The End of the Age of Petraeus

U.S. Air Force Senior Airman John Fitzgerald arm wrestles a young Afghan. (U.S. Air Force / Flickr)

The downfall of David Petraeus sent such shock waves through the policy establishment when it hit the news in November because the cause was so banal: the most celebrated and controversial military officer of our time compelled to resign from his dream job as CIA director as the result of an extramarital affair. Yet long after the headshaking details are forgotten, Petraeus' larger significance will remain, as his career traced one of the era's crucial strategic narratives -- the rise and fall of counterinsurgency in U.S. military policy.

As recently as 2006, the country's top generals were openly scorning counterinsurgency as a concept; the secretary of defense all but banned the term's utterance. One year later, it was enshrined as army doctrine, promoted at the highest levels of the Pentagon, and declared official U.S. policy by the president. Then, five years after that, a new president and new defense secretary barred the military chiefs from even considering counterinsurgency among the war-fighting scenarios used to calculate the military's force requirements.

The swerves reflected the changing courses of the wars being fought on the ground. The George W. Bush administration had invaded Afghanistan in 2001 and Iraq in 2003 with a "light footprint" strategy, designed to defeat the enemies and get out quickly to avoid getting bogged down. That approach, however, revealed its limits as Iraq began unraveling soon after the collapse of Saddam Hussein's regime, and by mid-2006, the country had slipped into a vicious, chaotic civil war. A desperate Bush decided to gamble on counterinsurgency in a last-ditch effort to head off disaster, and he picked Petraeus, the author of a new army manual on the subject, to lead the effort. The apparent success of the new approach in stanching the bleeding inspired commanders, including Petraeus himself, to apply it to the worsening conflict in Afghanistan as well. But its apparent failure there led President Barack Obama -- never a huge fan -- to back away from the strategy not only there but in general.

U.S. troops are now out of Iraq and being drawn down in Afghanistan, but the basic questions about counterinsurgency -- or COIN, as it is widely abbreviated -- remain. Did it really succeed in Iraq, and if so, how? Why did it not work in Afghanistan? Is it a viable strategy for dealing with contemporary insurgencies, and even if it is, can it be employed by a democracy, such as the United States, with little patience for protracted war?

THE ROOTS OF THE COIN CABAL

The revival of COIN in the Age of Petraeus -- a brief era, but worthy of the title, so thorough was his influence and the improbable fame he attained -- was in part the product of generational politics. In 1974, when Petraeus graduated from West Point, the Vietnam War was approaching its inglorious denouement, and the U.S. Army's senior leaders were determined never to fight guerrillas again, in the jungle or anyplace else. They turned their gaze instead to the prospect of a major conventional war with the Soviet Union on the wide-open plains of Europe and threw out the books on what they termed "low-intensity conflict." By the early 1990s, army scribes had come up with a still more dismissive term: "military operations other than war," abbreviated as MOOTW (pronounced "moot-wah"). And the feeling was, as General John Shalikashvili, former chairman of the Joint Chiefs of Staff, once muttered, "Real men don't do moot-wah."

Many of today's officers, however, rose through the ranks fighting precisely these "other-than-war wars" (as some called them), in El Salvador, Panama, Somalia, Haiti, and the Balkans, which didn't seem so low intensity and certainly felt like wars. Petraeus himself spent the early years of his career as an airborne infantry officer in France and Italy, where he happened upon a shelf-load of books touting what the French call "revolutionary warfare": Jean Lartéguy's novel The Centurions, Bernard Fall's Street Without Joy and Hell in a Very Small Place, and, most influential, David Galula's Counterinsurgency Warfare: Theory and Practice. Galula, who had observed and fought in several counterinsurgencies himself, was unlike any author Petraeus had read before. "Revolutionary war," his book stated, has "special rules, different from those of the conventional war." It's like a fight between a lion and a flea: the flea can't deliver the knockout punch, and the lion can't fly. The insurgent can sow disorder anywhere, whereas the counterinsurgent -- fighting on behalf of the government -- has to maintain order everywhere. Defeating fleas requires draining the swamp that sustains them; defeating insurgencies requires protecting, then wooing or co-opting, the population that sustains their cause. As Galula described, a soldier in a COIN campaign must "be prepared to become a propagandist, a social worker, a civil engineer, a schoolteacher, a nurse." Likewise, "a mimeograph machine may turn out to be more useful than a machine gun," and "clerks [are] more in demand than riflemen." These kinds of wars, Galula calculated, quoting Mao, are "20 percent military action and 80 percent political."

In the mid-1980s, Petraeus spent a summer as an aide to General John Galvin, head of the U.S. Southern Command. Central America was then blazing with the sorts of insurgencies that Petraeus had previously only read about; in El Salvador, U.S. military aides were devising something close to a counterinsurgency plan. Toward the end of his stay, Petraeus ghostwrote an article for Galvin titled "Uncomfortable Wars: Toward a New Paradigm," which called on the army to abandon its obsession with big wars and firepower and to recognize the prevalence of new kinds of warfare -- subversion, terrorism, guerrilla insurgencies. When he returned to the States, Petraeus elaborated on these points in a Princeton doctoral dissertation on the army's "myopic" post-Vietnam aversion to such conflict and its need to change its doctrine, tactics, and personnel policies accordingly.

In the mid-1990s, Petraeus served as chief of operations in the U.S.-led multinational peacekeeping force in Haiti, his first experience with full-fledged nation building. A few years later, in 2001, he was deployed to Sarajevo, as NATO's assistant chief of staff for operations and deputy commander of a clandestine unit called the Joint Interagency Counterterrorism Task Force. A briefing for his campaign plan emphasized the need both to go after the terrorists directly and to address the problem's root causes, tackling unemployment, the issue of sanctuaries, and a corrupt justice system.

In 2003, the United States went to war in Iraq, and the campaign plan devised by Secretary of Defense Donald Rumsfeld and Tommy Franks, head of the U.S. Central Command, had no use for any of this. It was intended to be a "shock and awe" campaign to topple the regime quickly and then hand over responsibility for the country to somebody, anybody else -- American allies, Iraqi exiles, untainted local leaders, whatever. Petraeus commanded the 101st Airborne Division during the brief but fierce drive to Baghdad.

It was after Saddam fell that Petraeus made his mark. Assigned to occupy the northern Iraqi province of Nineveh, including the city of Mosul, he applied all the lessons he had learned during his stints in Central America, Haiti, and Bosnia and from his readings of Galula and the other COIN classics (which he brought with him and consulted frequently). He sought out and worked closely with community leaders, vetted candidates for new local elections, got gas pumps working, reopened the university, even opened the province's border with Syria. Petraeus was doing all this on his own initiative. Few other commanders detected the rise of an insurgency; fewer still understood its implications. They had not read up on counterinsurgency strategy: it hadn't been taught at West Point or any of the army's war colleges recently, and a field manual on the subject had not been published in 20 years.

In 2005, Petraeus returned home to command the U.S. Army Combined Arms Center, at Fort Leavenworth, Kansas. He was determined to get COIN into the army's official curriculum, to see that it was part of the predeployment training programs for all units, and above all to write a new counterinsurgency field manual. At the same time, under the radar, a new generation of like-minded officers was rising through the service ranks, inspired by similar experiences and imbued with similar insights. Much of the new thinking was coalescing in West Point's Department of Social Sciences, known as "Sosh," where Petraeus had taught while finishing his dissertation. Sosh had long been the army's locus of unconventional thinking, run by professors determined to turn out "very broad-gauged individuals," not just battalion commanders. For the junior officers who had returned from the mootw wars of the 1980s and 1990s to teach in Sosh and West Point's history department, the curriculum and discussions reinforced what they had learned on the foggy battlefields in the developing world: that such fights had at least as much to do with politics and economics as with military tactics and that most of their senior officers -- still stuck on Cold War precepts stressing large maneuvers and heavy firepower -- were ill equipped to command irregular wars.

This group included John Nagl, who would go on to write an influential book on counterinsurgency, Learning to Eat Soup With a Knife; H. R. McMaster, who would command one of the Iraq war's most successful COIN campaigns, in the city of Tal Afar; Kalev "Gunner" Sepp, who would help set up a COIN academy for all incoming soldiers in Iraq; and others. During the early years of the U.S. occupation of Iraq, as they realized that the war there had morphed into an insurgency that American military and political leaders didn't recognize or know how to fight, these peculiar officers, along with a few outsiders, wrote articles for army journals, attended workshops and conferences, and formed a nascent community -- "the COIN cabal," or "the Sosh Mafia," as some called it. By the time Petraeus set out to write a new COIN field manual, this network was already in place for him to draw on -- and with Petraeus, it gained a leader with ferocious ambition, talent, and stars on his epaulets.

FROM FRYING PAN TO FIRE

Meanwhile, all hell was breaking loose in Iraq. The calm that Petraeus had achieved in Mosul quickly eroded when the 101st Airborne Division rotated out and was replaced by a smaller force with a conventional approach. Other parts of the country deteriorated faster and further. During the three years after the fall of Baghdad, an overtaxed U.S. civilian occupation authority, a hapless and underresourced U.S. military command, and Iraq's own fractious politics combined to produce vicious anarchy. The increasingly authoritarian national government in Baghdad was competing for power not only with insurgents led by disaffected Sunnis and radical jihadists in the west and north but also with separatist Shiites in the south. The capital itself became a killing field, with the conflict throwing up scores of mutilated bodies weekly.

On February 23, 2006 -- the same day that Petraeus commenced a workshop on his new COIN manual at Fort Leavenworth -- Sunni insurgents blew up the Golden Mosque, a major Shiite shrine in Samarra, sending Iraq to the brink of civil war and giving Petraeus a greater sense of urgency. Pushing the manual through a resistant army bureaucracy and corralling support for COIN among opinion leaders now appeared vital not only to shifting the military's broader view of warfare but also to avoiding catastrophe in Iraq. Petraeus had heard stirrings that in a year's time, he might be sent back to Iraq as the new U.S. commander there. To be able to impose his Nineveh strategy across all of Iraq, however, he would need the cover of officially sanctioned doctrine, which the field manual, if accepted, would provide.

General George Casey, then the U.S. commander in Iraq, had signed on to a COIN campaign plan the previous summer, influenced by his two strategic advisers, Sepp and Colonel William Hix, co-founders of the COIN academy. But by the time of the Samarra attack, Sepp, Hix, and their team of advisers -- mostly think-tank Ph.D.'s, nicknamed "doctors without orders" -- had rotated out. And Casey's support for COIN had always been shallow. He had served in Bosnia, but unlike Petraeus, he saw it not as a model for future wars but as a trap, in which the locals took advantage of the large and active U.S. presence to shirk their own responsibilities even as their resentment of the outsiders grew. Casey was also an institutional army man, and from that perspective, he saw Iraq draining the army of resources. Finally, he had his orders: Rumsfeld was telling him to lower the U.S. profile and get out of Iraq as quickly as possible.

So Casey responded to the Samarra bombing and the subsequent violence by returning to his pre-COIN position. The war, as Casey saw it, had degenerated into a battle for political and economic power among many ethno-sectarian factions, and with no single insurgency, it made no sense to pursue a counterinsurgency strategy. He reverted to the only alternative he knew, his original plan, written before Hix and Sepp joined his staff, which involved turning over authority to the Iraqi government and withdrawing rapidly.

To the COIN advocates, Casey was defining counterinsurgency too literally. Another phrase for such campaigns, after all, is "stability operations," and Iraq in the spring of 2006 was the very picture of instability. Handing responsibility to the Shiite-dominated Iraqi government would make things worse rather than better, since the government was itself one of the warring factions. The Interior Ministry's police weren't guardians of public order; they were death squads assassinating Sunnis in broad daylight. The Health Ministry's guards were refusing to treat wounded Sunnis in their emergency wards and, in some cases, were actually murdering them. Scaling back and pulling out would pour oil on the flames and possibly ignite a broader regional conflict.

The alternative put forth by the "COINdinistas" came straight from Galula. The task of a counterinsurgent army, he had written, was to push the bad guys out of one area at a time, and then to stay there, so they wouldn't come back. Meanwhile, it also had to train the local police and soldiers, so they could secure their country by themselves, and help the local government provide basic services, thus earning the allegiance of the people and drying up, or co-opting, their support for the insurgency. The shorthand term for this strategy (taken from a similar, although brief, effort during the Vietnam War) was "clear, hold, and build": clear the insurgents, hold the area, and build services and support.

Petraeus' COIN field manual was published on December 15, 2006, to much acclaim, some criticism, and a surprising level of curiosity. In part because of Petraeus' fame as the hero of Mosul and his knack for dealing with the press and Congress, 1.5 million people downloaded the manual's online edition in its first month on U.S. Army Web sites. But army doctrine was one thing, national policy another. Before it could have much effect on the war or the U.S. military at large, four changes would have to take place: there would have to be a new secretary of defense in Washington, a new U.S. commander on the ground in Iraq, more troops to implement the new approach, and a clear example of practical success to light the way forward. As it happened, in the month that the manual was published, the groundwork was laid for all four.

THE STORY OF THE SURGE

On November 7, 2006, the Republican Party had lost badly in the midterm elections; as Bush noted, the voters had given his administration and party "a thumping." In the wake of the defeat, Bush fired Rumsfeld, the chief theorist and most passionate advocate of the "light footprint" approach, paving the way for serious consideration of the alternative Iraq strategy that had been gaining ground among dissidents in various quarters.

The case for what would come to be known as "the surge" grew out of quantitative analysis but spread as a result of bureaucratic networking. During the course of 2006, the Iraq Study Group -- a blue-ribbon panel appointed by Congress to give advice on the faltering war effort -- held its deliberations about what the government should do next. As the fall progressed, Frederick Kagan, a defense analyst at the American Enterprise Institute, worried that the commission would recommend a rapid withdrawal rather than a commitment to stay and win. Kagan had spent the previous decade as a civilian professor of military history at West Point, teaching, among other things, the course on revolutionary warfare, the academy's one post-Vietnam concession to COIN theory. One of his colleagues during those years, and for a time his officemate, was McMaster, the commander of the COIN campaign in Tal Afar. McMaster was now in Washington, as one of 15 colonels secretly advising the Joint Chiefs of Staff on options for Iraq. (He had been recommended by Petraeus.)

Kagan and McMaster both agreed with Casey's critics, but Kagan needed some hard evidence to back up his case against the hand-over-and-withdraw plan. McMaster suggested that Kagan call two of his former aides from Tal Afar, Colonel Joel Armstrong and Major Daniel Dwyer, who had done the fine-tuned analysis that made the operation there a success, including crunching the numbers on how many troops were necessary, when, and where. (Both had since retired from the army.) Kagan and a few assistants had already gone through the open-source literature to pinpoint the Iraqi neighborhoods with the most violence, mainly in Baghdad and Anbar Province. Armstrong and Dwyer now called up overhead images of those areas on the Google Earth Web site and calculated how many troops would be needed to secure -- to clear and hold -- each area. The conclusion: five brigade combat teams and two regimental combat teams, about 24,000 extra troops in all. Dwyer then computed how quickly those units could be mobilized to Iraq, drawing on the army's "force generation model." (The model was classified, but Dwyer, to his amazement, found it reprinted on Wikipedia.) It turned out that five brigade combat teams and two regimental combat teams were exactly the number that could be spared for Iraq.

Kagan, Armstrong, and Dwyer prepared a PowerPoint briefing based on their analysis -- 55 slides in all -- and presented it at a conference in early December. But first, Kagan showed it to Jack Keane, a retired army general who was also growing worried about Iraq. As it happened, Keane had been called to a White House meeting the following week, as one of a handful of experts to discuss Iraq with Bush. Keane brought along a copy of Kagan's slides and gave them to Vice President Dick Cheney. During the meeting, all the experts urged Bush to fire Casey. Bush asked who should replace him. Keane mentioned Petraeus; others agreed.

Bush had recently appointed Robert Gates, a former CIA director, to be Rumsfeld's successor, and on his first full day in the position, Gates and a handful of his staff members flew to Iraq. Eric Edelman, then undersecretary of defense, had found out about the briefing by Kagan, who had worked long ago as his intern, asked for a copy of the slides, and showed them to Gates on the flight over. "The president has seen these," Edelman said. "You should, too." Meanwhile, Petraeus had been tapping into his own network, including McMaster, several other colonels and generals on the Joint Staff, and, not least, Meghan O'Sullivan, Bush's special assistant on Iraq, with whom Petraeus had established a back channel: she would use him for a reality check on Casey's reporting, and he would use her for updates on the state of play in the White House.

By the time Bush met with his senior national security advisers in Crawford, Texas, over Christmas to discuss Iraq, the fix was in place for the surge, a change of strategies, and the appointment of Petraeus as top commander. In a prime-time speech on January 4, 2007, announcing his plans, Bush declared that the situation in Iraq was unacceptable and that "we need to change our strategy." To facilitate the change, he said, he had decided to send "more than 20,000 additional troops" -- five army brigades to Baghdad and another 4,000 marines to Anbar. "In earlier operations," he noted, "Iraq and American forces cleared many neighborhoods of terrorists and insurgents, but when our forces moved on to other targets, the killers returned. This time we'll have the force levels we need to hold the areas that have been cleared." Counterinsurgency was now official policy.

STANCHING THE BLEEDING

While these maneuverings were playing out on the home front, something was happening in the epicenter of violence in Iraq. Seventy miles west of Baghdad, in the capital of Anbar Province, Ramadi -- a city of nearly half a million people where Sunni insurgents ran free and al Qaeda gunmen enjoyed unchecked control -- a mere 6,000 U.S. troops were turning the war around through classic COIN techniques, with little direction or even recognition from Baghdad or Washington.

The phenomenon came to be called the Anbar Awakening. It began when local Sunni sheiks concluded that the jihadists in their midst were stepping out of line: forcibly marrying their daughters and killing anyone who resisted, often dumping their bodies in fields rather than giving them proper Muslim burials. Sunni militiamen who had been shooting at the American occupiers a few weeks earlier now started asking them for help against a common -- and more dangerous -- enemy. Colonel Sean MacFarland, the army brigade commander on the ground, had been a Sosh cadet at West Point, had pored over COIN literature at Fort Leavenworth's School of Advanced Military Studies, and, more recently, had replaced McMaster as the commander in charge of Tal Afar. McMaster had briefed him fully on what he had done there; MacFarland added his own twists and later decided to apply the same principles when he was reassigned to Ramadi.

Ramadi was a tougher nut to crack. Once McMaster had driven the insurgents out of Tal Afar, for example, he built a fence around the city to keep them out, but in Ramadi, the insurgents were living in the city. So MacFarland had to find potential allies, recruit them into a police force, and hand out money for economic development projects, all while heavy fighting was still going on.

Still, by the time Petraeus returned to Iraq, the clear, hold, build approach was showing results in Anbar, and the new commander decided to extend it throughout the Sunni regions of the country. He called the program the Sons of Iraq, recruiting former militiamen to join the fight against al Qaeda, giving them weapons, and paying them out of his commander's discretionary fund (a move of borderline legality, but he persuaded his lawyers to approve it under the rubric of "site security"). "Cash is a form of ammunition," Galula had written, and Petraeus kept it flowing.

In that first year, as the surge took hold and the strategy evolved, casualties at first rose but then subsided. The cycle of violence -- the persistent pattern of Sunni attack sparking Shiite retaliation, provoking Sunni attack, and so on -- broke.

The Anbar Awakening had preceded Petraeus and the surge, and it was initiated by Sunnis, not Americans. But it took an officer of MacFarland's training and disposition to grasp its potential and respond to it shrewdly. Even then, it would have remained a local phenomenon had it not been for the surge and Petraeus. The surge provided the resources to spread the Awakening across the rest of Iraq; Petraeus knew exactly how to spread it.

Petraeus' success, throughout his career, had stemmed in part from his brazen assertiveness. During the Casey era, U.S. forces often faced attacks from a Shiite militia based in the Baghdad neighborhood of Sadr City, but Iraqi Prime Minister Nouri al-Maliki barred Casey from responding, and Casey complied. By contrast, Petraeus simply ordered his troops into Sadr City without telling Maliki ahead of time, then gave him intelligence materials showing that Muqtada al-Sadr, the militia's leader, was not the reliable friend that Maliki had believed.

The 2007 turnaround in Iraq was remarkable, but it was also oversold. It was not due entirely to the surge or to COIN or to Petraeus personally. There were other factors, which had little to do with anything the Americans had done (apart from invading Iraq and thus ripping its social fabric apart in the first place). Petraeus' command and the surge came late to the civil war. Many areas of violence had already been cleared through ethnic cleansing and the exile of hundreds of thousands of Sunnis. Moreover, the Sunnis would not have been so eager to split with the jihadists, much less ally with the Americans, had they not realized that they were losing the civil war against the Shiites. Similarly, Maliki consented to U.S. assaults on Shiite militias in part because he had no choice but also because he, too, had come to realize that his erstwhile partner Sadr was at least as much a threat as an ally.

There was also a larger issue. Petraeus frequently said both publicly and privately that the surge was a means, not an end. The idea was to give Iraq's factions a relatively calm breathing space in which they could work out a durable political settlement, one that reconciled Sunnis and Shiites so as to create truly national institutions, dealt with Kurdish claims of regional autonomy, divided up Iraq's oil wealth, and resolved property disputes in Kirkuk. Six years later, none of this has happened. The surge and the switch to COIN can be seen, in retrospect, as mere tactical successes at best. At the time, however, they were seen as much more than that, and so they helped shape decisions when it came to the next large-scale escalation.

LATHER, RINSE, REPEAT

During the 2008 presidential campaign, Obama repeatedly called for sending more troops to Afghanistan. Many thought he was playing politics: Iraq was Bush's war, therefore bad; Afghanistan was the war Bush had neglected, therefore good. There may have been something to this, but another factor was that one of Obama's foreign policy advisers was Bruce Riedel. A recently retired CIA analyst with a specialty in South Asia, Riedel possessed a deep knowledge of terrorist groups in Afghanistan and Pakistan, and the Democratic candidate heeded his briefings on their dangers.

After his election, Obama set about making good on his commitment, putting Riedel in charge of a quick review of policy toward what was now called AfPak. The results of the review were announced at the end of March 2009, enshrining goals for Afghanistan -- to "degrade, dismantle, and destroy al Qaeda"; accelerate training of the Afghan military; and build up the Kabul government -- and arguing that the best way to accomplish them, at least in the southern part of the country, was through "a fully resourced counterinsurgency strategy." As in Iraq, the new policy was accompanied by a dynamic new commander, General Stanley McChrystal, and some additional troops.

By his own admission, Riedel knew nothing about COIN beforehand and was influenced on this point by one of the members of the interagency group that worked on the review: Petraeus, now commander of Central Command. Most of Obama's senior advisers backed the idea, again under Petraeus' influence, but there were two major dissenters. Vice President Joseph Biden favored a "counterterrorism-plus" strategy -- just going after the insurgents (using drones strikes and Special Forces raids) and training the Afghan army -- on the grounds that COIN would take too long and exhaust the public's patience. The other skeptic was Gates, who had stayed on as defense secretary. He had been deputy director of the CIA when the Soviets crashed and burned on Afghanistan's forbidding terrain, and he worried that if the U.S. footprint got too large and intrusive, history might repeat itself.

By late summer, two things had changed Gates' mind. The first was an article in The Weekly Standard called "We're Not the Soviets in Afghanistan," by Kagan (who had made the case for the surge in Iraq), which noted that the Soviets had rolled in with brute force, that their arsenals contained no precision weapons, and that their soldiers had had no experience with COIN. The United States would do things differently. The second was a 66-page secret report by McChrystal, who had conducted his own policy review, with his own team of COINdinistas, after arriving on the scene.

The report was leaked to The Washington Post in September, just as the White House was in the midst of another review of its policy on Afghanistan. Obama's top aides were infuriated, seeing the leak as additional pressure from the generals to jump into a full-scale COIN campaign. At the end of the year, Obama announced a compromise decision: McChrystal would get his Afghan surge (33,000 U.S. troops plus 7,000 more from NATO allies), but they would have a brief window in which to operate; withdrawal would begin in mid-2011.

The new commander was not a newcomer to COIN theory, as some believed at the time. At West Point, where McChrystal was two years behind Petraeus, his favorite course had been the one on revolutionary warfare. But his recent experience lay in the kill-and-capture realm, which he had revolutionized during the Iraq war as head of the Joint Special Operations Command. When reintroduced to COIN as he took command in Afghanistan, he embraced it with the zeal of a convert.

McChrystal's tenure was not a success. He embraced COIN with a rigid literalism. Galula had defined counterinsurgencies as 80 percent political and 20 percent military action; McChrystal, in his official "COIN Guidance," put the split at 95/5, weighing down his officers with incapacitating rules of engagement. His intention was to put an end to the free-fire zones and strafing that his predecessor had encouraged. But he had an oddly mechanical view of the strategy's workings. Planning his first offensive, in Marja in early 2010, he thought the fighting would be over in a week or so, and then, as he told Dexter Filkins of The New York Times, "We've got a government in a box, ready to roll in." Nothing was so simple. The fighting persisted; the "government in a box" was illusory.

That summer, McChrystal resigned after Rolling Stone published an article quoting him and his staff making crude comments about senior U.S. officials. Obama appointed Petraeus to take his place. If anyone could make COIN work in Afghanistan, it was the general who had seemed to make it work in Iraq.

The problem was that no one could make it work in Afghanistan. Sometimes, Petraeus himself seemed to understand this. One chapter of Galula's book (which he continued to consult) is titled "The Prerequisites for a Successful Insurgency." The conditions it describes include a weak or corrupt government; a neighboring country that offers safe havens; a predominantly rural, illiterate population; and a primitive economy -- precisely the traits that marked Hamid Karzai's Afghanistan. In the PowerPoint briefing that Petraeus delivered to countless delegations visiting his Kabul headquarters, he titled one slide "Storm Clouds," listing all these factors as ill omens for the war's outcome.

But Petraeus' optimism burst through all too persistently. Petraeus recognized the obstacles, but surmounting obstacles was his specialty; it was what he did, most recently, and remarkably, in Iraq. Intellectually, he understood that the two wars and the two countries were very different; his PowerPoint briefing included a slide that read, "Afghanistan Is Not Iraq." But Iraq was what he knew best, so it was natural for him to view Afghan problems through an Iraqi prism. His instinctive reaction to each new challenge was to seek a parallel from his years in Iraq (We solved that this way in Mosul. . . . We did this when that happened in Anbar. . . . I said this when Maliki threatened to do that). Once he drew a comparison between Kabul and Baghdad during a conversation with Karzai himself. Afterward, one of his aides, who had worked in both countries, told him bluntly, "Don't talk about Iraq so much," adding, "It might be a great mental exercise for you to try not thinking about Iraq at all." Petraeus nodded and said, "I'm working on it."

The fact was that there were no parallels to what had facilitated COIN in Iraq: the sectarian and tribal divides were more complex; there was no foundation for, say, a Pashtun Awakening; the main enemy, the Taliban, was homegrown, not foreign. The problem wasn't Petraeus or McChrystal or even Karzai; it was that Afghanistan was not susceptible to COIN. Eventually, Obama recognized this. He had endorsed a surge and COIN in Afghanistan provisionally, giving the policy 18 months to produce results. The generals assured him, with more hope than analysis, that this was feasible. When it proved otherwise, he pulled out the surge troops and scaled back COIN to Biden's counterterrorism-plus. Meanwhile, he had killed Osama bin Laden and decimated al Qaeda's ranks, so he could do all this while declaring victory.

Back in February 2006, at the workshop that Petraeus held at Fort Leavenworth to discuss the COIN field manual, some attendees had questioned the whole enterprise. Were the historical precedents for COINCIA -- mainly, colonial wars against Maoist insurgents -- relevant against messianic jihadists in failed but sovereign states? If the main goal of a COIN strategy is to help the local government attain legitimacy, what does legitimacy mean and how can it be promoted? And is the necessarily protracted nature of a true counterinsurgency campaign plausible given the U.S. political system's demand for quick results? Petraeus had written in his dissertation that "Vietnam was an extremely painful reminder that when it comes to intervention, time and patience are not American virtues in abundant supply." Were he to update his thesis, he might note that Iraq and Afghanistan are now such reminders as well.

THE FIRE NEXT TIME

As the Age of Petraeus comes to a close, what lessons can be learned? A COIN approach did help produce stunning results in parts of Iraq and Afghanistan, most notably Mosul, Tal Afar, and Anbar Province, whose Awakening then spread to many other Sunni districts. The distinctive thing, however, was that in these areas, the Americans and the local authorities -- the mayor, the provincial council, or tribal elders -- shared common interests or at least common enemies. But in both Iraq and Afghanistan, the central governments, which COIN was ultimately supposed to strengthen, were another matter. If the identity and interests of a government obstruct the regime's willingness or ability to govern its people with legitimacy, and if the intervening power has little leverage to alter this fact, then a counterinsurgency campaign may be futile.

The hubbub over Petraeus and his COIN field manual was always overblown. Counterinsurgency is a technique, not a grand strategy. Field manuals are guides for officers preparing to fight in specific settings, and in that sense, a COIN field manual isn't so different from a field manual for mountain warfare, amphibious operations, or armored combat. If the setting is appropriate and the conditions are favorable, a good field manual can provide a road map for success. But if a mountain is too steep to climb, or a beach is too turbulent to storm, or a field is too cluttered for tanks to maneuver across, then even the best manual won't help much -- and it is a commander's responsibility to say so.

In assessing the prospects for a COIN campaign, if the insurgents are out of reach, or if the government being challenged is too corrupt to reform, or if the war is likely to take longer and cost more than a president or a nation is willing to commit, then here, too, it is the commander's responsibility to say so. Few U.S. presidents have plunged into a counterinsurgency on purpose, yet it still tends to happen, one way or another, every generation or so -- at intervals just long enough for the lessons of the last such war to be forgotten. It would be good, then, for this generation's officers, and politicians, to set down the lessons of these COIN wars, so that next time around, the United States might not only fight them more effectively but, more important, calculate more wisely whether to intervene in the first place.