The Bomb: Presidents, Generals, and the Secret History of Nuclear War

by Fred Kaplan

American intelligence agencies were watching these Soviet moves, just as Soviet agencies were watching the Americans. Ordinarily, when the Soviets took such actions, the agencies would notify the Joint Chiefs or the secretary of defense or, if things looked particularly dicey, the president—and onward and upward the escalation might spiral. But a three-star general named Leonard Perroots, the deputy chief of staff for intelligence at the U.S. Air Force’s European headquarters in Germany, decided to do nothing. He’d been apprehensive about Able Archer 83, viewed it as needlessly provocative, and so interpreted the Soviets’ actions not as a threat but as a rational, defensive response to what they saw as an American threat. Six weeks earlier, Lieutenant Colonel Petrov, the Soviet air defense officer who decided not to report the detection of an apparent U.S. missile attack, might have averted war by violating protocol. Now Lieutenant General Perroots, the American intelligence officer monitoring Able Archer, may well have done the same.

In the decades since, partisans have debated whether to credit the Cold War’s end to Reagan’s hostile rhetoric and arms buildup in his first term or to his turn toward détente and disarmament in his second term. In fact, it took both—Reagan the super-hawk and Reagan the nuclear abolitionist—and, at least as important, the rise of Gorbachev as his collaborator. They were the most improbable leaders of their respective nations, and the great change could not have happened without the doubly improbable convergence of their reigns. Gorbachev needed to move swiftly if his reforms were to take hold; Reagan exerted the pressure that forced him to move swiftly and offered the rewards that persuaded his Kremlin rivals and skeptics to take the risks. History sometimes moves in serendipitous patterns. If Yuri Andropov’s kidneys hadn’t given out, or if Konstantin Chernenko’s heart had kept ticking a few years longer, Reagan’s bluster would have come to naught and might even have exacerbated tensions; the Cold War could have raged on for years. Similarly, if Reagan hadn’t been president in 1985—if Jimmy Carter or Walter Mondale had defeated him, or if Reagan had died and his vice president, George H. W. Bush, had taken his place—Gorbachev would not have received the push that he needed; those other American politicians were too traditional, too cautious, too sensible, to have pushed for such radical arms reductions or to take Gorbachev’s radicalism at face value.

For all practical purposes, Ohlert told Miller, the war plan contained only one option that was thoroughly conceived and rehearsed: MAO-4, the ultimate Major Attack Option—the option to destroy as many targets, with as many nuclear weapons, as quickly as possible. To the extent some smaller attacks were written into the plan, they weren’t “limited” by any reasonable definition of the word: the Soviets would interpret it as an all-out attack and retaliate accordingly. Sometimes JSTPS would play word games to protect the plan. Since McNamara’s first revision to the SIOP, in 1962, every defense secretary had ordered an option that let the president refrain from attacking Soviet or Chinese cities. But, Ohlert told Miller, if a nuclear war happened and if the president selected that option, he would wind up bombing cities anyway. There were two reasons for this. First, the actual “no-cities” option, as devised by JSTPS, excluded just twenty-four “urban areas”—described as Soviet and Warsaw Pact “regional political centers”—from the target list. That meant lots of other, very large cities, besides those twenty-four, would still get hit. Second, even those exempted cities would get hit to some extent because the JSTPS defined an “urban area” as a circle within which 95 percent of a city’s population lived.

One of the most blatant instances turned out to be the planning for Limited Nuclear Options. Miller asked an analyst in the Defense Intelligence Agency, the Pentagon’s own intelligence bureaucracy, how many discrete objects the Soviets’ early warning systems could track—how many incoming missiles it could detect as individual missiles—before they all merged into a vague blob on the radar screens. The answer was 200. In other words, on the Soviets’ radar screens, 200 missiles and 2,000 missiles looked exactly the same. And yet in SIOP’s smallest “limited” attack option, the United States would launch nearly 1,000 missiles. To Soviet air defense officers, this would look like an all-out attack.

The point was to let some time pass, to see if the war could be halted and the bombers called back to their base, before inflicting the most devastating damage. The old all-out attack option, which unleashed every weapon against every target, was removed from the SIOP. Nor were factories and other structures to be destroyed per se, unless they directly supported the Soviet war effort.

Conversations with Miller had primed Cheney to notice peculiarities in the war plan, and he was particularly struck by one slide showing that the SIOP’s main attack plan hit the Soviet transportation network with 725 nuclear weapons. Cheney asked Chain why. Chain looked at the colonel who had been running the briefing’s slideshow and reading the script. The colonel shrugged. Chain told Cheney he’d get back to him on that. Afterward, Cheney brought Miller into his office and asked him what was going on. Miller said he didn’t know: Weinberger and Carlucci had given him a mandate to investigate whether the SIOP reflected the president’s policies, but the Joint Staff had barred him from examining targets—either the individual targets themselves or the way JSTPS calculated how many weapons to hit them with. Cheney told Miller that the bar was hereby lifted and had him write a memo, which Cheney then signed, ordering the SAC commander to give Miller and his staff access to all documents that they requested.

Klinger noticed redundancy throughout the calculations in the Blue Book, and, now that he was in the belly of the beast, he searched for more examples. He had no trouble finding them. He was especially struck by the way the plan went after the Soviet army’s tanks. The SIOP aimed a lot of weapons not only at the tanks themselves, but also at the factory that produced the tanks, the steel mill that supplied the factory, the ore-processing facility that supplied the steel mill, and the mine that furnished the ore. Klinger asked why they needed to destroy the entire production chain. His liaison officer told him that, if they left any part of the chain intact, the Soviets could reconstruct the factory and rebuild tanks. Though Klinger didn’t know it, this was the same unfathomable frame of mind that, back in the early 1960s, led General Curtis LeMay to insist on destroying not just the repair depot at a Soviet air base but also the factory that built the spare parts supplying the repair depot because, otherwise, the Soviets could reconstruct their air force—never mind that thousands of American bombs and warheads would have just reduced the entire country to a smoldering radioactive ruin.

From talking with these analysts and perusing the Blue Book, Klinger unraveled the mystery of why the SIOP laid down 725 nuclear weapons on the Soviet transportation network. It turned out that JSTPS had decided, for unclear reasons, to launch nuclear weapons against all railroad yards above a certain metric capacity and all railway bridges that stretched for longer than a certain distance. But, as the Defense Intelligence Agency’s logistics specialists told them, this standard was completely arbitrary; it had no bearing on the military value of a target. Some long bridges and large railroad yards weren’t used by the military at all; some very short bridges and small rail yards were militarily vital. There were other absurdities. For instance, the loading zones on the Russia-Poland border were clearly vital military targets: Russian and Polish rail tracks had different gauges; the loading zones were where Soviet armored vehicles would be switched from one track to the other before proceeding to the front lines of a war with NATO. The SIOP did target those loading zones—but it left unscathed some road bridges a couple miles up the river, because the JSTPS had called for attacking the Soviet “rail system,” not the “transportation system.” Bridges on roads weren’t targeted at all.

As he delved more deeply into the war plan, Klinger saw that much of it made no sense on any grounds. At one point, he had Wayne Lumsden print out the latitude and longitude of all the targets in the SIOP database. He then asked the graphics department to plot all those points on a huge map of the Soviet Union and draw circles around each point, signifying the area that would be destroyed by the specific weapon aimed at each target. The result was a vivid, color-coded display of massive death and destruction. It showed a staggering number of bombs and warheads hitting nearly every city in overlapping density. Hundreds of targets were so close to one another that, in many cases, one weapon would hit more than one target, and one target would get hit by three, five, as many as seven nuclear weapons. Klinger wondered how many weapons were aimed at Moscow and its environs. On the map, he drew a fifty-mile circle around the capital and counted the number of weapons that were aimed at targets within the circle. There were 689 of them, many releasing more than a megaton of explosive power. “Overkill” was a severe understatement for what had gone on for decades.

During one of his early trips to Omaha, Klinger asked one of the officers at JSTPS to analyze whether the treaty’s prospective cuts would affect their ability to fulfill their mission—whether they could continue to deter nuclear war and limit damage if deterrence failed. The officer replied that he didn’t do that sort of analysis. Klinger, thinking he wasn’t making himself clear, rephrased the question. The officer said that he understood the question perfectly well. He explained that JSTPS was prohibited from setting requirements or analyzing whether a certain kind of attack, with a certain number of weapons, would be militarily effective. This prohibition had been laid down in 1960, with the first SIOP, when Navy admirals feared that Air Force generals would use the war plan as a tool for controlling the Navy’s arsenal and budget. Klinger was astonished. The officers who drew up the war plan didn’t know—weren’t allowed to figure out—how many weapons they needed to protect the nation? What did they do then?

The implication was stunning. Back in McNamara’s time, the whiz kids had famously asked, “How much is enough?” The chief whiz kid, Alain Enthoven, later coauthored a book, a combination treatise and memoir, with that question as its title. Now, it turned out that nobody had ever asked the question in a way, or from a position, that mattered. The authors of the Defense Department’s directives on how to use (and not use) nuclear weapons may have thought they were asking the question; but they’d lacked access to the machinery of the SIOP—they hadn’t known how to ask the question in a way that the JSTPS officers could translate into their operational plan. Meanwhile, the JSTPS officers were barred from asking, or trying to answer, the question on their own. Washington and Omaha were running on parallel, sometimes wildly divergent tracks—which had been Curtis LeMay’s intention all along: he didn’t want civilians, or even rival branches of the military, meddling with his war plan, and he designed SAC explicitly to keep them out.

Powell concluded from the exercise that war with the Soviet Union probably would trigger nuclear war; that nuclear weapons, once used, couldn’t be controlled; and that, therefore, war had to be prevented at all cost. This was one problem Powell had with the whole business of Limited Nuclear Options. He told Miller, during one of their sessions, that there was nothing to support the idea that we could drop a few nukes on Soviet territory and say, “Let’s stop here.” There was nothing in the intelligence reports to suggest that the Soviets would play the same game—that they would do anything but retaliate, and with more than just a few nukes of their own.

In the end, they cut the requirements for another couple thousand bombs and warheads, leaving the number of strategic weapons at 3,500—the lowest number since 1962, when the nuclear arms race began.

On January 3, 1993, two and a half weeks before the end of Bush’s presidency, he and Boris Yeltsin signed the START II treaty, which prohibited land-based ICBMs from carrying more than one warhead apiece. The accord reversed the most destabilizing action in the history of the arms race. ICBMs loaded with MIRVs—multiple warheads, each of which could strike widely separated targets—were at once the most potent and the most vulnerable weapons in both sides’ arsenals. In a crisis, a desperate or risk-prone leader might be tempted to launch a first strike, if just to preempt the enemy from launching a first strike. In short, the very existence of MIRVs on ICBMs created a hair-trigger situation, in which both sides might feel an incentive to strike first. START II would remove that incentive.

Carter arrived in Seoul on June 13, crossed the Demilitarized Zone into North Korea soon after, and spent the next three days locked in conversation with Pyongyang’s leaders, including Chairman Kim Il-sung, eighty-two years old but still alert. As it turned out, both sides in the White House debate over whether Carter should make the trip had been right. After the long hours of meetings, meals, and a ride on Kim’s yacht, the North Korean leader backed down from his threats. And Carter went way beyond his instructions, negotiating the outlines of a twelve-point treaty without any authorization from—in fact, against the express directives of—the Clinton administration. Not only that, but as his trip was wrapping up, Carter called Gallucci, who happened to be at an NSC meeting in the White House Cabinet Room, to tell him that he’d reached an accord with Kim and would be announcing some of its terms in five minutes on CNN. (This was another surprise to Clinton’s people: Carter had brought along a camera crew from the cable news network.)

If Powell was embarrassed by Bush’s stance, Kim Dae-jung was humiliated. KDJ, as Korea-watchers called him, was a new type of South Korean leader, a democratic activist who, during his country’s authoritarian period, had spent years in prison for his political beliefs and had run for president promising a “sunshine policy” of opening up relations to the North. During the Clinton years, South Korea’s ruling party had been implacably hostile to the North; efforts to hold serious disarmament talks were obstructed at least as much by Seoul’s sabotage as by Pyongyang’s manipulations. Now South Korea had a leader who could be a partner in negotiations—but the United States had a leader who was uninterested in negotiating. In Bush’s view, articulated openly by his vice president, Dick Cheney, to negotiate with an evil regime was to recognize that regime, legitimize it, and—if the talks led to a treaty—prolong its reign. Bush told a reporter, on the record, “I loathe Kim Jong-il.” He held Kim Dae-jung in contempt as a naïf. KDJ had also committed what Bush took as a personal insult: before his Washington trip, he’d met with Russian president Vladimir Putin and issued a joint statement endorsing the preservation of the Anti-Ballistic-Missile Treaty.

In any case, by the early fall of 2002, signs of an impending enrichment program were unambiguous. On October 4, James Kelly, the assistant secretary of state for East Asian and Pacific affairs, flew to Pyongyang to confront officials with the evidence. To his surprise, they admitted the charge was true. But then, to the North Koreans’ surprise, for the next two weeks the Bush administration kept the meeting—and the evidence—secret. The Senate was debating a resolution to give Bush the authority to go to war against Iraq. The public rationale for war was that Iraq’s dictator, Saddam Hussein, possessed weapons of mass destruction. If it was known that North Korea was also making WMDs—and nuclear weapons, at that, not chemical or biological weapons, as was asserted (mistakenly, it turned out) about Saddam—it would have muddied the debate. Some would have asked why Bush saw a need to go to war against Iraq but not against North Korea. The Senate passed the Iraq war resolution on October 11. The Bush administration publicly revealed what it had known for months about North Korea’s enriched-uranium program six days later, on October 17.

When their approach to Richardson led nowhere, the North Koreans escalated the pressure. Over the next two weeks, U.S. spy satellites detected trucks pulling up to the site where the fuel rods were stored, then driving away toward the reprocessing facility. When Kim Il-sung threatened to take this step in 1994, Clinton warned that it would cross a “red line.” When Kim Jong-il actually did it in 2003, George W. Bush said and did nothing. Specialists inside the State Department and the Pentagon were flabbergasted. Once those fuel rods left the storage site, once reprocessing began, once plutonium was manufactured, the strategic situation would change: even if the North Koreans could be lured back to the bargaining table, even if they agreed to drive the fuel rods back to the storage site, no one on the outside could be certain that they’d completely disarmed; no one could ever know if they’d manufactured some plutonium and hidden it in a bunker.

This was the Bush-Cheney-Rumsfeld line: as long as the North Koreans were pursuing nuclear weapons, even to sit down with them would be an act of appeasement, succumbing to blackmail, rewarding bad behavior. Cheney once put it this way: “We don’t negotiate with evil; we defeat it.” Cheney was the prime mover in all these campaigns for regime change and for the resistance to diplomacy. Those who knew him from his days as secretary of defense, during the presidency of Bush’s father, were puzzled by his evolution from conservative but competent manager to blazing ideologue. Some attributed his tectonic shift in views to the 9/11 attacks; others thought the end of the Cold War, leaving America as “the sole superpower,” galvanized his imperial ambitions; still others wondered whether his heart attack shortly after the 2000 election—his fourth—might have finally done some slight damage to the flow of oxygen to his brain. Colin Powell, who had worked alongside him in the Pentagon, told friends that Cheney hadn’t changed at all: the difference was that, in his former job, he’d had “adult supervision”; now, with George W. ceding him authority on national security matters, he was in charge.

Soon after the nuclear explosion, Bush got serious about arms control talks with Pyongyang, largely at the urging of Condoleezza Rice, who was now his secretary of state. In the past, Cheney and Rumsfeld would have joined forces to quash such a move, but around this time Bush forced Rumsfeld to resign—for many reasons, most of them having to do with his failed policy in Iraq, which was now consumed by a sectarian civil war and an anti-American insurgency. As a result, Cheney found himself increasingly isolated. In early 2007, Bush let an assistant secretary of state, Christopher Hill, hold bilateral talks with North Korea in Berlin—the sort of talks that he had prohibited back when Kim Jong-il was merely threatening to detonate an atom bomb. Hill and his counterpart even signed a disarmament accord, but it was shot through with so many loopholes as to be meaningless. The restraints that Bill Clinton had imposed with the Agreed Framework, and had come close to nailing down in the talks on a missile ban, snapped off and blew away, perhaps forever. For the first time in nearly forty years, nuclear proliferation was a front-burner threat.

Obama thought it was crazy for any American president to use nuclear weapons first, under any circumstances. For one thing, given the U.S. military’s conventional superiority, it wasn’t necessary. For another, he had read enough history to know that the “nuclear taboo,” as some called it, was real: the destructiveness of the bomb was too enormous; the risk of escalation to global catastrophe was too high. Still, Obama recognized the distinction between believing in no-first-use and declaring it as national policy, and, in that context, he saw that Gates had a point.

The only ones that stood out, in that sense, were Russia and North Korea. The U.S. already had a well-defined security arrangement—an acknowledged state of mutual deterrence—with Russia. That left North Korea, and the one distinctive thing about North Korea was that, several years earlier, it had withdrawn from the Non-Proliferation Treaty. Obama suggested this for a policy: the United States will not use, or threaten to use, nuclear weapons against countries that had no nuclear weapons and that were abiding by the NPT. Neither he nor the others in the room knew or remembered, but thirty years earlier, during Jimmy Carter’s presidency, Paul Warnke, director of the Arms Control and Disarmament Agency, and his deputy, Spurgeon Keeny, had come up with the same idea and summarized it in a speech for Secretary of State Cyrus Vance. (A decade before then, back in 1965, Keeny had been staff director of the Gilpatric Committee, which first proposed an international treaty to stop the spread of nuclear weapons.) But the name they gave the idea—“negative security assurances”—was the opposite of catchy, and it went nowhere.

And so Obama ordered a detailed review of the SIOP. The idea was to delve into the nuclear war plan at least as deeply as Miller had—with one difference: this time, the civilians faced little resistance from the four-star general who led what was now called Strategic Command. The commander was Robert Kehler, and the difference between him and Jack Chain, the general who tried to block Frank Miller’s intrusions (and who said he needed ten thousand weapons because he had ten thousand targets), was enormous. They were a vast generation apart: Chain, rising through the ranks as a fighter pilot in Korea and Vietnam; Kehler, twenty years younger, coming up as a missile officer, then as a senior officer in the Air Force’s Space Command, and picking up political skills as a legislative liaison in the interim. More relevant, Kehler had served in the Joint Staff’s nuclear and chemical directorate when Frank Miller was waging his assault on the SIOP; and so he saw, in greater detail than most of his fellow officers, the full scope of bloat in the nuclear plan and thus in the nuclear arsenal. When he arrived as head of StratCom in 2011, two years into Obama’s term, he recognized that, while the bloat had thinned out considerably, there was still a lot of overkill; there was room for deeper cuts.

This led to a deeper debate over just what the point of the arsenal was. If the “sole” or “fundamental” purpose of nuclear weapons was to deter an adversary from launching a nuclear attack on the United States and its allies, then there was plenty of leeway for miscalculation and error; you could pass up a lot of Russian targets, and therefore cut a lot of American weapons, without degrading deterrence. But if you were concerned about what happened if deterrence failed, if nuclear war erupted, then there might be less leeway; and, in that case, you had to figure out what you wanted to do—how you wanted to fight this nuclear war. If you wanted to destroy the enemy’s remaining nuclear weapons and other military sites, then you would have less room for error. Tell me what you want to do, Kehler said to the group on a few occasions. Then we can figure out how many weapons we need.

Iran was working out very differently. Talks between Iran and the P5+1 group—the five permanent members of the Security Council (the United States, Russia, China, Britain, and France) plus Germany—were making slow but substantial progress. After two years of negotiations, they signed a meticulously detailed, 159-page accord with the most intrusive inspection procedures of any nuclear arms control pact in history, blocking all of Iran’s possible paths to a nuclear weapon. In exchange for dismantling its nuclear programs, the P5+1 nations would lift economic sanctions that they’d imposed on Iran after discovering its covert uranium enrichment plants, which were illegal under the Non-Proliferation Treaty. (Other sanctions, which the U.N Security Council had passed to penalize Iran for testing ballistic missiles and supporting terrorists, remained in place.) On August 5, 2015, as part of his campaign to promote the deal against strenuous opposition from the Israeli government (though not most Israeli security officers), from Sunni Arab leaders (who, like the Israeli political leaders, wanted to wage war on the Shiite government of Iran), and from most Republicans in the U.S. Congress (who opposed any diplomatic triumph for the young, black Democratic president), Obama delivered a speech at American University in Washington, D.C., drawing parallels with President Kennedy’s speech there fifty-two years earlier. Obama portrayed the Test Ban Treaty, which grew out of that earlier ...

Then Colin Kahl, Vice President Biden’s national security adviser, raised his hand. Kahl wasn’t a specialist on nuclear matters; in Obama’s first term, he’d worked in the Pentagon as the chief civilian official on Middle East affairs, and before then, he’d taught at Georgetown University’s School of Foreign Service. But Kahl thought the generals were missing the big picture. The minute the Russians drop a nuclear bomb, he said, we would face a world-defining moment—the first time an atom bomb had been used since 1945. It would be an opportunity to rally the entire world against Russia. If we restricted our response to conventional combat and diplomatic ventures, we could isolate and weaken the Russian leaders, policies, and military forces. However, if we responded by shooting off some nukes of our own, we would forfeit that advantage and, more than that, normalize the use of nuclear weapons. Breedlove seemed confused. He understood the debate over whether the United States should be the first to use nuclear weapons in response to a conventional attack; but it seemed perverse to consider using conventional weapons in response to a nuclear attack.

Trump, who was also known as the author of a ghostwritten best-selling book called The Art of the Deal, leaned in and shared his magic. Arrive late at the first session, he told Burt. Walk up to the main guy on the Russian side of the table, stick your finger in his chest, and say, “Fuck you!” Just over two years later, in July 1991, Burt managed to get a treaty signed, reducing each side’s nuclear arsenals by one third, without following Trump’s suggestion. That same month, Trump filed for bankruptcy—his third, this one on his Taj Mahal hotel-casino in Atlantic City.

On the Fourth of July, in what he called an Independence Day “gift to American bastards,” Kim launched a missile that, had it followed a different flight path, could have hit the West Coast of the United States. Three weeks later, he followed up with tests of another potential ICBM as well as four short-range missiles, which landed in the Sea of Japan. On August 8, ten days after those tests, Trump told reporters outside the clubhouse of his golf course in Bedminster, New Jersey, “North Korea best not make any more threats to the United States,” or else “they will be met with fire and fury like the world has never seen.” Kim Jong-un, he said, “has been very threatening, beyond a normal state, and, as I said, they will be met with fire, fury and, frankly, power, the likes of which the world has never seen before”—repeating the line, almost word for word, as if he’d rehearsed it after being told that this was how President Harry Truman described the effects of the atomic bomb that leveled Hiroshima.

In short, Mattis didn’t think there were any good military options in Korea, and he didn’t want to foster the illusion that there were. Mattis’s resistance infuriated McMaster, who was Trump’s point man on the task. The two officers did not get along. As the president’s national security adviser, McMaster thought he should be regarded as an equal by the secretary of defense, but he felt that Mattis treated him the way a four-star general treats a three-star—as a subordinate. McMaster too had been a decorated wartime commander—the leader of a tank squadron in the first Gulf War, a master counterinsurgency tactician in the Iraq War—but he never commanded a unit larger than an armored cavalry regiment, which consisted of about 3,500 troops. By contrast, Mattis had been commander-in-chief of U.S. Central Command, overseeing 200,000 personnel from all the services in military operations across the Middle East, North Africa, and South Asia. Mattis had also been senior military adviser to a secretary of defense, while McMaster had no prior Washington experience.

But Mattis also heard two new arguments. One was floated by retired General Bob Kehler, the former head of StratCom who had helped Obama’s security team plumb the nuclear war plan and who had been the main author of the letter to the Wall Street Journal defending the Triad. Kehler pointed out that, since President George H. W. Bush’s unilateral initiatives, none of America’s bombers had been on runway alert and few were loaded with nuclear bombs. If there were a nuclear war, it would probably be preceded by spiraling tensions, which would give the president time to put the planes on alert—though it was possible he wouldn’t take that step, knowing it might escalate tensions. In any case, on a day-to-day basis, the United States already had a dyad; eliminating the ICBMs would leave just one leg, the submarines, and if some technological breakthrough made the submarines vulnerable, the entire nuclear arsenal would be obsolete. Kehler’s argument—which he had outlined to a few previous defense secretaries from his perch at StratCom—had appeal to senior officials whose jobs swayed them toward extreme caution about security. The second argument, which had germinated inside the Pentagon’s Joint Staff, was known as the “sponge theory.” The idea was that, without the ICBMs, there would be only five “strategic” targets in the continental United States—three bomber bases (in North Dakota, Missouri, and Louisiana) and two submarine ports (in Bangor, Washington, and Kings Bay, Georgia)....

In this light, Miller was trying to shift the question from “Why produce these weapons?” to “Why not?” Gottemoeller’s objections were hardly radical. The previous March, in hearings before the House Armed Services Committee, General Hyten, the head of StratCom, was asked whether he needed new low-yield nuclear weapons to deal with the Russians’ “escalate to de-escalate” threat. Hyten replied that his arsenal, as it stood, had “a number of capabilities” that would provide the president with “a variety of options to respond to any number of threats.” He went further a few weeks later, at a speech to military reporters, assuring them that he was “very comfortable with the flexibility of our response options.” At the same House hearing, General Paul Selva, vice chairman of the JCS, noted, in answer to the same question, that the U.S. already had nuclear-capable aircraft in Europe, fitted with low-yield bombs, which would also “deter the Russians from escalating to nuclear warfare” if they found themselves losing a conventional war. Nonetheless, in the end, Miller won the debate, in part because he helped write the document.

The language of this discourse also obscured the fact that the “low-yield” Trident II warhead wasn’t, in the scheme of world history, so low. In fact, it would wreak more destruction than most people alive had ever witnessed from a single explosion. The conventional bombs that leveled buildings in Iraq, Afghanistan, and elsewhere in the first two decades of the twenty-first century had the explosive power of 2,000 pounds of TNT. The low-yield Trident II warhead would explode with the blast power of 8 kilotons—meaning 8,000 tons, or 16,000,000 pounds—plus the heat, smoke, and radiation that would spread like toxic wildfire.II

As the hearing went on, it became clear that the debate was ultimately beside the point. Kehler noted that “nuclear decision-making at the highest level” is “a consultative process,” involving many people—the head of StratCom, the secretary of defense, and other advisers—any one of whom could say, as he put it, “Wait, stop, we need to resolve these issues.” But, he also acknowledged, if the president wanted to launch one of the many pre-set attack options that were listed in the nuclear war plan, the fact that the option was in the plan meant that the military lawyers had already vetted it. Ultimately, he acknowledged, the “decision authority”—the legal power to launch nuclear weapons—“resides with the President.” Senator Markey, who had prompted the hearing, then made what he considered the central point: Absent a nuclear attack upon the United States or our allies, no one human being should have the power to unilaterally unleash the most destructive forces ever devised by humankind. Yet, under existing law, the President of the United States can start a nuclear war without provocation, without consultation, and without warning. It boggles the rational mind. I fear that, in the age of Trump, the cooler heads and strategic doctrine that we once relied upon as our last best hope against the unthinkable seem less reassuring than ever.

It was sobering to some senior officers who watched the hearing that not a single senator, not even any of the Republicans, bothered to dispute the claim, openly expressed by some of the Democrats, that President Trump was “unstable” or that his very presence in the Oval Office—his everyday access to the button—made nuclear war more possible. General Kehler, sitting as a witness, may have been most flustered of all. It was up to Congress to set the rules on the procedures and safeguards for the use of nuclear weapons. If the members of Congress thought the president was unfit to command, then they needed to take responsibility and change the procedures; they couldn’t just toss it in the laps of the military. Worse still would be if they raised provocative questions about the president’s fitness, then did nothing, leaving the generals with a chain of command that no one could trust. Yet, he thought to himself, that is what this committee was doing.

Nixon had told Henry Kissinger to warn the North Vietnamese that he was crazy and might use nuclear weapons if they didn’t end the war. The ploy failed, possibly because the North Vietnamese didn’t believe Nixon was mad in quite that way. Did Kim fear that Trump might really be a madman? President Moon certainly feared he might be. American officials put out the word—this time without pretending—that Trump was erratic, unpredictable, different from previous presidents, though all this was clear without these messages. This was what led Moon to accept Kim’s overtures at once, to schedule a summit, to urge Kim to ask for a separate meeting with Trump, and to urge Trump to accept the invitation.

In other words, after Trump’s most bellicose remarks, Kim conducted four tests honing his ability to strike America and its allies. Far from appearing intimidated by Trump’s threats, Kim seemed to be moving steadily along his path to acquiring a nuclear arsenal.

Months before his summit with Kim, Trump had withdrawn from the Iran nuclear deal, a highly detailed accord that blocked the mullahs of Tehran from every possible path toward acquiring a nuclear weapon, calling it “the worst deal ever,” even though Jim Mattis, after reading the text three times, characterized its verification protocols as “robust” and even though the International Atomic Energy Agency attested, repeatedly, after several inspections, that Iran was abiding by its terms.

In June 2018, shortly before his first summit with Kim Jong-un, Trump met in the White House with Shinzo Abe, Japan’s prime minister. Abe proposed that he and Trump get their respective national security teams to work out a joint negotiating strategy. H. R. McMaster, who’d heard the suggestion ahead of time, was enthusiastic about the idea, as was much of his staff: the two teams had similar views; the two countries had common interests. But Trump waved away the idea. There are two things you need to know, he told Abe. First, he said, “I am the greatest negotiator in the history of the White House.” Second, when he was about to make a big deal, he went on, he didn’t prepare with a lot of documents; he went into the room and let his gut take a reading of the guy across the table. For Abe, this was a turning point; he suddenly realized that, as long as Trump remained president, Japan could no longer count on the United States as a reliable ally.

Most presidents come to realize that nothing in their lives has quite prepared them for the weighty rigors of the Oval Office. Most react to the shock by hunkering down with their briefing books, listening to experts, convening their top advisers, consulting their predecessors, asking questions, balancing options, and developing a keen awareness of their own styles of thinking: their strengths and limits. Trump seemed never to reach that moment of awakening. He never thought he needed to learn much beyond what he already knew. To Trump, everything was a deal, deals depended on personal relations, and he navigated the nexus between the two through the finely tuned nerve endings of his gut. The bomb—the weapon of fire and fury that Trump threatened to rain down on North Korea to punish its leader and obliterate the entire nation—has coexisted with humanity for three quarters of a century with no catastrophes, as yet, since the two explosions, over Hiroshima and Nagasaki, that set the nuclear age in motion. Still, all along, the generals and their aides have kept churning out war plans and calculating the consequences on our side versus their side—the targets destroyed, the cities ravaged, the millions of people killed—spinning them in a way that might yield some kind of victory. They have seen this as their job: to seek ways to make these weapons usable, in case the commander-in-chief wanted to use them.

But Trump was a symptom, not the cause, of our common nuclear problem. At the dawn of the Republic, James Madison wrote in The Federalist No. 10, “Enlightened statesmen will not always be at the helm,” which is why he and the other Founders devised checks and balances to a potential autocrat’s power—a legislature, judiciary, free press, and (they hoped) an educated public. At the dawn of the nuclear age, Harry Truman wrested control of the bomb away from the generals, entrusting it to the top civilian authority, because he understood that, as he put it, “this isn’t a military weapon.” But neither Truman nor anyone since devised checks or balances to the fears, whims, or impulses of the unenlightened statesman whose rise Madison feared as inevitable. Just months after Hiroshima, Bernard Brodie, one of the pioneering nuclear strategists at the RAND Corporation, a scholar who lacked the ambition or temperament to move into official circles, wrote, “Everything about the atomic bomb is overshadowed by the twin facts that it exists and that its destructive power is fantastically great.” The story of the bomb ever since has been—and will continue to be—the story of presidents and generals grappling with these twin facts, figuring out the best way to maneuver around them or to reconcile with them.

With the spread of the bomb came a logic—a stab at a strategy—on how to deter its use in warfare. The logic involved convincing adversaries that you really would use the bomb in response to aggression; part of that involved convincing yourself that you would use it, which required building certain types of missiles, and devising certain plans, that would enable you to use them—and, before you knew it, a strategy to deter nuclear war became synonymous with a strategy to fight nuclear war. And when crises arose, the logic encouraged, almost required, escalating the cycle of threats and counterthreats, just up to the point where deterrence and war converged, in order to maintain credibility. The compelling, and frightening, thing about the logic was that, once you bought into its premises, you fell into the rabbit hole; there seemed no exit. The presidents who fell deep into this hole, who faced the abyss where the logic led, avoided its end point—avoided war—by scrambling out of the hole, snapping out of the logic, like snapping out of a bad dream.