Initial interpretations of Mesolithic Portuguese material as evidence for violence are questioned: the alternative suggestion is of accidental trauma, especially in childhood. A situation of undoubted extreme interpersonal violence in Kenya is shown to relate to withinpopulation conflict caused by external pressures. Intensified violence between North American groups of Northeast Woodland peoples in the seventeenth can also be related to external factors (and constraints on resources): though documented ethnohistorically, the osteological evidence for this violence is extremely limited, demonstrating that absence of evidence is not to be relied upon. A Neolithic Chinese site is examined next, as the focus of the paper, since it provides clear physical evidence of violence, despite the lack of archaeological recognition of conflict, pointing to the need to be wary of the interpretations in this sensitive area of anthropological study. The background of the Jiangzhai sample is described and it is noted that 64% of the victims were female. Meggitt's research on violence within and between groups is cited in the following discussion, and the conclusions focus around the need for care in concluding that violence has or has not occurred; the need to avoid simplistic explanations for evidence of violence and the necessity of setting that evidence within a broad context - one with chronological and social/geographical depth and breadth; and the need for sensitivity in recording and interpreting violence.
