Thursday, May 7, 2009

On the Nuclear Edge by Seymour M. Hersh

A Reporter at Large

Seymour (Sy) Myron Hersh (born April 8, 1937) is an American Pulitzer Prize winning investigative journalist and author based in Washington, D.C. He is a regular contributor to The New Yorker magazine on military and security matters. [Courtesy Wikipedia/The New Yorker]

On the Nuclear Edge by Seymour M. Hersh

March 29, 1993

On May 30, 1990, President Mikhail Gorbachev of the Soviet Union arrived in Washington for his second summit meeting with President George Bush. The Cold War was over, and the publicly announced agenda reflected that fact: the two world leaders would concentrate their talks on the future of unified Germany and on renewed negotiations to reduce long-range nuclear weapons. Most Americans were increasingly upbeat about the prospects for world peace. A Times/CBS public-opinion poll of more than eleven hundred Americans taken a week before the summit showed that fewer than one in five believed nuclear war to be likely by the year 2000—far fewer than those interviewed in earlier polls.

There was a fearful irony in the poll, because in the days before Gorbachev’s visit the Bush Administration became convinced that the world was on the edge of a nuclear exchange between Pakistan and India, as both nations continued their tug-of-war over control of the state of Kashmir, on India’s northern border, whose status has been in dispute since the collapse of the British Empire in India, in 1947. During months of increasing tension, India had massed two hundred thousand troops, including paramilitary forces, in Kashmir, and had deployed five brigades of its most sophisticated attack unit, the Indian Army Strike Corps, fifty miles from the Pakistani border in the south. Pakistan, against which the much larger India had fought—and won—three wars since 1947, openly deployed its main armored tank units along the Indian border and, in secret, placed its nuclear-weapons arsenal on alert. There would be no repeat of the disastrous two-week war of December, 1971, when Pakistan, outgunned and outgeneraled, was dismembered by an Indian blitzkrieg and lost what is now Bangladesh.

The American intelligence community, also operating in secret, had concluded by late May that Pakistan had put together at least six and perhaps as many as ten nuclear weapons, and a number of senior analysts were convinced that some of those warheads had been deployed on Pakistan’s American-made F-16 fighter planes. The analysts also suspected that Benazir Bhutto, the populist Prime Minister of Pakistan, had been cut out of—or had chosen to remove herself from—the nuclear planning. Her absence meant that the nation’s avowedly pro-nuclear President, Ghulam Ishaq Khan, and the Pakistani military, headed by Army General Mirza Aslam Beg, had their hands, unfettered, on the button. There was little doubt that India, with its far more extensive nuclear arsenal, stood ready to retaliate in kind.

Since the last week of May was summit week, President Bush and his top aides were preoccupied with Gorbachev and crucial questions about his status inside the Soviet Union. A full understanding of what could happen in South Asia during those days was thus most vivid to the men and women running American intelligence, who knew that Pakistan had long been America’s ally in the clandestine war against the Soviet Union. As early as 1950, the Pakistani government had effectively ceded remote areas of its northern provinces to the Central Intelligence Agency and to the National Security Agency—the larger and still more secretive group that, from its headquarters, at Fort Meade, in Maryland, is responsible for communications intelligence. It was from northern Pakistan that the N.S.A. eavesdropped on the Soviet nuclear facilities in Kazakhstan; it was Pakistan that provided secret bases for America’s U-2 spy flights; and it was Pakistan that served as a key jumping-off point for intelligence gathering and anti-Soviet activities by the C.I.A.

Pakistan was rewarded for its support with large amounts of American military and economic aid. The American intelligence community, to protect its investment and its continuing operations, spent many millions of dollars to recruit agents and to install technical equipment to learn as much as possible about the inner workings of its ally. Those agents and that equipment enabled America to ascertain, in the early spring of 1990, that Pakistan had gone nuclear, and that its leadership was fully prepared to use the weapons, if necessary, in a war against India.

Two of the men most deeply involved in the May, 1990, crisis recently agreed, for the first time, to discuss in on-the-record interviews what had happened. Richard J. Kerr, an even-tempered, low-key career intelligence officer, who, as deputy director of the C.I.A., coördinated the intelligence reporting in May of 1990, described the confrontation in stark terms: “It was the most dangerous nuclear situation we have ever faced since I’ve been in the U.S. government. It may be as close as we’ve come to a nuclear exchange. It was far more frightening than the Cuban missile crisis.” Kerr retired as deputy director last year, after thirty-two years in the C.I.A.

At the height of the American concern, President Bush called on Robert M. Gates, a former longtime C.I.A. official who was then serving in the White House as the deputy national-security adviser, to fly on his behalf to Islamabad and New Delhi and negotiate a stand-down between the two perennial enemies. Gates’s career at the C.I.A. had been mired by his closeness to William J. Casey, Ronald Reagan’s controversial C.I.A. director, who led the nation into the Iran-Contra morass. After Casey’s death, in early 1987, Reagan failed in an attempt to install Gates at the top of the C.I.A., and the young Casey protégé—he was forty-three years old at the time—was eventually moved into the White House job. His performance as deputy national-security adviser finally won him respect throughout the bureaucracy. Gates told me that he, too, knew that a holocaust was at risk in May. “The analogy we kept making was to the summer of 1914,” he said, referring to the inadvertent outbreak of the First World War. “Pakistan and India seemed to be caught in a cycle that they couldn’t break out of. I was convinced that if a war started, it would be nuclear.”

An enduring mystery is why the essential details of the India-Pakistan confrontation have remained a secret for almost three years, especially since it was the intervention of President Bush’s personal White House envoy that defused what looked to be inevitable warfare. Stopping a nuclear exchange seemed made to order for the public-relations machinery of the White House. A few newspaper and television reports since 1990—most notably in the Sunday Times of London and the Los Angeles Times and on NBC’s “Nightly News”—have given some details of the India-Pakistan nuclear crisis, but the reports, whose sources were not named, were debunked by the Bush Administration as exaggerations. Similarly, many officials assured me, in the course of my reporting for this article, that, as a retired Under-Secretary of State put it, “there was a lot of oversell going on.”

An obvious explanation for the high-level quiet revolves around the fact, haunting to some in the intelligence community, that the Reagan Administration had dramatically aided Pakistan in its pursuit of the bomb. President Reagan and his national-security aides saw the generals who ran Pakistan as loyal allies in the American proxy war against the Soviet Union in Afghanistan: driving the Russians out of Afghanistan was considered far more important than nagging Pakistan about its building of bombs. The Reagan Administration did more than forgo nagging, however; it looked the other way throughout the mid-nineteen-eighties as Pakistan assembled its nuclear arsenal with the aid of many millions of dollars’ worth of restricted, high-tech materials bought inside the United States. Such purchases have always been illegal, but Congress made breaking the law more costly in 1985, when it passed the Solarz Amendment to the Foreign Assistance Act (the amendment was proposed by former Representative Stephen J. Solarz, Democrat of New York), providing for the cutoff of all military and economic aid to purportedly non-nuclear nations that illegally export or attempt to export nuclear-related materials from the United States.

A second law passed that year, known as the Pressler Amendment (for Senator Larry Pressler, Republican of South Dakota), also affected the continuous flow of hundreds of millions of dollars annually in American aid to Pakistan. This legislation calls on the President to certify every year that Pakistan does not have nuclear weapons. Without such certification, Pakistan could not get any foreign aid. The certification process became farcical in the last years of the Reagan Administration, whose yearly certification—despite explicit American intelligence about Pakistan’s nuclear-weapons program—was seen as little more than a payoff to the Pakistani leadership for its support in Afghanistan.

Thus, many of the men and women in the American intelligence community who in May of 1990 found themselves monitoring what appeared to be the prelude to a nuclear exchange knew that the political leadership of the United States, flagrantly violating the law, had permitted Pakistan to buy restricted items inside the United States for its nuclear arsenal. Some of these officials, who are still working in the government, emphasized in interviews that many more nuclear-related goods were clandestinely bought inside the United States by Pakistan than by Saddam Hussein’s Iraq. This information was simply not known outside the Bush Administration early last year, when the American media and Congress became obsessed with Iraqi high-tech purchases.

There is indisputable evidence that Pakistan has been able to escape public scrutiny for its violations of the law because senior officials of the Reagan and the Bush Administrations chose not to share the intelligence about nuclear purchases with Congress. The two Republican Administrations obviously feared that the legislators, who had voted for the Solarz and Pressler Amendments, would cut off funds for the war against the Soviet Union in Afghanistan. It was yet another clash between a much desired foreign-policy goal and the law—a clash of the kind that emerged most publicly in the Iran-Contra scandals of the mid-nineteen-eighties, when the Reagan Administration diverted funds from illicit arms sales to the Iranians to get around a congressional ban on arming the Nicaraguan Contras.

The government’s ability to keep the Pakistani nuclear-arms purchases in America secret is the more remarkable because for the past four years the State Department, the Central Intelligence Agency, and the Defense Department have been struggling with an internal account of illegal Pakistani procurement activities, given by a former C.I.A. intelligence officer named Richard M. Barlow. Barlow, now thirty-eight years old, was hired by the C.I.A. in 1985 and quickly became one of the agency’s top experts on Pakistan’s nuclear program. In 1987, he was dismayed to learn, at first hand, that State Department and agency officials were engaged in what he concluded was a pattern of lying to and misleading Congress about Pakistan’s nuclear-purchasing activities. He resigned a year later, after senior agency officials attempted to bar him from working on Pakistan. In 1989, Barlow, then working as a proliferation analyst in the office of the Under-Secretary of Defense for Policy, was forced to resign, under threat of firing, in what he alleges was retaliation for his persistent investigations into nuclear smuggling and his heated objections to yet another misleading congressional briefing on Pakistan. His allegations, initially written off as the mutterings of a malcontent, lately have been taken with increased seriousness by government investigators. In a recent interview, Sherman Funk, the Inspector General of the State Department, described Barlow as “one of the most brilliant analysts I’ve ever seen,” and depicted his forced resignation as an injustice.

Barlow, who is currently a consultant for the intelligence community, agreed early this year to discuss his allegations in an interview, after initially attempting, he said, to work within the system. His allegations of a State Department and C.I.A. conspiracy to shield Pakistani nuclear-related purchases from members of Congress have severely tested the government’s system of oversight by Inspectors General. As of today, there have been no fewer than five I.G. reports dealing with Barlow, two of them in the Defense Department, with only Sherman Funk confirming some of his essential allegations. But Funk has recently urged his Defense Department colleagues to reopen their inquiries.

Barlow’s complaints of government wrongdoing got a boost when Paul C. Warnke, who served as an Assistant Secretary of Defense under President Lyndon Johnson and as director of the Arms Control and Disarmament Agency in the Carter Administration, reviewed the case and agreed to represent Barlow on a pro-bono basis. “He was doing his job as best he could, and was penalized for doing his job,” said Warnke, who is now a partner in the law firm of Howrey & Simon. “I don’t think that’s the way the government should operate. It’s as simple as that.” Warnke and a law-firm partner, James C. Duff, have assembled evidence indicating that the Defense Department’s I.G. report included mischaracterized—and perhaps fabricated—testimony. Their information has been turned over to the Senate Armed Services Committee, and on March 17th Senator Jeff Bingaman, Democrat of New Mexico, a committee member, formally asked the Pentagon to reopen its investigation.

In interviews for this article over a three-month period, many of Barlow’s former C.I.A. and State Department colleagues confirmed his essential allegation—that the full story of the Pakistani purchases was deliberately withheld from Congress, for fear of provoking a cutoff in military and economic aid that would adversely affect the prosecution of the war in Afghanistan. In addition, senior members of Congress, including Senator John Glenn, Democrat of Ohio, who is the chairman of the Governmental Affairs Committee, and former Representative Solarz, who was the chairman of the Asian and Pacific Affairs Subcommittee of the House Foreign Affairs Committee, acknowledged that they had not been formally briefed, as required by law, about any significant Pakistani procurement except for one publicly known case, which in 1987 forced President Reagan to invoke—and then waive—foreign-aid sanctions against Pakistan.

Furthermore, Senator Glenn, the former astronaut, who is widely respected inside the government for his discretion—he is a member of the Senate Select Committee on Intelligence—said that he had been told nothing about the nuclear scare in 1990. “They should have told me,” he said, citing legislation that compels the executive branch of government to keep him and other ranking members of Congress “fully and currently informed” on all proliferation-related activities.

It isn’t difficult to understand why Glenn had not been briefed, even in camera: he has been one of the most persistent public critics of Pakistan’s nuclear ambitions and of the failure of the Reagan and Bush Administrations to do something about them. Glenn told me he knew that many of his colleagues in the Senate and the House had viewed nonproliferation as a secondary issue, especially when they saw a chance to take on and defeat the Soviets on the battlefield, as in Afghanistan. “I always thought in terms of the bigger picture—the nonproliferation treaty,” Senator Glenn said, referring to the 1968 international agreement, signed and ratified by the United States, to halt the spread of nuclear weapons. “We made a commitment that we’d cut off aid to transgressors, and we had to keep faith with those Third World people who signed with us. I didn’t think I had any option but to press for enforcement of the law against Pakistan.” He was in the minority on that issue all through the Reagan and Bush years, he said. “The Administration would always come to me and say how important it is to keep the arms flowing through to Afghanistan. I’d take my case on nonproliferation to the floor and lose the vote.”

Richard Kerr and Robert Gates still decline to discuss specific intelligence about the nuclear crisis in May of 1990, but a detailed account of what the American intelligence community knew—and did not know—about it has emerged in scores of interviews with Americans who were closely involved in the crisis management, and also with some Pakistani and Indian military and government officials who had direct knowledge. Most of those men and women who agreed to be interviewed did so after being promised anonymity.

One surprising consensus was that the 1990 confrontation had its beginnings in a large-scale Indian military exercise, code-named Brass Tacks, that started near the end of 1986. Rajiv Gandhi was Prime Minister of India, and there was a new and aggressive Indian Army leader, General Krishnaswami Sundarji. Gandhi and Sundarji agreed to stage the largest military exercise in modern Indian history, involving all branches of the armed forces and as many as four hundred thousand troops, starting in December. The exercise would take place not in India’s far north, where the always tense state of Kashmir is situated, but in the desert area of Rajasthan, a few hundred miles to the south, and roughly a hundred miles from the Pakistani border, which—as the Pakistan government was sure to note—was an ideal location from which to launch a cross-border operation into the Pakistani state of Sind that could cut Pakistan in half.

Another feature of the exercise was a decision by General Sundarji to integrate India’s special weapons, including tactical nuclear bombs, into the day-to-day field maneuvers of the troops. American analysts concluded that the Indian operations, which at their height involved more than a thousand armored vehicles, were as large as some exercises carried out in Western Europe by the North Atlantic Treaty Organization. President Muhammad Zia ul-Haq, who was also the Army chief of staff, viewed the Indian maneuvers as a direct threat, and ordered his armored units to move to the front, where, by mid-January, the two armies stood within firing range along an extended border area. At the height of the buildup, the Indian leadership decided to provide full-scale briefings to the Indian media about the Brass Tacks exercise, in which General Sundarji declared that they were non-provocative—there had been no public discussion of the huge troop movements until then—and the crisis abated.

The American intelligence community understood that the Indian operation was potentially more dangerous than anyone in the public or the media realized. For one thing, there was evidence that General Sundarji, with the encouragement of some of the more hawkish elements in the Indian government, had toyed briefly with the possibility of simply moving his armies, which had gone into the field with extra fuel and ammunition, across the border and dismembering Pakistan. There was little question that the smaller and less mobile Pakistani Army would be unable to stop the Indian advance, and thus the landlocked Pakistan of the north, including the capital of Islamabad, could be cut off from the nation’s main port, at Karachi, on the Arabian Sea. “This is the Pakistani nightmare,” an Indian military analyst told me. At this point, Dr. Abdul Qadeer Khan, a German-educated metallurgist, who has directed Pakistan’s nuclear-weapons laboratories at Kahuta since the nineteen-seventies—he is known to some as the father of the Pakistani atomic bomb—gave a stunning interview to an Indian reporter, in which he explicitly claimed that Pakistan had the bomb. “What the C.I.A. has been saying about our possessing the bomb is correct,” he was quoted as saying. “They told us Pakistan could never produce the bomb and they doubted my capabilities, but they now know we have it.” His government did not want to use the bomb, Khan added, but “if driven to the wall there will be no option left.” His remarks created instant embarrassment and controversy in Washington—the Reagan Administration had reported only a few months earlier that there was no evidence of a Pakistani nuclear bomb—and Dr. Khan recanted his remarks within days, claiming that the reporter had tricked him into giving the interview.

Dr. Khan’s goal was not to embarrass the United States but to convey a not very subtle message to the Indians: any attempt to dismember Pakistan would be countered with the bomb. Two days after the Khan interview was published, Prime Minister Gandhi publicly criticized the United States for failing to stop Pakistan’s nuclear-weapons program, and for continuing to supply Pakistan with economic and military aid. As tensions diminished in the spring of 1987, there was one very troubling legacy for the American intelligence community: studies showed that the main intelligence services for both sides—India’s Research and Analysis Wing and Pakistan’s Inter-Services Intelligence Directorate—had exaggerated their reports to their senior leadership. It was clear that, in a crisis, both raw and isi were eager to provide incendiary intelligence without being sure of its reliability. Such eagerness could become deadly.

By 1987, the American intelligence community had become exceedingly proficient at learning what could be learned about another nation’s nuclear ambitions. From the early days of the Cold War and the first U-2 aerial-reconnaissance flights, special teams of analysts had learned to look for clues to nuclear proliferation. One telltale sign was the pouring of a thick concrete floor on an isolated construction site, which the Israelis had done in mid-1958 at Dimona, the site of their future nuclear reactor in the Negev Desert. C.I.A. and N.S.A. analysts immediately understood that Israel, public denials to the contrary, was going after a nuclear-weapons arsenal.

Sometimes the American analysts ignored the vast array of technical intelligence and relied on simple common sense. The C.I.A. had learned the advisability of continually monitoring the intellectual debate inside any nation suspected of being on the verge of going nuclear. Analysts correctly concluded that India had gone nuclear in the mid-nineteen-seventies, when its officer corps suddenly stopped publishing articles and essays on what had been a long-standing and occasionally vituperative argument about the strategic and tactical implications of such a move.

In Pakistan, the United States had done more than monitor an intellectual debate or spy on a suspected nuclear facility. For nearly two decades, the N.S.A. had maintained an extensive “watch list” of high-tech companies in Germany and Switzerland whose telexes and facsimile transmissions were routinely intercepted and translated for signs of nuclear trafficking with Pakistan, which was known to be an illicit purchaser of nuclear materials from the West. The constant spying paid off. C.I.A. operatives managed to come up with a complete set of the floor plans for an ambitious Pakistani uranium-enrichment plant at Kahuta, twelve miles west of Islamabad, while the plant was under construction. The plans showed that the hot, or radioactive, work areas at Kahuta had been built as many as five stories underground, to guard against a surprise Indian bombing raid. Furthermore, once Kahuta was operational the C.I.A. found a way to obtain firsthand information, in detail, about nuclear-weapons work there.

Pakistan, apparently in response to India’s Brass Tacks exercise, aggressively expanded its efforts in 1987. American satellites watched that year as a thick concrete floor was poured for what would become a second uranium-enrichment site—at Golra, also near Islamabad. It became known to West German intelligence that Pakistan had violated German law by buying a small plant for purifying and storing tritium gas. Tritium is a by-product of the irradiation of lithium-6 in a reactor; when tritium is inserted in a warhead, it provides additional neutrons at the moment of fission—in essence, jump-starting the weapon when it reaches critical mass. The result is a larger explosive yield with smaller amounts of enriched material—more bang for the buck.

There was widespread agreement inside the American intelligence community in 1987 that Pakistan had enough enriched uranium to put together perhaps six nuclear devices. But was it in a form that could actually be used in a warhead? In more precise terms, the unresolved question was this: Had Dr. Khan’s men converted enriched-uranium hexafluoride gas (UF6) into a metal? The question resulted in the design of a highly sensitive C.I.A. operation, which produced irrefutable evidence that Pakistan was capable of manufacturing weapons-grade enriched-uranium metal at a facility near Islamabad—but not at Kahuta. The metal could then be machine-tooled to fit into a warhead small enough to hang under an F-16 wing.

Despite such evidence, the Reagan and Bush Administrations certified Pakistan in 1987, 1988, and 1989 as not having a nuclear weapon, the rationale being that there was no specific evidence that Pakistan had indeed done what it was known to be capable of doing. “There is no question that we had an intelligence basis for not certifying from 1987 on,” Richard Kerr told me. The public American rationale for certification was that the continued flow of American weapons and ammunition to Pakistan would reassure its leadership that it could rely on conventional arms, and thus have no need to go nuclear. It was a very thin argument, as everyone involved knew. The C.I.A.’s role in all this, Kerr said, was merely to supply the political leaders with the best available information and for them to carry on from that point.

On August 17, 1988, President Zia, ten Pakistani generals, and Arnold L. Raphel, the United States Ambassador to Pakistan, were killed in the crash of Zia’s Air Force C-130 a few miles north of the garrison town of Bahawalpur. General Beg, Zia’s successor as the Army chief of staff, was known to be at least as deeply committed to Pakistan’s nuclear project as Zia, but was also viewed as being much less dependent on the United States. So was Pakistan’s new acting President, the seventy-three-year-old Ghulam Ishaq Khan, who had been Pakistan’s Finance Minister and was then head of the Senate. Zia and President Khan had been among the most enthusiastic political and financial supporters of Abdul Qadeer Khan’s expensive operations at Kahuta over the previous decade. Beg and President Khan emerged from the trauma of the Zia crash with firm control over the Pakistani nuclear project, and they maintained their control even after the general elections in November of 1988—the first democratic elections in Pakistan in eleven years—which led to the formation of a populist government headed by Benazir Bhutto. Ms. Bhutto’s father, former Prime Minister Zulfikar Ali Bhutto, who was hanged in 1979, after being convicted of murder (in a much disputed trial), is widely credited with being the founder of the Pakistani nuclear project, but the American intelligence community simply does not know whether Benazir Bhutto was ever permitted full access to her nation’s nuclear secrets.

A galvanizing moment for the new Pakistani leadership came in June of 1989, four months after the Soviet Union began its withdrawal from Afghanistan, thereby handing the United States and its Pakistani allies a major Cold War victory. With the Soviets gone, President Bush and James Baker, the Secretary of State, decided that the United States could no longer ignore the evidence of the Pakistani bomb. The question of what Ms. Bhutto knew, or didn’t know, about the bomb would be resolved when she paid a scheduled state visit to Washington in June; at that time she was provided with a detailed briefing by William H. Webster, the C.I.A. director. To dramatize the extent of American knowledge, Webster arranged for Ms. Bhutto to be shown a mockup of a Pakistani nuclear bomb. A day later, at a private meeting with President Bush, Ms. Bhutto was told that the United States would certify Pakistan as not having a nuclear weapon in 1989, but that no certification would be available in 1990—not unless she could assure the White House that her government would not take the final step of producing nuclear-bomb cores. To soften the blow, the President told her that he would authorize the sale of sixty more F-16s—planes desperately needed by Pakistan, whose Air Force was outmatched by India’s. She would have something to take back to the generals.

Ms. Bhutto, who was bitterly attacked in Pakistan last winter for her criticism of her country’s nuclear program, did not wish to be interviewed for this article. But Mark A. Siegel, a former Carter Administration official and close Bhutto associate, who served as Pakistan’s American lobbyist in 1989 and 1990, described her feelings of disbelief upon being briefed by Webster. “The briefing was more detailed” than any information that she, as Prime Minister, had been provided, Siegel said. “It also showed that the military was doing it behind her back.” Nonetheless, the ardently pro-American Ms. Bhutto went before a joint session of Congress and, to thunderous cheers, categorically promised that “we do not possess nor do we intend to make a nuclear device,” adding, “That is our policy.”

The tough, albeit late, new American policy toward a nuclear-armed Pakistan, as it was relayed by President Bush to Ms. Bhutto, did not play well in Islamabad. “The Paks understood us better than we understood ourselves,” one informed American official explained. “They knew that once the Soviets were whipped in Afghanistan we wouldn’t need them anymore. Would we unilaterally defend Pakistan? Never. Our relationship with Pakistan was to counter the Soviet-Indian relationship. The Pakistanis knew that time was limited. And that’s why they went balls out on the nuclear program. Benazir never had a chance. If they held back the bomb from us, they could have held it back from her as well. It was too serious.”

The official added that General Beg and his colleagues understood, as did the American intelligence community, that Pakistan could never stand up to a full-scale Indian assault. “The only way for the Pakistanis to deal with the Indians is to be able to take out New Delhi,” he said. “There’s no way that sending ten F-16s with conventional bombs is going to do it. Only the nukes could strike back.”

Not surprisingly, General Beg’s relationship with Robert B. Oakley, the new American Ambassador, had none of the warm rapport that had existed between President Zia and Ambassador Raphel. Beg and Oakley, one American recalled, were constantly at odds by 1990, and would “literally yell at each other.” The C.I.A. eventually concluded, in a highly classified personality profile, that General Beg was more closely attuned to the Islamic world than to the West. He was viewed, one American said, as “a fifty-year-old Muslim aristocrat who suddenly became very religious . . . and thought Iran was a savior.” In an interview, Oakley recalled that Beg’s tilt toward Iran had destroyed their relationship. “Beg came back in February of 1990 from Teheran and told me, ‘I’m greatly reassured,’ ” Oakley said. “He was pleased as punch. He said, ‘Now we’re in good shape. With the support from Iran promised me, we will win in case of war over Kashmir.’ ” The Pakistani General said as much in a meeting a few weeks later with General H. Norman Schwarzkopf, head of the United States Central Command, which has military responsibility for the Middle East and parts of South Asia. Beg’s “pro-Iranian kick was scary to us,” Oakley said. Oakley was interviewed a few days after his return from Somalia, where he served until early this March as the United States special envoy.

Another American who has been involved in nuclear intelligence for many years told me that his reservations about Pakistan and “an Islamic bomb” transcended General Beg: “What if they decide they truly want to be a player on the world stage? Look at the Pakistani ego. They want to be with the Big Boys—China, the United States, the Soviet Union. Pakistan is deadly serious.”

Dr. Khan’s few public statements have been unsettling to many American policymakers. “Western countries had never imagined that a poor and backward country like Pakistan would end their [nuclear] monopoly in such a short time,” Khan told a Pakistani newspaper in 1984. “As soon as they realized that Pakistan had dashed their dreams to the ground, they pounced at Pakistan and me like hungry jackals and began attacking us with all kinds of accusations and falsehood. . . . How could they tolerate a Muslim country becoming their equal in this field. . . . All Western countries including Israel are not only the enemies of Pakistan but in fact of Islam. . . . All these activities are part of the crusade which Christians and Jews have been carrying on against Muslims for about one thousand years.”

India’s military leadership has been worried for years about the combination of a new crisis in Kashmir and the Pakistani bomb. In a study published in 1979, D. K. Palit, a retired Indian major general, noted that the liberation of largely Muslim Kashmir could be “regarded as a jihad, a holy war of Islam, which would justify the threat of the use of the Islamic bomb against India.” What was unnerving was “not that having acquired a minimum capability Pakistan would automatically launch an invasion across our border, using nuclear weapons to bomb Indian targets,” Palit wrote. “Even an unstable military government would seek a more sophisticated scenario than that.” Much more frightening was the possibility that Pakistan “could declare that it intended to use its nuclear weapons in a defensive capacity only, that is, as a deterrent to an Indian attack on Pakistani territory.” In that case, Palit wrote, India’s conventional deterrent against a preëmptive Pakistani attack on Kashmir—its threat of a counteroffensive with the vastly superior Army, Navy, and Air Force—“would not be credible.”

Major General Palit’s concerns and assumptions came into play in 1990, as Pakistan and India engaged in their near-war over Kashmir. That western Himalayas area, filled with caverns, mountains, lush valleys, gorges, rivers, rapids—and twice as many Muslims as Hindus—was the only state in the South Asian subcontinent not allowed to vote on self-determination in 1947, after Britain’s withdrawal from its Indian empire triggered widespread religious rioting between Muslims and Hindus. Kashmir, which is about the size of Utah, was to remain under Indian control, but continued religious warring and demands for independence led the United Nations to arrange a ceasefire in 1948, providing for two-thirds of Kashmir to be a self-governing entity within the Indian Union, and for Pakistan to be nominally in control of the remaining third. The United Nations mandate also called for a plebiscite to allow the people of Kashmir to decide their own fate, but India has not permitted the election to take place, insisting that Pakistan must first withdraw its troops from Kashmir. Pakistan refused to do so unless India also withdrew its troops. India has taken advantage of the decades-old impasse to steadily increase its military and political control over its area of Kashmir, bringing itself into direct conflict with those Kashmiris who insist on independence.

Kashmir is still categorized by the United Nations as an occupied territory whose final fate is yet to be determined—a status similar to that of the West Bank and the Gaza Strip in the Middle East. As the birthplace of the late Prime Minister Jawaharlal Nehru, Kashmir remains an emotional battleground for India, which is determined to hold on to the area by force. India views the loss of the predominantly Muslim area as a potential threat to the country’s status as a secular nation. Pakistan has responded to the tenacious Indian hold on Kashmir by engaging in a large-scale covert effort to foment revolution, in part through the systematic use of insurgents. Neither India nor Pakistan is willing to see an independent Kashmir; the two nations have repeatedly gone to war to keep control of the territory. In the view of American intelligence, the weak governments in place in Pakistan and India in May of 1990 were willing to run any risk—including nuclear war—to avoid a disastrous military, and thus political, defeat in Kashmir.

There is no evidence that the Pakistani leadership or its intelligence services had advance knowledge of or were in any way involved in the initial uprising in Kashmir, which was triggered in January of 1990 when Indian police opened fire and killed fifty pro-independence demonstrators who were protesting what they viewed as the latest of India’s “puppet” governments. But Pakistan did move quickly to exploit the unrest. By early spring, the Pakistanis were known to be operating at least thirty camps in their territory for the training of Kashmiri Muslims and were prepared to aid them in slipping across the border. Mass Muslim-led demonstrations of more than a hundred thousand people were organized in Srinagar, the capital of Kashmir, to protest Indian rule. India blamed its enemy, Pakistan, for its problems, although there was evidence that the protests were in part indigenous. India responded to the angry protests with a heightened police and military presence throughout Kashmir, and that led, almost inevitably, to brutal repression and shootouts. Between January and late April, as civilian control ebbed, hundreds of Kashmiris were shot dead in the street. India moved its crack Strike Corps into Rajasthan, within fifty miles of the border, and once again, as in 1987, there were fears in Islamabad of a preëmptive Indian attack to cut off southern Pakistan from the north. Pakistan moved its 1st and 2nd Armored Corps toward the border.

The American intelligence community noticed an intense increase in Pakistani radar activity early in the year. Earlier reports showed that the Pakistani Air Force, working closely with officials from Pakistan’s nuclear-weapons program, had stepped up its F-16 training to practice what seemed to be the dropping of a nuclear bomb. Further intelligence, from Germany, reported that the Pakistanis had designed a nuclear warhead that could be fitted under the wing of an F-16, and that the design had gone through a series of wind-tunnel tests. Pakistan was also reported to have learned to program its in-flight computer system to provide the correct flight path for a nuclear-bomb run.

The language from New Delhi and Islamabad heated up. In mid-March, Ms. Bhutto visited a training camp inside the Pakistani-controlled part of Kashmir and pledged nearly five million dollars in support of “freedom fighters” —a phrase clearly aimed at provoking the Indian leadership. A few weeks later, Vishwanath Pratap Singh, India’s new Prime Minister (he defeated Rajiv Gandhi in national elections in November of 1989, with the aid of the militant Hindu Bharatiya Janata Party, or B.J.P.—the Indian People’s Party), publicly warned Pakistan in a speech to the jingoist Indian Parliament that it “cannot get away with taking Kashmir without a war.” Singh added that India would not allow Pakistan to achieve nuclear superiority. One leader of the B.J.P. went further and suggested to cheering members of Parliament that in case of war “Pakistan ceases to exist.”

Sometime in the early spring of 1990, intelligence that was described as a hundred per cent reliable—perhaps an N.S.A. intercept—reached Washington with the ominous news that General Beg had authorized the technicians at Kahuta to put together nuclear weapons. Such intelligence, of “smoking gun” significance, was too precise to be ignored or shunted aside. The new intelligence also indicated that General Beg was prepared to use the bomb against India if necessary. Precisely what was obtained could not be learned, but one American summarized the information as being, in essence, a warning to India that if “you move up here”—that is, begin a ground invasion into Pakistan—“we’re going to take out Delhi.” Washington suddenly joined Islamabad and New Delhi in realizing that a crisis had developed. “I began to scream that what’s going to happen is Brass Tacks all over again,” Ambassador Oakley recalled. He had been serving on the National Security Council during the confrontation in 1987, which was resolved by the direct intervention of two strong leaders, Pakistan’s President Zia and India’s Prime Minister Rajiv Gandhi. “Rajiv and Zia are gone, and this time they won’t stop,” Oakley said he warned Washington in 1990. “We’d better move before the momentum builds.”

Washington decided to augment the available technical intelligence. The orbiting flight paths of American satellites were adjusted to increase the hour-to-hour overhead coverage of South Asia. The joint American-British listening post in Diego Garcia, in the Indian Ocean, which monitors both Pakistan and India, was provided with additional electronic-eavesdropping requirements (a move known at N.S.A. field stations as “tasking”) and ordered to increase its coverage. Sometime in May, as conditions inside Kashmir worsened, an orbiting American satellite relayed photographs of what some officials believed was the evacuation of thousands of workers from Kahuta.

An American analyst who participated in White House situation-room meetings throughout the crisis described what happened next: “We thought the reason for the evacuation of Kahuta was that they expected a retaliatory attack by India, in response to a Pakistani first strike. We were keyed to an offensive ground strike”—into the Pakistani state of Sind—“by the Indians. The Pakistanis were going to cut it off with a nuke. We thought they’d go for Delhi.”

At this point, the American analyst recalled a secret meeting he had had with Abdul Qadeer Khan during an international energy conference in Europe in the mid-nineteen-eighties: “I was told by Khan in no uncertain terms ‘Never again. Whatever else occurs, even if we tell you we’ve terminated’ ”—ceased working on the nuclear bomb—“ ‘I can tell you that I will not be allowed to terminate, because we must continue to show the Indians that we have the ability to never again be defeated at their hands.’ ”

There was further evidence that May of Pakistani intent: “We’d spotted a facility in the mountains”—in Balochistan, in southwestern Pakistan. “If there was a nuclear-storage facility, that had all the signs of being it.” The Pakistani military were conducting high-explosive tests near the suspected storage facility. Such tests, which had been monitored by the United States since the mid-nineteen-eighties, were an essential element of the nuclear-weapons process; nuclear fission is achieved by triggering precise, simultaneous explosions inside a warhead, which instantly compress the core of enriched uranium or plutonium. Initially, the test area seemed to be a normal facility, the analyst recalled, except that the Pakistani military protected it with an unusually high degree of security. Eventually, the intelligence community learned from a source inside Pakistan that A. Q. Khan had gone to the site for a visit—Khan was under constant surveillance—and the test site and its nearby storage area were considered to be nuclear-related. If the Pakistanis had manufactured more than a few weapons—there was talk in the White House that spring of as many as ten warheads—and if they truly expected a preëmptive Indian attack, the weapons would have to be dispersed to Pakistani Air Force bases.

Just as expected, satellite and other intelligence later produced signs of a truck convoy moving from the suspected nuclear-storage site in Balochistan to a nearby Air Force base. “There were moving secure zones all over the convoy route,” the analyst said, similar to the way, years ago, that the American military cleared the highways in front of its convoys carrying nuclear weapons. Once the Pakistani convoy reached the airfield, the pattern was repeated. “They had beefed up security,” the analyst said. “People who normally worked on the flight line are all of a sudden sealed out of the area. They’re not allowed to be there. All of this simple stuff becomes important. We’ve watched the Pakistanis go through normal alerts and drills, and you can tell the difference. You don’t have to be a rocket scientist to understand what’s going on, because you’re dealing with the military, who work in set patterns. This is low-level human-intelligence crap. No big deal. We’re watching for things to happen that never happened before. Their big mistake was putting on more security than they needed.”

Eventually, the intelligence community picked up a frightening sight, the analyst recalled: “They had F-16s pre-positioned and armed for delivery—on full alert, with pilots in the aircraft. I believed that they were ready to launch on command and that that message had been clearly conveyed to the Indians. We’re saying, ‘Oh, shit.’ We’ve been watching the revolution in Kashmir, the internal problems in India, and we look at the Pakistani pre-positioning. These guys have done everything that will lead you to believe that they are locked and loaded.”

The Pakistani government had told Washington that it had intelligence—whose reliability was suspect, as was all Indian and Pakistani intelligence in moments of crisis—indicating that India was prepared to send its army across the border. “We were holding a session in the National Security Council conference room in the White House basement,” the analyst said. “And one of our guys”—a senior military man who had served as defense attaché with the United States Embassy in New Delhi—“said we ought to focus on getting the Paks to hit Tarapur, not Delhi.” A large reactor at Tarapur, north of Bombay, was known to be one of India’s main sites for the chemical extraction of weapons-grade plutonium. “In other words, he wanted to do a tit for tat: Kahuta for Tarapur. It was the old limited-war issue: If they do a strike, what could we do to get them to neutralize each other?”

Former Ambassador Oakley, in his interview with me, depicted the intelligence analysts in Washington as being overwrought about the imminence of war and a nuclear exchange. “We never had any hard indications that any nuclear warheads had been delivered to an airbase. You could guess that,” Oakley said. “isi was putting out all sorts of messages, but we had no evidence that a nuclear exchange was imminent.” In his view, Oakley added, full-scale war between Pakistan and India would not break out until early fall, the end of the rainy season. “Once it began, there was a strong likelihood they would have gone all the way,” he said. “I don’t blame people [in Washington] for being nervous. That’s the way intelligence analysts are supposed to work. But the evacuation of Kahuta is not necessarily evidence of war,” he said, because “Pakistan has always been afraid” of a preëmptive Indian raid on the nuclear-weapons center.

One American nuclear-intelligence expert subsequently described Oakley’s comments as a “classic case of what we were dealing with” in reporting to the State Department and the White House on an ally’s nuclear-weapons program: “It’s a warning situation, but they want smoking guns for everything. The guy is saying, ‘Wait until the bombs are dropped on New Delhi.’ ”

Richard Kerr was also persuaded by the intelligence that the chances of a nuclear war between India and Pakistan were great. “There’s no question in my mind that we were right on the edge,” the former C.I.A. deputy director recalled. “This period was very tense. The intelligence community believed that without some intervention the two parties could miscalculate—and miscalculation could lead to a nuclear exchange.”

The fears of the intelligence community were impossible to ignore, even amid the crunch of a summit week. It was at this point that President Bush ordered a White House airplane sent to Moscow to pick up Robert Gates—who was part of a high-level advance team preparing for the summit—and fly him to the rescue. One involved official asked rhetorically, “Why Gates? Because he went carrying the President’s personal imprimatur: ‘He’s my top adviser on nuclear matters. I’m sending my No. 1 career intelligence official. He knows. He can’t be bullshitted.’ ”

In his interview with me, Gates was quick to acknowledge that he was not an expert on India or Pakistan and had no personal relationship with the leadership of either country, but had indeed been sent to South Asia as the President’s man, carrying special messages to President Khan in Islamabad and to Prime Minister Singh in New Delhi. “The President wanted somebody who could go discreetly and without an entourage,” Gates said. “Somebody who spoke for him.” Gates was being briefed as he was moving into action—standard operating procedure for a senior White House aide—and not until he actually began the mission, he said, “did I have a full sense of the danger.” He added, “These guys”—in India and Pakistan—“felt their country was at stake.”

The major reason for his initial lack of alarm, Gates recalled, was the ho-hum attitude in Moscow toward the India-Pakistan standoff. Before leaving Moscow, he and Secretary of State James Baker “asked the Soviets to weigh in with the Indians,” their longtime ally—a step that would have been automatic in the Cold War days. But the Soviets “basically pooh-poohed” the notion that there was a crisis, Gates said. “It eventually became clear that they were not going to weigh in, because they didn’t want to expend the political capital and they didn’t have the information” that Washington had.

The high-powered mission began on an ominous note: Gates and his companions, who included Richard N. Haass, the ranking National Security Council aide for the region, and John Kelly, the Assistant Secretary of State for Near East and South Asian Affairs, were stood up by Benazir Bhutto. President Bush had written a private note to the Pakistani Prime Minister, with whom he had established a good rapport the previous year, and the Gates team was expecting confirmation of a meeting place during a refuelling stop at Athens. Ms. Bhutto’s reign as Prime Minister had sharply disappointed many of her supporters in Pakistan and elsewhere; they saw little evidence that she had the ability to run the government. There also were heated complaints of nepotism and corruption in her administration. Many Americans assigned to the Embassy in Islamabad and as military attachés sided with General Beg and President Khan in their intense power struggle against Ms. Bhutto. The infighting was often very ugly. Pakistan’s Inter-Services Intelligence Directorate, its most clandestine intelligence agency, was known to have bugged many of Ms. Bhutto’s meetings, and even videotaped her in private moments with her husband. Copies of some of that material had made their way to the large and very active C.I.A. station in Islamabad, whose agents—not unnaturally—discussed them with at least a few colleagues throughout the world. Many of these men, a well-informed congressional aide explained, saw Ms. Bhutto “as a nuisance,” and not—as did many in Congress—as someone capable of giving Pakistan “a chance for democracy.”

Ms. Bhutto’s popularity at home had revived with her ardent support for the Kashmiri freedom fighters, as she described the Kashmiri Muslims in a series of jingoistic speeches that spring throughout Pakistan. As Gates set out on his mission, she was in Yemen, one of her stops on an extensive tour of the Gulf states, to drum up Islamic backing for her nation’s demand for elections in Kashmir. She did not meet Gates, and even today he remains angry and perplexed about her slight. “First, they had said ‘Let’s meet in Cairo,’ ” Gates recalled. “I got word in Athens that we’d meet someplace else. She yanked my chain for four or five hours. Finally, she said, ‘Why not come to Yemen?’ I said, ‘Screw it.’ It was clear that she did not want to talk about it. She knew what I had to say.”

Gates’s main worry as he headed for South Asia was the essential instability of both governments, which he feared were “too weak to stop a war,” he said. “There was the view that both sides were blundering toward a war, and we were afraid that it would go nuclear. And who knew what the consequences would be?” In addition to taking along Presidential letters urging restraint, Gates was authorized to tell the leadership in Islamabad and New Delhi that the United States was prepared to share its most sensitive satellite intelligence with both sides, so they would simultaneously be able to verify troop withdrawals—if they could be agreed upon—from border areas.

Gates said that he had made no effort to be especially tactful during his meetings with General Beg and President Khan, who were, after all, representing a loyal Cold War ally and a key partner in the Afghanistan war. “I’m not much of a diplomat,” he told me. “We simply wanted to show them that they were bungling toward a war. The idea was to go in and voice our concerns, and also to show that we knew what we were talking about. I gave an explicit briefing to both sides—in excruciating detail.” He said that he told the Pakistanis, “ ‘Look, I’m not here to solve the Kashmir problem or discuss regional arms control. I’m here because we think there is a short-term problem that we want to defuse.’ I looked straight at Beg and said, ‘General, our military has war-gamed every conceivable scenario between you and the Indians, and there isn’t a single way you win.’ ” It was a tough message for a proud and bellicose Army chief of staff. Beg said nothing, and Gates detected no visible response. “He was very cool,” Gates said. “That was the high point.”

Gates was also accompanied by Oakley, who recalled, “The essential goal of the meeting was to get the message”—that Pakistan could not win a war with India—“to President Khan. It was quite clear that Khan had not been given a full briefing” on the risks of war by Beg, “who sat there looking smug. Khan became quite sober,” Oakley said, after Gates warned him that if war breaks out “Don’t expect any help from us.”

One of General Beg’s military colleagues subsequently gave me a blunt account of the Gates presentation. According to the officer, Gates angrily let the Pakistani government know that the United States had hard evidence that “we’d crossed the line” and developed nuclear warheads. The American criticism prompted a bitter protest from President Khan. He accused the United States of being hypocritical in its sudden concern over Pakistani nuclear intentions. “Now, after the Afghanistan war is over, you are squeezing us,” Khan was quoted as telling Gates. “You’re telling me what’s going on. Don’t you also know what’s going on in India and Israel? It’s a double standard.” Pakistan’s fear, as it was said to have been relayed to Gates, was that a full-scale ground assault by India could dismember the nation within two weeks, as had happened in 1971, leaving the leadership with, in the Pakistani officer’s words, “no option but to go nuclear.” In the Pakistani officer’s account, Khan and Beg did seek, with no success, a United States commitment to reward Pakistan for not using nuclear weapons in case of all-out war with India, by providing a full-scale airlift of conventional arms and ammunition, as had been done to bolster Israel in the early days of the 1973 Mideast war.

Both sides agreed, however, that the defining moment of the meeting came when Gates told the Pakistani leadership, “You guys are going to have to stop supporting terrorism in Kashmir.” President Khan then gave the American envoy what he needed for his trip to New Delhi: he assured him that the Pakistani training camps for Kashmiri insurgents would be shut down.

The Pakistani concession turned out to be essential, for as Gates left Islamabad events in Kashmir took a drastic turn for the worse. On May 21, 1990, Maulvi Muhammad Farooq, the senior Muslim cleric of Kashmir, was assassinated in his home by unidentified gunmen. Farooq was a moderate, who had campaigned peacefully for an independent Kashmir—an option rejected by both Pakistan and India. More than a hundred thousand mourners took to the streets of Srinagar, and Indian security forces, panicked, opened fire. As many as a hundred people were killed, and tension rose among both the Pakistani and the Indian forces on the border. Perhaps because of this renewed tension, Gates emerged from his May 21st meetings with the leadership in New Delhi with what he termed “a strong sense of their worry about the nation’s fragility. They were worried about Pakistani meddling in Kashmir, and worried about the future of their secular state,” he told me. “They talked of their concern about Hindu fundamentalism juxtaposed with Islamic fundamentalism—and the dangers both of those movements posed.”

Gates said his message to the Indians—presented in separate meetings with top officials—was essentially the same as the one he gave the Pakistanis: “You guys got to lay off each other.” In India’s case, that meant stopping its infiltration into the Pakistani border state of Sind, and taking immediate steps to improve the human-rights situation in Kashmir. From the Indian point of view, Gates said, “It was the first time an American had come out and treated Pakistan and India as equals.” The Indians responded with a significant concession: they agreed to let American military attachés of the United States Embassy go to the front in Kashmir and Rajasthan and see for themselves that no imminent invasion of Pakistan was in the works. The American attachés duly reported that the Indian units, including its vaunted Strike Corps, were in the process of closing down their exercises. That information was quickly relayed to the Pakistani leadership, and over the next few days both armies moved their troops away from the borders and both foreign ministries opened discussions on confidence-building measures. By the end of June, the crisis was over.

In its aftermath are the inevitable questions about Pakistan’s nuclear status—who knew what, and when? Oakley insisted, as did many former members of the Reagan and Bush Administrations, that the essential facts about Pakistan’s inevitable emergence as a nuclear power were known fully in the late nineteen-eighties to the United States Congress, whose members “acquiesced”—as Oakley put it—in continued foreign aid to Pakistan because of the Afghanistan war and the enormous personal popularity of Prime Minister Bhutto in the United States. Oakley’s point seemed to be that passive approval by Congress of bad policy somehow justified bad policy. In an interview, a former top-level Reagan Administration official went further and questioned the integrity of those members of Congress who publicly supported the Solarz and Pressler Amendments: “All this morality horseshit. We were caught in a dilemma, and I didn’t know how to solve it: there was no way to stop the Pakistanis. Pakistan had been fucking around for years on the bomb. They weren’t going to depend on us. What do you expect of the Pakistanis? All this talk about breaking the law—it’s just a morality play. Of course everybody in Congress knew. The Administration was carrying out a popularly based policy in Afghanistan. If we’d cut off the aid to Pakistan, would we have been able to withstand the political heat from Congress?”

Nonetheless, one intelligence expert who has closely followed the nuclear situation from Washington, when asked to reflect on the India-Pakistan crisis, said that it might be time for the senior policy people in the White House, the State Department, and the C.I.A. to learn that there are “people outside”—such as those perennially disregarded arms-control officials responsible for the nation’s nuclear-nonproliferation agenda—“who also understand the rules.” He went on, “This crisis demonstrates that the question of ‘Who out there can toss one of these?’ may be the truly fundamental question for the next generation.”

The expert’s fears have no reverberations, for the near nuclear war in South Asia has remained an unknown event—a crisis that wasn’t a crisis—with no lessons learned. In the fall of 1990, after Ms. Bhutto’s ouster, senior officials of the State Department handed Sahibzada Yaqub Khan, the Pakistani Foreign Minister, a letter demanding that his nation destroy the nuclear-bomb cores known to be in existence. The American solution to the potential for a nuclear holocaust was to try to put the genie back in the bottle, as had been done throughout the nineteen-eighties.

There were other signs of business as usual. President Bush continued his policy of invoking the Pressler Amendment and not certifying Pakistan to be nuclear-free in 1991 and 1992, as he had in 1990, but he also permitted Pakistan to buy American-made arms from commercial firms, thus nullifying the impact of the law. Pakistan and India, while still refusing international inspection of suspected weapons facilities, continue today to publicly deny, as they did throughout the nineteen-eighties, that they have nuclear weapons.

In a political-science honors thesis written in 1980, Richard Barlow, who was then a senior at Western Washington University, in Bellingham, Washington, analyzed the relationship between the collection of intelligence about Pakistan and the failure of American policymakers to stop Pakistan from going nuclear. The problem, he concluded, was not a lack of information about Pakistan’s intent, but “a clear-cut failure of policy perpetuated by the consumers of intelligence.” Correcting that failure would become his mission in government.

In 1985, after having worked two years for the Arms Control and Disarmament Agency, Barlow made it to the big leagues—he was hired as a proliferation-intelligence officer by the C.I.A. Barlow was smart, he was committed, and he had a tremendous capacity for absorbing information—traits found in the best analysts of the intelligence community. Barlow looked the part: slender, with hazel eyes, light-brown hair, and a movie star’s profile. And yet he also was very much unlike the public image of hard-nosed C.I.A. men; he was openly enthusiastic, and exuded an energetic reverence for his work and its responsibility. “He brought the agency a unique skill,” Richard Kerr, who was the C.I.A.’s deputy director for intelligence in 1985, acknowledged. “He was an investigative analyst. It was a different kind of skill from the kind we had historically used. We’re doing more of it now.”

Barlow quickly came to learn that there was solid information sitting in C.I.A. files that was not being acted upon. While rummaging through old State Department cables, he realized the significance of a case involving a California couple who had exported dozens of high-speed cathode-ray oscilloscopes and special cameras to Hong Kong, where they were picked up by Pakistani agents. Barlow knew that the devices were also widely used in nuclear-weapons development to calibrate the uniformity of compression on a nuclear core—the implosion that triggers chain reaction—from the explosives inside the warhead. There could be no other use, he reasoned, for the export of so many exquisitely engineered devices. He found an overlooked link between the purchase, which was shipped through Hong Kong, and the Pakistan Atomic Energy Commission, which, as the C.I.A. knew, was then playing a major role in the development of a nuclear bomb. There were consultations with Justice Department attorneys and the Commerce Department, whose agents were nervous but pleasantly surprised to get help from someone in the C.I.A. Barlow even personally briefed federal prosecutors on the technical aspects of the case, enabling them to make clear—without violating national security—that the high-speed oscilloscopes were of great importance to Pakistan’s nuclear-weapons work.

Even as Barlow began his digging, some senior State Department officials were worried that too much investigation would create what Barlow called “embarrassment for Pakistan and trigger the Solarz Amendment, which would cut off all aid.” Protecting the Afghanistan war had emerged as a major policy of the State Department’s Bureau of Near East and South Asia Affairs, which was responsible for Pakistani policy. It was also considered essential by the C.I.A.’s clandestine-operations directorate, whose officers were deeply involved in the “covert” war that, as everyone in Washington understood, wasn’t covert at all.

Nonetheless, at a high-level meeting in mid-1986 Barlow was designated the C.I.A.’s delegate to the Nuclear Export Violations Working Group (N.E.V.W.G.), a newly formed top-secret panel that brought together the policy, law-enforcement, and intelligence communities in an effort to stop illegal American exports to Pakistan and other non-nuclear nations. Barlow took his work seriously, and volunteered—with the approval of Fred McGoldrick, the director of the State Department’s Office of Nonproliferation and Export Policy, and the chairman of the N.E.V.W.G.—to help set up an undercover operation in an effort to stop yet another planned Pakistani purchase of nuclear-related materials. The State Department’s Near East Bureau was not told of the planned operation, for fear that the officers there would tip off the Pakistanis, as they had done in the past, by sending a formal diplomatic protest (known as a démarche) to the Pakistani government. “We were sure they’d manage to screw it up,” one N.E.V.W.G. member said of the Near East Bureau. “They were more guilty of clientitis than anyone.” Barlow and his law-enforcement colleagues were convinced that the State Department’s intent in sending the démarches, which were inevitably ignored, was simply to signal Pakistan that the cops in America knew what was going on.

McGoldrick, who is still chairman of the working group on nuclear-export violations, agreed to the delay in informing his colleagues in the Near East Bureau of the pending operation. “The only thing we had going for us”—he and other working-group members—“was the Pressler and Solarz Amendments,” McGoldrick said. “The absolute imperative in this building was to get the Russians out of Afghanistan. I can’t quarrel with that. But it has to be done in accordance with the law. There were many more reports of covert procurement”—of nuclear-related goods—“in the case of Pakistan than with Iraq. For all the hoopla, Iraq didn’t take all that much out of here.”

As McGoldrick’s panel reached more deeply inside the Pakistani operations in the United States, the atmosphere grew tense. At one point, Barlow recalled, the C.I.A.’s Dick Kerr summoned senior State Department officials to a meeting and initiated a pointed discussion about the steady flow of démarches and highly classified internal American intelligence analyses between Washington and Islamabad.

In his interview with me, Kerr acknowledged that the Near East Bureau had been extremely active, just as Barlow and others alleged. “What they were doing it for was to persuade the Pakistanis to stop,” he said. The Near East Bureau “was trying to change Pakistan’s behavior,” he went on. “They didn’t want the relationship to break.”

Others in the United States government had a less benign view of the State Department’s approach to Pakistan. “We still have cases on Pakistan and we still don’t tell State about it,” a senior Customs Service official told me recently, in anger. “The State Department constituted a security problem for us. One of the things about being a dumb gumshoe is that we didn’t have to worry about foreign policy. The law is the law—and the law is the only thing I go by.” The Customs Service official, who is still in the government, praised Barlow for “telling it the way it was at a time State was jerking me around,” and said, “The guy was as topnotch as any I met. He laid it on the line.”

Some superiors and peers, on the other hand, viewed Barlow as single-minded and difficult to control. “The most dangerous problem we have in our business is zealotry—people who have a mission,” Richard Kerr said. “Richard bordered on that. When others disagreed with his analysis, he made it a matter of integrity, as opposed to a matter of judgment.”

The N.E.V.W.G.’s undercover operation led to a highly publicized arrest, in Pennsylvania, in July of 1987, of a Pakistani-born Canadian named Arshad Z. Pervez: he was seized while attempting to buy twenty-five tons of a specially strengthened steel for use—as Barlow knew—in Pakistan’s uranium-enrichment program. There was now little doubt about Pakistan’s nuclear intentions. The arrest created a furor among those members of Congress committed to nonproliferation, including Senator Glenn and Representative Solarz. Congress, obviously more concerned about safeguarding aid for the Afghanistan war than about stopping the spread of nuclear weapons in South Asia, nonetheless approved a foreign-aid package of four hundred and eighty million dollars for Pakistan in December. A dismayed John Glenn subsequently told the Times that the threat of nuclear proliferation “is a far greater danger to the world than being afraid to cut off the flow of aid to Afghanistan,” and added, “It’s the short-term versus the long-term.” In January, six months after the arrest of Pervez, President Reagan finally invoked the Solarz Amendment, as he was compelled to do under the law, and then immediately waived its provisions, clearing the way for the American aid. The President was telling Pakistan that it could have its money—and its bomb.

Barlow received official praise for his work on the Pervez case, and spent hundreds of hours coördinating the inevitable array of disputes between the intelligence, foreign-policy, and law-enforce- ment communities. He was offered (and rejected) a job as a special agent by the Customs Service and given a note of personal commendation, relayed to Kerr by Abraham Sofaer, the State Department’s legal adviser. By mid-1987, Barlow had emerged as the government’s leading expert on illegal Pakistani procurement.

He produced an extensive analysis for the Department of Justice and the Customs Service on the hundreds of documents found in Pervez’s apartment. He was brought into a sensitive case involving possibly illegal State Department approval of licenses to the Pakistani Embassy in Washington for equipment whose export had previously been denied—for nuclear-proliferation reasons—by the Commerce Department. There were many other cases in which high-tech goods, approved for export by Washington, turned out to have a significant application to Pakistan’s nuclear program.

After Pervez’s arrest, Representative Solarz, who was obviously concerned about the violation of his amendment, requested a top-secret briefing by the C.I.A. and the State Department on Pakistani bomb procurement. Barlow was assigned to accompany David W. Einsel, the C.I.A.’s senior nonproliferation officer, to the briefing. It would be Barlow’s first appearance before Congress.

Before going, Barlow was instructed by Charles Burke, his immediate superior in the C.I.A., to tell the truth, if asked. Barlow recalled having no anxiety about the briefing. “I assumed Congress had been briefed by State on the extent of Pakistani procurement activity in the United States,” he said. But that briefing marked the beginning of his nightmare. “I wish I had called in sick,” he told me.

Barlow had worked closely with Einsel, a retired Army major general, who had been brought to the agency in 1985 by William Casey, and he knew that Einsel had seen the highly classified reports and analyses that were available on Pakistan’s extensive nuclear-related procurement activities inside the United States. “I knew his state of knowledge,” Barlow recalled. “I knew exactly what he knew. Einsel’s testimony was highly evasive, and deliberately so.”

Barlow was disturbed because Solarz made it clear that he had a broad definition of what constituted a violation of his amendment. “I don’t care if the Pakistanis ordered it”—American-made equipment—“by mail order from abroad, or if they went to Germany to buy it,” Barlow quoted Solarz as saying. “As far as I’m concerned, that’s a violation.” At one point, Barlow said, Solarz turned to him and asked a direct question: “Have there been other cases?” “I said to myself, ‘You don’t know?’ He started interrogating me like a prosecutor. I was told to tell the truth and did.”

Barlow testified that, indeed, the C.I.A. did know of “scores” of other Pakistani attempts to violate American export law. He also tried to tell the subcommittee something else it did not know: that the United States government had evidence directly tying one of Pervez’s collaborators, a retired Pakistani general, to Pakistan’s government and its nuclear-weapons project. At that point, Barlow said, “Einsel shut me up and I was never allowed to get into specifics.”

A senior government official who was outside the hearing room (he had not been cleared to attend the briefing) recalled what happened next: “I was outside when Robert Peck”—the Deputy Assistant Secretary of State for Near East Affairs, who died last year—“rushed out and said to me, ‘You’ve got to testify. That s.o.b. Barlow is telling them all sorts of things, and you’ve got to straighten it out.’ I went in to the hearing, but I didn’t know what to say.” He had credibility with Solarz and his subcommittee, the official acknowledged. Solarz declared that he was “appalled” at the extent of the intelligence reports. “I told him that not all intelligence reports are accurate,” the official said. He understood that his function was to somehow expunge Barlow’s testimony—to smooth it over—and he did his job. “I cooled it out,” he said. The official made it clear that it was not his proudest moment in government service: “I didn’t know what I was getting into.”

Einsel, who retired to Ohio in 1989, characterized as “nuts” Barlow’s allegation that he withheld information from Congress. “There is always information that is in process, that’s ongoing,” the former general said in a telephone interview. “You generally give Congress evaluated answers—not all the rumors that are flying. That’s where he and I get off on a tangent. The man was obsessed with passing on his latest rumors to the Hill, and whether they proved correct didn’t matter to him. That’s the problem.”

After his testimony, Barlow became a marked man in some offices inside the Central Intelligence Agency. “It upset me very much that I was forced to be involved in something like this,” he told me later. “I had no idea that Congress had not been informed of the cases.” There was an almost immediate payback. “Einsel went crazy,” Barlow said. “I was told that my personal behavior at the hearing had been unprofessional. I was accused of being unpatriotic and almost scuttling the Afghanistan program. I was viewed as being disloyal.”

Within days, the C.I.A.’s nonproliferation office was being pressured by Einsel and the State Department’s Near East Bureau to fire Barlow or, at the least, remove him from the Pakistan account. Six weeks later, Barlow’s formal job description was rewritten to remove him from any responsibility for Pakistan. Richard Kerr intervened and told him to “continue doing what I’m doing,” Barlow recalled. Not long after that, however, he was again offered a job with the Customs Service. A year later, he accepted the job and resigned from the C.I.A., with mixed feelings. “I still think it’s the best agency in the government,” he said.

In early 1989, Barlow was forced to change jobs again—this time for personal reasons—and joined the Pentagon’s Office of Non-Proliferation Policy. There were renewed contacts with the N.E.V.W.G., whose anti-Pakistani activities had dropped off precipitously the year before, and renewed contacts with law-enforcement officials. Barlow was once again chasing Pakistani nuclear procurement.

The Office supervisor, Gerald G. Oplinger, who was a longtime government expert on nonproliferation, also had his bureaucratic problems with Barlow. “He was smart, with a photographic memory,” Oplinger said. “A very persistent guy who doesn’t kowtow and who doesn’t like to do scutwork.” There was also no question that Barlow was knowledgeable and capable, and Oplinger, before retiring from the government in April, recommended him for promotion. Oplinger knew that Barlow had enemies from his days in the C.I.A., the bitterest of whom were those members of the clandestine service who had been involved in the Afghanistan war.

Barlow went about his business without incident, so he thought, until July of 1989. Then he learned that the United States government was once again distorting intelligence on Pakistan’s nuclear capability. He had prepared a comprehensive analysis on Pakistan’s nuclear capability for Secretary of Defense Richard Cheney and other senior officials. The paper cited what Barlow and many others in the intelligence community understood to be persuasive data showing that the F-16 aircraft previously delivered to Pakistan had been modified to deliver nuclear weapons. Barlow’s paper was complemented by a separate Defense Intelligence Agency study, which reached the same conclusion.

All this gave the government a big problem. President Bush, after his June meeting with Benazir Bhutto, had agreed to sell sixty of the aircraft, a sale of a billion six hundred million dollars that the Pentagon—and Pakistan—badly wanted. But that sale hinged on the Pentagon’s assurances to Solarz’s subcommittee that the Pakistani government would not modify the F-16s for nuclear delivery. Furthermore, Barlow was involved just at that time with investigating four major criminal cases involving senior officers of the Pakistani Army who had attempted to make illegal purchases in the United States or abroad of American-made nuclear-related materials, including highly enriched uranium. One of the cases involved evidence showing that Pakistan was attempting to obtain dual-use items for its nuclear program by claiming that the materials were to be used for its F-16 fleet. The State Department’s Near East Bureau had learned of Barlow’s activity only a few days before.

On August 2, 1989, Arthur Hughes, a newly appointed Deputy Assistant Secretary of Defense, testified before the Solarz subcommittee that the F-16s to be sold to Pakistan would be stripped of nuclear wiring before delivery. To deliver a nuclear bomb, therefore, Hughes said, “it first would be necessary to replace the entire wiring package of the aircraft.” Hughes assured Solarz, in response to a question, that the F-16s to be sold to Pakistan were not nuclear-capable unless the nuclear wiring, or some modification of it, was replaced.

Barlow knew that the Hughes testimony was totally contrary to the analyses that had previously been prepared. The fact that the Pakistani Air Force had practiced low-level F-16 delivery of nuclear weapons was widely known throughout the American intelligence community. The primitive delivery system required very little in the way of electronics or special wiring. Barlow informed his superiors, including Gerald Brubaker, his immediate supervisor, of his problems with the testimony, and urged that it be corrected.

Two days later, Brubaker called Barlow into his office and, with no warning, handed him a letter of termination. He stood accused, as he did not know at the time, of being a national-security risk to the United States. He was stripped of all his classified clearances and given three weeks to clear out of his office. Barlow decided to fight the dismissal. He spent the next eighteen months assigned to a Defense Department personnel pool, under surveillance by Pentagon security officers. He learned later that the “national-security risk,” according to classified documents released to him under the Freedom of Information Act, was the fear of Brubaker and his superiors that Barlow would go, without authorization, to Representative Solarz and tell him the truth. “Would I have done it?” Barlow recently said, in astonishment. “Hell, no. I was trained to work within the system.” Nonetheless, over the eighteen months, Pentagon investigators, on the pretext that they were dealing with a national-security threat, began looking into his personal life, his finances, his taxes, and false allegations that he had been fired from the C.I.A. and from the Customs Service.

Barlow has repeatedly stated that Gerald Brubaker and other officials made a concerted effort in 1989—as he was looking for other government and private-sector jobs—to impugn his reputation by spreading lies about his general reliability. Brubaker and the other officials named by Barlow categorically deny those charges. Barlow remains convinced, he said, that his firing and other adverse actions against him stemmed from his refusal to stop working with federal law-enforcement agents to investigate Pakistan’s illegal nuclear-related purchases in the United States.

Brubaker, in a telephone interview, maintained that Barlow’s dismissal resulted solely from his “poor performance,” and not from a government-wide conspiracy to do him in because of his stand on Pakistan. “He is incompetent and insubordinate—period,” Brubaker said. “He didn’t do what he was told, and he has parlayed this into a bigger issue.”

Two documents obtained by Barlow under the Freedom of Information Act challenge Brubaker’s assertions. An internal Pentagon memorandum shows that Brubaker—whose personal antagonism toward Barlow was widely known—did not complain of insubordination at the time of the firing but instead “expressed concern” that Barlow would provide “classified information” to Congress. A handwritten notation made four days later shows that Barlow’s top-secret clearances were to be suspended because he was a potential “national security” threat.

In May of 1990, after an eight-month investigation, the Pentagon determined that the security charges against Barlow were false. In a formal memorandum, Barlow was told that “after thorough investigation . . . any question of your trustworthiness for access to sensitive information was resolved in a manner completely favorable to you.” His top-secret security clearances were reinstated, but the Pentagon did not restore his clearances to compartmentalized intelligence, without which no intelligence officer can do his job. Barlow remained on duty in the personnel pool, arranging civic luncheons and obtaining computers for public schools in Washington. In August, the Defense Department offered Barlow and his attorneys a settlement of twenty-six thousand dollars. Acceptance would mean that Barlow could not take his complaints to Congress or to any executive agency. There was an impasse. Donald Mancuso, the director of the Defense Criminal Investigative Service, then offered Barlow a chance to become a special agent, or a consultant, if he agreed to disclose any potential criminal wrongdoing about which he had firsthand knowledge. He agreed to do so—reluctantly, he said—and thus became, in the eyes of many of his former colleagues, a whistleblower. He told Mancuso everything he knew about the C.I.A., the State Department, and the Pentagon coverup of Pakistan’s nuclear-purchasing activities in the United States. Barlow was interrogated for two weeks, and his allegations were referred to the Inspectors General of the Department of Defense, the State Department, and the Central Intelligence Agency.

Not surprisingly, all the initial I.G. reports absolved their agencies of any criminal wrongdoing, but pressure to reopen the case is currently being put on the Pentagon by Sherman Funk, the State Department I.G., as well as by Senator Bingaman, of New Mexico. Further doubts about the efficacy of the Pentagon’s inquiry have been raised in a sworn affidavit from Gerald Oplinger obtained by Barlow’s attorneys, Paul Warnke and James Duff. In its I.G. report, the Pentagon had accused Oplinger of deliberately inflating his annual evaluation of Barlow in order to avoid “an unpleasant personnel situation.” In his affidavit, Oplinger angrily responded, “This is a serious charge, and devoid of merit.” He further noted that one of the Pentagon I.G.’s reports, prepared by Navy Captain O. M. Horstman, was said to be based on an interview with him. “The fact is,” Oplinger wrote, “I have never been interviewed by Horstman nor by anyone identified to me” as representing Horstman’s office.

In his interview with me, Richard Kerr acknowledged that Barlow’s essential allegation was true: David Einsel did not give “full briefings to the Hill” on Pakistan’s procurement activities. “I think Einsel fretted too much over how far to go, and he did not know all the details,” Kerr said. He added that Barlow’s “knowledge on Pakistan was greater than anyone else’s in the agency,” and that “there was an inability of Einsel to connect to others on the staff.” The former C.I.A. deputy director added, generously, “I don’t think it was an attempt to conceal the facts, but a difference of opinion over what to share.”

Such talk of absolution does little to soothe Barlow, who has suffered through what he calls a “Kafkaesque hell” since August of 1989. “There was no legitimate national-security issue involved” in his dismissal, Barlow said, in explaining why he was continuing to press the issue. “If you let the Pentagon get away with this, any time they want to destroy someone who is threatening their interest all they have to do is to instruct his superior to tell Security that the person has an alleged intent to engage in classified whistleblowing, and unleash the machinery.”

No judicial proceeding can repay Richard Barlow for his lost years, his shattered career, and his lost self-esteem. And Congress can do little today about testimony that wasn’t presented in 1987 or 1989—or about testimony that was manipulated.

Stephen Solarz, who lost his congressional seat last November, after nine terms, acknowledged that he and others who cared about nonproliferation had been constantly trying to balance that concern with a desire to support President Reagan in the Afghanistan war. “There were legitimate concerns that the Afghan war might spill over to Pakistan, and I felt we needed to give the President flexibility,” Solarz told me recently. “I didn’t want us to be in a worst-case scenario in case the Soviets moved across the border. I thought I was being responsible at the time.” What he could not understand, he added, was how irresponsible the Reagan Administration had been. “If what Barlow says is true, this would have been a major scandal of Iran-Contra proportions, and the officials involved would have had to resign,” Solarz said. “We’re not dealing with minor matters. Stopping the spread of nuclear weapons is one of the major foreign-policy issues of the nation—not to mention the law of the land.”

And not to mention the fact that in May of 1990 some of the bombs that were not stopped in the Reagan and Bush years were almost put to use. And that there are many more still out there. And that India, frightened and besieged today by terrorist bombings and growing Hindu fundamentalism, remains always willing to fix the blame for its troubles on Pakistan. ♦

Courtesy: The New Yorker [Issue of March March 1993]


No comments: