Out of the Mountains

by David Kilcullen

  Max Weber, of course, famously defined the state as a political organization that “upholds a claim to a monopoly of the legitimate use of physical force in the enforcement of its order.”37 We might reformulate this, in the context of our discussion of competitive control, to say that a government is a political organization that has successfully outcompeted its rivals across the full coercion-persuasion spectrum, allowing it to establish an uncontested normative system over a given population or territory.

  The coercive end of the spectrum is critical, because it supports and enables the rest of the system: the persuasive and administrative parts of a normative system work (as Weber noted) only because they rest on the ultimate sanction of force, which a dominant actor can apply against those who break its rules. From the insurgent standpoint, Mao Zedong explained this fundamental truth in 1938:

  Every Communist must grasp the truth, “Political power grows out of the barrel of a gun.” Our principle is that the Party commands the gun, and the gun must never be allowed to command the Party. Yet, [by] having guns, we can create Party organizations . . . We can also create cadres, create schools, create culture, create mass movements. Everything in Yenan has been created by having guns. All things grow out of the barrel of a gun.38

  Clearly, however, the resilience of an armed group also depends on the capabilities it can bring to bear across the full spectrum of a competitive control system. As we’ll see shortly, groups that can only apply coercion may achieve temporary dominance over a population. But their control will be brittle, lacking resilience, because it depends on fear alone—in normative systems terms, it covers only a narrow band of the persuasion-coercion spectrum. A purely coercive actor can cast a spell of fear over a population, but as soon as this spell is broken, the population will turn on its tormentor with incredible speed and violence. By contrast, a group that applies a range of coercive, administrative, and persuasive means has a much stronger and more resilient control system. Such a group can respond to a setback in one part of the spectrum by increasing its efforts in another, and can therefore maintain greater and more flexible control over time.

  As we saw in the Afghan example with which we began, the creation of safe behavioral space, as part of a wide-spectrum normative system, has an attraction effect on an at-risk population, who tend to flock to it, drawn by the persuasive inducements and administrative benefits of the system, as well as by the fear of what may happen to them in the unsafe space outside it. This, I suspect, is one of the primary mechanisms for the support-follows-strength pattern that Stathis Kalyvas observed. Once in the system, however, people are corralled and prevented from leaving by the threat of coercion. We might call this the “fish trap effect,” since it induces people to enter a system which they then find extremely difficult and painful to leave. Thus, Fall’s “competitive system of control” can be seen as just one type of normative system, and its application to insurgencies as but one example among many, in which state and nonstate armed groups of all kinds compete to control population groups. To illustrate this in a way that might be closer to home for many readers, let’s consider the everyday example of road rules in a large city.

  Rules of the Road as a Normative System

  Think about the last time you drove your car in a large urban area. You may have been driving to work, or going into the city for a meeting or to go shopping. As you got behind the steering wheel of your car and drove onto the road, you entered a violent, dangerous, and unpredictable environment—across the United States in 2010, for example, roughly 33,000 people died in road traffic accidents; someone was killed on average every sixteen minutes.39 What makes you willing to commit yourself and your family to such a risky activity on a daily basis? Among other things, it seems to me, one reason is that the rules of the road give you a degree of order and predictability, and this sense of predictability gives you the confidence to function in a dangerous environment (with, perhaps, little conscious perception of risk).

  This is because, thanks to the rules, the driving environment—though it’s undoubtedly dangerous—is far from chaotic. There’s a designated side of the road on which all vehicles have to drive. There’s an approved speed limit. There are road signs, in a standardized format, that warn of hazards and prompt certain key behaviors—braking, yielding, or stopping. There are traffic lights that regulate intersections, and lines marked on the road that ensure each vehicle keeps within its own lane. There are television, radio, and billboard advertisements that publicize these rules, seeking to persuade motorists of the benefits of following them, and warning of the consequences (death and injury, speeding fines, loss of license) that follow from breaking the rules.

  And sitting behind this system, underpinning it though often not directly visible, is a government, with a traffic authority or transportation department that sets the rules, a police force or highway patrol that enforces them, a court system that tries those who violate the rules, a system of fines and penalties, and, ultimately, a prison system. People follow the rules (norms) for a variety of reasons—because they fear the punishments (sanctions) that correlate with breaking the rules, because they’re persuaded of the risks of speeding or driving drunk, because they fear public opinion (the embarrassment of having to catch the bus because of a suspended license, or a spouse’s disapproval because of speeding fines), or because they value the ease and convenience of efficient road transport that would be impossible without rules to tame the chaos. Even in a less formal system of road traffic control, where some of these visual cues and formal traffic control measures (signs, lane markings, and so on) may be missing, the cooperative behavior of motorists moving on a busy road is underpinned by the presumed or actual presence of an enforcer, in the form of police or a locally dominant group.

  What we’re describing here is of course a normative system, one that’s owned by the government, enforced by the police, and seeks to control the population of road users within a given territory. This normative system embodies rules and sanctions that create a safe area of behavioral space within which people can go about their business with a basic expectation of safety, and indeed without a great deal of conscious thought. This barely noticed system of control, however, rests ultimately on the power of an armed actor—the police—and on the coercive sanction of the courts and the correctional system.

  It seems to me that this is probably what Joseph Conrad meant in his novel Heart of Darkness when he described the ordinary citizen as living within a control system of which he or she is barely aware, with the police at one end of the spectrum and public opinion at the other. Kurtz, the novel’s antihero, finds himself wielding immense power, alone and unsupervised, in a jungle outpost:

  He had taken a high seat amongst the devils of the land—I mean literally. You can’t understand. How could you? With solid pavement under your feet, surrounded by kind neighbours ready to cheer you or to fall on you, stepping delicately between the butcher and the policeman, in the holy terror of scandal and gallows and lunatic asylums—how can you imagine what particular region of the first ages a man’s untrammelled feet may take him into by the way of solitude—utter solitude without a policeman—by the way of silence—utter silence, where no warning voice of a kind neighbour can be heard whispering of public opinion?40

  Heart of Darkness is, of course, fiction. But most analysts agree that it closely parallels Conrad’s real-world experience as captain of a river steamer, the Roi des Belges, in the Congo in the 1890s.41 In an earlier piece, written soon after his return from Africa, Conrad made a similar point:

  Few men realize that their life, the very essence of their character, their capabilities and their audacities, are only the expression of their belief in the safety of their surroundings. The courage, the composure, the confidence; the emotions and principles; every great and every insignificant thought belongs not to the individual but to the crowd: to the crowd that believes blindly in t
he irresistible force of its institutions and of its morals, in the power of its police and of its opinion.42

  And here lies another critically important point: as Conrad understood, it’s the predictability inherent in the existence of rules, publicly known and consistently enforced, not the content of the rules themselves, far less the popularity of a given government, that creates the feeling of safety that allows a normative system to function. Predictability, not popularity, is the key: you don’t need to like the police or agree with the speed limit for the road rules to make you feel safe.

  As long as people have a well-founded expectation that the police will consistently enforce the rules—that those who break the rules will be punished, while those who obey have nothing to fear—then the government’s popularity or otherwise is completely irrelevant. We saw this in Tivoli Gardens, where many people supported the local don because of the perception of predictable security that the System generated, while the don had a strong incentive to be consistent, fair, and predictable in enforcing his rule. The same set of incentives is at work in the Taliban court system described earlier. More broadly, in the kinds of normative systems that Bernard Fall, Stathis Kalyvas, and Joseph Conrad describe, people don’t have to support a group’s ideology in order to follow its rules. They do so because of the predictability the normative system creates, whether they like its owner or not, or indeed whether they’re even fully aware of the system or not.

  Al Q aeda in Iraq: Brutal and Brittle

  We noted earlier that a group that establishes a wider range of capabilities, covering more of the spectrum from persuasion to coercion, will be stronger and more resilient. This is worth exploring in some detail, by comparing narrow-spectrum to wide-spectrum groups.

  Al Q aeda in Iraq (AQ I) is a good example of a narrow-spectrum group that mainly applied coercion. Early in the Iraq war, AQ I cells moved into Sunni neighborhoods and established control over the population through acts of terror. They applied restrictive rules—banning smoking or the playing of music, prohibiting any kind of cooperation with the government of Iraq or the occupation forces, enforcing the strictest imaginable codes of Islamic dress and behavior upon both men and women, forbidding people from listening to tapes of moderate imams’ speeches, seizing control over economic activity in the district, and so on—and punished anyone who broke these rules in a brutally harsh, violent, and public manner. Thus, although they did establish a normative system (rules and sanctions) they focused their efforts entirely at the coercive, violent end of the spectrum.

  It wasn’t uncommon, for example, in towns such as Ramadi and Tal Afar, for the bodies of local civilians to turn up in the street with the two first fingers of the right hand cut off (a punishment for smoking); for non-AQ I religious leaders to be assassinated; for the children of tribal leaders who opposed AQ I to be tortured to death, their broken little bodies sent back to their families as a message; or for acid to be thrown in women’s faces as a punishment for wearing their veils pushed back too far.43 In farming areas, AQ I developed the habit of leaving the decapitated heads, or other body parts, of their victims in fruit boxes to be found by their families.44

  AQ I cells were thus cruelly capable and effective at the coercive end of the spectrum, but almost totally lacking in administrative and persuasive capabilities (as well as basic humanity). AQ I attempted nothing like formal governance, nor did it ever even try to provide any significant administrative services—it gave no tangible benefits to its supporters and provided no essential support or humanitarian assistance to the Sunni population. Its message to the population was, in essence, “Follow our rules, or we’ll kill you.” (In this, AQ I differed greatly from Shi’a militia groups such as Jaysh al-Mahdi or the Badr Organization, both of which—though also extremely violent—put a great deal of effort into winning support through social, economic, and humanitarian programs.)

  This is not to say that AQ I’s approach to violence was unsophisticated. On the contrary, the group cleverly established domination over the community through fear and through a carefully engineered cycle of sectarian violence, intimidation, and revenge. The cycle worked as follows. Having established a base in a Sunni-majority neighborhood and enforced their code of silence and fear on that community to give themselves a secure base of operations, members of an AQ I cell would set out to provoke a neighboring Shi’a community. They would kidnap Shi’a children, especially young boys, brutally torture them to death, and then dump the bodies in their families’ streets, or would attack Shi’a social and religious institutions, seeking to throw the blame on the Sunni community. These atrocities in turn would prompt retaliatory attacks by the Shi’a population against Sunni districts. Perhaps the most prominent example of this was AQ I’s bombing of the Samarra mosque, one of the holiest shrines in Shi’a Islam, on February 22, 2006. This attack prompted a tsunami of retaliatory bloodletting by Shi’a groups, targeting Sunnis across Iraq. At the local level, AQ I launched dozens of attacks against Shi’a districts from bases in Sunni-majority areas, including a notorious and bloody series of market bombings in 2006–7, which prompted Shi’a retaliation against Sunni neighborhoods.

  This approach worked for AQ I at first. Members of the Sunni community, attacked by Shi’ite vigilantes and hounded by representatives of the Shi’a-dominated Iraqi state (especially the Iraqi National Police, some units of which became notorious for extrajudicial killings of Sunni men and boys), felt they had nowhere to turn.45 Many began to see AQ I as the only thing standing between them and oblivion at the hands of the Shi’a, an ironic turn of events since (like a gangster in a protection racket) AQ I posed as the protector of the Sunnis—pretending to be the solution to a problem that AQ I itself was creating and exploiting.

  After the death of its first emir, Abu Musab al-Zarqawi, on June 7, 2006, AQ I’s new leadership began a limited attempt to translate its terror into broader political support, through the establishment of the Islamic Emirate of Iraq and the Mujahideen Shura Council. But AQ I’s approach was still fundamentally one of intimidation and fear, an approach that relied heavily on coercive means with little attempt at building administrative capability or persuasive activity. Thus, while AQ I did indeed establish a normative system that was capable of controlling the population, its capabilities all lay at the coercive end of the spectrum, and its control was therefore brittle. AQ I could offer its supporters nothing positive, and its violence and brutality against the population, along with its disregard of community leaders’ authority, its intolerance of traditional forms of Islam, and the fact that its leadership was largely urban or non-Iraqi, was building a groundswell of hatred against AQ I within rural and periurban Sunni communities. The only thing that kept these communities from turning against AQ I was the pall of fear the terrorists had cast over the population, and the expectation that anyone who opposed AQ I would die a slow, horrible, and publicly humiliating death.46

  The so-called Awakening, the uprising against AQ I from within the Sunni community that began in 2006 under a group of leaders that included Sheikh Sattar of the Albu Risha, was by no means the community’s first attempt to throw off these parasites. On the contrary, the 2006 Awakening was at least their fifth uprising.47 What made all the difference in 2006 was the U.S. troop surge. It wasn’t enough for the local community to hate AQ I: to rise up successfully, local civilians also had to believe they would survive the attempt. On every previous occasion, community leaders who went up against AQ I had been slaughtered. In 2006–7, the extra troop presence of the surge meant that for the first time, the coalition could hold and defend population centers on a permanent basis, support the Sunnis when they turned against the terrorists, and protect people against retaliation. The partnership between U.S. troops and the local community—arising in part from counterinsurgency tactics that emphasized protecting the people where they slept—gave the community the confidence to rise up again, and this time they succeeded. With
in a matter of weeks AQ I was destroyed, its control was swept away, and its cadres were mercilessly killed by the very population they had terrorized.48

  Thus, AQ I is an excellent example of the brittleness that can result from too narrow a spectrum of capabilities. AQ I established a terrifyingly effective ascendancy over the Sunni population, but because this dominance was based entirely on fear and coercion, it had no resilience. As soon as the surge created a minimal assurance for people that they would survive the attempt to turn against AQ I, and as soon as coalition forces in Anbar demonstrated that they could kill or capture members of AQ I cells, the myth of AQ I’s invincibility was shattered and the people turned on AQ I in a flash and swept it away. And because the terrorist group had little to offer but fear and intimidation, it had no way to counteract or bounce back from its loss of control.