Sunday Book Review: Innovation, Transformation, and War

James Russell's Innovation, Transformation, and War tracks doctrinal innovation in the US Army prior to the Surge. Russell, an associate professor at the Naval Postgraduate School, wrote the book based on face-to-face interviews of personnel operating in Iraq between 2005 and 2007, and close attention to primary documents associated with the development of counter-insurgency on the ground in Iraq. Russell gives what is essentially a bottom up account of doctrinal innovation prior to the Surge. Networks of officers and NCOs shared information about tactics that worked across units, leading to general increase in the institutional capacity of units to conduct the Surge. Russell's work runs orthogonal to many accounts of the Surge, which focus on top-down doctrinal innovation coming from the group of officers and civilian experts closely associated with David Petraeus and FM 3-24. We read Russell as part of my Spring 2011 Counter-Insurgency course at the Patterson School of Diplomacy and International Commerce.

Russell's unit of analysis is the brigade combat team, and he tracks the development of certain tactics across several brigades in Anbar and Nineva. These tactics include what have come to be known as COIN best practices, including census taking, engagement with local leadership, and maintenance of presence amongst the population. Russell finds that these best practices spread through low level interaction between officers and NCOs from units participating in the fight.

But here's the problem: Good social science typically needs some variation on the dependent variable. This is to say that in order to determine what causes learning, we need some examples of units that have learned, and some examples of units that have not. Russell, as far as I can tell, gives no account of units that have failed to learn. As Russell writes:

All the units studied here demonstrated a grasp of COIN principles and sought to build organizational competencies that could be operationalized in plans and operations. Senior leaders in each of the units studied here universally recognized that the ultimate objective in their operations was to secure the support of the local population and isolate the insurgents from local support.

One reason for this may be that Russell worked closely with many soldiers and marines, and was reluctant to suggest that any particular unit had failed to learn. Identification of the researcher with the subject is a problem that invariably affects social science projects that involve face-to-face interaction. But another possibility goes to the core of Russell's argument. When I'm presented with a situation in which there is no variation across units, I'm inclined to wonder whether a unit based (that is, bottom up) explanation is really necessary. We normally focus on unit level processes (and I use "unit" in the social science sense rather than in the "brigade" sense) when we see lots of variation across units. When all of the units act similarly, social scientist are inclined to look for a common cause. In this case, it's not clear that the development of counter-insurgency was really "bottom-up" as described by Russell. If it were so, we'd probably see variation across brigades, rather than uniformity. The uniformity that Russell describes is an indicator that the units all share a common environment (potentially a top-down oriented learning environment) that focuses their learning around certain goals. Networks may function, but they function as part of a hierarchical system that supports the learning of certain tactics, rather than others.

Long story short, while Russell argues that he's presenting a bottom-up doctrinal case, it seems to me that his evidence is quite consistent with a top-down account, albeit one rich in process-oriented detail. There's a lot at stake here, and not just for counter-insurgency; whenever any military organization recognizes a need to shift course, it's crucial to understand the precise nature through which doctrine changes. The bottom-up vs. top-down distinction is one of the arguments common to the literature, along with the civilian vs. military distinction. It's not clear to me that Russell is on the side he thinks he's on.

Nevertheless, the book was extremely helpful in the context of my counter-insurgency course. On the day we read Russell, Colonel James Crider (who recently took command of the 1st Brigade, 3rd Infantry Division) spoke to the class about his experiences in Iraq, and the development of counter-insurgency theory. We also read Colonel Crider's CNAS monograph on learning during the Surge, which supported Russell's argument. We also had the pleasure of Pete Mansoor, who discussed his book Brigade Commander at Sunset the week after Crider spoke. This allowed us to compare Mansoor's practices in 2003-4 with Crider's during the Surge, which was an exceptional informative exercise.

See also Frank Hoffman's review.