The Doomsday Machine

by Daniel Ellsberg

  Obama had been asked by an AP reporter whether there was any circumstance in which he would be prepared or willing to use nuclear weapons in Afghanistan and Pakistan to defeat terrorism and al-Qaeda leader Osama bin Laden. As USA Daily reported, “ ‘I think it would be a profound mistake308 for us to use nuclear weapons in any circumstance,’ Obama said, with a pause, ‘involving civilians.’ Then he quickly added, ‘Let me scratch that. There’s been no discussion of nuclear weapons. That’s not on the table.’ … When asked whether his answer also applied to the possible use of tactical nuclear weapons, he said it did.” (He meant for Afghanistan and Pakistan. He had elsewhere kept it on the table for Iran, like Clinton and John Edwards.) The AP account continued:

  Clinton chided her fellow senator about addressing hypotheticals.

  “Presidents should be very careful at all times in discussing the use or non-use of nuclear weapons … I don’t believe any president should make any blanket statements with respect to the use or non-use of nuclear weapons,” Clinton said.

  So what was obvious to the then front-runner in 2007—along with the generally agreed feeling that she had won this round—is that a real president, or someone qualified to be one, would not “telegraph” that he or she would not use tactical nuclear weapons in unilateral operations against guerrillas inside the territory of a politically unstable, nuclear-armed ally. Indeed, as Reuters paraphrased Hillary Clinton as saying in this exchange, “presidents never take the nuclear option off the table.”309 [emphasis added]

  That is undoubtedly what she meant to convey. And it is, simply, a correct statement about American presidents in the nuclear era, all of them so far.

  That holds not only for presidents, but also for aspirants to that office, including ambitious members of Congress. No major candidate in either party has ever been willing to undercut a current or future president’s “bargaining hand” by insisting that initiating or threatening to initiate a nuclear attack is not a legitimate “option” for the president of the United States or for any other national leader—for example, Vladimir Putin.

  This record was upheld in the most recent presidential campaign in 2016, during which Donald J. Trump’s repeated refusal to reject the option of first use occasioned considerable unfavorable comment, even consternation. Some interlocutors almost begged him to do so, in particular Chris Matthews, in a town forum he was moderating in Green Bay, Wisconsin, on March 30, 2016:

  MATTHEWS: Can you tell the Middle East we’re not using a nuclear weapon on anybody?

  TRUMP: I would never say that. I would never take any of my cards off the table.310

  MATTHEWS: How about Europe? We won’t use it in Europe?

  TRUMP: I—I’m not going to take it off the table.

  MATTHEWS: You might use it in Europe?

  [LAUGHTER]

  TRUMP: No, I don’t think so. But I’m not taking …

  MATTHEWS: Well, just say it. “I will never use a nuclear weapon in Europe.”

  TRUMP: I am not—I am not taking cards off the table.

  MATTHEWS: OK.

  TRUMP: I’m not going to use nuclear, but I’m not taking any cards off the table.

  Matthews’s pursuit of this issue, like that of nearly every other interviewer, seemed to reflect the simple, widespread ignorance of the reality that Trump was taking the same position of every president since Truman, and of every major candidate in that long period, definitely including his rival Hillary Clinton. She would surely have given essentially the same answers to Matthews’s questions as Trump did if she had been in that same forum, consistent with her stand in 2007. No candidate or president has ever come close to adopting and proclaiming a no-first-use policy (with Barack Obama being the only president311 to encourage serious internal consideration of it, especially in his last year, before rejecting it in face of opposition from his secretaries of defense, state, and energy and certain allies).

  Granted, other major candidates and presidents taking the same position have aroused less unease than Donald Trump, who, along with being unusually volatile and thin-skinned, has explicitly embraced a deliberate penchant for unpredictability, evident not only during the campaign but also while president. That was already on display in his exchange in March 2016 with Chris Matthews. On the one hand, he said, “I would be very, very slow and hesitant to pull that trigger.” But he followed that by asking, moments later, “Somebody hits us within ISIS, you wouldn’t fight back with a nuke?”

  Similarly, he said a bit further on, “I’d be the last one to use the nuclear weapon.” The assurance that might have provided was undercut by the sentence (widely known to be untruthful) that immediately preceded it: “I opposed Iraq.” Or this: “Look, nuclear should be off the table. But would there be a time when it could be used, possibly, possibly?” And if not, as Matthews’s response seemed to suggest, “Then why are we making them? Why do we make them?”

  Many mocked him for that question, though it seems a fair one. Many others shuddered at the implication that Donald Trump, presiding over a trillion-dollar makeover of our entire nuclear arsenal that he inherited as a program from Barack Obama, might feel that he could actually use some of these weapons. But of course he planned to use them, as he had clearly implied to Chris Matthews. He wants to use them like every other president: in “negotiation,” in threats, in exploiting uncertainty in our opponents as to whether he might launch “a nuke” in a stalemated armed conflict or a crisis, or perhaps in pique at what he experienced as humiliating provocation. Whether he would carry out such threats in any given circumstances, or otherwise use them in attacks, remains as uncertain, and as possible, as it has been for every other president in the nuclear era.

  Trump hinted strongly to Matthews, and he even came close to saying outright—“I’m not going to use nuclear, but I’m not taking any cards off the table”—that he would be bluffing. Most, if not all, of the time. Nevertheless, the last bargaining strategy mentioned above,312† advertising and exploiting his own unpredictability, deliberately creating uncertainty in an adversary by demonstrating impulsive, erratic, vindictive behavior—reminiscent, to many observers, of Nixon’s madman theory313—is especially worrisome to many in America and elsewhere because of a growing sense that this particular president actually may be mad.

  There’s ample evidence supporting that impression. Still, in some ways he has shown himself to be crazy like a fox, or he would not be president. He may yet elude his domestic pursuers and survive in office, and we and our democracy might survive that too. Or not.

  Yet what seems to me beyond question is that any social system (not only ours) that has created and maintained a Doomsday Machine and has put a trigger to it, including first use of nuclear weapons, in the hands of one human being—anyone, not just this man, still worse in the hands of an unknown number of persons—is in core aspects mad. Ours is such a system. We are in the grip of institutionalized madness.

  There is nothing new about that in human affairs. Among the aphorisms in Beyond Good and Evil, Friedrich Nietzsche wrote: “Madness in individuals is something rare; but in groups, parties, nations and epochs, it is the rule.” We Americans do have an unusual individual in the White House at this time. But both our parties, many nations, and this epoch are on track with Nietzsche’s rule. In the nuclear era, that means that we humans—above all, the nuclear weapons states and their allies—pose an imminent danger of near-extinction to ourselves and to most other terrestrial species.

  That ultimate outcome is threatened above all, as revealed by the nuclear winter studies, by the arsenals and policies of the two nuclear superpowers. Let me focus on our own country’s stance, as expounded by all major candidates in recent elections.

  First, it should be self-evident that so long as the U.S. government seeks to maintain the credibility of its first-use nuclear threats—both in declarations and more importantly by maintaining and “modernizing” a first-strike capability aimed at Russia that supports that cre
dibility but endangers most life—it cannot even participate in, let alone lead, a truly significant disarmament process or a campaign to delegitimize nuclear weapons’ possession and use. And without U.S. leadership—requiring a reversal of course by our government—no significant reduction in the danger to humanity from nuclear weapons can occur.

  Yet what is at issue here is more than the practical benefits of joining in the broad international consensus against initiating nuclear war, though that would seem urgent enough. It is important as well to regain a grasp of what might be called moral reality, a human perspective that transcends insiders’ obsession with agency, service, party, or national advantage. Plans and doctrines for the use of nuclear weapons, and reflexive, systemic resistance to the goal of eliminating them, raise questions about who we are—as a nation, as citizens, as a species—and what we have been doing and risking, what we have a right to do, or an obligation to do, and what we should not do.

  Speaking personally, I have always shared314 President George W. Bush’s blanket condemnation, under all circumstances, of terrorism, commonly defined as the deliberate slaughter of noncombatants—unarmed civilians, children and infants, the old and the sick—for a political purpose. On its face, that position is hardly controversial. Thus, for example, the destruction of the World Trade Center buildings with their inhabitants on September 11, 2001, was rightly recognized as a terrorist action and condemned as mass murder by most of the world.

  But in contrast, most Americans have never recognized as “terrorist” in precisely the same sense the firestorms caused deliberately by U.S. firebombing of Tokyo or Dresden or Hamburg or the atomic bombing of Hiroshima. These deliberate massacres of civilians, though not prosecuted after World War II like the Japanese slaughter in China at Nanking, were by any prior or reasonable criteria war crimes, wartime terrorism, crimes against humanity.

  Just like the bombs that destroyed Hiroshima and Nagasaki, any future attack by a single tactical nuclear weapon near a densely populated area would kill tens to hundreds of thousands of noncombatants, as those did. Thus, virtually any threat of first use of a nuclear weapon is a terrorist threat. Any nation making such threats is a terrorist nation. That means the United States and all its allies, including Israel, along with Russia, Pakistan, and North Korea.

  Indeed, going infinitely beyond any concept of terrorism or criminality, it is not merely a moral danger but a moral catastrophe that both Russia and America (with its NATO allies) are still threatening, deploying tactical nuclear weapons, carrying out exercises to execute first-use nuclear attacks against an opposing nuclear superpower “if necessary,” and implying readiness to impose on the rest of humanity a near-certainty of escalation to nuclear winter and omnicide. Speaking as an American, that must cease to be the case for the United States, and it cannot wait on others or, as of 2017, come too soon.

  To recover fundamental moral bearings, as well as to move urgently toward preserving human civilization and other life on this planet, the U.S. government—including the president, officials, and Congress, pressed by a popular movement and preferably backed by binding congressional legislation—should announce decisively that there is no “nuclear first-use option” on the bargaining table in our dealings with Russia, Iran, China, North Korea, or any other nation, because we as a people and our government recognize that nuclear first use would be a murderous, criminal action, not a legitimate “option” for the United States, Russia, or for any other country under any circumstances.

  a Endnotes present the more accessible references, generally after decades of secrecy and denial, from memoirs and other public sources to recently declassified documents.

  CHAPTER 21

  Dismantling the Doomsday Machine

  One way the RAND Corporation compensated the Air Force for the virtually complete freedom it had granted us to pursue our own self-generated research was for us to respond promptly whenever it occasionally asked us to evaluate some proposal from within the service. In 1960 my RAND secretary delivered a proposal from the Air Force for me to assess. It was, as usual, a photocopy of an original typed memo with, I believe, the title “Project Retro” from an Air Force officer. It had already gone through a number of Air Force offices. That was indicated by check marks and initials on a routing chart that was stamped on the first page, recording that it had been seen and in some way acted on by many of the agencies on the chart, Research and Development, Plans, Science and Technology, and so forth.

  There was also the usual routing chart within RAND. I seemed to be among the first in the building to see it, though it wasn’t obviously in my line—at first glance, it appeared to be more in the province of engineering—but I was known to be among those who were preoccupied with problems of SAC vulnerability to a Soviet surprise attack and the ability of our offensive forces to survive and retaliate.

  It was a classified proposal to deal with the possibility that a Soviet attack with ICBMs could eliminate our capability to retaliate with land-based missiles, primarily Minuteman ICBMs. This was in mid-1960, before the exposure of the missile-gap myth. I had been at RAND full-time for about a year, working on problems precisely like this.

  This scheme proposed in some detail to assemble a huge rectangular array of one thousand first-stage Atlas engines—our largest rocket propulsion engines, except for Titans, of which we had only a few—to be fastened securely to the earth in a horizontal position, facing in a direction opposite to the rotation of the earth.

  The officer originating this proposal envisioned that if our Ballistic Missile Early Warning System (BMEWS) radars detected and reported on the huge viewing screens at NORAD a large flight of missile warheads coming across the North Pole from the Soviet Union—aimed at our missile fields in North and South Dakota, Wyoming, Montana, and Missouri—the array of Atlas engines would be fired, as near simultaneously as possible, to stop the earth’s rotation momentarily.

  The Soviet missiles, on their inertial path, would thus bypass or overfly their intended targets. Our land-based retaliatory force would be saved, to carry out—presumably, when things had settled down and earth was again spinning normally—a retaliatory attack against the cities and soft military targets (their missiles having already left their hardened silos) in the Soviet Union.

  You didn’t have to be a geophysicist, which I wasn’t, to see some defects with this scheme. An awful lot of stuff would be flying through the air. Everything, in fact, that wasn’t nailed down, and most of what was as well, would be gone with the wind, which would itself be flying at super-hurricane force everywhere at once. Cities on the coasts and beyond would be wiped out by giant tsunamis as the oceans redeployed onto the continents.

  The Minuteman launch control officers, safe in their capsules deep underground, would have even less reason than in the foreseeable conditions of nuclear war either to launch their missiles or to come aboveground, since there would be nothing left to destroy on the surface of the Soviet Union, or the United States, or anywhere else on the planet. All structures would have collapsed, with the rubble along with all the people joining the wind and the water in their horizontal movement across the face of the earth, into space.

  All this was obvious enough. My first thought was, “Pretty funny.” It was the only piece of paper I had seen from the Air Force bureaucracy that showed a sense of humor. Even better, it was done perfectly straight, with no hint that it was anything but an ordinary secret official document. It looked absolutely authentic. I gave whoever had originated it (a RAND jokester?) credit.

  Then I looked again at the routing slip from the Air Force. It really did appear to have gone through a number of relevant official agencies and been passed on. Half the boxes were unchecked—it hadn’t gone to those divisions—but half acknowledged receipt. The signed initials were all different and looked real. No one had stopped it before it was sent to RAND, and I realized it was not a joke.

  I remember sitting at my desk, looking at that document, and as
king myself, for the first time: “Could I be in the wrong line of work?”

  I did show it to a couple of RAND colleagues to see if they had the same reaction. They both were dismissive of the scheme. One engineer made some rough calculations on the back of an envelope (RAND engineers really did do this literally on occasion, though there was a blackboard in every room) and said after a few minutes, “One thousand Atlas engines wouldn’t do it.”

  Another, a physicist, said, as I recall, “If you could actually muster enough power to stop the rotation for a second or so, it’s more likely that the earth’s surface would rupture from its core. The planet might break up.” Yes, Project Retro could surely be filed under “Crazy.”

  But the truth, in retrospect, was that most of the documents I read in my national security work, including many of those I wrote myself, were only marginally, if at all, less unbalanced than Project Retro. “Unbalanced” here being a euphemism for crazy, criminally insane.

  True, only Project Retro would have had the effect of wiping the surface of the earth clean of human structures, humanity, and all other terrestrial species, and dispersing the creatures of the lakes and ocean deeps to dry out, eventually, on what remained of the land.

  But, as I would soon discover, the Joint Chiefs’ estimates of the effects of carrying out their first strike plans, under a variety of circumstances, foresaw killing more than half a billion humans with our own weapons in a matter of months, with most of them dead in a day or two.

  How to describe that, other than insanity? Should the Pentagon officials and their subordinates have been institutionalized? But that was precisely the problem: they already were. Their institutions not only promoted this insanity, they demanded it. And still do. As do comparable institutions in Russia.

  RAND analysts, of whom I was one, sought to bring about less insane planning for nuclear war. We failed. That was in part because the civilian officials we were advising found it hard to get the military to adopt our proposals. But, in retrospect, our proposed strategies were totally unrealistic—as crazy as SIOP-62, SAC’s own plan in 1961. Or almost as crazy. Simply improving on current SAC war plans—the secret goal that a generation of RAND analysts, from Bernard Brodie to Kaufmann to me set themselves—was far too low a bar, and still held us prisoners within the realm of madness. If RAND recommendations, including my own, had been implemented by SAC in an actual nuclear war—in the way that SAC proceeded to interpret them and execute them operationally—they would still have resulted in total global catastrophe.