Annihilation from Within

written four months after Hiroshima. At that early date, Jacob Viner explained essentially all the benefits and problems of deterrence, and anticipated with astonishing foresight and beautiful clarity the transformation of international affairs that nuclear proliferation would cause. Since then, nuclear strategists became more and more fixated on “deterrence” as if it were a concrete, empirically observable entity. The American philosopher Alfred North Whitehead called such an excessive reliance on an abstraction “the fallacy of misplaced concreteness.”5
“Deterrence” theorizing postulates how a “potential” aggressor calculates whether to attack—a hypothesis hard to test empirically. By contrast, the nonuse of nuclear weapons is not a hypothesis; it has been evident since 1945. Indeed, it has prevailed even in wars that nuclear-armed powers have lost against non-nuclear nations. By the end of the Cold War, many strategists came to see nonuse and deterrence as two sides of the same coin: nonuse became the proof of successful “deterrence” and “deterrence” the strategy guaranteeing nonuse. Alas, 9/11 casts anguishing doubts on this proposition.
Since war never broke out in Central Europe during the forty years when the largest military confrontation in history divided the continent, it is tempting to assume nuclear deterrence preserved the peace. This view remains unproven; Soviet archival documents released so far do not provide an answer. It is, however, unquestionable that America’s nuclear superiority did not keep the peace on the Korean peninsula. Released Soviet documents confirm that Stalin authorized Kim Il Sung (the North Korean dictator) to attack South Korea in June 1950, although the United States had some 350 atomic bombs at that time and the Soviet Union only about five. Stalin even remained undeterred after the United States had come to the defense of South Korea: he supported North Korea with Soviet fighter pilots. Had nuclear superiority really been effective in deterring aggression, the United States would not have suffered 33,000 fatalities and 92,000 wounded in defending South Korea against North Korea and China, neither of which had nuclear weapons at the time. Nor was this the only occasion on which nuclear-armed nations have accepted stalemate or defeat in wars with enemies that did not have a single nuclear bomb. The United States did so in Vietnam, the Soviet Union in Afghanistan, and in 1979 a nuclear-armed China withdrew from its cross-border aggression into Vietnam. Then there is the case of the 1962 Cuban missile crisis. It would be a fallacy of misplaced concreteness to believe that nuclear deterrence safely prevented war on that occasion. That crisis was caused by the Soviet Union’s provocative deployment of nuclear missiles in Cuba and nearly triggered a full-scale, yet unintended nuclear war.6 Let us note, Cuba 1962 was not a crisis that could have been prevented by the “deterrent” effect of nuclear weapons—it was caused by nuclear weapons.
Lesson Three: We Were Lucky—So Far
Deterrence can ward off only deliberate attacks, it cannot dissuade an accident from happening or a madman from detonating a nuclear bomb. The world seems to have been close to an accidental nuclear war during the Cuban missile crisis, as well as on several less visible occasions. What averted the accidental Armageddon? Was it the prudence of all those managing the nuclear arsenals, was it the intervention of extraordinary good luck, or—seen in a transcendental way—the intervention of Providence?
I shall open the problem of accidental nuclear war by describing my own professional encounters with it. During the 1950s, I worked at the RAND Corporation, a think tank in Santa Monica, California. That decade was RAND’s most creative period—a time when its scientists, sustained by generous contracts with the Air Force, could pursue a range of innovative ideas. Many of us at RAND worked on nuclear strategy, elucidating the need for survivable retaliatory forces, credible response options, and other critical issues. Much of our thinking was recorded only in top-secret documents. Then, one morning in March 1955, the newspapers brought us Winston Churchill’s grand speech on deterrence (his last major speech in the House of Commons). I recall vividly our astonishment at RAND that day. Here the British Prime Minister had expressed all the key ideas of our secret work—and with such eloquence! Who had authorized him to go public?
In any event, I felt the workings of deterrence were becoming a rather crowded field of inquiry, and that it might be more useful to explore the possibility of a nuclear catastrophe that deterrence could not prevent. Might a nuclear weapon be detonated accidentally, or a missile launched because of an unauthorized act? With the help of RAND’s weapons experts, I was able to demonstrate that in many situations just one person (with authorized access) could trigger an unauthorized detonation, either “by accident” or “by design”—however much the design was conceived in madness. Together with Gerald Aronson, a brilliant psychiatrist who consulted for RAND, I reviewed then-current procedures for selecting personnel with access to nuclear weapons, as well as medical statistics on the occurrence and types of dangerous mental disorders among active military personnel. Aronson and I concluded that existing personnel procedures could not prevent mentally unstable individuals from gaining access to nuclear controls. Special personnel screening would help, we suggested, but it could not guarantee complete protection. Hence our recommendation: that two people always be in charge of the truly critical controls, and that coded safety locks be placed on all weapons and missiles exposed to the risk of unauthorized launch.
RAND endorsed these findings, and I was sent to Washington to brief a sizable audience of Air Force generals. As a young researcher at my first high-level Pentagon briefing, I delivered my rather simple message with trembling knees. But the next day, General Curtis LeMay, then Vice Chief of the Air Force, had heard about my study and asked for it. General LeMay is justly famous for having built the U.S. Strategic Air Command into the world’s most formidable nuclear deterrent; yet he also became known for his injudicious pugnacity regarding the use of nuclear weapons. We at RAND expected him