Moral Origins

by Christopher Boehm

  HOW “LOOSE” IS THE PROCESS OF NATURAL SELECTION?

  Apparently, natural selection hasn’t managed to set up very effective barriers to cut off such generous helpfulness whenever it is being directed at nonkin. Were egoism and nepotism the only forces driving natural selection for humans, and were the processes involved totally efficient and totally determined by biology, we would expect something quite different. Indeed, first we would expect ourselves to have evolved some foolproof means of identifying kin so as to avoid donations to nonrelatives—and also of making sure we helped our kinsmen only in proportion to the degree of mutual relatedness. And second, we should have evolved to never give costly, uncertainly reciprocated help to nonkin, for such “genetic self-sacrifice”27 defies the very notion of the efficient natural selection process that evolutionary biologists assume to be operating when they create mathematical models of gene selection.

  George Williams, himself a biological mathematical modeler par excellence, describes forcefully how natural selection’s lack of total efficiency could occur even with respect to something as basic as reproductive behavior, saying that “reproductive functions, perhaps to a greater extent than any other adaptations, are characterized by a considerable degree of looseness in timing and execution.”28

  In this context, he notes that homosexual behavior is widespread among animals. If such looseness applies also to benevolent helping of nonkin, Williams predicts that the help to “unrelated animals should never be more intense, and should usually be less intense, than the same behavior toward offspring.”29 Thus, in effect, the very substantial benefits of kin selection could be genetically subsidizing the occasional acts of wider generosity that go astray. Williams does not propose that this type of “misplaced reproductive function” could be the sole explanation for altruistic helping behavior, but he does point out “that when an animal actively assists an unrelated individual, it uses only those behavior patterns that are seen in a family setting.”30

  Although intriguing, misplaced reproductive effort is not mentioned very often as a possible solution for the mystery of altruism in human or other species. Although thousands of evolutionary scholars have been struggling to resolve this paradox for these past four decades,31 they’ve done so mainly through theories involving very efficient mechanisms that directly compensate altruists for the losses inherent in being altruistic.

  The effort to explain extrafamilial generosity as an important component of human cooperation continues in many directions, but I shall begin with this “looseness model” that I have just introduced.

  1. “MISTAKING” NONKIN FOR KIN

  Kin selection is a powerful agency that can sustain generosity to blood relatives in accordance with the strength of the blood tie, and this model readily accounts for a fair amount of the generosity seen in a hunter-gatherer band that usually consists of about 25 percent close kin.32 What the model doesn’t account for, however, is all the generosity shown to nonkinsmen—unless, somehow, some of this generosity is simply “spilling over” owing to natural selection’s inefficiency as Williams suggests. This might be called a slippage model, meaning in the case of altruism that overall the individual advantages of nepotistic generosity are so strong that moderate amounts of costly extrafamilial generosity could be “piggybacking” on them genetically,33 with no major harm done because the inclusive-fitness advantages of nepotism are so strong and the costs of being altruistic are weaker.

  For humans, one immediate agency that could facilitate such slippage is our cultural inventiveness in assigning people to social categories. By custom, people in bands sometimes use primary terms reserved for close kin, such as “mother,” “uncle,” “sister,” or “brother,” to refer to distant kin or to nonkin they are closely bonded with. If we assume that the use of such terms summons up sympathetic feelings, generosity to unrelated others could come into play because in effect band members are “tricking” themselves in ways that redirect their sympathetic generosity from kin to nonkin. In fact, you may be far more closely bonded with a generous nonrelative with whom you spend a lot of time collaborating in subsistence activities than you are with your own selfish cousin, whose niggardly habits tend to rub you the wrong way. In that case, you’re more likely to give assistance to the nonrelative in time of need even though, at the level of genes, you’re not being nepotistically compensated through inclusive fitness.

  The genes that make this “imprecision” possible might be considered multipurpose or, technically, “pleiotropic.”34 This explains how some moderately maladaptive assistance to nonkin can “piggyback” on the highly adaptive assistance to kin. However, as Williams suggests, over evolutionary time this slippage-based beneficence can continue only if these two behaviors in combination bring a net relative-fitness benefit to the generous individuals involved.

  2. CULTURAL DOCILITY AND GENEROSITY

  Economist Herbert Simon’s docility model provides quite a different kind of piggybacking possibility.35 This model requires no sympathetic feelings, at all. In Simon’s conception, there are enormous advantages stemming from more “docile” individuals being better set up to automatically copy useful behaviors from other culture members without having to engage in costly trial-and-error learning. For example, out on the Kalahari Bushman parents tell their children exactly where poisonous snakes are likely to be encountered. This lesson is best learned vicariously, and I was grateful to learn it that way when I traveled to central Africa to study conflict management among wild chimpanzees. (My personal and dreaded serpentine “favorite,” by the way, was the hyperaggressive black mamba, followed by spitting cobras.) When I arrived at Gombe, one of my first questions was about snake hazards, and the long list Jane Goodall told me about also included green mambas, boomslangs, night adders, water cobras, nonpoisonous but seriously sizable pythons, and tiny vine snakes that can barely bite but will kill a person with nerve poison if they do. There were just enough deaths on record to make it obvious that internalizing this information in advance was far better than relying on trial and error, and my cultural docility, combined with a truly major fear of snakes, helped me to stay safe.

  Consider now that among all the personally useful cultural patterns that group members pass on to the next generation, there will be pointed, golden-rule-type messages that call on individuals to be generous even to nonfamily. When a person automatically acts on such messages, the moderate altruistic costs will subtract from the very substantial overall individual benefits that come from being generally so spongelike with respect to cultural learning. But as an innately absorptive learner, the person will still be making a net gain.

  Of course, “nonconformists” who inherit faulty sponges can easily resist such messages to be altruistic, but they will also be losing out on the general advantages of conformity. For example, they’re more likely to die of snakebite. On the other hand, someone who is able to internalize all the useful rules of living but resist the cultural messages that promote altruism could be a free rider on this system.

  3. BEST GROUPS WINNING

  As we’ve seen, group selection provides a very different kind of explanation, and it need not involve the altruistic individuals’ somehow being compensated. Were group selection strong enough, it could straightforwardly support individually costly, empathy-based, group-useful cooperation among nonrelatives in the same band36—especially if all the band members were staying put for life.37

  Earlier opinions in evolutionary biology held that group selection could work only rather feebly,38 but there was no disagreement that its effect would have been to contribute to the evolution of sympathetic tendencies that help to motivate genetically costly generosity toward anyone in the group, be it nonrelatives or relatives. Just as Darwin said, this can lead to more surviving offspring in high-altruism groups than in their competitors, which exhibit less sympathy, less generosity, and, at the bottom line, less cooperativeness.

  One major argument of anti–group selectio
nists has been that not only is group selection inherently weak, but the models also are highly vulnerable to free riding.39 With respect to inherent weakness claims, technical simulation work by economist Sam Bowles40 demonstrates that for humans group selection could have been a major force prehistorically, because of major genetic differences between groups. With respect to free riders, from the standpoint of modeling the evolution of altruism they are a nasty piece of work, as it were. In past mathematical modeling they are basically “cheaters” who are designed to take advantage of gullible, vulnerable altruists by deceptively taking without giving.41 Thus, these freeloaders can cash in on the benefits of cooperation without paying any of the costs, which means that as individuals they will easily outcompete the altruists, whose genes thereby lose out and—in theory—all but go away. If this free-rider problem could be eliminated or seriously ameliorated, group selection’s power to support extrafamilial generosity would be increased. The arguments in this book will point in precisely that direction, not only for group selection but also for several other models I’ll be discussing here.

  4. RECIPROCATED “ALTRUISM” AS A THORNY QUESTION

  Biologist Robert Trivers’s reciprocal altruism model is quite beautiful, in its long-term, “tit-for-tat” symmetry. It’s appealing because in theory this model might account for a great deal of the extrafamilial generosity we see.42 However, this would be so only as long as the unrelated pairs involved were cooperating consistently over the long run, and only if certain other conditions were being met, namely, the exchanges would have to be reasonably equilibrated, which means no major cheating.43

  Given my definitions, there’s a semantic problem with calling such balanced reciprocation “altruism,” in that neither party is paying any special costs; indeed, they are both coming out ahead in comparison to individuals who do not pair up and thereby forfeit the advantages of cooperation. However, the real problem is that indirect reciprocity—which is what cooperative foragers actually practice as they help others in need or share their large game—is far from being either dyadic or anywhere near exact in the relative long-term contributions of different individuals or households.44 In fact, the amounts of communal meat that individual hunters provide to their bands over a lifetime can vary quite substantially.45

  Where might this elegant and seductive, game-theory-based model of Trivers’s apply, then? In humans, probably the closest thing to a real-life relationship in which major contributions of two unrelated partners can equalize out over time and bring major mutual benefits would be a stable, lifelong marital-procreative arrangement. In the absence of sexual cheating, enormous and quite equal reproductive benefits automatically accrue to both parties in the form of offspring. The main type of free-riding deception that would throw this kind of reciprocity seriously out of balance would be cheating by females, because this can result in the male partner’s investing heavily in parenting his unrelated genetic competitor’s child. Marriage also involves economic reciprocation, and with most of the prehistoric types of foragers we’ll be meeting with in the next chapter, husbands and wives make rather different but overlapping types of contributions to the family economy; this means that the costs and benefits would be cognitively difficult to balance even if the two unrelated partners were trying to count every last bean—which in fact they aren’t.

  It’s curious that so little has been written about this very special, sexualized, two-person version of reciprocated extrafamilial generosity, for whether a perfect give-and-take equilibration exists or not, marriage partnerships can bring very large reproductive advantages compared to not pairing up. As a preferred way of doing things, procreative pair bonding appears to have been universal among humans at least since we became culturally modern, and given the likely payoffs, its contributions to our altruistic potential might have been significant.

  Trivers’s famous model has been employed optimistically by many scholars to explain human cooperation in a variety of other contexts, where the fit with everyday behavior is, I believe, far less compelling. However, the use of his model to explain the affectively warm, generous, sexualized pair bonding of marital partners seems quite promising, and I think this merits further exploration. Obviously, free-riding issues, such as laziness or sexual cheating, do pose problems, but it’s worth noting that in general small foraging groups universally frown on major laziness or adultery and sometimes, at least, punish them harshly. In addition, divorce appears to be a forager universal, and it offers some protection against cheating.

  5. IMMEDIATE, WELL-EQUILIBRATED MUTUAL COLLABORATION

  Short-term mutualized collaboration resulting in balanced benefits involves two partners of the same species engaging simultaneously in one-shot cooperation in contexts so immediate that cheating can simply be set aside as a free-rider issue.46 Cooperation based on mutualism does occur in real life, as when two African foragers are smart enough to quickly gang up against and bluff away a big-cat predator that could easily have taken either person had each acted solo. As a nice example, the Hadza of Tanzania do their night hunting around water holes in pairs because lions will pick off single hunters.47

  Today, one-shot mutualistic approaches to explaining cooperation in various species have largely replaced Trivers’s very demanding long-term dyadic model, but in my opinion their potential for explaining ongoing extrafamilial generosity in human foragers seems limited. This is because the nutritionally important indirect reciprocity actually practiced by foragers is so far from being immediate or restricted to dyads or balanced; indeed, forager contingent meat-sharing and security-net systems last over lifetimes and involve bands of several dozen people—with families constantly changing bands.

  6. SOCIAL SELECTION AND GOSSIP-DRIVEN PREFERENCES

  It was in this socially flexible context that biologist Richard D. Alexander influentially coined the term “indirect reciprocity,”48 and he emphasized superior outcomes in mate choice as a major mechanism that could be supporting the substantial extrafamilial generosity involved in helping those in need according to one’s resources. The idea is that being generous makes you look competitively attractive as a partner in a situation of cooperation, be it marital or otherwise. Thus, when others choose you preferentially over someone else who does not display such generosity, this is good for your relative fitness, and the costs of being generous can be more than compensated by benefits coming from being chosen more readily as a partner in cooperation.49 The idea is that people who are able to partner up, and partner up well, will outreproduce those who fail to do so.

  Alexander did see cheating as a serious potential problem in that people could manipulatively showcase or even dissemble generosity. This means that cheating free riders would present problems here, just as they did with group selection and reciprocal altruism. But cheating aside, if it were strong enough, selection by reputation could go far in explaining the often highly contingent type of generosity that is inherent in social systems based on indirect reciprocity—and this is exactly what hunter-gatherers actually practice when it comes to sharing large game or when it’s time to help those who are sick, injured, snakebit, or otherwise seriously unlucky and in need of help.

  In 1987, Alexander’s big-picture “modeling” was anthropologically down to earth, meaning that it had the advantage of being keyed to actual behaviors in the same type of people who most recently evolved our genes for us. His favorite extant group was obviously the exceptionally well-studied Kalahari Bushmen and the !Kung in particular. In more recent years the selection-by-reputation model that Alexander formulated has been explored in the laboratory, using mainly student-based, game-theory experiments, while indirect reciprocity also has been investigated out in the field anthropologically, sometimes by examining meat-sharing systems that accomplish variance reduction, sometimes by focusing on “costly signaling,”50 and sometimes by studying social behavior that provides social safety nets.51

  People’s reputations are determined by what other
s see them doing, but even more by their being talked about. Language permits individuals in small groups to exchange such firsthand and secondhand information, the result being a thorough and very useful general knowledge of people’s reputations. Not only are good reputations known, but also bad ones,52 and both are taken into consideration when social choices are being made. For example, a person who is unusually generous would be given some preference as a subsistence partner or marriage partner, whereas an unusually selfish individual who is prone to bullying, cheating, or theft might be carefully avoided by those in a better position to choose.

  Add all of this up, and selection by reputation appears to have been a powerful agency in shaping certain behavioral aspects of the human gene pool. As a mechanism similar to Darwinian sexual selection, it favored costly traits involving self-sacrificial generosity but also others that were not costly. For instance, being personally dependable may not be costly, but it is an attractive trait. Also attractive is being hardworking, which in fact would be highly useful to individual fitness regardless of whether an individual was in a position of being chosen or not. Again, the theoretical joker in this selection-by-reputation pack would be free riders, who are able to fake the desired qualities and thereby make themselves as attractive as the real good guys.

  SUPPRESSING FREE RIDERS

  These six hypotheses for the support of extrafamilial generosity have been widely debated, but the free-rider suppression we’re about to discuss brings a new way of approaching the altruism paradox. For humans, the active, punitive social suppression of free riders does not directly select for altruism; rather, it disfavors these classically deceptive born enemies of altruists by either completely suppressing their behavior at the level of phenotype or by placing them at a net genetic disadvantage. These effects open the way for mechanisms such as selection by reputation or reciprocal altruism or group selection to support altruism more effectively.