Social Evolution Forum
FIND sef:
Counterproductive Way of War

Over the last two years I’ve written a number of blogs exploring the role of warfare in cultural group selection (you can see these blogs collected under the heading Ways of War in Popular Blogs and Series).

Today I’d like to return to the question, What makes war productive as an engine of cultural evolution? It’s not how many corpses are piled up. The key factor is how conclusive war is. When one society decisively overruns another one, or a big chunk of it, it creates the conditions for cultural change, whether by more brutal means (e.g., genocide or forcible assimilation) or by relatively gentle means.

Before proceeding further, I’d like to remind my readers that I have no intention of excusing or even glorifying war – my goal is understanding its role in human history. See Joe Brewer’s article Linguistic Framing of “War! What is it Good For? and my own Of Course, War Is Evil!

Sign up for our newsletters

I wish to receive updates from:

One factor that immediately comes to mind is the importance of geography. Human groups that live in treeless plains find themselves in a particularly perilous situation when they lose in war. They can cede their territory to the victors and retreat elsewhere. However, as the landscape fills up, and it inevitably does, there are fewer and fewer places to retreat to. Eventually, losing wars means being subjugated by victors, or worse.

Dense forests and swamps and bogs can make plains more defensible and, thus, blunt the intensity of cultural group selection. But the most difficult terrain for conducting offensive operations is hills and mountains. Mountainous areas are challenging even for modern armies, equipped with airplanes and helicopters, means of instant communications, and artillery or missiles that can kill enemy from 100 km away. Before modern technology, rugged terrain favored the defenders even more. A small force can defend a narrow defile against a much larger force. Occupying a high ground, defenders could roll boulders on enemies. Shooting missiles downwards extends their range and killing power.

Once again, what is important is not how many people are killed, but what is the effect of war on cultural group selection. Warfare in mountainous terrain is often vicious and bloody. An altercation in which someone is killed typically triggers a spiral of revenge and counter-revenge that can decimate family groups or whole clans. However, given their ability to retreat to their mountain fastnesses when pressed, entire cultural groups are much more secure from being overrun and culturally extinguished, unlike groups inhabiting plains. For this reason, in rugged terrain warfare is ‘unproductive,’ or even ‘counterproductive’ as a force of social evolution.

As an example, take warfare among the Enga, a horticulturalist society inhabiting the highlands of New Guinea. The large island-continent of New Guinea is probably one of the most rugged places on Earth. The difficulty of traveling through it is so great, that for centuries people who lived on its shores thought that there was nothing in the interior but mountains piled on top of other mountains.

Only in the early twentieth century, when airplanes began overflying the interior and gold-seekers crossed the forbidding ranges driven by greed, the Europeans discovered that the interior was far from uninhabited. During the 1930s the first explorers found fertile, densely populated valleys, divided by saw-tooth mountains. These valleys were inhabited by a total of more than a million people, still living in the Stone Age.

Round sweet potato beds in gardens at high altitude on the south road from Mt. Hagen to Tambul, Papua New Guinea. Source

One of the best studied societies on the island is the Enga of the central New Guinea. The Enga women cultivate sweet potatoes and raise pigs. It would not be much of an exaggeration to say that the chief business of men is war.

Warfare among the Enga was studied by the Australian anthropologist Mervyn Meggitt, and more recently by the American anthropologist Polly Wiessner. Meggitt began his field research in central New Guinea in 1955 on one ethnic group among the Enga, Mae Enga, at the time when the colonial authorities largely succeeded in imposing peace on the central highlands (which, unfortunately, did not last). However, his primary interest was on the warfare during the period before the arrival of the Europeans. The Mae Enga at that time numbered around 30,000 and were divided into a number of ‘phratries’ or tribes. A typical tribe, in turn, was subdivided into 7-8 clans. These clans were independent political units of roughly 300–400 people occupying 2–5 squared kilometers of land. As we can see, we are dealing here with a very small-scale society, indeed.

Because clans were the ‘sovereign’ political units, most warfare was between clans. However, any particular conflict could also involve other clans (usually from the same tribe) who allied themselves to one or the other of the belligerents. Occasionally the Mae Enga organized Great Ceremonial Wars—“tournament fights fought recurrently between entire tribes or pairs of allied tribes to demonstrate strength and brew the grounds and spirit for the feasts and exchanges of enormous proportions that followed” (Wiessner). Clans also sometimes split as a result of lethal conflict. The predominant form of warfare, however, was not at the subclan or intertribal level, but between clans.

The intensity of warfare was very high. Meggitt estimated that 35 percent of men were killed in war or died of battle-related wounds afterwards. War was the main killer of men. An additional quarter died from illness or accident, and only 15 percent grew old enough to die naturally (these percentage do not add up to 100 because Meggitt was unable to determine the cause of death for 26 percent of victims).

Constant warfare imposed a landscape of fear and mutual suspicion. Most people stayed within their clan territories, a few square kilometers, most of the time. Meggitt writes:

In the past all movement outside one’s own clan territory was hazardous, and in general men made such excursions only in armed groups and for compelling reasons, in particular to attend distributions of wealth, to negotiate exchange transactions, to trade, and to assist friends and relatives in battle. Casual social visiting by men was not common, not only because it exposed the wayfarer to the dangers of ambush and murder en route, but also because it violated Mae notions of personal privacy and group security. A man who unexpectedly appeared at the house of even a close kinsman in another clan was viewed with some suspicion by the latter’s clansmen as a potential spy, who might carry home information—about the clan’s defenses (palisades, ditches, secret escape passages, and the like) or about the disposition of its pigs—that could be used in planning a night raid or theft.

Warfare between clans, especially those belonging to different tribes, was very different from the ritualized tournaments of Ceremonial Wars. According to Meggitt, its distinguishing features were

(a) the execution of surprise attacks or invasions with the aim of achieving a total rout that opens the enemy’s territory to occupation; (b) the deliberate maximizing of property destruction (cult structures, houses, ceremonial grounds, trees, crops, and pigs), in the hopes of demoralizing the enemy; (c) the readiness to ignore the restraints of kinship and affinity, both as they moderate the intensity of violence and as they encourage acceptance of mediation or conciliation; (d) the occasional refusal to recognize non-combatant status; (e) the mutilation of fallen enemies; (f) and, perhaps, the longer duration of these confrontations.

In short, interclan warfare has all the earmarks of total war. The consequences of war could be quite dire for the losers. Of the 34 wars in Meggitt’s database, for which he could determine the final outcome, six resulted in an eviction of the losing group, with the losers usually dispersing among other clans where they could find friends or relatives willing to take them in. In nineteen cases the victorious group was able to increase its land holdings, and in the final nine cases fighting ended in a stalemate, in which neither side gained territory.

Yet, and this might come as a surprise, the highly intense and lethal conflict among the Mae Enga must be classified as completely ‘unproductive’ warfare because it resulted in no cultural evolution whatsoever. Ceremonial Wars between tribes were highly regulated and did not result in an extinction of a whole tribe. The real Mae Enga warfare was between very small-scale social groups—clans of 300–400 people occupying a few square kilometers of land. Such groups could become extinguished, when one clan defeated another and took its territory. But both the winning clan and the losing clan had precisely the same culture—they spoke the same dialect of the Enga language, they used the same weapons in battle, they grew the same crops and raised pigs in the same way, and they followed the same rules of social behavior. In short, more than a third of men, and a few women were killed every generation, some clans disappeared, and victorious clans expanded territory, but overall there was no change in the frequencies of cultural traits.

The Mae Enga warfare, thus, provides us with a striking illustration of the importance of cultural variation. It doesn’t matter how intense the selection on clans was (and it was very strong)—they did not differ in their cultural characteristics. Thus, no evolution. Or very slow evolution, because new cultural variants are, once in a while, invented (for example, new forms of ritual), or diffuse from outside (sweet potatoes, which are now the staple, arrived in the Enga area roughly 350 years ago).


Join the discussion


  1. EdwardT says:

    So why so little cultural variation? Perhaps cooperation can be limited when individuals cooperate too much due to geographical factors.

    “Too much” cooperation is a relative state that means very rarely someone would seek to change the dominant norms that bind the group together. In the opposite way, “too little” cooperation would be when there are many people frequently seeking to change the dominant norms. Thus – apart from any selective effect of warfare – the ideal situation for social evolution would be when there is not too much and not too little within-group cooperation.

    What might make individuals cooperate more in some geographical circumstances than others?

    I notice you mention forests. I think forest cover density is important for the understanding of warfare and social evolution. In theory, perhaps dense forests create an environment of almost total cooperation and high evolutionary fidelity.

    Bogs and swamps are only found at low elevation. Forests are special because they are present on plains, in valleys and on high slopes.

    The primary effect of forests on pre-modern societies was to inhibit ranged warfare – because arrows won’t hit their targets when there are trees in the way. This would be particularly helpful to the team that is running away. The only way to decisively defeat an enemy on a forested landscape was with up-close, hand-to-hand fighting.

    So in order to win a battle in heavy forest more extreme bravery and a far larger degree of inter-personnel cooperation was necessary compared to merely hurling stones from a cliff or firing thousands of projectiles from atop a hill.

    Therefore forested landscapes should favour extremely cooperative groups of individuals willing to put their lives at risk for their group. With the stakes so high in hand-to-hand fighting nobody will want to be the one who messes things up for the others, damage the team spirit by threatening the ideological system – however limited it is compared to more complex ones – and so there will be little cultural genesis for a selective effect (e.g. of warfare) to work on.

    • Richard says:

      Actually, in heavily forested areas, I’d think you’d just see more ambushes and surprise attacks.

      Yes, forests inhibit projectiles to some degree, but they make sneaking up on and ambushing enemies (as well as disguising strength or weakness of numbers) easier. Eastern North America was one giant forest, and the woodland Native Americans launched plenty of wars against each other.

      For that matter, so was northern Europe back in the day, and the Germanic (and before them, the Celtic) tribes fought all numbers of wars.

      I also don’t get why forests would inhibit the selective effect of war. Two groups from 2 different cultures could still fight in a forest (especially if both are use to living in and fighting in forests).

      • EdwardT says:

        would the groups that live in dense forest get devastated by surprise attacks or is this a problem for the groups that don’t live in a forest, and simply want to move through the forest? I don’t suggest that forests inhibit wars, but I would expect that bow technology is not best adapted for the heavily forested landscape. groups that live in forests would more likely fight at much closer range. wrong or right, that is what I would predict! the selective effect is not inhibited as much as there is not much selective effect at all because there is little variation to work on among forest groups, which is what I thought Peter s,just aid,

  2. Richard says:

    BTW, even now (or until relatively recently) in parts of New Guinea, the custom when two strangers meet is to ask about each other’s geneaology (in order to try to determine some sort of kinship), so that they’d have an excuse not to kill each other.

  3. spandrell says:

    There’s cultural variation in the west, where the Papuans meet the very different Indonesians. Do the West Papuans have more asabiya?

    • Richard says:

      Highlanders and coastal inhabitants seem to be very separated by geography. Until recent times, few coastal inhabitants would want to (or could) venture in to the highlands.