Anthropology and War

“Anthropology And War” Conference
Sheraton University Center, Durham, N.C.
28 January 1994

About the Speakers. The five anthropologists who made presentations at the conference exhibited a wide range of research interests. Robert L. Carneiro, Curator at the American Museum of Natural History, has done extensive field work among Indian groups in South America. Currently an adjunct professor at Columbia University, he received his doctorate at the University of Michigan. Carol R. Ember has been a professor of anthropology at Hunter College, City University of New York since 1970. She has published widely in the field, often collaborating with her husband (see immediately following). She earned her doctorate at Harvard University. Melvin Ember, formerly chairman of the anthropology department at Hunter College, has held the post of president of the Human Relations Area Files since 1987. He has lectured and published widely on the subject of the conference. He received his doctorate from Yale University. Jonathan Haas, vice president of the Field Museum of Natural History in Chicago, has spent some twenty-five years in archeological field work, mainly in the southwestern United States. He earned two masters degrees and a doctorate from Columbia University. Rutgers University’s R. Brian Ferguson also earned his doctorate at Columbia. Much of his field work has been done in Puerto Rico. He has published and spoken widely in the field — as indeed have all of the Conference presenters.

Conference Proceedings. With some ninety persons in attendance, Jonathan Haas began the conference with a talk entitled, “The Origins of War and Ethnic Violence.” Haas first questioned the assumptions that war is an inherent part of the human condition and that ethnic groups naturally see each other with fear and hatred. Observing that humans existed for hundreds of thousands of years before the advent of written histories, Haas turned to archaeological evidence from the North American continent to isolate the factors that he believes originated war and ethnic violence. The archaeological record shows no evidence of warfare or organized violence in the first 10,000 years that people inhabited the North American continent. Haas used a case study of the Anasazi peoples of the American Southwest to argue that the beginnings of settled agriculture and gradual ethnic differentiation were preconditions for the development of conflict. Yet, he argued, these differences alone were not enough to lead to war. It took a combination of unfavorable factors including climatic change, population pressure, and environmental stress to produce a competition for resources that, by the thirteenth century A.D., produced warfare. Haas concluded that violence was the last resort of people faced with imminent extinction.


When members of the panel began their comments, Brian Ferguson expressed curiosity as to how evidence of Anasazi cannibalism fit into Haas’ conceptualization of the origins of war. Haas replied that cannibalism appeared in his area of investigation before the advent of warfare and suggested that its extremely violent nature could indicate a ritual, rather than a subsistence function. Haas added, in response to comments from the panel, that the victims most likely were prisoners; he suggested that ritual violence may be part of the process by which warfare transforms “the other” into non-human, a process he views as connected with tribalism and linguistic differentiation. Professor Ferguson noted that this kind of cultural difference does not necessarily bring on war, and Haas agreed, observing that the reasons for beginning a war were often quite different from those for continuing a conflict; once war appeared in the archaeological record, however, it never went away. At the same time, he maintained, the incidence of conflict did decline as environmental stresses lessened.


After the floor was opened to discussion by conference participants, Haas responded to questions about variables he might have found to account for internal conflict or terrorism. He observed that many motives for violence exist that are difficult or impossible to infer from archaeological evidence. A conference participant asked Haas for his views on conflict between pastoral and agrarian cultures as an origin of war, especially when only one group had the use of horses. Haas responded that we cannot automatically assume cultures to be in conflict in the absence of other factors, but noted the advantage gained by the people of the North American plains when they began to use horses after contact with Spaniards. Haas then replied to several questions on the significance of cannibalism, pointing out that scholars are re-evaluating archaeological evidence, but that it remains difficult to discern possible uses of cannibalism. He cited as possible evidence of cannibalism used as a political tool the destruction of a whole village, an act that seemed to have been outside the material standard for the area. In response to a query about specific linguistic examples of cultures that have words for themselves as “people,” and others as “not people,” the panel generally agreed that cross-cultural studies are needed to determine the true incidence of these linguistic patterns and their significance for conceptualization of “the other.”

R. Brian Ferguson began his lecture on “Local and Global Factors in Generating War” by suggesting that anthropologists can build explanations for war in places such as Africa and Eastern Europe by examining war as a specific form of social and cultural organization and by using cross-cultural analysis to show the significance of a wide variety of factors. Local conflict, in Ferguson’s opinion, is inextricably linked to national and global processes. In order to understand the process of war, supra-local influences must be recognized as crucial. The most important of these, he maintained, are the pressures brought to bear by expanding states, with European expansion being the best known example. Even seemingly unaffected societies can be subtly changed by expanding states, especially in terms of increased war and indigenous violence. Ferguson then turned to a case study of the Yanamami peoples of the Brazilian rain forest, whom, scholars have maintained, are some of the most isolated but “most warlike” peoples on earth. Questioning whether the Yanamami are normally warlike, he asserted that internal violence may simply be a reaction to external state pressure. A market in western goods, in particular, has transformed social relations among the Yanamami. Material goods become the primary basis of social action, he argued, even when the Yanamami cite personal or moral justifications for conflict. Ferguson then applied this theory to Africa, noting that the United Nations is now having to deal with armed groups not associated with states. Because state building in Africa was artificial, superpowers propped up authoritarian governments and the World Bank managed their economies. For those reasons, he concluded that social relations within African states are inextricably connected to external powers. Thus, Ferguson hypothesized, there may be no such thing as a purely local war. Because ethnicity becomes linked to current economic or political relations, the international system may be unconsciously creating conditions that tear it apart.

The panel discussion began with a question from Carol Ember about possible biases in anthropological sources, as many early observers of non-industrial societies were military or commercial personnel. Ferguson replied that many first contact reports do not seem to be reliable; anthropologists can construct village histories and correlate them with changes in the contact situation. Panelist Robert Carneiro, acknowledging his long-standing differences of opinion with Ferguson, stated that he saw war in South America as an integral part of state-building, not as a result of contact. Ferguson stressed in his response that he was not suggesting that indigenous peoples were “noble savages” before contact, but rather that he was advocating the inclusion of external factors in models constructed to explain war. When asked about anthropologist Chagnon’s biological explanation of Yanamami warfare, which holds that warriors have the most reproductive success, Ferguson questioned Chagnon’s classification of warriors, in that it includes tribal leaders. In his view, truly warlike individuals died in battle before they could have many children; thus, they had a very low success rate in passing on their genes.

As the discussion period opened, a member of the audience put forth the counter-example of the Roman Empire, which achieved peace by wiping out the enemies of its Gallic allies. Ferguson remarked that this example reflects fundamental differences between older world systems and the modern one. Panelist Melvin Ember added that his cross-cultural research found only forty percent of cases in which initial contact resulted in pacification. Of the remaining sixty percent, eighty-nine percent of those cultures engaged in war at least every ten years. Jonathan Haas then suggested that the incidence of pacification may be related to the impact of superpower interests in a given situation. Several questions followed on ritual warfare, including competitions, games, sports, and tests of champions. Professor Ferguson noted that military sports correlate with the presence of war and that early contact reports often commented on war as ritual; competition between champions, however, does not seem to have been employed. Ritual, he argued, is the institutionalization of self-interest in a conflict situation. Returning to the question of sources, Ferguson responded to the suggestion that confusion over time might be common in periods of conflict by noting that ethno-historians could avoid misunderstanding by constructing a core of genealogy and village movements, for which times can be established by determining who was born in which location. A conference participant then asked how cultural models for warfare became dominant in view of the supposition that, genetically speaking, warfare is unprofitable and risky. In the discussion that followed, conference participants stressed the lack of genetic evidence for aggression and questioned the use of methodologies that correlate evidence with explanation on the basis of human rationality and assumptions of functionality. Panel members pointed out that theories of material rationality for warfare are still under active dispute in the anthropological field; they expressed a consensus that the evidence falls on the side of cultural explanations for war.

Melvin R. Ember led off the afternoon session with a talk on “Causes and Consequences of War: A Cross-Cultural Study.” Using statistical cross-cultural analysis to test hypotheses, Ember and his colleagues determined that a statistically significant result for a particular time should be obtained if the variables correlate. The advantages of this type of study, he argued, are that conclusions drawn from such world-wide comparisons come closest to expressing a universal phenomenon. He differentiated his study from cross-national comparisons, noting that his data compare societies as defined primarily by linguistic differentiation. His major findings point to an ecological explanation for war, which he defined as armed combat between units, usually not involving specific forces or politically unified societies. His multiple regression analyses found that, where unpredictable natural disasters destroyed food supplies, societies faced with the threat of scarcity chronically engaged in conflict. The actual incidence of scarcity was not important and victors took resources even when they did not have an immediate resource problem. Further, socialization for aggression did not seem to be a predictor of war, but socialization for mistrust was highly predictive. Thus, fear predicts war, whether fear of others or of nature. Professor Ember hypothesized that unpredictable disaster may make people less trustful and noted that his findings could be applied to complex industrial societies if disaster is defined in terms of complex economic interdependencies.

Discussion centered at first on issues of fear and security, one conference participant hypothesizing that fear could explain the tendency of nation-states to fight for security. Panelist Carol Ember noted that national security is a constant concern, but its importance varies in proportion to the magnitude of insecurities at a given time. A conference attendee underscored the correlation between economic downturn and war and held that economic difficulties can be seen as a kind of “natural disaster.” One of the panelists then commented that socialization for mistrust seems more a consequence of warfare than a cause. Professor Carol Ember proposed an environmental explanation for this type of socialization, noting specifically that temperate regions, with their unpredictable weather patterns, saw a higher incidence of socialization for mistrust. When questioned about the correlation between military technology and a willingness to initiate conflict, she cited a cross-national study which indicates that military parity makes war more likely than a vast inequality of armed strength. When asked to what extent their correlation assumed rational decision making, she noted that people’s stated motivations and their actions were often quite different; for example, protagonists may claim to be fighting for revenge, but end up taking land.

Carol Ember then turned to her presentation on “Democracy and Peace: A Cross-Cultural Study.” She began by explaining that her study was designed as a test of results from cross-national analyses that found a lower incidence of warfare between democracies — states which, however, were found to engage in conflict with non-democracies at a usual frequency. Scholars have questioned the assumption that people within a democracy respect the individual rights of others if those people are self-governing and are less restrained when considering non-democracies; a cross-cultural study would test whether this vision of “the other” indeed justifies hostile policies on the part of democracies. Because the cultures in their sample were not unified beyond the community level, Ember and her colleagues had to construct an analog of democracy based on various factors. These included levels of participation and decision making, concentration of political power, including rates of adult participation, levels of local decision-making, and the frequency of community fission. Thus, their revised hypothesis was that political units with wider participation and more willingness to agree to disagree go to war with each other less often than with other polities. Excluding cases where coders disagreed, excluding pacified societies, and controlling for community size and physical conditions, the cross-cultural results were consistent with the cross-national data. Low levels of political participation predict internal war, with multiple regression analysis showing that the local level is more important than the actions of higher political authorities.

The majority of questions from the floor concerned methodology, including how Dr. Ember and her co-researchers defined democracy and what they considered indicators of procedural democracy. A conference participant cited the example of the American Civil War as a case in which there were limits to dissension that a system could stand, but in which fission did not take place. Ember observed that disagreement seems to produce war or gridlock. India and Ireland were suggested as anomalous situations where lack of commonality produced internal conflict.

Discussion then turned to distinctions between internal and external war and the researchers’ definition of society based on language. A panelist expressed the view that the two studies presented in this session fit together, and a member of the audience wondered if it was possible to correlate fear with the perception of another country as a democracy. When asked whether there was a correlation between female participation and war, the Embers replied that societies with higher overall frequencies of war were more likely to exclude women from political participation.

Robert L. Carneiro rounded out the afternoon with his talk, “From Autonomous Village to the State: the Role of Warfare.” He began by drawing a distinction between dispersive warfare and aggregative warfare, which he saw as connected with the beginnings of settled agriculture, especially in borderlands with a less-hospitable environment. Aggregative warfare, he argued, produced simple and compound “chiefdoms,” the latter of which began to resemble small states. Thus, in his view, aggregative warfare is the process that underlies all political evolution. As evidence for this statement, Carneiro turned to a case study of state development in Spain between 2500 B.C. and 1479 A.D. Tracing Iberian evolution from the Neolithic peoples, to the three tribes noted by the Roman invaders, through Greek colonization and pre-Moorish barbarian invasions, Carneiro linked war to political unification. After the seventh century A.D., the interplay between the Moors and Spain’s Christian defenders resulted in the final unification of Spain under Christian kings. He emphasized the non-rectilinear development of a unified kingdom, and linked the phenomenon to the unwillingness of autonomous political units to surrender their sovereignty without a fight. In conclusion, Carneiro surmised, in light of the development of the European Community and other international phenomena, that there may be new state-building mechanisms at work which have taken over the aggregative function of warfare.

The discussion period opened with the presentation of several examples of amalgamation without warfare and the observation that some polities put together in conflict now appear to be breaking apart in conflict. A conference participant pointed out that it is easy, from a historical perspective, to find examples for several different points of view on the importance of war in state formation and noted that the pace of change during a given time period may be a factor that Professor Carneiro’s thesis does not take into account. The exchange of views then turned to the question of aggregation in response to an outside threat in contrast with aggregation after submission to an external authority. A conference participant noted that some groups do not submit to authority and that various degrees of autonomy can exist within a formally unified area. Another conference participant pointed out that Yugoslavia and the former Soviet Union were often paired as examples of state-formation through violence, but the actual circumstances of their formation and disintegration were quite distinct. Finally, several panel members observed that recent examples of consolidation without conflict may suggest new aggregative mechanisms operating within the international system.

In conclusion, members of the audience asked the panelists to summarize their views on how cultural anthropology, including the analysis of symbolism, rhetoric, and ritual, can influence studies of war, peace, and the military. The panel generally agreed that cultural anthropology offers a valuable perspective in that it tries to explain the meaning of war instead of focusing solely on its practice. Noting that symbolic systems, particularly languages, correlate to perceptions of how war occurs, the panelists agreed that more cross-cultural studies were needed to clarify the role of symbolic systems in war. One panelist noted the tendency of anthropologists to rely on their own interpretations instead of asking people inside a culture for theirs. This comment elicited several different opinions from conference participants on the relative significance of the ethnographer’s bias, the role of self-reflection, and the importance of bringing several viewpoints to bear in a given study. As a final comment, the panelists concurred that anthropology, more readily than other disciplines, can offer a useful cross-cultural perspective in the study of war.