Ian Spears sets himself some sound and needed tasks: to show that understanding the insurgents' self-interest is necessary as the starting point for conflict resolution and that such self-interest is a search for security. Security is shown as a rational basis for action, as opposed to primordial or reprehensible sources of conflict behavior. This security is pursued through a choice among three strategies: integration, domination, and separation, also a commonsense categorization. At this point, it would be interesting to see some kind of theory of conditional preferences, applying that fundamental question of social science analysis: Which when, why, and how? I am not aware of any convincing or even attempted efforts to tell which strategies will be chosen first, under what conditions, and so the challenge is important and the results would be of major analytical and practical use. Spears then addresses the “why” in outlining the factors that can influence strategic choice: distribution of power, previous experience, internal attributes (resources, demography, geography), and global and regional factors. In a word, everything, although personality, cost/benefit calculations, opportunity, and frustrations are explanatory factors that others have cited in attempting to answer the same question. But the “when” and “how” are not addressed systematically in even implicit hypotheses, a missed opportunity.

The main part of this book is devoted to three case studies: Eritrea's efforts to achieve independence from Ethiopia, unification and secession in Somalia, and the Angolan civil war. The cases are well-done, comprehensive accounts of the attempts of the variously defined groups to achieve their security in a number of ways. The groups are not simplistically equated to an ethnic protest; the struggles over identity in Eritrea, and the wavering role of the ethnic factor in Angola in Somalia and Somaliland, are carefully treated. A major omission is the role of the external enemy in Somali cohesion and identity; there may be other details of interpretations or small factual omissions that might be signaled, but they do not dominate the accounts. Spears has a good grasp of all three cases.

The analysis of a choice from among the three strategies is less clearcut and convincing, however. Jonas Savimbi supposedly shifted from integration to domination as a strategy, although the first choice is not clearly documented (nor could it be), and so the shift is neither clearly identified nor the reasons explained. The Isaaq of the Somali National Movement moved from overthrow of Mohamed Siad Barre to secession as a means of achieving security against Oganden dominance (p. 142), but there is much more to Somaliland's earlier integration with Somalia and then to its own characteristic intraclan fighting after its secession than simple single-minded Isaaq self-protection. The Eritean and Tigrean Peoples Liberations Fronts (EPLF and TPLF) made a cohabitation (scarcely marriage) of convenience to overthrow the Derg, but when the common enemy was toppled, each went its separate ways dictated by the basic identity, so that integration and secession are rather evidence as separate choices. What explains the internecine battles of the various Eritrean nationalist movements until the arrival of the EPLF on the top, other than the failure of one or another to capture the flag, to which little attention is paid? According to Spears, “Survival in many African states requires not simply coercion but vigilance [not one of the posited explanatory variables]. It means understand when one stands to benefit from compromise and cooperation and when one needs to remove challenges forcefully” (p. 978). But this would suggest that in the end (and the beginning), dominance is the only strategy that matters, determining the choice between integration through dominance and secession for dominance. That does not explain why any other strategy is tried, or when the light strikes about the unique need for dominance. But which groups are hell-bent on dominance? And does understanding that the insurgents' self-interest is necessary as the starting point for conflict resolution mean that every (self-declared) group needs to be given its dominance to settle conflict?

The conclusion does not tell us. It does not pull the three themes or other ends together. It tells us that the international community has three strategies, once the inevitability of African anarchy is recognized: intervene to end anarchy and control its harmful effects, recognize that it can do nothing in the face of local strategies, or invent something to replace the inadequate state. The conclusion states that all three approaches have their pros and cons; it does not refer back to the threefold strategic choice with which the study started, nor a fortiori address the “which when why and how” question. And so it leaves us with three good case studies as examples of different strategies adopted under various conditions, with domination through secession as the dominant strategy, and conflict management a pretty lost cause. That is realism, but it is not much of a guide for how to handle it.