In 1917 T. E. Lawrence wrote an essay called “Twenty-Seven Articles” summing up many of the lessons he had learned as an insurgent. What follows might be called Twelve Articles. It sums up the lessons of Invisible Armies.
1. Guerrilla warfare has been ubiquitous and important throughout history. Tribal warfare, pitting one guerrilla force against another, is as old as mankind and still exists in modified form in some parts of the world. A new form of warfare, pitting guerrillas against “conventional” forces, is of only slightly more recent vintage—it arose in Mesopotamia five thousand years ago. Therefore labeling guerrilla warfare as “irregular” has it backwards: it is the norm; interstate war is the exception.
Much of the world’s population lives in states whose current boundaries and forms of government were determined by insurgencies waged by or against their ancestors. Think of the United Kingdom, which was “united” by the success of the English in defeating Scottish and Irish insurgencies. That London rules much less territory than it did a century ago is in part the result of successful insurgencies by groups ranging from the IRA to the Zionists. Preceding all of them, of course, was the war waged by American colonists. The state they created reached its present borders only by waging three centuries of unremitting warfare against Indian irregulars.
These are just a few illustrative examples, which could be extended endlessly. It is hard to think of any state in the world that has avoided the ravages of guerrilla warfare—just as it is hard to think of any organized military force that did not have to spend, however unwillingly, a considerable portion of its time and resources fighting guerrillas.
2. Guerrilla warfare is not an “Eastern Way of War”; it is the universal war of the weak. Thanks largely to the success of Chinese and Vietnamese Communists in seizing power, there has been a tendency to portray guerrilla tactics as the outgrowth of Sun Tzu and other Chinese philosophers who were supposedly at odds with the kind of conventional tactics preached by Western sages such as Clausewitz. In reality, ancient Chinese and Indian armies were as massive and conventional in their orientation as the Roman legions. It was not the Chinese who had a cultural proclivity toward guerrilla warfare—it was their nomadic enemies in Inner Asia. For them, as for other tribesmen ranging from the Sioux to the Pashtuns, irregular warfare was a way of life. But even tribal peoples such as the Turks, Arabs, and Mongols, who enjoyed great success with guerrilla tactics during their rise to power, turned to conventional armies to safeguard their hard-won empires. Their experience suggests that few if any people have ever chosen guerrilla warfare voluntarily; it is the tactic of last resort for those too weak to create regular armies. Likewise, terrorism is the tactic of last resort for those too weak to create guerrilla forces.
3. Guerrilla warfare has been both underestimated and overestimated. Before 1945 the inherent value of guerrilla campaigns was generally underestimated. Because irregulars refuse to engage in face-to-face battle, they did not get the respect they deserve—notwithstanding their consistent ability to humble the world’s greatest empires, from the days of the Barbarian assaults on Rome to Haiti’s uprising against French rule in the 1790s. Since 1945, however, popular sentiment has swung too far in the other direction, enshrining guerrillas as superhuman figures who cannot be defeated by force. “Hence,” as one expert noted in 1967, “the modern guerrilla was almost invested with the nimbus of invincibility.”1 This is largely because of the success enjoyed by a handful of guerrilla leaders such as Mao Zedong, Ho Chi Minh, and Fidel Castro in the immediate postwar period that gave rise to the radical chic of the 1960s–1970s. But focusing on their exploits distracts from the ignominious end suffered by the great majority of insurgents, including Castro’s celebrated protégé, Che Guevara. In reality guerrilla warfare is neither invincible nor unwinnable. The truth lies somewhere in between: although often able to fight for years and inflict great losses on their enemies, guerrillas have seldom achieved their objectives. Terrorists have been even less successful.
4. Insurgencies have been getting more successful since 1945 but still lose most of the time. According to a database compiled for this book (see the Appendix), out of 443 insurgencies since 1775, insurgents succeeded in 25.5 percent of the concluded wars while incumbents prevailed in 63.6 percent. (The other 10.8 percent were draws.) Since 1945 the win rate for insurgents has gone up to 40.3 percent. But counterinsurgents still won 50.8 percent of post-1945 wars. And those figures actually overstate insurgents’ odds of success somewhat because many present-day rebel groups that are still in the field have scant chance of success. If ongoing uprisings are judged as failures, which they have been so far, the win rate for insurgents would go down to 21.9 percent in the post-1945 period, while the counterinsurgents’ winning percentage would rise to 68.7 percent. Balanced against this is the fact that some terrorist or guerrilla groups that do not manage to attain their stated goals can still call attention to their struggle and win some concessions from the other side. The IRA and PLO are good examples.
5. The most important development in guerrilla warfare in the last two hundred years has been the rise of public opinion. What accounts for the fact that guerrillas have been getting more successful since 1945? Much of the explanation can be found, I believe, in the growing power of public opinion brought about by the spread of democracy, schools and colleges, communications technology, the mass media, and international organizations—all of which have sapped the will of states to engage in protracted counterinsurgencies, especially outside their own territory, and heightened the ability of insurgents to survive even after suffering military setbacks.
Before the rise of public opinion, most guerrillas were largely apolitical tribesmen who may have excelled at hit-and-run raiding but had no conception of political mobilization. They did little or nothing to woo undecided people or to undermine the will of the opposing populace, save by brute force. Sometimes, as in the case of the late Roman Empire, the targets of their inveterate attacks were weak enough that they finally succumbed. More often, however, the tribesmen were beaten back by settled civilizations that could muster more powerful armies. The strategic advantage in those days lay with the defenders, notwithstanding the attackers’ immense tactical skills at fighting on horseback. As states got stronger from the seventeenth century on, the advantage titled further in their favor.
The balance of power began to shift toward insurgents because of the growth of public opinion, a term that first appeared in print, appropriately, in 1776. Notwithstanding their military setbacks in North America, the British might eventually have been able to restore control had not a parliamentary revolt in 1782 forced the downfall of Lord North’s ministry and the rise of a Whig government dedicated to negotiating an end to the conflict. A few decades later the Greek rebels of the 1820s benefited from public opinion in the West, where philhellenes rallied their governments to oppose Ottoman abuses. A similar strategy would be pursued by many rebels of the future, from the Cubans opposing Spanish rule in the 1890s to the Algerians opposing French rule in the 1950s and Hezbollah opposing Israeli power since the 1980s. A spectacular vindication of this approach occurred during the Vietnam War, when the United States was defeated not because it had lost on the battlefield but because public opinion at home had turned against the war. The same thing almost happened in Iraq in 2007.
Public and press opposition is particularly potent when brought to bear against liberal states whose will to continue a war is dependent on popular support. But the growing power of public opinion affects even the calculations of illiberal regimes that find it harder to repress rebellion because of the spread of technologies such as Twitter and YouTube, of media outlets such as CNN and Al Jazeera, and of organizations such as the United Nations and Human Rights Watch. Today insurgency and counterinsurgency have to be waged not only on the ground but in cyberspace and on satellite television. These are realms where innovative Islamist groups such as Al Qaeda and Hezbollah have excelled and more hidebound conventional militaries have lagged.
6. Conventional tactics don’t work against an unconventional threat. Regular soldiers often assume that they will have no difficulty besting ragtag fighters who lack the firepower or discipline of a professional fighting force. Their mindset was summed up by General George Decker, U.S. Army chief of staff from 1960 to 1962, who said, “Any good soldier can handle guerrillas.” The Vietnam War and countless other conflicts have disproven this bromide. Big-unit, firepower-intensive operations snare few guerrillas and alienate many civilians. To defeat insurgents, soldiers must take a different approach that focuses not on chasing insurgents but on securing the population. This is the difference between “search and destroy” and “clear and hold.” The latter approach is hardly pacifistic. It too requires the application of violence and coercion but in carefully calibrated and intelligently targeted doses. As an Israeli general told me, “Better to fight terror with an M-16 than an F-16.”2
7. Few counterinsurgents have ever succeeded by inflicting mass terror—at least in foreign lands. When faced with elusive foes, armies too often have resorted to torturing suspects for information and inflicting bloody reprisals on civilians. Such strategies have worked on occasion, but just as often they have failed. The point is well illustrated by revolutionary and Napoleonic France’s experience. The French revolutionaries killed indiscriminately and successfully to suppress the revolt in the Vendée, a region of France, in the 1790s. But French armies failed to pacify either Spain or Haiti in spite of their willingness to be just as brutal. Even in the ancient world, when there were no human-rights lobbies and no CNN, empires found that pacifying restive populations usually involved carrots as well as sticks. There were considerable benefits to the Pax Romana that won over subject populations; there was much more to Roman counterinsurgency than “they create a desert and call it peace.”
The brute-force approach is most successful when the rebel movement is very weak or, better still, nonexistent and the counterinsurgents are trying to pacify their own territory, where they have at least some degree of legitimacy, considerable knowledge of the human and geographic terrain, and can bring overwhelming force to bear. Stalin’s Great Terror in the 1930s and Mao’s Cultural Revolution in the 1960s are cases in point. But in many other instances, like those of the Nazis in the Balkans and the Soviets in Afghanistan, even the willingness of counterinsurgents to inflict genocidal violence was not enough to prevail; their atrocities simply drove more people into the arms of rebels who had external backing. That is why the political scientist Stathis Kalyvas, a leading student of internal wars, has concluded that “indiscriminate violence seems to be counter-productive, with the exception of situations where there is a high imbalance of power.”3
8. Population-centric counterinsurgency is often successful, but it’s not as touchy-feely as commonly supposed. The fact that the U.S. and other liberal democratic states cannot be as brutal as dictatorial regimes, or, more precisely, choose not to be, does not mean they cannot succeed at counterinsurgency, as is sometimes claimed;4 they simply have to practice a more humane style of counterinsurgency exemplified by Edward Lansdale and Gerald Templer. David Petraeus showed in Iraq in 2007–08 how successful population-centric counterinsurgency could be, at least in narrow security terms, even if the “surge” did not bring about a lasting political settlement. The same lesson was taught by President Alvaro Uribe who implemented a similar “democratic security” strategy in Colombia at the same time, and by the Israel Defense Forces during their successful operations to suppress the Second Intifada. Other notable successes for population-centric tactics occurred in Northern Ireland in the 1990s, El Salvador in the 1980s, and Malaya in the 1950s.
A more popular term for population-centric strategy is “winning hearts and minds,” a deceptive phrase first coined by General Henry Clinton during the American Revolution and popularized by Gerald Templer in the 1950s. This term, by now a cliché, suggests that the counterinsurgents are trying to win a popularity contest. Some governments have tried this approach: between 2003 and 2007 the United States spent at least $29 billion in development aid in Iraq5 in the hope that this would generate socioeconomic benefits that would cause the Iraqi people to reject the insurgents. This strategy failed because of a pervasive lack of security: if supporting the government entails a high risk of getting murdered, few will be enticed to do so because of the provision of better education or trash collection. In most conflicts, the majority of the population has sat on the fence until it was clear which side was likely to win, something that did not occur in Iraq until the surge took effect in 2007. Likewise during the American Revolution, successful British offensives caused Tories to become more active, whereas British retreats led more colonists to declare rebel sympathies. This underscores an important point made by Stathis Kalyvas: “Gaining control over an area brings collaboration, and losing control of an area brings much of that collaboration to an end.”6
The only way to gain the necessary control is to garrison troops 24/7 among the civilians; periodic “sweep” or “cordon and search” operations, even when conducted by counterinsurgents as cruel as the Nazis, fail because civilians know the rebels will return the moment the soldiers leave and exact a terrible revenge on anyone who collaborated with them. The populace will embrace the government only if it is less dangerous to do so than to support the insurgency, which is why successful population-centric polices aim to control the people, not to win their love and gratitude. As John Paul Vann, the legendary American adviser in South Vietnam, said, “Security may be ten percent of the problem, or it may be ninety percent, but whichever it is, it’s the first ten percent or the first ninety percent. Without security, nothing else we do will last.”7
9. Establishing legitimacy is vital for any successful insurgency or counterinsurgency—and, in modern times, that is hard to achieve for a foreign group or government. It is, in fact, the second-most important requirement after physically securing the population. Before the twentieth century it was easy for unelected rulers, even foreigners, to obtain legitimacy: rule by emperor, kings, and chieftains was the norm throughout much of history. The spread of nationalism and democracy in the intervening years has made it difficult for unelected regimes, particularly if imposed from abroad, to gain popular allegiance. Thus, when fighting insurgents abroad, great powers today are forced to buttress the legitimacy of homegrown regimes rather than simply impose their own colonial officials at bayonet point, as their ancestors might have done. For the United States, this task was relatively easy to accomplish in states such as Colombia and the Philippines where the United States was supporting established, democratically elected governments. It was much more of a challenge in South Vietnam, Afghanistan, and Iraq where the Untied States was trying to create legitimate governments from scratch. For the Soviet Union in Afghanistan, backing a militantly atheistic regime in a devoutly Muslim land, this was a mission impossible.
Legitimacy is a problem for insurgents as well. Anarchism faded out as a revolutionary threat because it was never able to establish its credibility as a governing creed. Leftist insurgencies faded away after the collapse of the Soviet Union and the market reforms in China discredited Marxism-Leninism. Today jihadism, meaning violent extremism not mainstream Islamism, is imperiled by its lack of popular legitimacy even in the Muslim world. Hamas and Hezbollah have done better than other armed Islamist groups because they have provided social services to win over the people of Gaza and Lebanon.
The ideology that has proven most popular and hence most durable as a motivating force for guerrillas and terrorists is neither liberalism nor anarchism nor socialism nor Islamism but, rather, nationalism. Its appeal may be judged from the fact that even though most terrorist groups of the 1970s failed, those that had a nationalist appeal, such as the PLO and IRA, managed to win important concessions, whereas those that advocated radical social change, such as the Baader-Meinhof Gang and the Weathermen, disappeared without a trace after the death or imprisonment of their leaders.
10. Most insurgencies are long-lasting; attempts to win a quick victory backfire. According to the Invisible Armies database, the average insurgency since 1775 has lasted ten years. The figure is even longer for post-1945 insurgencies—fourteen years. Interestingly there is little correlation between the length of a conflict and the insurgents’ chances of success. This flies in the face of the conventional wisdom summed up by the Vietnamese Communist strategist Truong Chinh in 1947: “To protract the war is the key to victory. . . . Time is on our side.”8 Protracting the conflict did favor the insurgents in Indochina, where their enemies were foreigners who would eventually tire of the conflict. But time is not usually on the insurgents’ side when fighting a homegrown regime. In the Invisible Armies database, there is little difference in outcomes between insurgencies that last fewer than ten years and those that last more than twenty years—the incumbents won 64.3 percent of the former and 63.9 percent of the latter.
The fact that low-intensity conflict tends to be “long, arduous and protracted,”9 in the words of Sir Robert Thompson, can be a source of frustration for both sides, but attempts to short-circuit the process to achieve a quick victory usually backfire. The United States tried to do just that in the early years of the Vietnam and Iraq wars by using its conventional might to hunt down insurgents in a push for what John Paul Vann rightly decried as “fast, superficial results.”10 It was only when the United States gave up hopes of quick victory, ironically, that it started to get results by implementing the tried-and-true tenets of population-centric counterinsurgency. In Vietnam, it was already too late, but in Iraq the patient provision of security came just in time.
A particularly seductive version of the “quick win” strategy is to try to eliminate the insurgency’s leadership. Such strategies do sometimes work. The Romans managed to stamp out a revolt in Spain by inducing some of the rebels to kill their leader, Viriathus, in 139 BC. The Americans managed to hasten the end of the Philippine insurrection by capturing its leader, Emilio Aguinaldo, in a daring 1901 raid. But there are just as many examples where leaders were eliminated but the movement went on, sometimes stronger than ever—as both Hezbollah and Al Qaeda in Iraq did. High-level “decapitation” strategies work best when a movement is weak organizationally and focused around a cult of personality. Even then leadership targeting is most effective if integrated into a broader counterinsurgency effort designed to separate the insurgents from the population. If conducted in isolation, leadership raids are about as effective as mowing the lawn; the targeted organization can usually regenerate itself.
Insurgents can also stumble if they attempt a premature offensive to achieve victory, as General Giap did in 1951 and 1968. Both insurgents and counterinsurgents would be well advised to heed Field Marshal Templer’s sagacious observation: “I have always said,” he wrote from Malaya in 1953, “that the complete cure of it all will be a long slog.”11
11. Guerrillas are most effective when able to operate with outside support—especially with conventional army units. From the guerrillas’ standpoint, the most important advantage they can enjoy next to having a popular cause is having outside support. Best of all is to be able to operate in conjunction with conventional units, either their own or an ally’s. This keeps a conventional army off balance. When it masses to fight main force units, it leaves its lines of communication vulnerable to guerrilla attacks. When it disperses to focus on the guerrillas, it leaves itself vulnerable to attack by the main force. In a few cases guerrilla leaders such as Mao Zedong and Ho Chi Minh have been able to build their own main force units. But such precedents are rare. More common is for rebels to work with foreign allies, the classic examples being American rebels cooperating with French forces against the British, Spanish guerrilleros cooperating with Wellington against Napoleon, and Arab rebels cooperating with Allenby and Lawrence against the Ottomans.
Even when they do not have main force units to work with, guerrillas greatly benefit from foreign funding, arms, training, and safe havens. No other factor correlates so closely with insurgent success—as demonstrated by examples as varied as the Vietcong and the Afghan mujahideen. One of the factors that greatly aided counterinsurgents in the ancient world was that insurgents were usually devoid of outside support. When insurgents do have substantial aid and it is cut off, the result can be catastrophic, as it was for the Greek Communist Party, which was cut off by Yugoslavia in 1948, and Angola’s UNITA, which was cut off by South Africa and the United States in the 1990s.
We must not, however, exaggerate the impact of foreign support. It is possible to win with little or no outside backing, as Fidel Castro showed in Cuba and Michael Collins in Ireland. It is also possible to lose even if you have substantial foreign backing, as Al Qaeda in Iraq, the Popular Front for the Liberation of Palestine, and other groups have discovered. However, even if foreign sanctuaries cannot necessarily lead an insurgency to victory, they can keep it from being totally defeated: that has been the experience of Colombia’s FARC, which was kept alive in large measure by support from Venezuela’s Hugo Chávez after it suffered a long string of defeats starting in 2002.
12. Technology has been less important in guerrilla war than in conventional war—but that may be changing. All guerrilla and terrorist tactics, from suicide bombing to hostage taking and roadside ambushes, are designed to negate the firepower advantage of conventional forces. In this type of war, technology counts for less than in conventional conflict. Even the possession of nuclear bombs, the ultimate weapon, has not prevented the Soviet Union and the United States from suffering ignominious defeat at guerrilla hands. To the extent that technology has mattered in low-insurgency conflicts, it has often been the nonshooting kind. As T. E. Lawrence famously said, “The printing press is the greatest weapon in the armory of the modern commander.” A present-day rebel might substitute “the Internet” for “the printing press,” but the essential insight remains valid.
However, the role of weapons in this type of war could grow in the future if insurgents get their hands on chemical, biological, or especially nuclear weapons. A small terrorist cell the size of a platoon might then have more killing capacity than the entire army of a nonnuclear state like Brazil or Egypt. That is a sobering thought. It suggests that in the future low-intensity conflict could pose even greater problems for the world’s leading powers than it has in the past. And, as we have seen, the problems of the past were substantial and varied.