American military planners hope alliances with Afghan tribes can save the war effort. But the reality is far more complex, Joshua Foust writes. Assessing the American effort in Afghanistan would discourage even the most stubborn optimist. Eight years into the war, soldiers are dying in record numbers; the insurgency has built up a "shadow" government in 33 of the country's 34 provinces; thousands of civilians are killed every year; and the Afghan government seems to invent new ways to demonstrate its appalling corruption on a daily basis.
In the 48 months since the return of the Taliban, American planners have announced a series of new plans intended to reverse their momentum, without much evident success The latest of these initiatives - what the New York Times grandly called "America's New Hope" in Afghanistan - is an attempt to bribe local tribes into battling the insurgency alongside American forces. Despite public declarations that they are doing no such thing, the US policy establishment clearly hopes to duplicate the Sunni Awakening in Afghanistan.
In an effort to showcase the Afghan version of an Iraqi grassroots rebellion, US military commanders in far eastern Nangarhar province announced at the end of January that they would support one tribe, the Shinwari, in its fight against the Taliban, pledging $1 million in development aid for tribal leaders in exchange for their allegiance. Unlike the Iraqis in Anbar, however, the Shinwari do not support the central government - suggesting that this "tribal engagement" and others like it may have detrimental effects on the legitimacy and stability of the administration in Kabul.
But the Shinwari pact is merely the most recent of America's attempts to exploit Afghanistan's tribes. There is a widespread belief in the military, policy community, and public that an intimate understanding of Afghanistan requires an intimate understanding of its tribes, and the US has tried to work with tribes since the earliest days of the invasion in 2001 - selecting charismatic leaders to lead their men into generously-funded battle. Despite a UN-led effort to demobilise these non-government militia groups, over the last eight years the US has continued small-scale experiments with what they now call community or local defence initiatives - to almost no effect. In many areas, especially in the south, these programmes have actually served to enrich and re-arm various Taliban groups.
Despite such a rich history of failure, one still finds a common idea in the testimonies, strategy papers and briefings of the policymakers in charge of America's Afghanistan strategy: Afghanistan is a "tribal society", and the exploitation of those tribal ties is the key to fighting the insurgency. Practically every pundit, soldier and official repeats this as an article of faith, to the point where it has strayed into tautology. Because the Taliban is Pashtun, and because Pashtuns are tribal, we therefore must understand the tribes to defeat the Taliban.
It is one of the most frustrating assertions about Afghanistan, directly contradicted by decades of academic research. Let us lay aside for the moment the fact that "tribe" in Afghanistan is a notoriously fuzzy and imprecise concept (in 1983, the anthropologist Richard Tapper wrote that "tribe - has almost ceased to be of analytical or comparative value" in Afghanistan). Scholars studying the country agree unanimously that Afghans do not identify themselves primarily along lines of tribal affiliation or make decisions (like whether to join the insurgency) on the basis of tribe. More importantly, the Taliban itself is organised along non-tribal lines, which means there is no readymade coalition of "anti-Taliban" tribes to be arrayed against them.
As Antonio Giustozzi documented quite thoroughly in his 2007 book Koran, Kalashnikov and Laptop, after 2001 the Taliban gained strength so quickly by exploiting conflicts between communities, which were only sometimes vaguely relevant to the question of tribe. Often, members of a local Taliban cell contain people from all the tribes in a given area. Even among the known senior leadership of the Taliban there are members of each major grouping of tribes - the better to increase the movement's broad appeal. This phenomenon is not a new one, and it predates the defeat of the Taliban in 2001: in the appendix to Ahmed Rashid's book Taliban, published in 2000, he lists the tribal affiliations of the movement's then-leaders - and they come from a wide range of tribal groups.
Misunderstanding the concept of tribe in Afghanistan highlights a critical, and unspoken, weakness of any attempt to turn tribal ties to the advantage of the United States: how does one do it? What do we learn from identifying an Afghan's tribe? Nangarhar's governor, Gul Agha Sherzai, is a Barakzai from Kandahar. Does that tell us anything about what decisions he will make in his province, 350 miles away from his tribe? Jalalulddin Haqqani, a leader of the insurgency, is a member of the Zadran tribe - but so is Pacha Khan, a prominent member of parliament. Which one is more representative of his tribe? And which one would other members of that same tribe be more likely to support?
All this talk of tribal engagement ignores the uncomfortable fact that tribe matters less and less as a factor in a typical Afghan's decision-making process. In his 1995 book, Fragmentation in Afghanistan, the noted scholar Barnett Rubin observed that "the biggest impact of the international system on local power structures in the early 1980s was the penetration of village and tribal society for the first time by political parties". That penetration disrupted the traditional tribal and local authorities, forcing communities to rely on the tanzims (political-religious parties that operated mujahidin militias from Peshawar) for revenue and contacts with the outside world.
The end result of this rapid and severe reordering of Afghan society in the 1980s was the domination of politically-motivated mullahs and militia commanders, and the diminishment of traditional local and tribal authorities. Previous Afghan rulers were able to "govern" the Pashtuns by exploiting tribal divisions, whether undermining individual leaders through tribal division or simply enacting harsh rules that held tribal leaders responsible for their community's actions. (This is one reason there are small Pashtun communities across the generally non-Pashtun north of the country - they were internally exiled for resisting the king.)
The current structure of the Taliban, as a de-tribalised Islamist resistance movement, means that normal methods of working within the tribal system are far less effective, if at all. It is why these initiatives to bribe tribes to fight for America fail so reliably - they just don't apply to how and why people are choosing affiliations and making decisions. Since the Taliban is a movement that is inclusive of traditionally rivalrous tribes, even rivalrous ethnicities, that rivalry cannot be exploited to undo the movement.
Which brings us back to why we even bother discussing this in the first place: there is still the pervasive sense in the foreign policy community that a tribal approach to counterinsurgency - as in Anbar - is the best way to fight the Taliban in Afghanistan. Such an approach could not be less effective if we designed it to be: social realities in the two countries are different enough to make analogy all but impossible.
The curious inability of policymakers to view Afghanistan in its own context - choosing to see it as it is, rather than what they wish it to be - reveals a worrying artifact of American policy. The political and military leadership of the war have eschewed a difficult, complex view of the country (despite always proclaiming that they understand its complexity), and seem intent on relying on assumptions and quick fixes instead. Until that mindset changes, until there is an appreciation for the unique features of Afghanistan's social fabric, we can expect the war to continue to unravel before our eyes.
Joshua Foust is a military analyst. He blogs about Central Asia and the Causasus at registan.net. The views expressed here are his alone.