Underneath our continuing debates on nuclear danger we begin to detect the possibility that we may be able to reach a new level of common understanding of what nuclear deterrence is, what it is and is not good for, and how it relates to morals, to politics, and to the prospects for peace. It is too soon, and probably too much, to try to set forth any “general theory” of deterrence, but it does seem possible to clarify some elements of its nature. Such an undertaking may have additional value in a season in which, as Emma Rothschild has powerfully demonstrated, the Pentagon under Caspar Weinberger is doing its best to give deterrence a bad name.1 Such consideration of Mr. Weinberger’s misconceptions is one powerful catalyst of new understanding, and another, itself refined by extended debate, is the pastoral letter of the National Conference of Catholic Bishops, overwhelmingly approved in Chicago on May 3. Still a third is the report of the Scowcroft Commission, in which good and bad arguments are mixed together with remarkable inattention to their intrinsic incompatibility.

The pastoral letter fully deserves the wide audience it seeks. It is a thoughtful and comprehensive effort to bring religious and moral principles to bear on nuclear weapons. In reaching their conclusions, the bishops have been unafraid to criticize existing official positions; in turn they have been treated first with suspicion and then with wary assertions of sympathy by the Reagan administration. The pervading concern of their letter is with the “unique challenge” presented by nuclear warfare to “the classical Christian position” on war and peace. The power of their response to this challenge makes the report a landmark in the changing pattern of American concern with nuclear danger, and an excellent starting point for a look at what can now be said about deterrence.

I begin from these unpleasant truths: nuclear danger exists; there is no escape from the requirement to coexist with it; and this requirement in turn implies some notion of deterrence. We do not know how to transcend the danger. This is not a welcome thought, since a quite modest awareness of the nature of thermonuclear weapons is sufficient to persuade most of us that if we could truly “ban the bomb” we should. But neither political nor technological reality now allows it.

The political reality is that the nations have never come near to any agreement that would reliably eliminate nuclear weapons. This melancholy history does not itself determine future possibilities; it may show only that we have not tried hard enough, and certainly there have been many moments at which a better effort might have been made. Yet a larger lesson of the record is that in the weapon-building decisions of nations, fear of the bomb itself has always been less powerful than fear of an adversary’s bomb. The weapons each government has sought have been those it found necessary in the light of what others had done or might do. Not one of the nuclear-weapons states has ever been ready to surrender that judgment to any other government or to any international authority. Flora Lewis was right to remind us recently of the comment of Arkady Sobolev, the senior Russian official at the United Nations when the Baruch Plan was put forward, that “the Soviet Union was not seeking equality, but, rather, freedom to pursue its own policies in complete freedom and without any interference or control from the outside.”2

This situation is not likely to change soon. It reflects the continuing and deeply rooted power of the idea of national sovereignty. The reality that the nations of the nuclear world are now inescapably interdependent—dependent on each other’s behavior for their very survival—has not changed the deeply entrenched determination of nearly every nation, and in particular every nuclear-weapons state, to make its choices on such matters through the decision-making process of a sovereign government in which authority, interest, hope, and fear are all defined in primarily national terms.

If anything could have led people to abandon their nation-centered ways, it should have been the bomb. The logic of this proposition has been compelling from Hiroshima onward. From Norman Cousins to Jonathan Schell, writers preoccupied by nuclear danger have fixed on national sovereignty as the basic obstacle to the exorcism of the specter. They are right, but the obstacle persists, and while there may be a special intensity to the state-centered dedication of the Soviet leadership, the beam is in other eyes too, including our own.

A second possible means of transcending the threat is technological. Hope on this front was rekindled, at least for some, by Mr. Reagan’s remarks in March offering “a vision of the future” in which there would be a reliable defense against strategic missiles. But is such a hope well founded? I think not. The difficulty here is not simply in the extraordinary cost and complexity of the new systems we do not yet know how to build or in the fact that there can be countermeasures to these systems. There is also the more basic problem that thermonuclear weapons impose a radically new calculus of advantage on anyone seeking to neutralize them: they make it necessary to achieve a kill rate very near 100 percent. Anything less is not good enough for safety—only good enough, at best, for deterrence.


In that sense the hope held out in the president’s address is deeply misleading. Even if all his assumptions about technical promise were granted, it would be impossible to believe we shall ever have a perfect defense against ballistic missiles, including those at sea, or to suppose that we could have a defense equally perfect against aircraft and cruise missiles too.

We have two bedrock realities: the units of political account are nation-states, and thermonuclear weapons are drastically different from any others. These two realities, mixed together in different ways, with different additional ingredients, underlie all assessments of nuclear deterrence. Even taken alone they require some notion of deterrence, of the way in which nation-states are to coexist, with nuclear weapons and without nuclear war.

On these two realities the United States and the Soviet Union, over the last thirty years, have built a third: two extraordinarily large, varied, and survivable sets of thermonuclear systems which are in themselves a most formidable reality. Even if the two governments were to reach agreement tomorrow on large-scale reductions—whether those of Ronald Reagan or those of George Kennan—they would still face each other with thousands of warheads so deployed that neither side could hope to make a large-scale nuclear attack on the other without a wholly unacceptable risk of receiving a catastrophic reply. Moreover, the path from any lower level of weaponry to a world truly free of nuclear weapons is still blocked by this enormous unanswered question: How will the two nations ever be able to trust each other not to hide a few—or a few hundred?

Nuclear danger and nuclear deterrence plainly exist between the superpowers and will not disappear soon. As the bishops put it, “Deterrence is at the heart of the US-Soviet relationship, currently the most dangerous dimension of the nuclear arms race.”

To this present reality the bishops give a highly qualified but clear-cut response. Balancing nuclear danger against the need to protect “the independence and freedom of nations and entire peoples,” they reach “a strictly conditioned moral acceptance of deterrence.” In our current debates their strict conditions may be more significant than their approval. The bishops have studied the question closely, and they know what they are against: nuclear war. They are not impressed by war-fighting theories of any sort. They have no difficulty explaining their firm opposition both to strategies that would entail deliberate attack on large populations and to strategies that would merely produce catastrophic loss of human life as an “unintended” consequence of megatonnage aimed at military targets. They know and report the medical assessment of what nuclear war would be like, and how it would obscenely multiply the “butchery of untold magnitude” that Pope Paul VI saw in Nagasaki and Hiroshima. They know the extraordinary uncertainty attending any attempt to “limit” any nuclear exchange and the incalculable risk involved in any first use of nuclear weapons. They explicitly withhold their approval from all of these forms of war-fighting. These and other conclusions and recommendations all follow from the bishops’ rejection of nuclear war, and we must return to them. Yet what must be noted first is that with all their conditions and caveats, the bishops accept deterrence itself.

The deterrence so conditionally approved is a deterrence essentially independent of any particular theory of war-fighting, a deterrence inherent in the deployments already made and sure to exist for many years to come even under the most optimistic assumptions. In the absence of very large changes in their survivability, the existing systems on both sides are now so powerful and varied that no political leader can have or hope to have any clear idea of what would in fact happen “if deterrence failed”—that is, if nuclear war began.

This difficulty is not escaped by any theory, because no theory can predict with any confidence the behavior of any government, friend of foe, in such a situation. Most “scenarios” for nuclear warfare between the Soviet Union and the United States reflect nothing more than the state of mind of their authors. Estimates of the ease or difficulty of limiting such a conflict, for example, are just that—estimates. Moreover, they are estimates of interacting behavior under conditions of unprecedented stress and danger, possibly in the midst of already appalling destruction. No one can have any certainty that credible communication would be possible between the adversaries even hours after such a conflict began. Such communication requires leadership that is mutually recognized, effective, and continuous, on both sides. Who can promise that? Yet without such communication who knows how to stop the horror?


Similar inescapable uncertainties apply to every other element of the situation. Which weapons would function with what efficiency? What intentions would be read into what actions? Can we tell how Soviet decision makers would regard the launching of a missile that could be aimed at either missile fields or nearby Moscow? How would an adversary interpret a temporary absence of response, when it could reflect almost anything: careful restraint, a plan for later coordinated retaliation, or the absence of any governing mind at all? Would the impulse to stop the slaughter be stronger than the impulse to kill the killers, and would it be the same or opposite on the two sides? Who can tell?

Because of all these uncertainties, serious leaders have known for a generation that there is no pat answer to the terrible question of what to do if deterrence fails. It is natural but unprofitable for advocates of one or another approach to deterrence to accuse each other of having profoundly immoral intentions in this event—plans for “prolonged nuclear war” or for “mutual assured destruction.” The bishops are quite right to reject both kinds of nuclear war, and their own concentration on deterrence should remind us that in reality our debates about hypothetical nuclear exchanges have seldom reflected settled convictions about what one would actually do in the dread event of nuclear war. Instead they reflect judgments on the kind of military capacities thought necessary for the self-confident deterrence of the other side.

The bishops do not address the question of what to do if deterrence fails, but it is at least a partial defense of their omission that a bad answer to this question is worse than none at all. My own belief is that if the test ever comes, most of our military scenarios and planning directives will prove irrelevant, and that the attention of both sides must be driven toward the literally vital need for ending the nuclear battle if possible, not winning it. But that is another subject.

The terrible and unavoidable uncertainties in any recourse to nuclear war create what could be called “existential” deterrence, where the function of the adjective is to distinguish this phenomenon from anything based on strategic theories or declared policies or even international commitments. As long as each side has large numbers of thermonuclear weapons that could be used against the opponent, even after the strongest possible preemptive attack, existential deterrence is strong and it rests on uncertainty about what could happen.

Existential deterrence has been strong in every major crisis between the super-powers since “massive retaliation” became possible for both of them in the 1950s. As everyone closely involved recalls, such deterrence was particularly powerful during the Cuban missile crisis. In those tense days neither government showed any desire to expose itself to the nuclear uncertainties I have noted; neither enjoyed its inescapable dependence on the sanity, and even the restraint, of the other. For the Americans it was not their nuclear “superiority” that allowed the necessary firmness, but rather the fact that the Soviet provocation could be dealt with by usable conventional superiority in the area. The lesson of this classic case is that existential deterrence makes other kinds of strength determinant. Nothing in the succeeding twenty years has weakened the force of this rule for decision makers on both sides.

Now that both strategic arsenals are redundantly destructive and amply survivable, we can say with still more confidence that existential deterrence is strong, and that its strength is essentially independent of most changes in deployment. Because no one can predict how these arsenals might be used, because these uncertainties create an enormously powerful existential deterrent, and because this reality is essentially unaffected by any changes except those that might truly challenge the overall survivability of the forces on one side or the other, it makes no sense to base procurement decisions on refined calculations of the specific kinds of force that would be needed for a wide variety of limited nuclear responses. All such responses would be full of unpredictable elements; all would be open to catastrophic misinterpretation; none could be expected to lead to a good result. The decision to have options is sensible, if only because there is no need to lend even marginal credibility to the destructive and erroneous notion that our presidents in recent decades have had no choices between surrender and all-out retaliation. But at the current levels of survivable strength and accuracy, to maintain such options is relatively easy; to let the pursuit of special capacities to “kill hard targets” drive procurement is wrong.

In view of what both sides have to-day, only the reinforcement of survivability can justify the expense of any large new system, and survivability itself is coming to depend increasingly on some degree of arms control. Its recognition of these two realities is what persuaded the Scowcroft Commission to adopt its wise conclusion that we need to emphasize limits on warheads (rather than launchers), and that we need to move away from vulnerable multiplewarhead systems—systems of low survivability that can be destroyed by relatively small attacking forces. Unfortunately, its recommendation that the ten-warhead MX be placed in vulnerable Minutemen silos goes in exactly the opposite direction. So emplaced, MX would add nothing to the overall survivability of American forces and would only complicate the very transition away from such forces that the commission recommends. In the Scowcroft report the future belongs to sanity, but the present still belongs to the mistake of the Air Force, which designed the wrong missile and then fell in love with it. Yet I think that in the end the sensible recommendations of the commission will overtake its illogical endorsement of MX, just as I think the basic conclusions of the bishops will prove more durable than the debate over the degree of their enthusiasm for specific proposals like the freeze or (to take one I myself resist) the idea of a “peace academy.” Already influential members of Congress are asking for new arms-control proposals, and as the assurances promptly offered by the president are tested by his real behavior in the months ahead, I think it will become clear that he cannot in fact meet the real concerns of Congress and keep anything like 100 MX in his plans.

Mr. Reagan has told influential senators, for example, that he is sympathetic to their idea of a “build-down,” under which any new strategic warheads would be “paid for,” in new arms control agreements, by withdrawing other warheads, two for one. He might vary the ratio, the president says but the senators, concerned over survivability, would clearly expect that MX warheads be paid for at two to one or even more. But who will ever support the notion of deploying 1,000 MX warheads at the price of withdrawing 2,000 others, all of them more survivable and so more valuable for the deterrent? It is really a great pity that no one ever told Mr. Reagan that the right escape from the MX was to call it not “Peacekeeper,” but simply “that other fellow’s mistake.”

Existential deterrence has other useful qualities. It deters quite impersonally; no provocative threats are needed to support its power. It deters both sides at once, since the unpredictable risk of catastrophe is essentially symmetrical. It makes full and impartial use of one of the great realities of nuclear weapons: that they are far more terrifying to adversaries than they are comforting to their possessors. When we review the record, looking at the reasons for the decisions of major governments, we find that while motive is not always clear and usually somewhat mixed, the dominant theme, starting with Franklin Roosevelt in 1941, is fear—if we don’t get it first, he (Hitler) may. The theme has many variations. He’s got it, so we must—Stalin in 1945; they are doing it, so must we—Britain and France; they have done it, and we will—Mao; they will try, so we will succeed—Israel. Less fearful motives are also in play: we will win; we will catch up; we will be great; we will even, perhaps, be safe. But fear dominates.

In the age of overkill, it still does. We do not dwell on what we can do to them, but on what they can do to us. We talk about it more than they do, but it would be wrong to suppose that we care more. What country has a longer and deeper awareness of the need to preserve itself from external danger than Soviet Russia? But the point is not centrally one of national attitudes; it is rather a consequence of the realities of nuclear exchange. A nuclear exchange in which our side inflicts ten times the damage that it suffers—no matter what target system either side chooses—is still a catastrophe for our country. The horrors that our land and people suffer are not a whit reduced by the fact that something even worse has happened far away. Indeed if the war somehow stopped at that point, the overhang of our greater guilt would heavily shadow an already grim future with the threat of some later revenge. As long as each side retains survivable strength so that no leader can ever suppose that he could “disarm” his opponent completely, nuclear war remains an overwhelmingly unattractive proposition for both sides. This is the reality from which moralists and political analysts must draw the same great conclusion: no to nuclear war.

The deterrence the bishops conditionally approve is deterrence only of nuclear war. They do not believe in what is called “extended deterrence,” the threat to use these weapons to resist or deter conventional aggression. They think it will never be morally right to be the first to cross the nuclear threshold, thus joining those of us who have argued, as a matter of political prudence, that it is time to move away from reliance on that suicidal threat. The bishops accept the logical possibility that such a change of policy, especially in NATO, might require new conventional forces, but they naturally prefer to emphasize the need for conventional-arms limitation and also for the reduction of tensions. It is interesting, if not surprising, that their German and French colleagues do not share their views on this point, but it takes an unnecessary and even unbecoming lack of American self-confidence to suppose that this particular difference makes the American bishops wrong.

Yet there can be no doubt that in denying the morality of any first use of these weapons the bishops run against deeply settled attitudes, not so much in the general public here or in Western Europe as among so-called elites. Accustomed for thirty-five years to think of the “atomic shield” as our ultimate defense against large-scale Soviet aggression, our governments, and above all our defense establishments, find it hard to re-examine their assumptions. There is movement even in these circles; it is now a conventional ritual to assert one’s concern for attending to the conventional balance, and there is increasing evidence that this particular task is well within our means.3 Yet people always fear the unknown, and not to rely on the nuclear threat is now the unknown. After a year of exposure to friends on both sides of the Atlantic who exhibit strong withdrawal symptoms when invited to consider no-first-use, I have found myself drawn to the conclusion that we do not think very well about the problem of avoiding war when we think only about nuclear deterrence.

Consider the situation of states without nuclear weapons or nuclear umbrellas held over them by friends. Such states too must have some notion of deterrence, a view of what is necessary to deter the use or threatened use of nuclear weapons. In this sense both Sweden and Yugoslavia have views of deterrence. Each has the Soviet Union as a neighbor. Neither takes Soviet benevolence for granted, but neither has chosen to seek safety in nuclear deterrence. Instead each relies on a combination of substantial conventional strength, solid political self-confidence, and avoidance of any threat to Soviet vital interests. Each government is persuaded that the cost of any conventional or nuclear attack, in the judgment of any sane Soviet leader, will heavily outweigh any possible advantage. In these two cases deterrence has worked just as well as deterrence in NATO.

To some degree, of course, deterrence in Sweden and Yugoslavia is related to deterrence elsewhere. While Swedish and Yugoslavian leaders do not frequently speak out about the problems they might face if they were alone in the world with the Soviet Union, one doubts that the question is absent from their thoughts. But what is most interesting about the way this kind of deterrence works is its relevance to other states that are not neutrals. All the forces that protect Sweden and Yugoslavia can also help to protect others, even allies of the United States. Consider the case of the Federal Republic of Germany. The Federal Republic already has in the field what is man for man the best army in Europe. It has troops of other major nations at its side. Winning a conventional war against those forces in Germany would not be simple today, and it can be made much harder; but what is worst of all, a Soviet victory in West Germany would not be the end of the war. As with Sweden or Yugoslavia, but with much greater speed and certainty, it would be only a terrible overture to a continuing confrontation with the unconquered nations, even if it did not lead to the abyss of nuclear war—as it always might, whatever governments may have hoped or even pledged beforehand.

What keeps the peace in Europe, then, is the whole range of consequences that aggression would entail, even if in some initial stage it might “succeed.” As long as the Americans cannot be defeated—as long as the deterrence of their survivable forces is strong—there is no attraction in the choice of war.

The members of the NATO alliance, like the neutrals, can properly rely as well on other kinds of deterrents. Their own political self-confidence is one. We too easily forget that the postwar failure of exactly this self-confidence in Western Europe was probably the primary motivation for both the economic effort of the Marshall Plan and the military guarantee of the Atlantic alliance. Once Europe was reassured, the Soviet government was by that very fact the more deterred. Michael Howard of Oxford correctly insists that reassurance and deterrence, properly understood, are mutually reinforcing.4 One powerful argument against complex and noisy plans for deliberate nuclear escalation is that they have a strongly negative effect on reassurance. Nor are Europeans reassured by American officials who talk as if the deployment of new American missiles in Europe were somehow a test of European loyalty to NATO. American strength unaccompanied by American moderation does not make for reassurance.

In similar fashion the refusal to pose any aggressive threat also contributes to deterrence in that it does not create any heightened incentive for Soviet leaders to accept the risks of war. In this sense the repeatedly demonstrated absence of any Western will or capacity for forceful liberation of the Soviet satellite empire has become a form of reassurance to the leaders in the Kremlin—a further demonstration to them, on a different plane, that peace is better than war—and indeed less threatening to their empire. Hating as we do the enforced denial of freedom in the Eastern European nations, and concerned as we should be to see their freedom increased, we are nonetheless required to recognize that the option of liberation by external force is foreclosed.

These suggestions of other elements in what keeps the peace of Europe are couched in the language of politics, not morality, but they remain in harmony with the pastoral letter. Like the letter, they rest on a lack of faith in nuclear deterrence as a cure-all, and also on a view of both ourselves and the Soviet Union that is less absolute than the one to which cold-war rhetoric and its Reaganite revival would accustom us. The bishops are not naive about the Soviet government: “The fact of a Soviet threat…cannot be denied…. Americans need have no illusions about the Soviet system of repression”—or about the Soviet nuclear buildup. They specifically note with disapproval the largescale Soviet deployment of potentially destabilizing missiles with many warheads during the last decade (although they do not make the Scowcroft Commission’s mistake of supposing that the right answer is a similar system of our own). But the bishops perceive that the Soviet leaders can be deterred, and in more than one way—by negotiations as well as by defensive strength, by awareness of our common humanity as well as our uncommon differences, and above all by the recognition “that everyone will lose in a nuclear exchange.”

Returning to existential nuclear deterrence, we must ask ourselves how it looks to Soviet leaders; it is certainly wrong to assume that what deters us necessarily deters them. Yet those who are most gloomy about Soviet intentions generally concede that the Russians do not in fact want a nuclear war. They are surely right, and both the behavior and the language of political leaders in the Soviet Union confirm that conclusion. As I have argued in these pages recently,5 there is no doubt that Soviet military doctrine and Soviet procurement reflect a belief that the best way to deter is to have a usable capacity to fight, but their underlying purpose, like our own, is not to have the war. Communists and Catholics obviously have fundamental differences on the nature of man, but on the nature of nuclear danger serious Soviet leaders and experts have repeatedly shown an understanding not essentially different from that which moved the bishops. In this sense existential deterrence is strong in Moscow too.

From this assessment we can draw a further conclusion which the Russians themselves might not wish us to reach. Where existential deterrence is strong, where there is an informed fear of nuclear war on both sides, the threat of nuclear blackmail can be contained. Direct threats to use nuclear weapons have not been frequent in the nuclear age, and even indirect threats have become fewer as the realities of existential deterrence have been recognized more and more widely. I think this is not an accident. When sanity requires leaders to act carefully, and when even the most bitter adversaries perceive each other as sane, threats to do something self-destructive become unpersuasive; they simply lack credibility. The coercive value of nuclear threats has never been as high as various temporary believers in “atomic diplomacy” have temporarily believed. For use between possessors of huge survivable forces they are certain to be ineffective unless one side has a foolish loss of nerve. Even against non-nuclear states, the coercive power of nuclear strength is low.

This will not prevent the government in Moscow from occasional attempts to gain by such threats, but it does limit the value of such pressure and can even make it politically damaging to the threat maker, as Gromyko should have learned when his attempt to influence the recent German elections boomeranged. George Kennan is right when he tells us that “it takes two to make a successful act of intimidation,” and that therefore “no one in Western Europe needs to be greatly intimidated by them unless he wishes to be.”6 Although the reliability of the current Soviet adherence to no first-use would certainly be doubtful in a moment of ultimate crisis, it is a powerfully self-deterring position at all lower levels of tension. It provides excellent political support for those who are determined not to be coerced by verbal threats.

Existential deterrence cannot do everything. In particular it cannot take the place of other kinds of strength. There is nothing new in this proposition, and in particular it is wrong to suppose that Soviet actions in Africa or Afghanistan or Poland are somehow caused by new Soviet strength in nuclear weaponry. For those old enough to remember Soviet behavior between 1945 and 1949, when there was no Soviet bomb, it comes as a surprise to see the level of Soviet expansionism equated with the level of Soviet missiles. As far as I know, no one has ever shown a positive correlation between the growth of Soviet nuclear capability and the external misbehavior of the Soviet Union. The real correlation is always the one between applicable Soviet influences of all sorts and the situation on the spot. Some of these situations have allowed Soviet influence to grow in the last decade; others have not. And it was precisely while their arsenal was growing most rapidly that the Soviets suffered their greatest reverses—in China, in Indonesia, and in Egypt. And on our side it is worth remembering that the one instrument of influence that is entirely irrelevant in Central America is the bomb.

The deterrence I have been describing is not as good as a world without nuclear weapons, and the bishops emphasize that they “cannot consider it adequate as a long-term basis for peace.” But what is only the short term for men with a sense of history rooted in the immense age of their church can be quite a while for the rest of us. It would be wrong to suppose that the bishops are telling us we can quickly escape from a world of deterrence. What they tell us rather is that we must try. To accept our present position as the best we can do would be to let it become steadily more dangerous. We have ample evidence that the processes of bureaucracy and technology in both great nations are in no way automatically geared to the carefully circumscribed concept of nuclear deterrence that is acceptable to the bishops and that I have outlined here in language that is somewhat different from but still compatible with theirs. To moderate our reliance on nuclear deterrence, while not denying its necessity, is itself a great first task. It in no sense lets us off the wider search for peace and understanding, but if we are to live one day without these weapons, we must in the meantime live with them. To that first task the American bishops have made a signal contribution.

This Issue

June 16, 1983