Rumsfeld"s response made it even clearer to Keane that the administration was digging in its heels. "My goodness," Rumsfeld began in his anachronistic fashion as he launched into a pa.s.sionate defense of his past deeds and of current policy. "History will make a judgment" on past decisions about troop numbers, he announced, as if to tell her that coming to such conclusions was above her pay grade.

As for an increase in troop levels, the prospect that Keane was mulling, Rumsfeld rejected the idea. "The balance between having too many and contributing to an insurgency by the feeling of occupation, and the risk of having too few and having the security situation not be sufficient for the political progress to go forward, is a complicated set of decisions. And I don"t know there"s any guide book that tells you how to do it. There"s no rule book, there"s no history for this." In fact, as Keane knew, there was ample history on just that point. The experiences of the British in Malaya, the French in Algeria, the Americans in Vietnam, and in a dozen other smaller counterinsurgency campaigns all taught the same lesson: You must protect the people and separate them from the insurgents, and to do so you had to live among the population. And doing all that required a lot of troops. Indeed, the manual on counterinsurgency that Gen. Petraeus was drafting out at Leavenworth as Rumsfeld spoke would make just that point.

But Rumsfeld was sticking to the existing plan, despite the multiple setbacks it had encountered. "The goal is to not have U.S. forces do the heavy lifting in Baghdad. There are many, many more Iraqi forces in Baghdad. The role of the U.S. forces is to help them." As Keane knew, there was growing evidence by this point that this transitional approach wasn"t working.

The defense secretary concluded by lapsing into his trademark rhetorical device of posing a question and then answering it himself. He didn"t seem to grasp how condescending this could be, especially with people like United States senators who tend to have a strong sense of the importance of their positions. He did so some ten times. Three of them were, "Are there setbacks? Yes. Are there things that people can"t antic.i.p.ate? Yes. Does the enemy have a brain and continue to make adjustments on the ground requiring our forces to continue to make adjustments? You bet."

Abizaid supported Rumsfeld. Going above the current level of 140,000 troops in Iraq, he said, would place "a tremendous strain" on the Army.

Gen. Pace, never particularly impressive as the chairman of the Joint Chiefs, had even less to offer that day. "Shia and Sunni are going to have to love their children more than they hate each other," he said. It was an almost despairing phrase, pointing toward no discernible strategy for the U.S. government. He apparently so liked the thought that he repeated it later in the hearing.

As Keane watched, the doubts he had been gathering for months coalesced and solidified. "My G.o.d, if we don"t do something different, we"re going over a cliff," he thought that night. He was not a man who came to a conclusion like that idly.

The next afternoon he sat in his living room, meditating on what to do and how to do it. He was there for so long that evening fell. Lost in thought, he didn"t turn on any lights, and his wife came into the room to find him sitting in darkness. She asked what he was doing. "Iraq," he responded. "Our strategy there is failing. We need a new strategy, and new people, "cause the guys doing it don"t think it"s failed."

"What are you going to do about it?" she asked. It was an unusual question to pose. She knew that many Americans had similar concerns, but that her husband was uniquely positioned to do something about it. After decades of service, he was an Army insider. In particular, he had been a mentor to two rising Army generals, David Petraeus and Raymond Odierno, and he thought they could be key to making the changes, even though neither was in Iraq at the time.

"I think I am going to try to change it," he told her. It wasn"t an idle response. He began to write some notes, outlining the problem.

Since retiring from the Army, he also had come to know influential strategic experts such as Kissinger and Eliot Cohen, other members of the Defense Policy Board who were growing increasingly worried about the direction of the war. In May 2004, Cohen had gone to Iraq for the board and come back to deliver a grim a.s.sessment. "There is no sense of a common vision or direction, a real operational or strategic level plan," he reported. A senior officer had told him that mid-2003 to mid-2004 was "a lost year." In addition, Cohen had concluded that Army and Marine doctrine for conducting a counterinsurgency campaign was badly outdated-an observation that may have encouraged the Army to send Petraeus to Leavenworth.

Keane agreed with many of Cohen"s worries. And while many active-duty officers shared his deepening concerns, he possessed an option they didn"t: If he felt he didn"t get a thorough and serious hearing, he could take his concerns public. He knew how to do it-he was a retired four-star general who maintained cordial relations with several defense reporters.

"DAVE, YOU"RE SHOT"

Perhaps most important, Keane had known Petraeus for years. An advocate of realistic training, Keane loathed seeing soldiers toss grenades as if they were outfielders hurling metal baseb.a.l.l.s, instead of in the context of how they would be used in combat, where people who want to survive don"t stand up in view of the enemy. So he had pushed for "live-fire" exercises, in which soldiers used real bullets while training and moved as if they were on a battlefield. One day in 1991 at Fort Campbell, Kentucky, Keane and Petraeus were observing just such an exercise, in which a squad was practicing taking down a machine gun bunker. Some soldiers provided suppressive fire while one of their comrades crawled forward from one side and, leaning to one side while still p.r.o.ne, lobbed a hand grenade into the bunker.

Under the cover of the explosion, the grenade thrower turned and ran as fast as he could back to his fellow squad members. He hit the dirt using the b.u.t.t of his M-16 rifle to break his fall, as he had been taught to do in order to get down quickly. But the soldier, probably distracted by his grenade throwing, had made two mistakes: He had his kept his finger on the trigger of his weapon, and the safety was off.

Petraeus, observing from 40 yards away, grunted and stepped back, but didn"t fall. Keane, standing next to Petraeus, looked over. "Dave, you"re shot," he said. The bullet from the soldier"s weapon had pierced Petraeus in the right side of his chest, just above the A A in PETRAEUS on his fatigues, and clipped both a lung and an artery. Keane laid Petraeus on the ground, then reached around him and felt for the exit wound. It was about the size of a half-dollar coin. He called for a medic. Then he looked down at Petraeus and said, "Dave, you know what"s going on here, we"ve got to stop the bleeding. . . . Then we have got to make sure you don"t go into shock." in PETRAEUS on his fatigues, and clipped both a lung and an artery. Keane laid Petraeus on the ground, then reached around him and felt for the exit wound. It was about the size of a half-dollar coin. He called for a medic. Then he looked down at Petraeus and said, "Dave, you know what"s going on here, we"ve got to stop the bleeding. . . . Then we have got to make sure you don"t go into shock."

Characteristically, while waiting for a medical evacuation helicopter, Keane took aside the commander of the company training that day and told him to continue the exercise. "What I was trying to teach them," he recalled, is that "in combat it"s going to be much worse than this, we are going to get our guys shot and get our guys killed and, one, we go on with the mission, two, we find out what the mistakes were after it"s over so we can fix it for the next time."

Keane held Petraeus"s hand on the short helicopter hop to Fort Campbell"s Blanchfield Army Community Hospital. A doctor there picked up a suction tube to clean the entry wound of strands of Petraeus"s uniform and dirt. "Colonel Petraeus," he said to his new patient, who was supine but still conscious, "I"ve got to clean this wound out, because when the bullet goes in there it takes all of that with it. I"ve got to get as much out of there as I can so it doesn"t start to get into your bloodstream." Not waiting to administer an anesthetic, he worried: "This is going to hurt like h.e.l.l"-and told some orderlies to hold him down. Then he jammed the tube into the b.l.o.o.d.y hole in Petraeus"s chest.

Usually, the doctor later told Keane in the hallway, the procedure inflicts so much pain that the body jumps up on the operating table and the patient "screams like h.e.l.l."

Petraeus just grunted. "That really is one tough soldier in there," the doctor said.

"Yeah, I know that," Keane replied. The chest operation that Petraeus would need required a second flight, this time by an Army Black Hawk helicopter to a hospital in Nashville, where he was met by Dr. Bill Frist, who had yet to enter politics but who would later become Senate majority leader. Frist, still in his golfing outfit, saw the small entry wound and wondered what all the fuss was about. Turning Petraeus over, he saw the exit wound and understood. Keane told him that the exit injury was typical of a high-velocity weapon, which was outside Frist"s usual cases of wounds made by cheap pistols and knives. Frist operated for more than five hours.

Less than a week later, Petraeus was back recuperating at the Fort Campbell hospital and growing impatient with it. "Dave was raising all sorts of ruckus because he wanted to get out of there and go home," Keane recalled.

A senior doctor went to see this troublesome patient. "Hey, Dave, you"re not going home so just leave my staff alone," he ordered. "You"re just out of surgery, you"re not going to be able to get out of here for a few more days."

Everybody heals differently, Petraeus argued. "I believe that I"m recovered enough to be able to go home," he said.

"That"s impossible-you"re not going home," the doctor said.

"Can I demonstrate to you the degree of my recovery?" asked Petraeus.

The doctor asked what he meant. "Just undo my tubes here," Petraeus said. "Don"t worry, I"m not going to do anything to hurt myself, just undo my tubes." He got down on the floor and counted out 50 push-ups for the doctor, who then allowed him to leave the hospital.

The life of a light infantryman is tough for soldiers in their twenties, requiring both strength and stamina. The weight, the rain, the stress, can combine to break the health of soldiers in their thirties. Petraeus left the hospital worried about regaining his strength. "What was really eating at me at the time was how well I was going to be able to run again," he recalled. He went to the base gym to work out on a stationary bicycle, "gently, just to keep the legs moving." But there was a running track just outside the gym. And the watch on his wrist had a stopwatch function. "One thing led to another," he said a bit sheepishly, explaining how he happened to find himself jogging on the track, trying to see if he could still breathe deeply. "It wasn"t the gunshot wound alone, I had thoracic surgery where they cut you, I have a scar that goes from here all the way around to there," he said, tracing a line from his chest, under his arm, and to his back. "There is a lot of scarring so the lung doesn"t glide in your chest the way it used to, so it feels like you are permanently taped up. I just wanted to get a sense of what it was going to feel like." The next round of X-rays revealed that that spontaneous bout of exercise had begun to fill the injured lung with fluid. The doctors told him to knock off running.

KEANE ON THE WARPATH.

On September 19, 2006, Keane was ready to make his case to Defense Secretary Rumsfeld. The two old bulls had had some differences in the past, but Keane felt there was mutual respect. "We could talk to each other," he said.

Gen. Pace showed up at the meeting, a bit to Keane"s surprise, because Pace"s reputation was that he was taking a hands-off approach to Iraq, on the grounds that his plate was full with the rest of the world and that two other four-star generals, Abizaid and Casey, were on the case. Keane worried that Rumsfeld would play to Pace, as was his wont. But the chairman remained quiet, just taking notes, and Rumsfeld stayed focused on Keane"s grim message. "We are edging toward strategic failure," Keane warned the defense secretary. "Despite capturing Saddam Hussein, killing his two sons, holding three elections, writing a const.i.tution, installing a permanent government, beginning to develop a capable ISF, killing Zarqawi-the level of violence has increased every year in the contested areas. Security and stability is worse today than it has been since the insurgency started. It threatens the survival of the government and the success of our mission."

So, Keane asked rhetorically, what is wrong with our strategy? His answer: "It is not designed to defeat the insurgency and therefore the insurgency thrives, and the violence is growing. It begs the question, how can you defeat it?"

Continuing to borrow Rumsfeld"s approach, he posed another question and then answered it: "How can we possibly obtain victory out of all that is happening?" First, Keane said, "you have to admit that you cannot defeat the insurgency by destroying their forces or simply transitioning to Iraqi security forces. You have to come to grips with that." Next, he said, start employing cla.s.sic counter-insurgency practice: "The only way to do this is the way that it"s been done in the past, using proven COIN [counterinsurgency] practices-and that is by protecting the people and permanently isolating the insurgents from the population." So, he told Rumsfeld, the U.S. Army needed to stop conducting mindless Humvee patrols out of big bases and instead start living among the people and patrolling small areas on foot. Set up traffic-control points, conduct a census, and issue ident.i.ty cards-all cla.s.sic measures to channel and track the movements of a population. His most controversial recommendation was that Rumsfeld order everyone to stop talking about drawing down troop levels in Iraq. Get some new generals in there, hold them accountable, and match your policies to your resources. To live among the people, and dry up the sea in which the insurgents swam, you are going to need more troops. And focus them on Baghdad. What Keane was saying was hardly novel. He had captured the core lesson of David Galula, the great French theorist of counterinsurgency, who argued in his influential Counterinsurgency: Theory and Practice Counterinsurgency: Theory and Practice that to defeat an insurgency, military units must live among the population. Indeed, Keane recommended that book to Rumsfeld. The defense secretary was "uncomfortable" during the meeting and opposed increasing the troop levels without offering a reasoning, Keane said as he read aloud from his extensive notes of the meeting. that to defeat an insurgency, military units must live among the population. Indeed, Keane recommended that book to Rumsfeld. The defense secretary was "uncomfortable" during the meeting and opposed increasing the troop levels without offering a reasoning, Keane said as he read aloud from his extensive notes of the meeting.

GRADING THE CHAIRMAN.

Pace followed up by asking Keane to come see him. Two days later, Pace began that meeting by asking Keane with a smile what grade he would give him as chairman of the Joint Chiefs.

"F," Keane replied. He wasn"t smiling.

Gen. Pace was taken aback. People didn"t usually talk to a top officer in the U.S. military that way, not even retired four-star generals. What do you mean? he asked.

"Well, Pete, the number one national security priority we have is Iraq, and it"s the number one priority in the Pentagon," Keane told him. But, he added, "you"re absorbed in so many other things." And, he continued, why the h.e.l.l are you, the chairman of the Joint Chiefs of Staff, asking a retired general about the situation in Iraq? How is it possible that you, with all the far-flung resources of the global U.S. military establishment at your fingertips, know less about the war than one guy who goes out there and asks a few questions? "You"ve got to get into this full time," Keane admonished. (One of Pace"s subordinates would later recall that at about the same time, Adm. Mike Mullen, then the chief of the Navy, sent a similar message, to the effect, "This is going down on your watch. You need to do something.") Keane also told Pace to put new people in charge: Replace Abizaid at Central Command with Fox Fallon, who as an experienced regional commander could step into the job quickly, Keane said. And replace Casey with Dave Petraeus.

Pace responded with alacrity, Keane recalled, canceling a planned trip to South America and instead starting up a group of officers to review Iraq policy. This new panel, dubbed "the council of colonels," was tasked first to look at the entire war on terror, but then decided that the first question had to be whether the strategy of the Iraq war was working. It first met on September 27, not long after Keane rang Pace"s bell. It began asking a series of questions that in the following weeks would make Pace and the Joint Staff reexamine the deteriorating situation in Iraq.

In November, Raymond Odierno, another Keane protege, headed out to Iraq to take over the number two position in the war from Lt. Gen. Pete Chiarelli. On November 20, a macabre note was struck when a leading Iraqi comedian was a.s.sa.s.sinated. It wasn"t clear what the point of the killing was, because Walid Ha.s.san had specialized in mocking the difficulties of life in occupied Iraq in a non-sectarian fashion. Three days later, a barrage of car bombs, mortars, and missiles. .h.i.t Sadr City, killing more than 200 people. It was the single deadliest attack in years. The next day, Shiite fighters retaliated with a citywide attack on Sunni mosques crowded for Friday prayers. Bombs, rocket-propelled grenades, and automatic-weapons fire hit crowds of worshippers.

As he did Petraeus, Keane knew Odierno well. (It is an interesting coincidence that all three men are from the greater New York area, which generally isn"t seen in American culture as a hotbed of generalship. Keane is from Manhattan; Petraeus from Cornwall, just up the Hudson; and Odierno from just across that river, in northern New Jersey.) In the summer of 2001, Keane, then the vice chief of staff of the Army, had gone to a meeting with Rumsfeld at which 40 slides were presented on various recommendations the defense secretary had planned to make. This was before the 9/11 attacks, and long before there was any serious thought of invading Iraq. Tucked among the slides, to Keane"s surprise, was a plan to reduce the active-duty Army to eight divisions from ten, and on top of that to cut four divisions from the National Guard. This represented a major reduction in the ground combat strength of the United States, and Keane, the number two officer in the Army, had received no advance notice that it was coming. He was stunned-that just wasn"t how a superpower was supposed to work. "Mister Secretary, I disagree with this strongly," Keane had said. He asked for 24 hours to develop a response. Back in his office, he summoned Odierno, then a bright young one-star general on his staff, and told him to brief Deputy Defense Secretary Wolfowitz on how the Army had arrived at its current force structure and how that related to the size of the Army that actually fights. He picked Odierno, he said, in part because "he had this intellect, this grasp, this power of persuasion." Also, he said, "a lot of guys become afraid to fail. You want them to push the envelope, and Odierno was doing that."

The next day, Odierno gave a presentation to Wolfowitz and others on why the Army"s numbers shouldn"t be cut. At the end of the briefing, Wolfowitz, who had been given the lead on the issue by Rumsfeld, said, "I"m convinced." He said he would take the recommendation to the secretary, who concurred.

Even so, Keane wasn"t entirely approving of Odierno. He had some issues with how he had led the 4th Infantry Division in 2003-4, during his first tour in Iraq. Nevertheless, Keane thought Odierno was a tough, intelligent officer who, unlike some of his peers, was willing to take risks for what he thought was right.

KEANE AND ODIERNO VS. THE WORLD.

Between Keane and Odierno, a kind of guerrilla campaign was launched inside the U.S. military establishment. Keane was in Washington and Odierno was in Baghdad, but they talked by telephone almost every day.

"We don"t easily jump the chain of command," retired Col. John Martin, a friend and adviser to Petraeus, said in another context. Making one of the most audacious moves of the entire war, Odierno did exactly that, bypa.s.sing two levels of command above him to talk to officials at the White House and aides to the Joint Chiefs of Staff. He was about to become the sole senior official in the active-duty military speaking out for an increase in troops, recalled a senior U.S. intelligence officer who privately supported such a full, five-brigade counteroffensive. "He was the only one in the chain of command-not MNF-I, not Central Command, not the Joint Staff," this intelligence official recalled. (Adm. Mike Mullen, who was then the chief of the Navy and in 2007 would become chairman of the Joint Chiefs, would later insist that he had supported the surge, but apparently meant that he had endorsed the smaller, two-brigade option that Pace would take to President Bush just after Christmas 2006.) "Odierno and I are having a continuous dialogue" at this time, Keane recalled. "He knows he needs more troops, he knows the strategy has got to change. His problem is General Casey."

Just as Odierno was beginning his epic end run around Casey and the rest of the senior leaders of the U.S. military establishment, the president a.s.serted in an interview with the Washington Post, Washington Post, "It"s important to trust the judgment of the military when they"re making military plans. . . . I"m a strict adherer to the command structure." "It"s important to trust the judgment of the military when they"re making military plans. . . . I"m a strict adherer to the command structure."

Ironically, it was only after Odierno stepped outside that structure, rejecting the views of his superiors and lobbying the White House on his own, that policy formulation began to work effectively, producing a workable strategy. Arguably, his actions amounted to insubordination. Casey seemed puzzled when told in a 2008 inverview that Odierno had grave doubts about the direction of the war back in December 2006. "Ray never came to me and said, "Look, I think you"ve got to do something fundamentally different here,"" Casey said.

"Courage takes two forms in war," observed Hew Strachan, the British military historian and interpreter of Clausewitz. "Courage in the face of personal danger, whose effects are felt in the tactical sphere, and the courage to take responsibility, a requirement of strategic success." By taking on his new boss, Odierno displayed that second, more elusive form of bravery. He was laying his career on the line. If the surge went wrong, he would be the first blamed by many inside the military who had made clear their profound concerns and objections.

Lined up against Odierno were the collective powers at the top of the U.S. military. Abizaid, the chief of the Central Command, told Senators John McCain and Lindsey Graham in a Senate Armed Services Committee hearing on November 15 that he and every general he had asked opposed sending more troops to Iraq. "I do not believe that more American troops right now is the solution to the problem," Abizaid emphasized. "I believe that troop levels need to stay where they are now."

Gen. Chiarelli, who was about to leave Iraq, was also questioning a troop escalation. He said at the time, "I happen to believe we have done everything militarily we possibly can." Asked about that in 2008, he said his concern had been "How are we going to source them? And I still thought, troops alone are not going to stop this problem, that we need to get the Iraqi government to act differently."

In early December, Pace, the chairman of the Joint Chiefs, was privately telling his colleagues that he didn"t see that 160,000 U.S. troops in Iraq could do anything that 140,000 weren"t doing. He agreed with the other members of the Joint Chiefs: What was needed, he griped, wasn"t a troop surge, but a new commitment by the rest of the U.S. government. In this view, the military was doing its part but had been left high and dry by the civilians. Like the rest of the Joint Chiefs, Pace believed there was no military answer to the situation in Iraq. "In the military sense, you"d only commit the reserves if you were exploiting success or salvaging failure," said one general involved in the discussions, explaining that he didn"t seen anything happening from a relatively small troop increase. "It isn"t that we"re opposed to doing it, it"s just that we don"t see the payoff." This was a rational calculation, because at the time no one could predict that the Sunni insurgency would largely come over to the American side, or at least to the American payroll, or that Moqtada al-Sadr would order his Shiite militia to stand down.

Even Colin Powell, who though retired from active duty for more than a decade remained the best known military figure in the country, spoke out against the notion. Gen. Casey already had tried a surge in Baghdad in the summer of 2006, the former chairman of the Joint Chiefs of Staff argued in a December appearance on CBS"s Face the Nation. Face the Nation. So, Powell said, "I am not persuaded that another surge of troops into Baghdad for the purpose of suppressing this communitarian violence, this civil war, will work." At any rate, he added, "There really are no additional troops." So, Powell said, "I am not persuaded that another surge of troops into Baghdad for the purpose of suppressing this communitarian violence, this civil war, will work." At any rate, he added, "There really are no additional troops."

By mid-December, the notion of some sort of troop escalation, or "surge," was a major topic of conversation in both Washington and Baghdad. Yet Maliki"s government seemed lukewarm at best on the idea. After Gates met with Iraqi officials in a house in the Green Zone at this time, I b.u.t.tonholed Abdul Qadir Muhammed Ja.s.sim, the Iraqi defense minister, as he was leaving and asked him in the driveway if he had told the Americans that he supported the surge. Somewhat inscrutably, he responded, "I didn"t say no." Maliki was said to favor a "donut" approach-that is, put more U.S. troops outside the capital and leave the city to him, perhaps so the ethnic cleansing of Sunnis in several Baghdad neighborhoods could be finished. "I think they wanted to present us with a fait accompli of a Shiia Baghdad," Kilcullen suspected.

One of the very few voices in American public life supporting an increase in troops was that of Senator John McCain, who was in the difficult position of arguing that the war had been poorly executed but that more troops would improve the situation. "Without additional combat forces, we will not win this war," he said in mid-November.

On the morning of Thursday, December 7, President Bush sparred with reporters over Iraq. One asked if he were in denial about the state of the war. "It"s bad in Iraq," he replied with a glare. "Does that help?"

Actually, it may have. Finally, and years later than he should have, the president was beginning to grapple with the ugly facts on the ground in Iraq.

ONE WEEKEND AT AEI CHANGES THE WAR.

The 2003 invasion of Iraq arguably was conceived at the American Enterprise Inst.i.tute, the right-wing think tank that is the mecca of American neoconservativism. Its boxy building across from the National Geographic Society"s headquarters in downtown Washington, D.C., was the roost of a variety of prominent hawks-Fred Kagan, Richard Perle, Gary Schmitt, Tom Donnelly, William Kristol. In the fall of 2006, they saw their war going down the tubes. The same building also houses the Weekly Standard, Weekly Standard, the torchbearer publication of neoconservativism, and the Project for the New American Century, an advocacy group for an aggressive interventionist foreign policy that was an early and persistent advocate of ousting Saddam Hussein. the torchbearer publication of neoconservativism, and the Project for the New American Century, an advocacy group for an aggressive interventionist foreign policy that was an early and persistent advocate of ousting Saddam Hussein.

For years, the American right had been far more conflicted over the war than liberals generally perceive. Kagan, who had met with Bush at Camp David in June, long had thought that the war was mishandled. He was especially wary of Rumsfeld, who he thought was overly absorbed with restricting troop numbers in Iraq and insufficiently focused on providing the troops and other resources needed to prevail. After the Abu Ghraib scandal broke in 2004, Kagan and others called for the defense secretary to step down-an action that he believed had made him persona non grata at the Pentagon. Like Keane, he had grown increasingly concerned during the course of 2006, especially after a new Iraqi government was seated and violence increased rather than tapered off, as the Bush administration had predicted.

Bob Woodward"s thorough but White House-centric The War Within The War Within reports that the president was settling on a surge by November 2006. There is little on the public record to support that a.s.sertion. On the other hand, there is strong evidence that Bush"s meeting with outside experts early in December had a strong effect on his thinking. reports that the president was settling on a surge by November 2006. There is little on the public record to support that a.s.sertion. On the other hand, there is strong evidence that Bush"s meeting with outside experts early in December had a strong effect on his thinking.

The road to that meeting began on Friday, December 8, when Kagan gathered a handful of like-minded a.n.a.lysts and military planners on the top floor of the American Enterprise Inst.i.tute"s 12-story building. The most important person invited to the conference was Gen. Keane. Kagan had heard about his dissent on the conduct of the war, which was becoming the talk of conservative Washington, but had no inkling that Keane was talking regularly to Odierno and Petraeus, who were about to take command of the war.

The three-day exercise wasn"t intended to change the course of the war, or even to add more troops, which Kagan didn"t think was possible. Rather, it was to see if it was possible to devise an alternative military approach for Iraq. At any event, Kagan thought, it was quite possible that the exercise would be purely academic, because the word from the White House was that the president would be giving a major speech on Iraq within just a few days. "We were disappointed with the quality of the debate over the military aspects of the war," Kagan said. "Baghdad is burning, Iraq is about to explode, and we are moving toward a primitive civil war. This is about to head off the cliff. So, the mandate was: Stop the bleeding in Baghdad."

An untold aspect of the exercise at the think tank was the involvement of some active-duty Army officers. These weren"t just any random military planners. Rather, the ones attending had served with the 3rd Armored Cavalry Regiment in Iraq under Col. H. R. McMaster. That"s meaningful for two reasons. First, they had partic.i.p.ated in McMaster"s maverick 2005 campaign in Tall Afar, the first place in the Iraq war where the U.S. military conducted a successful large-scale counterinsurgency effort. They had shown how to use more troops effectively. Their example had not been embraced by top commanders, probably in part because it wasn"t consistent with stated U.S. strategy, and perhaps also because the Army didn"t see how to replicate the effort in larger cities: The McMaster approach worked in relatively small Tall Afar, but to many seemed inapplicable to the larger, more important ones, such as Mosul, Kirkuk, and, most of all, Baghdad, home to around 6 million souls. At any rate, the Army seems never to have taken a serious look at the feasibility of following the Tall Afar example elsewhere.

On top of that, at the time the American Enterprise Inst.i.tute exercise was being held, McMaster was playing a significant role across the river at the Pentagon as a member of the council of colonels reviewing Iraq policy for Gen. Pace, the chairman of the Joint Chiefs. He had been suggested for service on that panel by Gen. Petraeus. In the course of working on the study, he met Gen. Keane, and the two hit it off. Tying all these connections in a neat package, McMaster had shared an office with Kagan when both taught at West Point, and the former thanks the latter in the acknowledgments to his influential book Dereliction of Duty, Dereliction of Duty, about the failures of the Joint Chiefs of Staff during the Vietnam War. about the failures of the Joint Chiefs of Staff during the Vietnam War.

Another officer helping Kagan was Maj. Joel Rayburn, also a veteran of West Point. His involvement wasn"t official, he said, but rather grew out of months of "writing and talking and arguing" with Kagan about the war. "It became sharper as we went along-and became urgent in the fall of "06."

Not all the officers attending bought into the American Enterprise Inst.i.tute"s hawks" view of the world. "They completely blew it," said retired Army Col. Joel Armstrong, the former executive officer of McMaster"s regiment. "All the a.s.sumptions [about Iraq] were wrong." But when Kagan, spurred by McMaster, asked Armstrong to fly in from Spokane, Washington, for the weekend exercise, he agreed because he thought the United States was sputtering toward defeat in Iraq. "We were basically heading for a loss, and I couldn"t see anything changing without something dramatic," Armstrong said. It felt odd, he added, to advocate a course of action that the generals leading the war opposed. "I felt kind of strange going against the chain of command, but I felt I had to."

The basic concept was to figure out how to redeploy American troops in Iraq so that they might protect the population, which had become a major theme of Gen. Keane"s. Establishing outposts with that mission was an issue Armstrong had thought about since early 2005, when he had suggested putting a small, company-sized base in Salman Pak, southeast of Baghdad. "I was told by division I was out of my mind," he recalled. Donnelly and others from AEI sat with Armstrong and retired Maj. Dan Dwyer, another officer who had served in Tall Afar. "We"d look at a neighborhood, and use the 3rd Armored Cav"s experience to say, "Yeah, we could do that,"" Donnelly said. The military officers unrolled maps and Google images of Baghdad and began discussing what sort of troop numbers might be needed in its neighborhoods. There were company-sized problems, they would decide, and there were battalion-sized problems.

Armstrong"s role was to make the partic.i.p.ants loosely adhere to the military-planning process, which is basically to pose a problem, figure out a solution, decide which tasks are the logical steps toward reaching that solution, and then calculate what troops and other resources are needed to execute those tasks. Ultimately, they concluded that to improve security in Baghdad and neighboring al Anbar Province, nearly seven brigades would be needed-five from the Army, almost two from the Marine Corps. The next question was how to find those extra troops. On midafternoon Sat.u.r.day, AEI"s Donnelly and Armstrong, the retired colonel, had a quick discussion about the actual number of troops they thought doable. Looking at the Army"s planned rotation schedule, which he had found posted on the Internet, Donnelly quickly figured out how many combat brigades the Army could send. "The five-brigade answer was immediately obvious, in terms of the max that could be done in a timely fashion based upon the current force generation model," he said. (The calculations would prove to be so accurate that when Keane and Kagan went to brief Gen. Richard Cody, the Army"s vice chief of staff, on their recommendations, he offered only one tweak, which was that they were three weeks off in their estimate of the availability of one of the surge brigades.) The next major step was for an active-duty officer who was attending to "red team" the planning-that is, to look at the proposed operation from the enemy"s point of view. He discussed how al Qaeda might react, what the Shiite militias would do, the steps other fighters might take to counter the American moves. This was the moment "when I really came to believe this could work," Kagan said. "He persuaded me that we had a pretty good feel for the operational patterns of the enemy."

Keane mainly sat and watched, absorbing their thinking. "He was pretty quiet," said Armstrong.

Keane said later that he was impressed by the quality of the information and a.n.a.lysis the group presented. "I was amazed by the intel they got from open sources," he said. "I was very current but they understand pretty clearly what was happening among the factions; the level of detail that they understood was amazing to me." Much of what they told him, he noted, he would use a few days later in a meeting with President Bush. By coincidence, the White House had called Keane on Friday and asked him to come by on Monday to talk to the president about Iraq. He already had an appointment that day with Vice President Cheney, who also had gotten wind of Keane"s concerns, so Cheney"s office said the two meetings would somehow be combined.

One major a.s.sumption of the exercise was that improved security would lead to a political breakthrough. At first, this kind of slipped in as an afterthought. Donnelly said he wasn"t focused so much on what the surge might produce as he was on just getting one. "It was more a sense of, if you don"t turn the security issue around, you"re about to lose," he said. He was just trying to figure out a way to keep the United States in the war.

But Keane had what he needed to take to the White House the next day: an informed look not just at why more troops were needed, but how they might be used differently. This what was the White House had wanted to know, but the Pentagon hadn"t bothered to look into. It wasn"t a subject that seemed to interest the Joint Chiefs of Staff-a lapse that, to borrow McMaster"s t.i.tle, amounted to dereliction of duty.

BUSH GETS BOTH BARRELS.

On Monday, December 11, President Bush hit a new low in his ratings, with only 36 percent of respondents approving even "somewhat" of his performance and a stunning 62 percent disapproving.

At a White House meeting that began at 3:20 that afternoon, Keane listened as Professor Eliot Cohen began on a frank note. About a dozen high-level note-takers-Karl Rove, National Security Adviser Stephen Hadley, and some of his staffers-sat in the outer circle. Cohen, the sole attendee from the Camp David meeting the previous June to be asked to this session, remembers it as being far harder edged than the desultory discussion six months earlier. "There was something in the air-more tension," he recalled.

Stephen Biddle, a defense expert at the Council on Foreign Relations, later said that going in, he had half-expected the session to be a photo opportunity, intended to demonstrate that the president did indeed talk to smart outsiders. Instead he found the tone of the hour-long meeting to be open, even confrontational. "It was obvious that the president and Cheney were taking this seriously," he said. "The president had a drawn face, was very subdued, looked depressed."

Cohen was determined to be clearer and more emphatic than he had been the previous June at Camp David. "Mister President, I"m going to be very blunt," he began. "I don"t mean to cause offense, but this is wartime, and I feel I owe it to you." He also owed it to his own family and friends: Not only were many of his former students at Johns Hopkins University"s School of Advanced International Studies military officers on duty in Iraq, so was his own son, a recent Harvard graduate who deployed there as an Army lieutenant specializing in military intelligence.

This time Cohen hit the issue of generalship squarely. It was high time to get a new team and a new strategy in Iraq, he advised. "It"s not enough to say these are good guys-of course they are good guys. The question is, are they the right right guys?" He said he didn"t think so. He urged the president to hold them accountable. guys?" He said he didn"t think so. He urged the president to hold them accountable.

He also talked about how presidents need to push their military advisers. "Generals disagree, sometimes profoundly," he said, citing a lesson he knew both from his academic work and from his time a.s.sessing Iraq for the Defense Policy Board. "Civilian leaders need to discover these disagreements, force them to the surface, and probe them. This is what Lincoln and Roosevelt did. LBJ"s failure in Vietnam was not micromanagement, but failure to force serious strategic debate." Cohen, who is steeped in military history, was on solid ground in rejecting the conventional wisdom that President Johnson"s error had been to meddle too much. Retired Army Brig. Gen. Douglas Kinnard, who surveyed Army generals who served in Vietnam for his study The War Managers, The War Managers, wrote that it is possible to argue that "there was not enough civilian partic.i.p.ation in terms of asking the big questions about what we were really doing in Vietnam." As Cohen himself had pointed out in wrote that it is possible to argue that "there was not enough civilian partic.i.p.ation in terms of asking the big questions about what we were really doing in Vietnam." As Cohen himself had pointed out in Supreme Command, Supreme Command, during World War II, Winston Churchill also injected himself into the smallest of issues, but while doing so he never lost hold of the big strategic picture. during World War II, Winston Churchill also injected himself into the smallest of issues, but while doing so he never lost hold of the big strategic picture.

Cohen, whom retired Gen. Barry McCaffrey, also at the meeting, sometimes called "Mr. Bow Tie," also questioned the nature of advice Casey was providing, in which the general sought to balance his needs in Iraq with the state of the Army. That wasn"t Casey"s job, Cohen said. There were plenty of people at the Pentagon paid to take care of the Army. Casey"s mission was very different, he said. His job was to win the war. "Not all generals are up to the task," he advised, knowing, for example, that well over a dozen division commanders had been relieved during World War II. Yet the Bush administration handled its generals as though they were all equally successful, interchangeable parts. "Not a single general has been removed for ineffectiveness during the course of this war." The Army needed a push here, he noted. "The current promotion system does not take into account actual effectiveness in counterinsurgency. We need not great guys but effective effective guys. Routine promotion and a.s.signment systems for generals in wartime is a disaster." guys. Routine promotion and a.s.signment systems for generals in wartime is a disaster."

Keane, speaking second, was also emphatic. "Mister President, to my mind, this is a major crisis," he began. "Time is running out." We need more troops, he said. And more important, he continued, we need to use them differently. "For the first time, we will secure the population, which is the proven way to defeat an insurgency," he explained. "In time the troops will be more secure, but I can"t hide from you that the casualties will initially go up. In any counteroffensive operation that we have ever done, from Normandy to the island-hopping campaign in the Pacific, Inchon in Korea, multiple ones in Vietnam, casualties always go up, because you are bringing more troops and more firepower to bear on the problem."

To Keane"s surprise, the two other retired generals at the meeting, McCaffrey and Wayne Downing, disagreed with his call for an escalation, while the two academics there, Cohen and Stephen Biddle, supported it.

"No more U.S. forces," argued Downing.

McCaffrey explained why. "Sir, I have known Jack Keane since we were young officers," he began. "I have great admiration for him. But this so-called surge is a fool"s errand. Yes, it will have a short-term impact. But it isn"t sustainable. The solution is Iraqi forces." But even as he spoke, McCaffrey began to suspect that the president really wasn"t listening to his view and had already made up his mind. Indeed, a study of personnel-mobilization issues a.s.sociated with the surge done by William Luti, a staffer on the National Security Council, already had concluded that a surge was doable both in terms of the effect it would have on readiness and on how long the troops would stay. "I think it went in one ear and went out the other," McCaffrey said later. "I don"t think the president was listening. Cheney was-he was taking extensive notes."

Bush asked them what to do with the advice, especially about selecting generals. "All well and good, but how am I supposed to know, and who I am supposed to pick?" Bush responded, according to Biddle, who spoke last.

"David Petraeus," said Cohen. His thinking, he recalled was that "all armies get it wrong at the beginning, as [the great British military historian] Michael Howard says-the question is who adapts fastest." Cohen believed that Petraeus was the general who while serving in Iraq had best shown the ability to adapt. Keane and McCaffrey seconded the idea. (McCaffrey believes he was first to mention the name, but others disagree.) All the invitees were in accord that Petraeus was the only serious candidate for the job.

Then why is he disliked by some people at the Pentagon? Bush asked, apparently referring to some supposed friction between Petraeus and Rumsfeld, who liked to be the smartest person in the room.

Don"t worry about that, the partic.i.p.ants said. "You got to go with this guy," McCaffrey responded, with Keane supporting him.

Cheney asked only one question during the session. As it ended, McCaffrey watched as the vice president took Keane down the hall with him, which he thought confirmed his hunch that "the fix was in." Asked about this, Keane said that he was just going to his existing appointment with Cheney to go through the details of how a new counterinsurgency strategy might be implemented in Iraq.

Not long after, Keane got a call from a White House official telling him that the meeting had had a decisive impact on the president"s thinking. A small group of NSC staffers had been pushing for a troop surge for weeks, pointing to the examples of Tall Afar and Ramadi. Now this group, which dubbed itself "the surgios," had been given ammunition by a respected group of outsiders.

Despite that, Bush continued to hold his cards close. He would say later in the month, "I haven"t made up my mind yet about more troops."

THE COUNCIL OF COLONELS UNLOADS.

Two days later, on Wednesday, December 13, Bush traveled to the Pentagon, where Gen. Pace briefed him on the ominous findings of the council of colonels.

Despite general discouragement, the group had been unable to find a consensus on the war, especially on whether to escalate. "The Air Force and Navy guys were clearly anti-surge. But Mansoor, H. R. McMaster, and the Marine, Colonel Greenwood, were for it," said one Pentagon official who worked with the council. These were the three most influential members of the council. After commanding a brigade in Baghdad, Peter Mansoor had become a counterinsurgency adviser to Petraeus at Fort Leavenworth. Tom Greenwood had served several years on the staff of the National Security Council and then done two tours commanding units in Iraq. H. R. McMaster was probably the most prominent colonel in the Army at this point. All three agreed with Keane: Put more troops into Iraq even if it means breaking the Army. That"s what you do in war, this official said: "You serve the national interest."

Nor did Pace, the chairman of Joint Chiefs, who had ordered the study, settle the argument. "Pa.s.sive Peter Pace, he was looking for the path of least resistance," this Pentagon official continued. "He brought no strategic vision, and no determined leadership-and the nation was at war. He is an honorable, genuinely nice man, but a tool for others."

The council ultimately recommended a small increase in forces, but nothing like the surge that eventually would occur. Yet the group"s minority view in favor of a bigger escalation, although put aside at the time, ultimately would have more staying power and impact, because its leading advocates, Mansoor and McMaster, would be asked by Petraeus to come to Iraq. The colonels" group, along with the barrage of pointed criticism being gathered for the development of the counterinsurgency manual, played a little noticed but helpful role in civil-military relations: Together they quashed the growing view among officers that the U.S. military had performed marvelously in Iraq but had been let down by the rest of the government, the Congress, the media, and the American people. "There had been a "stab in the back" school emerging, but then there was a point at which the Army turned introspective and said, "You know, we can do this better,"" said Lt. Col. Paul Yingling, a veteran of three tours in Iraq.

It wasn"t quite put to Bush in such blunt terms at the Pentagon meeting, but the council of colonels had concluded that the United States had invaded Iraq on the basis of a series of flawed a.s.sumptions-that Bush and others wrongly a.s.sumed that it would be a war of liberation, that Iraqis would take over power quickly, and that the country would remain more or less orderly, with a functioning police force. Likewise, the Bush administration was operating on some a.s.sumptions that badly needed to be examined. Can the Iraqi government survive? Is it under the control of Iran? Does it have staying power? Are Iraqi Security Forces truly a national inst.i.tution? Will they crack if the civil war spreads and deepens? Will neighboring powers, especially Iran, become more involved in Iraq? And, most important of all, are we past the point of no return? If so, how do we reposition ourselves to minimize the damage? If not, what do we do next? Does anybody in Iraq believe we will stick around? The point was that until these questions were thought through, the United States wouldn"t have a strategy; it would have only aspirations.

Those sharp questions weren"t put to Bush, but the group"s findings were. Even those conclusions were a sharp departure from Bush administration policy. Iraq was indeed in a "low-grade" civil war, the colonels said, and on the path to a bigger one. "We are losing because we are not winning and time is not on our side," they flatly stated, according to the doc.u.ment summarizing their work. Their warning that Iraq was experiencing "an intensifying civil war" was a finding that both the Pentagon and the White House had long resisted. (The slide containing these key findings on the Iraq war appears in this book"s photo insert.) The colonels also pointed to a basic problem with the approach then in place: "Current . . . troop strength was inadequate to secure the population (cannot execute "clear, hold and build")." In other words, the Americans were trying to implement a strategy for which they hadn"t devoted enough people. The point of that finding, recalled Greenwood, the Marine colonel who was a member of the council, was that, "unless troop levels were increased dramatically, the U.S. strategy would remain bankrupt."

The report amounted to a blinking red light for the U.S. effort in Iraq. Either the strategy had to change or more resources had to be devoted to making the current strategy more plausible. In any event, the colonels were saying, we should not try to stay the course. "In every measurable category we were failing, and in fact we were on the path to defeat," summarized Mansoor, another member of the group.