Normally (by ambition) I like to try and come up here with more practical analysis, but I do feel the inclination every now and then to post insights relevant rather in the internal world of the ivory tower. This is one of those occasions.
Given Nuristan's prominence in the news recently as a rationally abandoned, sparsely populated area (or a liberated area from the Taliban's perspective), I decided to read Richard Strand's study, Ethnic Competition and Tribal Schism in Eastern Nuristan (from 1978), thanks to Christian who highlighted it.
It is brief, fascinating reading, and I very much recommend it to everyone, but not really because it would point you out the sort of population faultline that can be exploited in the spirit of divide-and-rule COIN, for the marginalisation of all anti-government elements, all of a sudden. Rather because it doesn't: at least definitely not in a direct sense. Like I said, this post will be relevant from that ivory-tower perspective.
So, in six pages there, you get a picture of conflict in the Landay Sin valley over pasture use between Koms on the one hand, and Gujars and Meswanis on the other, the latter two referred to by Koms as "Gujirbandevol." It was a conflict in which the resource of state authority was called on even, on various levels, and it was complicated further by the presence of an entire "Gujar faction" among the Koms that showed more of a readiness to accept Gujirbandevols' access to certain lands, because of a number of reasons. These reasons included intercommunity marriages creating cross-faultline connections, economic interests, and also, interestingly, the power of the concept of being "Afghan" from which an acceptance of community between Koms and Gujirbandevol can stem. Highlighting the latter factor, Richard Strand demonstrates that the conflict is partly a modernisation-related one. His study could be one of many lessons in history to those who think that "Afghans" are these revenge-making, xenophobic, war-like beings who would never change.
As to change, fast-forward time: nowadays different things are the key determinants of armed conflict in the area. Land disputes are not gone, of course, and related issues of community formation could be as complex a subject to study today as ever. Yet today the reason why anthropologists are most likely to go this place embedded with the U.S. military is a new conflict involving radicalised Muslim militants intent on waging jihad against the superpower that once used to fund them. That, in a place formerly referred to by Muslims as Kafiristan...
In Peace and Conflict Studies, there is a commonly used scheme of the process of conflict as it leads from a more peaceful state of "social change" (usually constantly eroded by latent conflicts of all kinds) through escalation to all-out war, and how conflict resolution can bring about a transcendence of the original critical conflict starting from the dire state of warfare, leading back to the state of "social change" (like I said, never really free of conflict). It is a result of this, as a result of conflict dynamics, that the interaction of whole new actors and whole new incompatibilities may be constituting conflict in the same place as time progresses. Nuristan is a particularly spectacular example of this.
If this post is to have any policy relevance, it would be that one needs to get down and study hard. And produce policy-relevant posts only when there is really, surely, the spark of an intelligent idea. Too bad this standard is just practically impossible to enforce. Including for me, on myself.