The Promise of a "Disentangling Alliance": Collective Security and the Statecraft of Violence in the Age of Industrial Modernity

Abstract

In 1919 and, again, in 1945, the statesmen and policymakers of the victorious great powers resolved to rebuild international order, sundered by years of total war, around the principle of collective security. In this dissertation, I explore the reasons behind that decision and consider their implications for our understanding of the possibilities for, and limits to, decentralized violence cooperation. I shall argue that the founders of the League of Nations believed that the industrialization of international relations fundamentally transformed the logics of war and peace. They recognized that, with the advent of industrial modernity, the security of every state became dependent on the stability of the international system at large. And they devised collective security to cooperatively manage that dependence. But their expectations for collective security -- or, indeed, their diagnosis of the strategic ramifications of the industrialization of international relations -- cannot be understood in terms of familiar concepts and conjectures. To properly grasp them, and appreciate the broader possibilities for decentralized violence cooperation that the semi-centenarian experiment in collective security implies, we first need to enrich our conceptual vocabulary and causal grammar. One crucial addition that we require is the distinction between positional and systemic risks. In the particular case of violence, positional risks describe a state’s vulnerability to military attack, while the systemic risk of a strategy measures the cumulative expected instability in distributional settlements across all positional disputes in the international system that would ensue should that one strategy falter. Using this distinction, we can articulate the rationales of the founders precisely: they created collective security, I shall argue, to preserve their states against systemic risks induced by unreliable alliances. I shall further show that the distribution of systemic risks has been a crucial, if neglected, determinant of the patterns of rule-governed violence cooperation over the last two centuries -- even beyond collective security. This analysis has important implications for our understanding of opportunities for, and limits to, decentralized cooperation. Most crucially, for rule-governed violence cooperation to be successful, its institutional design and distributional implications must be tailored to expected volatilities in topologies of systemic risk.