Humans get along at astonishing scales, but it doesn’t take an ivy league historian to notice a recurring ethos: help your friends and hurt your enemies. That may be common sense, but serious questions remain as to why this approach exists and how it works. Can an overarching theory of human cooperation be verified scientifically?

A new study by anthropologists Carla Handley and Sara Mathew, published in Nature Communications, puts this age-old intuition to the test. Examining the complex relationship of pastoralist tribes in northern Kenya, they present empirical evidence that, under the pressure of intergroup competition, ethnic solidarity is the direct result of cultural group selection. They argue that the effects of cultural forces are far greater than genetic predisposition or geographic proximity in promoting cooperation with nonkin. Both researchers have done extensive fieldwork in the region, and their rich personal experience underpins a rigorous quantitative analysis.

The Tribes of Northern Kenya

For millennia, rugged pastoralist communities have struggled for survival in the arid Kenyan landscape, battling drought and disease, and most significantly, each other. Watering holes and grazing territories are severely limited. Brutal cattle raids are a feature of everyday life.

Handley and Mathew’s experiment centers on four ethnolinguistic groups in the region: the Turkana, Samburu, Borana, and Rendille. 793 individuals were interviewed from 9 different clans within these broader groups to identify the contours of their cultural differences. An initial survey identified variations on norms regarding cattle-raiding, cultural markers, cooperation, family dynamics, and crime and punishment. Of these, the greatest differences were related to raiding and cultural markers—especially codes of conduct relating to cattle-rustling.

The initial survey was comprised of 49 normative statements. Subjects were asked whether they agreed or disagreed with such culturally significant items as:

  • It is permitted to kidnap women or children of the enemy during a raid.
  • A raiding party should not depart if they have not received blessings from the elders or a diviner.
  • Only unmarried men should go on raids.
  • During a raid, if the fight is tough, you can leave a wounded man to be killed by the enemy.
  • It is good for a warrior to boast about the raids he has gone on through his bull song.

For Kenyan pastoralists, the ethical question is not whether to raid, but how. Each of the four groups under study varies dramatically on these norms. The one consistency is in regard to raiding one’s own group.

For instance, the Turkana are quite adept at snatching livestock from herdsmen caught off guard. As in much of the pre-modern world, such behavior isn’t thought of as “robbery,” it’s simply a way of life. Courage is of prime importance, and young men who shrink from battle are shamed and beaten by their peers. In the Turkana’s northwest territorial section, Handley and Mathews inform us that a staggering 50% of adult male mortality is the result of cattle raiding. The competition for survival is fierce. But the Turkana are not without mercy. As a matter of principle, they spare fellow Turkana.

To tease out each group’s implicit attitudes toward ethnic solidarity, the study’s subjects were presented 16 scenarios in which they would decide how to respond to various ethical quandaries. One narrative involves the discovery of fine pastureland followed by an encounter with another cattle-herder from a specific ethnic group. Do you let him in on the prime grazing or not? In other scenarios, subjects were asked to imagine an opportunity to steal an unguarded goat, or that they’ve come across a neighboring clan under attack. In each case, the ethnicity of the target is emphasized. Do you help? Do you harm?

If the scenario’s target was of the same clan, cooperation was basically a given. However, the greater the cultural difference, the more likely the respondent would withhold help or actively harm. While each group exhibits interesting variations, cooperation rates for targets of a different ethnolinguistic group hovered between 0.4 and 0.6, whereas cooperation rates for those of the same group were well above 0.8.

The results suggest that in an environment of constant intergroup competition, parochial cooperation will emerge. When your neighbors are prone to unprovoked attacks, it’s wise to reserve the milk of human kindness for your own. But don’t shrug if you knew that already. The study’s theoretical implications are significant.

Testing For Cultural Group Selection

Handley and Mathew are among the first to systematically test the theory of cultural group selection against quantitative data. In its most prominent formulations, this theory claims that natural selection may either favor or eliminate a human group’s shared culture. The process mirrors environmental pressures on biological variations, such as a spider’s venom or a leopard’s camouflage. In any given environment, some cultural norms will do better than others.

We can reason that in hostile social conditions, behaviors such as pure selfishness or indiscriminate kindness will tend to disappear. The loner will be outnumbered. Givers will have everything taken away. A successful strategy must lie somewhere in between.

Can we be sure that loyalty toward the ethnolinguistic group is the result of cultural similarity rather than a perceived genetic relatedness? Not entirely, and some bio-cultural interaction is quite likely. But there is reason to believe that cultural identity is the primary object of affinity or aversion rather than bloodline. In the case of north Kenyans, the genetic fixation index is far lower than the cultural differentiation. Outside of one’s immediate family, there’s not enough difference to perceive. Any sense of tribal kinship is largely fictive. In a broad population with a high degree of genetic relationship, one’s clan can only be recognized by cultural markers and idiosyncratic norms.

Handley and Mathew also rule out geographic proximity as a critical component of the Kenyans’ ingroup bias. Indeed, their findings suggest that ethnolinguistic affiliation transcends physical distance, despite the fact that their respondents show a greater tendency to cooperate with their geographical neighbors. Overall, the effects of cultural differentiation are far more significant.

According to the authors’ causal model, geographic distance leads to cultural variation over time. Given an environment of intergroup aggression, which is somewhat universal in pre-modern societies, cultural group selection will tend to favor a warrior ethos that shows a preference for co-ethnics.

In addition to isolating culture as the binding force underpinning cooperation with nonkin, Handley and Mathew also argue that the ethnolinguistic group is a particularly stable unit of social organization. In the Kenyan pastoral setting, this ethnic loyalty appears to be a much stronger center of gravity than the clan or the nation-state. I would add that this tendency isn’t limited to African culture. As we observe the European Union and other pan-cultural alliances fracture along ethnolinguistic lines, one might suspect that human cooperation tends to stabilize at this social scale.

Complicated Realities on the Ground

Strict commitment to the ethnolinguistic group is hardly a homogeneous cultural mode. Handley and Mathew are careful to note the subtle contours of each group’s response to the ethnic other. For instance, the Borana are far more likely to cooperate with others regardless of ethnicity. Given the pervasive influence of Islam among the Borana, it may be that a universalizing religion will, to some extent, subvert the parochial barriers that naturally arise in competitive environments.

Another interesting finding is that the militarily weak Rendille tends to cooperate with the more dominant Samburu—even above other Rendille clans. In this case, it appears that submission to a greater political power can also subvert tribal loyalty, at least so long as that inequality persists.

Regarding personal kindness, many subjects reacted with compassion to one of the study’s hypothetical scenarios. In it, the subject finds an injured man from a rival tribe who needs to be carried to the hospital. The overwhelming response from all four groups was to be the Good Samaritan and help him.

Still, exceptions do not refute the rule. Handley and Mathew’s study is consistent with a recent experiment conducted on workers in the US, Germany, and France. In each case, corporate teams who faced intense competition from rival companies were far more likely to put their egos aside and exhibit group solidarity against the other. Hopefully, more research will be conducted in this direction. Quantitative evidence of greater individualism and outgroup sympathy in noncompetitive societies will be a critical point of comparison.

If competitive environments tend to produce parochial cooperation by way of cultural group selection—and if this is an emergent principle in human nature—then the implications are enormous. On the one hand, it highlights the potential flexibility of human alliance. Effective cultural institutions can overcome the perennial allure of blood and soil. On the other hand, this theoretical model uncovers the dark origins of large-scale social bonds. If human cooperation is a successful evolutionary response to feuds between fictive kin, an aspiring teambuilder should probably leave peace and love to the side and emphasize bonds that have stood the test of time.

For more, read “Human large-scale cooperation as a product of competition between cultural groups” in Nature Communications.