The SAC admitted that for ‘four minutes’ on last November 24th, during the height of the Berlin crisis, it seemed as if the United States might be under a nuclear attack. The SAC issued a global alert to all its stations when a key electronic-system failed in the nerve-centre of the North American Defence command, and planes loaded with H-bombs were rushed on to the runways all over the world. General Powers, the SAC commander, said that only the President could initiate a missile count-down, and that he had not alerted the President as he suspected that there had been a signal failure in the attack-warning system.

(From the New York Times, April 20th, 1962)

There are no experts on Accidental War. There are only students who read the newspapers and the Rand reports with more or less vivid imaginations. One school of readers is satisfied, almost to the point of complacency, that the military services have developed ‘fail safe’ techniques that can prevent the unintended discharge of nuclear weapons. The other points to the 50 weapons failures that are known to have occurred so far, though without a nuclear detonation resulting. This school questions whether an enormous arsenal of diverse and sophisticated weapons can be maintained ‘intact’ for 20 years or more in a world that is daily—if not hourly—beset by international tensions.

There are two fundamental issues upon which all opinions concur. First, that the formidable thermonuclear striking-power now available to each ‘camp’ could deliver such terrifying punishment upon the other, if the opponent should have decided to strike first, that neither ‘camp’ will be easily tempted to begin an intentional war. Second, all opinions agree that the potential dangers of an unintended catastrophe are astronomical. Each nuclear Power will soon possess 40,000 missiles, satellites, submarines or bomb-carrying planes; the size of the global nuclear stockpile already stands at 70,000 megatons and is almost doubling annually; the number of Powers equipped with nuclear weapons will grow steadily during the 1960’s and will thus intensify the perils of international conflict. (Expert opinion has been quoted that China will probably test its first atomic weapons between late 1962 and early 1964. C. P. Snow has predicted that 12 countries will be able to manufacture fissionable material by 1965). Worst of all, as the destructive power of the opening salvo, or First Strike, grows apace with the development of 100-or 500-megaton bombs, the increasing speed of the missile-carriers will sharply reduce the warning-period that can elapse between the despatch and impact of a nuclear salvo—whether it was accidentally fired or not.

When von Clausewitz issued his famous dictum, that war could resolve international conflicts through a competitive resort to violence, there were two premises which he kept in mind. First, that wars would be preceded by several weeks of diplomatic manoeuvring and negotiating, during which the protagonist would be able to experiment with bluffing tactics and withdraw its provocative gestures if it judged this to be expedient. Second, that it would take so long to mobilise for an effective campaign that no aggressor could hope to overwhelm or totally destroy its adversary in the opening gambits of the war. These two premises were totally demolished when the manoeuvring period was reduced to a period of less than 15 minutes, and when the concept of ‘overkill’ was embodied in the ruthless calculation of a First Strike. The demolition implied, however, that war was no longer to be viewed as a rational pattern of behaviour, since any one accident of the moment might not only be irrevocable, but also infinitely catastrophic.

The logical deduction from this impossible situation was Deterrence theory, or the projection of a mutual intimidation into a Balance of Terror. This theory of daily catastrophe accepted an arithmetic possibility that Accidental War might break out since it accepted the fact that it was beyond the bounds of human control to devise a universal form of ‘fail safe’ precautions. If it were necessary to multiply, scatter, and ‘harden’ the number of weapons that were trained upon the adversary at each hour of the day, in a state of constant trigger-readiness, then it was obviously impossible to fit an interlocking safety mechanism upon these weapons that could immobilise each one of them until the very minute that it was needed. Realistically, it should be granted that there will always be a constant variable of imperfect intelligence reporting, of strategic miscalculation, and of political misunderstanding. Hence, it is reasonable to anticipate that the probability of an Accidental War will multiply at the same terrifying, geometric progression as the numbers and destructiveness of weapons that each ‘camp’ keeps adding to its thermonuclear arsenals. The anticipation should be based upon the likelihood not that defective weapons, themselves, can start wars, but upon the probable fallibility of human judgment in handling such massive powers of destruction.

It is impossible to quantify the probabilities of human error, mechanical failure or political misunderstanding that might unleash an Accidental War. It might be useful, though, to identify the types of categories of error that might occur, if only to distinguish between the most and the least accident-prone activities of the contemporary arms race. The ‘grand design’ of the Balance of Terror necessarily assumes that the status quo can be perpetuated for years to come if nuclear weapons can be sufficiently improved, dispersed, ‘hardened’, or concealed. Underlying this theory of mutual Deterrence is the expectation that the weapons can be kept under a permanent and almost foolproof control, even while they are multiplying in number and in sophistication. If it can be shown that this expectation is uncertain, and that provocative accidents are likely to increase, the theory of Deterrence will be deprived of one of its central supports. The theory, after all, is obliged to guarantee the international stability of a mutual-annihilation system if it is to remain as the basic mode of the Cold War. The military services have openly asserted that precautionary measures will invariably keep pace with innovations of technology, strategy and deployment, and that the lay citizen need not fear for a sudden unleashing of the powers of mass destruction. This assurance must be severely questioned. No control system can anticipate all the forms of aberrant behaviour that are likely to occur; nor can the varied possibilities and the increasing probabilities of Accidental War always be subjugated to the ingenuity of human control. It is the purpose of this paper, therefore, to investigate the assurance of the Deterrence strategists that the probability of an Accidental War is properly accounted for in their plans for a lasting Balance of Terror.