“We know that we are apes, but we cannot be classified simplistically as ‘naked apes’ or ‘killer apes’ or ‘moral apes.’…Our past is complicated; so is our present, and so will be our future.” – Paul Ehrlich (2000: 331)

“When we are bad, we are worse than any primate that I know. And when we are good, we are actually better and more altruistic than any primate that I know. ” – Frans de Waal

The Eagles headed back to their cabin feeling dejected after losing a tug-of-war contest to their rivals, the Rattlers. Along the way, one of the boys noticed the Rattlers had forgotten their flag on the baseball field, leaving it unprotected. Craig and Mason soon seized it, but struggled to tear it to pieces. McGraw then presented some matches and suggested they burn it instead. The group then hung the flag’s charred remains from the top of the backstop fence. Mason said, “You can tell those guys I did it. If they say anything I’ll fight ‘em.”

The above scene is from the psychologist Muzafer Sherif’s classic social psychology experiment at Robbers Cave, Oklahoma during the summer of 1954. Sherif divided twenty-two 11-year-old boys with comparable backgrounds into two even groups at nearby cabin sites, with the boys kept unaware of the other group’s existence.

After giving them a week to bond among themselves, Sherif introduced the groups to each other and announced that they would be competing for prizes in team sports and other events. Eventually the rivalry grew heated, and the boys turned to name-calling, flag-burning, and vandalizing each other’s cabins. The competition nearly escalated into serious violence, with sticks and rocks as potential weapons, before adults intervened.

Sherif’s experiment is sometimes cited as a depressing warning of how easily people can slide into “us versus them” hostilities, even if the groups are formed rather arbitrary, and even if we’re only talking about preadolescent boys with little at stake except ego and trivial prizes. There is truth to that warning. People can cling tightly to group identities, sometimes resulting in serious animosity toward outsiders.

Rivalries that fall along national, ethnic, ideological, or religious divides have the potential to escalate into more severe vitriol, violence, and even war: Sunni–Shia, Muslim Seleka–Christian anti-Balaka, Israelis–Palestinians, Kyrgyz–Uzbeks, Hutu–Tutsi, Russians–Ukrainians, Americans–Afghans, etc. Certainly, rivalries are not the entire story of human group interactions, but they occur often enough that many have wondered whether group conflict is an inevitable part of human existence. Others have asked if the roots of intergroup conflict stem from the base of our evolutionary lineage as a species, or possibly even earlier.

These are difficult questions, partly because our past is often mythologized, where it is tempting to make sweeping generalizations that venture beyond the facts. One famous example includes Barack Obama’s Nobel Peace Prize acceptance speech when he said that: “War, in one form or another, appeared with the first man.” Winston Churchill thought similarly.

However, such claims are often based on an imagined past, not evidence, and reconstructing an accurate picture of our species’ prehistory is a notoriously difficult task. To do so, researchers usually round up some of the usual anthropological suspects by studying: (1) our primate cousins, to ascertain possible shared characteristics we may have inherited from a common ancestor; (2) living hunter-gatherer populations as an (imperfect) model of our more recent evolutionary past; (3) cross-cultural variation, to assess the full range of human behavior; and (4) most directly, the archaeological record.

By now, it is fairly well-known that chimpanzees, one of our genetically closest relatives (the other being the peaceable bonobo) engage in lethal violence within and between neighboring groups. In a paper published last September in Nature, Michael L. Wilson and twenty-nine fellow primatologists reported that across eighteen chimpanzee communities in east and west Africa, there had been 152 killings over 426 total years of field observations (Wilson et al 2014). Furthermore, killings were observed in fifteen of eighteen of the communities, including those with minimal human interference. Most victims (63%) were from different communities than the attackers, and the authors concluded that killing among chimpanzees was adaptive and a means of removing unrelated rivals when one group had superior numbers and the potential risks from engaging in an attack were low.

What to make of this? On one hand, we can see tantalizing parallels between chimpanzee and human intergroup violent behavior. For one, chimpanzee killings are overwhelmingly a male activity, as they comprise 92% of participants in attacks. Killings also occurred over a wide geographic range, from Senegal to Tanzania, and were therefore not merely an aberrational phenomenon specific to one community or the result of human impact.

On the other hand, lethal violence was rather infrequent, as the majority of communities had rates of less than 0.2 killings per year. In some communities lethal violence had never been observed. These rates are not easily translatable to human societies today simply because we live in much bigger groups, and a single chimpanzee victim will be much more impactful. Still, if lethal violence among chimpanzees is adaptive, it appears to be only sporadically so, not a regular feature of daily life. Lethal violence is also more likely to occur under specific circumstances. For example, Wilson et al. found that chimpanzees were more likely to kill when there were more males around, and when population densities were higher. As is the case in humans, chimpanzee behavior is flexible and responsive to social and ecological variables. They are no more automatons than we are.

Elsewhere, Wilson wrote that he was agnostic about what chimpanzee lethal violence means for us, adding that “definitive claims about human behavior need to be based on data from humans.” But the human data are also tricky. The historians Will and Ariel Durant once suggested that: “War is one of the constants of history…In the last 3,421 years of recorded history only 268 have seen no war” (1968: 81). That seems rather conclusive: war is a part of who we are. However, the Durants’ calculation does not mean that war was present everywhere on the planet simultaneously, and if we cast our net wide enough it makes sense that we would find conflict somewhere. Nor does it necessarily mean that war was common in pre-history.

Contemporary hunter-gatherer societies are often viewed as stand-ins for our prehistory. While anthropologists emphasize that all societies change over time, the rationale is that prior to agriculture, humans acquired food in this manner for the vast majority (>90%) of our species’ history. When the anthropologists Douglas Fry and Patrik Söderberg (2013) looked at violence in twenty-one contemporary and historical mobile hunter-gatherer groups, they found 148 lethal incidents in the ethnographic record dating back to the early 1800s, with a median of 4 incidents per group.

Their findings were not very consistent with the idea that war and intergroup aggression were a consistent feature of mobile hunter-gatherer groups. Only a minority of lethal incidents occurred between different societies. Instead, 36.3% of lethal incidents were within the same local group (relatives, spouses, other group members), 48.7% were within the same society but not the same local group, and 15.0% were between different societies (neighboring cultures, missionaries).

Additionally, most killings were dyadic, consisting of one aggressor and one victim (that is, homicide). These often had personal motives such as sexual jealousy or revenge, rather than stemming from impersonal intergroup hostilities. Only 22% of the incidents involved coalitionary violence, where more than one person participated in killing more than one person. Roughly half of the groups (10 of 21) had no coalitionary violence events, while three had no lethal events at all.

However, the most obvious and direct way to understand violence in our prehistoric past is through the archaeological record. There are clear examples of skeletal trauma indicating that interpersonal and group-level violence appeared in prehistory well before modern nation-states. Our ancestors were not complete peaceniks. Like us, they were complex.

But this is a far cry from saying that war has been unending since the beginning of humanity. In fact, violence seems to have been relatively rare before the increasing populations densities and fixed settlements that accompanied agriculture. In a thorough review of prehistoric violence, the archaeologists Jonathan Haas and Matthew Piscatelli wrote, with emphasis, that:

“globally, at least 2,930 skeletal remains of Homo sapiens have been recovered at over 400 archaeological sites dating prior to 8000 BC/10000 BP… the small number of skeletal finds mentioned above showing ambiguous signs of conflict come from a comparatively small number of sites. Rather than demonstrating the commonness of ancient warfare among humans, consideration of the entire archaeological data set shows the opposite… The archaeological record is not silent on the presence of warfare in early human history. Indeed, this record shows that warfare was the rare exception prior to the Neolithic pressures of population densities and insufficient resources for growing populations” (2013: 182-3).

Prior to agriculture, there were several reasons that mobile hunter-gatherers would have little incentive to engage in warfare. These include small group sizes, an egalitarian social structure, kin and social ties to neighboring groups who were a source of mates and trade, low population densities that allowed easy mobility to new territory should tensions arise, and the relative lack of accumulated wealth that could be an incentive to attack one’s neighbors (see Fry and Söderberg 2013’s appendix for a fuller list).

So, were our ancestors naturally violent or peaceful? It doesn’t appear that war was ubiquitous in our hunter-gatherer ancestors. But perhaps a better answer to that question is “it depends.” As with chimpanzees, intergroup violence in humans likely ebbed and flowed situationally rather than being an inevitable outcome of human nature. Context matters, and any attempt to extrapolate clear trends of violence over human pre-history risks sacrificing nuance and understanding local variation. Along these lines, in a recent paper in the Yearbook of Physical Anthropology, the bioarchaeologists Debra Martin and Ryan Harrod write that patterns of violence via skeletal trauma “suggest variability, nuance and unevenness in the type, use, and meaning of violence across time and space and therefore defy generalizations or easy quantification” (in press).

Another reason the question is so difficult to answer is that humans are so behaviorally flexible, with a capacity not just for hate and conflict, but also for cooperation and love, even for non-kin and members outside of our group. Aggression and violence are part of our species’ behavioral repertoire, but at what frequency, and in what form, are more complex questions. We have the capacity to give someone the shirt off our back, and the kidney underneath it, but we can also pick up an automatic weapon and aim it at an ‘enemy’ or even participate in genocide. The challenge is not to figure out whether our species is inherently violent or altruistic, but why both extreme capacities are found within a single species, and what circumstances and social structures facilitate or impede those behaviors.

Finally, to return to the boys at Robbers Cave, the final stage of the experiment is at times forgotten. Sherif created situations that required the Rattlers and Eagles to put aside their differences in order to achieve “superordinate goals” that neither group could accomplish alone – pooling their money to pay for a movie (“Treasure Island”), fixing a damaged water tank that supplied both cabins, etc. By the end of camp, both groups agreed that they would travel home to Oklahoma City on the same bus. Sherif noted that the boys chose to sit intermingled, rather than strictly along group lines. On the trip home, they sang together (Oklahoma!), exchanged addresses, and at a refreshment stop the Rattlers volunteered to pay for malt drinks for both groups.

The lesson of Robbers Cave is not that the boys were biologically inclined to be vicious or virtuous. Rather, they were capable of moving in either direction. Same boys. Different circumstances. This makes sense. From an evolutionary point of view, flexibility is the very essence of behavior, as it allows organisms to respond to external conditions. After all, we can find violence in ‘peaceful’ bonobos, fruit-eating in alligators, and meat-eating in herbivorous orangutans. And perhaps one of our species greatest assets is our adaptability, to be flexible enough to live in a range of physical environments, consume a variety of diets, and respond to the other highly intelligent, behaviorally complex, humans around us. Sometimes we compete, sometimes cooperate; we come equipped for both.


Durant W and Durant A. 1968. The Lessons of History. New York: Simon & Schuster.

Ehrlich PR. 2000. Human Natures: Genes, Cultures, and the Human Prospect. Shearwater Books.

Fry D, Söderberg P. 2013. Lethal aggression in mobile forager bands and implications for the origins of war. Science 341(6143): 270-273.

Haas J and Piscatelli M. 2013. The prehistory of warfare, In War, Peace, and Human Nature. Pp. 168-90 in: D Fry (ed.) Oxford University Press.

Martin DL, Harrod RP. In Press. Bioarchaeological Contributions to the Study of Violence. Yearbook of Physical Anthropology.

Ridley M. 2003. Nature Via Nurture: Genes, Experience, and What Makes Us Human. New York: Harper Collins.

Wilson ML, Boesch C, Fruth B, Furuichi T, Gilby IC, Hashimoto C et al. 2014. Lethal aggression in Pan is better explained by adaptive strategies than human impacts. Nature 513(7518): 414-7. Link

Published On: January 18, 2015

Patrick F. Clarkin

Patrick F. Clarkin

Patrick F. Clarkin Associate Professor of Anthropology at the University of Massachusetts Boston. Dr. Clarkin is a biological anthropologist who has conducted research on the impact of social and evolutionary forces on growth, nutrition, and health. Specifically, he has focused on the long-term impact of war, refugee experiences, and poverty on the growth and health of Southeast Asian refugees (Hmong, Lao, Khmer). He is also broadly interested in in the causes and consequences of human conflict and cooperation.


  • Mark Sloan says:

    Patrick, interesting article about our ability to be either cooperative or warlike depending on where the advantage lies. The evidence that our arguably egalitarian hunter gatherer ancestors do not appear to have been predominantly warlike is encouraging (though they doubtless were violent one-on-one and perhaps as group punishment of norm violators who threatened the group).

    But we certainly seem to have, in perhaps excessive abundance, the mental machinery and inclinations for cooperating in hierarchies and in demonization of other groups that are necessary for war. It seems unlikely this sometimes costly mental machinery has evolved just since the emergence of agriculture which shifted the circumstances to favor war. For example, I have heard no one argue that present day egalitarian hunter gatherers who presumably have never practiced agriculture have different mental biology regarding cooperation in hierarchies, willingness to sacrifice to defend their groups, and even in demonization of out-groups.

    I have seen implied elsewhere that our hunter-gather ancestor’s way of life is best understood as the environment that shaped our war related psychology. Perhaps we might usefully understand our presumably egalitarian hunter-gatherer ancestors (with culture) as an almost anomalous phase regarding between group ‘warfare’. Then the emergence of agriculture coupled with our even older ability to cooperate in hierarchies (largely evolved in a pre-egalitarian and even pre-cultural phases closer to chimpanzees) would have put all the pieces in place for large scale war being ‘profitable’ against weaker neighbors.

    Now to figure out how to make it ‘obvious’ enough to all that war is now not just too often unprofitable, but too often disastrous, regardless of how attractive our mental machinery makes it appear.

    • Helga Vierich says:

      Mark Sloan raises a valid issue when he says “we certainly seem to have, in perhaps excessive abundance, the mental machinery and inclinations for cooperating in hierarchies and in demonization of other groups”.

      However I would respectfully suggest that the human tendency to demonize is not necessarily an attribute of intergroup relationships, hostile or otherwise. It seems more likely to me that the categorization of other people as not human got its evolutionary traction from something else entirely; form the dynamic of internal social control. While evidence for warfare is very scant in the archaeological and ethnographic record of foraging economies. What is however ubiquitous is the use of an escalating scale of social control. And it is not just within foraging economies that human beings shame and punish those who cheat, deceive, bully, or harm others. The punishments for various offences range from public humiliation through gossip and scandal, through beatings and lashings, all the way up to and including actual execution. Certain offences are severe enough to remove previous citizens from the “human” category, as public comment on trials of murderers or pedephiles will attest.
      It might just be part of evolved human nature to reject and dehumanize those whose behaviour has shown their behaviour to be untrustworthy to that extent.

      Humans have been subject to over two million years of positive selective pressure to enlarge the brain, especially the cortex, and, most particularly, the prefrontal cortex. The functional correlates of this have been resolved: the prefrontal cortex is the brain’s executive. It is the seat of reason and impulse control. What we may deduce from this is that self control, and the ability to find rational alternatives to cope with situations that give rise to strong emotions and urges, has been under positively draconian selection pressure throughout our evolution.

      I don’t think we will find the source of this selection pressure in looking for evidence of warfare between groups.

      I do not, however, agree that we might look at human cooperation in hierarchies as being sort of submerged in the “anomalous” phases of our evolutionary past when we were (as you say) “presumably” egalitarian hunter-gatherers. On the contrary, cooperative behaviour, participation in group activities, ritualized sharing, coalitional bonding that is strong enough to enforce “moral” behaviour and mutual aid even under trying circumstances, seems to me to have something to do with the selection for a bigger and better prefrontal cortex. Flexing the mind between impulses toward self gratification and compassion was not a contest between immorality and morality: it was a contest between being considered inhuman and being considered human. If you were to be fit company for other human beings, you had to be just, fair, loyal, courageous, kind, reasonably honest, and trustworthy. You could pretend to be these things and commit inhuman acts of meanness and cruelty (or worse) in secret, but when you were found out, your career as a respected member of society was usually over.

      A sharing economy based on egalitarianism is NOT incompatible with systems of ranking based on tiered responsibility for management and distribution of stored foods and other sources of human welfare.

      It is only incompatible with a system of coercive inequality, and with the kind of entrenched privilege that develops when leaders corrupt an interest in public service in favour of personal power and authority.

      In fact, among hunter-gatherers, when larger aggregations make more organized cooperation necessary, there are clear social controls in place to prevent highly regarded leaders from abusing positions of responsibility. Any indication that a person, so entrusted, is assuming the authority to coerce, is met with immediate resistance. People withdraw from cooperative activities. Some families may leave the area even if this means missing some important ceremonial events.

      If I did not know any better, I might even suggest that hunter-gatherers are alive to the dangers of the self-attribution fallacy that causes people to abuse power! It is worth noting that sensitivity to any hint of hubris, of deceit, of greed, and of coercive intent, are met with swift reprisals in the form of mockery and public denouncements. Such social controls are active even during the dispersed phases of the annual round.

      I would suggest that hierarchy, or ranking, based on reputation, is always latent in human groups. This is not some ancient behavioural algorithm from our deep evolutionary past. Like our disgust at immoral conduct and tendency to dehumanize, especially, those who murderer “in cold blood”, I would suggest that our ability to organize in groups, and to respond to the leadership based on honour, is fully alive in humans and has been put to use throughout our evolutionary past, to resolve conflict and ensure the welfare of the whole community.

      I agree that people do not go to war willingly or commit genocide unless powerful mental machinery has been activated. But that does not mean that this machinery was evolved for war.

  • Lesley Newson says:

    It would perhaps be useful as part of your effort to make sense of “human nature” if you defined what you mean by the phrase. Are you suggesting that there are some behaviours that are “typical” or “normal” in humans? Or is “human nature” just a buzzword that can’t be defined?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.