War Over War, Again



Writing in Scientific American, John Horgan states, “10,000-Year-Old Massacre Does NOT Bolster Claim that War Is Innate. A new report on a massacre of hunter-gatherers in Africa is consistent with the claim that war, far from being an inborn trait that evolved millions of years ago, is a recent cultural invention.”

What does it mean to call war an “innate behavior” or an “inborn trait”? This is an old-fashioned way of saying that some behavior is genetically determined. Actually, modern biologists stop using such terms quite a while ago. We now understand that human behavior is molded by a complex mixture of genetic predispositions, environmental influences, and culture (and don’t forget the key role that culture plays in shaping human behavior—for an excellent review of cultural evolution see Joe Henrich’s book The Secret of Our Success).

Furthermore, war is not individual behavior. How can you as a person have the trait of war, in isolation from other human beings? Even one group cannot have war; it takes two or more groups of people. So the whole framing of the debate as “Is War Innate?” is completely unhelpful. John Horgan should know better.

The debate over war (or War Over War, as I have referred to it in previous posts) is very important. We all want to eliminate war, but the best way to do it is not by wishing it away, but by using science to understand its evolution and the causes that make it wax and wane. This is one of the central concerns in my recently published book, Ultrasociety: How 10,000 Years of War Made Humans the Greatest Cooperators on Earth.

Below is an excerpt from Ultrasociety, in which I review the debate on war before states arose, and reject both the extreme position of John Horgan and the opposite extreme (currently occupied by Steven Pinker).



Excerpt begins:

What do we know about the incidence and intensity of war before the states arose?

Many of my readers will know that this is an extremely controversial question. There is a bitter ‘war over war’ in academia, which periodically spills over into the blogosphere and popular magazines. Because my conclusions critically depend on the answer to this question, I also need to enter this contested field.

There are two extreme positions, neither of which makes sense to me. The first one is the myth of the peace-loving “noble savage” of Jean-Jacques Rousseau. Even when these savages fought, their wars were somehow non-lethal and non-serious, even comic affairs (according to the Eurocentric notions of ‘primitive war’ that we discussed earlier).

The myth of peaceful savage was demolished by Lawrence Keeley in his ground-breaking book, War Before Civilization. Keeley writes, in particular, of how archaeologists “pacified the past” by refusing to see evidence of prehistoric warfare, sweeping such evidence under the rug when it “stared them in the face.” He collected data from archaeological and ethnographic sources and demonstrated that death rates (in other words, the probability of being killed in war) were an order of magnitude higher in pre-state societies than in our own.

The opposite extreme is the view that the distant human past was an unrelenting Hobbesian “war of all against all.” This position has been recently occupied by the psychologist and author of popular books Steven Pinker in The Better Angels of Our Nature: Why Violence Has Declined. Here’s how Pinker starts Chapter 1, A Foreign Country:

If the past is a foreign country, it is a shockingly violent one. It is easy to forget how dangerous life used to be, how deeply brutality was once woven into the very fabric of our lives.

The bulk of Pinker’s book is devoted to showing that the long-term trend for all forms of violence, including homicides, civil wars, and interstate wars, has been one of decline. There were some local peaks and valleys, but the violence curve starts very high and then gradually declines. It’s a “declining sawtooth.”

Pinker’s book triggered a lot of controversy, with both supporters and detractors dissecting the data on which his conclusions are based. Of particular interest to our goals in this book is the assessment of the Pinker thesis by academic anthropologists. One of the most thorough such critiques is War, Peace, and Human Nature, a collection of articles by a number of eminent archaeologists, anthropologists, and primatologists, edited by Douglas Fry.

In his summary of the evidence, Fry makes several excellent points. He agrees with Pinker that after the rise of large-scale states, or roughly over the last 5000 years, the overall trend in violence has been downwards. But Fry fervently disagrees with Pinker about the trajectory during the first 5000 years—after the adoption of agriculture, but before the rise of the states. He argues that violence, and especially warfare, actually increased, before it started to decline.

I agree. Multiple lines of evidence suggest that during the last 10,000 years the curve of war can be represented with the Greek letter Λ (lambda). Both the ascending and the descending trends are of course ‘jagged,’ because there were local increases and decreases superimposed on the long term Λ-trend. The peak position also varies among world regions, and generally coincides with late pre-state and early state societies.

However, Fry and others who contributed chapters to War, Peace, and Human Nature, go too far when they suggest that “war was simply absent over the vast majority of human existence” prior to 10,000 years ago.

Yes, during the climate chaos of the Pleistocene, warfare was probably rare. Human populations were in much greater danger of being wiped out by an advancing glacier than by another foraging band. When glaciers receded, enormous areas opened up for human colonization. Avoiding aggressors by moving away was both preferable and feasible. Yet there must have been periods of relatively stable climate when the local landscape would fill up. Nomadic foragers can be as territorial as farmers, and will defend rich hunting grounds or patches of valued plant resources. Once one group resorted to violence, war would spread: pacifist groups would be eliminated by natural selection. Such episodes of warfare could have been relatively rare during the Pleistocene, leaving no clear evidence in the archaeological record. If someone was killed by a well-thrown stone (or died later of the injury), how could we distinguish that from another unfortunate person who died in a hunting accident? In any case, we have very few skeletons from the Pleistocene, leaving us with scarce evidence for statistical analysis.

An additional problem underlying the prehistoric war controversy is that different people use different definitions of war. So let me be clear about what definition I use. My primary interest is in cultural group selection, and thus I define warfare as lethal group-on-group violence, no matter what forms it takes (battles, raids, ambush of stray individuals, etc.).

By this definition, both chimpanzees and wolves fight wars. Take Yellowstone National Park, where wolves were reintroduced in 1995, having been hunted to extinction earlier in the twentieth century. Once the wolf population increased to the point where all territory was divided up among the packs, between-pack violence flared up, ultimately becoming the chief cause of wolf mortality. We now have at least one example of a pack exterminating another one. It happened not in a single glorious battle, but by a thousand cuts, picking off one individual here, another there. The winning pack then expanded its territory and split into two.

When one pack of wolves exterminates another, I call it warfare, because it is between-group competition carried out by violent means. Other scholars use different definitions. Some insist that conflict should be “organized” to count. Others consider only large-scale conflict and exclude ‘primitive war.’ Such alternative definitions may be as valid as mine, being appropriate to the kinds of questions and conceptual approaches that other investigators use. But I am interested in warfare as a form of between-group competition.

We run into additional difficulties when trying to assess the prevalence of war in prehistory. Clearly we need to distinguish between interpersonal violence and group-level war. This can be difficult. A skull bashed in by a blunt object may indicate a death in battle or a murder resulting from a domestic dispute. For this reason, many anthropologists want to see additional indicators of group-level conflict before they can agree that it was warfare. Such archaeological signs could be fortifications or weapons specialized for man-on-man fighting (warclubs, swords). Bows and arrows, however, are equally useful in hunting and war. As a result, much warfare between small-scale societies, who tend to use ranged weapons and rely on raids and ambushes, will be invisible to archaeologists.

Let’s step back from this debate and consider how it affects the question we are currently investigating, the role of war in the rise of archaic states. While there is confusion resulting from alternative definitions, and a great degree of controversy about evidence and how to interpret it, all parties agree on one thing. Warfare was particularly vicious among pre-state farming societies. There is a lot of empirical support for what I called the Λ-shaped curve of warfare during the last 10,000 years. It is quite possible that the period after agriculture spread but before states arose was the most violent one in human history—at least when measured by the proportion of people who died as a result of war. If this is correct (and this is a very active research area, so we should expect more data soon, especially as the methods of forensic anthropology become better), it would strengthen the proposed link between war and the evolution of states. Watch this space.

Notify of
Most Voted
Newest Oldest
Inline Feedbacks
View all comments
Gene Anderson

Seems right to me, from a lifetime of reading about this. Don’t ever live in a chiefdom. The massacre near Lake Turkana was late in prehistory and the groups involved were apparently fairly large, and at least somewhat sedentary or localized around a lagoon–a choice bit of property. I expect these were moving in the chiefdom direction, not just tiny wandering bands.

Raymond Scupin

For those interested in these issues of warfare, I would recommend Azar Gat’s essay “Proving Communal Warfare Among Hunter-Gatherers: The Quasi-Rousseauan Error in Evolutionary Anthropology 24: 111-126, 2015. I think this essay is a very nuanced analysis of Fry et al. I can send the PDF to anyone interested.

al loomis

war is innate in hierarchical society. when one man’s ego is bloated by power over many, he will use the many to satisfy his greed.

Bryan Atkins

Greets Dr. Turchin,
Kind of just discovering thee, so my question / comment may be boring & redundant for you.
Re: “How can you as a person have the trait of war, in isolation from other human beings?”
(I get that innate is out of favor, and that multiple, complex factors are in play — with almost everything.)
But if group selection is true (think it is), couldn’t war be innate, per the following?
Stealing from Robert Kurzban and E. O. Wilson, I see war, rape, love, jealousy, cooperation, genocide, etc., as apps on file, part of the genetic repertoire for reality / relationship interface that may or may not be summoned per the variant conditions of the environs.
Think the war app is likely genetically structured in individuals (innate) and available for use by individuals in the group setting.
I see reality as largely the ongoing computation of information-in-relationship (omits time, etc.). Part of that computation yielded the war app for human groups.
Any thoughts?
Thanks . . .

Matthew Zefferman

I’m going to plug my recent paper with Sarah Mathew where we make a similar appeal to get away from arguments over war’s innateness.

“Long before Mead declared that “war is only an invention – not a biological necessity,” anthropologists have disagreed on whether war or peace is the natural state of humankind. Some
scholars present evidence that prehistoric small-scale societies were predominantly warlike, suggesting that humans have an evolved predisposition for engaging in lethal intergroup violence. Other scholars sharply disagree, citing numerous societies having little or no evidence of warfare. Both sides can marshal ethnographic and archeological examples, and have for decades. However, neither side has adequately explained why there are some societies where peace prevails and others, such as communities among the Turkana, in which 50% of adult male mortality is due to warfare.” (note that we are referring to modern-day Turkana.)

Zefferman, MR and S Mathew. (2015) An evolutionary theory of large-scale human warfare: group-structured cultural selection. Evolutionary Anthropology. 24:50-61.

Chris Kavanagh

John Horgan has a tendency to make unilateral assertions about what is and isn’t science and seems convinced that his experience as a journalist makes him better equipped than researchers to objectively assess entire fields. I don’t find that argument convincing nor do I think that he is particularly objective as he has a clear tendency to cherry pick studies that support his preferred position. As such, and as your post demonstrates, his article provides a good illustration of the problem with the black and white thinking that surrounds this issue. The extract from your book displays much greater nuance, it would be great if such a response became more widespread…

Lucien Chardon

Is Horgan disputing the conclusion that the rates of killing among pre-state societies may be higher than we previously thought, or is he denying that this conclusion could be deployed in service of the claim that war is a product of tendencies ineliminable in human nature?

Peter Moss

Pinker’s thesis is muddled. He conflates war with violence, and goes on to conclude that there is less of both now compared to the past. This is false. There is actually more violence now than ever before, albeit less war. Certainly there are more threats of violence, from all kinds of sources. The state is the most violent institution ever invented by the human mind.


How are you defining “violence”?

There certainly is a smaller chance of dying a premature death due to being deliberately killed by other people now than at most points of ancient human history or even a few centuries ago.


OK. But I still don’t see evidence of more violence now (even if you include corporal punishment) than centuries past.

Anyone care to argue that we now see more nonlethal violence against minorities, women, children and/or slavery, human sacrifice, witch hunting, cruel punishments, etc. than we did centuries ago?

What about lethal violence? Someone want to argue that we see more lethal violence now than we did centuries ago? I do not see that either. Just read Pepys’s diary. People got run through with swords or got their heads blown off by a cannon, etc. on a fairly regular basis.

Peter Moss

Violence is all around you, if you have eyes to see. Did you grow up 100% completely free of violence? Did you ever get into fights as a kid? Most young people get into scraps, sometimes serious ones. Were you abused as a child? Lots of people have abuse in their background. Were you ever bullied as a teenager? Were you ever assaulted? The victim of a crime? Of a police beating?

Think about laws. There are tens of thousands of laws on the books, governing every aspect of your existence. Each and every one of those laws is backed up by the violence and terror of the state. There are more people locked up in prison (most for nonviolent crimes) around the world than at any other time in human history. The existence of violence, or the very threat of violence, is omnipresent in modern civilization, on a scale unprecedented in human history. Everyone, and I mean everyone has either experienced violence at some point in their life, or else knows someone who has.

To make the grand sweeping claim, as Pinker does, that all violence, in all its forms, has slowly declined since the dawn of man, and then pretend to be able to somehow quantify this including the everyday nonlethal violence of even our homo habilus ancestors (as if this could be known) is so absurd, so ridiculous, so utterly lacking as a scientific hypothesis, it insults his reader’s intelligence.


Saying that there is violence now is a weak argument for the claim that the level of violence has not gone down during human history.

It seems very clear that you haven’t done much historical research or read stuff like Pepy’s diary or Dickens.

It’s rather clear to me that violence and cruelty in first-world countries today (even in the gun-happy/shooting-happy US) is much less than the violence and cruelty that existed in 17th century London or 19th century London.

Peter Moss

You haven’ addressed any of my points. Saying, “It’s clear to me…” isn’t an argument.

Mike Waller

In my view, the transcendent human trait is the desire to be well thought of by significant others. With infants it gives confirmation that care-givers consider the child a viable vehicle for the transmission of shared genes; with adults, that the individual is a viable breeding partner or, at least, worthy of cooperating with. As I have recently suggested to Peter, for reasons having to do with kin selection and familial reputation, if we are denied such emotional buttressing, the outlook is exceptionally grim.

Outside that, the distinguishing feature of humans is the sheer size of the cerebral equivalent of the computer’s Random Access Memory. This is unsurprising because we have evolved to become opportunistic problem-solvers, forever seeking out new ways to exploit our environments, including the socio-cultural contexts within which we find ourselves. In respect of war, this means that if the consensus amongst our reference groups is that it is either desirable (Spartans, Crusaders, and other “warlike” groups) or essential for cultural self-preservation (stance taken by Allies in both World Wars), all but the most independent-minded take up arms and consider it “right” to do so. Where there is an influential consensus against war (UK at the time of Munich; US re. Vietnam) peace movements afford the emotional buttressing for individuals to stand out against personal participation. On this basis, it seems to me that the lead article above has it bang on the money, war is not something we are hard-wired to engage in. It is simply something to which we may well resort if it seems to enough people either the only way to protect what we hold dear or simply the best way to further our own interests.


By that same logic, peace is not something that we are hard-wired to engage in either, no? (It is simply something to which we may well resort if it seems to enough people either the only way to protect what we hold dear or simply the best way to further our own interests.)

Mike Waller

I am sure that is right although one has to accept that there are exceptional individuals to who pacifism is so central to their self-image that they will avoid violence at all costs. This, presumably, is the literal meaning of “turning the other cheek”. Yet even amongst professing Christians it is obvious that, as a principle, THIS has been honoured far, far, more in the breach than in the observance.


It is a good point, to separate violence (individual vs. individual) and war (group vs. group) as different structural level processes. It will be very fascinating to look on a correlation between violence and war in historical prospective. Is it negative or positive?

Mike Waller

In the democracies, one of the fears of having had millions of their populations engaged in world wars was that this kind of savagery would be brought back home. My impression is that no such correlation was found, suggesting that they are independent entities. However, caution needs to exercise because (a) for all the millions in the armed services, it was a much, much smaller number who were directly engaged in the kind of “hand-to -hand” fighting which is the direct parallel with most criminal violence; and, (b) even amongst that much smaller group, many claimed never to have shot to kill. Whether you would get the same answer if your looked at elite shock troops is certainly worth investigating, but even there a distinction would have to be made between PTS induced violence and violence as a thought through means of achieving criminal ends or to meet some pre-existing psychological need..

That said, what would seem to me to be the key distinction is societal approval. To those with an interest in law, order and the protection of life and property, a criminal thug is a criminal thug. In contrast, a war hero is………a war hero; and as such treated very differently..

  1. Home
  2. /
  3. Cliodynamica
  4. /
  5. Regular Posts
  6. /
  7. War Over War, Again

© Peter Turchin 2023 All rights reserved

Privacy Policy