This is a study of individual and collective conflict. Social scientists have long noted a connection between the notion of honor and the various forms of violence that human beings use to defend their social positions. This connection is especially clear in feuding, a type of conflict in which well-defined groups in the same community engage in hostile competition over extended periods. But existing scholarship in this area tends to offer either universal, idealized models that disregard the context and intentionality of social behavior, or richly textured, particularistic accounts of cultural and discursive practices (conceptions of honor, social standing, masculinity and femininity, loyalty, and so on). This research adopts a perspective that takes seriously the intentional, interpretive aspects of feuding, but which nonetheless works with concepts that can be translated across cultural contexts. The approach connects what scholars already know about feuding with sociological research on the network context of collective action. This is an historical study of individual and collective violence in Corsica from 1850 to 1890, but the aim is to develop general theories that will apply today in many societies. The advantages of the chosen geographical and temporal setting are a broad time span, a clearly defined and stable study population, and access to information that is often confidential in contemporary court documents. Transcripts of roughly five hundred violent incidents will be subjected to textual and statistical analysis to illuminate the process by which individual conflicts become (or do not become) collective conflicts. Feuding is a significant cause of violence, injury, and loss of life within traditional communities. Insights gained by studying feuding can help us understand similar phenomena, such as clan-based political conflict, civil war based putatively on ethnic difference, and urban gang warfare. The most vexing feature of violence in such contexts is the tendency of individual disputes to become collective, vastly expanding the number of potential victims. This research will be guided by the premise that escalation is not automatic but depends on social processes that can be specified, understood and, in principle, circumvented.
