and by other compensatory mechanisms we have discussed.

  This useful self-inhibition wasn’t perfect. In today’s bands, we still have the occasional active bullies and cheats, who are prone to take major free rides because their own reckless optimism about what they can get away with leads them astray,26 because their dominance or deception is compulsive, or because the feedback their consciences provide them with is faulty. We’ve also seen that it was bullying that caused more of the social problems, and that it did so more frequently. Let’s reconsider the numbers. In Chapter 4, Table I, bullying (“Intimidation of group”) dominated the reported executions, and in Chapter 7, Table III, the most frequently reported acts of deviant social predation were acts of domination by intimidators (with 461 cites for the ten societies), whereas acts of cheating were mentioned only 42 times and for only half of the ten societies sampled even though the central tendency was quite noteworthy. Thus, even though cheating free riders have been academically center stage ever since Trivers made them so famous in 1971, for humans it appears that the major free rides have been taken by politically forceful selfish dominators, whose victims include not only generous altruists who equal them in power but are much less selfish, but also anyone else who is less inclined or less able to dominate others.

  Aggressively selfish behavior continues to be a widespread problem today among LPA foragers, and it’s obvious that individually variable, selfishly aggressive tendencies are still with us. If the people so endowed were free to express these propensities without social inhibition, then fair-minded, generosity-driven systems of indirect reciprocity simply would not work. With respect to human cooperation, it’s fortunate that these selfishly aggressive propensities have been susceptible to a remarkable degree of control from within and also from without. From within the human psyche an evolutionary conscience provided the needed self-restraint, while externally it was group sanctioning that largely took care of the dominators and cheaters who couldn’t or would’t control themselves.

  AN EVOLUTIONARY SEQUENCE

  At this point I’d like to set forth a historical sequence, starting at the ancestral beginning, as a way of summarizing much of what has been said in the preceding pages—and as a way of keeping my promise to make the natural history of moral origins more historical. To start with, most likely primitive “altruism quotients” in Ancestral Pan were as modest as they seem to be in today’s bonobos and chimpanzees. However, in comparing their degree of innate sympathetic altruism with ours, we must keep in mind that in groups these apes have no way to amplify their cooperation through golden rules.

  Our behavioral reconstruction also tells us that a noteworthy if rudimentary potential for (nonmoralistic) group social control existed in this ancestor, a potential that was being directed exclusively at bullies as a resented and readily identifiable type of selfish, competitive, exploitative free rider. This means that the selfish behaviors of certain aggressive individuals could be curbed, even though basically ancestral social orders remained quite hierarchical. This ancestor did have a significant capacity for self-recognition, but in the absence of an evolutionary conscience, its self-control was based just in fear of retaliation, and a capacity for submission.

  We cannot be sure how abruptly the next phase of moral evolution began, but it could have involved the escalation of similarly anti-hierarchical social control to a point that rather than profiting, stronger individuals more often were paying a significant price for their attempts at opportunistic, free-riding domination. This could have led to genetic selection in favor of an enhanced capacity for self-control that involved something new: although fear of retaliation by subordinate coalitions continued to slow these bullies down, as a protoconscience developed my hypothesis is that rules could no
w be internalized as well and that this led to a more sensitive adjustment of individual behavior to group preferences.

  It’s impossible to estimate when the capacity to identify emotionally with rules began to seriously affect overall patterns of social behavior and, hence, selection outcomes. Homo erectus with its relatively large brain is at least conceivable as a possibility. However, the position I’ve taken is that such social selection had to become quite decisive starting a quarter of a million years ago, when a still larger-brained archaic Homo sapiens—toward the end of its career—began to hunt large, hooved mammals and depend on their meat. For groups of people in bands to have been really efficient at this, it’s very likely that meat-sharing had to have been well equalized, and this efficiency could have been crucial for group or regional survival when climate change made local environments challenging. As we’ve seen, another possible effect of decisive sanctioning would have been that disempowered alphas would have had problems in reproductively controlling a band’s females. This may well have opened the way for monogamous pair bonding to develop, or to develop further.

  I’ve taken 250,000 BP as the magic number for the likely beginning of really strong conscience evolution, but new facts could make for an adjustment of this hypothesis. If it were discovered that some archaics were depending on intensive hunting 400,000 years ago, then that might be our date instead of 250,000 BP. The same would be true if some new Homo erectus site showed systematic hunting of large ungulates a million years ago, although the brains involved would have been much smaller. However, these hypotheses would be difficult to reconcile with the fact that archaic Homo sapiens went from multiple butchers to a single butcher between 400,000 BP and 200,000 BP.