by Steve Coll
Obama told his advisers, “I think I have to do this,” because the war was deteriorating so badly that if he didn’t “do something to at least create some space, this whole thing could collapse.” The troop surge was a holding mission necessary to allow the counterterrorism mission—drone strikes, primarily—to decimate Al Qaeda and to give Afghan security forces a chance to succeed. The Pentagon’s commanders had bigger ideas. They had been trained at West Point and elsewhere to win large conventional wars, not manage expeditionary counterterrorism missions rife with contradictions and subtleties. They also knew that fighting wars successfully required confidence, momentum, sacrifice, and ambition—a rallying cause. They chafed at constraints.
The next day, Obama spoke to Karzai by secure video. Karzai’s inauguration to a second term as president had lately reduced his state of aggravation. He had won power for five more years. As a practical matter, he needed the United States to survive, however distasteful and irritating that dependency might be to him. He now told Obama he was ready to publicly support the American troop surge, or, at least, this was how Obama’s aides understood him. He repeated once more, however, according to a contemporary record, that the “military and political dimensions of achieving peace in Afghanistan can’t be addressed unless the issue of sanctuary in Pakistan is made explicit and is a priority in the new strategy.” (Years later, Karzai would ascribe a different emphasis to his views about the Obama troop increase. His goal, he recalled, was to reduce the American military “footprint,” garrison the American troops on bases, and ensure that there were “no troops in our villages or towns.” In any event, he was out of sync with Washington; the counterinsurgency doctrine Obama had partially endorsed turned on pushing troops into villages, towns, and urban neighborhoods.)31
McChrystal flew that day to Rawalpindi to brief Kayani on Obama’s decisions. The Pakistani general said once more that the United States “lacked the time to accomplish all that was necessary.” Kayani also said again that he did not think the United States could build up effective Afghan security forces to carry on the fight against the Taliban after N.A.T.O. departed. McChrystal left thinking that “our chances were better than he believed.”32
On December 1, on prime-time television, Obama delivered a solemn speech at West Point. He announced, “As Commander-in-Chief, I have determined that it is in our vital interest to send an additional 30,000 U.S. troops to Afghanistan.” In the very next sentence, without any prologue or clarification, he added, “After 18 months, our troops will begin to come home.”
He spent the next several minutes emphasizing how agonizing his choices had been. Obama’s going-in-while-going-out decision reflected his temperamental search for middle ground and consensus, his reading of domestic politics, and his painstaking study of the Pentagon’s claims and plans for the war, about which he retained skepticism.
“He made an amazing decision,” Holbrooke explained to a visitor soon afterward. “To send thirty thousand troops and announce the withdrawals would start in eighteen months—no one’s ever done something like that.” In Holbrooke’s judgment, “The Taliban was not the audience. The primary audience was the American public and the Congress. The secondary audience was the Afghans, to tell them they had to really focus on Afghan National Security Force training. The more complicated audience was the rest of the world, starting with the Pakistanis and the Iranians, who were of course thoroughly befuddled. . . . It was, all things considered, the best course to take . . . a very interesting idea. So far as I can tell it came directly from the president.”33
In Kabul, on the morning after the West Point address, in the classified command center of I.S.A.F., McChrystal spoke to his high command. He sat in a secure briefing room filled with screens showing the faces of regional commanders around Afghanistan. Other generals and colonels sat with him around tables set as a hollow square.
“I think we’re at an inflection point,” McChrystal declared. “We have a new clarity in our mission.” That mission was to “provide our Afghan partners time, space, and resources” to defeat the Taliban eventually. “The success of this operation will be determined by the minds of the Afghan people,” McChrystal added. “It’s a war to give people a chance.”34
PART FOUR
THE END OF ILLUSION,
2010–2014
TWENTY-THREE
The One-man C.I.A.
During the first week of August 2009, while on vacation in the south of France, Barnett “Barney” Rubin took a call from a Saudi lawyer he had met. His contact occasionally did odd jobs for Saudi intelligence. The Saudi said that Rubin should travel to Dubai. Rubin was a political scientist with a doctoral degree from the University of Chicago who specialized in Afghanistan as well as conflict prevention and resolution. He now worked at the State Department as a senior adviser to Richard Holbrooke. Rubin called Holbrooke and asked what he should do. By all means go, Holbrooke said. Just don’t tell anyone what you are doing.
Rubin drove to Nice and flew to Dubai, arriving in the early hours of August 9. He took a cab to a comfortable hotel. His contact received him in his rooms a few hours before dawn. Rubin, then fifty-nine, was a stout man with thinning white hair and a bushy white beard. He had an “alarming physical resemblance to Leon Trotsky,” Holbrooke once noted, which was “not a big asset” at the Pentagon or the C.I.A. Holbrooke valued him because he offered “a level of intellectual quality you almost never see in the U.S. government.” Rubin had been traveling through Saudi Arabia and the smaller Gulf States for many years, while going in and out of Afghanistan and Pakistan. The leading South Asia specialists in the U.S. intelligence community generally acknowledged that Rubin was the premier scholarly American expert on Afghan politics. He spoke Dari haltingly as well as Hindi and Urdu, and had studied literary Arabic. As an undergraduate at Yale, he had majored in history and joined the Students for a Democratic Society, a leftist youth movement involved in the civil rights movement and the opposition to the Vietnam War. During the 1980s, as an assistant professor at Yale, he had been drawn to human rights issues in Afghanistan during the Soviet occupation. This led him to write an influential book, The Fragmentation of Afghanistan, about Afghan elites and the modern Afghan state. He was a liberal with a strong interest in peacemaking, an outlook that was in the mainstream on the Upper West Side of Manhattan, where he lived.1
He had gotten to know his Saudi friend recently, while trying to probe for Holbrooke the possibilities for dialogue between the United States and the Taliban. His contact had fought in the anti-Soviet jihad of the 1980s as a volunteer, aligned at the time with Saudi and American policy. He had known Osama Bin Laden. Later he gave up war and became a professional in the kingdom. Because of his credibility with Islamists, Prince Muqrin bin Abdulaziz, the head of Saudi intelligence since 2005, had asked him to play a liaison role between Saudi Arabia and the Taliban.
In his hotel rooms, as the multicolored lights of Dubai blinked outside, the Saudi revealed that in recent days, he had received a letter and two telephone calls from well-known Taliban representatives who said they were in direct contact with Mullah Mohammad Omar. They wanted to convey privately to Saudi Arabia and the United States that Omar had appointed a new, exclusive representative for political negotiations. The man’s name was Tayeb Agha. He was visiting Jeddah, the Saudi city on the Red Sea.
“There is one channel” for talking to the Taliban, the Saudi explained to Rubin, and “it should be protected and strengthened.” He said that he had told the Taliban that the Americans were ready to stay in Afghanistan for twenty years if that’s what it took. Yet there were nonetheless possibilities for negotiation, to separate the Taliban from Al Qaeda and coax them into power sharing and Afghan politics. “The U.S. does not want to lose, and the Taliban do not want to lose,” the Saudi interlocutor observed. “Both want a way out in which they do not lose.”2
He said he was about to fly to Morocco, where Pr
ince Muqrin maintained a summer palace, to tell the spy chief about this new development. As for next steps, he said, the Saudis were wary about becoming too exposed in a peace negotiation with the Taliban. The kingdom considered itself at war with Al Qaeda, which had staged terrorist attacks inside Saudi Arabia. The royal family, he went on, wanted a letter “plus a voice cassette from Mullah Omar” making clear that the Taliban do not support terrorist activities anywhere in the world and that the Taliban have had no contact with Al Qaeda for the past seven years. That was a tall order. Yet Saudi Arabia had been damaged by allegations that its royal family officially supported Al Qaeda and other terrorists; the kingdom would not broker talks with the Taliban if there was any risk that it would be accused again of such complicity. How could this anxiety be squared with Mansour’s acknowledgment that Tayeb Agha, an official Taliban envoy, in his account, was resting comfortably in Jeddah? Officially, the royal family took the position that, as custodians of the Two Holy Mosques in Mecca and Medina, they could not refuse entry to any Muslim seeking pilgrimage. Unofficially, the policy allowed the kingdom to maintain contacts with Islamist groups, to convince them that Saudi Arabia meant them no harm.3
Rubin flew back to France. He did not know it then, but his meeting had inaugurated what would become a four-year project of secret diplomacy between the United States and the Taliban, aimed at finding a political settlement to end the Afghan war or at least reduce its violence. The project would ultimately draw in President Obama, who would invest long hours of personal effort. Not since Vietnam had the United States undertaken a comparable effort to negotiate with its enemy secretly while in the midst of war.
On August 11, Rubin wrote privately to Holbrooke that Tayeb Agha “is the chief of finance for the Taliban Leadership Shura and frequently travels to Saudi Arabia to collect funds” from individuals, businesses, and charities. The United States faced a dilemma. It seemed apparent that the same Taliban envoy who traveled outside Afghanistan and Pakistan to raise funds for the movement was also a potential channel for diplomatic contact. If they sanctioned him for fund-raising, they would lose him as a potential point of entry to Taliban leadership. A few days later, at Holbrooke’s invitation, Rubin penned a second memo formally urging that any sanctions decision be postponed while they assessed this opening for engagement with Mullah Omar.4
Holbrooke did not share Rubin’s reports widely, if at all, in Washington. If Holbrooke surfaced the news about Mullah Mohammad Omar’s appointment of Tayeb Agha as a special envoy, he risked further isolation. Rubin had been reporting to Holbrooke periodically about all sorts of putative channels to the Taliban leadership. He was confident about Tayeb Agha’s probable legitimacy, but the overall picture of who represented whom and what Mullah Omar or the rest of the Taliban’s leadership wanted in possible talks was confusing. The best course was to wait and probe.
Holbrooke warned Rubin to be discreet and patient. Proposing talks for a “political settlement” with the Taliban was much too incendiary an idea for the Obama administration to digest right now. “Remember,” Holbrooke told his adviser, “your problem is not with the Taliban. Your problem is with Denis McDonough and the C.I.A.”5
On Saturday, September 12, at a meeting at the Brookings Institution, Holbrooke disclosed that the C.I.A. had recently completed its latest update of the District Assessments. A superficial examination of the colored maps showed that more of the country was under Taliban control than under government control, Holbrooke remarked, but a deeper look, accounting for areas that were contested or controlled by neither side, presented a less dire picture of the war. There was time to turn this conflict around.
Yet it frustrated Holbrooke that so much analysis of the war looked at the Taliban from the outside in. Where was Mullah Omar? With whom was he consulting about Taliban strategy? What did diverse members of the Quetta Shura think about the Obama administration’s escalation of the war? Recent intelligence reporting suggested that Mullah Omar had threatened to kill anyone who tried to reach out for political talks with the United States without his personal authorization, Holbrooke disclosed. If this was accurate, what was Mullah Omar worried about? Were splits emerging in the leadership?6
On Monday, September 14, Holbrooke met with Steve Kappes of the C.I.A. to discuss how the agency could probe the Taliban’s political outlook more deeply. Until now, the C.I.A. had been given no formal tasking from the White House to collect intelligence on political debates within the Taliban leadership or how the movement might think about peace negotiations with Karzai, Saudi Arabia, or the Obama administration. It seemed evident from military reporting that the Taliban had suffered heavy casualties in Kandahar and Helmand after 2006 but the impact of these losses on the movement was also a mystery. The C.I.A.’s mission in Afghanistan still prioritized counterterrorism and Karzai watching. Holbrooke and Kappes agreed that the agency would now undertake “more collection on the Quetta Shura and political views” within the Taliban leadership. It was not a brand-new tasking, but Kappes wanted to be fair to Holbrooke and the C.I.A. was in business to take deeper looks at such difficult targets.7
The C.I.A. responded that autumn by circulating intelligence reports derived from secret interviews its field officers conducted with Taliban leaders, current and former. Yet even as this intelligence seeped into the American system, the possibility of direct talks with the Taliban leadership was not on the table during the long review of Afghan war strategy that Obama oversaw in the Situation Room that autumn of 2009. General Petraeus, extrapolating from his successes in Iraq turning Sunni tribesmen against Al Qaeda, advocated for trying to induce battlefield defections of Taliban commanders and soldiers—a program he called “reintegration” of enemy fighters, from the bottom up, as opposed to “reconciliation” with Taliban leaders, from the top down, which might come later. Petraeus did not actually believe Holbrooke’s vision of reconciliation was viable because there was no way for the United States to threaten the Taliban leadership, who felt safe in Pakistan. Without such coercion, why would the Taliban negotiate seriously? Drawing on Bush administration policy, the White House did seek to give space for Afghans to talk among themselves about peace. The president’s West Point speech therefore contained only a single sentence about seeking accommodation with the Taliban: “We will support efforts by the Afghan government to open the door to those Taliban who abandon violence and respect the human rights of their fellow citizens.”8
In the absence of presidential support, Holbrooke used Barney Rubin as the forward probe of his own diplomatic operation, to identify and test Taliban negotiating channels. Rubin was an unconventional choice. He had never worked in government before Holbrooke drafted him, never mind on top secret, compartmented national security matters. Yet Doug Lute, the general who ran Afghanistan and Pakistan policy at the National Security Council, remained friendly with Rubin and protected him. As Rubin provoked the national security bureaucracy with his travel and discussions with Taliban interlocutors, Lute made sure the White House and the Pentagon did not cut him off.
On December 11, 2009, ten days after Obama’s West Point speech, Rubin arrived at the Foreign Office in London to talk about his project with British counterparts. A Taliban specialist from MI6 joined the discussion. “Change is in the air,” the British intelligence officer reported. A “substantial part” of the Taliban senior leadership “is interested in reconciliation” talks. Their motives were not benign—they wanted to divide the N.A.T.O. coalition in Afghanistan by demonstrating willingness to compromise. Yet it remained hard to read who within the Taliban high command was behind these gestures or where Mullah Mohammad Omar stood.
In a separate meeting during this period, a senior Foreign Office official asked an MI6 counterpart: If we wanted to pass a private message to Mullah Omar, do we have a reliable channel to do so? Not really was the reply.
Foreign Secretary David Miliband sought to push the Obama administration toward
negotiations. Yet Prime Minister Gordon Brown’s government did not think having Barney Rubin fly around the Gulf looking for Taliban intermediaries was the best approach. Drawing on experience in Northern Ireland and elsewhere, the British argued that the C.I.A. should open up initial direct contacts with the Taliban, to probe, collect insights, and build confidence in the event that serious negotiations evolved. The C.I.A. had cachet. Enemies of the United States often preferred to deal with the spy service directly. Also, Langley managed the American relationship with I.S.I., the Taliban’s historical patron.
As Steve Kappes, the C.I.A.’s deputy director, put it during one discussion, “You don’t think the I.S.I. doesn’t have the Taliban penetrated and will find out everything that is going on anyway? Why not collaborate with I.S.I. and guard our interests at the same time?” Rubin and other interlocutors heard repeatedly from Taliban leaders that they wanted to speak directly to the United States, free from I.S.I. interference. Yet it was likely that Kappes was correct about the penetration issue—the Taliban could not travel easily outside Pakistan without passports I.S.I. approved, unless they resorted to forged documents.
Holbrooke, however, didn’t want to lose control of any talks that might lead to negotiations to end the Afghan war—that was his turf, his potential legacy. He merely wanted the C.I.A. to collect intelligence about the Taliban’s leadership and internal politics. He preferred Rubin, whom he could control, more or less, and he feared that the C.I.A. had become imprisoned intellectually about Afghanistan by wartime and counterterrorism dogma.9
