Just as violence has long been taken to be a sign of the primitive, the savage or the uncivilized, or alternatively of the deviant, the individual and the unsocialized, so anthropology has long been concerned to show that violence obeys rules, is part of culture, and even fulfils certain social functions. Classic *functionalist accounts of institutions such as the feud (e.g. Gluckman 1956), stress that feuds bind people together, through the shared norms and expectations that participants invoke, even as they appear to divide them. But, despite this well-worn interpretive path, violence retains its capacity to unsettle and disturb.
Theoretically, violence lurks behind many important anthropological conceptions of the human and the social. Violence represents ‘natural’ drives which society must tame and repress if it is to survive: this broad idea can be found in Western political philosophy (classically in †Hobbes), as well as in Freudian *psychoanalysis, in †Durkheim’s notion of humans as ‘homo duplex’, in †Mauss’s implicit argument in his essay on The Gift that gifts are society’s means of overcoming the inevitability of war. From these perspectives emerges the linked notion of *society, or most often the *state, as the monopolist of ‘legitimate’ violence. The place of violence as a sign of the natural and unsocialized is even more marked in its prominence in *sociobiological arguments about human nature and genetics, such as those employed by Chagnon in the complex controversy about Yanomamö violence in lowland *South America (Chagnon 1988; Lizot 1994). Not surprisingly, such emphases have generated a counter-literature in which ethnographic examples are employed to suggest that peaceful sociability is the ‘natural’ condition (cf. Howell and Willis 1989).
Anthropology’s most useful contribution has probably been its documentation of the fact that violence is pre-eminently collective rather than individual, social rather than asocial or anti-social, usually culturally structured and always culturally interpreted. This was already implicit in functionalist interpretations of violence, but in recent years it has been greatly extended as anthropologists have reported the experience and interpretation of violence from the point of view of (among many others) paramilitaries in Northern Ireland (Feldman 1991), Indian riot victims (Das 1990), and torture survivors in Sri Lanka (Daniel 1994). Here the anthropology of violence becomes part of a new anthropology of the *body, in which the body becomes a privileged site for the inscription of signs of *power.
What is harder to escape is the assumption that questions about violence are inevitably questions about human nature. Simon Harrison (1989), writing about the Avatip of the Sepik River area of New Guinea, argues that the Avatip distinguish between two types of sociality linked to two different concepts of the *person. The unmarked type, so to speak, is one in which everyday social relations are lived in an idiom of peaceful equality; the other type of sociality, encountered in the world of men’s politics and men’s *warfare, is marked by assertion, aggression and potentially uncontrollable violence. This second type is not, however, treated as a natural property of men, but rather as something which has to be created and sustained in ritual action. In order to perform those acts of violence which warfare requires (and warfare itself is politically necessary if Avatip society is not to descend into entropy), Avatip men have to acquire the capacity to be violent.
Harrison’s argument is an excellent example of the way in which cultural accounts of other people’s ideas about violence, *gender and personhood can serve to undermine powerful Western assumptions about human nature. Such cultural accounts do not, though, clarify any of the definitional confusions in the analysis of violence. Even in societies with an explicit concept which we could translate as ‘violence’, not all acts involving the deliberate inflicting of physical pain, marking or damage to another’s body are defined as ‘violent’. Are *sacrifice, †circumcision, tattooing, fighting, and biomedical procedures ranging from appendectomy to electro-convulsive therapy, all usefully classifiable as acts of ‘violence’? Do we dismiss acts of *witchcraft and sorcery which are clearly intended to cause bodily harm, even if we doubt their efficacy? What of attempts to break down such literal assessments of violence, like †Bourdieu’s use of the term ‘symbolic violence’ to refer to acts of coercion which are usually unaccompanied by overt physical violence? One way to imagine an anthropology of violence is to see it as a kind of mapping of the different moral and aesthetic evaluations people in different contexts make of their actions on the bodies of others.
Precisely because violence has such a central position in Western theories of power and human nature, anthropological evidence has an important capacity to disturb and unsettle our received understanding of violence. It is hard, though, to imagine such disparate evidence leading to a new anthropological synthesis, in which violence remains as central and as unquestioned. Instead, the broad category of ‘violence’ seems to contain particularly valuable evidence which can help us to explore the links between two connected aspects of human life: what Mauss called the ‘techniques of the body’; and the intersubjective world of signs and communications.
