Posted on behalf of Michael Barnett.

David Keen certainly puts the complex into complex emergencies. Combining critical theorizing and detailed knowledge of conflict zones around the world, Keen challenges a mountain of received wisdoms, urban myths, and simplified understandings regarding collective violence, aid, reconstruction, and peace-building. Keen tells us that: not everything is as it appears; every solution can generate its own problems; every problem might be a solution to something for someone; steps might be missteps; and even, with some luck, stumbles might accidentally propel war-torn societies toward a more stable outcome.

While in genuine admiration of the book and Keen’s other path-breaking contributions to the study of violence, Complex Emergencies contains two tensions that represent a challenge to all of us who are interested in how well-intentioned international interventions, in all its forms, might advance rather than retard human welfare.

The first tension is between the search for generalizations and an appreciation of history’s multicausality, complexity, and contingency. Keen registers considerable frustration with those who are seeking to identify enduring patterns and, in the process, simplify complex causal processes. Exhibit A is Paul Collier’s work on the causes of civil wars. Collier’s work has been immensely influential in part because he advances a fairly elegant thesis about the causes of civil war and collective violence. Yet Collier has his many critics and those looking for a point-by-point rebuttal would do well to turn to Keen’s chapters on his work. Indeed, the chapters are so passionately and convincingly written that the reader is left to wonder why anyone paid Collier much attention.

Keen does not advance an alternative thesis but rather many different alternatives. As Keen runs through the complex and conjectural causes of civil wars he identifies a laundry list of variables; no single factor is responsible for the patterns of civil war, or even a single civil war, but rather lots of factors. Those like Collier who have advanced specific hypotheses are reminded by Keen of the stream of exceptions in places like in Angola, Sierra Leone, and Congo. Every possible generalization is rebutted by several counterexamples. Detailed historical explanation is privileged over any conditional generalizations and historically-bounded explanations.

We should not be forced to choose between the search for a master variable (greed, grievance, prisoner’s dilemma) that can generate, at times, dumbed-down understandings, or case-driven complexity and contingency that can generate descriptive richness but little cumulation. While I am not in the Collier camp, we should not forget that one of his considerable achievements was to bring more rigor to what could be a maddeningly case-specific approach. Collier and others of his positivist persuasion have done a real service in forcing us to try and identify what might be the principal and underlying causes of civil war. And, they have helped us identify which variables might be the most important, which might be direct or intervening variables, and thus advanced both of our understanding of causes of collective violence and possible policy prescriptions.

Indeed, a fair bit of the literature on civil wars avoids the tails of description versus explanation. And, while Collier dominates Keen’s chapters, he does not dominate the research on civil wars. There is now a fairly rich, systematic, and sophisticated literature on civil wars, much of it aimed at Collier and the greed thesis. This literature does not really figure in Keen’s discussion, but those like James Fearon, David Laitin, and Stathis Kalyvas have moved our understanding forward precisely because they have tried to identify contingent generalizations. Indeed, I see some real progress in our knowledge regarding the causes of civil wars (and don’t agree with Keen that there exists a dominant line of argument that civil wars are caused by irrational forces) and not convinced that embracing complexity represents a step forward.

The benefits of moving beyond the false choice of nomothetic and idiographic explanations is illustrated by the developing literature on peacekeeping and peacebuilding. Throughout much of the 1990s this literature was woefully unsophisticated, offering a choice between rather simplistic rules (peacekeeping is only effective when there is a peace to keep) and data driven by individual cases that never seem to lead to any knowledge cumulation because scholars were overly fascinated with the particulars of their case (Somalia is different from Angola is different from Mozambique is different from El Salvador and so it goes). Real breakthroughs came, though, when scholars became serious about rigorous empirical analysis, working through middle-range theories, and teasing out their generalized arguments in specific cases to identify causal mechanisms. We now can point to several outstanding pieces of scholarship, including those by Michael Doyle and Nicholas Sambanis, and Page Fortna, which provide a real advance in our understanding of the conditions under which outsiders might be able to make a difference.

The second tension is between critical theory, and critique in general, and practical recommendations. Practitioners should read Complex Emergencies because of their tendency to be captured by their existing worldviews, failure to go beyond surface impressions, reliance on their favorite historical analogy to guide their categorization of the situation and identification of policy options, and general lack of critical introspection. All practitioners who have been lulled into a false sense of confidence will be forced to think twice about what they think they know – always a good thing. Moreover, and to his credit, at various moments Keen suggests not just cautionary tales but also possible exit options for policymakers looking for ways to do good.

The problem, though, is that Keen knows too much – he is deeply aware of conflict’s contingencies and that policy options that might have a reasonable chance of success in one situation might do incredible harm in another. And we might not even be sure that we have defined the situation correctly, or rather, taken into account all the different ways that the situation is defined by the different participants in the situation. So, one step forward, one step backward, one step sideways.

This “yes, but” form of analysis points to a problem that is familiar to all, including myself, that have struggled with the inherent tensions in critical theory between critique and praxis. We want to expose the underlying power structures that produce and reproduce injustice and violence, which frequently lead us to structural variables that are not easily changed (think capitalism or discourse). In other words, something of a dead end. Yet we also want to identify ways to make small but appreciable changes that have the capacity to reduce suffering in the short and medium term. Sometimes this means sacrificing the perfect for the good. Keen’s chapter on humanitarian assistance, for instance, discusses ways to alleviate the possible harm of aid by suggesting that humanitarian agencies might have to make bargains with the very people that benefit from the emergency. Yet how do such compromises leave intact existing power structures and thus hinder genuine change? In fact, the chapter on aid fails to consider how strategies that are designed to minimize harm during humanitarian emergencies might, in fact, frustrate the causes of peace as explored in a later chapter. The book’s signature moments, in my view, are when he begins to pull together the critical theory with forms of pragmatism. At these moments we go beyond contingency to identify some thoughtful do’s and don’ts.

It is perhaps unfair to fault Complex Emergencies for not silencing the twin tensions of explanation vs. description and critical theory vs. praxis. I can hardly point to many examples of those who have (certainly not my work). But the challenge for those (of us) committed to an emancipatory social science is to wrestle with these tensions as much as possible, to search for underlying structures that can give us leverage over various events even as we appreciate that all cases are unique in their totality of elements, and to constantly ask how specific policy prescriptions not only are the “least bad” at a particular moment but also might open a space for positive change.