INTERVIEW WITH THOMAS SCHELLING, March 4, 1986 AT THE KENNEDY SCHOOL. PROFESSOR SCHELLING WAS A CONSULTANT TO THE STATE DEPARTMENT DURING THE MCNAMARA YEARS AND A PROFESSOR AT HARVARD UNIVERSITY. HE WAS ALSO AT THE RAND CORPORATION, 1958, 1959, AND THE SUMMER OF 1961. YOUR INITIAL EXPERTISE WAS NOT IN NUCLEAR WEAPONS. YOU HAD OTHER TYPES OF INTERESTS. HOW DID THOSE EVOLVE INTO AN INTEREST IN NUCLEAR WEAPONS AND THEIR USE OR THE PREVENTION OF THEIR USE.

I got interested in nuclear weapons problems very indirectly. I'd been interested way back in graduate school in bargaining theory and game theory and conflict theory. Then I spent 5 years in the government with the Marshall Plan and waking on NATO issues watching negotiation. And I got very much interested in how nations make threats and bluff and how they make promises that are believable. And when I left the government and went to Yale University, I decided I would discover the bargaining process. And I very quickly became aware that nuclear strategy was essentially a bargaining process. It was using potential weapons for purposes of persuasion rather than action. And I wrote a few things that attracted some attention. And I got invited to the RAND Corporation to think about these issues in a technically experienced, knowledgeable environment. Met a lot of stimulating people, Learned a lot and became committed to this as a career for a decade or so.

WHEN YOU FIRST STARTED THINKING ABOUT NUCLEAR WEAPONS AT THE RAND CORPORATION, THE STRATEGY OF THE UNITED STATES UNDER EISENHOWER WAS MASSIVE RETALIATION. DID YOU THINK THERE WAS SOMETHING WRONG WITH THE DOCTRINE OF MASSIVE RETALIATION AND HOW IT APPLIED TO THE UTILITY OF NUCLEAR WEAPON

Well, by the time I got to the RAND Corporation hardly anyone believed in massive retaliation. Massive retaliation as enunciated by Secretary Dulles was to be the threat that would prevent small and medium size aggression on the part of the Soviet Union or Communist China or anybody. And it was my...the man who is currently my colleague, Bill Kauffmann, who wrote one of the first and most persuasive articles pointing out that it was simply not going to be believable to the Russians or to the Chinese that we would launch a huge nuclear war in reprisal for some incursion into a place like Korea or Southeast Asia. And I think... I think that he was writing. People were talking about it by 1954 or 5. By 1958 when I got to the RAND Corporation, everybody felt that massive retaliation was really dead as a strategy and there had to be some alternative way to prevent incursions in Europe, in the Middle East, in Southeast Asia or anywhere else.

AND DID THIS NEW WAY OF INCURSION OPEN UP POSSIBILITIES FOR THE APPLICATION OF YOUR THINKING IN TERMS OF NUCLEAR WEAPONS AS BARGAINING ETC. TO PREVENT THESE AGGRESSIONS?

Yes. Because in... in any uh... in any effort to prevent either aggression from starting or to prevent some kind of conflict from escalating, each side was bound to be terribly concerned about where it could go if it got out of hand. About the potential for the use of nuclear weapons if things blew up. And just about the time I spent a year at RAND, there were people in the United Kingdom as in the United States who were developing ideas about how to keep a war limited if one broke out. They were studying the experience of the Korean War and asking what kept the Korean War essentially limited to that peninsula. And in fact limited the land of that peninsula. And uh... uh... so the thought was that uh... if there were limited aggression, by all means try to stop it by limited means. Don't get caught bluffing with thoughts that weren't credible about massive retaliation. But always recognize that one of the main deterrents, even to a small non-nuclear war would be the fear that it could indeed blow up into a nuclear.

AND HOW DO YOU USE THAT SORT OF THREAT LET'S SAY IN THE EVENT OF FOR EXAMPLE, KHRUSHCHEV MAKING TRUE HIS THREATS ON BERLIN?

I think, and I thought then, that these threats are not threats you have to make. They exist. There were nuclear weapons in Europe. There were nuclear weapons in American possession in Europe. There were even nuclear weapons within the reach of uh... nationals of other nations in Europe. And the question was not, what do you tell Khrushchev about when and under what circumstances you would use them or how many you would use. Or on what targets you would use them. There they were. And the question was, Could Khrushchev or any Soviet leader be absolutely confident that these weapons would never be used in a way that would either to destroy his country or oblige him to use something in return that could escalate further upward. These weapons were perceived as very dangerous weapons. And uh... while one doesn't like to think that the potential Soviet aggression if there were Soviet interests in aggression, that it was deterred by the fact that things could get out of hand. Indeed the great threat was that in ways that neither side might possibly understand or control, things could get out of hand. Especially if say, Soviet troops were overrunning American or NATO troops that had nuclear weapons in their possession up in the front lines. So the threat was there. You don't have to make it... verbalize it.

COULD WE GO THROUGH SOME OF THE CONTINGENCIES IF BERLIN, FOR EXAMPLE: HAD THE SOVEIT UNION STOPPED THE UNITED STATES AT A CROSS POINT OR SOMETHING. AND HOW THAT COULD HAVE ESCALATED TO A NUCLEAR EXCHANGE. WHAT SORT OF NUCLEAR EXCHANGE WOULD HAVE HAPPENED TO MAKE THE THREAT OF THINGS GETTING OUT OF HAND A CREDIBLE THREAT TO KHRUSHCHEV?

It's very hard to reconstruct a plausible story of what could have been done, first by the Soviets and by NATO or by East Germans and by West Germans that would lead to the use of nuclear weapons. Uh... one can imagine what used to be called accidental use. Some unauthorized use of an occasional weapon. And it was thought that maybe that would generate a feeling that the war has started and all the nuclear weapons would go off. I doubt whether that was of any significant likelihood, but it could have been. I think the crucial thing was that even then, there was uh, a really appalling recognition on both sides that the first nuclear weapon used in anger after Nagasaki was going to be a watershed of a kind that no head of state, the United States of the Soviet Union, would want to embark on. Lyndon Johnson expressed it very well. I can't quite quote his words but he said uh...Any use of nuclear weapons is a presidential decision of the highest undertaking. No one knows what lies down that road once the first nuclear weapon is exploded. And I think that expresses his sense that uh... any use of nuclear weapons might get out of hand. There was no idea of the way to practice in advance on the control of nuclear weapons. There was no way to exercise the command and control system so that it could be contained. And it may well be that it was precisely the uh... the fear that they couldn't be controlled that made them so effective as threats. Precisely because they were so dangerous.

BUT HOW DO YOU MAKE A CREDIBLE THREAT THAT'S SO DANGEROUS?

One possibility is to put nuclear weapons under the command of somebody whose troops may be in danger of being slaughtered or captured, and let it be understood that physically he has the capacity to let his troops use nuclear weapons in self defense. It has always been considered the President's decision. But not by everybody. During the Kennedy/Nixon electoral campaign...

THE QUESTION IS, THE THREAT OF NUCLEAR WEAPONS BEING SO LARGE, HOW DO YOU AGAIN HARNESS IT TO BE AN EFFECTIVE POLITICAL BARGAINING TOOL?

I think it is not easily harnessed. It is almost never significant.

(REPEAT QUESTION)

The question of how one can use a nuclear threat is complicated. I think it's almost impossible to make a nuclear threat to make an adversary perform some overt act. I think it is much more likely that the threat can be made credible that if the adversary, let us say, engages in a massive attack somewhere, that the nuclear weapons will be used whether they were intended to be used or not. It has generally been considered in this country that only the President can authorize the use of nuclear weapons. On the other hand there are people who have proposed that authority be delegated to a commander in the field. And even during the Kennedy/Nixon administration some number, I think eight former senior officials of the defense department, both civilian and military, advocated publicly that in the event, or to take care of the event of communication failure between the White House and the commander of our troops in Europe, that commander should have the authority, on his own, to order the use of nuclear weapons in order to stop a Soviet advance, or to save his own troops. Now I think that was a poor idea because it meant that the Russians would know that nuclear war was on before the Americans knew it. That is to say, the Strategic Air Command wouldn't have warning. The President wouldn't have warning. Precisely if you do not have communications when you do not want the whole United States government to be taken by surprise because somebody in the field found it expedient to use nuclear weapons. But, if somebody had told Khrushchev that no supreme allied commander in Europe would ever use nuclear weapons without the authority of the President of the United Sates, I think Khrushchev might have said he had no guarantee that the President hadn't already delegated authority. And he had no guarantee that S(?) wouldn't on his own order the use of nuclear weapons. And therefore the fact that he was confronted with huge military forces in possession of nuclear weapons essentially made the threat. Now, it didn't make the threat clear out of Eastern Europe. It wouldn't even have made the threat leave the East Germans alone in case of an uprising. But it very clearly made the threat, Don't march west into this hornets nest because there are nuclear weapons here and who knows what happens with them if our troops begin to need to them in desperate self defense.

IS THIS THE ADVICE YOU GAVE PRESIDENT KENNEDY IN 1961 ON THE BERLIN CRISIS?

I didn't give him advice. I'm told that a few things I wrote may have reached him. And one of the things I argued that I think did reach President Kennedy, was what probably deterred the Soviets if they needed deterring was not a declaratory policy of the United States about what we would do if they did such and such. But rather the fact that the President, himself, couldn't exactly control what would happen. And therefore there was no guarantee that if the Soviets attempted let us say, even to conquer Berlin, that there wouldn't be a sequence of actions at the end of which one side or the other would be strongly tempted to use nuclear weapons. Maybe even use them just for demonstration, but it's hard to know what happens next if they're used for demonstration.

HOW DO YOU USE NUCLEAR WEAPONS FOR DEMONSTRATION?

Probably, badly. You... The question of how do you use nuclear weapons for demonstration. Well, it depends on whether you want to make a good demonstration or a bad demonstration. Uh... It was Secretary Forestal, I think, who asked Winston Churchill whether it wouldn't be a very good American demonstration to sail a cruiser and 2 destroyers into the Black Sea to show the Soviet Union that the American Navy couldn't be stopped from any...going anywhere on the high seas. And Winston Churchill said, if you want to sent the whole fleet into the Black Sea, it's a good idea. But you send a puny flotilla consisting of 2 or 3 ships. That will demonstrate to the Russians that you're afraid to go into the Black Sea. Now that, I think, is pertinent to the people who used to use the concept...the nuclear concept of the shot across the bow. Just before the Soviets begin to uh... let us say, in an invasion of West Europe, just before it's time to introduce nuclears on a huge scale, to uh...stop them, you fire the shot across the bow which is, I don't know, to fire 1, 5, 10 nuclear weapons or something to show them that we are serious. Whether that shows them that we are serious, I don't know. This came up in a controversy between uh... Secretary Haig and Secretary Weinberger early in the Reagan administration where the question arose: Did NATO have a policy of firing a shot across the bow before engaging in a substantial use of nuclear weapons. And uh...these of course are not the things that intelligent people debate in public. But the question of how one makes an effective nuclear demonstration uh...is something that nobody can know well because it has never been done. And it's very hard to know what any demonstration looks like to leaders on the other side. It's hard to know just how to gauge it so that it looks appropriately firm, determined, not too reckless. And not too much as though you're afraid of your own weapons.

WERE YOUR IDEAS ABOUT THE CONTROLLED AND DELIBERATE USE EMBODIED IN THE MCNAMARA ANN ARBOR SPEECH OF 1962 OF NO CITIES COUNTERFORCE?

The question of whether I had ideas about the controlled use of nuclear weapons an in particular of uh... not striking cities in and intercontinental war. This is a question I find difficult to talk about. I never particularly liked the idea of the controlled use of nuclear weapons. What I like is the idea that in the event they ever were used, God help us, by all means let it be controlled rather than uncontrolled. And therefore this was not devising strategies of how to use nuclear weapons effectively. It was how to keep them from being used in a way that everybody would regret instantly afterwards when it was too late to recall some bad decisions. And I was one of those several people to whom it occurred late in the 1950s, early in the 1960fs that maybe in the event even of a major war, the high priority thing to do was not to destroy enemy cities. In fact, quite the reverse. The important thing was to keep the enemy cities alive as hostage and make sure that the enemy knew that that was what you were doing so that the enemy might recognize that if he didn't equally spare cities on the Western side, the war would get disastrously catastrophic. The notion being, if you really want to destroy a hundred million Soviet citizens who live in urban areas, to punish them for the misdeeds of their leaders, you can wait 24 hours and see whether you can't stop the war instead. And I think several people had access to Secretary McNamara who must have got authority from President Kennedy to make this the announced Kennedy administration policy toward cities in the event of an intercontinental nuclear war. This essentially implied that any war should be nuclear...should be limited especially in the nuclear age.

THIS IS TAPE E05020. INTERVIEW WITH PROFESSOR SCHELLING. PROFESSOR SCHELLING, YOUR MAIN INTEREST YOU HAVE CONVEYED TO US WAS TO INTRODUCE INTO A POSSIBLY NUCLEAR WAR THAT COULD BREAK OUT TO STOP SHORT OF A HOLOCAUST. COULD YOU TALK TO USE ABOUT THIS PROCESS AND WHAT KIND OF DELIBERATIONS BETWEEN THE TWO PARTIES WOULD BE UNDERTAKEN AT A TIME?

It seemed to me important that, even in the event of an intercontinental nuclear war, one should plan in advance for where there might be a pause in hostility so that the war might possibly be brought to an end.

STOP TAPE. WE'RE GOING TO HAVE TO HOLD THAT DOWN... PEOPLE HAVE SAID THAT THE IDEA OF A LIMITED NUCLEAR WAR HOLDING CITIES FROM THE FIRST EXCHANGE, AWAY FROM THE FIRST EXCHANGE IS A WAR FIGHTING STRATEGY. AND THAT IT REFLECTED AN INCLINATION ON THE PART OF THE EARLY MCNAMARA TOWARD SEEKING WAYS IN WHICH TO FIGHT A NUCLEAR WAR. COULD YOU COMMENT ON THIS?

People have said that the idea of withholding attacked from cities in case of nuclear war constituted a war fighting strategy and it represented perhaps Secretary McNamara's notion that there was some you could safely fight a nuclear war without threatening the populations on both sides. I think that's very wrong. I think Secretary McNamara was always, not just opposed to frightened of nuclear weapons...

COULD WE START THAT...

I think Secretary McNamara was always not just opposed to but quite frightened at nuclear weapons and the notion of sparing cities in the event of a major war, I think was a kind of desperate idea on his part, that conceivably war could break out, perhaps not in any premeditated way, and if it did, it would be terribly important to try to bring it to a close before all the damage was done on both sides. The idea had some appeal to military planners. Military people do not like the idea of waging campaigns against civilians. They may not bend far over backwards to save the cities, but at least they don't like the idea that the purpose of their conduct in war is to incinerate tens of billions of ordinary folks on the other side. And I think what Secretary McNamara had in mind was if there's any possibility of a war's being stopped before massive damage is done on both sides, it'll depend on their having been some idea in advance that it can be stopped and it must be stopped. And it should be stopped before it became an orgy of reciprocated city destruction. And I think that was his idea but I think he also knew that very, very little planning was ever done on how to bring a nuclear war to a close. I headed a team in I think the summer of 1962 that looked into strategic problems for the forthcoming decade of the 1970s and one of the things we got interested in was the possibility that a decade later, by the 1970s, there might be much better command and control over nuclear weapons so that a war indeed, even if it got started, could be kept from just becoming a spasm destruction. And one of the questions we raised was how do you bring the war to a close? We couldn't find anybody in 1962 who had really thought about that. We even asked the question, how do you know when the war is over? And there hadn't been much thought given to that. And that was when many of us began to believe that unless you think in advance about where in the conduct of nuclear war there might be a stopping place, what are the things that once committed can't be recalled and what are the things that once committed are finite in amount, when if ever, is there a last chance to get the war under control and perhaps stopped before it becomes all out. These questions needed attention. I can't tell you whether they ever got adequate attention. Whether they still do. There's something very challenging about how to get off to a good start in a war and something so depressing, I think about how to bring a war to a close. And bringing a war to a close suggests bringing in politics and negotiation. If one side has to surrender, it sounds like being prepared to engage in surrender negotiations. And one can always then be accused of being prepared to do the surrendering rather than asking for the other side's surrender. And I think therefore that important as it is, and as important as McNamara recognized it to be in the early 1960s, to plan on keeping the war away from certain targets and to plan on being able to get it stopped while those targets were still in tact, I have no confidence that adequate attention has ever been given to that. I don't think adequate attention has given in the design of our weapon systems for their ability to participate in a cessation in the middle of a war.

WAS MCNAMARA MORE CONFIDENT IN 1962 THAT THIS MISSION COULD BE ACCOMPLISHED TO STOP A WAR AND TO LIMIT IT THAN HE WAS TOWARD THE END OF HIS TENURE AT THE DEFENSE DEPARTMENT?

I think Secretary McNamara became more and more pessimistic during the several years he was Secretary of Defense. Both about keeping a nuclear war under control and about exercising his own control over the planning of military campaigns, nuclear campaigns on the design of weapon systems and so forth. I think he felt that the technical and technological demands on strategic target planners made it almost impossible for them to design the more sophisticated kind of campaign that Secretary McNamara had talked about and I remember even one of his assistant secretaries telling me once that it probably would have been necessary to send an assistant secretary to SAC Headquarters, not so much to watch what the military planners did, but to communicate with them continually on what it was that the President and the Secretary of Defense had in mind. In short of that, I think it was just not possible for him to exercise the kind of control over the strategic target planning that he may have thought he could do back in 1962 and 1963.

SO HE ABANDONED OR HE BECAME MORE DISCOURAGED ABOUT THE IDEA BECAUSE OF THE DIFFICULTY IN THEIR IMPLEMENTATION IN TERMS OF DESIGN AND OPERATIONAL ADVISE OR OPERATIONAL INTERFERENCE WITH SAC OR WOULD YOU SAY ALSO BECAUSE HE MOVED MORE INTO THE QUESTIONS OF ASSURED DESTRUCTION AND OTHER SORTS OF CONSIDERATIONS?

I think the growing pessimism on the part of Secretary McNamara should not be attributed entirely to logic, analysis, technological developments. I think it's important to realize that the war in Vietnam was weighing terribly on him and by 1967, he was quite depressed about what was happening to US military forces abroad. What was happening to the United States. He was getting depressed about the difficulty of intelligently conducting a war of any kind. And I think part of his pessimism was from seeing that the war in Vietnam was not behaving at all the way he had earlier, more optimistically hoped it might go. And I think this colored his view of what a nuclear war in Europe would be like, or what any kind of nuclear war would be like. But I'm quite sure even back at the beginning of his stewardship at the Department of Defense, he was very skeptical about the use of nuclear weapons.

THE PROBLEM WITH ADJUSTING, THE COMMAND CONTROL AND THE DESIGN OF WEAPONS, ETC. TO FIT THIS MORE CONTROLLED TYPE OF ESCALATION, WAS IT A PROBLEM OF THE DIFFICULTY OF MCNAMARA OR ANYONE FROM DEFENSE GOING AND GIVING DIRECTIVES TO THE MILITARY PEOPLE OR WAS IT MORE A PROBLEM OF HOW DIFFICULT IT IS TO ADJUST THE PLANS, ETC. TOWARD A MORE FLEXIBLE AND CONTROLLED USE OF THE WEAPONS. COULD COMMENT ON THAT FOR US?

I've always been puzzled at why it seems so difficult for the Secretary of Defense to get his policy or philosophy properly incorporated into the strategic planning, planning for the strategic bombing of the Soviet Union. I think, it's several things. First, it is technically an exceptionally demanding task. If you just imagine especially back in those days when it was to be done with bombers that had to be flown all the way and they had to meet tankers to refuel. They had to navigate to their targets. They had to meet all kinds of unexpected interference along the way. It wasn't just a matter of aiming your missiles in peacetime and pressing a button. So that it was a demanding job and not the kind of job in which the experts would make an outsider feel welcome. I think it's also the case that the military believe that fighting wars is their business. And there are limits to how much meddling there should be. I remember a story shortly after Pearl Harbor, President Roosevelt asked the Secretary of War, the Secretary of the Navy, and the Secretary of State to sit down together and plan America's and make a statement of America's war plans. And the Secretary of State excused himself on grounds that it was none of his business. There has been a strong tradition in this country that peace is in the hands of the diplomats and war is in the hands of the military. That surely wasn't so in Vietnam. It wasn't completely so in Korea. The Cuban missile crisis was a case in which the President was literally in voice communication with people who commanded the ships in the Caribbean. And that is not the way the Navy thinks Presidential authority should be exercised. It should go through the chain of command and when you get to-the point where your dealing with something as militarily complex as designing the strategic bombing of the Soviet Union, you don't bring in a couple of civilian consultants and have them pose a lot of radical new ideas about some neat way to make sure that the war won't be as bad as it otherwise would be. Now, I don't know to what extent this was essentially stonewalling on the part of the military to what extent it was their belief that with civilians meddling in, only harm could result. I don't know to what extent it was just a sheer failure of communication. I think McNamara had around him a lot of people who early on somewhat offended senior military officers and it may have been that a little more diplomacy between the office of the Secretary of Defense and the Strategic Air Command could have gone a long way. But I am not confident that I can diagnose just what the problem was.

WAS THE EFFORT TOWARD A MORE, BETTER DESIGNED COMMAND AND CONTROL AND BETTER WEAPONS, MORE ACCURATE WEAPONS DESIGNED TO END THE WAR AND FOR MORE CONTROLLED USE, ABANDONED DURING THE 1960s AND PICKED UP LATER ON IN 1970s? OR WAS THIS TREND CONTINUED AFTER THE ANN ARBOR SPEECH IN 1962, TOWARD A BETTER COMMAND CONTROL SYSTEM?

I think the interest in better command and control has always been there, has always been strong. But that's for several reasons. In the 1960s, worldwide command and control began to look important and interesting. It became possible to think then about satellite communications... It became possible to think of linking the Atlantic Command the Pacific Command, the troops in NATO, SAC headquarters, the President, wherever he might be, and this looked important whether you are fighting a medium sized war in Europe, a small war in the Middle East or a global nuclear war against the Soviet Union. So that command and control has always been of interest. The particular kind of command and control that's necessary to bring a nuclear war to a halt, is essentially the question of what are the commands that the President wants to exercise, not what are the communications systems or the locks on the weapons that will, in effect, technically provide the communication and the control. And there you're always up against this main limitation. There is no President of the United State who will ever have the time to think about what he might want to command in the event of a nuclear war. It would be unimaginable that you or I could consider ourselves experts on the conduct of a nuclear war unless we had spent months or years of thinking about it. No President will ever spend days thinking about it. Presidents-, have other things to do. And ah—I don't know whether, in the chain of command a Secretary of Defense will ever go through 72 hours of simulated brink of nuclear war crisis in order to find out just what it is, just what the commands are that he would like to be able to issue and have obeyed at that time. So much depends on the plans in advanced. You see, by the time the President says, here's what I want to do, the answer is likely to be, you can't. We didn't design it that way. And it's too late now. So command and control is partly a matter of anticipating what commands a President would want carried out and designing not just the communication and control system, but the weapons themselves, even the... locations of military forces so that those commands make sense in the event. And this is a tall order and I think one of the reasons Mr. McNamara became pessimistic was that he saw that perfectionism would be required to make any of this work and perfectionism was just not involved where nuclear planning was concerned.

WHAT ABOUT IN WEAPONS DESIGN? DID HE BACK AWAY FROM THOSE OPTIONS OR REFINEMENT? IN TERMS OF THE DESIGN OF THE WEAPONS THAT WOULD BE REQUIRED TO CARRY THIS LIMITED MISSION, WAS THERE A BACKING AWAY FROM THE DESIGN AND THE PROCUREMENT OF THE TYPE OF WEAPONS THAT WOULD BE ADEQUATE FOR THAT?

On the design of weapons, it would make it possible to keep a war from going nuclear or it could bring a nuclear war to a halt. Secretary McNamara first tried very, very hard to get the European NATO allies not to feel a great reliance on nuclear weapons and to build up so-called conventional forces so that the war wouldn't have to go nuclear. He also did his best and so did President Kennedy, to keep General DeGaulle from proceeding with a French strategic nuclear force. And he used the argument, in his Ann Arbor speech he used the argument, conducting a nuclear war will require so much careful control and coordination that it will have to be done from one central headquarters. We cannot have a French nuclear war and an American nuclear war going on. Therefore, a French contribution could be only disruptive. But he didn't talk the French out of it. I guess nobody's ever talked national governments out of having nuclear weapons on ground that... you don't know how to use them, leave them to big nations like us. But he really did try very hard to, well, he tried to get electronic locks on nuclear weapons so that they would be less available for misuse. He tried very hard, not so much to avoid having nuclear weapons in Europe, but to avoid reliance on them by building up bigger European troop strengths. And with respect to strategic weapons, he was one of those who recognized that the worst danger on the brink of war would be the belief that we were tempting the enemy to strike first because we were vulnerable to attack. And in order to avoid that we should strike first, because they too, were vulnerable. And he therefore not only made a very strong pitch that all US strategic nuclear weapons should be made as nearly invulnerable as possible, but that and he said this to the Congress, he said, I would like to see the Soviets spend their money on making their own weapons capable of surviving an attack, because only then will they know that needn't fear a preemptive attack by us and therefore themselves desperately need to preempt. So his view was designing weapons that least invited preemptive attack and he perceived that this was a reciprocal thing. That it could be in our interest that the Russians never mistakenly believed that we were about to attack them preemptively.

INTERVIEW WITH PROFESSOR SCHELLING. PROFESSOR SCHELLING, THE NOTION OF NO CITIES AND A CONTROLLED WAR DEMANDED OR REQUIRED COOPERATION FROM THE SOVIET UNION TO BEHAVE LIKEWISE IN THE EVENT OF A NUCLEAR EXCHANGE. WERE YOU ANY CONFIDENT AND WAS SECRETARY MCNAMARA ANY CONFIDENT THAT THE SOVIETS WOULD AGREE TO SUCH A SET OF RULES FOR FIGHTING A WAR?

It's an interesting question whether, when Secretary McNamara proposed fighting a war without deliberately striking cities, the Russians were listening.

CUT.

It's an interesting question whether, when McNamara adverted to a no city strategy, the Russians were listening and if they were listening, what they thought they heard and how they thought about it. It's absolutely correct that the strategy required reciprocity. The whole idea was that Soviet cities were worth more as hostages in our hands than as burning embers. But the purpose was to make clear to the Soviets that even if war had started, no matter how bad the war appeared to have got, it could get much worse if we didn't reciprocally spare each other's cities. So this is not a policy that has much unilateral appeal, unless you simply like the idea of not gratuitously killing the foreigners, but I'm sure what Secretary McNamara had in mind was saving Americans, not saving Russians. It is interesting that by now, everybody I talk to in this business takes for granted...

WE HAVE TO CUT. I'M SORRY.

It is interesting that the strategic literature nowadays, and all the people I talk to, take for granted that the Russians would naturally have something that is usually called a reserve force. And as best I can understand it, what people mean by a reserve force is a well-protected force, a force not easily attacked and destroyed, that as a last resort would be available to destroy American cities. A reserve force whose function is essentially to see that the Americans are deterred from the worst all out attacks, even in the middle of war itself. That means that in the course of twenty years, Americans in the strategic nuclear business have gone from considering the no city strategy a preposterous one to one that is so obvious that it's taken for granted that the Soviets reciprocate the general idea. Whether this is based on any knowledge that the Soviets actually do, I don't know. My own feeling is that this is an idea that made much more sense to the Soviets than to the Americans. I think the Americans typically have rather formal, grand and honorable ideas about warfare and I think Soviet leaders are much more aware of the role of violence, brutality, ugly diplomacy, both in their internal politics and in dealing with other nations, and I don't think they have nearly as much traditional baggage about the way to use military force in war and I think if they saw that it suited their purpose to treat American cities as hostages in order to keep us from attacking their homeland populations, it might appeal to them much more quickly than it would appeal to a typical American. One other interesting thing. Um, when Secretary McNamara proposed that a war in Europe need not go nuclear, and he therefore wanted European countries to raise more non-nuclear forces, the Russians always made fun of him for years and years. They argued that no war in Europe can be kept high explosive. It's bound to go nuclear, it's bound to escalate all the way. And yet, by the end of the 1960s and all through the 1970s, the Soviets were spending enormous amounts of money on military forces in Europe, especially tactical aircraft, that would have been virtually worthless in a nuclear war, as though they were counting on a very important possibility that a war, if it started would be non-nuclear. I don't believe they've ever come out and declared that they too reciprocated the notion that in the event of a European war, it could be a large non-nuclear war. They played around with the idea of no first use declarations. But that's a little different. This would be a declaration that a war could actually get large, that is, require the available military forces without going unclear. And yet it looks as though they must, I wouldn't say, be confident, but they must put very high stakes on the notion that even a big war in Europe would be non-nuclear or that there would be some reason to keep it non-nuclear as long as possible. I find this to be what you might call, 'unconscious arms control'. It looks as if the Soviets reciprocated eventually the McNamara ideas of the 1960s. And if you think that arms control is a notion that includes limits on the use of weapons and forces, as well as limits on their design and deployment, than this is what you might almost call an unnoticed, unacknowledged, almost an unconscious reciprocated arms limitation, namely spending lots of money to make it possible to keep a large war non-nuclear.

DID WE SPEND THE SUFFICIENT AMOUNT OF MONEY AND EFFORT IN DESIGNING THE WEAPONS REQUIRED TO HAVE A CONTROLLED AND SOPHISTICATED STRATEGY? WHAT KINDS OF FORCES ARE NECESSARY TO CARRY OUT A NO-CITY CONTROL CAMPAIGN? IS ACCURACY NECESSARY AND, DID THE UNITED STATES, SECRETARY MCNAMARA INVEST ENOUGH ATTENTION AND MONEY IN THE DESIGN OF THE APPROPRIATE WEAPONS TO DO THIS?

What kinds of weapons you need in order to conduct a no-cities strategy depends on what you want to be doing if you're not striking cities. And that really has become the most controversial strategic issue that we have. There are those who believe that it should be possible to wage even a war to ultimate victory against Soviet military forces, even Soviet industrial capabilities. Trying very hard to spare cities, but mainly in the sense of not killing any cities deliberately, but also not bending to far over backwards to spare them, with the belief that ultimately the war will be over and not all the cities will be destroyed. And it's better to do the most aggressively effective job possible on Soviet military forces of all kinds. Then there are other people who believe that that kind of war is hopeless because it would entail striking so many targets with so many weapons in so much of a hurry that there'd be no likelihood that anybody could keep on top of what was happening or even notice whether cities were being deliberately struck or occasionally struck or accidentally struck. And I think at the present time, starting say with ah, maybe Secretary Schlesinger, there has been the notion that if you're not going to strike cities, you've got to do something. And the question is what is that something? There are people now who argue that Soviet leaders don't care about their people, they care about their power. Their power is their army and other military forces. If you attack, if you threaten to destroy their army, you threaten to destroy what they love. If you threaten to destroy their people, well they can breed more people, they don't have to worry about that. I tend to think that is nonsense, but it also implies that if we like people and they like armies, ah, they may be holding our cities hostage against their power base, and to go attack their power base would be the equivalent of their attacking our cities. I never hear this discussed very reasonably these days, but there is a very puzzling question if you're not hitting populations deliberately, what are you doing as the war goes on? And ah, there seems to be a strong compulsion to do something that is pretty active and noisy. My personal thought is that cities would never be completely off limits, that ultimately when the military forces have done what they can to each other, then the war gets dirty, and the war gets serious. Because then is the time when they're going to negotiate over the state of the world when the war is over and what they're going to negotiate with is these prime counters that they have been saving in reserve, namely people in cities and the weapons to destroy them. And if, which I don't believe possible, but if the war should reach that stage, where there has been a vigorous military campaign waged on both sides and there is not much more to be done militarily, then I'd say, we're down to where we're bargaining over people and that's the part of the war that counts. And...

WE HAVE TO CUT...WHAT IS YOUR PERSONAL THOUGHT ABOUT WHAT SHOULD BE HIT IF CITIES ARE AVOIDED AND HOW THAT WAR WOULD DEVELOP OR NOT?

The question of what targets you should be attacking in the Soviet Union in a nuclear war if you're not attacking cities is not one that I can really answer for you. I think that there are going to be two kinds of targets. There are probably going to be a lot of urgent military targets that somebody is going to insist on destroying and then I think if the President manages to keep control, he may end up engaging in the same kind of counter population threats and destruction that throughout history has characterized some of the most vicious wars, whether they are insurgency wars or terrorists wars or wars waged so that people will become fatigued of the war, like the US Civil War. And I think that at that point, the President's going to begin asking himself, know that I have a hundred million Russians in my hands and the have a hundred million Americans in their hands, know how do we bargain in order to arrive at a military stand down, at a situation that is not susceptible to double cross surprise attack. And on what kind of a time schedule can we bring about some kind of armistice and erect a capacity to monitor it? But ah, I can't claim that I put myself in the position of a President at that stage to think about just what he would do. It's a little like the ending of Dr. Strangelove.... There is a president who believes momentarily that he has just destroyed a Soviet city and that the Soviets in return are going to destroy at least an American city and he's having to ask himself how do you play this game to that most of your cities are left when we brought it to a close?

DO YOU BELIEVE EVERY WAR WILL INEVITABLE ESCALATE TO THE POINT OF A TOTAL DESTRUCTION ON EACH OTHER'S SIDE?

Don't say every war, I don't expect, I don't expect any such nuclear war, but I certainly don't expect more than one. Yeah. I tend to believe that US forces have been fairly well designed so that they don't have to go off in a hurry. I believe that both military and civilian leaders in the United States have spent forty years learning that the Russians aren't eager to jump into a great big nuclear war, that the Russians know that they could provoke the United States into a sequence of actions that could escalated upward into nuclear war. That the United States doesn't have to prove itself every time there appears to be some kind of challenge. And that nuclear war is to be avoided at almost all costs. And I have a hunch that the Soviet leaders feel pretty confident that the Americans are not about to launch nuclear war in any kind of crisis that is at all like any we've ever seen during these forty years of post-war history. So, I think the expectation of war on the part of Soviet and American leaders, is pretty calm, not much expectation. And I would like to see it to stay that way. As a result, I think escalation to nuclear war is extremely unlikely. I have participated in some gaming exercises back in the 1960s that began on the premise that almost any war would bubble up into a nuclear war and the question was to study how the beginning of a nuclear war might be. It turned out after many, many attempts at exercises, with all kinds of people, military and civilian, German's, French and British, as well as Americans, that in these exercises, it was almost impossible to get-anybody to make decisions that would cause this game war to go nuclear. And I think it's because everybody is aware as, even Eisenhower was aware, despite the fact that he occasionally said that nuclear artillery is no different from other artillery, it's just a little more economical. I think he knew that nuclear weapons were different. That's what the test ban was all about. There was a huge symbolism about how nuclear weapons are truly different. And I think on both sides there's a feeling that nuclear weapons are a watershed that people stay away from just as long as possible. Therefore I tend not to believe that every war escalates nuclear and becomes a holocaust.

I HAVE TWO QUESTIONS. ONE IS DO YOU BELIEVE A NUCLEAR WAR, ONCE YOU CROSS THE THRESHOLD INTO NUCLEAR WAR, WILL IT BE LIMITED OR WILL INEVITABLE ESCALATE INTO MUTUAL DESTRUCTION? ... ONCE YOU CROSS THE NUCLEAR THRESHOLD, AND YOU HAVE THAT FIRST NUCLEAR EXCHANGE... TODAY, DO YOU THINK IT'S POSSIBLE TO KEEP THAT WAR LIMITED AND END IT? OR WILL IT ALMOST INEVITABLY ESCALATE INTO THE DESTRUCTION OF BOTH SIDES?

Will any nuclear war, no matter how it starts, or where it starts or on what scale it starts inevitably escalate to a huge intercontinental war? Certainly not inevitably. I really think it's doubtful whether even a nuclear war that began in some theatre would escalate to a large-scale intercontinental nuclear exchange. I don't think this subject gets quite the attention it deserves. But it's very hard to study the process. It's very hard to simulate in your own mind or otherwise, what things would be like in the event that nuclear weapons began to be used. But, you see, if you just ask the question, would anybody initiate the use of nuclear weapons on a small scale, if he expected it to escalate, the answer must be no. If you expect it to escalate, you're wasting the opportunity to start the big war on your own terms. Your simply giving the enemy the chance to reciprocate in a manner of his choosing. Therefore the mere use of nuclear weapons, whether by us or by the Soviets, ought to be a pretty convincing demonstration that the war is not expected and not intended to get a whole lot larger. And that should put both sides on notice that we've now got a nuclear war that we're going to have to get stopped. But to jump to the conclusion that because nuclear weapons have been used on the high seas or somewhere in the Middle East, or even on a small scale, in central Europe, to jump to the conclusion that now we have lost complete control, now the President has no more decisions to make, now somebody simply pulls the chain and lets a huge nuclear war happen, would be a complete, besides being an abdication of responsibility, would be a, I think an absolute utter misreading of the possibilities of a stopping a war that has, in some theatre begun to go nuclear.

PROFESSOR SCHELLING, YOU CONDUCTED SOME WAR GAMES IN THE EARLY 1960s, ESPECIALLY AROUND THE QUESTION OF BERLIN AND ACCESS TO BERLIN. COULD YOU TALK TO US ABOUT SOME OF THOSE GAMES AND PERHAPS EVEN REFERRING TO PARTICULAR PEOPLE WHO PARTICIPATED AND THE OPTIONS THAT YOU OFFERED, AND HOW THE GAMES NEVER ESCALATED INTO A NUCLEAR WAR?

The games that I participated in, back in the early 1960s, involved military and civilian people from the US government, occasionally from other NATO governments. They were essentially decision games. The players were provided with a scenario that was a story that got them up to the point where something was going on they had to make some drastic decisions. And they were in separate teams, typically called red and blue team, with the control team to monitor what they were doing. And the object was to try to confront them with circumstances in which they would have to take a fairly aggressive action. And then study the process by which the two teams, in effect, manipulated in imagination, the forces they disposed of. And I first thought that it would be easy to generate situations that would escalate even up to the brink of nuclear war. And indeed the first such— the first such game I was involved in was intended to study what happened back at SAC Headquarters on grounds that the important things would be there rather than in Europe, in the belief that the war would escalate nuclearly, so rapidly. Well, after many, many games, nothing much ever happened at the SAC Headquarters because the wars you could generate in the games rarely got close, even to any kind of use of nuclear weapons. And I think that was for several reasons. One, ah, back then, in addition to the tradition that nuclear weapons were very different than other weapons, there was a very strong tradition that Soviet and American troops shouldn't ever shoot at each other. That once Soviet and American forces in uniform, under authorized command, engaged each other militarily, that was war, and I think there was a much greater willingness to foresee the Red team to use East German troops to confront the Americans than to use their own troops. And there was a tendency for American leaders to believe that they don't dare annihilate us, because if they do, then the United States will have to do something in retaliation that may get quite out of hand. Usually these scenarios would begin with something like an East German uprising, and the Soviets would move in forces, into East Germany. Berlin would be in the middle of it. The Soviets would find it necessary to use their own troops to take over Berlin and Berlin had the American and the British and the French garrisons, and this would mean there was an outpost of maybe seven thousand Americans and five or six thousand French and British troops that might have to be preserved, rescued, reinforced or something. At which point, various things, like the things that President Kennedy actually did once, ah, sending troops down the Autobahn with orders to keep moving, remove any obstacles and fire when fired upon and go all the way to Berlin, things like that would be attempted. And these were essentially were games of what you might call games of chicken, games of daring, games of attempting to put yourself where the next dangerous move was up to the other side, who would then back down, rather than take that dangerous move. One thing we learned, very often, was that whatever you did with whatever you did, promptness was often more important than the scale on which you did it. That there were many things that could be accomplished by a platoon in terms of confronting the Soviets with a... but if you didn't have a platoon available in twenty-four hours, you needed a company. But if you didn't have a company available in seventy-two hours, you needed more and therefore the speed of response, readiness for quick response, was interesting. Another thing we discovered, I enjoy this because it was involved on the French, the British, the Germans, and the Americans did a game together, It was to test a new procedure, the ah, there had been an agreement that crisis would go in stages. And there was agreement on what was to happen in stage one, stage two and stage three. And the game was to test how much difference it made to have predetermined decisions for stage one, stage two and stage three, and when the game was over after about two and a half days, everybody agreed that the main thing that this new procedure had done was to change the argument from what do we do next to what stage do you think we are in now? And they argued the same arguments in terms of whether these are the conditions under which we are pre-committed to do certain things. Ah, the control team, as it was called, typically tried to throw in lots of mischief. It tried to throw in a lot of misinformation about what each other was doing, in order to get the two teams to take more aggressive action. It rarely worked, and I think it's for two reasons. I think the kinds of people who populate high levels in the US government are very cautious, conservative, responsible, even in a play environment, unwilling to do things that they genuinely think would be reckless in real life. And I think it also reflected the fact that those same Americans attributed very similar behavior to the Soviets. And I think this, I hope it suggests that real escalation is hard to get going. But at least it suggested that the kinds of Americans who were in high military and civilian positions in the 1960s, were the kinds of people who couldn't even in a game bring themselves to be the least bit reckless, even with nuclear weapons or with any other kinds of weapons.

HOW WOULD YOU ENVISION, COULD YOU IMAGINE FOR US HOW A WAR IN THE 1960s OVER BERLIN FOR EXAMPLE, WOULD ESCALATE INTO STRATEGIC NUCLEAR CONFRONTATION, GOING THROUGH THE TACTICAL STAGE PERHAPS A SCENARIO?

Could I design a plausible scenario, according to which something could break out in Europe, say around Berlin, that would entail the introduction of military force by both sides on larger and larger scales until nuclear weapons were actually used?

RIGHT.

I can't. That doesn't mean it happened—it couldn't happen. It means that it would have to happen in a way that is unpredictable I don't believe there's any predictable way it could go. If there were some point at which it became now inevitable, some final step that made it inevitable, that the war would become a large nuclear war, nobody would take that step. These would have to be compounding steps, each of which had implications that were mistakenly perceived or so dimly perceived that people could actually take them. It would require what has traditionally been called, the "fog of war" to get people to make decisions that could escalate that way. There's nothing at stake—there wouldn't be anything at stake that would be worth a major nuclear war. I don't mean Western Europe's survival isn't worth a major nuclear war, I mean the Soviets would not want military conquests of Western Europe at the sever risk of a major nuclear war. And therefore, I don't think they would take any steps that would provoke us to initiate a major nuclear war. We would have to misread them completely in order to do that. And they may be afraid that we would, but what I'm saying is that if a nuclear war did come about, and I can't guarantee that it won't. If it did come about, it would come about through a process that was essentially out of control. And not understood, with people not completely in command of what was going on and what they were doing. And with, somewhere along the way, decisive events occurring, or decisive steps being taken. The implications of which were not correctly perceived.

BUT IF, FOR EXAMPLE, WARSAW PACT FORCES WERE TO OVERRUN NATO DEFENSES OF BERLIN, LET'S SAY 1962, WOULD HAVE THE ALLIES THAN ESCALATED TO A TACTICAL USE OF NUCLEAR WEAPONS AT THAT PARTICULAR TIME?

If the Soviets had launched a premeditated major attack on Western Europe in 1962, and had breeched conventional defenses, would nuclear weapons have been introduced? I feel pretty sure they would have. It would have depended a little perhaps on who was President, but I think at that time, no President could have withheld from troops in Europe the weapons that their commander said were needed to protect his troops, and to protect the territory of Western Europe to which by treaty we were committed to protect, just as if it were the territory of the United States. In principle, our treaty obligation, our military plans were we would use nuclear weapons to defend West Germany the same as we would use them to defend Cape Cod. And I think a President would indeed have authorized nuclear weapons along the way. He might have authorized them, not by merely releasing them to the commander to use as he pleased, but as we discussed earlier, through some kind of effort to demonstrating to the Soviets that if you don't think we would use nuclear weapons, you are wrong. Stop what you're doing. You just going to oblige us to use nuclear weapons and then the war will get worse for both sides. Now, the difficulty, if you believe that, the difficulty is to find a plausible reason why the Soviets would have embarked on that kind of an adventure, down the road at the end of which they were going to face nuclear war. And if they saw that the President of the United States, with several divisions of American troops about to be destroyed, captured, isolated, with West Europe about to be conquered, that the President wouldn't use nuclear weapons, then I think you'd have to be a very optimistic Soviet leader to think you could count on an American president with nuclear weapons available, not to allow them to be used and it would have been a very rash gamble if the Soviets thought that they could win anything without running a dreadful nuclear risk. And no reason to suppose if it went nuclear that they could come out of it with anything to show for it. So, we're back to the beginning, How do you devise a plausible way that the Soviets got engaged in this attack that poses the question, would nuclear weapons have been used? And one of the reasons why in these games we used scenarios like an East German uprising or a lot of Polish sabotage against Soviet troops moving in to reinforce East Germany or something on the order of the invasion of Hungary that maybe transposed to Yugoslavia, where there might be some notion, well, maybe if the Soviets invaded Yugoslavia, NATO troops would attempt to ah, some kind of counteraction. We tried to make these up, and it was extremely difficult to generate a plausible story, a believable story of how the Soviets or the Warsaw Pact and the NATO treaty powers could step by step, without doing foolish things on both sides, get to where nuclear war either was on or was inevitable. This shouldn't surprise anybody. Once you recognize that the nuclear war will be regretted by both sides, then clearly they could only get there by a process that included mistakes.

SO THE STRATEGY OF FLEXIBLE RESPONSE IN EUROPE, WHICH THREATENED TO ESCALATE A STEP BY STEP, FROM A CONVENTIONAL TO A TACTICAL NUCLEAR, THEN TO A STRATEGIC NUCLEAR WAS NOT TO BE IMPLEMENTED AS SUCH?

I think the so-called strategy of flexible response was many, many things. I believe the first use of the term 'flexible response' was in President Kennedy's first budget message, early in 1961, in March or April. And various people thought it meant different things. Some people who were concerned about strategic nuclear weapons thought it meant something like the no-cities strategy and that that's all it meant. Other people thought it meant let's not use nuclear weapons in Europe unless we have to. Ultimately, there were people who thought it meant we must have everything from Green Beret up to strategic intercontinental weapons. And there were occasionally people who thought flexible response meant you must be able to escalate in as fine tuned, as nice a fashion as possible, so that whatever the enemy does, you can top him ever so slightly. Ah, I am convinced that Secretary McNamara never thought that there was some comfortable strategy where you simply lay it out conventionally. If that doesn't work, you use a few... if that doesn't work you use a few more nuclears. I think, as I said, flexible response meant different things to different people according to whether they were primarily concerned with intercontinental war, primarily concerned with European, primarily concerned with Vietnam or Taiwan and ah, I think when most people tried to talk about how the war in Europe would go nuclear, as I said a moment ago, they couldn't figure out a sequence of rational steps by which it would go nuclear. But they could see that there were a sequence of not so rational steps by which it could go nuclear, and maybe the flexibility was that the nuclear weapons were there and could get out of hand if conventional escalation went too far.

COULD YOU REPEAT THAT LAST STATEMENT, BUT WITHOUT SAYING "AS WE DISCUSSED EARLIER"...

Those who thought of flexible response in a nuclear context, I think rarely had in mind that there was some well defined point at which you moved from a high explosive war to a nuclear war. Or at which you moved to a small nuclear war and another point at which you moved to a larger nuclear war. There were people who debated whether once nuclear weapons were used, they should be used only in defense of West Europe or, only as many kilometers east of the East/West German border as they were used West. There were those who proposed that using nuclear land mines to stop advancing troops was a non-provocative use of nuclear weapons. There were people who talked about whether nuclear weapons should be used on Soviet soil, on say, air bases in the Western Soviet Union. These were talked about, but I don't think anybody ever had any careful notion of exactly what it was that the other side might do that would provoke one or another of these widening concentric circles. I think the strategy of flexible response was essentially a ah, first it was something of a euphemism, keep control on nuclear weapons, be prepared to fight conventional as long as possible until the Soviets see that this is getting too dangerous to keep up. And second, if you do introduce nuclear weapons, don't for goodness sake, feel obliged to introduce them by a massive attack on the Soviet Union. See if you can confine it to Western Europe. See if you can confine it maybe to the point of contact, the front lines in Western Europe. Keep it limited. Don't let it get out of hand. And recognize that once the war goes nuclear, the Soviets may be as interested in bringing it to a stop as everybody else, and make sure you give them a chance to bring it to a stop. But this would have been improvisation. This would not have been an articulated strategy.