What is state failure? See my conceptualisation of state failure on the right flank below.

Monday, February 25, 2008

Debate: On applied strategy in a tribal society (Part two)

After the long silence of the week-end, I'm returning here with a pair of posts in which I'm happy and honoured to have the chance to publish parts of William S. McCallister's response, with his permission, to a lengthy post of mine that I wrote at the beginning of February. I came up with that blogpost in reaction to a debate between McCallister and Afghanistanica. I will collect some new thoughts to continue this discussion. But I will have to take my time doing that, since, just like in the case of William S. McCallister's earlier piece at the Small Wars Journal's blog (on which Afghanistica commented), while reading his thought-provoking remarks published here, a whole list of important research questions emerged in my mind. And trying to give but a preliminary answer to those will require some thinking through and delving into from my part, or anyone else's for that matter.
To give away something in advance, just consider what questions this discussion may structure about tribal relations in southern Afghanistan as well as the current ISAF/OEF counterinsurgency approach there (while the original discussion took place specifically concerning Pakistan's tribal areas so far, its ambition of universal relevance was never denied).
So, for those of you who have not been following this discussion, here are the earlier contributions to it.
And so below here are some excerpts from William S. McCallister's letter, dated February 22. This is part two as I have cut it. It elaborates specifically on how tribal coordinating messages and operating codes affect refugee movements and the social organisation of life in a refugee camp.
* * * * * *
(from hereon you're reading excerpts from William McCallister's response)
I would like to conclude with addressing the role of tradition or the break-down of tradition and refugees and refugee camps. The first step is to clarify if tradition evolves or can be imposed by a group that decides that it has found a better and in the western sense a more rational way of doing something. I believe this is implied in many arguments that seek to place the Taliban or AQ outside the social system, distinct and separate. When I first read your point concerning the break down of tradition I imagined a distinct break in traditional behavior and a new tradition being introduced. A position that as you already know, I disagree with. The Taliban and AQ are subject to the same simple rules, the DNA by analogy which governs behavior in the social organism and the social system as a whole. The Taliban and AQ may find weakness or gaps to exploit in the near-term but in time, the social organism will respond defensively in the manner that has served it well for so long. The simple rules are in effect.
I am very pleased that you liked how I conceptualized how a handful of Islamists can set themselves up in tribal areas. The point I was making with this conceptualization is that AQ must follow this traditional pattern. It will fail if it does not. Secondly, and more importantly, this pattern is intuitive and not the result of rational planning. The social system will not respond if a different manner of approach is initiated. Please ask yourself why many of our earlier approaches in Iraq failed and to a point continue to fail today. The simple answer is that they do not resonate with the local population and the reason they do not resonate is because they do not take into account the simple rules, the cultural operating codes and coordinating messages that determine, coordinate, regulate and synchronize behavior among the community’s component parts.
I believe that the refugees and refugee camps variable is a valid point but I have to ask myself are refugees and refugee camps a 20th-century phenomena. When the Mongols sacked Baghdad in the mid 13th century and then moved south to sack Basra twice did these military actions result in refugees and refugee camps? How did the social organism respond then and am I correct in assuming that the response was different from today? Has Afghanistanica considered this historical pattern of behavior? Was the existence of a historical pattern of behavior even considered? Was the question even asked if something akin to a historical pattern of behavior in terms of refugee behavior exists? If patterns exist do they differ from those in the 13th century and 21st century? Were refugees susceptible to “revolutionary” thoughts in the 13th century as we imply they are today? The question is akin to asking “how and where does a single thought begin”. There has to be a point in phase-space where numerous variables intersect to create that one single thought. In regard to refugees is the flow the result of the action of one person who decides that he has had enough, pack his bags and depart an area? How does this single action translate into thousands of people on the road? Or, are there a number of people that have the same thought at the same time and act in the same manner i.e. pack their bags and depart the area and contagion takes over? What are the simple rules that govern this behavior? A number of theories such as “concept of elasticity” and “expectancy” play a role in this phenomenon. I spent much time studying how refugee flows are initiated and how they play themselves out. Based on these studies I developed a methodology for military planners to assist in identifying humanitarian assistance options. The goal was to provide a predictive analysis tool, imperfect as it may be, but since it was the only product on the street, the best available at the time. The methodology is captured in “Appendix I (Techniques in Dislocated Civilian Operations) in Army Field Manual 3-05.401, Civil Affairs Tactics, Techniques and Procedures, 23 September 2003”.
Lastly, I’d like to provide a short example of how my cultural operating codes and coordinating messages methodology could be applied; in this case as a baseline study for patterns of behavior in refugee camps in Pakistan. I’ll caveat my pontifications by stating that I have not been to Pakistan nor seen or experienced a Pathan refugee camp. I recommend that you access the Small Wars Journal and look at the COIN and Irregular Warfare in a Tribal Society Primer for a more detailed explanation of the four cultural operating codes and the two coordinating messages. To recap, the codes are shame and honor, segmentation of tribes, patronage and territory. The two coordinating messages are “no stability without us” and “what have you done for me lately, what will you do for me tomorrow?” I will not dwell on the shame and honor operating code and recommend James Bowman’s book “Honor: A History” instead.
A refugee flow is initiated. A head of household decides to move to Pakistan. He and his immediate family pack up and depart the area. Others in the village observe this behavior. I will not delve into the contagion factors of coordination, integration, regulation and synchronization but in time more families join the exodus. Based on the segmentation principle, families, clans, and as more and more persons depart a given area and resettle in a refugee camp will do so based on kinship and alliance relationships. The camp will in time self-organize and form itself into districts and sections i.e. territory based on kin and tribal affiliations. I can now map the camp which in time will grow into a city (another living-breathing organism) and color code the sections or territory by tribal affiliation. The dynamic doesn’t stop for those who remained behind in Afghanistan will, once the decision is made to depart for Pakistan, likely seek out kin and allied tribal groups. The same phenomena occurred in our cities and continues to do so today i.e. “little Italy”, “Chinatown”, etc. The application of the segmentation of tribes operating code has provided a means for structuring the analysis and to develop a baseline understanding of the tribal demographic and territorial control in a given refugee camp. The patronage operating code is the mechanism that binds the groups and facilitates patterns of coordination, integration and synchronization. Individuals and specific families in tribal society are responsible for a given community group as a whole and are therefore obligated to seek access to influence and resources. Kinship ties and patronage relationships are exploited to compete for influence and resources. Patronage relationships will also reveal the existence of political ties above and beyond blood ties. Some leaders are better than others at gaining access to influence and resources vital to “maintain” cohesion within the primary group or to “attract” others to form a greater community group. The segmentation code helps explain why alliances emerge; the patronage code helps explain why they last. Once we have an idea of who controls what territory, existing patronage relationships and alliances, we can begin to develop indicators as to how well the various groups are faring. Is the leadership able to provide for the group or are there gaps and opportunities for others like the Taliban or AQ to exploit by providing “honor, guns and money”. Honor in accordance with the shame and honor code, guns and money in accordance with the patronage code which in time may evolve into conquering “territory” if the traditional leadership is unable to provide for the group. We can now ask the question: who is able to “attract” allies. What access to influence and resources are offered in exchange? How can we counter this initiative? In my opinion the key tenet of COIN or irregular warfare is to effectively communicate intent within the target audience’s cultural frame of reference. If we chose to counter the opponent’s initiative we must do so within the cultural frame of reference. I can now update my color coded map of the refugee camp to reflect this pattern of behavior. I can also use this information to begin targeting outsiders seeking to influence patronage relationships and territory inhabited by segmented groups. I can choose to support one or the other tribe in its ability to “attract” others i.e. increase its patronage building capacity, or assist in defending territory.
(...) we are able to identify that a “refugee camp” mentality is still based on some simple cultural operating codes such as shame and honor, segmentation, patronage and territory and that it really doesn’t matter what (“worldview of the 1990s”, “new order”, Pakistani, Soviet Union or U.S. facilitation) shapes the dynamic but that the dynamic is based on some simple rules. The Taliban and AQ for that matter have to be able to create alliances (segmentation), “attract” (patronage), and control territory. The Taliban and AQ are members of the social system and subject to its rules. AQI adhered to the rules early on in Anbar and was successful. In time, AQI thought it was above the rules and we can see the result of hubris.


Joshua Foust said...

"we are able to identify that a “refugee camp” mentality is still based on some simple cultural operating codes such as shame and honor, segmentation, patronage and territory and that it really doesn’t matter what (“worldview of the 1990s”, “new order”, Pakistani, Soviet Union or U.S. facilitation) shapes the dynamic but that the dynamic is based on some simple rules."

What a clever way of discounting all the exogenous factors that presently make FATA a unique and particularly vexing environment... to say nothing of the historical ones. McCallister is on to something, but he's not quite there yet. Generalizing Pashtun behavior from simple rules reeks of filtering everything through Pashtunwali, which is a mistake.

William S. McCallister said...

Dear Mr. Foust,

I would like to start by placing the Pakistan article into context. The article itself was in response to a question whether the Anbar model would be appropriate in the North-West Frontier. I personally believe that the Anbar model may not apply in total whether in the rest of Iraq or the North-West Frontier. Each situation is different, period. On the other hand I do believe there are components that might be applicable but we must first conduct a very detailed area and target audience analysis before implementing this or any initiative. The intent of the article was to highlight the risks involved in becoming too deeply involved in a kinetic relationship with the Pathans since their social and historical experiences differ so very much from our own. In short: look before you leap.

The article also sought to encourage greater contemplation on whether the target audience would be receptive to our current types of kinetic or non-kinetic forms of communication. Therefore, we seem to agree that the Anbar model as defined by recruiting tribes or "extended families" to the cause is inappropriate for the North-West Frontier.

I am a bit surprised that a number of respondents have focused on disproving the cultural operating codes and coordinating messages model and its universal application; a point I never made nor believed to be true. Please also note that there is a difference between the "Anbar model" and my approach to providing a means to structure the analysis. Note that I do not retreat from my assertion that the codes and coordinating messages model or methodology is applicable in Anbar province. We can continue to debate the specifics of words, definitions, or my dilettante approach to defining human nature but the methodology seems to work for us in the western province of Iraq. Having said the above, we are still faced with developing a workable mental model for our soldiers and Marines operating in the areas under discussion. In my opinion, the critical requirement is to develop an “operationally relevant” mental model.

My target audience is the squad and platoon leader, company commander, Battalion Operations Officer, Battalion, Brigade and Task Force Commander and not academics. The Marines and soldiers I seek to assist are in the people business and when they complete their mission analysis, concepts such as “establish democracy”, “restore stability” or “secure” a given area translate into X amount of trucks, X gallons of fuel, a specific task organization and X amounts of 5.56 and 7.62 mm NATO ball. This in turn translates into MEDVAC frequencies and Quick Reaction Forces (QRF), but you get the picture.

A model in my mind is a simplified construct of the real world. A methodology on the other hand provides the analyst a means to structure the analysis. It provides for a common terminology and vocabulary and hence a baseline for continuous and more detailed study and analysis over time. Please note the phrase "continuous and more detailed study and analysis over time". Since the military is in the execution business a baseline terminology and vocabulary is needed.

Your comment highlighting the clever way I discount all the exogenous factors and historical ones that presently make the FATA a unique and particular vexing environment highlights the difference in our thinking and hence is akin to me discussing apples while having to answer questions about oranges. I've been studying the North-West Frontier of and on since Christmas 1979. Not that this means anything except that I actually appreciate the regions vexing idiosyncrasies.

My simplistic approach for understanding the people with whom we are interacting at the tactical level in the FATA is to study Pashtunwali as a baseline from which to adjust our behavior so as to effectively communicate intent within the target audience’s cultural frame of reference. Please note that I recommend gaining an appreciation for Pashtunwali from which to adjust our behavior. I never made the statement to "filter everything through Pashtunwali" those are your words. I advocate "generalizing" to a degree since I do believe we can all accept the truism that the devil is in the particulars and that we can spend much time on debunking this or that specific statement in what appears to me an attempt to throw out the baby with the bathwater. I only ask how much specificity can we endure before analysis paralysis sets in. A military organization can not afford analysis paralysis. The model is intended to preclude this effect and seeks to provide a mechanism to highlight the causal behavioral processes at work and to recognize patterns of behavior.

While I truly enjoy our discussion I will continue in my attempts to provide soldiers and Marines an operationally relevant methodology to assess the cultural operating environment. All the tribes or as it has been pointed out to me “extended families” I’ve studied, whether located in South Africa, Iraq or in the FATA seem to share a number of basic behavioral traits. These are the cultural operating codes and coordinating messages which appear to provide the framework for the causal processes at work and are repeatedly expressed in patterns of behavior.

What I am not attempting to do is unequivocally state universal truisms such as all leadership in tribes or extended families are based on bloodlines or kinship. Iraqi tribes may forego a given bloodline if the “other” is more able to provide for the good of the group. Fictive bloodlines exist in Anbar province. Nor do I seek to advertise this model or methodology as universally applicable or a silver bullet for any and every case.

In my mind there is a difference between efficiency and effectiveness. The model fits into the same category. If it works; use it. If it doesn’t; adapt it. If you can’t adapt it; discard it. In my mind it is as simple as that and the rationale for my “don’t throw out the baby with the bathwater” comment. That comment of course is based on the assumption that the model or methodology has merit. If on the other hand you believe the model or methodology has no merit whatsoever then we are engaged in a completely different conversation.



Joshua Foust said...

Mr. McCallister,

Thanks for the detailed response. I guess I'm a bit confused, however. You say you don't mean for these kinds of codes or rules of behavior to be generalizable, yet you seem to be pointing to the success of the Anbar model as a reason to generalize them. Am I misreading that, or is there some other dynamic you're reaching for (such as a tribal or sub-tribal understanding of the target population)?

Similarly, I appreciate -- perhaps more than you realize -- the need to target your methodology for the appropriate audience. However, as neither a military agent nor an academic, I also appreciate rigor: an inconsistent model or methodology helps no one in this kind of a context.

In fact, I believe you're talking about a framework (if you're just sticking to common definitions). A methodology would imply a standard set of actions (i.e. a means for inquiry), which could also imply a model -- something I think you disavow in this context, given its propensity for over-generalization. Understanding Pashtunwali is great, but the problem is that Pashtunwali itself is a fluid concept that is inconsistently applied on a situational basis. It can be followed or not followed, recompense can substitute for revenge, and there are no universal or clear rules for which concept applies when. I'm sure you realize this, but I don't understand why, then, you think Pashtunwali is still an appropriate baseline.

To clarify my comments: I didn't mean to accuse you of "filtering" through Pashtunwali, since we both agree and know you didn't say that. I just said the same process "reeks" of it, by which I mean it lends itself to the same analytical pitfalls. Far from "analysis paralysis," this is at the heart of the problem for how the U.S. military and civilian government continue to misunderstand the tribal areas. Thinking Pashtunwali is easily definable beyond generalities strikes me as a trap to be avoided, not a convenient shortcut to be exploited.

I don't intend to be so negative here, because I truly do think you're onto a good idea. I want you to succeed, in other words, but I've just seen similar attempts to generalize behaviors backfire in a major way. Especially in an area like FATA, where even the very idea of identity is fluid, generalizing back to simple rules poses huge risks. Think about which level of identity -- super-tribe, tribe, sub-tribe, division, clan, super-family, family, and so on -- someone chooses to identify himself. That is itself fascinating information (think of the Suleimankhel in the context of the Ghilzai), and can indicate something about where he considers his place in his community. But that level isn't standard, and from village to village people identify themselves according to an inconsistent metric that is as trivial as it is unpredictable.

So I guess what I'm getting at is, in stark contrast to the military's preferred modus operandi, when in particular discussing FATA there needs to be an understanding for a certain (I would argue "great" but couldn't do so comprehensively at the moment) degree of not just uncertainty but inherent unpredictability included in the discussion: These people live here and say they're this, but they could also be that, and for these reasons.

Relying on simple behavioral codes might allow for this sort of flexibility, but you should also realize that in a practical sense (i.e. when utilized on the ground) it probably will not.

That is the pitfall I'm trying to highlight. Otherwise, I'm a big time fan of trying to refine this idea further.


Péter MARTON said...

The above discussion is continuing here: