In the late summer of 2006, Brian Glyn Williams, an assistant professor of Islamic history at the University of Massachusetts, was sitting on the back porch of his Davis Square home in Somerville when a contractor for the C.I.A. telephoned. The caller said that analysts at the agency’s Counterterrorism Center were seeking insights from scholars about a wave of suicide bombings remaking the battlefields in Iraq and Afghanistan. Williams agreed to talk further. The C.I.A.’s analysts, he soon learned, were puzzling over elemental questions. Why had the targets and use of suicide attacks changed so suddenly? Who were the recruiters, trainers, and financiers? What were the ages and backgrounds of those persuaded to kill themselves? Was there evidence of “diffusion” or “contagion” of suicide bombing from Iraq to Afghanistan? The agency wanted a large, multicountry study and hoped independent academics would participate.
Williams was intrigued. He had spent much of his earlier years in Florida and then made his way into Middle Eastern studies. He spent a semester in the Soviet Union during the 1980s and met soldiers returning from the war in Afghanistan; they told him that the war was “nothing but horror.” Later he published histories of Chechnya and Central Asia. Since 2001, he had been drawn into analysis of the Chechen war and the upheavals in Afghanistan. He was forty years old but looked younger. He smiled often, displaying perfect white teeth that reflected upon the profession of his Turkish-born wife, a dental hygienist.
She was unenthused about the project. “Can’t they send someone else?”
He agreed that field research for the C.I.A. in Afghanistan might be a little “beyond my pay grade.” But he found suicide bombing to be repugnant. If he could help reduce it by studying it, he should.1
There could be no doubt that autumn that suicide bombers had become a newly destabilizing feature of Afghanistan’s reviving war. Their increase also represented a departure from Afghanistan’s recent military history. During the 1980s, across a long, bitter uprising against the Soviet Union, Afghan mujaheddin had not participated in suicide attacks. The mujaheddin’s prideful, family-supported ethos of jihad emphasized individual bravery and, where possible, living to fight another day. When the Taliban conquered Afghanistan during the late 1990s, Mullah Mohammad Omar and his commanders did not employ suicide bombers, either. (The Arab volunteers who blew themselves up while assassinating Ahmad Shah Massoud belonged to Al Qaeda.) During the three years from December 2001 through December 2004, there had been eight suicide bombings in Afghanistan, all in Kabul. Yet by 2006 suicide bombers struck two or three times a week around the country, shattering public confidence and forcing Afghan and N.A.T.O. forces to hunker down behind walls and checkpoints manned by nervous pickets.
The Taliban announced their new plan in a policy editorial published in Al Samood, in January 2006:
Let the Americans and their allies know that even though we lack equipment, our faith has been unshakable. And with the help of Allah the Almighty, we have created a weapon which you will not be able to face or escape, i.e. martyrdom operations. We will follow you everywhere and we will detonate everything in your face. We will make you terrified, even from vacant lands and silent walls. We know we are inevitably heading towards death, so let it be a glorious death by killing you with us, as we believe in the words of the Prophet (Peace Be Upon Him): “The heretic and his killer will be united in the fires of hell.” We have thus prepared many suicide operations that even will involve women, and we will offer you the taste of perdition in the cities, villages, valleys and mountains with Allah’s help.2
The most visible spokesman of the Afghan Taliban’s initiatives was Dadullah Lang, a one-legged Taliban military commander who bragged about suicide operations on Al Jazeera. He called his young recruits “Mullah Omar’s missiles” and “our atomic bombs.” He told another interviewer in the summer of 2006, “We like the Al Qaeda organization. We have close ties and constant contacts.”3
The C.I.A. struggled to provide hard evidence to the national security cabinet about why this had evolved so suddenly. The agency offered to buy out Williams’s salary at UMass for a semester and to pay an additional $30,000 for his fieldwork and a research paper. He spent the autumn and winter working with a student researcher to build a matrix of suicide attacks in Afghanistan since 2001, to identify patterns and questions that he could examine when he traveled to the country. On his Excel spreadsheet, he recorded dates of suicide bombings, their locations, what category of target had been struck, the number of casualties, and details about the strikes, if they were available from media or other accounts.
As his grid of evidence morphed into color-coded patterns, two mysteries presented themselves initially. Williams realized that he was recording a surprising number of “zeros” in the casualty column, meaning that the suicide attack had failed altogether. There were about a dozen such cases in 2006, for example, about 9 percent of the total. Moreover, the most common outcome of an Afghan suicide attack that year was a single casualty—the bomber himself and no one else.
This was not what Williams had seen when he had previously studied suicide bombings in Iraq. There, failures were abnormal and high death tolls were common. This led to a second mystery about the Afghan pattern, concerning the bombers’ targets. In Iraq, suicide bombers typically struck crowds of civilians to sow terror by inflicting mass casualties, including women and children. The context for these attacks was often sectarian: Al Qaeda–influenced Sunni suicide bombers struck Shiite civilian marketplaces or mosques, and vice versa. In Afghanistan, however, suicide bombers most often struck military targets, such as heavily armed American or N.A.T.O. convoys moving on roadways. Remarkably often, only the bomber died. Williams and his researcher marveled morbidly about how incompetent some of these suicide bombers seemed to be. One had strapped on his vest, traveled to say goodbye to his parents, and accidentally detonated his device during the visit, taking his own life and theirs. But when Williams reflected on it, the pattern seemed tragic. Presumably such failures indicated how many suicide bombers recruited to die in Afghanistan might be coerced, naïve, illiterate, young, or disabled.4
The C.I.A.’s contractor and an analyst at the agency’s Counterterrorism Center arranged for Williams to travel to Afghanistan. He departed in the early spring of 2007. Williams connected with Hekmat Karzai, a cousin of the Afghan president who ran a Kabul think tank. He provided Williams with a base of operations. As to the risks he would take traveling to provinces to meet local police, investigate bombing case files, and speak with affected Afghans, he was largely on his own. His C.I.A. supervisors told him, “If you get caught, we don’t know you.” He wasn’t sure how serious they were, but their instructions were clear: “We appreciate your service, but don’t call us.”5
—
The C.I.A.’s suicide bombing study coincided with a renewed stirring of interest in the Afghan war at the White House. The Afghan Inter-Agency Operations Group at the National Security Council remained the main vehicle for policy and budget decisions. The group was bureaucratically weak and often ignored. John Gastright of the State Department now cochaired the effort. His modest rank—deputy assistant secretary of state—signaled his operation’s low standing in the Bush-Cheney-Rumsfeld hierarchy, an apparatus focused almost entirely on the worsening war in Iraq. Yet at least Gastright’s group was organized across government to think about what American policy in Afghanistan required, a change from several years before.
“We ought to take a look at where we are and what we’re doing,” Gastright announced one afternoon in mid-2006 at the State Department, where the Operations Group met. In Kabul, Ronald Neumann had succeeded Zalmay Khalilzad as U.S. ambassador. Gastright would typically arrive at Foggy Bottom around 6:30 a.m. and the first thing he did was call Neumann. They agreed that the levels of aid to the country were grossly inadequate and that the U.S. military needed to mount a bigger effort to prevent the war’s further deterioration. They also knew that to persuade President Bush and the Office of Management and Budget to change course and spend more in the midst of the Treasury-draining fiasco in Iraq would require a formal policy review. Meghan O’Sullivan, the senior National Security Council adviser who worked mainly on Iraq but also kept track of Afghanistan, brought the idea for a review to Bush, who approved it.6
Neumann had tried with little success to persuade the State Department and the White House to radically increase investments in Afghanistan. He regarded the plan for a new policy review with skepticism. “Searching for a new strategy seems to be policymakers’ recurring default reaction to problems,” as he put it. Instead of properly resourcing the strategy the administration already possessed, it authorized a search “for a new idea. Certainly ideas are more easily come by than money and soldiers.”7
In the autumn of 2006, Gastright and his cochair, Tony Harriman, the N.S.C. senior director for Afghanistan, commissioned classified study papers on every major policy subject that seemed relevant: the Afghan justice system, the police, narcotics, the Afghan National Army, and civilian aid. The final drafts of these papers contained recommendations. In essence, they identified the need for more resources, more spending. Harriman liked to point out to his colleagues that there were more combat aircraft controlled by the State Department operating in Colombia, to support the fight against coca growing and Marxist guerrillas there, than there were combat aircraft controlled by the Pentagon in Afghanistan.
“We cannot win in Afghanistan on the cheap,” Neumann wrote to Washington. Michael Waltz, the Special Forces officer, rotated from Afghanistan to a policy position at the Pentagon late that year. He was “pleased to find some growing awareness . . . of the worsening situation” in Afghanistan, yet many of the Afghan hands he encountered were “increasingly frustrated. Even though our senior leaders recognized the growing problem, the default response was to turn to the Europeans to do more.”8
The American electorate soon came to the aid of their cause. In early November 2006, in mid-term elections, voters repudiated President Bush and handed control of Congress to the Democrats. Bush accepted Rumsfeld’s resignation in the aftermath and appointed Robert Gates as secretary of defense. Gates had been at the C.I.A. and the White House when the United States abandoned Afghanistan in the early 1990s. He was determined not to preside over another failure in South Asia, yet he was skeptical that the United States could fix the war. His “historical perspective . . . screamed for caution” yet American generals in the field insisted they needed more forces and resources.
On November 13, 2006, the N.S.C. forwarded to President Bush a ten-page classified executive summary of “The Afghanistan Strategic Review.” Bush approved its findings on December 10. The plan called for more money, a reenergized effort to improve security and governance, a new push to undermine the booming drug economy, and more American troops on the ground. The hypothesis of the paper promoted the need to “connect the Afghan people to their government.” It proposed more schools, roads, and electricity. Of these recommendations, only the call for more American troops failed to materialize quickly. Early in 2007, Bush’s decision to authorize a “surge” of troops into Baghdad under the command of General David Petraeus effectively used up all of the Pentagon’s available personnel and then some. With about 150,000 American troops now in Iraq, there were no additional units available for Afghanistan. The Afghan review did reverse the administration’s parsimony about reconstruction aid, however. Now the O.M.B. approved a supplemental allocation to Afghanistan of more than $1 billion, although to Neumann’s disappointment, the White House split the funding across two fiscal years, diluting its impact. The ambassador “got pretty cranky,” as Gastright put it, and cabled in protest, but the decision stood.9
The Bush administration’s total spending on security in Afghanistan would nonetheless be greater during 2007 than during all previous years combined. The allocations included major new classified investments at Amrullah Saleh’s N.D.S., for the Afghan National Army, and for the police. Paul Miller, the C.I.A. analyst who moved to the White House that year to work on Afghanistan, estimated that total security assistance, classified and unclassified, ballooned toward almost $8 billion a year, at least four times the levels of the “Mr. Big” era when Khalilzad was ambassador in Kabul.10
In February 2007, Bush appeared at the American Enterprise Institute in Washington to speak about the global war on terror. He gave over much of his speech to Iraq, but about halfway through his talk, he disclosed that his administration had just finished “a top-to-bottom review of our strategy” in Afghanistan. Bush laid out renewed aims: “To help the people of that country to defeat the terrorists and establish a stable, moderate, and democratic state that respects the rights of its citizens, governs its territory effectively, and is a reliable ally in this war against extremists and terrorists.” He admitted, “Oh, for some that may seem like an impossible task. But it’s not impossible.”11
In fact, the war on the ground was deteriorating by the month. Its challenges had at last attracted the White House’s attention. Yet the Bush administration’s new strategy remained informed by undue optimism, not least because Afghanistan still looked much better than Iraq. Bush was defensive about the comparison. He told the Joint Chiefs, “Many in Congress don’t understand the military. ‘Afghanistan is good. Iraq is bad.’ Bullshit.”12
—
In the parking lot of the Kabul International Airport, Brian Williams met Humayun, the tall driver-cum-bodyguard with whom he would work for the next two months. He was from Kandahar. He mentioned that he had not told his family about his work with the Karzai regime in Kabul, as this would only attract the Taliban’s ire. He kept a pistol in his car. He told Williams that he appreciated what the United States was trying to do in Afghanistan. “The day you leave, the Taliban will be back,” he predicted.13
Williams settled into a room in Hekmat Karzai’s walled compound. Each morning he descended to the common area for tea, naan, and cheese, and to check for news of the latest suicide bombings. If possible, he and Humayun would drive out to crime scenes. Sometimes they arrived when there was still blood on the ground.
Five days after he arrived, Williams drove to meet U.S. Army officers at Bagram Airfield. He was stunned by the self-imposed isolation of the American soldiers and military intelligence officers he met. By 2006 Bagram had acquired some of the amenities common on other American military bases worldwide. There was a bowling alley, a Burger King, and an Orange Julius, Williams discovered. He learned nothing from his meetings except that the American officers knew nothing about why the rise of suicide warfare had occurred or where it was heading. The officers he met seemed to consider “everything beyond their barrier to be a red zone.”14
At N.D.S. headquarters, Williams met Amrullah Saleh’s staff and several of his senior officers. They were better informed. They had a theory of how suicide bombings had accelerated. Their insights were derived mainly from arrests and interrogations. The Iraq war was one factor, they told Williams. Arab technicians fashioning suicide vests and vehicle bombs in Iraq were highly sophisticated, in comparison with the typically illiterate Pashtun commanders of the Taliban. The international jihadists were trying to export their suicide bomb technology to Afghanistan, but with mixed results. In Nimruz, a large province in the southwest, N.D.S. had arrested several Arabs crossing from Iran who were transporting prefabricated suicide belts for the Taliban. N.D.S. believed the majority of bombers striking inside Afghanistan were Pashtuns from the Federally Administered Tribal Areas of Pakistan.
Security officers at the United Nations, who were responsible for the safety of development and political officers scattered around Afghanistan, handed Williams a PowerPoint deck showing about fifty photographs of the severed heads of deceased bombers. There could have been a touch of dubious phrenology in the United Nations analysis of the pictures of the heads—a confidence about what the contours of the faces showed about the bomber’s ethnic or national origins that would not pass as science. But the photos did make clear that the bombers were all very young. To Williams, they appeared to be Pashtuns, not Arabs.
Williams slowly developed a composite picture of a typical suicide bomber in Afghanistan. The Taliban bombers were often young, as young as twelve or thirteen. They typically had little experience at driving, never mind at speed racing along roads in a battered, bomb-rigged Toyota Corolla. Under pressure, in the last seconds of their immature lives, they failed. Police officers summarizing interrogations of detainees in N.D.S. custody explained that the Taliban paid the families of suicide bombers in the range of two thousand dollars to ten thousand dollars, a small fortune in Waziristan or the rural south and east of Afghanistan. The boys were recruited from madrassas where they had been enrolled for years in a curriculum of suffocating political-religious instruction.15
—
In Pakistan, during the first half of 2007, separately from Williams’s inquiries, Pakistan Army officers, psychologists, and social scientists confronted a parallel rise in suicide bombings against Pakistani targets. These attacks, too, emanated from Waziristan. Al Qaeda, Afghan Taliban, Pakistani Taliban, and Punjabi radicals had turned against the Pakistani state and I.S.I. itself on the grounds that it had betrayed Islam through alliance with the United States.
Abu Bakr Amin Bajwa, a Pakistani brigadier posted to Waziristan, stumbled into an abandoned Taliban suicide bomber school while on patrol. In Razmak, a district of North Waziristan near the Afghan border where Arab Al Qaeda units had embedded, he visited a jannat, or paradise facility for suicide bombers in the final stages of preparations. The recruits studied in concrete rooms whose walls were painted with murals of the afterlife, “channels of milk and honey, fruit trees, green mountains, street lights, and animals like camels and horses.” Another room depicted paintings of virgin girls. One of the young women was shown filling a water jug at a pond. The faces of the virgins and the animals had been crudely brushed out in keeping with Taliban proscriptions against living figures in art. On the walls of the final room, Bajwa saw the names of suicide bombers who had fulfilled their missions, “written in blood.” Across the way he found a water tank “where the Taliban used to execute or torture the people who would oppose them or spy on them.” Family finances, enticements of paradise, and torture created an environment that no thirteen-year-old boy was likely to withstand independently.16
Bajwa discovered further that the boys were injected daily with the antianxiety medicines Valium and Xanax and weekly with Penzocine, an analgesic. Trainers preferred boys who were either seriously ill, “mentally challenged,” or else had a reason for revenge, such as a family member killed in military operations. Bombers typically operated in pairs, monitored by their trainers. Those known to be suffering from mental illness or paralysis could be remotely detonated.17
Shazadi Beg, a British barrister who worked on programs to rehabilitate such boys in Pakistan after they were captured by authorities, encountered a fourteen-year-old in a Pakistani jail. He had been arrested after a failed suicide attack. The boy was shackled by the ankles and handcuffed. His eyes darted around the room. Beg asked what he was looking at. “I’m trying to see if there is something I can kill you with,” he answered.
The boy had been recruited from his family in South Waziristan but said he did not want to see them again. He had been arrested at an army checkpoint, wearing a suicide jacket, just two hours from his home; it was the first time he had ever left his village. “I am in dishonor. I failed in what I was supposed to do.” He said that if he was released he would complete his mission. His eyes brimmed with tears when Beg asked if he missed his mother, but he tried very hard not to show emotion and said that his mission was to kill unbelievers. He refused to watch television because he believed it was the devil’s instrument and designed to deceive him. He had memorized the Koran yet could not translate the meaning of any verses from Arabic to Urdu or Pashto.18
Psychiatrists who counseled former suicide bomber recruits in Pakistan found that the kids were bullied, assigned menial tasks, and beaten if they did not accomplish them. They were provided marijuana as a reward. They adapted. They had no alternative.19
—
The images of afterlife at the bomber school in Razmak—horses and virgins filling water pots from a pond—suggested how Taliban commanders had refined Al Qaeda’s ideology to communicate with rural Pashtun communities. To Brian Williams this helped to explain why so many Afghan Taliban suicide operations targeted armored military convoys or armed guards at the perimeter of walled security bases, rather than marketplaces, mosques, or other soft targets, where they might kill many more people and sow terror, even if their victims were civilians. During 2006 and 2007, Taliban suicide bombers did attack civilians in settings such as dog fights where gambling took place, but they did not often follow the Iraqi or Pakistani sectarian pattern of killing large numbers of civilians deliberately. To maintain legitimacy among Pashtun families sacrificing sons, the suicide operations had to be honorable. It was acceptable to die in an attempted military attack on a U.N. convoy, even if the attack was difficult. It was not as honorable to blow up pious Afghan Muslims in a bazaar. This pattern would dissolve in the years to come, particularly as I.S.I. and the Haqqani network sought to destabilize Kabul, and as sectarian feeling hardened in the intensifying war, but at the beginning of the suicide bombing wave, Williams’s research showed, the recruits and their families seemed to require a traditional military purpose.
As Williams completed his research, the number of Afghan suicide bombers who changed their minds at the last minute also impressed him. Afghan police showed him suicide vests that boys had torn off, dropped, and run away from. By now Williams empathized with the bombers and felt he understood their ambivalence. They really believed in the righteousness of their struggle. The attacks of September 11 meant nothing to them or their families; they did not even have televisions. The families that accepted payment for their sacrifice and glorified their martyrdom “really believe in the Taliban and believe in the war and the goodness of it,” Williams reflected.20
Yet suicide bombing should not be understood as an indigenous aspect of the Taliban’s revival, he concluded. In Gardez, he found D.V.D.s stacked for sale in a market. They were designed to inspire suicide bombers. The programs were Iraqi productions, originally produced in Arabic but dubbed in Pashto for the Afghan market. They presented calls to martyrdom amid naheeds, or Islamic vocal works. This was the clearest evidence Williams had yet encountered of “the Iraq effect.”
In fact, the number of suicide bombings in Afghanistan declined in the years after Williams’s study, while the number of land mine or improvised explosive device attacks increased more than sixfold. As N.A.T.O. and the U.N. imported armored vehicles and took greater precautions, powerful land mines were a more effective tactical counter for the Taliban. Suicide bombings constituted about 4 percent of all Taliban bombings in 2007; three years later, they constituted less than 1 percent, although the number of assassinations and mass casualty attacks against civilians increased.
Williams felt more convinced than ever before “of the rightness of the U.S. going into Afghanistan, that the U.S. needed to be there to fight off the Taliban,” as he put it. Yet his support was tempered by his shock at how badly the war was going. Back home, he wrote up classified and unclassified versions of his findings. (Because he did not yet have a top secret security clearance, once he turned in the classified paper, he wasn’t allowed to look at it again.) Williams presented his findings to analysts at the C.I.A.’s Counterterrorism Center. The contractor who had recruited him to do the study also asked him to a conference at a Virginia hotel, to present his work to about eighty C.I.A., military, and other intelligence analysts.
“I don’t think this is organic” to Afghanistan, Williams told them. The C.I.A. analysts he encountered accepted that, but those from the Defense Intelligence Agency challenged him. The back-and-forth went on for ten minutes. How do you know suicide bombing did not evolve intrinsically from Afghan culture? the more skeptical analysts asked.
Well, Williams said, all he could say was that the pattern of attacks he had documented showed that Taliban suicide squads were made up of poorly qualified, often coerced youngsters. Many in his audience “didn’t like” his overall conclusion, which was: “The Iraq war had destabilized the Afghan war.”21