Studies of the 1994 Rwandan genocide have moved, Scott Straus argues, beyond simplistic interpretations in terms of “tribal” or “ancient” hatreds (interpretations that were, in truth, more those of the media and politicians than of the early academic literature) toward a “new consensus” that this was a modern genocide based on elite planning, nationalist ideology, and media manipulation. Straus argues that while this is not wrong, it does not go far enough to explain why genocide happened and why so many Hutus were mobilized to kill their Tutsi neighbors as well as moderate Hutus. Emphasizing the need to link the national, elite level on which most study has focused with the local level in the rural areas where most killing was done, Straus undertook a unique study, interviewing more than two hundred confessed and convicted male perpetrators in Rwanda's jails.

On the basis of this study, the author argues for a new theory of the genocide that prioritizes three main factors. The first is the civil war, and especially the new phase that broke out between the government and the Rwandan Patriotic Front (RPF) in early 1994: “[W]ithout a war in Rwanda, genocide would not have happened.” It “provided the essential rationale for mass killing: security” (p. 7). Indeed, genocide was predicated on threat; war legitimated killing, created an atmosphere of insecurity, and involved specialists in violence (soldiers, gendarmes, and militias) who were central to the violence against Tutsi civilians. The atmosphere of insecurity was greatly heightened by the assassination of President Juvénal Habyarimana (about which Straus inclines toward another new consensus, blaming the RPF, rather than the genocidists). The second factor is the nature of Rwanda's state institutions, historically centralized (Rwanda was a single kingdom before the arrival of colonializers), based on unified culture and language (Hutus and Tutsis are not differentiated in these ways), with considerable mobilizing capacity in rural areas (manifested in the pregenocide period by mobilization for compulsory labor), and heightened by favorable geography (Africa's most densely populated state has few inaccessible areas). Thus, Rwanda's strong state defies the African stereotype of weak or even collapsing state apparatuses operating within artificial colonial-era boundaries. The final main factor is a particular ideology of ethnic categorization that conflated the Tutsi population in general with the armed “Tutsi” enemy, the RPF, and so labeled all Tutsis “the enemy.” This, rather than historic ethnic difference, is the significance of ethnicity for the genocide.

Straus's unique empirical study carefully, credibly, and fully backs up this interpretation. Despite the obvious dangers in interviewing perpetrators, especially under conditions of confinement—of which he is fully aware and takes account—Straus produces unprecedented information about the local-level spread of violence, its nature, and the reasons for participation in it, which will be invaluable to scholars. Some findings are counterintuitive: Contrary to the stereotype of very young men, even children, mobilized in Africa's armed conflicts, he finds that those who were in their teens and 20s at the time were (like, less surprisingly, the over-50s) underrepresented (compared to their proportions in the population); perpetrators were predominantly in their 30s. The explanation is the way that the elite sponsors of the genocide mobilized the population, following existing patterns of organization and recruiting participants by household. Another finding is that, though Rwanda's was the most mass-participatory of major genocides, “only” 14% to 17% of Hutu men participated. Killing was still a minority activity and, although local instigators were considerably motivated by the threat of the Tutsi RPF and the effect of the assassination, most participants were motivated more by fear of intra-Hutu intimidation and violence. Another interesting feature is the inclusion in Straus's sample of communes of the only one where genocide did not take place: Here he finds that the arrival of the RPF in a neighboring commune stalled the rapid spiral of escalation that typified killing in the localities (though varying considerably among them, in ways that he plausibly explains). This case is significant, he argues, for the prospect of intervention to halt the killing: A very rapid international response could have made a difference.

This is but a small sample of many fascinating and important findings. Straus's study is comprehensive, thorough, and cogently and carefully argued, and it engages stringently with the literature. It is altogether an impressive work that will be compulsory for specialists and invaluable for students. Straus is a former journalist and his writing is a model of clarity and economy; this book will be accessible to most readers. Generally, The Order of Genocide supports the emergent theme in genocide studies that war is crucial to causation; in terms of the debate on participation, it supports the position of Christopher Browning (Ordinary Men, 1993), rather than Daniel Goldhagen (Hitler's Willing Executioners, 1996), emphasizing group pressure rather than racial ideology.

I have two criticisms, one of which is serious. The lesser point is that the title emphasizes the extent to which genocide was produced by the “order” that Rwanda's deep state penetration produced, and to which it mobilized in accordance with given patterns of obedience, while the argument in the end prioritizes disorder and insecurity, which are not reflected in the title. The more serious point is that Straus's definition of genocide equates it with killing. This has only minor methodological and analytical consequences in this study, as when he excludes those who looted but did not kill from his category of “perpetrators.” But from the point of view of comparative study this is a narrow definition, which would exclude episodes where perpetrators did not simply and unremittingly focus on killing—like the “ethnic cleansing” of Bosnia-Herzegovina that was contemporaneous with Rwanda's genocide—from the scope of genocide studies. It is unfortunate that such an exemplary study should sustain a misleading idea of the field and its concepts.