My aim in this paper is to propose ways of overcoming obstacles to the development of a theory of violence that can bridge previously fragmented domains of inquiry. The solutions proposed are to (1) neutralize the definition of violence by (2) locating the phenomenon of violence within the conceptual framework of conflict such that (3) the same set of basic assumptions may be employed to analyze instances of violence exhibited by the powerful or the powerless-instances which vary in their degree of social acceptability. The assumptions presented move progressively from purported relations between 1) unequal distribution of resources, 2) asymmetric social relations, and 3) conflict-to the conditions that affect a transition from a) latent to manifest conflict, b) nonviolent to violent manifest conflict, and c) from to violence. Conditions that limit the applicability of these assumptions are discussed, followed by illustrative applications of the approach to two forms of interpersonal violence-parental violence and differential rates of male and female violence. The phenomenon of violence cuts across numerous disciplines and fields of inquiry. The study of violence, therefore, tends toward a fragmented inquiry into this or that form of violence. There are specialists in criminal violence, collective violence, family violence, political violence, media violence, social control violence, self-inflicted violence, international violence, social change violence, and so on. Such fragmentation undermines the development of the motivation and the requisite expertise to build more general theoretical frameworks. The most serious fragmentation has occurred between explanations of deviant violence or violence that is socially defined as illegal, unacceptable or immoral (e.g., parents beating their child) and normative violence or violence that is socially defined as legal, acceptable or moral (e.g., parents spanking their child). My aim in this paper is to propose ways for us to overcome this fragmentation and other obstacles to the development of general theories of violence. The specific obstacles considered are (1) biased definitions of violence that have led to (2) the assumptions that deviant or socially disapproved violence is caused by some personal or social system deficit while normative or socially approved violence is caused by normal processes of conflict, change and control. I propose that we (1) neutralize the definition of violence so that its social definition is treated as a variable rather than a phenotype; and that we (2) locate the phenomenon of violence within the conceptual framework of conflict such that (3) one set of basic assumptions can orient analyses of all forms of social violence.