Book Review Essay

Bundy examines the few instances in which more-or-less explicit nuclear threats were made, and also several of the more numerous occasions that involved vaguer nuclear "threads"-when the possibility of escalation was not stated but was nevertheless implicitly woven into the fabric of the conflict. His primary purpose is to demonstrate how distant these instances remained from nuclear confrontation, how much "more prudence than menace" was present even then. That case he makes compellingly.

Yet, Bundy points out, in several early instances nuclear "threads" made for success, or at least history cannot be said to demonstrate that they failed. At Geneva in 1954 the Western powers won more than they expected in negotiations over Southeast Asia, and Bundy, agreeing with Anthony Eden, concludes that "the existence of thermonuclear weapons in the hands of the Americans increased the appeal of a peaceful result for the Russians and perhaps also the Chinese"-even if no threat was made. So, too, while Eisenhower's more explicit nuclear threats over Quemoy and Matsu look, especially now, like strategic risks out of all proportion to the prize, "within their limited terms they cannot be called unsuccessful."

During the 1973 Middle East war, the American alert contained a nuclear thread that was less a threat than a signal of resolve. It might have been redundant, but it, too, cannot be called a failure, Bundy concludes. Other thoughtful studies reach a similar conclusion: when nuclear weapons have been used to convey political signals, in the explicit hope that no fighting of any sort would occur, the signals often have succeeded in their specific purpose.1 Nor has "success" been markedly less likely for the United States since the Soviet Union came to be its nuclear equal.

Bundy's "superficially paradoxical" conclusion about Geneva in 1954 fits this broader set of cases as well:

The very weapon whose real availability seems quite uncertain, given the hesitancy and secretiveness surrounding suggestions for its use, was still a force in the process by which a temporary settlement is reached.

Indeed it is hard to escape the conclusion that these nuclear threads seemed to work precisely because the superpowers were so aware both of the nuclear danger and of its distance in a particular case. Because the danger was so remote there was space for a safe resolution. Because the danger was so terrible, the signal was understood: "make no mistake, I care very much about this issue."

The most elusive of these slippery nuclear threads is whether the state of the nuclear balance itself has an effect on national behavior independent of particular threats or signals. Henry Kissinger's changes of mind illustrate just how elusive this question can be. As a student of nuclear matters in the 1950s he fretted about Soviet "blackmail" deriving from nuclear superiority. As a defender of SALT I in the 1970s, however, he mused about what possible meaning "superiority" could have when both sides had thousands of nuclear weapons. And still later in the same decade he had turned around again, fearing, as did the Committee on the Present Danger, that Soviet superiority "must exponentially increase" Moscow's willingness to take risks.

Happily, Bundy notes, that last fear has turned out to be unfounded: Soviet risk-taking has not increased in the last decade, during which the nuclear "window of vulnerability," if it existed a decade ago, remained as wide open as ever. More likely, the window was never real: the Reagan Administration wisely closed it in 1984, at least in the language of the Scowcroft panel. The rhetorical closing can be seen as an implicit acknowledgement that the window never existed.

The nuclear stalemate is robust; a thousand weapons, more or less, do not upset it. And neither do they produce usable political advantage. Still, as Bundy observes, it does a good argument no good to push it too far. If he and his colleagues during the Cuban missile crisis were not as comforted by American nuclear superiority as outside commentaries would have it, especially those made through the rosy glasses of hindsight, neither would those men have wanted to trade nuclear places with the Soviet Union. Geneva or Quemoy and Matsu may be evidence of the political effect of dramatic inferiority, something the United States has never suffered. But marginal differences do not matter.

It is always easy and almost always unfair for reviewers to ask authors for something more. Still, I wish Bundy had turned his conclusion into an exploration of the relation between the nuclear decisions that are his subject and the attitudes of the citizens in whose name they were taken. His own calm understanding seems very far from the tenor of our recent debate. So, too, while I hold no brief for my fellow strategists as a profession-there have been many, in uniform or out, promoting half-cocked schemes-the best of them have seen the nature of nuclear reality clearly from the beginning of the nuclear age.

Thus, the first American nuclear strategist, Bernard Brodie, writing in 1945, captured that reality in a paragraph that has not been improved upon since:

It seems hardly likely, at least as among great powers at some distance from each other, that an attack can be so completely a surprise and so overwhelming as to obviate the opponent's striking back with atomic bombs on a large scale. For this reason, the atomic bomb may prove in the net a powerful inhibition to aggression. It would make little difference if one power had more bombs and were better prepared to resist them than the opponent.2