War Before Civilization

by Lawrence H Keeley

  Thus pacifistic societies seem to have existed at every level of social organization, but they are extremely rare and seem to require special circumstances. The examples of Sweden and the Semai demonstrate that societies can change from pacifistic to warlike, or vice versa, within a few generations or, (as with the Semai) within the lifetime of an individual. As these examples and the case of the Polar Eskimo establish, the idea that violent conflicts between groups is an inevitable consequence of being human or of social life itself is simply wrong. Still, the overwhelming majority of known societies have made war. Therefore, while it is not inevitable, war is universally common and usual.

  THE FREQUENCY OF WARFARE IN STATE AND NONSTATE SOCIETIES

  How frequent are primitive wars, and do nonstate societies engage in warfare less frequently than states or civilized societies? These questions are related to the question of how intense primitive warfare is. Again turning to cross-cultural research, we find That many of the myths about primitive war are untrue.

  The three cross-cultural surveys mentioned earlier also include data on the frequency of warfare. All these studies show that warfare has been extremely frequent among primitive societies.19 In the sample of fifty societies, 66 percent of the nonstates were continuously (meaning every year) at war, whereas only 40 percent of the states were at war this frequently. In this survey, warfare was therefore found to be less frequent in state societies. The larger sample of ninety societies, however, indicated that the frequency of war increased somewhat with greater political complexity; 77 percent of the states were at war once a year, whereas 62 percent of tribes and chiefdoms were this war prone. Nevertheless, 70 to 90 percent of bands, tribes, and chiefdoms went to war at least once every five years, as did 86 percent of the states. All these figures support yet another survey, which found that about 75 percent of all prestate societies went to war “at least once every two years before they were pacified or incorporated by more dominant societies” and that warfare was no more frequent “in complex societies than in simple band or tribal societies.” In the sample of U.S. western Indian tribes, which consisted wholly of nonstate societies, 86 percent were raiding or resisting raids undertaken more than once each year. And such high frequencies of fighting were not peculiar to North America.20 For example, during a five-and-a-half-month period, the Dugum Dani tribesmen of New Guinea were observed to participate in seven full battles and nine raids. One Yanomamo village in South America was raided twenty-five times over a fifteen-month period. These independent surveys show that the great majority of non-state societies were at war at least once every few years and many times each generation. Obviously, frequent, even continuous, warfare is as characteristic of tribal societies as of states.

  The high frequencies of prestate warfare contrast with those of even the most aggressive ancient and modern civilized states. The early Roman Republic (510–121 B.C.) initiated a war or was attacked only about once every twenty years. During the late Republic and early Empire (118 B.C.-A.D. 211), wars started about once every six or seven years, most being civil wars and provincial revolts.21 Only a few of these later Roman wars involved any general mobilization of resources, and all were fought by the state’s small (relative to the size of the population), long-service, professional forces supported by normal taxation, localized food levies, and plunder. In other words, most inhabitants of the Roman Empire were rarely directly involved in warfare and most experienced the Pax Romana unmolested over many generations.

  Historic data on the period from 1800 to 1945 suggest that the average modern nation-state goes to war approximately once in a generation.22 Taking into account the duration of these wars, the average modern nation-state was at war only about one year in every five during the nineteenth and early twentieth centuries. Even the most bellicose, such as Great Britain, Spain, and Russia, were never at war every year or continuously (although nineteenth-century Britain comes close). Compare these with the figures from the ethnographic samples of nonstate societies discussed earlier: 65 percent at war continuously; 77 percent at war once every five years and 55 percent at war every year; 87 percent fighting more than once a year; 75 percent at war once every two years. The primitive world was certainly not more peaceful than the modern one. The only reasonable conclusion is that wars are actually more frequent in nonstate societies than they are in state societies—especially modern nations.

  MOBILIZATION

  The informal and voluntary mobilization for war supposedly characteristic of tribal societies is often cited as evidence of the lack of importance and effectiveness of primitive versus civilized war. The idea is that if war really represented an important activity, instead of just a sport or dangerous pastime, these primitive societies would muster all of their strength.

  Figure 2.1 shows some selected information on the size of war parties or armies in relation to the male populations of the social units from which they were drawn. While in most nonstate societies every male over the age of thirteen or fourteen is a potential warrior, not all of them participate in any particular war, battle, or raid. In general, tribal military formations are “all-volunteer” and usually muster proportions of their potential manpower similar to those achieved in the volunteer armed forces of states. Although modern conscript armies during active warfare generally represent a high percentage of the male population, on many occasions nonstate societies mobilize a higher proportion of their manpower. In World War II, neither the Soviet Union nor the United States, despite the tremendous power of coercion enjoyed by modern states, managed during the whole war to mobilize any greater proportion of its manpower than have some tribes and chiefdoms.

  The reasons why mobilization cannot be complete are essentially the same for any society. Many males are too young, too old, too ill, or temperamentally unsuited to endure the stresses of combat. Because the sexual division of labor in most societies trains men and women to be proficient in different tasks, a society’s economy may not be sustained if it is denuded of men to hunt, tend stock, clear gardens, or do whatever other essential work lies on the male side of the division. Although women may be able to take over some of these tasks, training and developing skill at them take time. It may be very unwise to focus all of a group’s manpower at one point on a border or beyond it, leaving women, children, and property vulnerable to attack from another quarter.

  Figure 2.1 Percentages of male populations mobilized for combat by various tribes, ancient states, and modern nations (see Appendix, Table 2.1).

  Women have very rarely engaged in combat, but have often played auxiliary roles in mobilization and logistics.23 Before hostilities commenced, they might shame cowards, taunt the hesitant, and participate in dances of incitement. Among some groups, women have accompanied war parties to carry weapons and food. During combat, they might serve as a cheering section, supply first aid, or collect spent enemy missiles to resupply their own warriors. In some cases, either by choice or by necessity (such as when the enemy breached their fortifications), some women might actually fight. For example, female warriors were apparently not unusual in northern South America. In general, though, women’s role has been to maintain the home front, tend gardens and stock, and nurse the wounded. While war may be everyone’s business, it has usually been men’s work.

  In civilized war, ancient and modern, tremendous manpower (and woman-power) is required just to equip and supply military formations. The higher “warrior” proportions of modern wartime states in Figure 2.1 disguise the fact that only a fraction of the men mobilized actually engaged in combat.24 In Napoleon’s armies, at any given time, only about 58 to 77 percent of his soldiers were “effectives.” The rest were convalescing, in training, garrison troops, or members of support units. During World War II, only about 40 percent of American servicemen served in combat units. The others were involved in administration, logistical support, and training; and an even smaller percentage carried a rifle, sailed on a warship, or flew in a warplane. The “tooth to ta
il” ratio between combat and support troops was 1:14 for the U.S. Army in Vietnam and is now about 1:11. This diminution in the proportion of actual combatants in armies means That no modern state army can or does engage all of its mobilized manpower. These proportions reflect the huge geographic scale of modern military operations and the heavy, complex technology involved. Of course, every person mobilized is lost to the home economy and peaceful pursuits, but the fact remains that very few of them actually engage in combat. By contrast, in ancient armies and primitive war parties, almost every participant was an effective. If mobilization figures are modified to reflect the higher proportion of noncombatants in modern armed forces, the mobilization for combat of tribal societies would compare even more favorably with those of modern states. This finding also implies that males in nonstate societies are far more likely to face combat than is the average male citizen of a modern nation. By the measure of manpower mobilization, then, war is no less important to tribes than to nations.

  PREHISTORIC WARFARE

  With regard to prehistory, nothing comparable to the surveys of historical and ethnographic societies cited earlier exists as yet. Any attempts to survey 2 million years of human prehistory for evidence of violence and armed conflict face several daunting difficulties. The first is that most regions of the world are poorly known archaeologically—the rare exceptions being Europe (especially the west), the Near East, and parts of the United States. The most unequivocal evidence of armed conflict consists of human skeletons with weapon traumas (especially, embedded bone or stone projectile points) and fortifications. However, humans have buried their dead for only the past 150,000 years or so; before this, the human remains that have been found were often disturbed and fragmented by scavengers and natural forces. Even during the past 150,000 years, many prehistoric peoples disposed of their dead in ways—for example, cremation and exposure—that left no remains for anthropologists to study. Only among some peoples—those for whom the use of stone- and bone-tipped weapons (which can survive embedded in or closely associated with human skeletons) was commonplace—is it easy to distinguish accidental traumas from those inflicted by humans. The use of these weapons occurred only during the past 40,000 years, and in many regions perishable wooden and bamboo spears and projectiles continued to be used until modern times. Until humans began living in permanent villages, fortifications would have not repaid the labor required to construct them (Chapter 3). But humans seem to have become sufficiently sedentary only during the past 14,000 years, and permanent villages are common in most regions only after the adoption of farming (8000 B.C. at the earliest). Thus it is possible to document prehistoric warfare reliably only within the past 20,000 to 30,000 years and in a only a few areas of the world. Granting these limitations, what does the archaeological evidence say about the peacefulness of prehistoric peoples?

  Some authors have claimed that the evidence of homicide is as old as humanity—or at least as old as the genus Homo (that is, over 1 million years).25 But many of the traumas found on early hominid skeletons have been proved by subsequent investigation to have had nonhomicidal causes or cannot be distinguished from accidental traumas of a similar character.26 For instance, the paired “spear wounds” found on some South African Australopithicine skulls are now recognized as punctures created by leopard canines as the predator carried these luckless ancestors of ours, gripping their heads in its teeth. As another example, Neanderthals seem to have been especially accident prone, compared with the modern humans who followed them. Neanderthals’ bones evidence many injuries and breakages (one study determined that 40 percent of them had suffered head injuries). Which, if any, of these injuries were caused by human violence cannot be determined. Since the heavy musculature and robust bones of Neanderthals imply that their way of life was much more strenuous and physically demanding than that of more recent humans, it seems probable that most of the traumas in question were accidental. Why they so often “forgot to duck” remains a mystery, however.

  Whenever modern humans appear on the scene, definitive evidence of homicidal violence becomes more common, given a sufficient sample of burials.27 Several of the rare burials of earliest modern humans in central and western Europe, dating from 34,000 to 24,000 years ago, show evidence of violent death. At Grimaldi in Italy, a projectile point was embedded in the spinal column of a child’s skeleton dating to the Aurignacian (the culture of the earliest modern humans in Europe, ca. 36,000 to 27,000 years ago). One Aurignacian skull from southern France may have been scalped; it has cut-marks on its frontal (forehead). Evidence from the celebrated Upper Palaeolithic cemeteries of Czechoslovakia, dating between 35,000 and 24,000 years ago, implies—either by direct evidence of weapons traumas, especially cranial fractures on adult males, or by the improbability of alternative explanations for mass burials of men, women, and children—that violent conflicts and deaths were common. In the Nile Valley of Egypt, the earliest evidence of death by homicide is a male burial, dated to about 20,000 years ago, with stone projectile points in the skeleton’s abdominal region and another point embedded in its upper arm (a wound that had partially healed before his death). The one earlier human skeleton found in Egypt bears no evidence of violence, but the next more recent human remains there are rife with evidence of homicide.

  The human skeletons found in a Late Palaeolithic cemetery at Gebel Sahaba in Egyptian Nubia, dating about 12,000 to 14,000 years ago, show that warfare there was very common and particularly brutal.28 Over 40 percent of the fifty-nine men, women, and children buried in this cemetery had stone projectile points intimately associated with or embedded in their skeletons. Several adults had multiple wounds (as many as twenty), and the wounds found on children were all in the head or neck—that is, execution shots. The excavator, Fred Wendorf, estimates that more than half the people buried there had died violently. He also notes that homicidal violence at Gebel Sahaba was not a once-in-a-lifetime event, since many of the adults showed healed parry fractures of their forearm bones—a common trauma on victims of violence—and because the cemetery had obviously been used over several generations. The Gebel Sahaba burials offer graphic testimony that prehistoric hunter-gatherers could be as ruthlessly violent as any of their more recent counterparts and that prehistoric warfare continued for long periods of time.

  In western Europe (and more poorly known North Africa), ample evidence of violent death has been found among the remains of the final hunter-gatherers of the Mesolithic period (ca. 10,000 to 5,000 years ago).29 One of the most gruesome instances is provided by Ofnet Cave in Germany, where two caches of “trophy” skulls were found, arranged “like eggs in a basket,” comprising the disembodied heads of thirty-four men, women, and children, most with multiple holes knocked through their skulls by stone axes. Indeed, some archaeologists, impressed by the abundant evidence of homicide in the European Mesolithic, date the beginnings of “real” war to this period.

  Indications of conflict, as reflected by violent death and the earliest fortifications, become especially pervasive in western Europe during the ensuing Neolithic period (the era of the first farmers, ca. 7,000 to 4,000 years ago, depending on the region).30 Some archaeologists have argued that real warfare begins only when hunters become farmers. This mistaken point of view does have some especially grim support in the remains of Neolithic mass killings at Talheim in Germany (ca. 5000 B.C.) and Roaix in southeastern France (ca. 2000 B.C.). At Talheim, the bodies of eighteen adults and sixteen children had been thrown into a large pit; the intact skulls show that the victims had been killed by blows from at least six different axes.31 More than 100 persons of all ages and both sexes, often with arrowpoints embedded in their bones, received a hasty and simultaneous burial at Roaix. The villages of the first farmers in many regions of western Europe were fortified with ditches and palisades. Several of these early enclosures in Britain, after being extensively excavated, yielded clear evidence of having been attacked, stormed, and burned by bow-wielding enemies. The early
agricultural tribes and petty chiefdoms of Neolithic Europe were anything but peaceful.

  Interestingly, the historically blood-soaked Near East has yielded little evidence of violent conflict during the Early Neolithic.32 Although extensive and elaborate fortifications were erected during this period at Jericho, they became common in the Near East only in the later Neolithic and in the Bronze Age.

  When we turn to the United States—specifically to those areas that have been subject to intensive archaeological scrutiny and where large samples of human burials have been excavated, such as the Southwest, California, the Pacific Northwest Coast, and the Mississippi drainage—violent deaths are at least in evidence and, in some periods, were extremely common.33 Fortifications were constructed at various times and in various regions by prehistoric farmers in the Mississippi drainage and in the Southwest, as well as by the prehistoric sedentary hunter-gatherers of the Northwest Coast.34 As with the best-studied regions of the prehistoric Old World, the prehistoric New World was also a place where the dogs of war were seldom on a leash.

  In each of these regions, the indications are that warfare was relatively rare during some periods; nothing suggests, however, that prehistoric nonstate societies were significantly and universally more peaceful than those described ethnographically. The archaeological evidence indicates instead that homicide has been practiced since the appearance of modern humankind and that warfare is documented in the archaeological record of the past 10,000 years in every well-studied region. In the chapters that follow, it will become clear that archaeological evidence strongly supports ethnographic accounts concerning the conduct, consequences, and causes of prestate warfare.