First of Six Articles

Sigfried Hecker watched contentedly as President Clinton pressed his face to the window of their slow-moving limousine. All along East Road, on fences and porches, sidewalks and rooftops, hundreds of townspeople in Los Alamos, N.M., had turned out to greet Clinton on his first trip to the birthplace of the atomic bomb.

"Did you see all those people?" Clinton asked his host in delight.

Impressing politicians was part of Sig Hecker's job as director of the Los Alamos National Laboratory, the country's preeminent facility for nuclear weapons testing. Clinton's visit on May 17, 1993, came at a remarkable juncture in the 50-year history of nuclear weapons. The breakup of the Soviet Union had transformed the nuclear landscape, reopening debates that had been dormant for many years. Soon Clinton would face a series of decisions, including one of profound importance to Los Alamos and the rest of the world: whether to ban testing of American nuclear weapons.

Clinton had come to New Mexico with Secretary of Energy Hazel R. O'Leary, who was preparing a recommendation on nuclear testing for the president. Hecker already knew that O'Leary was skeptical about new tests. He would now have a chance to tell Clinton what he and nearly everyone else at his lab believed: Testing made the United States and the world a safer place.

As their car cruised along a stark, craggy mesa toward the lab's main campus, Clinton spotted protesters waving a hand-lettered banner: "Stop Nuclear Testing Now."

Clinton turned to Hecker and asked, "How do your people up here feel about that issue?"

Hecker chose his words carefully, acutely aware of O'Leary's presence. "People here feel very strongly about that," he later recalled saying.

Testing, he said, was the best scientific tool to maintain a reliable nuclear deterrent in an unstable world. He knew some argued just as strongly that new tests would anger other countries, even drive them to pursue their own nuclear bombs. Clinton and O'Leary would have to decide how to strike a balance, he said.

Hecker had framed a larger dilemma confronting the post-Cold War world. For decades nuclear weapons have defined the global military balance. Today, generals, nuclear weapon scientists and national security policy-makers are asking the same questions their counterparts did after nuclear arms were invented in the 1940s: What is the best way to deter a nuclear war? What role do nuclear weapons play in U.S. national security? Is it possible to do away with nuclear weapons? Can the emergence of new nuclear weapon states and threats be prevented?

Next week, a historic conference opens in New York that will help define the nuclear order of the 21st century. Delegates from more than 170 nations will meet to decide the fate of the nuclear Non-Proliferation Treaty (NPT). That 25-year-old treaty limited the world's nuclear powers to five -- the United States, Russia, China, Britain and France -- in exchange for their pledge to work toward the eventual elimination of all nuclear arms, including their own. It is almost certain that the treaty will be renewed in some form. The United States has taken the position that anything short of an indefinite extension will undermine the stability created by the treaty; many Third World nations are resisting an open-ended extension, arguing that this would eliminate any pressure on the five powers to give up their weapons.

The last two years have been a time of intense debate and decision-making, both here and abroad. Pentagon planners have sought new weapons to defend against enemy nations or terrorists armed with a nuclear bomb, while some U.S. diplomats have denounced that proposal. Military leaders have scrambled to update their nuclear targeting to address new dangers. But during a confrontation last year with North Korea over its nuclear program, Pentagon leaders realized they did not know how to deter a small and militant country that might use nuclear bombs. Meanwhile, the United States, Russia and China have opened new, sometimes secret channels of contact in search of a new nuclear balance.

This series of articles, based on more than 200 interviews, will explore these changes and events. It will describe, too, some of the people making the decisions -- their thinking, their visions and their conflicts.

In January 1993, Clinton and his advisers had arrived in Washington without fixed ideas about what sort of nuclear future to pursue. Though it is hard to generalize about nuclear weapons ideology in this time of its reinvention, two rough groupings appeared in the administration's ranks.

Traditionalists appointed by Clinton to key national security posts believed that a smaller but nonetheless robust arsenal of nuclear weapons would remain essential for the foreseeable future to deter nuclear-armed foes old and new, to discourage regional powers armed with chemical or biological weapons and to assure longtime allies in Europe and Asia that the United States would remain a force for global stability.

Other Clinton appointees were convinced that U.S. nuclear weapons had become more dangerous than useful. They worried that if the United States continued to modernize its nuclear stockpile, other countries might seek their own nuclear forces to offset U.S. dominance. The wiser alternative, they had concluded, was to lead the world toward a marginalization of nuclear weapons -- even, in time, toward a "zero option," a total elimination. For years, the zero option had existed primarily in the heads of anti-nuclear activists and in the preamble of the NPT. That the zero option had any legitimacy at all in the internal debates of the new administration was a sign of how much things had changed.

In the short term, nuclear testing loomed as a crucial issue. Except for a moratorium from 1958 to 1962, the United States had been testing its nuclear weapons for nearly 50 years. Testing allowed scientists to experiment with new kinds of nuclear weapons and to assure that older ones still worked.

By the time Clinton visited Los Alamos, conflict within the administration was intensifying. On both sides, the debate had a searing, emotional aspect. Nuclear weapons had a way of getting under people's skins.

In the early spring of 1993, Tom Graham tracked back and forth from his office to the West Wing of the White House, aware that he stood nearly alone in the Clinton administration in his belief that the United States should no longer test nuclear weapons.

Graham knew well the arcana of nuclear bombs. For 25 years he had succumbed to what he called the "fatal attraction" of nuclear arms control. Graham observed that most people who got into arms control never got out. It had happened to him. He was 59 now, at the end of a long career as a treaty negotiator at the Arms Control and Disarmament Agency (ACDA). During the Reagan years, Graham had taken heat from fellow Republicans for being too interested in arms control and not tough enough on the Soviets. But he had weathered those storms and had even survived the changeover to a Democratic administration.

The Princeton-educated son of a Kentucky ward politician and horse racing enthusiast, Graham harbored something of a gambler's spirit behind his detached, vaguely aristocratic demeanor. He was preparing now to put his convictions on the line.

Graham's opportunity came when Clinton's national security adviser, Anthony Lake, circulated a Presidential Review Directive appointing an interagency group to take up the issue of nuclear testing. Candidate Clinton had pledged to pursue a worldwide test ban treaty. But negotiations had not been organized. A treaty looked years off, at best.

In the meantime, Congress had authorized up to 12 U.S. underground tests at the Nevada Test Site, provided they enhanced the "safety and reliability" of the U.S. arsenal. The legislation also permitted three British tests in Nevada. Russia and France had adopted temporary test bans. Should the United States join them and push for a permanent global ban?

In April 1993, officials from the State Department, Energy, Defense, the Joint Chiefs of Staff, the Central Intelligence Agency and the National Security Council met at the White House to debate the issue. Graham was there as acting director of ACDA.

Secretary of State Warren Christopher reminded the interagency group that winning permission from Congress for new tests had been tough. A deal had been cut to allow them. The Clinton administration should now go ahead, Christopher said.

In his soft Kentucky drawl, Graham suggested that continued testing would jeopardize the NPT vote. The NPT was too important to risk, Graham said. Therefore, testing should stop. Now.

The power of Graham's argument lay in his dedication to the NPT itself. Among arms control documents, the NPT boasted a rare distinction: In its 25-year life, it had shaped the world's balance of power.

The invention of atomic weapons by Los Alamos scientists during World War II began a race among the world's industrialized powers to acquire nuclear technology -- and with it, the ability to make nuclear arms. By the 1960s, the rise of civilian nuclear power industries across Europe and in Japan meant that a dozen or more countries had the makings of a nuclear bomb within their grasp. President John F. Kennedy worried aloud that within 20 years, the world might see two dozen nuclear-weapon states, each with missiles pointed at its enemies in a global gridlock of nuclear confrontation.

The NPT emerged as a statement against such instability. The treaty offered a bargain: Countries that had tested nuclear weapons by 1968 could keep them, but all others would forswear them. International inspections would ensure the nuclear have-nots kept their pledge. Few worried in those days about militant states such as today's Iran, Iraq or Libya. The main fears then were of a nuclear-armed Germany, Japan, Sweden, Italy, even Switzerland.

Europe's and Asia's rising economic powers were not happy about being second-class nuclear citizens under the NPT. They did not particularly want nuclear weapons -- they were costly and dangerous to acquire and maintain -- but they did not want to close off their options forever either. As the treaty moved toward signature, they extracted painful concessions in return for their support. The United States and the Soviet Union had to promise to seek a worldwide nuclear test ban and pursue "negotiations in good faith" for the total elimination of nuclear arms. The superpowers also had to accept that the treaty would last only 25 years and would require a new vote for renewal. The treaty entered into force in 1970, thus the extension vote in 1995.

Graham saw the NPT regime as a cornerstone of U.S. and global security. The treaty was not perfect -- Israel, India and Pakistan had always refused to join, and countries such as Iraq and North Korea sometimes sought to evade the NPT from within. But for the vast majority of nations, the NPT had become the key framework for nuclear order. In Graham's view, winning indefinite extension of the NPT should be the Clinton administration's number-one arms control priority.

The task did not look easy. Graham saw no trouble from the European countries or Japan; they had come to appreciate the bargain and now backed its indefinite extension. But many Third World nations, whose voices had been muffled the first time around, wanted some tangible proof that the five weapon states had meant what they said about giving up nuclear arms. For example, the NPT had declared that nuclear testing should stop. It hadn't. Was the United States now ready to lead the way by voluntarily ending its own tests? Judging by the reaction to Graham's arguments within the administration, it seemed unlikely.

For weeks, Hazel O'Leary watched Graham fight alone in the administration's testing debate. At a meeting in April, the interagency group had been ready to ignore Graham and go ahead with a round of new tests. They seemed to take it for granted that O'Leary, who supervised the labs, would give her support. Every other energy secretary had done so.

But after listening to Graham and looking over the lab's proposals, O'Leary had announced: "I'm new on this job. I want to review what the labs have been recommending."

Lake and other senior officials at the meeting were unhappy with O'Leary. They stormed past her as she stood in the White House driveway.

O'Leary decided she needed more information. The United States had conducted more than 1,000 tests during the Cold War. What did the labs hope to learn from new tests? She wanted to hear from both sides. She summoned Hecker and other important figures in the nuclear community to a secure, classified vault in the basement of the Department of Energy's Forrestal Building for two days of debate.

On May 18, 1993, the day after Clinton's visit to Los Alamos, Hecker arrived with his counterparts from the Livermore and Sandia nuclear weapons labs in California. Also present were two longtime advocates of nuclear disarmament -- Frank Von Hippel, a Princeton scientist recently appointed as a White House adviser, and Ray Kidder, a nuclear weapons scientist at Livermore. "It was like a Quaker meeting," Von Hippel recalled. "Anybody could say what was on his or her mind."

Hecker knew what he wanted to say. He believed in preserving a strong nuclear deterrent. Testing was the best way to do that. If the United States did not keep a nuclear shield over Europe and Asia for a while longer, he argued, countries such as Germany might be tempted to protect themselves by going nuclear. Hecker knew the dangers of German militarism: He was born in Poland 50 years earlier in the midst of World War II. His Austrian-born father was a German army conscript who disappeared on the Russian front soon after Hecker's birth.

Anti-nuclear activists saw Hecker as an affable but tragic figure, a man without a mission who was willing to say anything to keep nuclear weapons and his lab alive. Even a majority of Los Alamos scientists, in employee surveys, said they did not think Hecker had a vision for the lab's future. But Hecker thought his critics were more imprisoned by Cold War thinking than he was. He had a reputation among his management team as a forward thinker. He could see from the lab's shrinking budgets that its mission was changing. He wanted Los Alamos to adapt by converting aspects of its weapons work to aid U.S. industry.

In the basement vault, Hecker and his fellow lab directors warned O'Leary that the United States faced new nuclear threats in the post-Cold War world. Livermore's director, John Nuckolls, particularly worried that terrorists might get hold of a loose nuke from Russia; he declared that scientists could design a test that might provide insight on how to disarm a terrorist bomb.

There were strong scientific reasons for continued testing as well, Hecker and the other lab directors said. It was likely, they argued, that thousands of U.S. nuclear weapons would remain deployed for decades. Scientists at Los Alamos and Livermore didn't know what might happen to nuclear bombs that sat idle and untested for so long. Nobody had ever tried it before.

Nuckolls had brought along a dramatic prop to drive home the point that nuclear bombs were complex creations: a full-size MX warhead -- without, of course, any fissile material. Cut open to reveal its inner workings, it sat on a nearby table throughout the two days of meetings.

Hecker made clear, however, that he was not suggesting that the United States keep testing nuclear weapons indiscriminately. Congress had already ensured that would not happen, he said, by putting limits on the tests.

Hecker acknowledged that some countries might object to continued U.S. testing as a violation of the spirit of nonproliferation. But when you got down to the hard cases, Hecker said, what does Iraqi President Saddam Hussein care about whether the United States and Russia conduct tests? "As far as I'm concerned, he doesn't care at all," Hecker recalled saying. "Because that's not his sphere of influence. He's worried about the people around him. Whether we test or not makes no difference."

Von Hippel and Kidder responded that Hecker might be right about Saddam, but there were many, many other countries watching U.S. nuclear weapons policy. Allies such as France had already halted testing; it was important to support them, they said.

Besides, Von Hippel argued, some tests proposed by Hecker and his colleagues would not provide useful information. Other tests, he said, could be simulated with computers -- without full-scale nuclear explosions.

Whatever she decided, O'Leary told them, she ought to be able to explain her choice clearly to ordinary people. Her grandmother should be able to understand her decision, O'Leary said. Her grandmother would understand why scientists might want to conduct new tests to learn more about how to disarm terrorist bombs. But her grandmother would not understand why tests were necessary to make idle American weapons on subs and missiles more "reliable."

The lab directors were "sort of stunned when she introduced this grandmother thing," Nuckolls recalled later. Nonetheless, Hecker thought that O'Leary had run a fair debate.

O'Leary's own views were still evolving. She thought Washington had an obligation to lead the world toward a future free from the threat of nuclear arms. But she did not want to justify a testing ban strictly on moral grounds. She needed to demonstrate that her thinking was informed and reasoned.

After listening to the discussion, O'Leary decided that the scientific value of more testing was outweighed by the broader political and diplomatic benefits of ending the tests. She told Lake by telephone that she had made up her mind. She would recommend that Clinton ban any further U.S. nuclear testing.

The interagency group, including several Cabinet officers, reassembled on June 1, 1993. The atmosphere was electric, tense, according to people familiar with the meeting. Lake asked who wanted to speak against testing. Graham delivered the same speech he had been making all year. O'Leary also registered her opposition, and so did Clinton's science adviser, Jack Gibbons. Christopher, Defense Secretary Les Aspin and Joint Chiefs of Staff Chairman Gen. Colin L. Powell favored more tests.

But Powell seemed to waver, those familiar with the meeting said. It was the secretary of energy's job to decide if nuclear tests were necessary, Powell said. If O'Leary doesn't think we need tests anymore, then why test? he asked.

In truth, the uniformed military that Powell represented showed signs of ambivalence about the future of nuclear weapons. Some generals wondered why the United States should spend so much money on weapons they almost certainly would never use. The proposed tests would help develop new safety features for the Navy's W88 nuclear warhead. But deployment would cost the Navy about $2 billion or more. By June, Navy officials were saying privately that they were not sure they wanted to spend that kind of money on a new nuclear warhead -- with or without testing.

The interagency group was split. After the meeting, Lake asked his staff to sample the mood on Capitol Hill, since the administration was required to justify to Congress a decision to go ahead with new tests. Lake's staff sprayed telephone calls to a dozen offices, including members of the armed services committees in both houses. The word came back from Sen. Sam Nunn (D-Ga.) and others: Opponents might seek to block new tests, prompting a tough fight. That could hurt the administration's chance to win support for its ambitious domestic program, which was then pending before Congress.

Lake forwarded O'Leary's recommendation to Clinton as the best option. Clinton decided to declare that the United States would halt all testing until September 1994. This moratorium would continue indefinitely, as long as no other nation conducted any tests.

The president chose the July 4 weekend to make his historic announcement. His words -- measured, careful, hedged -- reflected the divisions within the administration on the issue.

"After a thorough review," he said in a July 3 radio address from the White House, "my administration has decided that the nuclear weapons in the United States arsenal are safe and reliable. Additional nuclear tests could help us prepare for a test ban and provide for some additional improvements in safety and reliability. However, the price we would pay in conducting those tests now -- by undercutting our own nonproliferation goals and ensuring that other nations would resume testing -- outweighs these benefits."

Having helped to achieve the moratorium on testing, Tom Graham wanted clear marching orders so he could begin the difficult job of lobbying for indefinite extension of the NPT. The orders came in the form of three Presidential Decision Directives on U.S. nuclear weapons policy.

The "PDDs," in White House parlance, pulled together months of thinking and debate inside the administration. Like most national security documents, they circulated to only a small circle of top officials; even the numbers in their titles were deemed secret. The documents bore Clinton's signature, but they were written by mid-level aides at the National Security Council.

Taken together, the directives -- PDD-11 in July, PDD-13 in September and PDD-15 in November -- set the goals and priorities of U.S. nuclear weapons policy under Clinton. They also provided language that gave comfort to both the marginalizers and the traditionalists.

The first directive, PDD-11, ratified the president's moratorium on nuclear testing. But it said that if another country tested nuclear weapons, the United States would prepare quickly to resume testing. The directive also said that Washington must preserve a strong, effective nuclear deterrent force and asked a new interagency group to develop a plan to do so without nuclear tests. At the same time, it laid out a plan to secure a global test ban treaty.

PDD-13 outlined U.S. nonproliferation policy. It said that stopping the spread of weapons of mass destruction had become a key national security mission for the United States after the Cold War. The directive named indefinite extension of the NPT as a central foreign policy goal. It also instructed the Pentagon to take the threat of nuclear proliferation into account in U.S. defense planning.

PDD-15 endorsed a specific program to preserve U.S. nuclear weapon capabilities in the years ahead, to maintain current nuclear forces and to hedge against unfavorable changes in global politics. The document mandated funding for a "stockpile stewardship" program, in which Los Alamos and other weapons labs would preserve for the foreseeable future the technical and intellectual ability to design and maintain nuclear weapons.

The directives had something for everyone. Graham could tell governments around the world that Clinton had endorsed the NPT regime as a cornerstone of U.S. foreign policy. Hecker, though disappointed by the testing decision, won a mandate for preserving a strong U.S. nuclear deterrent and overseeing the nuclear stockpile.

It soon became clear that the new directives should not be taken too literally. In September 1993, the administration learned that the Chinese government was about to test a nuclear weapon in that country's northwestern desert. The interagency group assembled as required under PDD-11, but agreed that the Chinese test posed no immediate threat to national security. There was no desire to replay the spring's emotional battles. They chose to ignore it.

Clinton's three directives proved in time to be like verses from the Bible: They were texts, but the texts could be interpreted in varied ways. The directives soon became disputed sources of authority in the still-evolving struggle over how best to protect the United States and the world from uncertain nuclear threats.

NEXT: How real is the threat?

Copyright 1995 The Washington Post By David B. Ottaway; Steve Coll

Ashton Carter, the 40-year-old whiz kid of President Clinton's Pentagon, sat in the audience at the National Academy of Sciences in Washington on Dec. 7, 1993, in the proud but nervous manner of a father about to witness the birth of a child.

For months, Carter had been crafting a military mission to counter the Pentagon's newest fear: the spread of nuclear weapons to terrorist groups or nations that saw the United States as their enemy. As the Pentagon's point man for nuclear security, Carter had helped draft a classified Presidential Decision Directive, PDD-13, that gained Clinton's backing for the new initiative. He had guided an announcement speech through interagency nitpicking. He had scrambled to rewrite charts at the last-minute insistence of the White House. Now Carter listened intently as his boss, Defense Secretary Les Aspin, announced the plan to the crowded room.

"We've recognized a new problem," Aspin declared ominously, "and we're acting to meet it."

The breakup of the Soviet Union meant that "loose nukes" might fall into unfriendly hands, Aspin warned. Instead of guarding against an all-out nuclear war with the Soviets, the U.S. military now faced dangers such as "fighting a Desert Storm kind of war with the opponent actually having a handful of nuclear weapons," he said.

To defend against the threat of nuclear disorder, Aspin said, the United States needed to develop new weapons that could penetrate underground bunkers hiding nuclear facilities. It also needed to gather better intelligence about emerging nuclear threats. To handle this new initiative, Aspin had created a high-level position that Carter would fill -- assistant secretary of defense for nuclear security and counter-proliferation.

Carter's anointment that day as the Pentagon's master of counter-proliferation -- a new word for post-Cold War dictionaries -- marked an important turn in the U.S. government's struggle to redefine the role of nuclear arms in a rapidly changing world. Counter-proliferation became the Pentagon's response to a terrifying question: Can countries such as Iraq, Iran, Libya and North Korea -- all hostile to the United States -- acquire nuclear weapons or provide them to terrorists?

For much of the past two years, that question has consumed Carter and others in the Clinton administration. The Soviet Union's collapse means that tens of thousands of nuclear weapons and tons of nuclear materials no longer are as tightly controlled. Moreover, the growth of global commerce and the spread of industrial technology mean that more and more nations might be able to build their own nuclear capabilities.

In tackling this challenge, the administration has found itself embroiled in heated debate at home and abroad over whether the best response to this emerging threat is to develop new American weapons and military doctrine or take a diplomatic, cooperative approach -- or perhaps try both at once.

Carter had been preparing for his role in this debate for years before coming to Washington. Trained as a theoretical physicist and medieval historian, he saw national security as a field where he could leave his mark. In meetings, his ardor coursed through his hands; rarely at rest, they grew more active the longer he talked.

During the 1980s, while teaching at Harvard, Carter had joined 15 different U.S. advisory boards, councils and think tanks involved with nuclear weapons policy. He received high security clearances, and, as the Soviet Union collapsed, he ran or participated in several Pentagon studies on nuclear weapons, including one on the role of intelligence gathering in fighting the spread of nuclear arms.

Carter was not a hawk, but neither did he fear nuclear weapons. His published writing took a pragmatic approach, emphasizing rationality about a subject -- nuclear war -- that was in many ways irrational. Carter had been a critic of the Star Wars missile defense program during the 1980s on the grounds that it simply would not work. His articles and speeches repeatedly asked, "Where is the logic in this?"

Upon his arrival in Washington, Carter found a Pentagon swirling with anxieties about new nuclear dangers. The Pentagon's war games fueled the concern. In one exercise, first reported by Brookings Institution scholar Bruce Blair, a rogue Russian general launched a nuclear attack at the United States; some 500 nuclear weapons were detonated before the "limited war" ended. Another game, run by the Federal Bureau of Investigation, took place in the streets of New Orleans. Squads of government agents swarmed through the city, searching for a nuclear bomb hidden by terrorists. The agents saved the day, but only after being given clues about where to look for the hidden explosive.

Some intelligence analysts considered the Pentagon war game managers as a kind of research and development division for post-Cold War defense budgets. In the analysts' view, the war game leaders kept dreaming up terrifying threats so Congress would cough up funds for new weapons and new missions, such as counter-proliferation.

Carter, however, considered the proliferation threat to be real and immediate. As he prepared to join the Pentagon in 1993, Carter wrote that the possibility of a nuclear detonation set off in anger was now more likely than during the Cold War.

Preventing this, in his view, required shedding old habits of mind -- especially the reliance on slow, formal strategic arms control talks. Carter carried to Washington numbered "agendas for action," steps that could be taken quickly to shore up global nuclear security, especially in the former Soviet Union. He did not want to be remembered as just another intellectual with a theory for his time. He wanted to charge out and do something. Sometimes this got him more attention than he wanted.

Counter-proliferation proved to be just one example. In Washington and abroad, Clinton administration colleagues and European allies worried that the new initiative might disrupt or damage the world's diplomatic, treaty-based system for controlling the spread of nuclear weapons. Indeed, some worried that the plan might trigger a new kind of nuclear arms race.

From his winged headquarters building above the Danube River in Vienna, Hans Blix followed the Pentagon's plans with fascination and dismay.

A lively, pragmatic Swedish lawyer, Blix was director general of the International Atomic Energy Agency (IAEA), the world's nuclear watchdog. Founded in the 1950s to help regulate the spread of nuclear technology worldwide, the IAEA's most important mission in the 1990s was to detect and prevent the spread of nuclear arms. It did this through a system of inspections and safeguards at nuclear facilities around the world.

Where Pentagon planners and CIA analysts saw new and terrible nuclear dangers, Blix saw progress. Brazil, Argentina and South Africa had renounced nuclear weapons after the Cold War. Dozens of Latin American and African nations were concluding regional nuclear weapons-free treaties backed up by IAEA-style safeguards.

The IAEA's role in nuclear police work was advanced dramatically by the 1968 nuclear Non-Proliferation Treaty (NPT). The treaty, which allowed five nations to keep their nuclear weapons, called upon the IAEA to ensure that all other member nations did not try to acquire nuclear arms. Blix embraced not only the NPT's enforcement rules but also its idealism: Every NPT signer had pledged to work toward the total elimination of nuclear arms, and Blix had given much of his professional life to this cause.

Blix traveled the globe proselytizing that the Cold War's end offered a historic chance to move toward this goal. At the age of 68, Blix remained highly ambitious. He worked long hours; colleagues thought he still dreamed of becoming U.N. secretary general. Whatever his personal goals, he now pressed his nuclear disarmament views more boldly than ever. "At no time since nuclear disarmament talks began has the political climate seemed more favorable than now for far-reaching agreements," Blix exhorted. "It is time to act. The opportunities must not be lost!"

In Washington, key members of Clinton's nuclear policy team, including senior arms control negotiator Tom Graham, wanted to help Blix. Some of Graham's colleagues compared the IAEA's post-Cold War importance in fighting the spread of nuclear weapons to the North Atlantic Treaty Organization's earlier role in containing communism.

But Blix knew that while some officials in Washington saw his agency as part of the solution to emerging nuclear threats, others regarded the IAEA as part of the problem. Their doubts arose from the IAEA's performance in Iraq.

After the Persian Gulf War ended in 1991, the United Nations discovered that Iraqi President Saddam Hussein had pursued a $10 billion secret nuclear weapons program under the noses of the IAEA's inspectors. Blix and his agency took a lot of heat in Washington. "Iraq often successfully manipulated the inspections," concluded a 1993 study by the Pentagon's Defense Nuclear Agency, based on confidential interviews with returning inspectors. "The lesson here is that any potential violator in the future may be able to `out wait' the inspection process through delays and denials."

Several Americans who had worked at the IAEA publicly attacked Blix's inspectors as weak-kneed, milquetoast U.N. types. Some U.S. nuclear weapons scientists returned from IAEA-supervised missions to Iraq grumbling that Blix wouldn't know a nuclear bomb if he tripped over one. "Blix, an attorney, would argue with me as a scientist that we weren't understanding what we were seeing," recalled Jay Davis of the Lawrence Livermore National Laboratory. "He was not perturbed by transparent Iraqi deceptions."

These attacks infuriated Blix. His face turned red and he ranted for days after reading criticism of the IAEA's inspectors, colleagues recalled. He accused the IAEA's critics of ignoring the challenges of running a cooperative nuclear inspection system in scores of different countries -- all of them sovereign nations with their own laws and security interests.

Yet Blix knew his nuclear inspection system had flaws. In the original NPT bargain, European nations and Japan had insisted on the most passive inspections possible. IAEA inspectors could not look or ask questions beyond a few "strategic points" in any nuclear facility. They could not walk into a nearby building that looked suspicious or wipe samples from the floor of a plant to see if the nuclear traces showed signs of clandestine bomb activity. IAEA inspectors had become like bank auditors: They depended heavily on the good faith of their clients, so they were vulnerable to a devious embezzler intent on committing a crime.

Worried that his critics might seek to replace or augment the IAEA with a new posse of global nuclear police, Blix worked to revitalize his agency as a tough, credible inspection force. His first step was to make a revolutionary agreement with the very Washington agencies that housed his harshest critics, the CIA and the Pentagon.

Blix believed that better access to intelligence information might have saved the IAEA from embarrassment in Iraq. He asked his Board of Governors, which included several members from Third World countries who were wary of Western dominance at the IAEA, to approve the agreement. He overcame the skeptics and obtained permission to receive what the board euphemistically called "information" from governments worldwide.

CIA and Pentagon officials decided that sharing such intelligence could advance American interests by giving inspectors the tools to catch and isolate nuclear cheaters. Other countries also decided to share, but they had less information and conveyed it less often.Throughout 1993, Blix and a few senior aides drove regularly from IAEA headquarters to the U.S. Mission in the 19th district of Vienna. Security guards checked them in and escorted them to a secure "bubble." There, CIA and Pentagon intelligence analysts displayed satellite photos, line drawings and other evidence of potential nuclear weapons threats in North Korea, Iran and Iraq.

CIA analysts carried sensitive satellite photos to a closed IAEA board meeting on the simmering North Korean problem early in 1993. In November, IAEA inspectors traveled to Tehran with CIA ink drawings of new, suspect buildings under construction at Iranian universities and nuclear training centers.

Blix accepted such intelligence gratefully -- even greedily, in the view of some IAEA colleagues, who saw the program as a Faustian bargain. The intelligence briefings raised a danger that the agency was "being captured by the Americans," said former IAEA safeguards director Jon Jennekens. "There were almost weekly briefings, some involving many briefers from Washington and the national laboratories. Their presence was very much felt."

Another problem was the CIA's split personality. One side of the agency, the one Blix dealt with, analyzed and shared intelligence. The other ran covert operations to collect intelligence. Blix's allies at the State Department feared that the CIA viewed the IAEA as a ripe target for collecting data on foreign nuclear programs. Not only did the IAEA have rich files on foreign nuclear programs, its Vienna headquarters attracted scores of nuclear scientists from around the world. It only made sense for CIA agents to try to infiltrate the place.

State officials did not know what, if anything, CIA agents might be doing covertly at the IAEA -- the CIA refused to confirm or deny its intelligence collection work, even in arguments with administration colleagues. But State officials worried nonetheless that the CIA would get caught with its hands in Blix's cookie jar. That could badly damage diplomatic efforts to control nuclear proliferation, they argued during internal Clinton administration debates.

None of this debate involved Blix. His problem with the CIA and the Pentagon was their analysis of the danger of proliferation. In stark terms, they were pessimists and Blix was an optimist. Blix could not seem to convince them that his inspection system should be the centerpiece of a renewed campaign to prevent proliferation. "It seems clear that the dominant opinion in the CIA is that there is a growing -- not diminishing -- threat of nuclear proliferation," Blix said at a seminar for German security experts in Bonn. "It seems clear that some of the critics feel that the only real protection against proliferating `rogue states' is [military] hardware."

In other words, more missiles and more bombs, a handful of which cost more than the IAEA's annual safeguards budget. To Blix, it just didn't add up.

Ash Carter admired Blix's drive to sharpen the IAEA's teeth. But as he set to work in his five-sided office at the Pentagon, Carter did not think he could rely solely on the IAEA to defend U.S. soldiers and territory from the potential threat of new nuclear-armed enemies. Carter's job was to think about battlefields and wars -- places where diplomacy and the IAEA already would have failed. The last thing he wanted was to meet a nuclear-armed North Korea in war without knowing what to do about it.

But the next questions were the tough ones: what kind of weapons, for what military purpose and at what diplomatic price?

At the Pentagon, Carter inherited studies that sought answers mainly in the lessons of the Persian Gulf War. That war had introduced a revolution in military thinking. Much of the public attention paid to high-tech weapons after the gulf war had focused on conventional arms. But at the Pentagon, the war had spurred a reassessment of the United States' nuclear arsenal that produced some startling ideas.

Beginning in mid-1991, the Pentagon, the Air Force and the nuclear weapons laboratories initiated several studies on new Precision Low-Yield Weapons Designs, also known as "mininukes," "micronukes" and "tinynukes." The Pentagon also carried forward work on High-Powered Radio Frequency bombs, nuclear-driven weapons designed to disrupt electronics and communications. The studies reflected the rising concern about future small wars where enemies might use weapons of mass destruction against U.S. troops.

After the collapse of the Soviet Union, the U.S. military began withdrawing its smaller, tactical nuclear weapons from around the world. In the main, the Pentagon's deployed arsenal after the Cold War was gigantic -- intercontinental missiles and heavy bombers with very powerful warheads.

Those big nuclear weapons would not do much good in a small war, some of Carter's Pentagon colleagues argued. To persuade the Saddam Husseins of the world never to use nuclear or chemical weapons, U.S. forces might need nuclear weapons small enough to be a plausible deterrent on a regional battlefield. In April 1993, the Joint Chiefs of Staff issued a doctrine for the conduct of nuclear war that said "a selective capability of being able to use lower-yield [nuclear] weapons" was a "useful alternative."

Small nuclear weapons might also be valuable in attacking deeply buried enemy targets. After watching U.S. warplanes drop bombs down chimneys in Iraq, potential Third World enemies such as North Korea and Libya bought cement and dug deep, hardened bunkers for defense, according to intelligence assessments that circulated in the Pentagon. Conventional bombs might bust some of those bunkers, but not the deepest ones. In technical terms only, mininukes would do the job better.

Carter had been thinking about these questions for years. At Harvard, he had written that the United States should either get rid of tactical nuclear weapons entirely or else rationalize them into a convincing, well-managed "nuclear expeditionary force" for regional conflict. But Carter discovered quickly that very few people outside the Pentagon wanted to think about the future of nuclear deterrence in regional wars. Even less did they want to miniaturize Washington's nuclear arsenal.

In the spring of 1993, nuclear disarmament activist William Arkin disclosed the mininuke studies. Several months later a Democrat-controlled Congress, outraged that the Pentagon sought new weapons so soon after the end of the arms race, reacted by banning any further research.

At the end of 1993, Carter flew to Brussels to sell NATO allies on the new counter-proliferation initiative. With communism gone, NATO needed a new mission. The alliance still managed a sizable nuclear force built to deter the Soviets from invading Europe. But NATO's High Level Group, a committee of defense officials who set nuclear policy for the alliance, had not come up with any new ideas. Carter, who could be impatient, complained to colleagues after attending NATO debates on nuclear policy, "Oh my God, what does this have to do with the post-Cold War world? What can we do to move this forward?"

During this trip and again a few months later, Carter carried his colored charts to NATO's elephantine concrete headquarters near the Brussels airport. "If counter-proliferation is good for us, it's good for NATO," Carter told his Pentagon colleagues.

But European ambassadors who assembled in a conference hall to hear Carter made clear they did not want to touch anything that smacked of preemptive strikes or new kinds of bombs. That would undermine the upcoming May 1995 vote by more than 170 nations on whether to extend the nuclear Non-Proliferation Treaty, they said.

With the NPT vote on the line, the Europeans declared, the time had come to emphasize diplomatic programs to curb nuclear proliferation -- not new weapons.

Carter did persuade NATO to begin a step-by-step program to develop its own counter-proliferation mission and to pool intelligence about new nuclear weapons threats around the world. The allies agreed to embark on a large-scale classified analysis on new threats, with Carter and a French colleague as co-chairmen. After that, they would consider whether they needed new weapons to fight the threats, the ambassadors agreed.

The most basic questions remained unresolved: Who were the enemies? What was the purpose and scope of U.S. and European nuclear weapons policy? As difficult as those questions were to answer in Brussels and Vienna, they were even more daunting -- and more urgent -- in Moscow and Beijing, the capitals of America's two potential 21st-century nuclear rivals.

NEXT: The nuclear triangle Copyright 1995 The Washington Post By Steve Coll and David B. Ottaway

Sig Hecker, director of the Los Alamos National Laboratory, hurried from the airport to a Santa Fe restaurant on Feb. 23, 1994, for an unprecedented meeting with six senior officials from China's nuclear weapons program.

Their tour of U.S. nuclear weapons labs had been years in the making. As the day grew nearer, classified memos ricocheted from Washington to the labs and back as officials fretted about how much to show the Chinese visitors, what could be discussed safely and what goals might be achieved.

The Chinese, including the chief engineer of China's nuclear test site and a general who ran the country's main weapons lab for many years, arrived without publicity at San Francisco International Airport to a nervous welcoming party of Chinese diplomats and U.S. security officers. No senior officials from China's nuclear weapons program had ever been invited on a guided tour of the labs, and everyone wanted to make sure nothing went wrong.

The Clinton administration had asked Hecker to advance the United States' quiet nuclear engagement with China. Administration officials weren't sure where the evolving relationship was going, but they thought new cooperation might persuade Beijing's leadership to help Washington contain the spread of nuclear arms worldwide. Yet there were political risks -- not least because Beijing's human rights record provoked anger in the U.S. Congress.

Hecker's main mission was to focus on cooperative, unclassified science projects, such as sophisticated computer modeling of China's energy and environmental problems. Inevitably, sensitive issues had arisen in the new relationship, and the Chinese seemed eager to pursue them. One was the future of nuclear testing worldwide. Another was the security of Chinese nuclear bombs.

The Chinese wanted to obtain closely guarded U.S. nuclear weapon locking systems called "permissive access links," or PALs, which keep bombs secure from unauthorized use. China had informed U.S. officials in Washington that during the 1989 Tiananmen Square democracy demonstrations, Beijing's leaders feared their army might split over the decision to crush student protests. According to accounts provided to the U.S. government, the Chinese had worried about losing control of their nuclear arsenal. American PALs, the Chinese had said, would reduce any such risk in the future. Washington could not decide what to do about the Chinese request.

The PAL issue -- and Hecker's mission to develop mutual trust -- reflected a much larger question: How will the United States, China and the world's third strategic power, Russia, redefine their triangular nuclear balance in the 21st century? Will they be partners or rivals?

For decades, the global strategic nuclear balance has been shaped by the Cold War's competition between communism and capitalism. But the Soviet Union's demise and stunning economic growth in China have created uncertainty in the world's nuclear order. The United States, Russia and China all have nuclear-armed missiles that can strike one another's territory in 30 minutes or less. While U.S. and Russian arsenals dwarf China's, the Chinese economy is expanding much faster, meaning Beijing might acquire the financial capabilities to match its rivals.

Since the fall of the Berlin Wall, each country has separately pursued talks with the other two -- sometimes quietly, sometimes formally. The talks have centered on vital questions: Can Russia and the United States undo the Cold War's balance of nuclear terror -- the reciprocal targeting of thousands of strategic nuclear weapons? Will China seek to modernize its arsenal and become a nuclear superpower, or will it cap its force levels and accept a proposed global nuclear test ban?

These contacts include a series of extraordinary discussions between U.S. and Russian commanders of nuclear forces that go well beyond what their civilian leaders have been willing to consider. In addition, exchanges among nuclear scientists in Russia, China and the United States have played a part in efforts to draw up new nuclear weapons treaties. Few details about these talks have been previously reported. In some cases, the new channels have been opened in total secrecy.

The new contacts have been hindered by Cold War-style suspicions. Each side worries that the others might seek a competitive advantage in nuclear security. Among other things, this tension could undermine existing nuclear arms control agreements such as the 1972 accord banning anti-ballistic missiles.

In the United States, the push to explore change has come often from the very people responsible for maintaining the defunct Cold War order: the scientists at the weapons labs and the generals in command of strategic nuclear missile forces.

For years they had gone to work each day contemplating their rivals on the other side of the nuclear divide. When the Cold War collapsed, they rushed forward to find each other almost instinctively, like athletes embracing on the field at the end of a long, hard game.

In December 1993, Gen. Igor Sergeyev, commander of Russia's nuclear missile forces, flew to Offutt Air Force Base near Omaha, headquarters of the U.S. strategic nuclear arsenal. For decades, Omaha had been a prime target on Sergeyev's nuclear missile maps. Now he was about to walk through Offutt as an honored guest.

His host was his American counterpart, Gen. George Lee Butler. A compact, intense man who had commanded the United States' nuclear forces since 1991, Butler belonged to conservative, small-town America. He believed deeply that the United States had to defend its values by all military means necessary. Yet the destructive power of the nuclear weapons under his command sobered him. On some mornings, Butler went out of his way to drive by a depot where nuclear weapons were stored, just to remind himself of the scale of his responsibility.

If the nuclear equation between the United States, Russia and China was going to change, this was one of the places where it had to begin. Offutt housed U.S. targeting computers and missile control links still able to send thousands of nuclear warheads across the Arctic Circle and into Russia within minutes of the U.S. president's command.

Butler and Sergeyev shared a giddy awareness of the historic importance of their one-on-one encounter. They could become emotional when they talked about the chance to transform the nuclear relationship between their two countries, according to people familiar with their meetings. They remembered one conversation in particular.

It began with a question from Sergeyev: What are your plans for modernizing your nuclear weapons forces for the 21st century? he had asked Butler.

"I don't have any," Butler answered. Butler said he had already agreed to terminate several U.S. nuclear modernization programs. Still in the works was a new guidance system for the Minuteman III missile and the completion of a couple of dozen strategic bombers and a similar number of nuclear submarines.

"What will happen in 2010?" Sergeyev asked, referring to the limited shelf life of present U.S. strategic nuclear forces.

"I'm betting on more cuts and on the U.S. and Russia getting closer and Russia changing more, and both sides wanting to put an end to nuclear arms," Butler told him.

"That's my bet too," Sergeyev said. "And I'm not going to lose the money. We're beginning to understand for the first time that the Cold War has ended."

Before the two men parted after their two-day visit, Sergeyev presented Butler's mother with a symbolic gift: an antique Russian egg painted with the image of Christ's ascension to Heaven. "How wonderful it is to be able to live and come and see you and to rediscover our religion from the minds of our grandmothers and our mothers," Sergeyev said. He wept. So did some of the American generals and admirals present, one of them recalled.

One of Sergeyev's senior aides, Maj.-Gen. Vladimir Dvorkin, commented later that everything had changed with this visit, and yet nothing had. It was one thing for uniformed generals to embrace each other after decades of confrontation, to build new trust and to talk about a future world free from the danger of nuclear war. It was quite another thing to unplug the complex, interlocked nuclear computer and targeting systems constructed so painstakingly during the Cold War -- or to convince civilian politicians in Washington and Moscow that such change was prudent and desirable.

Indeed, generals such as Butler, Sergeyev and Dvorkin were more willing to seek change than some of their uniformed colleagues or civilian overseers. Dvorkin, for instance, had proposed collaboration between the U.S. and Russian nuclear establishments to fight the spread of nuclear arms worldwide.

There had always been a streak of radical disarmament thinking about nuclear weapons in the U.S. uniformed military. Military commanders were paid to fight wars; they adhered to a tradition of honorable combat. Nuclear weapons did not fit well in this tradition.

When the Soviet Union collapsed, old questions rose to the surface. Air Force Gen. Charles "Hank" Horner, who ran the U.S. Air Force campaign during the Persian Gulf War before becoming the chief of U.S. missile defense, declared publicly that he did not think the United States needed nuclear weapons anymore.

Yet at the Pentagon, the White House, the Russian General Staff headquarters and in the Kremlin, there was little interest in such radical ideas. Senior officials in both governments, influenced partly by domestic politics, feared going too far too fast. Nuclear weapons had been a fountainhead of political power ever since their invention.

For Russia, with its economy in collapse and its conventional army in tatters, nuclear weapons provided one sure way to retain influence on the world stage. Horner said he ran into this reality when he met with his senior Russian counterpart, Gen. Vladimir Ivanov, in Colorado in 1994. Over shots of vodka, Horner said, he told Ivanov that the Cold War nuclear standoff "was stupid. . . . There is no reason in the world I'm going to fight you. First of all, you are too far away and I'd freeze to death getting to you and you'd freeze to death getting to me. So this idea about us fighting is bullshit."

Ivanov replied: "If it's the last thing, we will keep our nuclear weapons to keep our superpower status."

By early 1994, Sig Hecker knew as much as any American about the changing shape of the world's strategic nuclear triangle. He was building ties to nuclear weapons scientists in Russia and China -- simultaneously.

These contacts provided Hecker with a strategic mission at a time when his laboratory sorely needed one. Los Alamos's budget was falling. Good scientists were leaving the weapons labs for universities or retirement. Hecker seized on a new slogan late in 1993 to define and justify the lab's post-Cold War mission for Congress and the public: "Reducing the nuclear danger." He cited his evolving partnerships with nuclear weapons designers in Russia and China as prime examples.

The Russian program had been born in response to a fear: that the Soviet Union's collapse would lead Russian scientists to sell their services to aspiring nuclear powers such as North Korea, Iran, Iraq or Libya.

"Look, if you want to find out what it takes to get their scientists to stay home . . . why don't we go ask them?" Hecker asked Adm. James Watkins, then secretary of energy, at a Leesburg, Va., retreat in mid-December 1991.

Several months later Hecker found himself jogging what he called "a victory lap" around Red Square.

That journey closed a chapter of his family's hard history. His Austrian father had been drafted by Hitler's army. Hecker was born in Poland in 1943, where his father had been assigned to the Nazi occupation force. Soon afterward his father disappeared in combat on the Russian front. As a small boy in Austria after the war, Hecker watched his mother search vainly for news of his father's fate. For Hecker and his playmates, whose fathers had also vanished on the Russian front, the word "Siberia," site of Soviet prisoner of war camps, connoted awesome mystery.

Hecker emigrated to the United States at 13; in his early twenties, he joined the staff at Los Alamos. He rose through the ranks to become director in 1986. Throughout the Cold War, Hecker had spurned invitations to travel to Poland or the Soviet Union. He wanted to return only when the Cold War had ended -- as a victorious American, not an ambivalent emigre. By 1992, he could.

"This may not be just a short window," Hecker told his colleagues. "We may actually be able to change the way the two countries function."

After his stop in Moscow, Hecker traveled to Arzamas-16, the closed city that housed Russia's premier nuclear weapons lab. Hecker was amazed to find himself within its gates. Even the city's name had been classified top secret in the United States until the late 1980s. But now Hecker met and embraced the Arzamas director, Vladimir Belugin.

The next morning, he went out for a jog around the site, but armed guards would not let him leave their sight. Hecker complained. Belugin reprimanded his soldiers, and relations between the two men quickly warmed.

What could we do to make sure there isn't a loss of weapons science knowledge from Russia? Hecker asked Belugin one day.

Collaborate with us on science and physics, Belugin answered emphatically.

On his return to Los Alamos, Hecker told his colleagues, "These scientists will not go to Iraq. They will not go to North Korea. That's not what they want from life."

Belugin sent his deputy, Yuri Trutniev, to Los Alamos in April 1993, for a reciprocal visit. Trutniev attended the 50th anniversary celebrations of the laboratory's founding. He gave Hecker a carved Russian bear, and another Russian nuclear weapons scientist handed over a plaque with a piece of a dismantled Soviet nuclear warhead. The inscription said, "From Russia, with love."

That spring's celebrations of the atomic bomb's birth atop Los Alamos's 6,000-foot mesa reminded the Russian and American physicists that they belong to a special brotherhood. In its geology and its scientific culture, Los Alamos had the air of a world apart. It often lay shrouded in clouds above the flat New Mexican plains. Its physicists craved contact with colleagues who spoke their arcane language; the visitors from Arzamas did.

With the first visits completed, it was time to turn words into deeds. Belugin delivered an 11-point list of possible joint scientific projects. Some involved unclassified physics, some touched on classified areas. But when Hecker's colleague Steve Younger, deputy director of nuclear weapons technology at Los Alamos, sought approval for even the unclassified proposals, he met stiff resistance in Washington. One strain of thinking at the Pentagon held that the best thing to do with Russia's nuclear weapons establishment was to let it freeze in the dark.

Belugin grew frustrated with American delays. "Americans talk, talk, talk, but never do anything!" he ranted at one meeting before storming out of the room.

In August 1993, Younger flew to Arzamas. Belugin said he wanted to speak privately. He threw out his security guards and asked a senior aide to interpret.

"I want to level with you and tell you how things are," Belugin said. "We're desperate. We're not getting money from Moscow. The Americans are not delivering. . . . You're driving us into the hands of the Chinese."

Indeed, American visitors saw increasing signs of collaboration between Russian and Chinese nuclear weapons scientists. Russia's military had provided a PAL command-and-control system to the Chinese, though the Chinese later informed Washington that they would prefer U.S. devices. In the long run, if China acquired Russian ballistic and warhead technology, it might become a formidable nuclear rival.

As Younger departed from the Arzamas train station that August, a Russian scientist embraced him warmly and gave him a notebook. The Russian could not speak English well, but he kept saying "Sakharov, Sakharov." On the train, Younger opened the book and discovered it held original scientific notes jotted down by Andrei Sakharov, father of the Soviet hydrogen bomb and later a powerful symbol of resistance to Soviet communism. The notes concerned a method Sakharov had toyed with during the 1950s to achieve fusion energy ignition, an unproven way to turn the power of thermonuclear bombs into a peaceful, sustainable source of electricity.

The possibilities wowed Younger: American and Russian nuclear weapons scientists collaborating to find a replacement for fossil fuels, inspired by the scientist who helped destroy the Soviet Union's moral authority.

Washington had to help, Younger thought, as he later recalled. He returned to Los Alamos and banged out a stream-of-consciousness memo. He flew to Washington and made a nuisance of himself, charging from agency to agency saying, "Don't give me this crap about the programs in place. . . . You have done NOTHING."

Washington yielded slightly; approvals and limited funds for the Sakharov-inspired physics experiments came forth.

But Younger kept repeating a message: These are some of the most formidable nuclear weapons scientists and strategic generals in the world. Their nation may rise again. Can't we find a new way forward based on mutual respect and collaboration?

Hecker's lab-to-lab contacts represented a new form of freelance nuclear arms control developing among Russia, China and the United States. They formed a back-channel complement to traditional, formal treaty talks still ongoing among the world's three most important nuclear powers.

As a result of his new role, Hecker found himself thrust into negotiations for a treaty about which he had serious reservations. In January 1994, talks on a Comprehensive Test Ban Treaty opened in Geneva. Of the five nuclear powers, two offered the stiffest resistance to a deal: France and China. By now the United States, Russia and Britain were largely committed to a test ban. If China did not sign on, it would be free to modernize its nuclear forces while the others had to stand still. That would kill the treaty.

For their part, China and France worried that they would lose militarily if all five powers stopped testing. Testing proved the viability of new, more powerful or more efficient warhead designs. The United States, Russia and Britain had sophisticated techniques to simulate full-scale nuclear testing. Chinese and French physicists seemed to fear that U.S. and Russian superiority in simulation technology would give them a leg up in warhead maintenance and modernization in a testless future.

Appeasing French fears wasn't too difficult; the French were U.S. allies, and by 1994 Washington was prepared to share important, classified test simulation technology with them to help secure a global test ban.

China was a tougher case. Washington's relations with Beijing were always tense and sometimes hostile. Sharing classified test simulation technology with China would be unacceptable to the U.S. Congress.

For decades, the United States had known little about China's nuclear weapons establishment. That changed in 1990 when a Los Alamos scientist, Dan Stillman, made contact with Chinese counterparts and forged an extraordinary back-channel opening to China's nuclear weapons labs.

Stillman's success made Washington nervous. Were the Chinese plumbing deviously for U.S. secrets, or were they offering new cooperation, or both? By 1994, Washington had a reason to find out: the desire to conclude the proposed global test ban.

Hecker found his Chinese counterparts remarkably friendly when they arrived at Los Alamos in February 1994. They said their labs had orders from Beijing to contribute to economic growth; they were busy making commercial products from their weapons technology. Hecker too wanted to push Los Alamos into commercial collaborations. Turning bombs into money proved an easy talking point.

After the Los Alamos visit, the Chinese invited Hecker to their main nuclear design facility in what they called Science City, near Mianyang in the mountains northwest of Chengdu. Hecker delivered a lecture on "reducing the nuclear danger" before the elite of China's nuclear weapons establishment. He could hardly believe how fast relations were changing.

Afterward, there were a few awkward questions from the audience about nuclear test simulation technologies. Hecker fudged his answers. The guidance from an interagency task force in Washington was clear: Build friendly relations but offer no classified data to the Chinese. To show good faith, Los Alamos scientists plied the Chinese with unclassified papers about test simulation. But they returned home uncertain about whether China intended to sign up for a global test ban.

Meanwhile, officials in Washington debated whether to provide PAL nuclear bomb locks as the Chinese had asked. Some argued that providing PALs would make the world safer by lowering the chance of an unauthorized nuclear detonation if, for instance, a Chinese weapon were stolen. Others asserted that even though the older technology requested by the Chinese was no longer in use by U.S. nuclear forces, providing PALs might help the Chinese learn to pick U.S. nuclear locks. The debate would remain unresolved into 1995.

So would other key questions about the nuclear triangle. Most ominously, amid all the purposeful talking in Geneva about new agreements among the three powers, one treaty lay dormant like an undetonated mine. The document was the Anti-Ballistic Missile Treaty of 1972, a relic from the worst days of the U.S.-Soviet nuclear arms race.

The issue was the future of missile defense. After suffering Iraqi Scud missile attacks during the Persian Gulf War, the United States sought to convert the remnants of its Star Wars missile defense project into a viable system of missile interceptors to protect soldiers in regional wars. Russian generals worried the system could upset the world's nuclear balance. Debate raged in Congress about whether to make concessions to the Russians. But the debate was uninformed by a fact that U.S. negotiators learned privately from the Russians during Geneva treaty talks.

The Russians said that if they developed their own regional missile defense system to match that of the United States, they would retain as an option putting small nuclear warheads on their interceptors -- an alarming possibility that the Pentagon had ruled out for its own weapons. It wasn't clear the Russians could afford to develop this kind of system. But if missile interceptors did become a feature of 21st-century defense and the Russians were able to equip them with warheads, it would mean the deployment of hundreds or even thousands of new nuclear-armed missiles -- a new kind of nuclear proliferation.

Missile defense involved unproven technologies and deployments still years away at best. But the Russian disclosure symbolized how much lay unresolved in the Cold War's nuclear legacy. Generals and civilian leaders in the United States, China and Russia had begun an extraordinary effort to change the old order. But as 1994 progressed, they also discovered that sometimes the old system had a momentum all its own.

NEXT: Retargeting the missiles Copyright 1995 The Washington Post By David B. Ottaway and Steve Coll

William J. Perry and Ashton Carter swept in a helicopter above the Ukrainian countryside, speeding over villages and long rectangular fields. The U.S. secretary of defense and his chief aide for nuclear weapons policy were flying toward a destination that neither could have imagined a few years before: the inside of a former Soviet nuclear command center.

They thundered down, boarded an elevator and sank 35 yards underground, led by their hosts from Ukraine's military. Two missile control officers lived below, equipped with cots and a cooking pot. Menacing photographs of American nuclear weapons lined the walls. To one side stood a map of the United States with lights marking targets for the center's Soviet-era nuclear payloads.

Some of the map's target lights had been turned off, Carter noticed. But though it was March 1994, and the Soviet Union had ceased to exist two years earlier, other target lights still shone. The lights were vivid reminders that while the political landscape had changed dramatically, the missiles systems that define the Cold War balance had not.

Carter, who had joined the Pentagon the year before with strong ideas about the role of nuclear weapons in a changing world, journeyed that day to the very center of the Cold War's nuclear deterrence and war machinery -- the vast network of strategic missile launch centers and computers that for decades had targeted the United States with thousands of nuclear warheads.

A generation of Americans, Carter among them, had been raised in awed contemplation of these nuclear war systems. Throughout the Cold War, dozens of novels and popular Hollywood films -- from "Dr. Strangelove" to "The Day After" to "WarGames" -- tried to imagine what would happen if by accident or malice U.S. and Russian strategic nuclear missile systems were ordered into action and total nuclear war began. New phrases crept into the popular lexicon to describe the terrible, automated power of the system -- to have your "finger on the button" connoted a literal ability to destroy the world.

The Cold War's demise has given the two nuclear superpowers a chance to turn the old system off -- or at least reduce its dangers drastically. Since 1991, the United States and Russia have managed to dismantle parts of the old nuclear structure that lay in former Soviet republics such as Ukraine. For its part, the United States has carried out a secret but wide-ranging inventory of its war plans for nuclear weapons, eliminating thousands of strategic targets. It has also undertaken the first top-to-bottom rethinking of the size, structure and purpose of Washington's strategic nuclear weapons system in more than two decades.

Although these reforms have ushered in significant changes, the United States' primary nuclear war plan still targets Russia and provides the president an option to counterattack within 30 minutes of a confirmed enemy launch.

Rendering this system safe after the Cold War had been a central goal of Carter's writing and political lobbying before Clinton invited him to leave Harvard for the Pentagon. Moreover, by the spring of 1994, Carter had a franchise at the Pentagon to chase these heady goals: He had been put in charge of a thorough study of U.S. nuclear strategy that the Clinton administration called its Nuclear Posture Review.

The assignment reflected Carter's rise to new heights of influence. Perry, the newly appointed defense secretary, was an old friend and mentor from days in academia. He gave Carter new responsibilities, adding supervision of contacts with the former Soviet Union to his earlier brief of nuclear policy, security and targeting.

Because Carter was young, dynamic and smart, outsiders to the U.S. government -- especially those who favored rapid reduction or elimination of nuclear weapons -- tended to invest their hopes in him. Indeed, he was becoming a minor cult figure in Washington's small but intense community of nuclear weapons specialists -- the scholars, pamphlet writers, ideological lobbyists and journalists who tracked changes in the world's nuclear order.

This was not a role Carter sought. But neither could he avoid it.

A physicist by training and a pragmatist by inclination, Carter had a scientist's desire to pull things apart and see how they worked. Like Perry, he was as much engineer as intellectual. As he set to work, his engineering principles led him to an obvious conclusion, shared by many others: The very first thing to do was to ensure that only one nuclear-armed successor state -- Russia -- emerged from the Soviet Union's fall. That required removing nuclear weapons from Ukraine, Kazakhstan and Belarus.

Ukraine proved the toughest case. By January 1994, Carter had helped to broker the historic Trilateral Accord, under which Ukraine agreed to surrender its nuclear weapons. When he traveled to the Ukrainian missile base the following March, he and Perry saw the progress firsthand.

The Ukrainian challenge was only one among many. Carter also had to concentrate on Washington's unanswered questions about the role of its Cold War-era strategic nuclear system. This was the purpose of the Nuclear Posture Review that Carter supervised at the Pentagon. For several years now, civilians and generals at the Pentagon had been trying to challenge the system from within. They had found, as Carter was soon to discover, that the old structure did not yield easily to anyone preaching change.

On Jan. 26, 1991, his second day as the commander of the United States' strategic nuclear forces, Gen. George Lee Butler descended to the Air Room, an underground computer center at Offutt Air Base near Omaha. The center housed the Pentagon's targeting and nuclear war planning system constructed to deter the Soviet Union from a surprise nuclear strike -- and to obliterate the enemy if deterrence failed.

What Butler found there confirmed what his civilian colleagues in the Pentagon had told him. The nuclear war plans Butler began to examine that day -- accessible only to a small minority of officers and civilians at the Pentagon -- seemed to Butler an almost farcical mess.

To understand why, it is necessary first to understand the assumptions and machinery of the United States' nuclear war planning system.

The last presidential guidance on the use of nuclear weapons in war dates to Oct. 2, 1981, when President Ronald Reagan signed a classified document called National Security Decision Directive-13. The directive's key point is that the president should have a varied menu of nuclear war plans to choose from during a crisis. Among these should be plans for limited nuclear exchanges that would allow a chance to pause and negotiate.

Down in the Air Room at Offutt, military officers translated Reagan's directive into concrete nuclear war plans. They wrote and oversaw a huge, deeply complex computer program. The program assigned targets to U.S. nuclear missiles, submarines and bombers; it calculated flight times and trajectories so U.S. missiles did not crash into one another; and it wired the whole menu of plans on a speedy time line so the president had an option to order a strike within minutes of a confirmed Soviet attack. The computer program had a name: the Single Integrated Operational Plan (SIOP).

The SIOP's architects labored in the strictest secrecy. For years they rejected oversight from Pentagon civilians or even the Joint Chiefs of Staff. And because the SIOP writers relied so heavily on mathematical formulas, their war plans owed much to the numbers and assumptions typed in at the start.

In 1989, the fall of the Berlin Wall galvanized Richard Cheney, secretary of defense in the Bush administration, to study the SIOP. The end of communism offered a chance to negotiate drastic reductions in the U.S. and Soviet nuclear arsenals, but Cheney couldn't sign off on the reductions if he did not know how they might affect U.S. nuclear war plans. Cheney, a conservative Republican, kept asking questions about the SIOP that his aides could not answer. On Nov. 25, 1989, Cheney sent out a two-paragraph memo ordering a full review of the plan by his civilian aides in Washington.

Cheney concluded the SIOP was not a nuclear war plan, people involved with the review recalled. There were no clear objectives in the SIOP -- it seemed like just a jumble of processed data. Instead, every time the Pentagon had bought a new nuclear weapons system to match the Soviets' during the Cold War arms race, Omaha had simply found targets for the added warheads and rearranged the SIOP math formulas. This had gone on for years as captains and majors who wrote the SIOP rotated in and out of the Air Room. Nobody had ever asked, How much is enough? How many nuclear weapons does the military really need to meet specific war and deterrence goals?

Worse, computer software writers in different parts of the Air Room had inadvertently programmed multiple nuclear warheads to strike the same target. The programmers organized all their work around four categories of Soviet and Warsaw Pact targets: enemy nuclear forces, conventional military forces, leadership and war-supporting industry. One programmer would work exclusively on the Soviet rail system, for instance, while another thought only about Kremlin leaders. Nobody checked to see if the targets they programmed independently were right next door to one another.

Some Russian industrial cities with different kinds of factories had absurd numbers of weapons aimed their way -- the software writers had assigned attacks for each factory in the area, without realizing that all the buildings stood within a few miles of one another. Nearly 40 weapons were assigned to the mid-sized Ukrainian capital of Kiev alone. Some Peacekeeper warheads, which have a radius impact of 3 1/3 miles, were targeted less than a mile apart.

The SIOP did include Limited Attack Options, which were designed to impress Moscow's leaders with Washington's restraint in a controlled nuclear exchange. But even these smacked of dangerous fantasy, in the view of Cheney and his aides. For instance, the programs assumed that a nuclear strike on a heavily populated suburb would not be seen as indiscriminate by Moscow's leaders -- as if a Soviet strike on Prince George's County would be seen by a U.S. president as an act of restraint because the target had not been downtown Washington.

The bottom line was that for all the attempts to create a flexible SIOP with options for limited nuclear war, Omaha had built a nuclear war plan that would obliterate Russian civilization no matter which button the U.S. president pushed. American presidents had been misinformed: They were told that they had a way to show restraint during nuclear war, but as a practical matter, they did not.

Cheney, the Joint Chiefs of Staff chairman, Gen. Colin L. Powell, and their aides threw thousands of targets out of the SIOP, helping to reduce the total from its Cold War peak of more than 40,000 to around 10,000 by 1991. When Cheney appointed Butler to take command in Omaha in 1991, the general charged into the Air Room and joined the work.

Butler reviewed each target one by one, tossing many out like a spring cleaner in a cluttered attic. One day, Butler eliminated about 1,000 targets in newly liberated Eastern Europe by declaring to colleagues, "We do not target democracies," people involved remembered.

That principle only went so far. By the time Butler retired in February 1994, he had helped hack the SIOP to about 2,500 targets. But the SIOP still primarily targeted Russia -- now a democracy, if a very fragile one.

In meetings with Russian counterparts, Butler dreamed aloud of a future where mutual nuclear targeting between Russia and the United States would no longer be necessary. But by 1994 Butler was also a bruised man. His enthusiasm for change had met with stiff resistance from colleagues and earned him some enemies within the military. He was a candidate to succeed Powell as chairman of the Joint Chiefs of Staff, but he was not selected. He retired at the age of 54.

The SIOP Butler left behind had been altered in more ways than size. Indeed, Butler's goal was not to destroy the SIOP. He wanted the SIOP to adapt and survive. The name Butler gave to his vision had a Darwinian ring: "The Living SIOP."

Butler wanted Omaha's nuclear war-planning machine to become faster, smarter and more globally flexible for the 21st century. The SIOP's computers should be able to target unpredictable threats around the world on short notice with both nuclear and conventional missiles, Butler believed.

He approved computer acquisition programs to equip Omaha with the hardware and software needed to meet these goals before the year 2000. He reduced the time it took to write a new SIOP from 18 months to one year. He sought "an adaptive planning process that will offer the President a variety of options in response to any crisis," he wrote in an unclassified briefing paper. "The goal is to provide viable options in less than 24 hours."

He also authorized development of the Silver Book, a list of existing chemical, biological and nuclear weapons facilities in Third World countries such as Iraq, Iran, Libya and North Korea that might threaten U.S. troops during a regional war.

When news of the Silver Book leaked to Jane's Defense Weekly after Butler's retirement, it provoked a hail of criticism from nuclear disarmament groups such as Greenpeace. They argued that the Silver Book implied an expanding role for the U.S. nuclear arsenal. Embarrassed by the leak, Butler's successors sought to play down the Silver Book's importance, saying it had only reached a prototype stage and was not being pursued seriously.

Whatever the truth, one thing was clear: Butler had achieved many of his most important goals. The SIOP was no longer a bloated mess. Nor had it become extinct. It lived.

What Butler left undone, Ash Carter proposed to study in the 1994 Nuclear Posture Review. For the first time in decades, everything about U.S. nuclear policy was up for grabs. How many nuclear weapons should the United States deploy early in the next century? For what purpose?

Carter appointed six working groups staffed by uniformed officers and civilians to hold closed-door hearings in the Pentagon on these and other questions. He urged them to free themselves from Cold War thinking and consider the issues creatively.

Carter had worked on the SIOP while at Harvard as part of a panel that aided Cheney's reform drive in Omaha. He was interested especially in the hair-trigger alert status and 24-hour operations required of U.S. nuclear forces by the SIOP.

His critiques dissected a key assumption at the heart of U.S. nuclear war planning. During the Cold War, politicians in Washington and Moscow each assumed their enemy might launch a surprise, massive nuclear strike if it were militarily possible. The machinery was carefully designed on each side to convince the enemy that no such first strike could succeed without a terrible retaliation. To do this, both countries had to arrange their nuclear forces to survive the best first punch the other side might conceivably throw.

The easiest way to beat a surprise first strike was to launch your own nuclear forces quickly, before your opponent's incoming forces landed. It took about 30 minutes for an ICBM to fly between Russia and the United States. This meant Washington and Moscow each had to operate its nuclear forces so they could launch within half an hour of a confirmed strike by the enemy. This alert posture was called "launch on warning."

The United States had no firm policy to fire nuclear missiles quickly if the Soviets struck first. That decision belonged to the president. He might have just a few minutes to decide as Russian ICBMs hurtled toward the United States. The president might well choose to let the incoming missiles land, wait to see how many U.S. nuclear weapons survived and then retaliate. This more passive option was termed "retaliation after rideout."

Before Carter arrived at the Pentagon, the United States and Russia signed an arms control treaty that would help each side ease off the hair trigger. Under the START II treaty, signed by Washington and Moscow in January 1993, both countries pledged to eliminate land-based nuclear missiles with multiple warheads -- missiles so powerful and dangerous that they caused each side to keep its forces on high alert. But Carter wanted to go further. While at Harvard, he had written that the Cold War's demise offered "freedom from dependence on alerting and warning, and above all, freedom from reliance on prompt response."

At the Pentagon, Carter found a soul mate in Butler, who had drafted a highly classified list of about 30 proposals that would gradually strip launch on warning out of Washington's nuclear war planning system -- in cooperation with Russia. In January 1994, Washington and Moscow announced agreement on one idea on Butler's list: Beginning in May, the two governments said, the United States and Russia would no longer target nuclear missiles at one another, but would aim them at oceans instead. But the deal was only symbolic; the missiles could be retargeted within minutes.

Butler's ideas list included other, more radical options. He proposed taking the warheads off land-based missiles and exchanging inspectors with Russia to help ensure that neither side could strike quickly. Carter made certain that these ideas, some of which were very similar to his own, became a focus of the Nuclear Posture Review's closed-door hearings.

But unplugging the nuclear war machinery was not just an engineering problem; it required trusting Russia. And attitudes toward Russia hardened at the Pentagon in early 1994. The blustering ultranationalist Vladimir Zhirinovsky had won a shocking number of votes in a Russian election. Meanwhile, ornery diplomatic disputes had erupted with Moscow.

In Omaha, Adm. Henry G. Chiles Jr. succeeded Butler as commander of U.S. strategic nuclear forces, and he worried about Russia's ability to change. Chiles thought the United States had already gone far enough in reducing the size and strength of its nuclear deterrent. He was cautiously encouraged by the new contacts with Russian nuclear missile commanders, but he thought the United States should move carefully until Russia complied with existing treaties such as START II.

When he heard ideas like the proposal to take warheads off missiles and exchange monitors to verify zero alert, Chiles immediately asked questions such as, "How do I know my guy isn't tied up and left in the corner? How does he know that his guy isn't tied up in my launch control facility and left in the corner? Do I have to have a monitor from each side in every launch control facility?"

Carter may have misjudged the number of officers at the Pentagon who shared Chiles's concerns. When the posture review's working groups first produced draft options for change in the United States' nuclear forces, Carter thought they had failed to come up with new and innovative options for the president. Overall, the working groups envisioned a 21st-century nuclear force that looked a lot like the Cold War-era force, only smaller.

Carter rewrote the proposals and included some new, bold options, including a plan to eliminate land-based nuclear missiles, one leg of what was called the "nuclear triad."

The land-based ICBMs contributed most to the hair-trigger aspects of the nuclear war system, some argued, because land missiles could not survive a first strike as easily as submarines and aircraft, the triad's other two legs. Eliminating land missiles would move the United States away from launch on warning.

But Carter's proposal leaked from the Pentagon to conservative senators. "The United States appears determined to go out of the nuclear weapons business," declared Sen. Strom Thurmond (R-S.C.) at a Senate hearing. Four other conservative senators wrote President Clinton to criticize Carter by name and denounce the proposal as "extremely ill-advised" because of the continuing danger of a surprise nuclear strike by Russia or China.

In the end, Perry did not support the proposal. He believed that START II's elimination of land-based nuclear missiles with multiple warheads, as well as other changes in nuclear operations, had made the need for further reform less urgent than before. By April it was clear that the study's most radical proposals had been rejected.

The review ended the following September when Clinton signed a classified Presidential Decision Directive, PDD-31, outlining U.S. strategic nuclear policy until early in the 21st century.

The directive concluded that the United States needed a robust nuclear force for the foreseeable future and preserved the nuclear triad. At the same time, it ordered some reduction in alert levels and new safety features. It left open the possibility of faster and deeper reductions in the U.S. nuclear arsenal if the political situation in Russia made this possible.

Perry called the new policy "leading and hedging." The United States would continue to lead the way toward smaller nuclear arsenals and lower alert levels, but would hedge by maintaining its ability to rebuild nuclear forces quickly and by keeping some of its nuclear missiles on Cold War-style alert.

This proved hard to explain to the Russians. When Perry and Carter first outlined the new policy to Russian Defense Minister Pavel Grachev at a meeting in New York, an interpreter translated "lead" as "dominate." There was not an obvious Russian word for "hedge"; sometimes it came out as "shrubbery" and other times as "ability to break out from treaty commitments." The mix-ups soon became lore in Moscow.

Even when the language was clarified, some Russians still saw the new policy as a threat. Among other things, it bothered the Russians that U.S. nuclear-armed submarines could still patrol Russian waters on high alert.

Carter tried to ease the tension by traveling to Moscow and delivering an extraordinary briefing to 60 Russian generals. He used virtually the same charts that had been shown to Clinton a few weeks earlier.

Carter's trip had been eagerly awaited by pro-Western politicians and generals in Moscow. They had hoped the review's outcome would provide proof to entrenched Russian conservatives that the United States no longer saw Russia as its main nuclear enemy. A drastic reversal of U.S. nuclear posture toward Russia might deprive the conservatives of their argument that the United States was an aggressive power that could not be trusted.

Instead, the briefing "made clear that we still have mutual deterrence and nothing was new," Alexei Arbatov, a member of the defense committee in Russia's parliament, said later. "It provoked a sigh of relief on the part of those people who didn't believe that our relations were changing."

As Sergey Rogov, president of Russia's Center for National Security Problems, put it: "I talked to Ash. I was trying to warn him. Unfortunately, some of my best friends in this administration turned out to be part of the problem, not part of the solution."

Carter rejected such complaints. He saw the briefing as evidence of incredible progress in the U.S.-Russia relationship. Never before had a U.S. official briefed Russian generals about U.S. nuclear policy. Even if they did not like everything they heard, he told colleagues, this candor proved that a new nuclear order was emerging.

NEXT: The North Korea challenge Copyright 1995 The Washington Post New Threats Create Doubt in U.S. Policy;N. Korean Crisis Creates Doubt in U.S. Policy;Talking Tough;Cutting a Deal;A Complex Challenge By Steve Coll and David B. Ottaway

Defense Secretary William J. Perry telephoned Sens. Richard Lugar and Sam Nunn on Saturday, May 21, 1994, to ask them to travel privately to North Korea and meet with its aging dictator, Kim Il Sung.

I want you to find out what is really on his mind, Perry said.

A week earlier, Kim had brought the simmering North Korean nuclear crisis to a boil. His government announced that it had begun withdrawing plutonium-rich fuel rods from a small nuclear reactor. Within months, North Korea could extract enough plutonium from the rods to make about five nuclear bombs, in addition to the one or two it might already possess. Perry feared that North Korea might build more than 100 nuclear weapons by the year 2000. In the meantime Kim threatened, if provoked, to turn Seoul into "a sea of fire."

"In many ways Korea poses the greatest security threat to the United States and the world today," Perry had declared on the night North Korea announced its nuclear fuel withdrawal. "We have to regard the situation as very dangerous."

What was theory before had suddenly become fact. Perry and his Pentagon protege, Ashton Carter, had written thousands of words about the potential dangers of a breakdown in the post-Cold War nuclear weapons order. They had developed a new military mission, called counter-proliferation, to prepare for the threat. But the uncomfortable truth, Perry later recalled, was that he had no way to know what Kim really thought about nuclear weapons. Nor did he have any way to be certain how the United States -- despite its vast nuclear arsenal -- could stop North Korea from using nuclear bombs in war.

Perry said later that he had "no reason to believe that our nuclear forces would be capable of deterring the use of their nuclear forces."

Nuclear deterrence had been a foundation of the Cold War; now, at least in Perry's mind, it had crumbled in Korea. For decades the United States had used its nuclear weapons to prevent the Soviet Union from launching a nuclear war. Now Kim had caused Perry to doubt whether, in a world no longer governed by superpowers, U.S. nuclear policy would be effective in deterring a nuclear attack.

Indeed, North Korea's bid to develop nuclear weapons shows that the United States has yet to resolve basic questions about the purpose and use of its nuclear weapons in a rapidly changing world. The U.S. military is uncertain about whether an American president would ever use nuclear arms against a regional enemy such as North Korea, even if that enemy struck first.

Fear that nuclear deterrence might fail on the Korean peninsula has pushed the United States toward a negotiated compromise with North Korea over its nuclear program. Meanwhile, Pentagon planners have been running war games during the past year to study how nuclear weapons might be used in a Korean conflict, searching for ways to prevent a nuclear exchange and to win a war if one occurs.

Yet even as the Pentagon has concluded that nuclear deterrence cannot be relied on in North Korea, it has declared that it must preserve its nuclear force to deter Russia's still sizable nuclear arsenal.

Perry had spent hundreds of hours during the Cold War thinking about how nuclear deterrence worked between the United States and the Soviet Union. Gradually, over decades, he became confident that a stable understanding had developed between the two sides. But Perry saw no way to apply that model of nuclear standoff to North Korea -- the two sides did not have any relationship that allowed them to read each other's intentions in a time of crisis. Perry did not know how willing Kim might be to risk his country in a nuclear exchange, or how Kim viewed Washington's willingness to use nuclear weapons in response to a North Korean chemical or nuclear attack.

In other words, Perry did not know many of the important things a secretary of defense would like to know when considering war against a potentially nuclear-armed enemy.

Perry planned for the worst. In May, he studied contingency plans. One was a preemptive military strike on North Korea's nuclear reactor. Another involved the deployment of more than 10,000 U.S. troops to South Korea to bolster the 37,000 U.S. soldiers already there. He also reviewed several full-scale war plans, involving the mobilization of hundreds of thousands of soldiers -- with detailed schedules that listed the specific transport planes to be used in ferrying the troops to Korea.

On May 19, Perry went to the White House to join Gen. Gary Luck, commander in chief of U.S. forces in Korea, in a briefing for President Clinton. Luck had done his own review and had concluded that even a conventional war on the Korean peninsula would take at least a million lives, including those of 80,000 to 100,000 U.S. soldiers. In direct expenses to the U.S. government and in its impact on the economies of South Korea, Japan and China, war might cost upwards of $1 trillion, Luck estimated. As he put it later at a congressional hearing: "Bottom line: Unbelievable hardships would occur."

The next day, Perry conferred at the White House with national security adviser Anthony Lake and Secretary of State Warren Christopher. With so much at stake, they agreed that they needed more insight into Kim's thinking. Did North Korea's dictator grasp the dangers of brinkmanship in the nuclear age? What sort of deal would stop him?

Top White House officials decided to ask Nunn and Lugar -- both knowledgeable about nuclear security issues and both recent visitors to South Korea -- to seek answers directly from the hermitic North Korean leader. Could they set up a meeting quickly, and without publicity? The senators agreed to try.

Carter, Perry's chief deputy for nuclear affairs, helped arrange hurried briefings for the senators. He and Robert Gallucci, the administration's point man on North Korea, assembled a thick briefing book on the North Korean nuclear program. Through multiple diplomatic channels, they informed the North Korean government that Nunn and Lugar were ready to leave Washington before the end of the week.

Early on Wednesday morning, May 25, the senators arrived at their offices with their bags packed. A military transport plane stood ready at Andrews Air Force Base. Before they could depart, they needed official clearance to cross into North Korean airspace.

Permission did not arrive. The hours ticked by. Word filtered back through second-hand sources that Kim was too busy to meet with the senators just now. An order went out to Andrews: Stand down the pilots. The senators went home that night with their luggage.

Not only was Perry still mystified by Kim's thinking, he couldn't even find a way to talk to him -- at least for now.

That spring, thousands of miles away in his spacious office atop the International Atomic Energy Agency headquarters in Vienna, Hans Blix was in a defiant mood. He had put his personal prestige on the line by pushing North Korea to live up to the nuclear Non-Proliferation Treaty (NPT), under which North Korea had pledged never to acquire nuclear arms.

As IAEA director general, Blix had been leading the world's campaign to contain North Korea's nuclear program. Now, Kim had broken his earlier agreements with Blix by withdrawing enough plutonium to make five bombs within months.

As the crisis swelled, ambassadors from many countries tromped in to see Blix. He told them he had been spending time with his legal pad, crafting points about the North Korean nuclear danger that "hadn't been said or said enough." From his pad, he read out strong words about North Korea's duplicity and its defiance of the NPT.

The thrust of Blix's message was: They shouldn't be allowed to get away with this.

U.S. officials listened to his fuming, and they worried about something they would not have dreamed possible two years earlier: Blix now talked tougher about North Korea than the Pentagon.

As with Perry, Kim tested Blix's assumptions and programs for a 21st-century nuclear order. Blix had been preaching that the world's nations could stop the spread of nuclear weapons through cooperative diplomacy and that the IAEA had the backbone to police the NPT regime.

From the time North Korea signed the NPT in the mid-1980s, years after most nations, Kim had pushed the limits of Blix's considerable patience. Blix took special care to nurture North Korea's cooperation with the IAEA, traveling personally to the capital, Pyongyang, to inaugurate its adherence. But Blix also was on his guard. He toured North Korea's nuclear facilities with a skeptical veteran of France's nuclear weapons program, then haggled and won a commitment from North Korea to allow nuclear inspections "anywhere, anytime."

Months later, IAEA inspectors found discrepancies in Pyongyang's nuclear inventory, and Blix concluded that at least 90 grams of plutonium were missing. When he pushed for answers, Kim got so fed up that his government telexed Vienna to declare that North Korea would withdraw from the NPT. Blix jabbed and feinted artfully with the North Koreans, winning concessions one month, giving ground the next. Eventually, North Korea announced it would "temporarily suspend" its withdrawal.

Then in May 1994, Kim changed the game on Blix with his decision to pull plutonium-filled rods from an IAEA-supervised reactor. Until this point, Blix and his inspectors had been focused on the past -- whether Pyongyang had hidden plutonium in the late 1980s and then lied about it. Now Blix had to worry about the future. Whether Kim already had a nuclear bomb was a mystery; that he now had enough plutonium for a sizable nuclear arsenal in the future was a certainty.

Perry, Carter and others at the Pentagon had supported Blix's hard-line approach all along. But the prospect of a war, and Perry's doubts about whether the U.S. nuclear arsenal could stop North Korea from using nuclear bombs, meant the situation had changed radically. So the Pentagon had to rethink its goals and approach.

Luck, the U.S. commander in Korea, and other Pentagon planners assumed that if North Korea had one nuclear bomb and used it, the United States would answer with a massive retaliation, whether conventional or otherwise. Clinton had said as much while standing on the border of North and South Korea in July 1993. But, Luck advised, if North Korea could assemble 20 or 50 or 100 nuclear weapons by the early 21st century, it would pose a far different and more terrifying threat because it might then be able to fight a war of multiple nuclear exchanges.

Maybe North Korea had one or two bombs already, or maybe not. The CIA and the State Department's Bureau of Intelligence and Research traded opposing analyses, with the CIA saying probably yes and State saying probably no. Luck advised Perry and Clinton that even if North Korea already had a bomb or two, that wouldn't make much difference to how he would fight a Korean war in the summer of 1994 -- if it became necessary.

Perry worried about the future. What if North Korea did acquire a large arsenal over the next few years? The United States might still be unsure how to deter North Korea from using its many nuclear weapons. Meanwhile, North Korea would grow into an even more credible nuclear threat, possibly able to blackmail South Korea into reunification. Worse, he feared, North Korea's nuclear-capable neighbors -- Japan, South Korea, Taiwan, even Indonesia -- might decide to defy the NPT and acquire nuclear weapons to defend themselves.

In May 1994, these ideas informed Washington's response to Kim's raised ante. Bolstering Blix and his IAEA inspection regime had started at the top of the United States' priority list when the North Korean crisis began 18 months earlier, but by May that goal had fallen to the very bottom. War was the issue now.

As one of the authors of the ballyhooed Counterproliferation Initiative in 1993, which set forth military options for curbing the spread of nuclear weapons, Carter had argued that the Pentagon needed doctrines and hardware to prepare for a regional war in which U.S. soldiers faced an enemy armed with weapons of mass destruction. Now such a war seemed a real possibility. Did the Pentagon have the tools it needed, and if so, would they do any good on a prospective Korean battlefield?

As the Korean crisis was unfolding last May, the brain trust of counter-proliferation gathered in a conference room at Los Alamos National Laboratory in New Mexico -- top aides to Perry, senior nuclear missile generals, admirals, several U.S. senators and nuclear weapons scientists from the lab.

The lab's director, Siegfried Hecker, had organized the meeting to raise counter-proliferation's profile in the nuclear weapons establishment and to help win funds from a stingy Congress. Kim's bid for a nuclear arsenal showed Hecker that he and his nuclear weapons lab still had a role to play after the Cold War.

Hecker's aide, Walt Kirchner, the Los Alamos scientist in charge of defense programs, escorted Nunn and some military officers on a tour of the lab's nuclear target modeling computers.

As they watched, the machines generated images of North Korea's nuclear reactor, processing plant and other facilities. With the push of a few buttons, lab scientists simulated an Air Force attack on the plants with different types of weapons, displaying the odds of success and showing how the mission might be accomplished without spreading radioactive fallout.

Impressive though the mock attack was, Kirchner and his colleagues knew that knocking out North Korea's nuclear capability was not as easy as it looked on a computer screen. An attack might disable North Korea's reactors. But what if Kim had one or two nuclear bombs hidden away already?

Thousands and thousands of tunnels lay beneath North Korean soil, dug by Pyongyang's leaders in the knowledge that war might some day drive them underground. If they had hidden a nuclear weapon down one of those tunnels, Kirchner believed, the only way to destroy the bomb with certainty would be to launch scores of U.S. nuclear weapons -- an unthinkable approach.

Perry had similar concerns. Back in Washington, as May turned to June, he studied very carefully one of the options modeled on Kirchner's computers: a precision, conventional military strike on North Korea's main nuclear reactor.

By mid-June, Perry reached two conclusions, he later recalled. First, he decided after examining the strike plan that the United States did indeed have the technical ability to take out North Korea's reactor without spreading radioactive fallout. But he also decided that he would not recommend such an attack to Clinton.

There were several reasons. South Korea, Taiwan and Japan all argued vehemently against any act that might provoke Kim into war. And Perry concluded that he had to take the wild threats of total war issued regularly by North Korea's press agency at face value. Perry decided that without confidence about what Kim might be willing to risk in conflict with the United States, it would be unwise to view his threats as mere bluff.

Perry decided instead on a step-by-step approach that would raise pressure on North Korea without, he hoped, starting a war. The Pentagon would send more than 10,000 additional troops to South Korea while signaling Pyongyang publicly that the reinforcements should not be seen as a sign of aggression. Meanwhile, planning for total war -- if North Korea misread the U.S. moves -- would continue.

On June 16, Perry and Luck returned to the White House to recommend this plan to Clinton.

But Kim confounded them again. That week, almost a month after turning away Nunn and Lugar, the dictator received in Pyongyang an American mediator of his own choosing -- former president Jimmy Carter. It was an unusual situation: The administration had given Carter information but did not make him an official emissary. As Perry sat in the White House briefing his military buildup plan, Carter telephoned from North Korea. Kim had agreed to freeze his nuclear program in exchange for U.S. diplomatic concessions, Carter reported. The former president said he would momentarily explain the whole deal live on CNN.

Perry, Luck, Gallucci, Vice President Gore, Clinton and others sprawled around a television set in the White House, watching Carter announce an agreement on behalf of a U.S. government that he did not formally represent -- and which had been ambivalent about Carter's visit to Kim.

The deal outlined that day took months to conclude. The United States agreed to provide North Korea with oil, diplomatic contacts and light-water nuclear reactors, which would not easily produce enough material for a nuclear arsenal. North Korea agreed to freeze its nuclear operations immediately, dismantle its dangerous reactors over a period of years and open its past activity to inspection after five or six years. And the IAEA was appointed to supervise the agreement -- in the name of the NPT.

The deal looked shaky, but on paper at least, it deprived North Korea of the large nuclear arsenal that Perry feared he could not deter. At a minimum, it gave Perry the interim peace he desired, if not the sort of reliable peace he might have preferred. The deal had a role for Blix too, though not the role he might have preferred -- by postponing special inspections for years, North Korea had broken the NPT's rules.

On July 9, Kim died. His son, Kim Jong Il, declared himself his father's successor. Perry knew little more about Kim Jong Il's analysis of nuclear deterrence after the Cold War than he had known of his father's.

The momentary threat of war in Korea had passed. But the challenge of rethinking what role nuclear deterrence would play in U.S. national security remained.

That much was evident in the very first chart in a briefing package Perry and his aides carried around Washington last fall. The charts described the conclusions of the Pentagon's Nuclear Posture Review, the first top-to-bottom study of the purpose of U.S. nuclear weapons in decades.

"Why We Have Nuclear Weapons," the chart declared, and it quoted Clinton as saying that the United States would retain enough nuclear weapons "to deter any future hostile foreign leadership" that armed itself with nuclear bombs. Washington's aim was to stop such an enemy "from acting against our vital interest and to convince it that seeking a nuclear advantage would be futile."

The United States would use its nuclear weapons, Perry's deputy, John Deutch, told Congress while presenting the same charts, to "hold at risk those assets valued by the leadership of the hostile state. . . . This is the essence of nuclear deterrence."

But this was also the essence of Washington's problem in North Korea just a few months earlier, at least as Perry saw it.

Nuclear deterrence required confidence about the "assets valued" by a nuclear-armed enemy. By threatening those assets with nuclear weapons, the United States might prevent war, or at least nuclear war. But if, like Kim Il Sung, enemy leaders were opaque -- or if they were crazy, valuing nothing, not even the future of humankind -- then nuclear deterrence collapsed.

At the Pentagon and at U.S. strategic nuclear forces headquarters in Omaha last fall, generals in command of U.S. nuclear forces began running another series of classified war games in search of solutions.

One of them, code-named "Nimble Dancer," posited simultaneous wars in the Persian Gulf and on the Korean peninsula. On the Korean front, a fictionalized Kim Jong Il launched chemical weapons at South Korean ports, closing off landing points and poisoning thousands of American soldiers. How might U.S. forces respond?

One option considered was to launch a few tactical nuclear weapons at North Korea in retaliation. But the game players could not agree that any U.S. president would ever authorize such a nuclear retaliation. Nor was a chemical riposte conceivable.

A second option was to launch ICBMs loaded with conventional warheads into North Korea. But the game players balked at being the first to launch an ICBM in anger since the nuclear arms race began -- even if the missile did not have a nuclear warhead. To reach North Korea from the United States, an ICBM would have to overfly China. What would the Chinese think? How could the United States be sure China's leaders would not misread the flight pattern as a nuclear attack on Chinese territory?

The war game continued for months. Some participants who came and went grew increasingly confused about the future of nuclear weapons. But that was not a complaint limited to military officers.

In Washington and scores of other capitals, diplomats and political leaders were equally uncertain. As they prepared for a historic conference that would decide the future of the NPT, they had to deal with a basic question about the evolving role of nuclear arms. What if there were 20 North Koreas? That was an issue few would have taken seriously 10 years earlier. Now delegates from more than 170 nations had to think about it -- and decide what was the best way to avoid it.

Copyright 4/95 The Washington Post