The colonels had split over the way forward. Mansoor advocated sending more troops as part of a plan to eventually have a long-term but smaller U.S. military presence in Iraq, an approach he called "Go Long." But the majority of the group, and especially its members from the Air Force and Navy, opposed the idea of a broad escalation-that is, a troop surge. "What the group said was, if we spike, it should be for a specified purpose, in order to, say, take down the Jaysh al-Mahdi in Sadr City," Mansoor said, referring to Moqtada al-Sadr"s Shiite extremist militia, also known as "JAM." "They would have opposed just an increase in counterinsurgency forces." Even more broadly, there were doubts in the group about whether the U.S. military really had enough troops available to escalate sufficiently to make a difference.

The council had laid three choices before the Joint Chiefs of Staff. "The options we had were "Go Big," "Go Long," and "Go Home,"" said Mansoor, who was one of the group"s most influential members. His "Go Long" approach eventually would become U.S. policy.

The next day, Thursday, December 14, Keane and Kagan went to give a more detailed briefing to Cheney. Being a former defense secretary, Cheney asked questions that were informed and precise. But being Cheney, he gave away very little. "I didn"t get any feedback," Kagan shrugged.

Military resistance was still running high after the pair"s two White House meetings. Underscoring that, the next day Keane and Kagan were scheduled to fly down to Central Command"s rear headquarters in Tampa, Florida. At the last minute, with their tickets in hand, they were told that their visit was canceled. They were offered no explanation. Treating a respected four-star general such as Keane in that backhanded way was a strong indication of how the military chain of command felt about having its thinking second-guessed by White House staffers, think tankers, and a retired general.

Indeed, Casey, still in place as the top American commander in Iraq, would argue for another two weeks against a big troop increase. He didn"t know that he already had lost both the debate and his job. He had been planning to leave Iraq in the spring, but was told in December that he was being moved out in just a few weeks. "I left not really understanding what the h.e.l.l had happened." Even in 2008, he added, "The whole process with the surge and how that came to be" remained murky to him. Bush told others that he didn"t want to appear to be blaming Casey for executing a strategy that the president had approved, so Casey was to be given a "soft landing." The general would be named Army chief of staff-ironically, the same position given Gen. William Westmoreland upon his removal from command of a failing war in Vietnam in 1968.

A few days later, Vice President Cheney called Keane and asked if he would consider coming back on active duty and taking command in Iraq. Keane declined but said he was close to Petraeus, and knew and liked Odierno, and was willing to travel to Iraq as needed to advise both generals. This outcome was a victory for Keane, but it raised troubling questions about the ability of the chairman of the Joint Chiefs of Staff and his peers at the Pentagon to carry out their mission of figuring out how to fight and win. "Why did the American military establishment so fail to come up with a war-winning strategy that it was up to a retired general and a civilian think tank, AEI, to do their job?" asked retired Army Col. Bob Killebrew. "This is a stunning indictment of the American military"s top leadership."

SADDAM IN THE AIR.

Before dawn on the morning of Sat.u.r.day, December 30, Saddam Hussein was taken to the former headquarters of the Iraqi military intelligence service. As an official read aloud his death sentence, he interrupted, "Long live the people, long live jihad, and long live the nation." He added, "Down with the Persians and the Americans." The hangman arrived, and Saddam met with a Muslim cleric.

His bizarre ending was illicitly captured by someone in the room using a cellular phone camera. The condemned man, wearing a white shirt, black pants, and newly shined shoes, stepped to the gallows platform, his hands cuffed behind his back and his legs tied together. He appeared stone-faced, trying in his last act on earth to salvage whatever dignity he could from the day. A yellow noose was placed around his neck. It was Saddam"s turn to be interrupted. Someone yelled, "Moqtada, Moqtada, Moqtada"-a shout of triumph invoking Moqtada al-Sadr, the radical Shiite cleric who rose to great power in the wake of the American invasion.

Saddam, who had never been free to witness the rise of Shiite power in post-Saddam Iraq, responded sarcastically: "Moqtada?"

"Go to h.e.l.l," someone shouted.

"Long live Muhammed Bakr Sadr," someone else shouted, referring to Moqtada al-Sadr"s uncle, a founder of the Dawa Party.

"The man is facing execution," pleaded Saddam"s prosecutor, Munqith al-Faroun, who was also present. "Please don"t."

Saddam dropped through the scaffold"s trapdoor as more than one thousand pounds of torque snapped his neck. A few minutes later he was p.r.o.nounced dead. In both its lethality and unruliness, the event was, somehow, a fitting end to 2006 in Iraq. (Two weeks later, when Saddam"s half brother was hanged, his head came off, provoking more disapproval.) But that worst year of the war had one more ugly day left in it. On New Year"s Eve, two soldiers were killed by a roadside bomb while on patrol in Baqubah and a third was killed in Baghdad, eerily bringing the U.S. military death toll in the war to precisely 3,000.

A STRATEGY IS BORN.

(Winter 2006-7)

When he left Iraq at the end of his first tour of duty early in 2004, Raymond Odierno, then a major general, had believed the war in Iraq was going well. "I thought we had beaten this thing," he said several years later, as he began his second tour. "I could walk down the streets of Kirkuk, Tikrit, Baqubah, Samarra. I can"t do that now." Returning in the fall of 2006 on a preliminary trip as a lieutenant general, a corps commander preparing to take command of the day-to-day operations of the war, he was stunned by how much worse Iraq felt. In the two big operations to improve security in Baghdad, "They had cleared areas but were unable to hold them."

He also began to worry that he was being set up to suffer defeat on his watch. "I felt like we were in the process of leaving, no matter what the consequence," he said in one of a series of candid interviews. "I felt very disconcerted about that, because I had no control over that."

Oiderno"s strongest trait is determination. After that preliminary visit, he returned to his base at Fort Hood, Texas, thinking about how not to lose the war. On November 28, 2006, he and his key staff began the long flight from Texas to Iraq. Maj. James Powell, whose longish shock of hair combed sideways lent him a resemblance to a 1940s W. H. Auden, was exhausted and went to sleep. Somewhere over the Atlantic, Powell opened his eyes to see his boss, Col. Martin Wilson, the head of the corps" planning office, standing before him and staring at him. Wilson had come to Powell directly from a conversation with Odierno, who had given him and his other key staff officers these orders: Come up with a plan to retake Baghdad.

"When I got here, the situation was fairly desperate, frankly," Odierno said. "The only thing I thought would decisively change it was doing something in Baghdad, and the only way to do that was to increase forces. I didn"t know if it would work."

If Jack Keane was the spiritual G.o.dfather of the surge, Odierno was its biological parent. Petraeus, arriving in Baghdad two months later, would become its adoptive father. In order to position the U.S. for a new strategy-that is, get additional brigades and use them differently-Odierno had to take on his direct superior, Gen. Casey, who in turn was backed by the entire chain of command. It is extraordinary to consider that the new strategy that would be implemented by the U.S. military in Iraq in 2007 was opposed by the U.S. military in both Baghdad and Washington. With the exception of Odierno, it came from outside the military establishment.

LT. GEN. RAYMOND ODIERNO.

Odierno was an unlikely savior to appear in the midst of the Iraq war. There was little in his past to indicate that he would buck his entire chain of command and push the U.S. military in Iraq in a radically new direction. After growing up in New Jersey, he attended West Point and played varsity baseball there, most notably pitching against Ed Kranepool and Dave Kingman of the New York Mets in an exhibition game. He graduated in 1976 and became an artillery officer. He gained a reputation as the best of the conventional thinkers-intelligent, industrious, ambitious, and focused on using the tools the Army provided him, rather than discovering new and different ones. He deployed for the 1991 Gulf War and again for the Kosovo campaign eight years later.

During his first tour in Iraq, commanding the 4th Infantry Division in the middle of the Sunni Triangle in 2003-4, he and his subordinates earned a reputation as heavy-handed, kicking in doors and rounding up tens of thousands of "MAMs"-military-aged males. As I wrote in my book Fiasco, Fiasco, about the first three years of the Iraq war, when Brig. Gen. Barbara Fast, the top Army intelligence officer in Iraq, questioned the 4th ID"s indiscriminate approach, she was told by its intelligence officer, according to a subsequent Army report, that Odierno didn"t care. "The division commander did not concur with the release of detainees for fear that a bad one may be released along with the good ones," Maj. Gen. George Fay wrote. about the first three years of the Iraq war, when Brig. Gen. Barbara Fast, the top Army intelligence officer in Iraq, questioned the 4th ID"s indiscriminate approach, she was told by its intelligence officer, according to a subsequent Army report, that Odierno didn"t care. "The division commander did not concur with the release of detainees for fear that a bad one may be released along with the good ones," Maj. Gen. George Fay wrote.

Fast said in a statement to investigators that Odierno"s att.i.tude was "We wouldn"t have detained them if we wanted them released." She asked retired Army Col. Stuart Herrington, a veteran intelligence officer specializing in interrogation, to review the way intelligence was gathered from detainees. Herrington concluded that some U.S. commanders, in seeking to shut down the insurgency in their areas of operations, were using tactics that effectively made them recruiting sergeants for it. Herrington was especially bothered by the actions of Odierno"s division. "Princ.i.p.ally due to sweep operations by some line units-the 4th ID was consistently singled out as the major offender-the number of detainees" was rising steadily, he wrote in his report to Fast. He emphasized that point five pages later: "Some divisions are conducting operations with rigorous detention criteria, while some-the 4th ID is the negative example-are sweeping up large numbers of people and dumping them at the door of Abu Ghraib."

"Odierno, he hammered everyone," said retired Army Lt. Gen. Joseph Kellogg Jr., who was serving with the U.S. occupation authority in Baghdad.

"The 4th ID was bad," said an Army intelligence officer who worked with them. "These guys are looking for a fight," he remembered thinking. "I saw so many instances of abuses of civilians, intimidating civilians, our jaws dropped."

Lt. Col. David Poirier, who commanded a military police battalion attached to the 4th ID, said he found the division"s approach to be indiscriminate. "It became a philosophy, "Round up all the military-age males, because we won"t know who"s good or bad." "

Col. Alan King had a similar view that the division helped the insurgency. "Every male from sixteen to sixty" that the division caught was detained, he said, "and when they got out, they were supporters of the insurgency."

A subsequent review by the Army inspector general said interrogators reported "detainees arriving at the cage badly beaten. Many beatings occurred after the detainees were zip-tied by some units in 4ID. Some units wouldn"t take THTs [Tactical Human-Intelligence Teams] on raids because they didn"t want oversight of activities that might cross the line during capture."

The most striking instance of abuse in Odierno"s division occurred in January 2004, when some 4th ID soldiers forced two handcuffed detainees to jump into the Tigris River. One reportedly drowned. Lt. Col. Nathan Sa.s.saman, their battalion commander, believed, as he put it later, that "we had to convince the Iraqi people that they should fear us more than they feared the insurgents." He obstructed Army investigators trying to determine what happened with the detainees at the Tigris. Lt. Jack Saville would testify at his court-martial that he had discussed with Sa.s.saman how to mislead those investigators. Despite that, Sa.s.saman received only a written admonishment from his commanding general. "On 7 January 2004, you were briefed . . . that soldiers of the 1st platoon pushed two Iraqi men into the Tigris River causing one of them to drown," Odierno wrote. "You ordered them to deny that the men were pushed into the river and to say that they were dropped off at the side of the road. Your conduct was wrongful, criminal and will not be tolerated."

Despite that conclusion, Odierno let Sa.s.saman remain in command of the battalion for months, an outcome that shocked Poirier. "When you have a battalion commander who leads his staff in rehearsing a story about a murder-and he"s still in command?" Poirier said. "That"s not right."

It isn"t clear what led Odierno to reconsider his approach. He spoke at Basin Harbor, an academic conference on counterinsurgency run by Eliot Cohen in June 2005, and left the experts on the subject there distinctly unimpressed. He began by congratulating those present for holding such a conference because, he said, according to several partic.i.p.ants, "we in the Army don"t think that much about counterinsurgency"-a comment that provoked some eye rolling. "He indicated he knew there were different ways of dealing with insurgency but also showed he was not certain that population-centered COIN was the way to do it," said retired Marine Col. T. X. Hammes, one of the specialists who attended the meeting. In comments the following morning, some said they found him utterly conventional and so not really focused on the task at hand in Iraq.

Between his first and second tours in Iraq, Odierno served as a.s.sistant to the chairman of the Joint Chiefs of Staff, a position that has evolved into being the role of the U.S. military"s amba.s.sadorship to the rest of the national security establishment. For example, the general holding the job frequently accompanies the secretary of state on overseas travel, in order to provide a military perspective as needed and also to keep the Pentagon attuned to U.S. diplomacy. This may have been the work that best prepared Odierno for the role he would play later in Iraq, as it broadened his view and acquainted him with the personalities and levers of power at State, CIA, and the White House. Notably, this was during a period when a schism was developing inside the government about the trend in Iraq, with State and CIA being far more pessimistic than the military or the White House. Odierno would return to Iraq fully aware that many well-informed officials thought the war was going badly, far worse than the White House, or even the generals in Baghdad, seemed to understand.

Odierno was prepared, even determined, to operate differently on his second tour of duty in Iraq. "I think that before, he thought that through force you could achieve anything in the world," said Emma Sky, a pacificistic British woman who would have an important role on his staff the following year. "I think he now has a much more sophisticated understanding of how society works," she said. "I think he"s learned a tremendous amount." Still she said she didn"t agree with how his time commanding the 4th ID was portrayed in Fiasco. Fiasco. She didn"t believe, she said, "that he somehow has turned from an ogre into a good man." She didn"t believe, she said, "that he somehow has turned from an ogre into a good man."

He didn"t much want to talk about how his transformation occurred, brushing aside the question. "I think everyone"s changed," he said. "We"ve all learned. We came in here not thinking about counterinsurgency." He didn"t like to dwell on those old errors, but sometimes in discussing details of current operations, he would refer to them. "One of the mistakes we made early on-and one of the mistakes I made-was not understanding the importance of the tribes." He also pointedly and ruefully remembered being told by one of his superior officers in May 2003 that the fighting likely would last as long as 45 days. Odierno gives the impression that he doesn"t much care what reporters and politicians think about him, but very much what his peers in the Army do.

Part of the change may stem from his knowledge that he most likely would be reporting to Petraeus, who had been steeped in counterinsurgency theory. Odierno is "a good leader, charismatic, and a tactically competent soldier," said Donnelly, the defense expert at the American Enterprise Inst.i.tute. "I think that intellectually, he has learned from his mistakes. And I think that working for Petraeus makes all the difference in the world. It is very different from working for [Lt. Gen.] Sanchez and getting no education."

Another impetus to change, Odierno agreed in an interview, was the severe wounding of his son in August 2004. Lt. Anthony Odierno, then in the 1st Cavalry Division, had been leading a patrol near Baghdad"s airport when a rocket-propelled grenade punched through the door of his Humvee, severed his left arm, and then mortally wounded his driver. "It didn"t affect me as a military officer, I mean that," Odierno said one evening in Baghdad much later. "It affected me as a person. I hold no grudges. My son and I talked a lot about this. He was doing what he wanted to do, and liked what he was doing." But he said it did deepen his determination. "I was going to see this through-I felt an obligation to see this through. That drives me, frankly. I feel an obligation to mothers and fathers. Maybe I understand it better because it happened to me."

But the most important part of it was likely the growing recognition in the fall of 2006, as he prepared to redeploy, that the war was heading toward defeat, and it might occur on his watch. He didn"t want to lose and realized that meant taking on his new boss, Casey. On a predeployment trip to Iraq in August, he had been told that the plan was to cut combat troop levels by as much as one-third in 2007. On December 4, he was briefed on the plan he would be implementing, called the Bridging Strategy, which of course was referred to in U.S. military circles as "TBS."

This was an all-important briefing for Odierno, because it amounted to his overarching orders for the next year. Officially, his job was not to a.s.sess and challenge these goals, but to figure out how to achieve them. The key points were: "Move outside all major cities" and establish a handful of even bigger bases along key roads leading into Iraq, deploy U.S. forces to Iraq"s borders in order to limit outside influence, speed up the transition to Iraqi security forces, and let Iraqis handle fighting in the cities.

Together this plan (the briefing is reprinted in the appendix) amounted to a half withdrawal, not leaving Iraq but hanging on its periphery. The more Odierno and his planners considered this plan, they less they liked it. They feared that it got ahead of the ability of the Iraqis to do the job, and so, in keeping with the American pattern in Iraq from 2003 on, would likely amount to one more rush to failure. He was, he recalled, "very nervous" about the course of U.S. strategy. He decided he would formally oppose any additional troop cuts. He wasn"t even thinking about a surge, because, he said, "I didn"t think I could get more."

After taking command in December, Odierno and a small group of advisers met almost every night for several weeks, trying to figure a way out of the jam. Ultimately, they would decide on a course almost the opposite of the plan given them. Instead of moving out of the cities, they would deploy more into them.

Instead of consolidating their base structure, they would establish scores of smaller outposts. Nor would they withdraw to the borders. And most emphatically, they would slow the transition to Iraqi forces. He realized that to take all those steps he would need more troops-something Keane had been telling him in telephone calls. "Odierno was standing up to Casey, and he deserves a lot of credit for it," said Keane, who was trying to get to Iraq. "I was trying to go see him, but Casey wouldn"t let me come in." Keane called Casey directly, but the general put him off, saying, "It"s a really bad time."

Petraeus also was talking to Odierno about the notion of a major increase in troops. When Casey got wind of it, he had his executive officer call Petraeus with a message: "Hey, man, don"t be calling Ray." Petraeus responded that he had been asked by the chairman of the Joint Chiefs to look at Iraq and so needed to talk to senior officers there.

By mid-December it was clear to subordinates that Casey and Odierno were at odds. "Casey fought it all the way," recalled Brig. Gen. Joe Anderson, Odierno"s chief of staff. Planners were beginning to scratch their heads. Odierno was telling them to figure out how they might employ several brigades. "We would backbrief one general and get one set of guidances, and then brief the other and get a different set," remembered one senior Army planner in Iraq.

Keane, who was talking to Odierno once or twice a week by telephone, told Odierno that he should ask for five brigades. But when Odierno raised that number with Casey, his commander threw cold water on the notion. "He said, "You can do it with two brigades," " Odierno recalled. "I said, "I don"t know."" Casey eventually said that they could deploy only one brigade a month anyway, so he would commit to two brigades, and then make a decision about each additional one. Casey"s notion, coming off the disappointment of the Together Forward operations the previous summer, was to stiffen the backbone of Iraqi forces operating in Baghdad by putting them in some joint outposts with American troops. He figured it would take about two U.S. brigades-perhaps 7,500 soldiers-to do that. He wasn"t particularly interested in the other three brigades.

Hearing about the two-plus-three compromise, Keane first hit the roof and then called the White House. Odierno "started with one brigade, and Casey was fighting him over the first brigade," Keane recalled, "and then he finally gave him one. And then he fought for a second one and then they came up with this G.o.d-forsaken strategy of having three on hold in the U.S. and advancing them one at a time as needed, advance to Kuwait and then go from Kuwait to Baghdad. I knew what was going on here. Its pretty obvious-Casey doesn"t want his strategy dramatically changed."

Keane, believing that Casey"s incremental approach was unsound both militarily and politically, warned Hadley, the national security adviser, that having three units on hold would mean that there would be a separate debate about each one being sent. "Just think about what"s going to happen," he told Hadley. "You are not going to be effective in bringing down the violence with only two additional brigades, therefore you will call for an additional brigade three separate times, each time because we do not have sufficient troops. The media will be all over you for failing three more times. Meanwhile, the president is going to bite this bullet; he should only bite it once. He shouldn"t bite it one time and then three more times."

Hadley agreed: "Yeah, this makes no sense."

Keane told White House officials that "there was a huge struggle going on inside the command." He also broke the news to Hadley that the surge wouldn"t conclude by midsummer. "Can we get that wrapped up in, what, five or six months?" he recalled Hadley asking him. "That"ll be the summer and we can hold on to the wobbly Republicans because we"ve made some progress."

h.e.l.l no, Keane responded. "It will take all that time to get them in there. . . . It"s going to take twelve to eighteen months before we can realistically start reducing these forces."

Odierno told his planners to think about how they would use five additional brigades of combat troops. The planners were puzzled-they didn"t think he would ever be given that many. They didn"t know how he had come up with that number. "That stuff just kind of shows up," said Lt. Col. Jeff McDougall, one of his top planners. "You don"t know, sitting in our little dungeon, where that stuff comes from."

One day in December, Odierno told his planners, "We have to secure the population, first thing. We have to get back out into Baghdad." They and Odierno thought that they really needed about eight brigades but knew that no more than five were available, and that it would take months to get them all to Iraq. Odierno thought that the shortfall could be made up somewhat by adding other, smaller units, such as Marine battalions, a helicopter squadron, and some Special Operations troops.

Casey"s resistance was being supported by the chief of the Army and other members of the Joint Chiefs of Staff. The notion of escalation had been floating around, and most of the chain of command was against the idea. "I was not supportive at the time because of several concerns that I had," said Gen. Pete Schoomaker, the then chief of staff of the Army. "First, that no one had articulated or had established a clear purpose for the surge, including how to know when that purpose had been achieved." Also, he thought it would hurt the readiness of the Army without any clear payoff.

Casey finally agreed to an increase of two Army brigades and two Marine battalions, plus one more brigade on tap in Kuwait, recalled Maj. Gen. Barbero, who at the time was watching from the Pentagon as the J-33, or deputy director for current operations, for the staff of the Joint Chiefs. (A brigade has about 3,500 soldiers, while a Marine infantry battalion has about 750.) This was more or less where the council of colonels had come out. Barbero thought it was exactly the wrong answer and said so at Pentagon meetings. "This is the worst course of action we could do," he argued, "enough to put a strain on the force, but not enough to do anything."

Barbero had once commanded a battalion in the 101st Airborne, and he also was talking to his old division commander, Jack Keane. "He was the first person I heard the number five brigades from," he said. And, typical of the web of personal connections that every military career produces, he also had been the a.s.sistant commander of the 4th Infantry Division from 2003-4-serving under Raymond Odierno.

As Odierno"s planners worked on a big surge, the political context of the war began to change-and that tipped the balance in their favor. Back in Washington, the Iraq Study Group recommended accelerating the American turnover to Iraqis. This was far different than what Odierno was advocating but was pretty much what Casey had told him to do, except faster-that is, get U.S. forces out of combat and speed the transition to Iraqi security forces. The biggest difference was that the group recommended a tighter deadline, of getting U.S. combat forces out by early 2008.

But, coming just a month after the November 2006 elections, the political effect of the Iraq Study Group may have been more significant. "The situation in Iraq is grave and deteriorating," began its report, released on December 6. This grim finding, the consensus view of political and military experts from across the spectrum, ultimately was the real contribution of the group. "It stopped all the happy talk about how well things were going and how the press was reporting it wrong," noted former defense secretary Perry, one of its members.

Just before the elections, the president had said, "Absolutely, we"re winning." On December 6, the same day that the Iraq Study Group released its report, a new defense secretary, Robert Gates, who had been a member until his nomination, was confirmed by the Senate. He was sworn in December 18. The next day, Gates"s first full one at the Pentagon, President Bush said for the first time, "We"re not winning, we"re not losing in Iraq"-a striking turnaround from his formulation the previous month. (He also said that he was planning even before the election to replace Rumsfeld.) A month later, the president would go another step and say that the course he was on in 2006 was "maybe a slow failure."

Gates is a surprising man. The white-haired former CIA director is calm, quiet, and soft-voiced, reserved almost to the point of seeming humble-a presentation that blurs his tough-minded nature. One key to his nature is an observation he offers in his 1996 memoir, From the Shadows, From the Shadows, that some of the most effective U.S. officials he had seen were Henry Kissinger, Zbigniew Brzezinski, and George Shultz. The commonality among them, he concluded, was that they were "basically hawks who drew extensively on the ideas and initiatives of the doves." that some of the most effective U.S. officials he had seen were Henry Kissinger, Zbigniew Brzezinski, and George Shultz. The commonality among them, he concluded, was that they were "basically hawks who drew extensively on the ideas and initiatives of the doves."

In office Gates would prove both less windy and more decisive than his predecessor. "The interesting thing about Rumsfeld is that he didn"t make decisions," defense expert Anthony Cordesman said after Rumsfeld"s firing. Referring to the defense secretary during the Vietnam War to whom Rumsfeld was often likened, he added, "McNamara at least made decisions. Rumsfeld micromanaged, but he didn"t make decisions." Gates also would bring a skeptical view of the Iraq war shaped by his time on the Iraq Study Group. He had been at the Pentagon just two days when Gen. Abizaid, the head of Central Command, announced that he planned to retire a few months later. "He was told, "It"s time to go, we need some fresh air here,"" recalled an officer who was at Centcom at the time. Abizaid disputes that, saying that he had submitted his request to retire months earlier, and that it had been approved by Rumsfeld.

In any event, it would soon become evident that behind the new defense secretary"s slight smile lurked a very sharp set of teeth. In the spring of 2007, he would fire Army secretary Francis Harvey and the Army surgeon general over their sluggish handling of a scandal about the poor treatment of recuperating soldiers at the Walter Reed Army Medical Center. That June, he would ease out the chairman of the Joint Chiefs, Gen. Pace, who at that point was the last senior military figure a.s.sociated with the botched handling of the first three years of the war who was still in office. In March 2008, he would throw overboard Adm. William "Fox" Fallon, Abizaid"s successor. Three months later he would simultaneously oust the Air Force chief of staff and the Air Force secretary over lapses in that service"s handling of nuclear weapons. He executed these decisions in a soft-spoken way, showing no emotion. It was the opposite of Rumsfeld, who barked but appeared to have no bite.

Indeed, Gates was greeted as a liberator at the Pentagon simply because he wasn"t Rumsfeld. "You can already feel the stability," said retired Air Force Gen. Charles Wald, formerly the deputy U.S. commander in Europe, who deemed the appointment of Gates to be the Bush administration"s best move in years.

In the following months, he would be greeted with similar relief in Baghdad. "He seems to be just about everything you want in a sec def," said Charlie Miller, an aide to Petraeus in 2007 who would sit in on several video-teleconferences with Gates. "It"s too bad we didn"t have him from the get-go, instead of the other guy."

The day after Gates took office, he asked Petraeus to come talk to him about Iraq. "I"m going to Iraq in two hours," the new defense secretary said. "What should I look for?"

Petraeus had been thinking for months about the war and about how to change it. He had been talking to staff members on the National Security Council, among others, and also had "reviewed very carefully the AEI findings" from the exercise that Kagan and Keane had done. So when Gates asked, he was ready. Both men understood that the context of the conversation was that Petraeus was being considered for top command in Iraq, and at a crucial time. "You should focus on whether or not the approach is working," Petraeus said. "Get a sense of whether the emphasis on transition to ISF [Iraqi Security Forces] is working or not," and also on whether that emphasis is "appropriate, given the level of violence."

Another thing that emerged in the conversation was that the two men shared a preference for candid, even blunt a.s.sessments that would lead to strategic clarity. Both understood the need for more precision about the U.S. mission in Iraq.

Gates had another question. Zalmay Khalilzad, then the U.S. amba.s.sador in Iraq, had concurred with an intelligence a.s.sessment that the U.S. effort was failing. What, Gates asked, did Petraeus think of that? I haven"t been in Iraq for a while, Petraeus responded, but from what I can tell, it is correct.

Then Gates left for Iraq, bringing with him Gen. Pace. In Baghdad he met with Abizaid, Casey, and Odierno. The first two generals were at loggerheads with Odierno, the newer, younger, and junior officer pushing hard for more troops. Gates was soft-voiced and guarded. He listened but gave no indication of which way was he was leaning. Anderson, Odierno"s chief of staff, recalled that Pace offered the whiff of a possibility of meeting Odierno"s request. "What if we gave you more?" he asked in a meeting. "What if we gave you five brigade combat teams?" He seemed to be asking if they really had anything to show: What makes you guys smarter than Casey, Abizaid, and the Joint Chiefs of Staff?

Gates, to his credit, also breakfasted with a group of young soldiers and found far more agreement there on the need for more manpower. "Never mind all the generals standing around," he began, according to a tape recording of the meeting, which reporters didn"t attend. While Odierno and Pace listened, Spec. Jason Glenn, a drone aircraft operator for the 1st Infantry Division, told him that when he flew his aircraft over insurgents, they would just look up at it, "so I really think we need more troops here, with more presence on the ground. More troops might hold them off long enough to where we can get the Iraqi army trained up."

Pfc. Ca.s.sandra Wallace seconded that view. "I think we do need more troops over here," the Texan said. "More troops would help us integrate the Iraqi army into patrols more."

On the long flight home to Washington in a C-17 military cargo jet, Gates disappeared into his mobile home in the jet"s belly with Gen. Pace and a bottle of California cabernet sauvignon. A few days later, Odierno got the word: Gates wants you to have all five brigades. "The surge really began the day that Gates visited," Odierno later concluded.

But the issue was still in the air. While Gates and Pace were traveling, Bush gave an interview to the Washington Post Washington Post in which he began by emphasizing his intention to increase the size of the Army and Marine Corps. "The reason why is, it is an accurate reflection that this ideological war we"re in is going to last for a while, and that we"re going to need a military that"s capable of being able to sustain our efforts," Bush explained. It was an odd statement to make, coming more than five years after the 9/11 attack. Yet it hinted at the major strategic shift the president was contemplating: dropping the pattern of overoptimistic a.s.sumptions and instead moving to a strategy for long wars in Iraq and Afghanistan. in which he began by emphasizing his intention to increase the size of the Army and Marine Corps. "The reason why is, it is an accurate reflection that this ideological war we"re in is going to last for a while, and that we"re going to need a military that"s capable of being able to sustain our efforts," Bush explained. It was an odd statement to make, coming more than five years after the 9/11 attack. Yet it hinted at the major strategic shift the president was contemplating: dropping the pattern of overoptimistic a.s.sumptions and instead moving to a strategy for long wars in Iraq and Afghanistan.

When Pace got back to the Pentagon, he sent down word to a.n.a.lyze how to get two to five additional brigades into Iraq, in terms of both transportation and troop availability. Barbero, on the Joint Staff, received the a.s.signment. He liked the idea. "We"re finally going to try to do something more than just hold on until the Iraqis can take over," he thought.

Lt. Col. Charlie Miller, then working under Barbero, was surprised. Two weeks earlier, he recalled, neither the Pentagon nor the U.S. headquarters in Iraq had been particularly interested in a surge. In fact, when Casey"s command had been asked, he said, its response had been indifferent: "Sure, you can send an extra brigade or two, if you want." The new order was "interesting," he said, because the impetus clearly "didn"t come from inside the building. And it didn"t come from Iraq."

Even as Odierno"s planners were fleshing out the plan in late December, they secretly thought there was no way they were going to get five brigade combat teams, or BCTs. There was good reason to doubt it. Pace, who as the top U.S. military officer was the president"s direct adviser of military affairs, still wasn"t signed on to a full surge. Just after Christmas he flew to Texas to see the president at his ranch in Crawford and proposed the compromise of sending two brigades, plus keeping three more on tap ready to go.

Keane had gotten wind of this odd "minisurge." He also heard that Pace was telling officers on the Joint Staff who thought that two brigades was insufficient, "Don"t tell me what is wrong with this plan, tell me how to sell it at Crawford." Keane called John Hannah, who had replaced the disgraced Scooter Libby as Cheney"s national security adviser, and told him that "the force level you are going to see presented at Crawford is inadequate and destined to fail." He asked Hannah to give this question to Cheney or Bush to pose to Pace: "Is this a decisive force to succeed?" Privately, Keane thought that Odierno and Pace really needed eight to ten brigades, but he knew that only five could be made available.

After the Crawford meeting, Keane got a call from a White House official: You"ll get the five Army brigades, plus two Marine battalions. It would amount to about 20,000 additional combat soldiers and eventually would include another 8,000 support troops.

Pace returned to the Pentagon with a new urgency. Rather than the cool response he heard weeks earlier in discussions of the surge, Miller, the Joint Staff officer, recalled, it was, "Hey, we need to do this today." today."

The planners in Baghdad suddenly found themselves. .h.i.t by an avalanche of RFIs-requests for information. Officials at the White House, the office of the Secretary of Defense, and the staff of the Joint Chiefs were bypa.s.sing several echelons in the chain of command to call Odierno"s headquarters directly and ask colonels there a question they now considered crucial: What would you do if you had more troops? Frequently those officers in Baghdad would zip back a quick e-mail-and only later realized how much influence they were having. "We would get a ping," said Maj. Kent Strader, "we"d respond, and twenty-four to forty-eight hours later, it would show up in the national debate."

They also were getting questions for which they had no answers. "They would ask, you"ll get one brigade a month, where will it go?" said Maj. Powell, one of Odierno"s planners. "We didn"t know. It would depend on the evolving situation on the ground."

WWSHD?.

Once it was decided that the surge would have five brigades, the question was what to do with them. The plan devised at the American Enterprise Inst.i.tute under Kagan"s purview had been simply to flood Baghdad with troops. "The Kagan plan had four brigades going into the city, two or three into Anbar, and none into the "belts,"" or the areas surrounding Baghdad, said Powell, the planner. The plan as ultimately implemented would be more nuanced. The seed had been planted with Odierno by his next-door neighbor back at Fort Hood, Maj. Gen. James Thurman, who had commanded the U.S. troops in Baghdad in 2005-6. In a series of conversations, Thurman had pa.s.sed along two lessons: Don"t give up any ground and do something about the car bombs coming into the capital from the surrounding countryside.

Odierno laid down several new principles to his planners: This wouldn"t be just Baghdad. He told the planners to figure out how to cut the roads, dirt paths, and riverways the bombers used to move into the city, what the military calls the "lines of communication." His decision was influenced not just by Thurman, but also by heeding the revelations of generals who worked for Saddam Hussein. American a.n.a.lysts, studying the deployment of Republican Guard troops in 2002-3 outside the capital to the areas west and south of it, had a.s.sumed that this was done to reduce the ability of commanders to launch a coup. No, they were told: The elite troops were kept there, rather than in Baghdad, because that was where the trouble was. "I looked back: What would Saddam Hussein do?" Odierno said. "He would use Republican Guards to control those areas." So Odierno decided that, as much as possible, he would deal with the Baghdad belts as much as the city itself. As every one of the five surge brigades arrived, he would ask himself whether to use it in the city or in the belts. The first two went into the capital. The next three went mainly into areas around the city. Ultimately, the 21 battalions (five brigades plus some Marine units) that came in as part of the surge would be divided about evenly between Baghdad and its outskirts. "I looked back: What would Saddam Hussein do?" Odierno said. "He would use Republican Guards to control those areas." So Odierno decided that, as much as possible, he would deal with the Baghdad belts as much as the city itself. As every one of the five surge brigades arrived, he would ask himself whether to use it in the city or in the belts. The first two went into the capital. The next three went mainly into areas around the city. Ultimately, the 21 battalions (five brigades plus some Marine units) that came in as part of the surge would be divided about evenly between Baghdad and its outskirts. Don"t make a move unless your presence is sustainable, and once you take an area, don"t leave it uncovered. "Don"t give up terrain," he ordered his commanders. "Don"t try to do too much." This emphasis on tactical patience was consistent with Petraeus"s counterinsurgency manual, which had just been published in December but was new to the U.S. military in Iraq. "One principle I went by was that we would never give up anything we"d gained," Odierno said. "So when I wanted to move a unit [out of an area], I"d ask, Can the Iraqis hold it?" If the answer was probably not, the unit would stay put. This also meant that the counteroffensive would take much longer than the American Enterprise Inst.i.tute planners had thought or Bush administration officials believed. In order to really protect the population, Odierno would have to go after more than just al Qaeda, because Shiite militias were intimidating Sunni civilians. Donnelly, a member of the AEI group, applauded the changes to the Kagan approach. "I don"t think we thought beyond making the Baghdad security situation better and establishing a rationale for keeping the United States in the war."

In December, Odierno also went to visit Col. MacFarland to check out what he had heard about strange doings in Ramadi. He also looked up Sheikh Sittar, who was leading the turning of the tribes in al Anbar province away from al Qaeda and toward the Americans. "I spent quite a long time speaking with him, and he told me how his mind-set changed," he recalled. There was a lot of that going around.

PETRAEUS, MOVING AROUND Washington and the Army to explain his new manual, was also quietly looking at a new way to operate in Iraq. Lt. Col. Douglas Ollivant had coauth.o.r.ed an article the previous summer for the Army"s Military Review Military Review advocating moving troops out into the population. t.i.tled "Producing Victory: Re-thinking Conventional Forces in COIN Operations," it stated as a principle: advocating moving troops out into the population. t.i.tled "Producing Victory: Re-thinking Conventional Forces in COIN Operations," it stated as a principle: For the local people to feel secure and provide intelligence, they must have 24-hour access to the counterinsurgent force. Units with control over an AOR [area of responsibility] should live in that neighborhood. . . . Having a fortress mentality simply isolates the counterinsurgent from the fight.

On January 9, Petraeus e-mailed Ollivant to ask two questions: "Do you still believe what you wrote in "Producing Victory"? And can MND-B [the American headquarters for Baghdad] do it?"

Ollivant wrote back: "Yes, with caveats." He was in a position to know, having been for the previous 45 days the chief planner for the American division in Baghdad. Now he had top-level backing for the idea of putting outposts in the city.

In January 2007, Keane flew out to Baghdad and met with Odierno"s planners for three hours. They were joined by some intelligence officers and H. R. McMaster. The purpose was to persuade Keane that it was crucial for the surge plan to deal with Baghdad belts, where they told him a fierce fight continued between Shiite expansionists and al Qaeda, which was casting itself as the defender of local Sunnis. They succeeded. At the end of the session there was a pause, recalled Wilson, Odierno"s chief planner. Then Keane spoke. "He kind of got a glowing look in his face and said, "You know, you folks get it."" Keane"s six-month fight to mount a counteroffensive was over. After this point, his role in the conduct of the war became smaller. As he put it, "They were in the execution phase." His job was done.

On the evening of January 10, Odierno"s chief planner, Col. Wilson, went to his room in Camp Victory"s military version of a trailer park and set his alarm for 5 A.M. As it rang, he rolled over in his bunk and reached out to switch on his small television. Half asleep, he watched President Bush tell the world about what Wilson and his office had written.

BUSH RISES TO THE OCCASION.

As Col. Wilson watched in his aluminum hootch, George W. Bush delivered what may be remembered as the most impressive speech of his presidency. There was no dodging, no divisiveness, none of the brusque claims of extraordinary success that Vice President Cheney tended to fling about. Rather, as he stood awkwardly alone at a lectern in the White House library, Bush somberly accepted responsibility for a badly handled war, confessed that the course he had pursued hadn"t worked, and laid out a clear new plan, differentiating it from past efforts.

He had looked, finally, into the abyss. "The situation in Iraq is unacceptable to the American people, and it is unacceptable to me," he said early on. "Our troops in Iraq have fought bravely. They have done everything we have asked them to do. Where mistakes have been made, the responsibility rests with me." This was exactly the right way to begin, setting the tone for everything that would follow. "It is clear that we need to change our strategy in Iraq," he continued, logically. He laid out with some precision how he envisioned it working, with a new Iraqi command structure for Baghdad and new Iraqi troops injected to conduct patrols and operate checkpoints.

Then he got to the news: To make this work, he said, he would send "more than twenty thousand additional troops to Iraq." It was only at this point that Col. Wilson and Odierno"s planners really began to believe that they would get all five brigades. In fact, they would get more, as a helicopter brigade and support troops were added three months later, eventually pushing the total addition to nearly 30,000. The mission of those new forces, Bush continued, would be "to help Iraqis clear and secure neighborhoods, to help them protect the local population, and to help ensure that the Iraqi forces left behind are capable of providing the security that Baghdad needs." In other words, after more than four years of fighting, the "kill and capture" mind-set that had led to Haditha and to a losing effort was being replaced with cla.s.sic counterinsurgency theory: The people are the prize.

Immediately, he addressed those who doubted his plan-the majority of his audience, according to polls taken after the speech. "Many listening tonight will ask why this effort will succeed when previous operations to secure Baghdad did not," he said. Using the word "listening" was a good touch, evoking the radio days of President Franklin Roosevelt"s World War II "fireside chats," as when he confessed during the dismal days of February 1942, "We Americans have been compelled to yield ground, but we will regain it." The difference, unfortunately, was that FDR faced the hard facts three months into World War II, while it took Bush more than three years to reach a similar point in his strategic thinking.

Bush then laid out what he thought the difference was between the new effort and the failed Together Forward operations of 2006: "In earlier operations, Iraqi and American forces cleared many neighborhoods of terrorists and insurgents, but when our forces moved on to other targets, the killers returned." Also, he said, the Iraqi government had blocked some operations-a reference to putting Sadr City generally off limits, along with some other key targets across Baghdad. This time, Bush said, "Prime Minister Maliki has pledged that political or sectarian influence will not be tolerated."

He then turned to the political goals he envisioned: pa.s.sing a law to share oil revenues, stepping up Iraqi spending on reconstruction, and holding provincial elections later in 2007. The a.s.sumption here was that improved security would lead to such political breakthroughs. This would prove to be the weakest part of a strong speech.