Victorious traditional societies are observed to obtain many benefits. Listing some of the major benefits alphabetically without any pretense of ranking them in importance, they include children captured, cows, food, heads (for head-hunters), horses, human bodies to eat (for cannibals), land, land resources (such as fishing areas, fruit orchards, gardens, salt pools, and stone quarries), pigs, prestige, protein, slaves, trade rights, and wives.
But the motives that people give for going to war, just like the motives that they give for any other important decision, may not coincide with the observed payoffs. In this as in other areas of life, people may be unconscious of or not frank about what is driving them. What do people allege as their motives for war?
The commonest answer is “revenge” for killings of fellow tribespeople or band members, because most tribal battles are preceded by other battles rather than by a long period of peace. Examples from the Dani War of Chapter 3 are the craving for vengeance by the Wilihiman after the battles or deaths of January, April 10 and 27, June 10, July 5, and August 16, 1961, and by the Widaia after April 3 and 10 and May 29.
If revenge is the main motive cited for continuing a war, what motives initiate a war? In the New Guinea Highlands, common answers are “women” and “pigs.” For men from New Guinea as from other parts of the world, women give rise to escalating disputes by being involved in or victims of adultery, desertion of husbands, kidnapping, rape, and bride-price disputes. The Yanomamo and many other peoples similarly report women as a or the major cause of war. When anthropologist Napoleon Chagnon had occasion to tell a Yanomamo headman about people of Chagnon’s “group” (i.e., Americans and British) “raiding” their enemies (i.e., Germans), the headman guessed, “You probably raided because of women theft, didn’t you?” That motive no longer applies to modern large-scale state societies. However, the origins of the Trojan War in the seduction of King Menelaus’s wife Helen by King Priam’s son Paris testify that women remained a casus belli at least until the times of small ancient states.
As for New Guineans ranking pigs on a par with women as causes of war, recall that pigs to a New Guinean are not mere food and the largest available source of protein: they are the main currency of wealth and prestige, and are convertible into women as essential components of bride-price. Like women, pigs are prone to wander and desert their “owners,” are easily kidnapped or stolen, and thus provoke endless disputes.
For peoples other than New Guineans, other domestic animal species, especially cows and horses, replace pigs as prized measures of wealth and causes of disputes. The Nuer are as obsessed with cows as New Guineans are with pigs, and the main goal of the Nuer in raiding the Dinka and other Nuer tribes is to steal cows. Nuer cows also lend themselves to disputes over trade and compensation (“You didn’t pay me the cows that you promised”). As one Nuer man (quoted by Evans-Pritchard) summarized it, “More people have died for the sake of a cow than for any other cause.” Horses and horse theft played the role of cows and pigs in triggering wars among Indians of North America’s Great Basin and among peoples of the Asian steppes. Many other types of material things besides women and animals have led to wars by being coveted, stolen, or subject to dispute among other peoples.
Small-scale societies go to war not only to acquire women as wives, but also to acquire other individuals for other purposes. The Nuer captured Dinka children to raise as Nuer and to incorporate into their own numbers. The long list of head-hunting peoples that went to war to capture and kill enemies for their heads included the Asmat and Marind in New Guinea, the Roviana people in the Solomon Islands, and various peoples of Asia, Indonesia, the Pacific islands, Ireland, Scotland, Africa, and South America. Cannibalistic peoples who ate captured or dead enemies included Caribs, some peoples of Africa and the Americas, some New Guineans, and many Pacific islanders. Capture of enemies to use them as slaves was practised by some complex chiefdoms and tribal societies such as northwest New Guineans, western Solomon Islanders, Native Americans of the Pacific Northwest and Florida, and West Africans. Slavery was practised on a large scale by many or perhaps most state societies, including ancient Greece, the Roman Empire, China, the Ottoman Empire, and European colonies in the New World.
There are at least two other frequently offered reasons that traditional people themselves mention as motives for war. One is sorcery: it’s routine in New Guinea and many other small-scale societies to blame anything bad that happens (such as an illness or a death that we would consider natural) on an enemy sorcerer, who must be identified and killed. The other is the common view that one’s neighbors are intrinsically bad, hostile, subhuman, and treacherous and thus deserve to be attacked whether or not they have committed some specific evil deed recently. I already quoted an example for New Guinea in Chapter 3: a Wilihiman Dani man’s answer to a Dani woman about why he was trying to kill the Widaia Dani (“Those people are our enemies, why shouldn’t we kill them?—they’re not human”).
In addition to all these conflicts over people and animals serving as motives for war, land conflicts are regularly mentioned as motives. A typical example is the land dispute that I described in Chapter 1, between my New Guinea mountain friends and the neighboring river people over the ridge-line between their villages.
Ultimate reasons
This enumeration of motives offered by members of small-scale societies for going to war—women, children, heads, and so on—doesn’t exhaust the list. However, it already suffices to make clear why these named motives by themselves aren’t a satisfying explanation for traditional warfare. Everybody’s neighbors have women, children, heads, and edible bodies, and many or most traditional neighbors have domestic animals, practise sorcery, and can be viewed as bad. Coveting of these individuals and things, or disputes about them, don’t inevitably trigger wars. Even in especially warlike societies, the usual response to a dispute arising is to try to settle it peacefully, e.g., by payment of compensation (Chapter 2). Only if efforts at peaceful resolution fail does the offended party resort to war. Why, then, are compensation negotiations more likely to fail among some peoples than among others? Why are there such differences, when women and the other claimed motives for war are ubiquitous?
The ultimate factors behind a war aren’t necessarily the factors that the participants themselves understand or enunciate at the time. For instance, one theory of Yanomamo warfare debated by anthropologists postulates that its ultimate purpose is to acquire scarce protein by assuring abundant availability of game animals to be hunted. However, traditional Yanomamo don’t know what protein is, and they persist in citing women rather than availability of game animals as their motive for making war. Hence even if the protein theory of Yanomamo warfare were correct (which it probably isn’t), we would never learn about it from the Yanomamo themselves.
Unfortunately, understanding ultimate factors that you can’t ask people about is much more difficult than understanding proximate motives that people can describe to you. Just reflect on our difficulties in establishing the ultimate cause(s) of World War I, despite the availability of enormous quantities of relevant documents to whose study hundreds of historians have devoted their lives. Everyone knows that the proximate cause of World War I was the assassination of Archduke Franz Ferdinand, heir to the throne of the Habsburg Empire, by the Serb nationalist Gavrilo Princip at Sarajevo on June 28, 1914. However, numerous other heads of state and heirs apparent have been assassinated without such dire consequences, so what were the ultimate reasons why this particular assassination did trigger World War I? The debated candidate theories of World War I’s ultimate cause(s) include the pre-war system of alliances, nationalism, threats to the stability of two major multi-ethnic empires (the Habsburg and Ottoman empires), festering territorial disputes over Alsace-Lorraine and transit through the Dardanelles, and Germany’s rising economic power. Because we still can’t agree about the ultimate causes even of World War I, we can’t expect it to be easy to understand the ultimate causes of
traditional warfare either. But students of traditional warfare enjoy a big advantage over students of the two world wars, in that we have an almost infinite number of traditional wars to compare.
The ultimate factor most often proposed for traditional warfare is to acquire land or other scarce resources such as fisheries, salt sources, stone quarries, or human labor. Except in harsh fluctuating environments whose conditions keep human populations periodically or permanently low, human groups grow in size to utilize their land and its resources, and can then increase further only at the expense of other groups. Hence societies go to war to seize land or resources belonging to other groups, or to defend their own land and resources that other groups seek to seize. This motive is often proclaimed explicitly by state governments going to war to acquire land and labor. For instance, Hitler wrote and spoke of Germany’s need for Lebensraum (living space to the east), but Russians and other Slavs lived to the east of Germany, so Hitler’s goal of acquiring eastern living space for Germany led him to invade Poland and then Russia in order to conquer, enslave, or kill the Slavs who lived there.
The most extensive test of this theory that land and resource shortages lead to war was by Carol and Melvin Ember, using a cross-cultural sample of 186 societies. From ethnographic information about these societies summarized in the Human Relations Area Files (a large cross-cultural survey), Ember and Ember extracted measures of several causes of resource shortages: the frequencies of famines, of natural disasters such as droughts or frosts, and of food scarcity. It turned out that these measures were the strongest predictor of war’s frequency. The authors took this finding to mean that people go to war to take resources (especially land) from their enemies, and thereby to protect themselves against unpredictable resource scarcity in the future.
Although this interpretation is plausible, it doesn’t operate so straightforwardly that all scholars accept it. While some traditional wars are indeed followed by the losers fleeing and the victors occupying their land, there are also cases of the vacated land being left unoccupied for some time. It is not the case that traditional wars are consistently fiercer in more densely populated areas, because some habitats and subsistence modes can comfortably support much higher population densities than can other habitats and subsistence modes. For instance, hunter-gatherers living at a density of 5 people per square mile in a desert feel more resource-starved and pressed to expand than do farmers living at 100 people per square mile in fertile, warm, well-watered farmlands. That is, what counts is not population density itself, but population density in relation to resource density, resulting in actual or potential resource shortages. If one compares traditional peoples with similar subsistence modes and living in similar habitats with similar resources, the frequency of warfare does increase with population density.
Other ultimate factors proposed to explain traditional warfare are social factors. People may go to war to keep troublesome neighbors at a distance, to get rid of the neighbors altogether, or to acquire a bellicose reputation and thereby to deter the likelihood of attacks by neighbors who wouldn’t hesitate to attack a group with a reputation for not defending itself. This social interpretation isn’t incompatible with the previous theory in terms of land and resources: an ultimate reason for wanting to keep one’s neighbors at a distance may be to maintain secure control of one’s land and resources. But it’s worth mentioning social considerations as a factor separate from resource considerations, because one’s desire to maintain distance from neighbors may cause one to take actions far more extreme than what others would consider necessary just for securing resources.
For example, until around 500 years ago Finland’s population was concentrated on the seacoast and Finland’s forested interior was sparsely inhabited. When individual families and small groups began moving as colonists into the interior, they tried to live as far as possible from each other. Finnish friends told me a story to illustrate how those colonists hated feeling crowded. A man cleared for himself and his family a small farm by a river, pleased that there were no signs of any neighbors. But one day he was horrified to observe a cut log floating down the river. Someone else must be living somewhere upstream! Enraged, the man started walking upstream through unbroken forest to track down the trespasser. On his first day of walking he met no one; on the second day, again no one. At last, on the third day he came to a new clearing, where he found another colonist. He killed that colonist and marched three days back to his own clearing and family, relieved that he had once again secured his family’s privacy. While that story may be apocryphal, it illustrates the social factors that cause small-scale societies to have concerns even about distant “neighbors” far out of sight.
Still other ultimate factors proposed involve benefits, to the individual rather than to the social group, of being warlike. A bellicose individual or war leader is likely to be feared and to gain prestige for his war exploits. That can translate into his being able to win more wives and to rear more children. For instance, the anthropologist Napoleon Chagnon calculated, from Yanomamo genealogies that he gathered, that if one compares Yanomamo men who have or haven’t killed people, the killers have on the average over two and a half times more wives and over three times more children. Of course the killers are also more likely to die or to be killed at an earlier age than are non-killers, but during that shorter lifespan they win more prestige and social rewards and can thereby obtain more wives and rear extra children. Naturally, even if this correlation does apply to the Yanomamo, I’m not recommending it to all you readers, nor can it even be generalized to apply to all traditional societies. In some societies the shorter lifespan of warlike men is likely not to be compensated by an ability to attract more wives per decade of their shorter life. That is the case for Ecuador’s Waorani Indians, who are even more warlike than the Yanomamo. Nevertheless, more zealous Waorani warriors don’t have more wives than do milder men, and they have fewer rather than more children surviving to reproductive age.
Whom do people fight?
Having thus addressed the question why small-scale societies fight, let’s now ask: whom do they fight? For instance, are tribes more likely to go to war against tribes speaking a different language than against speakers of their own language? Do they fight, or do they instead avoid fighting with, tribes with which they trade or intermarry?
We can place the answers in a more familiar context by first asking the same questions about modern nations going to war. A distinguished British meteorologist named Lewis Richardson, whose official career focused on mathematically analyzing complex patterns of atmospheric winds, spent two years during World War I attached to a motor ambulance convoy transporting sick and wounded soldiers. Two of his wife’s three brothers were killed during that war. Possibly impelled by those experiences and by his own Quaker family background, Richardson developed a second career of mathematically studying the causes of wars, in the hope of drawing lessons about how to avoid wars. His method consisted of tabulating all wars that he could learn of between 1820 and 1949, recording their numbers of deaths, dividing his table into five sub-tables according to those numbers, and then testing questions about when and why different nations went to war.
During that period of 1820–1949 the number of wars in which a country was involved varied greatly among countries, from over 20 each for France and Britain down to 1 for Switzerland and 0 for Sweden. The main source of this variation was simply the number of nations with which a given country shared a common frontier: the more neighbors, the more wars averaged over the long run; the number of wars was approximately proportional to the number of adjoining states. Whether neighboring states spoke the same or different languages had little effect. The sole exceptions to this pattern were that there were fewer wars in which both sides spoke Chinese, and more wars in which both sides spoke Spanish, than expected statistically from the total world number of speakers of Chinese languages or Spanish. Richardson speculated about what cultural factors apparently make S
panish-speakers especially prone, and Chinese-speakers especially unlikely, to go to war. His speculations are intriguing, but I shall leave it to interested readers to read Richardson’s analysis for themselves, on pages 223–230 and 240–242 of his 1960 book, Statistics of Deadly Quarrels.
Richardson did not test statistically the effect of trading between countries on the probability of war. However, because war is disproportionately between neighboring countries, which are also disproportionately likely to be trade partners, one would expect trade relations and war to tend to be associated with each other. It does appear, at least from anecdotal impressions, that modern nations that are trade partners fight more often than those that are not. Presumably that’s partly because the apparent correlation of trade with fighting is really just because both trade and fighting are in turn correlated with propinquity; and partly, too, because trade often gives rise to disputes. Even for nations that aren’t neighbors, the biggest modern wars have pitted trade partners against each other. For instance, in World War II Japan’s two main targets of attack were its leading source of imported materials (the U.S.) and its leading export market for its goods (China). Similarly, Nazi Germany and Russia were trading right up until the eve of Germany’s invasion of Russia on June 22, 1941.
With that discussion of nations as background, let’s now consider the same questions for small-scale traditional societies. We don’t have available for analysis a tabulation of all recent traditional wars, corresponding to Richardson’s table of modern state wars. Instead, we’ll have to content ourselves with anecdotes. These suggest that small-scale societies, even more than nations, fight their neighbors, because they lack the capacity for long-distance transport that enabled Britain to send troops halfway around the world in the mid-1800s to fight New Zealand’s Maori. There is little evidence of small-scale societies differentiating between neighbors speaking the same or different languages in matters of war. Most traditional wars were between neighbors speaking the same language, because neighbors are more likely to speak the same rather than different languages. Everyone involved in the Dani War of Chapter 3 spoke the Dani language. The long list of other societies that fought societies speaking the same language include the Enga, Fayu, Fore, Hinihon, Inuit, Mailu, Nuer, and Yanomamo; the list could be extended indefinitely. One partial exception, however, is that, while Nuer tribes fought other Nuer tribes as well as the Dinka, they fought the Dinka more often, and they observed restrictions in fighting the Nuer that they didn’t observe in fighting the Dinka. For instance, they didn’t kill Nuer women and children, they didn’t carry off Nuer as captives, and they didn’t burn Nuer huts; they limited themselves to killing Nuer men and stealing Nuer cattle.