The Doomsday Machine

by Daniel Ellsberg

  The question was, would Gilpatric be allowed to say it—to make such a revelation—given that the administration had not chosen so far to do so? Tim had been given no assignment to make such a revelation, and his draft when I first saw it was the standard Pentagon boilerplate that every official in the Defense Department put in almost every speech—about our military buildup and how much we had added to the reserves and our offensive forces, plus some talk about Berlin. In adding, as he did, nearly all my points, he totally changed the tone and bearing of his draft. Some of my language didn’t get in, but this did:

  Our confidence in our ability to deter Communist action, or resist Communist blackmail, is based upon a sober appreciation of the relative military power of the two sides. We doubt that the Soviet leadership has, in fact, any less realistic views, although this may not always be apparent from their extravagant claims. While the Soviets use rigid security as a military weapon, their Iron Curtain is not so impenetrable as to force us to accept at face value the Kremlin’s boasts.

  The fact is that this nation has a nuclear retaliatory force of such lethal power that an enemy move which brought it into play would be an act of self-destruction on his part.

  That was the key point. My intended message was, for informed ears in the Kremlin and NATO, “We’ve discovered they are bluffing!” And for the American and European public: “We’re staying in Berlin, and there’s going to be no war.” I thought of it as calling Khrushchev’s bluff. I even rammed it home:

  The Soviet’s bluster and threats of rocket attacks against the free world—aimed particularly at the European members of the NATO alliance—must be evaluated against the hard facts of United States nuclear superiority which I discussed earlier.

  And with my new confidence that U.S. patrols along the Berlin corridors would not be obstructed, I felt free to underline our commitment in the final paragraph:

  The United States does not seek to resolve disputes by violence. But if forceful interference with our rights and obligations should lead to violent conflict—as it well might [though I no longer believed this]—the United States does not intend to be defeated.

  Gilpatric gave the speech on October 21, 1961,89 including these passages and others written by me, which were then quoted by the New York Times report of the speech. In fact, all the passages quoted there and in virtually every other journalistic or scholarly account since then were among those above, proposed and written by me.

  In line with the subtitle to this book, a confession is in order here. For decades after my work in the sixties on nuclear planning, I would have said that I had never proposed or been party to a threat of a nuclear first strike or first use in a crisis. I’m confident I could have passed a lie detector test on that assertion. Yet that would have been false. What else was I saying in my draft passages for the Gilpatric speech but that if the Soviets blocked our enlarged patrols along the Berlin corridors with some of their armored divisions in the neighborhood, they would have been taking an unacceptable risk of U.S. first use of nuclear weapons against those forces. Moreover, I was implying that we could do so in confidence that the Soviets would not respond with their own plentiful short-range nuclear weapons, because we would then exploit our “nuclear superiority” in strategic weapons to disarm and destroy the Soviet Union itself.

  How could I have failed to notice or recall, over the years, these implications of my own speech-drafting in the fall of 1961? Well, I have to conclude, the same way most humans manage not to recognize or remember discordant or unpleasant aspects or consequences of their own behavior. Like everyone I worked with (with the possible exception of Abe Chayes), I wanted to hold on to West Berlin. At the same time, like my closest colleagues, I would have been appalled to achieve this goal by initiating nuclear war on any scale. Yet—without making a deal with Khrushchev to recognize East Germany, something not within my ken—there was never any way to safeguard Berlin from Soviet conventional and nuclear-armed forces in East Germany except to threaten nuclear weapons and express a readiness to escalate to a nuclear first strike.

  So far as I was concerned, that ought to have been a total bluff. But in the giddy euphoria of the new intelligence, it seemed to me a bluff that was sure to work. That made it easy for me not to notice, or to forget, that it was, after all, a first-use and first-strike nuclear threat.

  That did not go unnoticed in the Soviet Union. The day after Gilpatric’s speech, Minister of Defense Rodion Malinovsky told the Twenty-Second Congress of the Communist Party in Moscow that

  [Gilpatric had] addressed a meeting of the Business Council90 in Virginia, presumably not without President Kennedy’s knowledge, and, brandishing the might of the United States, threatened us with force. What is there to say to this latest threat, to this petty speech? Only one thing: the threat does not frighten us!

  They are threatening to reply with force to our just proposals for a Germany peace treaty and the ending of the abnormal situation in West Berlin.… A realistic assessment of the picture would lead one to believe that what the imperialists are planning is a surprise nuclear attack on the U.S.S.R. and the socialist countries.

  I was gratified to see the Soviet defense minister responding so promptly and directly to my words, and so defensively, as it seemed to me. His claimed interpretation of what I had written went right past me, as Soviet hyperbole. After all, I knew, and I supposed he knew, that we had no intention or plans for a “surprise nuclear attack.” The speech said nothing explicitly about U.S. first-strike capabilities or conceivable intentions. And I, the drafter of the offending comments, had no desire to see nuclear war under any circumstances at all.

  What was the “threat” Malinovsky was complaining about? In his words, it was merely a threat of force, not nuclear attack of any kind. More precisely, it was a warning that we would force our way, with conventionally armed patrols, past an attempt by East Germans to block our access to West Berlin. As I thought of it, then and later, we were simply puncturing their baseless claims of nuclear superiority and their threats of cutting us off from Berlin on the basis of their genuine conventional superiority in the area. Still, as I’ve just now acknowledged, there was more to what he was claiming about what I’d written than I admitted to myself.

  What of the relation of President Kennedy himself to these threats? Most accounts of the origins of the speech attribute the substance of the passages I wrote to presidential initiative. As the historian Michael Beschloss puts it:

  The President, Bundy, Rusk, and McNamara collaborated with Gilpatric91 on a text for his Business Council speech.… Drafting of the speech was assigned to Daniel Ellsberg.

  Not so. All those officials undoubtedly signed off on the finished speech before it was delivered, and some or all may have earlier encouraged a strong statement about our military buildup and relative superiority, which Tim Stanley was drafting when I approached him. Probably none of them, starting with Gilpatric, knew of my role at all; none communicated with me before or after; I had no assignment. I’ve described above the actual sequence of events. Beschloss goes on to correctly recount my earlier exchange with Kaysen, including my suggestion that we give Khrushchev the exact coordinates of his four ICBMs. But he puts it incorrectly after I had supposedly been given this assignment to draft the speech, something that never happened. As in the New York Times account, every one of the five passages from the speech that Beschloss quotes on the balance were by me, taken from my handwritten notes originally drafted for JFK and given to Stanley at my initiative.

  I note this not from pride of authorship. As I’ve mentioned above, it makes me uncomfortable to realize that I’ve misinterpreted for more than half a century what I was really promoting in this instance. I’ve long thought, as I did at the time, that I was simply warning that we would act confidently with conventional force to assert our “rights and obligations” with respect to access to Berlin. We would not be deterred from that, I was saying, by the Soviet nuclear bluffs I was impl
icitly exposing. But they had never been threatening nuclear first use, over Berlin or anywhere else. We were. It didn’t stand out in my mind that Khrushchev’s bluffs had been precisely to counter our nuclear first-use threats, on which we were relying in the face of their real conventional superiority in Germany. I was participating in those nuclear threats without acknowledging it to myself.

  All right, I’d joined the crowd. But in retrospect, it was much worse than that. From the perspective of one year later, my initiative and my provocative words were near disastrous. That wasn’t clear right away—quite the contrary. The threats appeared at the time to have worked with spectacular speed. When I first learned soon after that Khrushchev’s ultimatum on signing a peace treaty with East Germany, giving control of access to the East Germans, had been withdrawn during the Party Conference, I and some others in the Pentagon assumed that the Gilpatric speech had led to this decision. That was very gratifying to me to think. I felt for a long time afterward that I had contributed to ending the Berlin crisis in 1961.

  I was disconcerted almost forty years later when an account written by my friend Seymour Hersh brought to my attention that “Khrushchev had publicly withdrawn his ultimatum that America negotiate a postwar peace treaty with Germany by the end of 1961” four days earlier than the Gilpatric speech, in his opening speech to the Party Congress. Thus, “the Gilpatric speech seemed to be Kennedy’s response92 to the Soviet retreat.”

  Michael Beschloss had pointed out even earlier:

  By asking Gilpatric to make this speech93 [sic], Kennedy may have strengthened his own domestic political standing and reassured American allies, but he also provocatively undermined Khrushchev’s position in the Kremlin and in the world.

  The Chairman’s entire domestic and foreign strategy was based on creating the illusion of Soviet nuclear might. Now, as the world learned that the emperor had no clothes, Khrushchev must have imagined that the Third World and perhaps even Soviet allies, previously mesmerized by Soviet power, might begin turning away from Moscow.… Khrushchev had fashioned an illusion of Soviet strength most of all so that the United States would treat his country as an equal. Now Kennedy seemed to have deliberately chosen to humiliate him.

  Khrushchev’s first reaction was to go ahead with a thirty-megaton nuclear test explosion two days after the speech, soon followed by a fifty-eight-megaton explosion, the largest ever.

  The thirty-megaton blast and Malinovsky’s tough language94 may have temporarily consoled the Party Congress delegates, but the deeply serious problems created for Khrushchev by Gilpatric’s speech remained. It pressured him to do something spectacular to change the world’s perception of the nuclear balance between the Soviet Union and the United States.

  Beschloss concludes:

  The speech violated the President’s own rule against backing an enemy into a dangerous corner. Kennedy never gave sufficient thought to how Khrushchev might receive the speech.

  Whether or not that was true for all the high officials who approved the address, I have to say that shoe fits me.

  Khrushchev almost certainly wondered95 why the President had decided to publicly humiliate him by rubbing his nose in the fact of Soviet inferiority, and amid a crucial Party Congress. Did the address foreshadow an American first strike against the Soviet Union?

  Khrushchev knew that his Kremlin and military critics would now demand that he relax his opposition to a huge Soviet military buildup. The forces set in motion by the Gilpatric speech and Kennedy’s other efforts to demonstrate superiority compelled Khrushchev to look for a quick, cheap way to remake the nuclear balance of power.… As Khrushchev might have put it, by authorizing the Gilpatric speech, the President of the United States was playing with fire.

  Within a few months Khrushchev had thought of a cheap, quick way to repay the humiliation and restore the balance. That wasn’t the only or even the main aim or the triggering cause of his deployment of nuclear weapons to Cuba in 1962 (see here).

  Nevertheless, in October 1961 I had done my part in greasing the skids toward the Cuban missile crisis.

  * * *

  When I visited Adam Yarmolinsky in his office next to McNamara’s in early June 1962—after spending the first half of the year writing my Ph.D. thesis at RAND—he mentioned that he had been assigned to draft a commencement speech for McNamara to deliver in July at the University of Michigan, in Ann Arbor, his alma mater. McNamara had decided for this occasion96 that he wanted to give an unclassified version of a speech he had presented at the NATO conference in Athens on May 5. That had been drafted, Adam told me, by Bill Kaufmann.

  Adam had rewritten Bill’s speech considerably. He handed me his draft and asked me to read it for comments. I asked for and got the source document as well, the Athens speech, which was classified Cosmic Top Secret (NATO’s Top Secret). McNamara had laid out, for the first time for the ears of our NATO allies, the “no cities,” counterforce, coercive strategy first promoted at RAND and by Kaufmann himself, underlying my draft guidance that McNamara adopted the year before. The United States had concluded, he said, that in a nuclear war stemming from a major attack on the alliance, “our principal military objectives should be the destruction of the enemy’s military forces”—not its civilian population—while retaining reserve forces capable of threatening its urban-industrial targets. This would give the Soviets “very strong incentives … to adopt similar strategies” to avoid urban targets of the alliance, providing the best hope of preserving the fabric of societies in the course of the war.

  When I compared Yarmolinsky’s draft for Ann Arbor with the Athens speech, I had a number of negative reactions. The least of these was that I thought the logic of the approach was spelled out more clearly in Kaufmann’s original speech than in the new version. Adam’s draft not only left out, necessarily, the classified figures for NATO and Soviet military forces and capabilities but, in an effort to reach a lay audience, it seemed to me to blur the rationale for the new approach, which was, after all, a striking departure from U.S. strategic planning heretofore.

  Second, I questioned the diplomacy of the speech in alliance terms. That applied to the Athens version as well. Kaufmann, with little guidance given to him initially for a classified speech to NATO on the new strategy, had chosen to present it implicitly as an attack on the French independent nuclear force, which Charles de Gaulle was in the process of constructing. The speech emphasized the importance of centralized control to a strategy aimed at avoiding destruction of civilians in urban areas on both sides, leaving them unhit in initial attacks but threatened by U.S. forces held in reserve. Without naming the French force explicitly—which de Gaulle had no intention of coordinating with U.S. forces and which was known to be aimed solely at a few Soviet cities, principally Moscow—Kaufmann drew attention to the contradiction that such a force posed for the very possibility of the U.S. strategy, which purported to have the best (or only) chance of “preserving the fabric” of Allied societies in nuclear war. For an uncoordinated French attack to destroy Moscow and some other cities at the outset of the war would mean “the destruction of our hostages—the Soviet cities,” assuring catastrophic Soviet attacks in kind on cities of the alliance.

  Kaufmann had added to this point a characterization of the projected Allied forces that seemed almost meant to offend the French, if not the British as well: “In short, then, limited nuclear capabilities, operating independently, are dangerous, expensive, prone to obsolescence, and lacking in credibility as a deterrent.”

  I didn’t see the point of gratuitously insulting the two allies even in a classified speech among their colleagues, let alone in a public address. But both the critique of the role of the independent forces and the aggressive language persisted in Yarmolinsky’s version.

  When I questioned this, Adam told me that this followed McNamara’s explicit guidance for his draft for Ann Arbor. McNamara had liked both Kaufmann’s general frame and his specific language. He had delighted in sticking
it to the French in Athens, and though he knew the French were unhappy about this, he wanted to do it again at Ann Arbor. (I never learned his motives for this. Like Kaufmann, he must have felt angry at de Gaulle.)

  I felt strongly that this subject was totally unsuitable for public discussion. Even our Allied military commanders had never heard in any detail whatsoever how the United States actually proposed to fight a nuclear war. After all, General Curtis LeMay had spent more than a decade keeping all such matters as secret as he could even from the Joint Chiefs of Staff, as well as from all civilian officials.

  I told Adam that I expected the American public, hearing of this for literally the first time, would be appalled.

  My reaction may need a little explaining. Yes, I had been proud of the effort I had made the previous year in helping shape this very strategy that was being expounded. (Earlier this same week I had reviewed for Gilpatric the new JSCP-63—submitted for his approval for the first time by the JCS—and found that it had incorporated, in language at least, virtually all the changes I had proposed, over his signature, in 1961.) But that had been because I was rejecting and replacing an Eisenhower-era plan that seemed to me to be unequivocally much worse.

  Moreover, in the spring of 1961 I had been working on guidance for what the JSCP explicitly said was for use in the event of a Soviet non-nuclear assault in Europe. I assumed, however, that it was actually for use as a retaliatory, second-strike plan, since I still believed that the Soviets had either a predominant strategic missile force or a parity in second-strike capability, either of which practically ruled out U.S. escalation to a first strike, regardless of our alliance commitment. So, for a desperate situation, “my” plan was, in my own understanding, the least awful way to respond to a Soviet surprise attack.