For years it was argued by North American scholars that violence and coups in Latin America were aberrations from a more desirable or more "natural" course of political change: the agreement to abide by election results. More recently, astute observers have demonstrated that coups and coercion are often not unusual or even disparaged, but rather an institutionalized way of doing political business. In various political systems, certain levels of violence may become considered appropriate, if not almost de rigueur, a regular feature of the political landscape (Horowitz, 1968:45-70; Anderson, 1967: Chap. 4; Payne, 1968; Fagen and Cornelius, 1969:383-419). Perhaps the pendulum has swung too far. We have become inclined to regard some degree of violence as too natural in Latin America without examining the different origins and catalysts of violence in different political contexts. In some countries certain levels of violence associated with political activity may be endemic, the periodicity and ranges of violence rather predictable. In others, high levels of violence may represent rather unusual political business. To distinguish between violence as the customary partner of political acitivity and violence as an egregious "outside agitator" is to raise an issue often obscured by cross-national studies of violence. Certainly cross-national studies have revealed a good deal about the socioeconomic and demographic factors associated with different levels and varieties of violence (Feierabend and Feierabend, 1966; Bye, 1968; Tanter, 1967). What those studies have rarely investigated, however, is the process by which violence is linked to political activity. This essay explores that process by examining a case study: the phenomenon of la violencia, a wave of unusually savage, widespread