The Logic of Violence in Civil War Book

A very solidly argued examination of the sources of violence in civil war. Although it was a significant undertaking to get through, this may be one of the best-structured political science books I've read to date ' the author lays out his concepts, proceeds methodically through alternative hypotheses, spells out his theory and then drills down into a detailed case study analysis based on quantitative and qualitative research, closing out with two chapters further unpacking the theory's implications. A chapter at the outset on the biases that confound the study of violence is worth the read on its own as a caution to any researcher, as is the subsequent chapter dissecting alternative hypotheses that attempt to explain the prevalence of violence. For the most part this is also clearly written throughout, and clearly the product of a tremendous literature review of other studies of civil wars.

The main argument posits that varying levels of incumbent and insurgent control ' from total to equally contested by both ' predict levels of indiscriminate and selective violence in a civil war. This is, Kalyvas argues, because violence in civil war is a product both of the traditional combatant organizations (who provide the means of violence) and information provided by civilian actors living in their areas of control. In areas of total control, civilians who might theoretically hope to the defect to the other side have little means of access the opposing party, and are effectively forced to collaborate or otherwise accommodate themselves to the ruling power, who has only limited need for coercive violence in the absence of meaningful alternatives. Opponents who lack access to the area are reduced to using indiscriminate violence. In areas of dominant but incomplete control, civilians have the opportunity to defect to the opposing side, but face high risk of selective violence in response. In areas of parity where both sides of the conflict can potentially carry out reprisals, residents attempt to straddle the fence, limiting the provision of information that would allow the combatants to use selective violence.

The book's biggest contribution is in examining, at length, how civilians in war are active participants, not just a field upon which the insurgents and incumbents act. The chapter on denunciations and their motivations was fascinating, and immediately brought to mind many of the cases from Afghanistan recounted in Anand Gopal's book, among other studies. The chapter on the linkages between national-level political actors and cleavages and local-level disputes was also excellent. The point that external conflict actors offer local residents the means to carry out malicious retribution and exacerbate local conflicts may not be novel, but it is thoroughly documented here.

While Kalyvas tests his theory with case studies from the German occupation of Greece and post-occupation civil war period, his coding of the varying levels of control he observers is a rare area where he does not elaborate his reasoning at length (though he acknowledges elsewhere that there are no precisely agreed measurements for territorial control and suggests this as an area for further study). In "Inside Rebellion", which I read prior this, Jacob Weinstein argues that Kalyvas' theory is not an accurate predictor of the use of violence. He instead attributes the use of indiscriminate or selective violence to the organizational structure of insurgent groups and their corresponding dependence on civilian actors, who Kalyvas presumes both sides of a civil war inevitably turn to for local information. I think a plausible case could potentially be made for integrating the two arguments (though I'm not fully sold on the importance Weinstein attaches to economic endowments in directly shaping organizational practices either) but this is an area that I'd need to read further on, since I'm not closely familiar with the case studies they're working from.

Still lots to process from this book; even if not applicable in all cases, it's very much worth the read.