Library at Trininty College, Dublin.

At the Sié Chéou-Kang Center I am working on my book manuscript, “Understanding Militant Splinter Groups.” The project explores variation in the survival and radicalization of groups, like the Islamic State and the Real IRA, that break away from established militant organizations. This is an important topic. To begin with, in protracted conflicts it is common for militants to break down and split apart, with new organizations continually emerging from the ranks of existing groups. In the new data collected for this project I find that nearly a third of militant groups form as splinters, suggesting that it is a central pathway by which militants come to exist. Historically, this process of organizational splintering has created some of the deadliest and most well known militant groups as well, including Al Shabaab in Somalia, Black September in Palestine, and the Real IRA in Northern Ireland. Yet at other times these splinter organizations quickly disappear, failing to impact the conflict in any meaningful way.

Existing research commonly assumes that splinters will universally grow deadlier and more radicalized than their predecessors. However, my findings challenge the conventional wisdom and I show that understanding their trajectory requires us to first understand their organizational split. Drawing on insights from studies of voter behavior, political party formation, and party identity, I devise a new theory that can explain the divergent trajectories of breakaway groups. In brief, I argue that two aspects of an organizational split are most important: first, the number of internal disagreements; and second, what these disagreements are about. Ultimately, my research shows how militants’ long-term trajectories are strongly shaped by their initial formation, which has important implications for how researchers understand the evolution and inner workings of violent – and even nonviolent – nonstate actors. These findings also help policymakers understand and anticipate the behavior of specific groups, and also for how they respond to increasingly fragmented conflicts around the globe.


Splinters emerging from a single, shared internal disagreement are best prepared to survive. These groups form with clear organizational objectives that attract a highly cohesive base of recruits. This is largely due to their focused identity that reduces uncertainty among potential supporters about their goals and their methods, and that appeals to a more narrow, homogenous subset of recruits. Their internal cohesiveness, or preference-alignment, bolsters their survivability by lowering the odds of defection and infiltration. This makes it easier to decentralize their operations and, ultimately, to survive government repression. Conversely, organizational splits stemming from a variety of unfocused grievances create diffuse group identities that appeal to a broad segment of the population, leading some people to join for one reason and others for another. This has the effect of lowering internal cohesion which destabilizes the organization and amplifies the efficacy of government repression and restriction.

Additionally, I find that the nature of internal disagreements is also important. Militant groups break apart for a variety of reasons, and these disagreements influence how splinters will eventually behave. Strategic, ideological, and leadership disputes each uniquely shape splinters’ organizational goals and who they appeal to, causing them to behave in very different ways. For instance, strategic disagreements, of the type that produced the Islamic State and the Real IRA, tend to attract hardliners away from the parent organization, leading to increasingly radicalized group behavior. Disagreements over ideology or leadership should produce new organizations that do not substantially diverge from the tactical and strategic choices of their predecessor. Instead, they should continue to rely on what they know since they have little motivation to change their ways.


I test my theory using a mixed-methods approach: first, I analyze an original data set of organizational fragmentation among a random sample of 300 militant groups from 1970 to 2014 which provides robust cross-national support for my theory. To construct this new data set I researched each of the three hundred groups to determine how each formed and, if it is a splinter, from whom it broke away, and why – in essence, creating a family tree for each group. With this data I am able to analyze the behavior of splinter groups both in relation to one another, but also in relation to the broader population of militant organizations.  Second, I explore the theory’s causal mechanisms through an in-depth case study of republican militants in Northern Ireland. To this end I spent three months in public and private archives in Belfast, Dublin, and London gathering declassified government documents, interviews, letters, and internal group communications.  Taken together, these investigations provide strong support for my theory, demonstrating that the durability and behavior of militant splinter groups is strongly influenced by the internal disagreements that led to their formation.