Like slavery war has a long history in our species, but it does have a history, a beginning, and perhaps an end. That is the sort of message you can take away from a paper such as Zefferman and Mathew’s An Evolutionary Theory of Large-Scale Human Warfare: Group-Structured Cultural Selection. War is a culturally mediated human phenotype, and one which requires particular contingent conditions to flourish, and others to diminish. Perhaps the best survey I’ve read on this topic is Azar Gat’s War in Human Civilization (which is $9.99 on the Kindle, so I just got a copy even though I already read this book!). But Gat doesn’t seem to come to any definitive conclusion as to the nature of war from what I could tell (the book is long and meandering). And that makes sense, because war is complicated, and the behavioral phenotype isn’t clear and distinct. What exactly qualifies as war? World War II clearly does. But how about the feud between the Hatfields and McCoys? Obviously there is a continuum, and we have to draw lines. The phenomenon which Zefferman and Mathew focus on is the paradox of mass intergroup conflicts which are defined by clashes between coalitions of unrelated males. More concretely, in wars between nation-states you have old men sending young men to their deaths in large numbers. Why do these young men put themselves in harm’s way?
The authors focus on two species as a contrast with humans, common chimpanzees and social insects, Argentine ants, which have been known to engage in war. War here can be thought of as coalitional intergroup conflict. Chimpanzees are informative toward any discussion of human evolution because they are phylogenetically close to our own lineage, while social insects are not, but like humans are highly complex in their organization (they even farm!). But, there are important contrasts between the wars of chimpanzees and social insects, and those of humans. Chimpanzee wars are of small scale, on the level of the band, and always opportunistic. That is, they occur in a manner which could be modeled as competing firms acting in their own rational interests. When two bands interact, and one of them is much larger, then the larger band proceeds to attack the smaller. Chimpanzees do not engage in conflict by and large when there is parity between two bands. The attackers take on little risk, to the point where there hasn’t been a documented instance of casualty on the part of attacking bands in field observation. Social insects are very different. The scale of their warfare is on the same order of that of humans, millions of ants for example may be party to conflict. But, unlike humans the coefficient of relatedness of the opposing coalitions are such that it can be explained via traditional inclusive fitness theory.
The figure above, from the paper, illustrates how humans are different. Humans engage in high intensity conflict despite most of the genetic variation being partitioned within the groups. In other words, unrelated individuals (almost always, but not exclusively, men) are fighting and dying for each other. In the examples above over ~20% of the genetic variation across Argentine ants is partitioned across the groups. You may know that across geographical populations a similar order of magnitude of human variation is partitioned across groups; for example, ~10% of the variation between Chinese and English populations is between the two groups. But for the ants we’re talking about adjacent groups, not those geographically distinct. As the figure above makes clear in the vast majority of cases where conflict might arise because of competition over resources or simple opportunity the genetic distance between the groups is very small. That is because a small amount of gene flow can quickly equilibrate differences between populations (1 migrant per generation across groups is sufficient to prevent genetic divergence if they separate). ~7,000 years ago the amount of variation partitioned in Europe between Mesolithic hunter-gatherers and the first farmers in parts of Central Europe which were co-resident was ~10%. Today, the average difference between national groups (e.g., Czech vs. Portuguese) is on the order of ~1%. Gene flow quickly removes variation, which is important, because without variation natural selection can not operate. Heritable variation is its raw material. Evolutionary pressures can maintain intergroup conflict on a massive scale among social insects because of the high degree of relatedness within colonies and supercolonies. In other words, they’re superorganisms. Similarly, moderately social behavior among chimpanzees manifests in a manner such that elegant individual level evolutionary dynamics such as inclusive fitness and reciprocal altruism are sufficient to explain interactions, in addition to the close relatedness of the males because of patrilocality. Larger groups of chimpanzees attack smaller groups of chimpanzees because it is the rational behavior in the context of minimal risk to self. Higher level group explanations are not really necessary.
The answer to the question for why humans often, but not always, engage in warfare from Zefferman and Mathew is “group-structured cultural selection.” Often this phenomenon goes by labels such as “cultural group selection” or “multi-level selection,” but the authors assert that these terms are somewhat fraught, so they seem to be presenting their own so that there is a precise and distinctive understanding of what they are getting at without old baggage. As you can see above the Fst between cultural groupings is far higher than that for genetics. Why? Intuition can lead us to the answer easy enough: genetics is a straightjacket in terms of the nature of inheritance, while cultural is more flexible. If, for example, one group defeats another in war and kills all the males and older females, but takes the younger females as slaves, then the genes of the defeated may persist. But often the culture will go totally extinct. Conversely, a group victorious in war may increase demographically through amalgamation, while preserving to a great extent its cultural distinctiveness and identity. As an example, the Zulu were two centuries ago simply one clan among the Nguni. Today, they are one of the major tribes of South Africa thanks to the victories of Shaka and his successors. The recent genetic results coming out of Britain, which suggest that Anglo-Saxons had an impact, but a secondary one genetically, illustrates how a demographic minority can drive a cultural rupture among a conquered populace. The language of the people of Devon, which was once Dumnonia, is a sibling to that of the Germans across the North Sea, with no relationship to the Brythonic Celtic of their ancestors (I choose Devon because this is a region of the British Isles with very little Anglo-Saxon genetic footprint; Dumonia was even conquered after the Anglo-Saxons had become Christian, and so postdates the sub-Roman era).
If war as a behavioral pattern is selected on the level of a culture, a group, what does that imply for its innateness? Probably that there is not a “war instinct,” and, that war may not be primal as an ancestral character for our lineage. Rather, war is a social phenomenon which emerges out of the constellation of other cognitive traits which we have as part of our ancestral heritage. In that way, it may be like organized religion or representational sculpture, aspects of evoked culture. Given particular conditions war may bubble up out of the possibilities for human behavior “naturally,” and then be selected upon as Zefferman and Mathews imply (its benefits are made explicit in works such as Ian Morris’ War! What is It Good For?). The reason that there are hundreds of millions of people who are the cultural descendants of Romans and their Latin allies, rather than Etruscans, is that the latter were defeated in war and absorbed by the Romans. Clearly there was a benefit to the Romans as a culture for developing social institutions which made them incredibly effective as a nation at arms. Some of the glory of victory was likely demographic, insofar s Roman colonies spread far and wide, but most of it was in terms of posterity and memory. Though the French may conceptualize themselves today as the descendants of Gauls (genetically this is probably correct), their language and religion come down to them because of the Romans.*
If memes, rather than genes, are the targets of selection, and groups are the units, how is it then that the genes for males who engage in highly risky behaviors persist? Shouldn’t cheaters have a higher fitness? Some of this is likely explained by the benefits to the group. Even though risk is entailed in war, the fitness benefits can be quite great for successful males. But there’s something else going on too. Zefferman and Mathew refer to the tendency toward conformity. This is an innate psychological bias which humans exhibit, and it allows for rapid change in cultural norms and expectations. Twenty years ago Bill Clinton did not hesitate to sign the Defense of Marriage Act in the interests of his political ambition, while today he wouldn’t hesitate to term someone opposed to gay marriage/marriage equality a “bigot.” In Daniel Schacter’s The Seven Sins of Memory he recounts how white American Southerners who came of age from the 1960s to 1980s often remember themselves as being ahead of the times when it came to segregation and race relations, a recollection belied by longitudinal studies. Psychological conformity as an individual level trait allows for rapid homogenization of cultural norms very fast and across wide swaths of the population. in relation to warfare it can explain why the Japanese can shift their national consciousness from that of being very militaristic on the whole to being opposed to the existence of a standing army which engages in force projection. Culture is protean because of a fixed aspect of human nature, social intelligence which fosters group conformity. Those who lack the tendency toward conformity exist, but they are often ostracized, and are considered disagreeable. Though cheating may be beneficial, it may be that the same psychological trait which allows for conformity and agreeableness means that few will wish to betray their “little platoon” for self-interest. War’s raw material exists because it is highly beneficial at a lower level of organization in terms of human social interaction. Humans sacrifice as a consequence of being human as we understand it, embedded in a network of trust, friendship, and affinity.
Human psychology also means that John Horgan’s hope in The End of War that this phenomenon may have a “sell by” date is not futile. Perhaps the best analogy here is to slavery, an institution which arose during the Neolithic in a mass form which many thinkers across time took as a given, whether it was for the good or bad (Aristotle believed there were natural slaves, while other thinkers accepted it as an inevitable evil). And yet a cultural shift, enabled likely by economic and social forces, occurred over the past few centuries, and de jureslavery is now all but abolished, and de facto chattel slavery exists only furtively in the hidden places beyond the reach of modern human norms. Like war the institution of slavery was universal among complex societies, though its magnitude and manifestation varied in detail. With the emergence of radical stratification, with some individuals deified, it stands to reason that the lowest form of abasement within a society would be to utterly dehumanize. Mind you, the anthropological evidence seems clear that dehumanization is common among humans, even in the primal condition, but this was of the Other, those outside of the tribe. Though slavery often had a tribal connotation (e.g., blacks or non-Muslims could be enslaved), the key difference between it and dehumanization in the generality is that slaves became human tools integrated into the body of society, part and parcel of the fabric of human cultures. The abolition of slavery was the revenge of the dignity of the individual, as the circle of human empathy was expanded outward and totality. Hierarchy and inequality persist, but they are dampened by novel cultural institutions and ancient intuitions.**
I do not know how war will end exactly, or, if it will end. History is not filled with many inevitable end points. But, the evidence in Steven Pinker’s The Better Angels of Our Nature is moderately compelling to me. Humans do have a nature, and it explains why war emerges so often in our history. But that nature also holds the keys for why it may diminish into irrelevance. If psychological conformity can be aligned with non-zero sum interactions then we may see an opportunity to avoid the conflicts that arise because of disputes over finite resources and opportunities.
* The French self-conception has gone through several iterations, some of which are much more philo-Roman than others.
** Do unto others as you would have done unto to you is simply a formalization of an iterated game.
Citation: Zefferman, Matthew R., and Sarah Mathew. “An evolutionary theory of large‐scale human warfare: Group‐structured cultural selection.” Evolutionary Anthropology: Issues, News, and Reviews 24.2 (2015): 50-61.