Over the past two decades, attention in the social sciences increasingly has been drawn to the problem of violent civil conflicts, a problem that has disproportionately affected Africa more than any other region. Two approaches to this problem have come to dominate the field: attempts to understand the root causes of civil conflict and attempts to understand the dynamics of its violence. Critics of the former approach have elaborated the ways in which the etiological agenda itself makes, and then politically mobilizes, the reality it claims to find. The goal of this article is to elaborate a similar critique for the latter agenda by examining the productive and destructive interaction between theoretical assumptions and empirical realities that have informed attempts to understand the Algerian massacres of the late 1990s. The overall intention is not to promote a new understanding of those atrocities. Rather, it is to gain a deeper insight into the processes by which episodes of mass civil violence become objects of scientific analysis—and thus objects for political utilization—despite their having emerged from an empirical milieu of contested, ambiguous, and indeterminate realities.