Many thanks to Paul Staniland for his thoughtful assessment of my book. He identifies and parsimoniously conveys the book’s main contributions. It is such a pleasure to have a smart, incisive, concise review of one’s work!
Staniland develops three important critiques. First, he argues that even benign-sounding political discourse constitutes a practice of power. Political narratives of pluralism, integration, and inclusion—which, I argue, make the strategic choice of genocide less likely—may alienate some in society and elide ethnic hierarchies. Agreed. I see now how more attention to this matter would have been useful.
That said, some context is in order. The book compares genocide cases to non-genocide cases. The latter are not Denmark. They are situations that could result, plausibly, in genocide. They are countries with deep political instability in a civil war with states that have committed significant human rights violations, and with militias that have sown havoc and violence. I find that certain political narratives restrain elites from choosing to respond to military threats through mass categorical violence. The implication is not that the situation in these countries is hunky-dory or that these narratives are unequivocally “good.” The claim is that these narratives tip dangerous situations away from mass violence and toward political accommodation. While I see how Staniland develops the implication he does, that is, that I am making normative claims about “good” and “bad” political discourse, the intention was to limit the argument to the ways in which elites develop strategies of violence.
Second, Staniland argues perceptively that the case studies show “political leaders neither fully including nor completely excluding” social categories. Again, agreed. Indeed, after spending time with the empirical record, I concluded that the reality is not a simple dichotomy of “exclusionary” versus “inclusionary” narratives, as Staniland suggests. Rather, genocide is more likely when a founding narrative establishes a primary identity-based population whom the state serves (p. 66). Through time, in the genocide cases, elites sometimes provided citizenship rights and afford limited political power to non-core populations. However, those leaders always maintained a hierarchical vision of political community and political power; they associated an identity population with the core political community and with the state’s rightful ruler—in contrast to the non-genocide cases, where elites did not associate the state with a core identity population. In short, I agree with Staniland’s reading of the empirics, and I sought to represent that nuance in the book.
Third, Staniland raises a concern that my theoretical emphasis on restraint requires better specification. The book develops two claims. First, the book makes the theoretical case for the reasons that restraint should factor into the analysis of violence, and the book offers a number of plausible sources of restraint. Second, the book finds empirically that three sources of restraint mattered, in addition to some narratives: capacity, economic structure, and external intervention. Capacity is a notoriously slippery concept, and thus I disaggregate it into coordination, identification, control, and infliction. I argue that all are necessary for genocide (with some rare empirical exceptions that I discuss). That said, Staniland’s point is well taken. Restraint is something of a residual category; no single source of restraint reappeared across the cases. I find that restraint matters, but I would be delighted if future research refines the claims.
I am grateful for Staniland’s careful reading of Making and Unmaking Nations. I hope the contributions he identifies will resonate for readers. Other readers may benefit from different parts of the book, in particular where I discuss the concept of genocide, the legacies of leadership in Africa, a framework for atrocity prediction and identification, and a template for atrocity prevention and response.
