We live in a world laced with forms of political violence. Kristen Renwick Monroe's latest work develops an interesting social psychological account of the conduct of victims, perpetrators, and bystanders of the most extreme form of violence—genocide. It also employs an interesting narrative approach that contributes to broad methodological discussions in political science about the ways in which subjective experience can best be understood. We have thus invited a diverse group of political scientists and historians to comment on the book's analysis of political violence and on its broader approach to the study of politics.—Jeffrey C. Isaac, Editor

While it is widely accepted that negotiated settlements are prone to breakdown, our understanding of the processes through which signatories defect lacks precision. A growing qualitative literature recognizes the potential for rebel group fluidity, yet the conflict field’s converging reliance on dyadic data obscures pathways of defection that result in splintering or merger in quantitative studies. An in-depth case study of a failed peace process in Uganda—which is misclassified in the extant data—helps to illustrate the ways in which excluded groups can lower the opportunity cost of defection for splintering factions, resulting in a strategic alliance. I test the generalizability of this argument against the full sample of rebel parties to settlements in Sub-Saharan Africa (1975–2015) using a large-N qualitative analysis of causal process observations (CPOs). The aggregated results provide strong evidence that the defection-by-alliance pathway is much more prevalent than previously recognized, accounting for more than one-third of all defections in the sample. Where settlements create shared incentives for stakeholders inside and outside the peace process to spoil, rebel elites appear more willing to bear the costs of an alliance with a rival, rather than surrendering under adverse conditions.