US strategy in Afghanistan sits upon a precipice. Eight years after the United States invaded the Taliban-controlled portion of Afghanistan in response to the horrific events of September 11, 2001, Washington and its allies find themselves waging a war of unknown character, for unclear purposes, against an ill-defined enemy. As the Obama administration grappled with the Afghan nettle during its first year in office, a plethora of policy papers and commentaries advocating one approach or another appeared. The consensus settled on pursuing a classic counterinsurgency campaign. But problems of fraud in the 2009 Afghan presidential election, the worldwide financial crisis, domestic American political priorities and war weariness, in particular among Democrats, have undermined that consensus. The resulting cacophony of recommendations in official and non-official circles has only underlined the fact that US policy is adrift.
Such listlessness reflects the very real possibility that it may well be too late for the US to salvage “victory” in Afghanistan. In particular, the battle for the “hearts and minds” of the Afghan people, crucial to the success of counterinsurgency as conventionally understood, seems to have been lost. But was this battle the one to join? Many refer to Afghanistan’s history to support their preferred course of action, yet few have detailed knowledge or understanding of it. There are indeed lessons to be drawn from the region’s past, but those lessons do not reflect well on the strategy pursued today.
“Hearts and minds” is a concept deeply intertwined with the fate of European imperialism, specifically that of the British Indian Raj on its famed North-West Frontier. A term first employed along the Baluch marches by Sir Robert Groves Sandeman, “hearts and minds” was shorthand for an approach to governance that sought to coopt rather than coerce its targets. It was a strategy of rule for a cash-strapped empire. As such, it should not be seen today as an all-purpose palliative for the US military’s pains in the contemporary Muslim world. “Hearts and minds,” and more broadly the counterinsurgency doctrine it is a central component of, does not fit the circumstances facing the US and its allies in Afghanistan. It is not simply the application that has been flawed, but the entire concept that is wrong.
The Sandeman System
The oft-quoted phrase “hearts and minds” is almost universally misattributed to Sir Gerald Templer, the British high commissioner in Malaya between 1952 and 1954. Templer is thought to have coined it to describe his counterinsurgency campaign against the Malayan Communist Party, then waging a bloody war along the length of the peninsula. Many strategic thinkers and military personnel, including the authors of the Counterinsurgency Field Manual (2006), believe the rules laid down by Templer to be a golden formula for waging and winning an asymmetric conflict. Yet Templer borrowed the phrase from a colonial administrator of a previous generation.
Writing to a friend in 1891, Sir Robert Groves Sandeman, agent to the governor-general in Baluchistan, argued that “to be successful on this frontier a man has to deal with the hearts and minds of the people and not only with their fears.”  What came to be known as the “Sandeman system” was based on the recruitment of local tribesmen into the edifice of imperial governance. The tribesmen were paid to police the territory they lived in, keeping it safe for imperial communications and limited government by clearing the roads of robbers and providing a mail service between military garrisons. Tribal elders selected members of the tribal militias, their authority and powers of patronage thus reinforced by imperial backing. The monies paid to the tribesmen for these services supported the frontier tribes. But it also made them vulnerable. The state acquired a “bond for the good conduct of all” of the tribe’s members, a colonial official observed, “as the ill behavior of any portion of the tribes will cause the entertainment of all the members of that tribe being immediately annulled.”  The British hoped exposure to imperial service would have a “civilizing” influence, with cash subsidies binding the tribes to the colonial political and economic order.
Sandeman’s system defined debates about how to rule what the British saw as the “wild” frontiers of empire until the end days of the Raj. But his influence was not limited to British India. His ideas permeated Lord Lugard’s actions in Nigeria in the early twentieth century, as well as the doctrines adhered to in French possessions from Morocco to Syria and Indochina. They shaped not only the form and substance of imperial rule, but also the content of imperial retreat. Sir Harold Briggs, author of the “Briggs plan,” the British counterinsurgency strategy in Malaya, had served in Baluchistan during the 1930s with C. E. Bruce, son of R. I. Bruce, Sandeman’s first assistant. Thus, when Templer spoke of winning “hearts and minds” in Malaya, he was echoing Robert Sandeman’s voice.
Sandeman’s system apparently has regained favor among US and NATO forces in Afghanistan, although today’s adherents are likely unaware of its origins.3 The “Sunni Awakening” in Iraq is considered a possible model for a tribal resurgence in Afghanistan, and both the British and Hamid Karzai’s government have pushed for the creation of networks of tribal militias, known as arbakai. But such a policy is problematic at best. The Sandeman system was designed to govern the colonial frontier as cheaply as possible, to keep the peace without the expense of military expeditions. In this aim, Sandeman and his system were largely successful. Baluchistan, where the system was used from the mid-1870s until British withdrawal in 1947, lacked the violence common along the easterly and northerly reaches of the Afghan frontier and was considered a quiet backwater of imperial administration.
Rather than governing the tribes directly and risking resistance, Sandeman established a system of indirect rule. He looked to the tribes’ “natural leaders” and “traditions” to ensure their quiescence and mediate their relationship with the colonial state. But it was Sandeman and like-minded frontier administrators who “knew the tribes,” who identified, endorsed and patronized these “traditional” institutions and leaders. Thus Sandeman’s system was based on a circularity of colonial administrative logic whereby tribesmen were ruled indirectly by “traditions” validated by the colonial state.
The effect of this logic was to isolate the tribes from the pressures of the modernizing state and instead encapsulate them in a colonially regulated version of “tradition.” Ruling through tribal elders meant that indigenous hierarchies of power were erected in lieu of imperial ones in order to govern the tribes cheaply. But neither Sandeman nor subsequent administrators anticipated the consequence of this system, namely the exclusion of the tribes from the colonial social and political order. On the contrary, Sandeman understood his system as part of the colonial “civilizing project,” believing it would facilitate the political subjugation of the tribes and their cultural education. In keeping with contemporaneous ideas of political economy, the tribal subsidy was seen as key, creating a cash-based economy linked with the rest of British India. Once ideas of property and productivity were introduced in the form of tribal subsidies, the inexorable pull of money would draw the tribesmen into the webs of colonial society. The tribes would be integrated into modern social and state structures through a kind of cultural osmosis, rendering them into pliant colonial subjects. So the theory went.
The subsidies rendered the tribesmen dependent on, but not integrated into, the colonial order. The tribesmen were instead cut off and isolated. Rather than being transformed into “civilized subjects” of the colonial state, they became its wild bastards, outcasts who were generally ignored and always kept at arm’s length. While this policy may have served imperial objectives, it has proven a disaster for the tribesmen themselves, as well as the successor states to the Raj, notably Pakistan. Many of the inhabitants of Pakistan’s Northwest Frontier Province remain outside the political space of the nation-state, inhabiting the legal no man’s land of the Federally Administered Tribal Areas. Scarcely recognized as citizens or subjects by the state, they can only play a disruptive role.
The US and its allies are replicating much of Sandeman’s system, albeit unwittingly. Yet they are doing so with substantially dissimilar aims and expecting considerably different outcomes. The outcomes, however, will likely be the same as those Sandeman achieved. Sandeman sought to buy political order on the cheap by governing the tribesmen through their “traditions.” But doing so did not integrate them into the colonial sphere as hoped; rather it excluded them from it. It is hard to see how US reliance on similar methods, working through “natural” local leaders and “traditions,” will do anything but alienate the tribes and thus presage the “failure” of the state.
Evidence from the southeast indicates that where “tribal” institutions and leadership remain strong, such as among the Shinwaris, the Taliban insurgency is correspondingly weak. The Americans are attempting to tap into those resilient “tribes” by paying them to fight the Taliban in their locality, much as Sandeman did previously. By recruiting tribesmen into informal militias as an addendum to the Afghan National Army and other state security forces, the US hopes to win the war against the Taliban. While they may well prevail in combat, such a strategy will not win the peace. Rather than coopting these tribes into state structures, such efforts both politically validate and give practical sanction to their existence outside those structures, ultimately weakening state authority.
To execute this strategy, the US has invested in a number of programs designed to give it “cultural knowledge” of the battlespace. The most well known of these programs is the Human Terrain System, which deploys social scientists with combat units who are supposed to “know the country” as Sandeman and his fellow imperial administrators presumed to. Yet such programs leave unaddressed fundamental disconnects between US-led efforts at state stabilization and the local experience of Afghans. For example, it is a common refrain to hear Afghans insist that the US actually supports the Taliban. While many allied commanders find such a belief incomprehensible, if not laughable, it is quite understandable. The Taliban has demonstrated its ability to fill quickly the vacuum of authority left in the wake of clearing operations undertaken by the US and its allies. Locals see NATO troops destroying, often unwittingly, local systems of authority through their operations, leaving the Afghans defenseless against the Taliban in the midst of NATO withdrawal. Thus the US and the Taliban must be secretly allied. Until the US both understands and addresses these types of concerns, American efforts will very likely come to naught.
A Lesson in Contrasts
Despite the popularity of the term “hearts and minds” with the Pentagon, Sandeman’s system was not a counterinsurgency doctrine, in the contemporary sense of prosecuting a war against irregular combatants. Rather, it was about preempting and thus preventing insurgency by enlisting tribesmen in the service of the colonial state.  Winning “hearts and minds,” however, is essential whether the goal is to defeat the rebellious “natives” or to get them to police themselves. Interaction with the local population was crucial because, in the words of one of Sandeman’s admirers, “To be ‘accessible’ means a great deal.”  In a sense, the Sandeman system and counterinsurgency doctrine are two points on a continuum of governance that corresponds to a sliding scale of violence. War is, after all, politics by other means.
Both the Sandeman system and current American strategy in Afghanistan seek to “indigenize” security and political elements. The Afghan National Army and tribal militias are manifestations of the US desire to recruit Afghans into the security apparatus at the same time that a political concordance with “traditional” leaders is forged. This rehash of Nixon’s Vietnamization is a recycling of British colonial strategies of collaboration.
Counterinsurgency critically relies on a permissive environment, an element almost universally ignored by theorists, practitioners and proponents. Historically speaking, that environment takes one of two forms — either a colonial or a domestically authoritarian political environment. No foreign power has fought a successful counterinsurgency save in a colonial or neo-colonial environment. Those national governments that have successfully prosecuted counterinsurgencies have been decidedly undemocratic. The obvious and oft-cited examples clearly bear this out. Malaya, Kenya, Yemen, Indonesia, and the Philippines all fall into one of these two categories, if not both. The current lull in violence in Iraq is hardly a success, except in that it has met the short-term US aim of being able to stabilize the situation in order to declare victory and withdraw, regardless of Iraq’s ultimate political stability.
The US, however, does not seek to perpetuate its own rule in Afghanistan. Washington’s stated aim is to build a stable and centralized Afghan state that cannot be used as a “safe haven” for the likes of Osama bin Laden. To facilitate this aim, the US has sought to develop “increasingly self-reliant Afghan security forces,” in the words of a 2009 inter-agency white paper on Afghanistan policy. It also remains publicly committed to building a formally democratic state, despite the controversy surrounding the 2009 presidential election. Afghanistan is thus neither an authoritarian nor a colonial environment de jure, though some may argue it is both de facto.
Even if one were to label US efforts “imperial,” there is an important distinction between colonial and imperial that carries implications for counterinsurgency campaigns. As the entrenched, paramount power, a colonial regime is a central part of the established order and political landscape. It is plugged into local information networks and has deep ties of patronage through which it draws upon a collaborating elite. It “knows the country” and uses that knowledge to embed itself into the colonized society while simultaneously attempting to reshape that society.
Colonial regimes relied on alliances with local holders of authority. These were central to British “successes” in places like Malaya and Kenya, and, conversely, to French failures in Algeria and Indochina. Those cooperating with the British and the French in these places knew that the colonial state had a vested interest in the outcome of counterinsurgency efforts. In Kenya, there were white highland settlers; in Malaya, rubber plantation owners agitated for action. The pied noirs of Algeria facilitated the downfall of the Fourth Republic with their refusal to compromise. There is no comparable constituency driving US actions in Afghanistan today; rather US interests are almost exclusively strategic. As a consequence, long-term American investment, of both political and financial capital, has been meager by the standards of state building. No matter what the Obama administration says, local actors know they simply need to bide their time, for the US and NATO forces will eventually leave. The previous US abandonment of Afghanistan, as well as the withdrawal from Iraq, reinforces this sentiment. Simply put, there is no constituency of pied noirs to force this fight in Washington.
In contrast to colonial regimes, imperial ones lack the roots and interests in local society that give the former staying power. At best, an imperial power may have networks of patronage or information through which it can exert its influence. The US, however, does not enjoy these advantages in Afghanistan. Ties to local networks of patronage are shallow and weak. American influence largely derives from the money it deploys as a positive sanction, and the force it deploys as a negative one, but neither effort has been sustained over the eight years of the US military deployment. With only a limited, and diminishing, ability to coopt a collaborative elite, and little knowledge of local society, the US has fewer resources with which to fight a successful counterinsurgency than did the colonial powers it emulates.
American attempts to acquire knowledge about the country are, in the main, ineffectual and misguided. The lack of language training for military forces, a consequence of the relatively short rotations in country, is a glaring handicap. The Defense Department’s Human Terrain Systems program is also of questionable efficacy. The program tends not to attract specialists in the languages and cultures of the region, many of whom are skeptical.  More importantly, the program has a utilitarian bent that fundamentally compromises its value. It is designed to make local knowledge readily accessible to ground commanders, ostensibly offering commanders information they may fit into their pre-existing worldviews and tactical plans. The program is not designed to challenge the commanders’ preconceived notions. This facile effort at “knowing the country” fails to recognize that the knowledge obtained will be highly contrived. Like Sandeman, who constructed a “tribal” reality of “tradition” which fit within his imaginative limits, current US efforts produce a reality that accords with pre-existing assumptions, constructing a hopelessly “tribal,” inherently violent and backward society.
The hitherto lackluster counterinsurgency efforts in Afghanistan have masked the fact the US and its allies are involved in a bloody and costly war. Between 2001 and the end of the 2009 fiscal year, the US had spent $228 billion on the conflict. At present, it is spending roughly $11 billion a month to cover the cost of Gen. Stanley McChrystal’s surge. Over 1,700 allied soldiers have been killed, and an unknown, but considerably larger number of Afghans have been as well, not to mention the refugees created by the ongoing fighting.
By dubbing its Afghan operation a counterinsurgency, the US has severely circumscribed its freedom of action. It has created a rhetorical trap for itself, with the Obama administration simultaneously talking about “getting serious” in Afghanistan, while at the same time indicating that the engagement is limited, written on a postdated check. The NATO secretary-general, Anders Fogh Rasmussen, captured the seeming paradox, if not the glaring contradiction, in a speech at the Atlantic Council on September 28, 2009. There he stated that the US would not bear the burden in Afghanistan alone, but also noted “this mission…cannot continue forever.” War is a bloody thing, and the Iraq debacle has largely soured the American public on the proposition — more so if the proclaimed aim is to establish a representative Afghanistan governed by the rule of law and respect for human rights. There is a growing belief among US allies, in Congress and even within the Obama administration itself, that the costs of the mission in Afghanistan outweigh the commitment, and may exceed the return.
A Dirty Little War
The United States now contemplates a Gordian knot in Afghanistan, but one it helped to tie. In the wake of the displacement of the Taliban at the end of 2001, the US had the opportunity to reshape the Afghan political landscape, largely with the support of the war-weary Afghan populace. Instead of disarming the militia commanders who had ravaged the country for 20 years, however, the US saw governing through them as the cheapest option. Winning the peace was not the same as winning the war, a lesson ignored at the time. So now the US presides over a fractured political and security environment with a weak regime dependent on its patronage and an “insurgency” fusing together little understood elements of a supposedly globalized Islamist militancy with local grievances. The conflict in Afghanistan has become one of America’s dirty little colonial wars.
The US hopes to spark a “tribal resistance” to counter the growing threat it faces in Afghanistan. Advocates point to the “Anbar Awakening” in Iraq, but ignore the specter of increasing sectarian violence that is proving one of its effects. Further, there is little effort to relate such short-term tactics to long-term strategic goals. A “tribal resurgence” may indeed stand firm against the “Taliban,” but how will arming the “tribes” affect the fortunes of the central state? Sandeman’s experiences along the Baluch frontier stand as a warning of the unintended but highly likely consequences of such a strategy.
One may justifiably doubt the success of any US strategy at this point. Critics contend that Afghanistan is the “graveyard of empires,” without recognizing the irony that the country is a wholly imperial construct. “Victory” is bound to remain illusory as long as it remains undefined, or at best poorly defined. Perhaps it is a matter of mitigation of harm, founded upon the formal declaration of victory while acknowledging its substantive abandonment. What is clear is that US efforts will likely come to the same conclusion as their historical antecedents. Counterinsurgency may be common currency in Washington, but the hard price of its failings will continue to be paid by American soldiers and Afghan citizens on the ground.
 T. H. Thornton, Colonel Sir Robert Sandeman: His Life and Work on Our Indian Frontier (London: John Murray, 1895). Letter dated April 19, 1891, title page.
 Commissioner of Sind to Governor of Bombay, June 17, 1870, Political Department, no. 772, vol. 132, Maharashtra State Archives, Mumbai.
 See, for instance, Greg Bruno, A Tribal Strategy for Afghanistan (New York: Council on Foreign Relations, November 2008).
 I thank Dane Kennedy for this observation.
 Davidson to Bruce, August 21, 1901, Bruce Collection, India Office Records, London, Mss.Eur.F163/1.
 The American Anthropological Association opposes the program. See the statement of its executive board, dated October 31, 2007, and available online at: http://www.aaanet.org/about/Policies/statements/Human-Terrain-System-Statement.cfm.