Cannibals and Kings

by Marvin Harris

  A factor promoting the shift to or the intensification of banana and plantain production may have been the European pacification and extinction (possibly due to malaria and other European-introduced diseases) of the Arawak and Carib groups who previously dominated all the navigable rivers in this region. In aboriginal times, large gardens with fruiting trees would have constituted an inviting target for these more populous and better-organized groups. An important point to keep in mind is that Yanomamo wars are being fought mainly between villages that have broken off from common parental settlements. The Yanomamo are expanding into territories formerly occupied by more powerful riverine peoples.

  I have suggested that in general the adoption of a new means of production—steel tools and bananas and plantain gardens in this case—leads to population growth, which through intensification leads to depletions and renewed pressure against resources at a higher level of population density. The average size of the villages studied by Chagnon has more than doubled—to 166 in the twelve groups reported on. Smole indicates that the typical village in the Parima highland core of Yanomamo territory has between 65 and 85 people and that “populations much over 100 are exceptionally large.” Other estimates place the average pre-contact villages in the 40-to-60 range.

  What resources have been depleted by permitting villages to grow to 166 people instead of the previous limit of 40 to 85? With the exception of the groups who live along major streams and who depend on narrow flood plains for their garden lands, Amazonian band and village peoples’ most vulnerable resources are not forests or soils—of which there are vast reserves—but game animals. Even without much hunting by human beings, tropical forests cannot support an abundance of animal life. As I have said, in pre-Columbian times large Amazonian villages were situated along the banks of the major rivers, which provided fish, aquatic mammals, and turtles. The Yanomamo have only recently occupied sites close to such rivers, and they still lack the technology for exploiting fish and other aquatic resources. But what about Chagnon’s statement that the areas between villages are “abounding in game”? In earlier observations, Chagnon gave the opposite impression:

  Game animals are not abundant and an area is rapidly hunted out, so that a group must constantly keep on the move.… I have gone on five-day hunting trips with the Yanomamo in areas that had not been hunted for decades, and had we not brought cultivated foods along, we would have been extremely hungry at the end of this time—we did not collect even enough meat to feed ourselves.

  Chagnon could easily have gained a false impression of superabundance if his later observation pertains to the “no man’s lands” between village territories. Such an impression is exactly what one would expect if these lands are functioning as animal sanctuaries where breeding stocks are preserved.

  I do not claim that there is an actual decline in the Yanomamo’s per capita protein ration as a result of the depletion of animal resources. By walking longer distances, capturing smaller animals, collecting insects and grubs, substituting plant protein for animal protein, stepping up the rate of female infanticide (slowing the rate of population growth as the village breakup point is neared), people can avoid actual clinical symptoms of protein deficiency. Daniel Gross of Hunter College has pointed out that such symptoms have seldom if ever been reported for Amazonians who maintain their aboriginal way of life. The lack of such symptoms has led some observers to underestimate the causal significance of animal proteins in the evolution of band and village societies. Yet if Yanomamo warfare is part of a population-regulating system, the proper functioning of that system is to prevent populations from reaching densities at which adults become malnourished and sickly. Hence the lack of clinical symptoms cannot be taken as evidence against the existence of acute ecological and reproductive pressures. Gross has estimated that the animal protein intake per capita per day for tropical-forest village groups averages 35 grams. Although this is well above minimal nutritional needs, it is just about half of the 66 grams of animal protein consumed per capita per day in the United States. Americans would reach Gross’s average animal protein intake estimate by eating one large (5.5 ounce) hamburger once a day. For skilled hunters living in the midst of the world’s greatest jungle, this is not a very impressive comparison. How much meat do the Yanomamo get? William Smole has the only definite statement on the subject. While hunting is indispensable to the Yanomamo life style, and all are very fond of eating fresh meat, Smole reports:

  It is not unusual for days on end to pass during which no men from a shabono [village] are hunting or when little or no meat is being eaten.

  The fact is that under tropical-forest conditions an enormous amount of land is needed to ensure even the modest animal protein intake of 35 grams per capita per day. Moreover, the proportionate increase in area essential for maintaining this level of consumption is greater than any increase in the size of the village. Large villages cause proportionately greater disturbances than small ones since the daily level of activity in a large village has an adverse effect on the availability of game for miles around. As a village expands, its hunting parties have to travel increasing distances to find game in reasonable abundance. A critical point is soon reached when in order not to return empty-handed the hunters must stay out overnight and this is not something they like to do in a region of intense warfare. As a result, the villagers are forced either to accept cuts in their meat rations or to split up and disperse. They eventually choose the latter.

  How do the Yanomamo react to the pressure against protein resources and how do they translate it into the actual breakup of a village? Chagnon emphasizes the fact that village breakups are preceded by a crescendo of fights over women. From the account of Helena Valero, a Brazilian captured by the Yanomamo, we know that wives make a point of taunting their husbands when the supply of game falters—a practice common among many other tropical-forest groups. The men themselves, after returning empty-handed, become touchy about real or imagined insubordination on the part of their wives and younger brothers. At the same time the failure of the men emboldens wives and unmarried junior males to probe the weaknesses of husbands, seniors, and headmen. Adultery and witchcraft increase in fact and fancy. Factions solidify and tensions mount.

  The breakup of a Yanomamo village cannot take place peacefully. Those who move away inevitably suffer great penalties since they are forced to transport heavy banana and plantain cuttings to new gardens, seek refuge among allies, and pay for food and protection with gifts of women, while waiting for new trees to mature. Many attacks by one village on another represent the prolongation of intra-village disputes. Raids between unrelated villages also increase as tensions mount within villages. As hunting expeditions range over greater distances in pursuit of dwindling game resources, incursions into buffer zones between villages and even into enemy gardens become more frequent. Tensions over women lead to more frequent raiding for women, as an alternative to adultery and as validation of masculinity and threatened headman statuses.

  I will not attempt to describe in detail all the mechanisms that serve to announce and transmit the threat of depletion of animal resources and that mobilize the compensatory behavior of village breakups and dispersion. But I believe I have provided enough evidence to show that the case of the Yanomamo strengthens the theory that band and village warfare is part of a system for dispersing populations and slowing their rate of growth.

  6

  The Origin of Male Supremacy and of the Oedipus Complex

  The practice of warfare is responsible for a widespread complex of male supremacist institutions among band and village societies. The existence of this complex is a source of embarrassment and confusion to advocates of women’s rights. Many women fear that if male supremacy has been in existence for so long, then perhaps it really is “natural” for men to dominate women. But this fear is groundless. Male supremacist institutions rose as a by-product of warfare, of the male monopoly over weapons, and of the use of sex for the nurturanc
e of aggressive male personalities. And warfare, as I have already shown, is not the expression of human nature, but a response to reproductive and ecological pressures. Therefore, male supremacy is no more natural than warfare.

  Unfortunately, feminists have tried to counter the view that male supremacy is natural by denying that it existed among the majority of band and village peoples. Among non-anthropologists this has led to a resurrection of mystical theories about a golden age of matriarchy when women reigned supreme over men. Anthropologists themselves have found nothing that justifies the exhumation of this nineteenth-century corpse. Instead they have tried to show that the extent and intensity of the male supremacist complex has been exaggerated. In more extreme instances feminists have recently insisted that the reported high incidence of male supremacist institutions is an illusion created by the sexist minds of the male observers who were responsible for writing most of the descriptions of band and village life.

  Those who believe that male supremacist institutions are no more common than female supremacist or sexually balanced institutional complexes display a lack of understanding of the bias that actually dominates and directs the professional careers of cultural anthropologists, be they male or female. This bias reflects an almost irresistible temptation to claim that one has done field work among a group whose customs are sufficiently removed from the ordinary to justify the effort and expense entailed in learning about them. (I well remember my own chagrin at having chosen to do field work among the Bathonga, a patrilineal group in southern Mozambique, when with a little more foresight I could have convinced the Ford Foundation to let me go to a more exotic and hence professionally more rewarding matrilineal culture slightly to the north.) Far from being inclined to overlook the existence of institutions that moderate male power and authority, most ethnographers can envisage nothing more rewarding than to be able to write journal articles about “uxorilocal postmarital residence” or a nice case of “matrilineal descent with polyandry.” With this in mind, I find it impossible to believe that the overwhelming statistical regularities indicative of virtually universal structural biases against women are nothing more than motes in the eyes of male field workers.

  There are 1,179 societies listed in George P. Murdochs Ethnographic Atlas. In three-fourths of these societies, when women get married they must ideally move to the home of either their husband or their husband’s paternal relatives, whereas in only one-tenth must grooms go to live in the home of either their bride or their bride’s maternal relatives. The reckoning of the descent of one’s children shows a like asymmetry. In the same 1,179 societies children are regarded as members of the father’s descent group (lineage or clan) five times more often than they are regarded as members of the mother’s descent group; that is, patrilineality is five times more common than matrilineality. And only in about one-third of the cultures where descent is in the maternal line do married children remain with the mother. In another third of such cultures married male children stop living with the mother and take up residence in her brother’s household. This pattern, called avunculocality (residence with the avunculus, the Latin word for “mother’s brother”), implies that it is the mother’s brother who controls the kin group’s children and property even though descent is in the female line. Remarkably, the opposite pattern is nonexistent, though its absence has not prevented anthropologists from using the word “amitalocality” to identify it. If amitalocality did exist, a married male in a society which had patrilineal descent would be obliged to accompany his wife to the residence of her father’s sister. This would imply that despite the reckoning of descent in the male line it was the father’s sister who controlled the kin group’s children and property.

  Marriage types also attest to the dominance of males in domestic affairs. Polygyny (one husband, several wives) occurs over 100 times more often than polyandry (one wife, several husbands) and is the marriage form functionally best-suited for using sex and women as rewards for aggressive “masculine” behavior. Polyandry, on the other hand, is the form that would be best-suited for a society dominated by women and in which servile husbands were the rewards for fierce, competitive womanhood. Such societies would have little chance of success in warfare against enemies among whom robust, aggressive males were the military specialists. This suggests why so few band and village societies encourage women to collect husbands the way so many of them encourage men to collect wives.

  Another common marriage-related institution provides still more evidence of culturally induced male supremacy related to warfare and ultimately to ecological and reproductive pressures. At marriage a transfer of valuables from the groom’s family to that of the bride is extremely common. This transfer, known as “bride-price,” compensates the bride’s family for the loss of her valuable productive and reproductive services. A striking fact is that the logical opposite of bride-price—groom-price—virtually does not exist. (A single case, recently brought to my attention by Jill Nash, is that of the Nagovisi of Bougainville, where economic compensation is given by the bride’s sisters and mother to the groom’s sisters and mother for the loss of his valuable productive and reproductive services.) The term “groom-price” should not be confused with “dowry,” which is still another form of wealth exchange at marriage. Dowry occurs in patrilineal societies and is given by the bride’s father and brother to the groom or his father. Yet it is not considered compensation for the loss of the groom’s productive and reproductive services. Rather, it is intended to help cover the cost of maintaining an economically burdensome woman or as payment for the establishment of political, economic, caste, or ethnic alliances valuable to the bride’s father and brothers.

  These male-biased marriage relationships lie behind French anthropologist Claude Lévi-Strauss’s theory that marriage is the “gift” of women exchanged between men. “Men exchange women; women never exchange men,” insists Lévi-Strauss. Lévi-Strauss, however, has never offered an explanation of why this should be so.

  Political institutions in band and village societies also tend to be dominated by males. Patrilineal societies always have village headmen rather than headwomen, and religious leadership in most band and village societies is also male-centered; there are some female shamans—those adept at dealing with supernatural forces—but they are almost always less numerous and less prominent than their male counterparts.

  Band and village societies consider women to be ritually unclean during menstruation. They regard menstrual blood as a pollutant. Yet they use semen in rituals aimed at improving the group’s health and well-being. Throughout the world, males menace women and children with bullroarers” (noisemakers whirled on a string), masks, and other paraphernalia whose nature is kept a secret from the women. Men’s clubhouses, in which these items are stored and from which women are excluded, are also part of the same complex. Women, on the other hand, seldom ritually menace men and I know of no village which has a clubhouse where women gather to protect themselves against pollution given off by their husbands.

  Finally, in almost all band and village societies, male dominance is evident in the division of labor. Women do the drudge work, such as weeding, seed grinding and pounding, fetching water and firewood, carrying infants and household possessions, and routine cooking.

  My argument is that all of these sexually asymmetric institutions originated as a by-product of warfare and the male monopoly over military weaponry. Warfare required the organization of communities around a resident core of fathers, brothers, and their sons. This led to the control over resources by paternal-fraternal interest groups and the exchange of sisters and daughters between such groups (patrilineality, patrilocality, and bride-price), to the allotment of women as a reward for male aggressiveness, and hence to polygyny. The assignment of drudge work to women and their ritual subordination and devaluation follows automatically from the need to reward males at the expense of females and to provide supernatural justifications for the whole male supremacist complex. />
  What has prevented others from seeing the causal connection between warfare and all of these male-biased institutions? The stumbling block has always been that some of the most warlike village societies seem to have either very weak male supremacist complexes or none at all. The Iroquois, for example, are well-known for their incessant warfare and their training of males to be immune to pain. They are also well-known for their merciless treatment of prisoners of war. Captives were forced to run a gauntlet, their fingernails were pulled out and their limbs were hacked off, and they were finally decapitated or roasted alive at the stake—after which their remains were consumed in cannibalistic feasts. Yet the Iroquois were matrilineal, matrilocal, paid no bride-price, were more or less monogamous, and had no elaborate religious complex for intimidating or isolating women. Many societies display a similar pattern of intense militarism combined with matrilineal rather than patrilineal descent and weak rather than strong male supremacist institutions. (Bear in mind, however, that matrilineal societies constitute less than 15 percent of all cases.)

  In fact, the association between matrilineal institutions and a ferocious form of militarism is much too regular to result by chance. If one weren’t already convinced that warfare was responsible for patrilineal-patrilocal complexes, a logical conclusion would be that it was also somehow responsible for matrilineal-matrilocal complexes. The solution to this quandary, of course, is that there are different types of warfare. Matrilineal village societies tend to practice a brand of warfare different from that practiced by patrilineal village societies such as the Yanomamo. William Divale was the first to show that matrilineal societies typically engage in “external warfare,” that is, penetration by large raiding parties deep into the territories of distant enemies who are linguistically and ethnologically distinct from the attackers. Warfare among patrilineal band and village groups like the Yanomamo, on the other hand, is called “internal warfare” because it involves attacks by small groups of raiders on nearby villages in which enemies speak the same language and probably share a fairly recent common ancestry—thus the term “internal warfare.”