As a critical medical anthropologist, my terrain if the "body," which, like the vexed designation "coloured,"(1) requires a standing set of quotation marks to indicate the body's contested status as both bio-existentially "given" (the source of all certitude, as Wittgenstein  saw it) and just as surely "made up" (and the source of all doubt). My specific set of concerns, derived from an early and indelible reading of Foucault, is the "everyday violence," the little routines and enactments of violence practiced normatively on vulnerable bodies in families, schools, hospitals, medical clinics, in various administrative and bureaucratic settings (from mayor's office and public registry office to the public morgue and the graveyard) -- all of which, in Franco Basaglia's (1986) sense of the term, are "institutions of violence."
While studying the madness of everyday life in the mid-1970s in a small, quiet (but nonetheless psychologically "violent") peasant community in western Ireland, I was largely concerned with interior spaces, with the small, dark psychodramas of scapegoating and labeling within traditional farm households that seemed to be driving so many young bachelors to drink and to bouts of psychotic depression and schizophrenia (see Scheper-Hughes, 1979; 1982). I paid scant attention to the activities of little Matty Dowd, from whom we rented our cottage in the remote mountain hamlet of Ballynalacken, and who used our attic (with our silent consent) to store a small arsenal of guns, rifles, and explosives that he and a few of his Sinn Fein buddies were running to Northern Ireland. Consequently, I left unexamined the possible links between the political violence in Northern Ireland and the tortured family dramas in West Kerry that I carefully documented.
Since then, I have studied other forms of what I call routine or "everyday" violence: the abuses of medicine and psychiatry when they are practiced in bad faith against the weak, the mad, and the hungry, as well as the social indifference to child death in Northeast Brazil that allows political leaders, priests, coffin makers, and shantytown mothers to rather casually dispatch a multitude of hungry "angel-babies" to the afterlife (Scheper-Hughes, 1992). Yet even in Brazil, I did not begin to study the structure and meaning of political violence itself until (beginning in the late 1980s) the spouses and half-grown children of my comrades,(2) friends, and neighbors in the shantytown of Alto do Cruzeiro began to "disappear," their mutilated bodies turning up later, the handiwork of police-infiltrated local death squads (see Scheper-Hughes, 1992: Chapter 6; 1994b).
Until then, I had believed that a running analysis of the political violence occurring within the context of military dictatorships, police states, or in times of transition during and after civil wars and wars of liberation was best handled by political journalists. For one, anthropologists were too slow, too hesitant, too reflective, and our knowledge was too local, too embedded, while political events were altogether too fast, too volatile, too unstable, so that by the time anthropologists had something to say, it was usually irrelevant or obsolete. Since the Brazilian newspapers insisted on running stories about the "dangerousness" and "violence" of shantytown dwellers (especially poor young Black men and boys), a perceived threat that made the work of the death squads seem to be a necessary defense against the anarchy of the favela, I saw that anthrological interventions were necessary to correct the manipulative (even exploitative) half-truths of the media. Despite our relentless self-critique of anthropology, the greater danger resided in leaving the analysis of violence to the journalists. With shades of Levi-Strauss' (1963) comment on Quasalid, the Native American shaman, our anthropological truths might be false (or at best partial), but they were certainly less false than those of the media. …