Sinja Graf has written a perceptive and generous appraisal of The Problems of Genocide that accurately delineates its analytical categories and normative standpoint. Especially in view of the book’s length, no author can ask for more. She presses me on three points, which I address in turn.

The first is the question of the “(dis)continuities between Christian and secularized constructions” of the “thresholds” that, if crossed, “shock the conscience of mankind.” Where she stresses discursive rupture in her book, I underline continuity. I do so because the Christian universalism of the Salamanca school of theologians is not the origin of the notion of the “universal crime.” It lies in antiquity, with the Greek stoics like Diogenes and Zeno, whose notion of universal citizenship or a single human community was taken up by Roman thinkers such as Cicero, who quoted Socrates’s declaration that he was “a citizen of the world.” This notion did not imply a relativistic universalism, as sometimes supposed, but distinguished between those deemed worthy or able to join their polities, whether Greek or Roman—namely, the wise—and those who were not worthy. The notion was thus hierarchical from the outset and is the basis of both Christian and secular conceptions of “mankind” or “humanity” that were invoked in European-Western global dominion for the next half-millennium. Not that one should distinguish too emphatically between the Christian and the secular. If Pax Britannica was spread by the British Navy, it was constituted by Christianity, as well as by commerce. Missionaries were central to the civilizing mission even if they often clashed with colonial officials.

Second, Graf asks me to delineate the concept of “permanent security” with greater precision, proposing to contrast it with what she calls “transient security.” I would analogize this with the distinction between vigilance and hypervigilance. The former is rational, when states engage in armed conflict where necessary but do not seek to wipe rivals from the map. The latter, by contrast, is irrational, a paranoid threat perception that embarks on once-and-for-all “solutions” to geopolitical security. To be sure, the line may be difficult to draw in some cases. Looking for indicators like preemption—anticipatory self-defense—and levels of wanton civilian destruction may be helpful in determining the relevant modality of security.

Third, Graf teases out a seeming tension in the argument about permanent security: on the one hand, I focus on the brutal fact of civilian destruction rather than state intent to highlight blindspots of the latter, which permit tremendous amounts of killing if motivated by military objectives; and on the other, I condemn permanent security as the state intention to perpetrate final solutions. I am in fact arguing that we can discern the deeper logic of permanent security in both its liberal and illiberal versions by beginning with the impact of the state and parastate violence: the killing of civilians, whether in short, intense bursts, as in genocide, or in longer-term campaigns and occupations in which smaller numbers are killed serially but mount, tragically, over time. My wager, finally, is not to conceptually stretch genocide to include the latter outcomes but to discard the notion of genocide and replace it with permanent security.