Military history



(Fall 2005-Fall 2006)

In February 2006, Petraeus convened a meeting at Fort Leavenworth of about 135 experts on irregular warfare to discuss a new manual for the Army and Marine Corps about how to conduct counterinsurgency operations. When he called the session to order, he looked out across a tiered classroom in Tice Hall, a squat, one-story brick building in a corner of the base, not far from Leavenworth’s forbidding old gray federal penitentiary. Usually used to train National Guard commanders, on this day the classroom held not just military officers but also representatives from the CIA and the State Department, academics, human rights advocates, even a select group of high-profile journalists. It was instantly clear that this wasn’t going to be the standard Army manual written by two tired majors laboring in a basement somewhere in Fort Leavenworth. “I thought the most interesting thing was the range of attendees, which spoke volumes about Petraeus,” said Eliot Cohen. The two had known each other since the mid-1980s, when Petraeus, then a major, was teaching at West Point, and Cohen at Harvard.

What wasn’t clear was whether the manual would be produced in time to make a difference in the Iraq war, which the attendees knew was heading south fast. One of those present, Kalev Sepp, was fresh from doing his study in Iraq for Gen. Casey of how well U.S. commanders in Iraq had absorbed counterinsurgency theory. His worrisome conclusion had been that 20 percent of them got it, 60 percent were struggling, and 20 percent were trying to fight a conventional war, “oblivious to the inefficacy and counterproductivity of their operations.” In other words, more than half of the U.S. war effort was wasted, and a good part of it was actually hurting the cause.

Petraeus, aware of that troubling finding, began the conference by noting a fundamental difference. In the past, he said, the Army had taught officers what to think. Now, he said, it needed to teach them how to think. Then he sat down next to Sarah Sewall, director of Harvard University’s Carr Center for Human Rights Policy. Just that act in itself made it clear that this effort wasn’t going to follow the usual way the Army devised doctrine.

Conrad Crane, the Army historian, kicked off the discussions by handing out more than a hundred small, hard pieces of green stones with red veins in them. It was coprolite. “They’re pretty, polished, like gem stones,” he told the audience. But, he explained, coprolite is actually fossilized dinosaur excrement. This, Crane warned, was what he didn’t want the new counter-insurgency manual to be: a new polishing of old crap. “There has never been an Army manual created the way this one was,” he said later. “It was truly a unique process.”

Until Petraeus arrived at Leavenworth, its magazine, Military Review, had been a backwater even in the sleepy world of official military publications. Under his command, Col. Bill Darley, its editor, quickly turned it into a must-read bimonthly dispatch from the front lines. It opened its pages to the views of young officers angry over how generals were fighting the Iraq war. The magazine sometimes made news itself. One of its most controversial articles had been by a British officer, Brig. Nigel Aylwin-Foster, who accused the American military in Iraq of cultural ignorance, moralistic self-righteousness, unproductive micromanagement, and unwarranted optimism. He specifically charged that the Americans displayed such “cultural insensitivity” in the war that it “arguably amounted to institutional racism” and may have spurred the growth of the insurgency. Most relevant to Petraeus’s purposes, he contended that the U.S. military’s handling of Iraqis “exacerbated the task it now faces by alienating significant sections of the population.” The meeting kicked off with Aylwin-Foster standing up to review and expand on his explosive charges.

The human rights specialists present were upset by a passage in an early draft of the counterinsurgency manual that was ambiguous about the use of torture in interrogations. It seemed to say that sometimes extreme measures might be deemed necessary, but they were still immoral, so any commander allowing such practices should then confess to a superior officer. Crane and his confreres already harbored doubts about that section and immediately agreed to strike it.

One purpose of the meeting was to ensure that the manual would stand up to such criticism; another was to build support for it. “I think that is always the way to go—inclusion is generally the appropriate course of action,” Petraeus said later. “Frankly, if you can’t convince 95 percent of the rational thinkers, perhaps the concept needs to be reexamined.” He also saw it as a team-building exercise, he added, for the people who would be writing sections for Conrad Crane to get to know one another and how they thought.

Petraeus watched and listened while Crane “played ringmaster,” running the discussions. “I was both physically and mentally exhausted by the end” of the two-day session, Crane said.

The manual that would be produced in the following months adhered to the classic tenets of counterinsurgency—yet in doing so it was prescribing a radical shift for the U.S. military. Historically, Americans have liked to use “overwhelming force,” which under Gen. Colin Powell’s influence was elevated to a first principle. But counterinsurgency, according to David Galula, the French officer who while at Harvard in 1963 wrote what is probably the best book on the subject, requires that the minimum of firepower and force be used. Galula also admonished that the people are the prize. “The population . . . becomes the objective for the counterinsurgent as it was for his enemy,” he wrote, drawing on his experiences in World War II, the French war in Indochina, and the Algerian war, as well as his firsthand observations of the Greek civil war and Mao Zedong’s Communist campaign in China.

If the manual were to have the desired effect, it needed to offer a sharp critique of how the U.S. Army had operated in Iraq for the previous several years. But pointing out the flaws in the American approach was delicate, because this could complicate the task of getting the Army to follow the manual. Many of the generals implicitly skewered in its analysis were still in the Army, and some were running it.

Just a month after the conference, four experts—Crane, Cohen, Lt. Col. Jan Horvath, and Lt. Col. John Nagl, who had studied under Petraeus at West Point—sent up the first flare to the Army and those watching it, signaling that the new counterinsurgency manual would be very different from the usual, small-bore stuff of Army doctrine. The heart of their article in Military Review was a list of the “paradoxes of counterinsurgency.” (This emphasis certainly was influenced by five years of American experience in Iraq—and it is interesting to note that playing with paradox is one of the hallmarks of the classical Arab literature produced in Baghdad at its zenith under the Abbasid caliphs.)

Among the conundrums the article explored:

• “The more you protect your force, the less secure you are.” In other words, it said, you need to get out among the population, because in the long run, that is the way to improve security. “If military forces stay locked up in compounds, they lose touch with the people who are the ultimate arbiters of victory.”

• “The more force you use, the less effective you are.” That is, you are trying to establish the rule of law, and the way to get there is through restraint, whenever possible. Aim for normalcy.

• “Sometimes doing nothing is the best reaction.” This was perhaps the hardest lesson for the can-do, gung-ho U.S. military to take on board. Don’t let yourself be provoked into action, because it may be counterproductive.

The article made clear that Petraeus and the people around him were seeking not only to change the way the Army was fighting in Iraq but also to change the Army itself. Its last paragraph began, “We are at a turning point in the Army’s institutional history.” Petraeus was out to alter how the Army thought about war—a major intellectual, cultural, and emotional shift for a huge and tradition-minded organization.

In June, Crane circulated a new draft of the manual around the Army and Marine Corps. “This was the six-hundred-thousand-editor stage,” he said, referring to the combined active-duty size of the two services. There clearly was a thirst out there for a new approach, reflected in the “thousands” of comments he received.

The manual also would borrow liberally from the work being done that summer by Australian army Lt. Col. David Kilcullen, a quirky infantryman with a Ph.D. in the anthropology of Islamic extremism, a wicked wit, and experience fighting in Timor. Kilcullen came to Petraeus’s attention when he wrote an essay breezily titled “Twenty-Eight Articles: Fundamentals of Company-Level Counterinsurgency”—that is, going one better than Lawrence of Arabia’s famous “Twenty-Seven Articles” on how to fight in the Middle East in 1917.

At the time, Kilcullen’s principles seemed astonishing, and not just because they were articulated with a directness not often seen in public in the U.S. military. His third principle set the tone of the essay: “In counterinsurgency, killing the enemy is easy. Finding him is often nearly impossible.” He also was quite willing to disregard military hierarchy if that was what was required to prevail. “Rank is nothing; talent is everything,” he advised in his eighth principle. “Not everyone is good at counterinsurgency. Many people don’t understand the concept, and some can’t execute it. It is difficult, and in a conventional force only a few people will master it. Anyone can learn the basics, but a few naturals do exist. Learn how to spot these people, and put them into positions where they can make a difference. Rank matters far less than talent—a few good men led by a smart junior noncommissioned officer can succeed in counterinsurgency, where hundreds of well-armed soldiers under a mediocre senior officer will fail.”

His tenth principle poked another stick at top U.S. commanders in Iraq. “The most fundamental rule of counterinsurgency is to be there. . . . This demands a residential approach: living in your sector, in close proximity to the population rather than raiding into the area from remote, secure bases. Movement on foot, sleeping in local villages, night patrolling—all these seem more dangerous than they are. They establish links with the locals, who see you as real people they can trust and do business with, not as aliens who descend from an armored box. Driving around in an armored convoy, day-tripping like a tourist in hell, degrades situational awareness, makes you a target, and is ultimately more dangerous.”

Also, he counseled in rule 26, don’t obsess on fighting your foe. “Only attack the enemy when he gets in the way. Try not to be distracted or forced into a series of reactive moves by a desire to kill or capture the insurgents.”

Petraeus read the cheeky essay and sent it rocketing around the Army via e-mail. Even before the manual appeared, it would begin to affect how some officers thought, perhaps reflecting the pent-up eagerness for change among younger soldiers. One young officer with the 1st Cavalry Division, Lt. Rory McGovern, later recalled that while he was preparing in the fall of 2006 to deploy to Iraq, he was told to read Kilcullen’s “Twenty-Eight Articles.” It changed the way he thought about intelligence operations, he said.

A year later, Petraeus would bring Kilcullen to Baghdad as his adviser on counterinsurgency. There the Australian would explain his role with the memorable comment, “Just because you invade a country stupidly doesn’t mean you have to leave it stupidly,” a comment that Barack Obama adopted in somewhat modified form for his stump speech during the 2008 primaries.

As the writing of the counterinsurgency manual neared completion, Petraeus began editing key portions word by word—and not just once. He made, he remembered, some “twenty or thirty edits.”

Again, this was not the way that the Army usually worked. “I can’t think of a precedent for a commanding general to be so involved in writing doctrine,” said Keane. “It is usually driven by bright young majors.” The hands-on approach helped Petraeus move the product along quickly, and also would make it far more readable—and influential—than most Army manuals.

Published at the end of 2006, just 11 months after the meeting at Leavenworth, the new manual had two striking aspects: It was both a devastating critique of the conduct of the Iraq war and an outline of the approach Petraeus might take there if ever given the chance. In political terms, it amounted to a party platform, the party in this case being the dissidents who thought the Army was on the path to defeat in Iraq if it didn’t change its approach.

• Think twice before launching a raid, it recommended, and consider its consequences. “An operation that kills five insurgents is counterproductive if collateral damage leads to the recruitment of fifty more insurgents.” This was a calculation that had eluded many commanders in Iraq.

• Don’t hole up in big bases, as the U.S. military increasingly was doing in Iraq. “If military forces remain in their compounds, they lose touch with the people, appear to be running scared, and cede the initiative to the insurgents.”

• Don’t let yourself be provoked, as President Bush and other U.S. officials were by the killing of four Blackwater contractors in Fallujah in March 2004. “Often insurgents carry out a terrorist act or guerrilla raid with the primary purpose of enticing counterinsurgents to overreact.”

• Don’t abuse your prisoners, as had happened with the 2004 Abu Ghraib detainee scandal and in many other instances in the war. “Treat detainees professionally and publicize their treatment.”

• Don’t take relatives of suspected insurgents hostage, because it is both illegal and unethical. “At no time can Soldiers and Marines detain family members or close associates to compel suspected insurgents to surrender or provide information.”

• Don’t waste time and money attempting to build a local replica of the U.S. military. “Have local forces mirror the enemy, not U.S. forces.”

• And don’t concentrate on big, capital-intensive reconstruction projects. “Remember, small is beautiful.”

Even discussions that didn’t appear to be about the Iraq war carried clear messages about it. The first “vignette” in the manual—a box inserted in the text that tells an instructive story—is about Marine Gen. Anthony Zinni, who had been a fierce critic of the war and by 2006 was seen by the Bush administration as a political foe. Even more remote, but at the same time even more pointed, is a discussion of Napoleon’s mishandling of his campaign in Spain in 1808. “Little thought was given to the potential challenges of subduing the Spanish populace. . . . Napoleon believed the conquest of Spain would be little more than a ‘military promenade.’ . . . The French failed to analyze the Spanish people, their history, culture, motivations. . . . Napoleon’s cultural miscalculation resulted in a protracted occupation struggle that lasted nearly six years.” All these missteps, of course, were also ones the U.S. military had been committing in Iraq.

The manual also pointed toward the very different approach Petraeus might take:

• “Remain alert for signs of divisions within an insurgent movement.” By the time Petraeus arrived to take command in Iraq, Sunni insurgents were willing to talk to Americans about cease-fires, and he would seek ways to expand on that trend. “Encourage insurgents to change sides.” Sitting down to talk with the evil-doers, as President Bush tended to portray them, would be a radical departure for the American effort in Iraq.

• “At the strategic level, gaining and maintaining U.S. public support for a protracted deployment is critical.” Petraeus would devote much of his time and energy in the coming years to what he called “putting more time on the clock,” especially in his 2007 and 2008 appearances before Congress, which would be the highest-profile occasions of military testimony in decades.

• Most important, “The cornerstone of any COIN [counterinsurgency] effort is establishing security for the civilian populace.” In 2007 this insight would become the starting point for U.S. strategy in Iraq.

Ambassador Ryan Crocker, a veteran diplomat, would read the manual early in 2007 as he prepared to go to Iraq as Petraeus’s civilian counterpart. “If only that had been published in 2002,” he thought to himself.


While the counterinsurgency manual was coming together, Iraq was falling apart. When Iraq’s parliamentary elections were held at the end of 2005, they seemed to many to mark a major turning point. Bush administration officials, buoyed by the photographs of smiling Iraqis holding up fingers inked with the purple of the ballot booth, eagerly greeted the election as a major victory. Vice President Cheney, who 10 months earlier had declared the insurgency to be in its “last throes,” used the occasion to make his first postinvasion visit to Iraq. “I think we’ve turned the corner, if you will,” he told a group of Marines. “I think when we look back ten years hence, we’ll see that the year ’05 was in fact a watershed year here in Iraq.” Gen. Casey, the top U.S. commander in Iraq, said it looked to him as though American troop levels could begin coming down in the following months.

In hindsight, the December 2005 elections were part of the problem, not the solution. “We needed elections in the worst kind of way in 2005—and we got them,” Maj. Gen. David Fastabend, who would become Petraeus’s chief of strategy in Iraq, wrote in a memo to his boss. Most notably, because Sunnis largely boycotted the vote, they planted Shiite-dominated governments in majority Sunni areas—and it would be those areas that would become most resistant to the Baghdad government. Also, less noticed, the elections encouraged U.S. commanders and planners to become overly optimistic. They began formulating plans for major drawdowns in 2006. Most significantly, by holding national elections without any other political structures in place, the U.S. government inadvertently herded Iraqis toward sectarian identification. In the 9 primarily Shiite provinces, the leading Shiite party, the United Iraqi Alliance, won 70 of 81 seats.

The Kurds swept the 35 seats in their region, and Sunni parties won 15 of 17 seats in al Anbar and Salahuddin provinces. The election results in Baghdad, Nineveh, Diyala, and Kirkuk also resembled the sectarian makeup of each province. This may have helped light the fuse of the small civil war that exploded in Baghdad months after. As Petraeus himself would put it much later, “The elections hardened sectarian positions as Iraqis voted largely based on ethnic and sectarian group identity.”

Neither Cheney nor anyone else knew it, but 2006 would, in fact, prove to be the crucial year for the war—but not in the way that American officials had hoped or wanted. Rather, 2006 would be the year that American policy ground to a halt, the Bush administration finally conceded that it was on a path to defeat, the American civilian and military leadership in the war was jettisoned, and a new set of commanders—Petraeus and Odierno—installed to execute a radically different strategy. “Iraq came pretty close, I think, to just unraveling in the course of that year,” Ambassador Crocker said.

It would take agonizing months—indeed, the whole of 2006—for that process of assessment and adjustment to occur. Many observers, both Iraqi and American, think the key event was the bombing on February 22, 2006, of the Golden Dome Mosque in Samarra, one of the most important Shia shrines in Iraq, and, indeed, in the world. Maj. Jeremy Lewis happened to be in Samarra, 65 miles north of Baghdad, at 6:44 that morning. He was preparing to go on patrol with some Iraqi National Police when

...we heard an explosion. All eyes turn toward the explosion. You see this plume of smoke going up, and the plume of smoke was right next to the mosque. I guess that was an initial charge, one of the minarets they had blown up or something like that. . . . Then all of a sudden, the mosque just explodes. You blink and shudder and hunch down. You’re thinking, “What the heck happened there?” . . . It was kind of a cloudy day, overcast. Now there’s this huge plume of smoke, a monstrous cloud, and it’s kind of yellowish and black. My gunner says, “Sir, it’s fucking gone! It’s gone!” I’m like, “No, it’s not gone, it’s not gone.” But then the wind carried the plume of smoke away and you just saw the rebar and everything from the mosque.

Lewis and his comrades battened the hatches. “Every last one of us said this was the beginning of the civil war in Iraq,” he recounted.

Hundreds of Iraqis would die in sectarian fighting in the following weeks, and many American commanders came to see the mosque bombing as a major turning point. But some officers and many observers argue that the incident simply was the culmination of a worsening trend that top officials weren’t grasping, in part because of their focus on developing Iraqi security forces rather than on the situation of Iraqi civilians, and also because they didn’t have troops living among the people. As James Miller, a former Pentagon war planner, put it, “The mosque bombing was just gasoline on a fire that already was burning pretty well.” Indeed, according to the U.S. military’s database of “significant acts,” violence had increased at a steady pace since March 2005 and would continue to increase at about the same pace after the mosque bombing until peaking in June 2007. In 2005 and the following year, hundreds of thousands of Iraqis had fled the country. Many of them were doctors, lawyers, teachers, and journalists, the professionals who are part of the glue of a modern civil society. Because they had both money and education, they were targeted for kidnappings and murders by criminal gangs and political extremists alike. “The situation in the last six months has gotten so bad, we couldn’t continue,” Dr. Omar Kubasi told the Washington Post’s Doug Struck as the flow of refugees increased.

The mosque bombing “wasn’t a trigger, it was an indicator,” concluded Col. Christopher Holshek, a civil affairs veteran of Iraq. “All that did was expose some of the weaknesses in our approach.”

The real effect of the bombing, added Jeffrey White, a former analyst of Middle Eastern affairs, was that it compelled U.S. commanders to deal with reality. After that day, it would become harder for them to argue that there were enough troops, because they had been given the additional mission of containing Shiite militias, on top of the existing tasks of countering the Sunni insurgency and training Iraqi security forces.

One Army officer who recalled buying into the optimism of late 2005 and early 2006, when he was a commander in Iraq, reluctantly agreed. In retrospect, he said, the situation had been far worse than he and his peers had understood it to be. It was the Samarra bombing that led him to believe that Iraq was indeed caught in a civil war: “What Samarra came to mean for me was a defining point in time, almost like a teaching point, where the real face of the Iraq war became clear.”

Kilcullen, the Australian counterinsurgency theorist, was at Petraeus’s conference at Fort Leavenworth when the Samarra bomb exploded. He immediately left for Baghdad, catching an Air Force cargo plane for the last leg. As he landed two days later, he said, “I felt like I was riding a C-130 into Hell. I mean, everything was burning.”


One of the questions raised by the bloodshed of 2006 was whether it was a revealing preview of what would happen if the U.S. military withdrew altogether. Gen. John Abizaid, at the time the chief of the U.S. Central Command, had argued for years that the U.S. military presence was an irritant to the Iraqi population. Yet as U.S. forces had pulled back from Baghdad in 2005, as part of a consolidation effort, violence actually had increased. There were 130,000 U.S. troops in Iraq, but they were becoming increasingly irrelevant as fighting swirled around the tall walls of their bases. To a surprising degree, they were offstage and ill informed. U.S. military intelligence gathering tended to focus on two sorts of events: anything that affected American troops and the killing of Iraqis. Other actions affecting Iraqi civilians—kidnappings, rape, robberies, acts of extortion, and other forms of intimidation—didn’t appear to be on the U.S. military’s radar. As one soldier in the 4th Infantry Division dismissively put it, all that was “background noise.”

As 2006 opened, there were almost no U.S. troops present on the streets of Baghdad, which U.S. commanders were trying to turn over to Iraqi forces. “We have become reactive,” warned Capt. Zachary Martin. “With our fortified bases and our few secured major supply routes linking them, we have immobilized ourselves and cut ourselves off from the battlefield—the populace of Iraq.” Many commanders, he charged, seemed more concerned with “force protection” than with winning.

American officials believed they were turning security over to Iraqi forces. But those Iraqi troops and police weren’t able to control the streets, which meant that the American commanders really were abdicating responsibility and letting the streets become a battleground for sectarian groups. The Americans wouldn’t enforce order and the Iraqis couldn’t—even if they wanted to. As one Army major put it early in 2006, the capital resembled the pure Hobbesian state where all are at war against all others and any security is self-provided. Iraq appeared to be slipping steadily toward chaos. On one day, 22 civilians would die in a car bombing in Baghdad. On another, 17 policemen would be slain. By the third anniversary of the war, on March 19, it was clear that Iraq—or at least greater Baghdad, home to about a quarter of the nation’s population—was on the edge of a civil war.

Enemy attacks were growing in both number and sophistication. In March, 17 policemen and security guards were killed in a dawn raid on a police station in Muqdadiyah that also released 33 prisoners. The attackers, numbering about 100, also set fire to a courthouse, destroyed 12 police cars, and held off an outnumbered U.S.-led counterattack. A month later, the enemy double-bombed an American outpost in the same province. The attack began with a truck exploding against a wall, clearing an opening for a second truck to barrel into the base and detonate against a security wall, which, improperly placed, fell over onto a building, killing 9 American troops inside.

In some U.S. Army units, commanders seemed simply to be keeping their heads down and plodding forward. “It is like we are on a combat patrol and what we see are all the indicators of an ambush—and yet we continue forward as if we had not been trained to detect, avoid, or take preemptive measures,” said one Army colonel in Iraq who was versed in counterinsurgency theory.

Despite Gen. Casey’s efforts with his new Counterinsurgency Academy, abuses by American soldiers, while less common than in 2003-4, still occurred. Even Petraeus’s old division, the 101st Airborne, which had posted a nearly spotless record while under his command in northern Iraq in 2003-4, ran into ugly trouble in 2006. In March, two of the division’s soldiers raped and murdered a fourteen-year-old Iraqi girl and killed her parents and sister to cover their crimes. In May, other 101st soldiers killed three detainees they had captured and handcuffed. Ultimately, four of them would be charged. One of their comrades, Pfc. Bradley Mason, would later testify that they had threatened to kill him if he reported their action.

The court-martial took an illuminating turn: The accused cited the aggressive tone set by their brigade commander, Col. Michael Steele, whose ham-fisted approach long had raised eyebrows in the Army. Retired Army Col. James Hallums, one of his predecessors in commanding the same unit, and himself a veteran of much combat, commented, “The supermacho image that Steele projected permeated his unit, and in my opinion, led directly to atrocities.” When the brigade deployed, Steele, whose role in the fighting in Somalia in 1993 was captured in the book and film Black Hawk Down, had given a speech that was captured on videotape by documentarians following the unit. “Anytime you fight, you always kill the other sonofabitch,” he had told his soldiers. “Do not let him live today so he will fight you tomorrow. Kill him today.” When you go to Iraq, he added, “You’re the predator.”

The fight would be won, Steele told his men, by those who “get violent the fastest.” The counterinsurgency manual then being written advised almost the opposite: “Sometimes, the more force is used, the less effective it is.” The manual also would recommend that prisoners be treated kindly, the better to obtain information from them and perhaps even get them to change sides. Steele was from the old school, telling his soldiers that ensuring that prisoners were shaded from the sun and given water was “bullshit.”

The documentary, made by John Laurence, a veteran war correspondent, captured how one of Steele’s sergeants interpreted that approach. Speaking to his soldiers before a raid, the sergeant instructed them, “We are not bringing anyone back alive.”

BY MID-2006 insurgents were detonating about 1,000 roadside bombs every week, according to the U.S. Central Command. Much of the U.S. effort was focused on countering those attacks. Meanwhile, large numbers of Iraqis were being slaughtered almost daily. Insurgents who later changed sides would report that during 2006, primacy in their movement shifted from former members of Saddam Hussein’s regime, some of whom were running low on cash, to al Qaeda, which “came in with a lot of money and bought away the young men,” reported Maj. Joel Rayburn, an intelligence officer who would later work for Petraeus.

On May 7, car bombs killed about 30 people in the Shiite Muslim holy city of Karbala, 60 miles southwest of Baghdad. One of the car bombs was heading to a major mosque when it exploded in traffic—an echo of the Samarra attack 10 weeks earlier. On the same day, 51 bodies were found in Baghdad, handcuffed, blindfolded, and shot. A week later, 9 bombs detonated in the capital, killing 37. Six days later, a pickup truck loaded with explosives blew up in a crowd of day laborers in Sadr City, the huge slum on the eastern side of Baghdad. Called “Saddam City” until the American invasion, it was almost immediately renamed for the father of Moqtada al-Sadr, the radical Shiite cleric. It became the heart of the son’s growing power. The bomb that day killed or wounded 99 people. Also, 6 policemen were killed by a bomb in the town of Qaim, and police found 40 bodies.

The slide into anarchy accelerated at the end of May after Nouri al-Maliki, a second-tier Shiite politician who was the grandson of a pre-Saddam minister of education, became the compromise pick for prime minister, ending months of stalemate between Shiite leaders. Gen. Chiarelli, the number two 2 U.S. commander in Iraq at the time, argued later that the U.S. effort went off track not because of February’s Golden Dome Mosque bombing but because of the six months of drift that occurred when the elections didn’t quickly lead to the selection of a prime minister. During that time, he said, Americans kept saying that the government, when it formed, would be a force for reconciliation. “We had said it so long, we believed it,” he said, the regret evident in his voice. Instead, he explained, the selection of Maliki may have been the starting gun for a small civil war because the Shiites no longer felt they needed to be on their best behavior with the Americans. They finally held power, and also had the Americans backing them up. So why hold back?

“All hell broke loose,” Chiarelli recalled. On May 30, another 51 people were killed in bombings. Inexplicably, American officials blithely continued to talk about drawing down the U.S. troop presence and turning over control of security to Iraqi forces. Such talk begged the question: If well-equipped and well-trained American troops couldn’t control the situation, why would a new, divided, and distrusted Iraqi police and army be able to do any better?

In the following 12 months, the Army’s 24th Transportation Battalion sent over 400 convoys north from Kuwait and across Iraq, and was hit 170 times. “Every time you left the gate, it was a greater than one-in-three chance that you were going to get hit,” said Maj. Dan Williamson, the battalion’s executive officer.

The U.S. intelligence community warned at this time that a cycle of “self-sustaining violence” had begun in Iraq, recalled Feaver, the NSC aide. As criticism of the U.S. strategy mounted, he added, “I was finding it harder to answer these critiques.”

What was happening was that the “strategic edifice” of the American effort in Iraq was collapsing, Col. Peter Mansoor, Petraeus’s executive officer, later observed. But, soft-spoken in his steel-rimmed glasses and short graying hair, he added with dry underestimation, “It took a few months longer to realize it.”


Back in Washington, the feeling of deterioration in the war was intensifying. One Pentagon official recalled the dysfunctional dynamic of the Bush administration that spring. “The president would say, ‘Get this done,’ and leave the room,” he recalled. “And then Rumsfeld would start squabbling with Condi—‘We’re not gonna secure your PRTs!”’—a reference to the State Department-led Provincial Reconstruction Teams that were at the heart of the strategy of rebuilding the economy of Iraq from the bottom up in order to improve security and so eventually reduce the American military presence. His thought on Rumsfeld at that point, he said, was, “Well, you fucking idiot, that’s your ticket out of Iraq.”

Officially, all was going well. “Iraq is making steady progress in meeting the president’s short-term and medium-term security goals,” said a bizarrely cheerful assessment of Iraq released by the Bush administration in April. The elections and subsequent security operations had led to “a political process that now includes all of Iraq’s major communities for the first time.”

But behind the scenes, a rift was developing between senior commanders in Baghdad and their bosses back in Washington about how to see the war. “It was clear to me that it had shifted from an insurgency against us to a struggle for power, that it wasn’t any longer totally about us, it was about them,” Gen. Casey said in 2008 as he looked back at that time. He was sitting in his Pentagon office under a portrait of Gen. Ulysses S. Grant, the Old Testament patriarch of today’s Army. When he sought to convey that sense of an altered war to officials at the Pentagon and White House, he got blank stares. “We tried until we were blue in the face to get folks [in Washington] to understand that the struggle had fundamentally changed. . . . I always felt I wasn’t conveying it in a way that people could grasp it.”

Officials at the White House were likewise beginning to lose faith in their military interlocutors. “We could just see the Samarra thing spiraling, and no progress on getting a government,” recalled Feaver, who had become one of the key NSC staffers working on Iraq issues. “As we see the situation eroding, there’s a growing question: Do we have the right strategy, is it going to work? And is it time for Rumsfeld to go? This is at the White House staff level—this is people talking at the water cooler. The nub of this is, who can replace him?”

The continuing lack of realism in official statements was one of the factors that precipitated the “revolt of the generals,” which really was just a few retired officers going public with concerns, albeit ones that had grown fairly widespread in their peer group. The wave of criticism began on March 19, when retired Maj. Gen. Paul Eaton, who had been the first overseer of the Iraqi military training effort, wrote an opinion piece for the New York Times that essentially made the Army’s case against Rumsfeld. More troops were needed in Iraq, the senior military officers around Rumsfeld were too pliant, and the defense secretary was “incompetent strategically, operationally and tactically,” he wrote. There wasn’t much new in these assertions. Their significance was that they were made by a general who had been on active duty in Iraq. If Bush was simply heeding the advice of his generals, as he had so often asserted, then why was this one calling for his defense secretary to be fired?

The next officer to jump Rumsfeld’s ship was retired Lt. Gen. Gregory Newbold, who went public with a similar critique in Time magazine. Like Eaton, he had been on active duty during the run-up to the war, serving in the key position of operations director for the Joint Chiefs of Staff. Arguing that the Iraq war was a mistake, he took aim at the entire Bush administration, which he blamed for a series of failures. Among them were distorting intelligence, micromanaging the war, alienating allies, failing to retain the Iraqi army, and denying that an insurgency existed. Again, this came not from any one of a thousand retired officers, but from someone who had seen Rumsfeld up close.

Other retired generals decided it was time to speak out. The third blow came from another officer with credibility gained from firsthand experience, both in the Pentagon and in Iraq. Retired Maj. Gen. John Batiste had been the senior military assistant to Paul Wolfowitz, who as deputy secretary of defense had been one of the leaders of the drive to war. Batiste then had become commander of the 1st Infantry Division, leading it to fight in Iraq in 2004-5. It also was widely known that he had been offered a promotion to lieutenant general to return to Iraq as the number two officer there but had declined because he no longer wished to serve under Rumsfeld. “I think we need a fresh start” at the top of the Pentagon, Batiste said. “We need leadership up there that respects the military as they expect the military to respect them. And that leadership needs to understand teamwork.”

Reporters soon found more generals willing to criticize the administration. Retired Maj. Gen. Charles Swannack, who had commanded the 82nd Airborne Division in Iraq, laid responsibility for the Abu Ghraib scandal at Rumsfeld’s feet, saying it was the result of top-level pressure to step up interrogations. Retired Gen. Anthony Zinni and retired Maj. Gen. John Riggs, who previously had questioned the handling of the war, now were recast as members of a growing group of dissident officers. Riggs made the point that three years into a war that was intended to end in weeks or months, it was growing increasingly difficult to believe the Bush administration’s explanations of events. “I think they’ve made fools of themselves, and totally underestimated what would be needed for a sustained conflict,” he said.

Rumsfeld’s response was to sidestep the substance of the criticism and instead belittle the critics. Rather than respond to the fact that people who had seen him operate firsthand were offering heartfelt—if angry and tardy—commentary, he acted as if they were a few inevitable if inexplicable malcontents. “I don’t know how many generals there have been in the last five years that have served on the United States armed services—hundreds and hundreds and hundreds. And there are several who have opinions. And there’s nothing wrong with people having opinions. And I think one ought to expect that. When you’re involved in something that’s controversial, as certainly this war is, one ought to expect that.” That anodyne comment was fundamentally dishonest because it didn’t answer the nagging question. Nor did Rumsfeld’s dismissiveness serve his president well. Bush had been saying since the start of the war that he relied on the judgment of his generals, and these were generals whose opinions mattered because of their personal experience in Iraq or with Rumsfeld. To swipe aside their collective judgment could only deepen the public’s lack of faith in Bush and those around him.

The generals’ revolt may have been most significant for the irritated response it provoked from a peeved President Bush a few days later. At the end of a press conference to announce the appointment of a new director of the Office of Management and Budget, Bush was asked about talk that Rumsfeld might be forced out by the officers’ criticism. “I don’t appreciate the speculation about Don Rumsfeld,” Bush said. “He’s doing a fine job.” As for the generals, he said, “I listen to all voices, but mine’s the final decision. . . . I hear the voices, and I read the front page, and I know the speculation. But I’m the decider, and I decide what is best. And what’s best is for Don Rumsfeld to remain as the secretary of defense.” With that awkwardly put comment, Bush gave Rumsfeld another seven months in office.

Just as the generals’ revolt was simmering down, the killings at Haditha the previous November erupted into a major news story. Investigations had been under way since February and had intensified after Time magazine ran a thorough article in March that cast doubt on official accounts. But the incident jumped onto page one because of comments made during a press conference by Representative John Murtha, a Pennsylvania Democrat and a Marine veteran of Vietnam who had turned strongly against the Iraq war. Midway through the conference, he blurted out that the incident was “much worse” than was understood. “Our troops overreacted because of pressure on them, and they killed innocent civilians in cold blood,” he said. In the following days, amid talk of another Abu Ghraib scandal that would inflict a strategic wound on the war effort, other members of Congress used less inflammatory language but expressed grave concern. The commandant of the Marine Corps, Gen. Michael Hagee, flew to Iraq to address his troops.

Andrew Krepinevich, who had written the Foreign Affairs article that laid out much of the strategy that the United States eventually would adopt, but much later, observed that the center of gravity in Iraq was the Iraqi population. The task is to convince the population that you will protect them, and also that you will win. So, he concluded, allowing an incident such as Haditha to occur, and then dismissing it as routine, as the Marine chain of command had done, was tantamount to “losing a major battle.”

Oddly, the revelations and allegations about the killings provoked less reaction in Iraq—but not for reasons that were good for the American cause. Some Iraqis said they hadn’t heard the news because they lacked electricity. “We live in darkness,” said Muhanned Jasim, an antiques seller in Baghdad. At any rate, he added, “What’s the big news about Iraqis getting killed? We’re powerless to change the situation.”

As Ghasan Jayih, a pharmacist, ruefully and correctly observed, “It’s normal now to hear twenty-five Iraqis are killed in one day.”

Feaver, whose official title on the staff of the National Security Council was Special Advisor for Strategic Planning and Institutional Reform, thought it was time to confront the president with the bad news. The cheerful son of a Lehigh University classics professor, Feaver himself was a political scientist with a full professorship waiting for him back at Duke University, which gave him a bit of freedom. He was in no position to oust Rumsfeld, but he could and did write memos to the national security adviser, Stephen Hadley, urging that the president hear from some sympathetic but worried outsiders. Feaver and a fellow staffer put that meeting together for a day in June at Camp David, the presidential retreat in the hills of Maryland. One of those on his list was his old friend Eliot Cohen, whom he had known since both were at Harvard in the early 1980s.

But then Feaver’s plan was undercut by an unexpected stream of good news arriving from Iraq. It all had started weeks earlier with the arrest in Jordan of Ziad Khalaf al-Kerbouly, a Jordanian customs worker, who confessed that he had helped smuggle cash and supplies to Abel Rahman, who was believed to be the spiritual adviser to Abu Musab al-Zarqawi, the head of al Qaeda in Iraq. Rahman also was thought to be the liaison between the Iraqi Sunni religious leadership and Zarqawi. U.S. Special Operations teams then began using his contact information to find Rahman and monitor his movements. After three weeks of watching, in the late afternoon of June 7, the spiritual adviser was tracked to a farmhouse in a palm grove in the village of Hibhib, about 35 miles north of Baghdad. U.S. forces went on high alert because Rahman had performed certain tasks—the specifics were never disclosed by the U.S. government—that he usually did before meeting with the terrorist leader. An F-16 jet that had been refueling was dispatched to the area. At 6:12 P.M., its pilot released two 500-pound laser-guided bombs that obliterated the hideout. American troops came upon Zarqawi as he lay near the rubble. He was suffocating as his lungs, torn and bruised by the bombs’ blast waves, ceased to function. He died at 7:04. It was a surprisingly swift and merciful end for the man who was believed to have been behind much of the car bombing of Iraqi civilians in the preceding weeks and months, attacks that had killed and maimed hundreds of innocent men, women, and children. Soldiers from a military intelligence unit found not only Islamic religious material, as they had expected, but also a copy of the May 2nd issue of the Arabic edition of Newsweek.


In June 2006, the presidential meeting with those sympathetic war critics came together at Camp David, atop a ridge in the Maryland foothills just southwest of the Gettysburg battlefield. Eliot Cohen, Michael Vickers, Fred Kagan, and Robert Kaplan—the first three men, smart national security experts; the last, an influential journalist—were generally supportive of the war but critical of its conduct. They were invited to tell the president how it might be better run.

Kagan went to the meeting hoping that it would be “a major turning point.” He had believed for years that the war was being mishandled. “Doing the right thing the wrong way” was the phrase that came to characterize the views of his faction of hawks who thought that the decision to invade Iraq had been correct but who were troubled by the U.S. performance since the fall of Baghdad. “Do we have enough troops?” he asked at the meeting.

Cohen, who on the advice of Feaver had given up his customary bow tie for the meeting, agreed that this wasn’t the time to discuss troop cuts, as the generals were doing, but thought Kagan was fiddling too much with the tactical level of operations and wanted the president instead to focus on strategy. “You probably need more people, but the real question is what you do with them,” he said. He also urged the president to get the rest of the U.S. government beyond the military more seriously engaged in the effort in Iraq. Cohen knew that Bush had read his Supreme Command. He wanted to make Bush think about how to deal with his generals—and consider replacing some. For him, the heart of the matter was “different commanders and a different approach.” After the meeting, he would lash himself for not hitting this point as hard as he should have. Also, he said, “You know, the Army is in worse shape than you think.” Bush didn’t respond. Defense Secretary Rumsfeld and Gen. Pace, the chairman of the Joint Chiefs, also listening, squirmed a bit. Pace had proven a weak chairman, seemingly unwilling to stand up to Rumsfeld when other generals thought he should and instead trying to simply ease the discord at the Pentagon between uniformed military and its civilian overseers. He had a reputation for being a good and decent man, but too pliant. His accomplishment may have been of another sort—keeping the Joint Chiefs from going off the reservation when they split with the president later in the year over whether to change the strategy in the war.

Nor had Pace been much of a presence in discussions of that strategy. It was the major war on his watch, but he tended to defer to Casey and Abizaid, the two four-star officers directly involved in its prosecution. The irony of all this was that policy formulation was following the prescribed method, with the hierarchy being observed and all the correct bureaucratic players involved, but the system wasn’t really working. That is, it looked good, but it wasn’t leading to a robust discussion by top officials of the necessary strategic questions. Nor were leaders held accountable and quizzed on their failures. It was only months later, when the prescribed system was subverted and the chain of command bypassed, that a rigorous examination of American strategy in the Iraq war would get under way.

Kaplan used his time to talk about counterinsurgency practices. “Get rid of periodic presence patrols and provide twenty-four/seven security, get out of big bases and deploy smaller units in neighborhoods,” he said. He was ambivalent about increasing the number of troops because he believed that those already in Iraq were being used incorrectly.

Vickers, a former CIA officer, had played a key role in outfitting the Afghan mujahadeen in the 1980s in their struggle against the Soviet occupation, a role later immortalized in the book and movie Charlie Wilson’s War. Reaching back to his time then, he emphasized what strategists call the indirect approach—that is, helping a local ally fight rather than having Americans carry the combat load. Draw down your conventional forces and rely more on elite Special Operators, he said. “You’re on borrowed time with the direct approach,” said Vickers, according to people who attended the meeting.

The conversation flowed freely, and the president enjoyed the brisk dialogue, said Feaver, the NSC aide who helped conceive and arrange the meeting. But it didn’t work as he had intended, which was to confront the president and his key advisers with the worried critiques of loyalists. Bush was riding on good news. Not only had a new government been seated, but just a few days earlier, Zarqawi had been found and killed. And as Bush was listening, he knew something his four visitors didn’t—that he would be slipping away from Camp David just minutes later to make a secret trip to Baghdad, his first since Thanksgiving 2003. (“He was almost a little bouncy,” Kagan said. “I now recognize that he was very excited about the trip he was about to pull.”) So rather than lead to a much-needed review of strategy, the three events effectively combined to reenergize the president’s commitment to the existing one, Feaver said.

Kagan agreed with that assessment. “I think it [the meeting] had no effect. It certainly didn’t change the minds of the principals. It didn’t generate any follow-up.” Rather than a radical change in strategy, he said, “we continued to drift.”

Returning from Baghdad, Bush gave a tempered but upbeat assessment. “I sense something different happening in Iraq,” he said in a Rose Garden press conference. “The progress will be steady toward a goal that has clearly been defined. In other words, I hope there’s not an expectation from people that, all of a sudden, there’s going to be zero violence—in other words, it’s just not going to be the case. On the other hand, I do think we’ll be able to measure progress.”

In fact, the Camp David meeting would have a far greater long-term effect than anyone could know at the time. In the following months, three of the four worried loyalists who had trekked to the presidential retreat would become deeply involved in revamping Iraq strategy. Cohen took the position of counselor at the State Department, where he became a major strategic voice in the government, not just advising the secretary of state but also officials at the Pentagon and at the White House. Vickers, another of Cohen’s former students, became chief of overseeing Special Operations and strategy at the Pentagon. Bush, still taken with Vickers’s role in arming the Afghan rebels, pinged the Pentagon twice to hurry the clearance process for him. Kagan wouldn’t go into the government but would help redesign U.S. strategy in Iraq, both figuring out what to do and then helping sell the new approach to top White House officials.


After the Camp David meeting the situation in Iraq turned sharply worse. The period from mid-2006 to mid-2007 would prove to be the bloodiest 12 months that American troops had seen thus far in the war, with 1,105 killed. Iraqi civilian deaths are harder to determine but were clearly a multiple of that figure. In the summer and fall of 2006, Shiite militias carried out a concerted campaign that pushed Sunnis out of much of Baghdad, which until then had been a mixed city, with Sunnis predominating west of the Tigris River and Shiites to its east.

The battle of Baghdad effectively began at sunrise on Sunday, July 9, when Shiite militiamen, some of them masked, appeared in the Sunni neighborhood of Jihad, near the Baghdad airport. They set up checkpoints on main streets and killed those passersby whose identity cards indicated they probably were Sunni. They shot up a vegetable market. They also went into homes they believed were occupied by Sunnis. All told, about 50 people were slaughtered. “This is a new step. A red line has been crossed,” said Alaa Makky, a Sunni member of parliament. “People have been killed in the streets; now they are killed inside their homes.”

The next day, Monday, Saleh Muhammed, a resident of the Sunni neighborhood of Amiriyah in far western Baghdad, called the police emergency line to report that the leading Shiite militia, the Mahdi Army, was attacking the quarter’s Malouki mosque. He was surprised by the dispatcher’s response: “The Mahdi Army are not terrorists like you. They are people doing their duty. And how could you know that they are the Mahdi Army—is it written on their foreheads?”

A wave of Sunni retaliation followed. Two car bombs exploded in Sadr City, the city’s biggest Shiite neighborhood, home to about 2 million people, killing or wounding nearly 30. On Wednesday, gunmen kidnapped a group of people, apparently Shiites, at a bus station in Muqdadiyah, and murdered 22 of them. The following Sunday, a café filled with Shiites was blown up north of Baghdad, killing 26. On Monday morning, death squads assaulted the marketplace in the mostly Shiite southern Baghdad suburb of Mahmudiyah. They fired heavy machine guns, burned cars, threw grenades, and entered a café to shoot 7 elderly men in the head. At least 40 people were killed. On Tuesday, a minibus loaded with explosives blew up near a Shiite mosque in Kufa, killing 53 day laborers and wounding at least 130 more. Hundreds more Iraqis were dying in smaller incidents. Police in the tough southern Baghdad neighborhood of Doura said 425 people were killed in that area alone during the week after the Jihad marketplace massacre. Altogether, more than 3,000 Iraqis were slain during July, the United Nations estimated. It was the deadliest month in three years.

Army Capt. Don Makay, who fought in southwest Baghdad, recalled that during his tour, every Sunni mosque in his area was attacked, in one case, he thought, with the involvement of the local commander of the National Police. From July through October, the number of murdered bodies dumped near Sunni districts “rose considerably,” wrote another Army captain, Michael Comstock, in his study of the ethnic-cleansing operation. Other Iraqis were luckier, receiving “night letters” that contained a bullet and an order to vacate their homes within a day or two.

The core of the Iraqi state was rotten. The Iraqi army was heavily Shiite, and even worse, the National Police were thoroughly infiltrated by Shiite militias. These forces didn’t have to carry out the cleansing themselves. All they had to do was go into a Sunni neighborhood and demand in the name of pacification that all heavy weapons be relinquished. After that was accomplished, they could tip off the Shiite militias, who might arrive that night or the next morning, ready to take on the newly defenseless population. As one foreign diplomat in Baghdad summarized the legitimate complaint of Sunnis, “You come and denude us of weapons, and the next day the militias visit.”

Nor did the gunmen need to kill everyone—just enough to intimidate the rest. This is how Capt. Eric Haas summarized the tactics of Jaysh al-Mahdi, Moqtada al-Sadr’s radical Shiite militia: “JAM/Shia militia group kidnaps a Sunni male from a mixed-sect market; takes Sunni male to the edge of Sunni-dominated neighborhood; takes Sunni male from the vehicle shot in the back of a head with a pistol; Shia militia drives off.”

Crueler tactics, such as using power tools to drill holes in the kneecaps or heads of victims, also became common. “People are killed here every day, and you don’t hear about it,” Capt. Lee Showman told the Washington Post’s Josh Partlow. “People are kidnapped here every day, and you don’t hear about it.” As the ethnic-cleansing campaign intensified, the number of Iraqis seeking refuge in neighboring nations spiraled, with an estimated 2 million leaving the country. An equal number were classified as internally displaced, with much of that movement occurring in 2006.

But the militias’ work was hardly done once the Sunnis had been driven out. The next step was to turn the neighborhood into a paying concern. First the vacated houses would be rented to Shias. Then kidnapping and extortion rings would raise money from shop owners and other holders of wealth. Shiite party banners would festoon the altered area. Local police would be intimidated, co-opted, or replaced with Shiite militia members who would cooperate. The explicit support and assistance of all civilians in the area was demanded. “Leave, join or die” was the summary offered by Army Capt. Josh Francis. At this point the area might become less violent, but that wasn’t necessarily a positive sign. Instead, it might just mean that the job was done and that the newly quiet neighborhood then could be used as a base from which to begin launching attacks on adjacent Sunni areas.

Sgt. Victor Alarcon watched as his battalion in the 1st Infantry Division lost 20 troops in an unsuccessful effort in 2006 to prevent the destruction of what had been a bustling middle-class Sunni neighborhood. “I don’t think this place is worth another soldier’s life,” he said near the end of his tour.

Maj. Mark Gilmore gave this dismal summary of his time in one Baghdad neighborhood: “When we got there, it was mixed Sunni and Shia. When we left, it was Shia. . . . When we left, it wasn’t even worth counting the Sunnis because there weren’t that many left.”

THE FIGHTING in Iraq wasn’t just sectarian. Two other major players in the tragedy of Iraq were also escalating their activities at this time: al Qaeda in Iraq, and Iran.

In August, Col. Peter Devlin, the senior Marine intelligence officer in Iraq, filed a secret report concluding that the U.S. military had lost al Anbar, in western Iraq, and that al Qaeda was now the dominant factor in the province. “The social and political situation has deteriorated to a point that MNF [MultiNational Forces] and ISF [Iraqi Security Forces] are no longer capable of militarily defeating the insurgency in al Anbar. . . . Underlying this decline in stability is the near complete collapse of social order in al Anbar.” What’s more, al Qaeda in Iraq, which was mainly made up of nihilistic Iraqi religious extremists but also included some foreign fighters, who frequently were used in car bombings, had elbowed aside other centers of power in the province and made itself Anbar’s “dominant organization of influence.” (Devlin’s assessment is reprinted in full as the first document in the appendix.)

To the north of Baghdad, al Qaeda in Iraq, sometimes referred to by the U.S. military as “AQIZ,” launched a swift and viciously effective campaign. “Using a small, localized cell of hardcore believers, AQIZ successfully coerced and intimidated the local populace over time through a four phased plan: clandestine organization, psychological preparation of the people, expansion of control, and consolidation of power,” Army Capt. James Few wrote in a study of the terrorist takeover of the town of Zaganiyah.

A mukthar, or town elder, sought an audience with al Qaeda leaders in the town of Nukisa to complain about the behavior of the organization’s recruits. He was beaten in public, the humiliation intended to demonstrate that there was a new sheriff in town. In November 2006, while American forces were focused on the deterioration of security in Baghdad, al Qaeda members in the town made their move, launching a complex attack on the local Iraqi police station, with a car bomb followed by an assault by fighters. “Over the next two weeks, ISF stopped patrolling the area, and CF [coalition forces] designated Zaganiyah as ‘No-Go’ terrain,” Few wrote. The al Qaeda cell then consolidated its hold, destroying the home of an Iraqi working as an interpreter for the Americans and beheading a captured Iraqi soldier and a local Shiite. They also dug fighting positions around the town and deeply buried more than 160 bombs, establishing a defensive belt.

Meanwhile, Iran, capitalizing on the cover provided by violence and counting on the Americans to be distracted, quietly launched its own offensive in Iraq. Devastating “explosively formed projectiles,” the most lethal type of roadside bomb, began appearing in great numbers in late 2006. These high-tech bombs operate by melting a disk of metal into a spray of high-velocity drops that cut through armored vehicles, frequently killing three or four soldiers in one blast. U.S. intelligence officials said all the devices were imported from Iran. During 2007 they would become the greatest threat to U.S. troops, inflicting 73 percent of all American casualties.

Asked what he would do differently in 2006 if he could, Abizaid, the top American commander for the Middle East, said, “We didn’t react quickly enough to Shia and Sunni violence,” or, he said, to the misdeeds of the Iraqi police.


Finally, in the summer of 2006, the U.S. military and its Iraqi allies launched a major counteroffensive aimed at improving security in the capital. Dubbed “Together Forward,” the operation involved some 42,500 Iraqi police and army personnel backed up by 7,200 U.S. troops. The Iraqi forces were instructed to erect new checkpoints, enforce extended dusk-to-dawn curfews and new restrictions on carrying weapons, and step up the frequency of their foot patrols. Posters were distributed showing an Iraqi soldier in tan battle fatigues holding the hand of a smiling Iraqi boy. But the Americans were operating more and more from big bases, removing themselves from the population and from the civil war being waged beyond the tall cement walls of their isolated bastions. They also continued to judge their actions all too often by input, such as the number of patrols conducted, rather than by output, such as the reduction in violence.

The offensive never really got off the ground. “They were dead in the water by midsummer,” said Krepinevich, the counterinsurgency expert.

For Brett McGurk, a staffer on the National Security Council who was in Baghdad that summer, the failure was a turning point in his view of the approach the U.S. military was taking. “Gaziliyah was probably the best example of a clearly failing strategy,” he recalled. “We go in, MNF-I reports its metrics (buildings cleared, violence reduced), we leave, and violence in Gaziliyah hits all-time highs.” His conclusion was that “it was clearly a failed recipe—the question was whether we could do anything about it.”

Fred Kagan, the defense analyst who had been at Camp David as the summer began, later said the offensive was doomed from the start, because it relied excessively on Iraqi police forces, which he said were part of the problem, not the solution. “They were not and could not be effective bulwarks on their own against sectarian violence of which they were a part,” he wrote.

American commanders would, in fact, blame Iraqi units for the failure. “The loyalty of the Iraqi security forces, particularly the police, was the overriding issue that kept this from being a success,” Gen. Casey said in an interview. A secondary flaw, he said was the slowness with which the Iraqi government moved to conduct follow-on economic aid projects. “It was never clear whether it was incompetence or sectarian bias.”

Chiarelli, the number two U.S. commander in Iraq, added,“ I was under the impression that we would get two additional Iraqi brigades, and they didn’t show up.” In addition, Abizaid said, the American liaison connection to Iraqi forces needed to be strengthened.

Some in Iraq said that Chiarelli and Casey should have known that the Americans couldn’t rely on Iraqi forces to carry a large part of the burden. “They ordered these Kurdish units to come down,” recalled Maj. Matt Whitney, who at the time was an adviser to the Iraqi Ground Forces Command, a top headquarters. “One of them mutinied. They look for troops in the south and they wouldn’t come either. They looked for two more units from the north and they didn’t come.” He wasn’t surprised by this, because many Iraqi units thought they were supposed to defend the area where they were based. They had neither the training nor the equipment to pack up and move around the country. “General Casey was frustrated because he couldn’t get Iraqi units to deploy, although we never built that army to deploy. Somehow he was surprised by this.”

Together Forward not only didn’t work, it backfired on Gen. Casey, because it undercut the confidence of Bush administration officials in his ability to deliver. “In July, when Baghdad Security Plan One tanked, they said, ‘We didn’t have enough reliable Iraqi units, they didn’t show up,”’ recalled Feaver, one of the National Security Council staffers working on Iraq. “Over the summer, doubts began to grow among White House officials working on Iraq. By September the NSC staff initiated a quiet but thorough review of strategy with an eye to developing a new way forward.”

McGurk, the NSC staffer, returned to the White House with doubts not just about the approach but about the people implementing it. He “had lost all faith in our security strategy. MNF-I and the embassy were locked in a corrosive cycle of finger-pointing . . . with nobody asking serious questions about what to do differently.”

A new iteration, Together Forward II, was launched on August 8. It did nothing to stop the big bombings. Casey called in additional troops from his theater reserve and sent those reinforcements to help clear the city, block by block. The notion was that Iraqi forces then would hold those areas. “Clear, hold, and build” was a phrase that grew out of Col. H. R. McMaster’s successful campaign in the northern Iraqi city of Tall Afar, one of the few bright spots in the war that year. A visiting State Department official picked up the phrase and passed it along to Secretary of State Condoleezza Rice, who used it in congressional testimony. Rumsfeld resisted the phrase, even after the White House adopted it. He argued that it was the job of the Iraqis or the State Department to oversee holding and building, but grudgingly seemed to accept the idea as at least a rhetorical necessity.

There was little reason to believe that the plan to clear, hold, and build in Baghdad would work any better the second time around. Brig. Gen. John Campbell started in Baghdad as the assistant commander of the 1st Cavalry Division on the day Operation Together Forward II began. He watched as attacks rose steadily despite U.S. efforts. “They went through and cleared, and tried to hold that with Iraqi forces,” Campbell said. “The issue was, we didn’t have enough ISF, both in quantity and quality.”

The failure to hold meant that the U.S. military was simply repeating the pattern of 2003-5 that Kilcullen, the Australian counterinsurgency analyst, had labeled “kiss of death” operations, in which American forces moved into an area, found cooperative locals, and then, after some improvement of security, pulled out of the area. “Then,” Kilcullen grimly concluded, “insurgents kill those who cooperated with us.”

White House officials were also concluding that the government of Prime Minister Nouri al-Maliki was impeding success, especially because it wouldn’t allow actions to be taken against Shiite militias, Feaver said. Indeed, after U.S. Army units launched a raid into Sadr City in early August, resulting in a two-hour-long firefight, Maliki angrily appeared on television to apologize for the operation. “This won’t happen again,” he promised. Chiarelli said that Maliki constantly impeded U.S. operations during the summer and fall of 2006. Near the end of the year, for example, U.S. Special Operators would pick up in Baghdad one of the most senior leaders of the Iranian Revolutionary Guard’s Quds Force, the guard’s wing for foreign Islamic revolutionary operations. He was believed to be involved in planning attacks on U.S. forces and was found at the compound of Abdul Aziz al-Hakim, head of an influential Shiite political party that was a big part of Maliki’s ruling coalition and one of the most prominent politicians in Iraq. U.S. officials were furious when a few days later, Maliki’s government sent the Quds man back to Iran.

August ended with two days of ferocious bombings, with 27 people killed in the Shorja market, Baghdad’s largest bazaar, on the thirtieth, and then 66 killed the next day as a huge explosion flattened an apartment building in a Shiite neighborhood. Meanwhile, Shiite militiamen battled U.S. troops both in Sadr City and in the southern city of Diwaniyah. The same month saw a 33-day Israeli war with Hezbollah in southern Lebanon end in what was widely seen as a military and political setback for Israel, an outcome that only further worried analysts assessing the American position in Iraq.

Despite the growing violence, Casey continued to insist on a policy that emphasized transition to Iraqi forces. Late in August, he predicted that Iraqi forces would be able to provide security in the country pretty much on their own by late 2007 or early 2008. “I can see—over the next twelve to eighteen months—I can see the Iraqi security forces progressing to a point where they can take on the security responsibilities for the country with very little coalition support,” he said.

Chiarelli, the number two officer in Iraq, and so commander of day-to-day operations, occupies an ambiguous position in this tale. As the commander of the 1st Cavalry Division in Baghdad during his previous tour, he had done a far better job than most in understanding the principles of counterinsurgency. There were rumors of disagreement between Casey and him over the way forward. Publicly he was entirely supportive of Casey, reiterating in mid-September the view that sending additional U.S. troops was not the answer. “I feel that given the conditions we’ve got in Baghdad, we’ve got the force posture exactly where it needs to be,” he said. On the face of it, there would appear to be little else that he could do, given his subordinate position. Yet just a few months later, Lt. Gen. Odierno would arrive to take over from Chiarelli in that number two slot and effectively challenge Casey as Chiarelli had never done, conducting his own strategic review that ultimately would reverse almost every tenet of American strategy in Iraq. Chiarelli struggled with the number two position; Odierno would redefine it.

Sadi Othman, who would become one of Petraeus’s closest advisers in Iraq, said that in retrospect, neither American officials nor Iraqi leaders understood just how dangerous the situation was in 2006. “I think people knew the situation was bad, but they didn’t know it was very, very bad,” he said in 2008. “The Americans didn’t get out of the Green Zone. The government of Iraq didn’t get out. And we didn’t have troops on the streets. So when people said things were okay, they weren’t lying. They were innocent.”

Chiarelli, in a 2008 interview, disputed Othman’s assertion. In fact, he said, he had gone to Maliki in July “to tell him how bad it was.” The prime minister’s chilling response, he recalled, had been “It was a lot worse in Saddam’s time.” American officers interpreted this to mean that Maliki didn’t intend to do anything to curtail the violence, which was Shiite payback against Sunnis for what had happened before.

Meanwhile, political pressure was building for a radical shift away from Casey’s approach. In September the Iraq Study Group, which had been appointed by Congress to review policy in the war and to make recommendations to improve it, arrived in Baghdad to check its views against the thinking in the Green Zone, the heavily guarded enclave in the center of Baghdad that housed the headquarters of the American effort in Iraq. Many of the study group’s members, such as former congressman Lee Hamilton and former secretary of state James Baker—its two chairs—were more familiar with politics and diplomacy than warfare. Another member, Robert Gates, was destined to become defense secretary just four months later, but nobody knew that then. At the time, there were just two members of the group who knew the military establishment well: former defense secretary William Perry and former senator Charles Robb, a Marine veteran of the Vietnam War and a longtime member of the Senate Armed Services Committee. Both Perry and Robb had come to think that current U.S. strategy couldn’t continue and were mulling advocating a troop escalation.

Before heading to Baghdad, Perry had distributed a memorandum to the group making an argument for such a “surge” in troops in Iraq. “We thought we couldn’t get enough troops to surge the whole country, but we could maybe have an effect in Baghdad,” he recalled. (President Bush said in 2008 that when he interviewed Robert Gates in November 2006 about becoming defense secretary, Gates told him that he also had favored such an increase.)

When the group met with Casey and Chiarelli, the generals threw cold water on the idea of a troop increase. “They were very explicit,” Perry said. “Both Casey and Chiarelli said this would not be useful, as they saw the problems in Iraq.” The officers offered three arguments: First, it would give the Iraqi government the impression that the Americans would solve their problems. Second, it would decrease the leverage the Americans had. Third, whatever improvement it provided wouldn’t last. “They made the point that wherever you put American troops, it would stabilize the situation—but when they left, it would destabilize the situation.”

Perry worried that the group was being given what he called a “party line,” so he asked for separate one-on-one meetings to get the generals’ personal views. In those private sessions, he said, “Both stuck to their guns.”

Faced with the opposition of the top U.S. military leadership on the ground, Perry withdrew the idea. When he wrote the first draft of the military section of the group’s report, he left out the idea of a surge. “It would have been in there if they had responded differently,” he said. Ultimately, the group’s report straddled the idea, rejecting a major increase but conditionally supporting “a short-term redeployment or surge of American combat forces to stabilize Baghdad . . . if the U.S. commander in Iraq determines that such steps would be effective.”

Chiarelli said later that he wasn’t against getting additional forces. “In fact, I already knew where I would put a brigade,” he said in an interview at the Pentagon in 2008 shortly before he pinned on a fourth star and became the vice chief of staff of the Army. But, he added, he knew that it would take time to bring in additional troops. “I thought we could push violence down a lot faster if we went to Maliki” and delivered a strong message: Your policies, such as not delivering services to Sunnis, are exacerbating sectarian tensions. “We need to use our leverage with Maliki,” was his recommendation. At any rate, he remembered, when he arrived in Iraq at the beginning of 2006, he had been told that during that year, the U.S. combat presence would be nearly halved, from 108 bases to 50, and from 15 brigades to as few as 8.

In sum, Casey and Chiarelli were sticking to their approach, even though there was little evidence of it working. The U.S. strategy, concluded Anthony Cordesman, a defense analyst at the Center for Strategic and International Studies, was “deeply flawed in timing and resources. It was based on a grossly exaggerated estimate of political success, an almost deliberately false exaggeration of the success of the economic aid effort and progress in developing the ISF.”

Francis “Bing” West, a former Marine and Pentagon official who had a son fighting in Iraq, put it even more bluntly: “The strategy was a hope posing as a plan.”

By late 2006, agreed Philip Zelikow, who at the time was counselor at the State Department, there was essentially “a strategic void” in Iraq.

Oddly, the White House also decided that this was a good time to attack critics of the war as appeasers and worse. Rumsfeld said they were morally and intellectually confused, not unlike those who had opposed confronting Hitler in the late 1930s. Cheney said those who disagreed with his administration’s approach were abetting terrorists. Bush, a mite more generously, conceded that the detractors were “sincere” and “patriotic,” but said “they could . . . not be more wrong.”

That White House move was an inept political tactic, because it made it appear that the president was divorced from the realities of Iraq and dismissing the legitimate worries of those who believed—with ample evidence—that the war was being mishandled and that it, in fact, was rapidly spinning out of control. During the winter of 2005-6, there had been about 500 attacks a week on U.S. and allied forces. By late in the summer of 2006, there were almost 800. Some 1,200 roadside bombs were detonated in August. The number of roadside bombs was at an “all time high,” conceded Maj. Gen. William Caldwell IV, the U.S. military spokesman in Iraq. The bombings continued like a daily drumbeat, contributing to the capital’s monthly civilian death toll of about 1,000.

In one of the most horrific incidents, on September 23, a bomb exploded as people waited in line to buy gasoline, sending women engulfed in flames running through the streets. Witnesses reported that two young girls embraced each other as they stood in the inferno burning to death. “This deployment, every patrol you’re finding dead people,” Staff Sgt. Ian Newland told Army Times. “It’s like one to 12 a patrol. Their eyes are gouged out. Their arms are broken. We saw a kid who had been shot 10 to 15 times.” Newland’s company arrived in Baghdad in August, and over the next 15 months it would lose 14 men, the most of any Army company to fight in Iraq. In the first week of October 2006, some 24 soldiers and Marines were killed, most of them in Baghdad, and nearly 300 more were wounded. The violence was also spreading, with Shiite militias fighting Iraqi police to the south of the capital and confronting Sunni militias to the north.

Internal Army surveys of the morale of soldiers underscored the feeling of loss. In both 2004 and 2005, studies by an official Mental Health Advisory Team had reported that morale was improving among troops involved in combat. But a September 2006 assessment found a sharp decline.

On October 19, Gen. Caldwell, the U.S. military spokesman, acknowledged that the renewed security effort in the capital was failing. “Operation Together Forward has made a difference in the focus areas, but has not met our overall expectations of sustaining a reduction in the levels of violence,” Caldwell said. “We find the insurgent elements, the extremists, are in fact punching back hard. They’re trying to get back into those areas. We’re constantly going back in and doing clearing operations again.”

Caldwell’s admission might have been the worst point of the entire war, at least so far. The U.S. military had played its ace in the hole—“the sole superpower” had asserted itself in Iraq’s most important city—yet had not been able to reverse the deteriorating security of the capital. What’s more, not only had U.S. commanders taken their best shot and failed, they apparently were going to continue on the same unpromising course of handing off control to Iraqis who didn’t seem competent or much interested in the stability the Americans wanted.

In the midst of all this, in the fall of 2006, Iraqi army and police forces finally hit their targeted size of about 325,000 total—but the U.S. wasn’t able to stand down as they stood up, as the president for years had said would happen, repeating the phrase as late as June of the year. Paradoxically, as the number of Iraqi soldiers and police grew, so did the violence in the streets of the capital. From August through October 2006, the number of attacks in Iraq grew by 22 percent, according to the U.S. military database, which almost certainly undercounted the total but probably was accurate in tracking the direction of the trend.

The Americans seemed to have run out of both troops and ideas. The one possible bright spot in that bankrupt approach was that it created the conditions for the strategic surprise that Petraeus and Odierno would launch a few months later, as they showed both new flexibility and determination. Given the track record of the previous four years, no one in Iraq saw that one coming.

The downward trend continued. In October 2006 an American soldier was kidnapped. American intelligence officials suspected he was being held in Sadr City, the stronghold of Moqtada al-Sadr’s militia, the Jaysh al-Mahdi, or JAM. The U.S. Army, searching desperately for the missing soldier, erected a series of checkpoints along the Canal Road, the broad boulevard that parallels the south side of a densely packed neighborhood slum. Maliki told Casey to lift the checkpoints. “If that’s your order, we’ll do it,” Casey responded. “But people will say you don’t care about American soldiers, and that you kowtowed to Sadr. Third, the Sunnis will read this as a pro-JAM action. Can you accept that?” Maliki said he could.

Casey was reading a history of the Vietnam War at the time and thought of the weak and chaotic governments that American officials had dealt with in Saigon back then. “How do you save a head of state when he is diametrically opposed to the policy you are trying to save him with?” he thought to himself.

Then he called his deputy, Chiarelli, and told him to lift the checkpoints. “This was going on all the time with Maliki,” Chiarelli recalled. “We had certain things we could do in Sadr City, but not what we needed to do.”

Maj. Gen. David Fastabend, hearing about the order to remove the checkpoints, called another general and said, “This is the singular moment of defeat. If you want to know when we lost, this was it.” The ethnic cleansing continued as Shiite militias pushed Sunnis westward. “You’d find dumped bodies every day,” recalled Maj. David Voorhies, who was advising an Iraqi army unit that he believed was infiltrated by Shiite militias. “You’d see murders, a lot of extra-judicial killings, a lot of kidnappings, a lot of demonstrations would arise. Eventually those areas would collapse . . . on Amiriyah and Gaziliyah, which were really the last two big Sunni neighborhoods in west Baghdad.”

The failures of the summer and fall of 2006 may have given the U.S. military establishment the push it needed to realize that everything it had tried over several years wasn’t working, and that—despite the assurances of commanders in Iraq—a very different approach was needed. A major split was developing inside the military about what the next step in Iraq should be. Some called for an accelerated transition to Iraqi control, but others said that would just lead to an intensified civil war. Others called for backing out of Iraq and letting the Iraqis sort it out, and others responded that that move could lead to regional war. And a few, here and there, were thinking about increasing the number of troops and using them differently. One of the significant consequences of this split was that, really for the first time in the war, the Bush administration could no longer blandly state that it was following the advice of the military. By late 2006, there simply no longer was a consensus view to follow. “We may need more resources, but first we need a strategy,” Eliot Cohen and Francis “Bing” West would write a few months later.

Even more significantly, the doubts White House staffers had held about the top American general in Iraq had reached the president. Bush usually was affable in his conversations, but in mid-November, “the president was noticeably cold,” Casey recalled. So, after three years of war, Bush and his aides would be forced into a serious review of their strategy in Iraq. Finally, they would begin to ask some of the basic questions that they had neglected to address before the invasion.


Back in Washington, Jack Keane, the old general who was more influential in retirement than most officers are while on active duty, was growing increasingly concerned as he watched the two Baghdad security operations sputter to a halt. “We had two bites of this apple in Baghdad, and we failed both times,” he said. “I knew that our chances to succeed in Iraq were just slipping by us.” He decided it was time to share his worries with the Bush administration.

The White House was ready to listen to him. Gen. Casey may not have known it, but the failures of the Together Forward operations were the beginning of the end for his command in Iraq. Behind closed doors, the outlook appeared even worse. “Even in the military, there’s a concern right now that wasn’t previously,” said one worried Marine colonel. “Folks that took things at face value in the past are asking more questions.”

Pressure was clearly building for an overhaul of American strategy in Iraq, but a major obstacle stood in the way at the top of the Pentagon. Not long before he was fired, Defense Secretary Rumsfeld insisted that the strategy of passing responsibility to the Iraqi forces was working and needed no change. “The biggest mistake would be to not pass things over to the Iraqis, create a dependency on their part, and instead of developing strength and capacity and competence,” he said at a press conference the day after Caldwell spoke in Baghdad. “It’s their country. They’re going to have to govern it, they’re going to have to provide security for it, and they’re going to have to do it sooner rather than later. And that means they’ve got to take pieces of it as we go along, even though someone may inaccurately characterize it as a strategic mistake, which it wouldn’t be at all.”

Bush would back up Rumsfeld, saying he was flexible about tactics but wasn’t contemplating a change in strategy or goals. “Are we winning?” asked a reporter at an East Room news conference a few days later.

“Absolutely, we’re winning,” Bush insisted. At the same time, he said, “I know many Americans are not satisfied with the situation in Iraq. I’m not satisfied either. And that is why we’re taking new steps to help secure Baghdad and constantly adjusting our tactics across the country to meet the changing threat.”

Feaver, the White House aide, cringed at Bush’s “winning” comment. “That wasn’t the way it felt from where I sat.” He recalled that at this time, Karl Rove, the president’s political adviser, was also speaking up, telling others, “We need a new face on Iraq”—by which he apparently meant that Rumsfeld should leave.

Support for the war was eroding rapidly among the Republican Party faithful. Back in February, John Warner, the courtly Virginia Republican who was chairman of the Senate Armed Services Committee, had expressed “a high degree of confidence” that a new government would take charge and that by the end of the year the conflict “won’t be the same.” But as October opened, Warner returned from Iraq with a far grimmer assessment: “The situation is simply drifting sideways.”

Something had to give, said Senator Olympia Snowe, a centrist Maine Republican. “I don’t believe we can continue based on an open-ended, unconditional presence.”

Senator Lindsey Graham, a conservative South Carolina Republican and a close friend of John McCain’s, was mulling a different strategy. “The American people are beginning to wonder if the Iraqi people can get this right,” he said. “People have begun to wonder about the basic premise, that the Iraqi people are capable of solving their problems politically. We’re at a real crossroads. The level of violence in October just shows you we don’t have enough security to ensure long-term success.”

Others argued that the situation was even more dire than that. “Basically, the bottom has fallen out of support with the general public,” former Republican congressman Vin Weber said later that October, just before the election. “The public is on the verge of throwing up its hands over Iraq. They are right on the edge of believing that success isn’t possible.”


Near the end of Gen. Caldwell’s press conference on October 19, a few minutes after the spokesman had announced the failure of the Baghdad security plan, one reporter had inquired about some odd reports coming out of Ramadi, 60 miles to the west of Baghdad. Specifically, inquired the man from Reuters news agency, why were armed civilians marching in the streets? What was going on out there? Caldwell responded that he hadn’t heard about that and would look into it.

It was a good question, because Ramadi had been one of Iraq’s most dangerous cities for years. This time, to the astonishment of anyone focused on Baghdad, the armed men were not members of al Qaeda in Iraq but allies of the Americans, albeit tentative ones. Ramadi, the capital of turbulent al Anbar Province, had begun to provide a counterexample to Baghdad. That turnaround, led by Col. Sean MacFarland, would take place even as the senior Marine intelligence officer in the country pronounced the province lost. Ramadi in 2006 would become the link between the first successful large-scale U.S. counterinsurgency campaign in Iraq, in Tall Afar in 2005, and the “surge” counteroffensive in Baghdad in 2007.

By chance, MacFarland’s unit first had been assigned to replace the 3rd Armored Cavalry Regiment in Tall Afar, in the far northwest of Iraq, and had spent several months there before moving south to Ramadi. What MacFarland and his subordinates had seen there was very different from how the U.S. military had operated in Iraq for several years. The new approach made sense to him. Under Col. H. R. McMaster, an innovative officer unafraid to chart a different course, the 3rd Armored Cavalry Regiment had slowly and patiently approached Tall Afar, a medieval feeling town of about 250,000. After the U.S. military reduced its presence in northern Iraq in 2004, Islamic extremists had begun to seep in from Syria and make contact with local allies. By mid-2005 they had intimidated the locals with terror tactics and made the town a base from which to send suicide bombers and other attackers 40 miles easy to Mosul, the most important city in northern Iraq. “Give the enemy credit,” said Maj. Chris Kennedy. “As soon as we started pulling back, the enemy identified that as a weak point.”

McMaster, who is both a rugby player and a Ph.D. in history, began by telling his soldiers to treat Iraqis with dignity and respect. “Every time you treat an Iraqi disrespectfully, you are working for the enemy,” he instructed them—neatly summarizing counterinsurgency theory in a way that any nineteen-year-old infantryman could grasp. In a marked contrast to the attitude found in some other units, his standing orders required his soldiers to “Treat detainees professionally; do not tolerate abusive behavior.” He met with sheikhs and clerics who had ties to the insurgency and apologized for past American mistakes: “When the Americans first came to Iraq, we were in a dark room, stumbling around, breaking china. But now Iraqi leaders are turning on the lights.” And, he added, the time for honorable resistance had ended.

Then, after months of preparatory moves in the desert around the city, cutting off lines of retreat and safe havens, McMaster attacked Tall Afar. Rather than just stage patrols from his big base outside the city, he moved his people into it, establishing 29 outposts in its neighborhoods. In sum, it was a model of a counterinsurgency campaign, the first large-scale one conducted in the war. It was an example the U.S. military needed badly. In far northwest Iraq, a Marine battalion commanded by Lt. Col. Dale Alford carried out a similar campaign, establishing outposts in the area of al Qaim and cutting deals with local sheikhs. However, these examples weren’t imitated by other commanders, probably because they were at odds with the strategy set by Gen. Casey and his boss at Central Command, Gen. John Abizaid. Working on the theory that the U.S. military presence was an irritant to Iraqi society, the generals were trying to oversee a transition to Iraqi forces and so wanted an ever-shrinking American “footprint.” By contrast, McMaster injected thousands of U.S. troops into the middle of a city, implicitly saying that they were not the problem but part of the solution, that American troops weren’t the sand irritating Iraqi society, but could be the glue that held it together.

McMaster’s organization also began to grasp the significance of Iraqi tribal power. One of MacFarland’s officers, Capt. Travis Patriquin, a bright, bushy-haired, Special Forces veteran who spoke Arabic, Spanish, and Portuguese, was particularly intrigued by this. Lt. Col. Paul Yingling, his counterpart in the 3rd Armored Cavalry Regiment, told him about an officer’s encounter with a sheikh of the Shammar tribe. “Sheikh, why do you smuggle sheep and benzine in from Syria?” the officer asked.

The sheikh had responded, “Why did you put the Syrian border in the middle of my sheep? We were here first.”

Yingling told Patriquin about the Shammar tribe’s view of the world. “He understood it very well, and got a good laugh out of the story,” he recalled.

A few months later MacFarland was ordered to move his unit, the 1st brigade of the Army’s 1st Armored Division, to Ramadi. A soft-spoken officer from an Irish Catholic neighborhood in Albany, New York, MacFarland knew that every brigade assigned to that violent provincial capital had lost about 100 soldiers during its tour of duty, even as the city steadily declined into chaos. “I’ll be goddamned if I lose one hundred soldiers here and have nothing to show for it,” the brown-haired cavalryman, a 1981 graduate of West Point, vowed to himself.

His orders were to “fix Ramadi but don’t do a Fallujah”—a reference to the intense battles for that city just to the southeast in 2004. “But I really wasn’t sure how I was going to ‘fix Ramadi.’ ”

All the conventional responses had been tried and none had worked, so three years into the war, MacFarland was willing to take a gamble on something different. Anbar Province had at first been all but ignored in the planning for the 2003 invasion, then treated as an “economy of force” operation, and then saw two bruising battles for control of Fallujah in 2004. In a low point just before MacFarland’s brigade arrived, a protest broke out at a graduation ceremony for 978 Iraqi soldiers, most of them Sunnis, at nearby Camp Habbaniyah. Provoked by word that they would be ordered to deploy outside their home province of al Anbar, some soldiers began tearing off their uniforms before the astonished eyes of the Iraqi and American officials in attendance for the event, which they had hailed in speeches as a major step in the formation of the Iraqi army. At the time, U.S. military spokesmen attempted to minimize the significance of the event. “It was actually a very small number of graduates,” claimed one, Army Lt. Col. Michael Negard. But Carter Malkasian, a counterinsurgency adviser to the Marine Corps in Anbar, later disclosed that a full two-thirds of the soldiers refused to deploy, and more than that ultimately deserted.

On top of that, the Iraqi battalion that MacFarland was counting on for help had mutinied upon being informed that it would be deployed to Ramadi. Of several hundred men in the Iraqi unit, only about 140 showed up, he recalled—and most of them refused to leave the base to go on patrol. “We basically just sent them home,” he said.

MacFarland’s audaciously different approach to Ramadi ultimately would become an out-of-town tryout for the surge that came eight months later in Baghdad, not so much in troop numbers, but—far more important—in the strategy of moving into the population and the tactics of how to do that successfully. The two major differences are that Ramadi is overwhelmingly Sunni, and so didn’t have sectarian fighting, and also is a fraction of the size of the capital.

In 2005 al Qaeda in Iraq had mounted a ferocious campaign against about 12 tribal leaders who competed with the terrorist group for the loyalty of al Anbar’s population by forming the Anbar People’s Council. “This was the first broadly based opposition to al Qaeda,” recalled Marine Brig. Gen. John Allen. “Al Qaeda recognized the threat and attacked almost immediately,” conducting a focused and efficient assassination campaign. In one month, half the sheikhs in the council were dead, with the remainder fleeing the country. The Americans really hadn’t come to the aid of the sheikhs, who had multiple ties to the Sunni insurgency.

“There was a large safe haven there. . . . Al Qaeda was calling the shots,” MacFarland said. “Zarqawi was known to go out there, for instance. I mean, this was where al Qaeda went when they got pushed out of Fallujah.” In retrospect, he estimated that he faced perhaps 5,000 fighters in the city.

When MacFarland’s unit arrived in Ramadi, it was hit by bombs, grenades, mortars, and rifle fire an average of 25 times a day. It was replacing a unit from the Pennsylvania National Guard that had retreated from parts of the city. “My predecessor was just trading artillery fire with the rocket and mortar fire,” he remembered. “Al Qaeda had the run of the town. . . . The enemy basically controlled the center part of the city.” Every night, insurgents were planting an average of eight roadside bombs in and around the town. The National Guardsmen had stopped patrolling in areas where they had been hit hard, he said, leaving parts of the city map that, he joked, were labeled, HERE BE MONSTERS.

The city wasn’t even on life support. “There was no mayor, there was no city council, and there were no communications like we had in Tall Afar,” he said. “Basically, all services had stopped.”

Sheikhs were telling reporters that they no longer felt safe being around Americans. “Today, there is no tribal sheikh or a citizen who dares to go to the city hall or the U.S. base, because Zarqawi issued a statement ordering his men to kill anyone seen leaving the base or city hall,” said the head of one tribe, Bashir Abdul Qadir al-Kubaisat. The U.S. military assessed that of the 21 tribes in the area, only 6 would cooperate with it.

Desperation may be one of the stepmothers of invention. “There was really no place to go but up,” MacFarland recalled. “I was willing to try whatever made sense.” Other units were moving away from the cities, concentrating their forces on big bases. He decided to go in the opposite direction. His commanders, who were Marines, were skeptical, having seen dialogues with tribal leaders start up and then peter out before, but they let him take a flyer. “I had the backing of my bosses, but not a lot of guidance. I felt like if it failed, it would be my failure.”

Sterling Jensen, who was working as an interpreter for MacFarland’s brigade and had become deeply involved in tribal issues, recalled the Marines’ being even more negative. “They’d say, you guys don’t know what you’re doing. You’re way too arrogant. You’re going to get yourselves killed.” The Marines had tried several times to reach out to tribes, only to see al Qaeda assassinate sheikhs who turned. Senior Marines also thought that MacFarland was dealing with third-rate sheikhs who didn’t hold real power. What MacFarland wasn’t seeing was that some Marine generals had noticed that there was a quiet, almost secret war under way in Anbar between some tribes and al Qaeda. The Marines were reaching out to some of the harder hit sheikhs, offering them help.

On the upside, MacFarland’s superiors were willing to give him what he needed—a Marine infantry battalion, snipers from two Navy SEAL platoons (dubbed “Task Force Bruiser”), and even four 40-foot-long armored Marine riverine boats to cut off the enemy crossing points on the Euphrates River and stealthily insert patrols. “They were fast, they were quiet, they were heavily armed, and they could carry a squad and put them ashore,” he said. “They could kind of run up on the beach, dump them off, back off, and then provide fire support.”

Interestingly, among the Marines deployed to Ramadi was Cpl. Jimmy Webb, son of James Webb, the novelist (Fields of Fire) and former Navy secretary who in 2006 was running a long-shot campaign to become a U.S. senator from Virginia. While home on leave, the corporal asked his father why his opponent, Senator George Allen, made cowboy boots the symbol of his campaign “when Virginia doesn’t have any cowboys.” Webb was intrigued. His son also pointed out that he and his father both had worn combat boots in wartime. He gave his father his own boots, which he had worn in the streets of Ramadi. Webb would wear them throughout his campaign.

MacFarland and his staff began by thinking about the “metrics” they should use. If the goal was to protect the population, as they had seen in Tall Afar, then that is what should be tracked somehow. They also knew they would have to confront the skepticism of local leaders, who had seen Americans come and go for more than three years, making promises that often weren’t met or were forgotten by successor units. MacFarland began to spread the word that the Americans weren’t leaving anytime soon.

Knowing that Americans had put in office a generation of leaders, and then seemed unable to keep alive those police chiefs, mayors, and governors, MacFarland made protection of local leaders a top priority. He stationed tanks at key intersections near their houses and put drone aircraft circling over their homes to keep an eye out for attacks. He also asked sheikhs for advice on where to place new police stations and outposts, calculating that they would put them near their homes.

He named the Arabic-speaking Capt. Patriquin as his liaison to the sheikhs. Together they tried to sort out who was a real sheikh, with big wasta, or influence, and who was a lightweight. They also realized that years of fighting had created an opening: Not only had some sheikhs been killed, many others had moved to Jordan—and so a new generation of tribal leaders was emerging. “It was like going into Don Corleone’s house—you can tell who has wasta,” especially by following who moderated the discussion, he observed. The first sheikh with whom he began to work closely was Abu Ali Jassim, whose tribe was based out in the desert.

Following the example of the 3rd Armored Cavalry Regiment in Tall Afar, MacFarland began to establish small bases in the city. In the past, U.S. units had operated from a large FOB, or forward operating base, outside it. “They exited the FOB, drove to an objective or patrolled, were attacked, exchanged fire, and returned to base,” he wrote later.

The first step under the new approach was to send Special Operations sniper teams to sneak into the building he wanted to occupy. Then he would have a “route clearance” team work its way through the roadside bombs to the building, followed immediately by a company of Army troops or Marines to occupy the building. Upon arrival they would begin building a new combat outpost. The snipers would move out to the surrounding area to disrupt counterattacks. Overnight, the outpost would appear, with living spaces and walls and barriers to limit the damage from car bombs. They even figured out how to use a crane to immediately deposit a steel “crow’s nest” on top of a building, so they could begin with a well-protected observation post without having to divert troops into filling and carrying sandbags to the roof. (Learning that filling sandbags between patrols was wearing out the troops in the outposts, MacFarland instituted a new policy on his base. Everyone had to fill two sandbags before every meal. “No work, no food,” he said. “We could generate ten thousand to twelve thousand sandbags a day on Camp Ramadi and push them out to the combat outposts.”) Quick steps to establish combat readiness in the outposts were necessary because new outposts were almost always assaulted within two or three days.

Four benefits, much of them unexpected, flowed from the redisposition of troops into the small new bases, which eventually would total 18. In the most successful ones, Americans and Iraqi soldiers lived and ate side by side. This meant Iraqis and Americans could learn from each other—about Iraqi culture, about weapons maintenance, about leadership. Also, Iraqi soldiers living on American rations began to show more energy. “You’d be surprised at how much work you can get out of an Iraqi if he has had enough calories to eat,” he said. Another immediate benefit of this redeployment, he found, was that his soldiers became less predictable. No longer could Iraqi fighters simply watch the front gates of an American base to know when a patrol was coming. “Because we now maintained a constant presence in disputed neighborhoods, the insurgents could no longer accurately trace and predict our actions.”

Most important was the political effect of the new outposts. MacFarland laid down a rule that once one was established, they wouldn’t let themselves be driven from it. “You never give it up,” he said. “More than anything else, that was what persuaded the sheikhs we were there to stay.” In the past, he said, American commanders had said, “Don’t worry, we’re leaving.” He decided to say the opposite: “We’re staying until we win this fight.” It helped that once he had an outlying base, he would begin spending reconstruction funds in the surrounding neighborhood. All told, he estimated, he would dispense more than $2 million in 2006 and early 2007.

He sought to keep up the pressure, so that the enemy, once knocked off balance, couldn’t regain the initiative. “What can I do to make life miserable for al Qaeda today?” he would ask himself. “We tried to have an operation every few days. Can I put up another combat outpost? Should I start an adult literacy class? Can I throw in the kitchen sink?” Figuring that the local al Qaeda fighters might move to the outskirts, he set up Iraqi police stations in the rural tribal areas. Police were always recruited locally, which gave them extra incentive to stand up to the terrorists, he noted. “The IPs [Iraqi police] refused to be intimidated because they were defending their own homes,” he said.

By the end of July 2006, he was beginning to sense that the new approach was working, even though it brought new risks. The commander of the Marine battalion attached to MacFarland’s Army brigade told him that west Ramadi was quieting down. Top Marine commanders began to be convinced that what was happening in Ramadi was different from previous sheikh-led pushes against al Qaeda. Even so, there were days when MacFarland had his doubts, especially as the enemy launched a counteroffensive. At the end of the first week of August, he thought to himself, “My God, I’ve lost ten guys.” Two weeks later, on August 21, Sheikh Jassim, his first ally in the tribes, was assassinated. “I couldn’t have protected him if I wanted to,” MacFarland said. The sheikh’s killers hid his body for four days, a pointed violation of the Muslim custom of quick burial. On the same day, a new Iraqi police station, in the Jazeera neighborhood, and manned mainly by members of Sheikh Jassim’s tribe, was bombed. All told, MacFarland lost two dozen vehicles—a few tanks, but mainly trucks—as he moved into the city.

But, he said, the local reaction to the August attacks indicated that al Qaeda might have overplayed its hand: They drove some fence-sitters into the American camp. One sheikh, Sittar albu-Risha, was particularly angry. “Sittar has lost enough family members that he was ready to throw away caution.” This sheikh, a minor tribal leader who had a reputation for running a thriving cross-border smuggling business, called a meeting for September 9. More than 50 sheikhs and other notables showed up. They created what they proposed calling “The Awakening Council.” They had a platform with 11 planks. “Ten of them I would have written for them almost exactly the same way they wrote them,” MacFarland recalled. The last one was problematic, in that it implied they might have to kill the governor of al Anbar Province. He suggested they modify it.

As MacFarland parleyed with sheikhs, the energetic Capt. Patriquin worked the other people in the room. “He was very extroverted and friendly and was very popular among the tribes because he was the officer who identified little things we could do for them, like attention for a sick child.” Sheikh Sittar eventually gave the captain the honorary tribal name Neshan Abu Risha, which some Iraqis say means “a warrior of the Albu Risha,” the sheikh’s tribe.

That day was a turning point for MacFarland—and as it would develop, for al Anbar Province and Iraq. “To me, it was the first real clear vindication of the strategy we were pursuing, that we were beginning to turn the tide.” The meeting encouraged more sheikhs to come in and work with the Americans, and with them came a “snowball effect” on recruiting of local police and other tasks, MacFarland said. “Whenever a tribe flipped and joined the Awakening, all the attacks on coalition forces in that area would stop, and all the caches of ammunition would come up out of the ground. If there was ever an attack on us, the sheikh would basically take responsibility for it and find whoever was responsible, and this happened time and again. So it was incredibly effective and they were as good as their word.”

MacFarland had come to terms with the fact that some of those newly forthcoming sheikhs had participated in attacks on Americans. “I’m a product of Catholic schools,” he said, “and I was taught that every saint has a past and every sinner can have a future.” Sittar reported that he had several thousand volunteers who didn’t qualify for the police, because they were illiterate, underage, or overweight, so he was allowed to create three “emergency battalions” to employ them. MacFarland armed them with captured weapons and had his SEAL teams give them a one-week training course. The prevailing American theory for years had been that improvements in security would lead to progress in politics. This was the opposite—political change leading to improvements in security.

That decision also took the United States into the dangerous and complex new territory of supporting an armed group that was opposed to the government in Baghdad that the United States also supported. As Carter Malkasian, the counterinsurgency adviser to the Marine Corps in al Anbar, put it, “For all intents and purposes, the government was permitting Sittar and his movement to have their own militia.” But, as Petraeus and Odierno would do the following year, MacFarland had decided it was time to take some risks, especially given that the alternative appeared to be failure.

The Shiite-dominated Iraqi government, in the midst of a small civil war in Baghdad between Shiites and Sunnis, wasn’t happy with what it was hearing out of Ramadi about the Americans cutting local cease-fire deals with Sunni sheikhs. Here again, MacFarland found that American experience on the ground in Iraq helped. His deputy commander, Lt. Col. James Lechner, had spent time as an adviser to the Iraqi military and “knew how to work the system to get guys paid.” Among other things, MacFarland noted, “That built up my wasta with the sheikhs.”

Faced with skepticism from his superiors, and from journalists who were being told by Iraqi officials in Baghdad that he was arming Iraqis to fight the Iraqi army and police, MacFarland had Patriquin create a briefing to explain what he was trying to do. Far from the usual razzle-dazzle of U.S. military PowerPoints, the briefing was written breezily, almost in the style of a children’s book, with stick figures. It was titled “How to Win the War in Al Anbar, By CPT Trav.”

Capt. Travis Patriquin’s briefing was both perhaps the most informal one given by the U.S. military in Iraq and the most important one. “This is an American Soldier,” it began. “We’ll call him Joe. Joe wants to win in Al Anbar. But sometimes it seems like other people don’t share that idea.” This made Joe sad. Then the briefing posed the key question: “How can Joe win in Al Anbar? By fighting the insurgents?” The briefing didn’t say so, but the answer the U.S. military had given for three years had been: Well, of course, yes.

A subsequent slide identified the problem with that approach: “Poor Joe can’t tell the terrorist from the good Iraqis.” The smiling stick figures look all the same to him. The solution, the brief said, was to talk to the sheikhs about making local militiamen members of the police force. “The Iraqi Policeman can tell the difference. And the insurgent knows that. See, that’s why he’s sad.” This makes everyone else happy. “The sheikh brings more sheikhs, more sheikhs bring more men. Joe realizes that if he’d done this three years ago, maybe his wife would be happier.”

The theory was working, but the fighting continued. On September 29, Michael Monsoor of Garden Grove, California, a twenty-five-year-old member of a Navy SEAL team, threw himself on a hand grenade while his team was being attacked, an act of valor for which he was posthumously recognized with the Medal of Honor. He already had received a Silver Star for rescuing a wounded comrade under fire four months earlier.

On November 25, about three dozen al Qaeda men with weapons drove into Sufia, home of the Albu Soda tribe, just east of Ramadi. The small tribe, which had only about thirty men of military age in that area, had rebuffed MacFarland’s recruiting efforts, he said, because it wanted to be neutral. But as part of that effort, it had established checkpoints to keep out al Qaeda, which antagonized the terrorist group, because the tribe lived along the main corridor from Fallujah to Ramadi. After the gunmen opened fire, some tribal members escaped in boats across the Euphrates and ran to an Iraqi army base. Soldiers there called an Iraqi interpreter for an American officer, who called MacFarland’s headquarters. Capt. Patriquin and Sterling Jensen, the interpreter, began gathering information. “We’re being wiped out,” the tribe’s beleaguered sheikh told Jensen. “People are killing us.” The sheikh’s sister had been killed, and al Qaeda men were dragging the body by ropes behind a pickup truck. MacFarland postponed another operation and sent the units involved in that to the aid of the tribe, even though it had held him at arm’s length. A drone reconnaissance aircraft was sent to circle over the fight. Patriquin called the sheikh of the tribe. “Hey, look,” he said. “We can’t tell who is who. Could you have your guys wave towels over their heads so we can identify friend from foe?” That done, Marine F-18 warplanes rolled into bomb those without towels, and then arriving U.S. Army tanks began to fire on fleeing al Qaeda automobiles.

After the fight, MacFarland went to talk to Albu Soda’s bloodied, combat-shocked leaders. “They were kind of battle-fatigued, had lost a lot of family members. At the same time, it was like a switch had been flipped. Guys who had been reluctant to talk to us were saying, ‘Would you please build a combat outpost near our home?’ and telling us where al Qaeda was in their area.” That day was the tipping point, he said. After that, he was flooded with tips and recruits. “After that tribe flipped, the kids were running around, it was like liberated France, it was like Rumsfeld imagined it would be in 2003.” Also, a major insurgent route into the city had been cut.

One of MacFarland’s enlisted men studied street life and concluded that people in Ramadi didn’t read newspapers or even listen to the radio much, but that they did pay attention to the messages from loudspeakers on the minarets of mosques. So, beginning at one platoon-sized base, Combat Outpost Firecracker, MacFarland’s soldiers put up loudspeakers to broadcast, every day but Friday, the Muslim sabbath, sports news and weather reports—and occasionally slip in information about al Qaeda attacks. “The news was pulled from places like Al Jazeera and Al Iraqia and news sources that people would know were not ours,” he said. Some of it was just helpful tips: “The UN warehouse has a new shipment of rice.” Occasionally another message would slip in: “Last night al Qaeda killed a family of five in their home.”

But even as late as December 6, 2006, MacFarland would be thrown for a loss. On that day, Spec. Nicholas Gibbs, a twenty-five-year-old from Stokesdale, North Carolina, was killed by small-arms fire. “Part of me died along with him,” his mother told a reporter. “I will never be the same.”

The same day, Sgt. Yevgeniy Ryndych, a 1998 émigré from Ukraine to Brooklyn, was killed by a roadside bomb. His fiancée received her engagement ring in the mail from him the same day. “He was one of those people who not a lot of people liked because he sat home the whole day and read books,” said his brother Ivan. “He was like a genius kid.”

And Marine Cpl. Dustin Libby, from Presque Isle, Maine, was manning a machine gun on a roof in Ramadi when he was shot.

Three other soldiers died that day when their Humvee was hit by a bomb. The first, Spec. Vincent Pomante, was from Westerville, Ohio. The second, Marine Maj. Megan McClung, became the highest-ranking female to die in combat in Iraq. A triathelete, McClung had organized a marathon for troops in Iraq. A journalist in North Carolina remembered that when a police officer pulled over McClung, she proved she was sober by doing a backflip on the side of the road. A graduate of the Naval Academy, McClung had left the Marine Corps but went back on active duty in order to serve in Iraq. Her name would become the first woman’s to be added to the marble tablet at the academy that memorializes graduates killed in action. “Please don’t portray this as a tragedy,” her mother requested of a reporter. “It is for us, but Megan died doing what she believed in.”

The sixth loss that day perhaps hit MacFarland hardest: Capt. Patriquin, the soldier who had reached out to the tribes, had been sitting next to McClung.

The next morning MacFarland found his staff and commanders downcast. “Everybody was kind of looking at their feet.” He told them about how Gen. Ulysses S. Grant handled that first terrible day at Shiloh in April 1862. The Confederates had pushed the Union troops back to the Tennessee River, where thousands huddled terrified below the bank. Thousands more lay dead and wounded on the battlefield above them. That night Grant met with his commanders next to a log house being used as a hospital, reviewing the day’s losses as men under the surgeon’s knife screamed and died nearby—probably not the best place to locate a command post. At midnight, Grant went out to smoke a cigar, taking refuge from the driving rain under a tree. There, MacFarland told his soldiers, Gen. William T. Sherman found him. “Well, Grant, we’ve had the devil’s own day, haven’t we?” Sherman said to his dripping friend.

MacFarland reminded them of Grant’s laconic response: “Yes. Lick ’em tomorrow, though.”

The stoic, taciturn Grant was an inspiration to MacFarland throughout the year. “I felt I was fighting my way through the Wilderness Campaign,” he said, referring to Grant’s running battle through rough ground in northern Virginia against Robert E. Lee in May 1864. “I was taking a lot of casualties.” MacFarland was hardly alone. In recent years, as the Army has come to grips with Iraq, Grant seems to be enjoying a resurgence in popularity with today’s officers, probably because he is its patron saint of the long, hard slog.

By the onset of winter it was becoming clear that something fundamental had changed in Ramadi. “In the latter half of December, it was like the fever broke,” MacFarland said. “Up until then, when we threw a punch, they threw a punch.” The death rate for U.S. forces began to decline after that incident. By the end of that month, 12 of the tribes in the area were deemed cooperative, and 6 neutral, leaving just 3 classified as “uncooperative.” By mid-2007 it wasn’t uncommon for a month to go by with no U.S. losses. Al Qaeda, meanwhile, was reeling. As David Kilcullen, Petraeus’s counterinsurgency adviser, later put it, “In Anbar, we’ve got the tribal vengeance structure working in our favor.” That is, where Armericans once had been the target of Iraqis seeking revenge, now they were helping direct that impulse against al Qaeda and its allies in the insurgency.

Not only were the roadside bombs less numerous, they were becoming less sophisticated. “They went away from the remote-controlled IEDs to subsurface command wires to just hastily throwing out IEDs with pressure plates because that was all they could do,” MacFarland said. “Because we were keeping such pressure on them, they just weren’t able to get the big IEDs and get them all set up. So, we knew we had them on the run when we started to see those kinds of things evolve and their attacks became smaller and smaller and less and less effective.”

The attacks would continue, though. All told, MacFarland lost 83 soldiers in Iraq—but he had something to show for it. In February 2007, Gen. Petraeus, newly arrived in Iraq, would come to see him and ask some questions about his methods and metrics. “Sean had obviously done something extraordinarily important,” Petraeus said later. “What you had there was the first really significant example of the concept of reconcilables and irreconcilables.” Petraeus already knew that he wanted his troops to go out and protect the population. In Ramadi, he learned that “a key way of implementing that is not just living with them, it is also . . . literally separating them, protecting the population from the irreconcilables. That means you have to know who the reconcilables are and who the bad guys are, and then of course try to achieve some separation and protect the one from the other.”

Chiarelli, the number two U.S. commander in Iraq in 2006, said that MacFarland’s operation marked the first time in the Iraq war that a counterinsurgency campaign had been conducted and then had been sustained by the succeeding unit. “Sean was the first guy who did it and it stuck for the guy who followed,” he said.

Upon arriving in Iraq, Odierno would seek to build on what MacFarland had started. “He’s the guy who put this together”—that is, how to operate differently and more effectively in Iraq, Odierno said later. “Once they cleared Ramadi, and they stayed in Ramadi with a significant amount of force, that was the tipping point. The whole province seemed to turn over.”

But Baghdad would be more complicated. Not only was it at least 10 times larger, it also had both the Shiite militias that weren’t active in Ramadi, which was homogeneously Sunni. Tribes were less significant in the cities and among Shiites. Securing Baghdad in 2007 would make MacFarland’s experience in Ramadi in 2006 look relatively simple.


In October 2006, Petraeus was in Washington, partly to lay the groundwork for rolling out his counterinsurgency manual a few months later, but also because Gen. Pace, the chairman of the Joint Chiefs of Staff, had sent word that Defense Secretary Rumsfeld wanted to see him.

Petraeus didn’t know what the meeting would be about. But he could see his time at Leavenworth coming to an end, and he was eager to get back into the fight in Iraq. Every indication was that a radical change in the handling of the war there was urgently required. He felt ready to lead that charge.

As he prepared, he contacted Lt. Col. Charlie Miller, whom he had known since he himself was a lieutenant colonel and Miller was a green officer in his battalion in the 101st Airborne, to ask him to go for a morning run. In 2006 Miller was a strategic planner on the Joint Chiefs of Staff, and like many of his peers, was anxious about the state of the war. The next day the two met at Petraeus’s hotel. It quickly became clear that Petraeus wanted to talk about Iraq. “He was very spun up on the war, knew what was happening,” Miller recalled.

As they ran along the sandy paths of Washington’s mall toward the Capitol, Petraeus posed a series of questions. “The nation has to decide what it is going to do—is it going to do what it takes, or is it going to get out?” he began.

After the run, Petraeus said to Miller, “What are we trying to accomplish there? And what resources do we need to do it?”

This was magic to Miller, a native of Virginia’s Shenandoah Valley who was trained in strategic thinking and who had believed for years that more troops needed to be sent to Iraq. “This had been a major frustration for me,” he said.

“We have undertaken a major national project and put it on the backs of a small group of volunteers.”

These also were the basic questions any strategist would ask about a war—especially if he suspected he might about to be put in charge of that war.

When Petraeus went to see Defense Secretary Rumsfeld, he thought that perhaps he would be offered command in Iraq. But as he was walking up the Pentagon stairs with Pace to Rumsfeld’s office, the Joint Chiefs chairman turned to him and said, “Don’t be surprised if this is about the Afghanistan job.” That was not a bad command, but it was still a relative backwater compared to Iraq. As it happened, Rumsfeld, who could be extremely noncommittal, didn’t offer Petraeus anything.

If you find an error please notify us in the comments. Thank you!