Michael Wallace received his BA and MA at McGill University and his PhD in political science from the University of Michigan in 1970. He worked there with David Singer, a peace researcher who founded the
CANDIS: You've directed your efforts lately to studying the risk of accidental nuclear war. When this is mentioned, many people envision a "bolt out of the blue."
WALLACE: Or that someone will walk into a missile silo with a machine gun or something. Lots of movies have been based on that sort of thing. That's not going to happen. Oh, I guess there could be an inadvertent release of one weapon, say it's very remotely possible. But not a coordinated attack that leads to a strategic exchange. There are just too many locks on the system for that to happen.
CANDIS: What's the real danger, then?
WALLACE: During a crisis, they take the locks off. In peacetime, there are so many locks on these things, it takes an incredible effort to pry them off. That's the way it should be. But in a crisis, you have to take those locks off because you have to be prepared to use them. That transition is fraught with danger. This is most dramatic in the European theatre, where plans call for an early release of the locks in the event of any European crisis that could be the start of a conventional war. When these locks (which are called "PALS"-for "permissive action links") are off, every tactical commander from Hamburg down to the Swiss border has the right to use nuclear weapons. It doesn't take much calculating ability to figure how long it would take one of these guys to actually use a nuclear weapon. It would be days, not weeks. Once the nuclear weapons start flying, nobody can predict what would happen. I don't see any way to stop escalation at that point.
CANDIS: In a crisis, many people will have control of nuclear weapons. How reliable are those people?
WALLACE: In a period of eighteen months, over 6900 U.S. people lost their access to nuclear weapons either because of daily use of cocaine, heroin, or hallucinogenic drugs; the commission of a major crime such as assault with a deadly weapon or murder; alcoholism; or what's called "deviant acts of behavior"-which can include some pretty crazy things-basically people who are mentally ill. I don't know about you, but the thought really bothers me that 6900 people had access to nuclear weapons who were either using hard drugs, alcoholics, felons, or mentally ill.
CANDIS: What about false alarms in the NORAD early warning system?
WALLACE: The Center for Defense Information in Washington pried loose some numbers from NORAD about false alarms, using the Freedom of Information Act. The facts were very scary. We're talking of hundreds of false alarms per year and one extremely serious one per year on average. These are the ones that actually change the alert status. There were a couple in particular that were so bad-in 1978 and 1979--that a joint bipartisan Senate Committee, chaired by Senators Hart and Goldwater, looked at these alarms and came to the conclusion that there had been really major malfunctions, caused by totally trivial events. In one case, a fifty-six cent computer chip failed and in the other case a human operator put a computer tape on the wrong tape drive. When such a trivial, easy-to-occur event causes a catastrophic malfunction, this is when you have to start getting worried.
Let's put this into perspective. The Cuban missile crisis altered the alert status of both superpowers for thirty days. Of these, 13 days were the period of deepest crisis. Now if one of these false alarms were to occur during the crisis, the feeling was that there would have been a war.
From talking with people on the inside, I can say this:
Under certain conditions in a crisis, a single unsubstantiated tactical warning can cause a war. They would have to have some antecedent reason (either political or from the early warning sensors) to think that the other side is going to launch. If at that time something comes along, bang that would be it!
CANDIS: Bruce Blair and Robert Aldridge have both stated that the U.S. is probably on a launch-on-warning condition during crises. What do you think?
WALLACE: Yes. The flight times of the Pershing IIs and the Soviet close-in missile-firing submarines are so short, you have to fire almost instantly upon receiving the warning, in order to protect your command centres. That means there's no time to check and see if you have the real thing or a false alarm. You're talking of a decision time of less than a minute. Besides, much of the decision will be automated and that's not subject to any intelligent judgment. Frankly, whether you have one, five, or even ten minutes under this situation is not all that relevant. Human decision-making times are not in that scale.
CANDIS: Would it alleviate danger to keep a human in control? "Keep a human in the loop," as they put it?
WALLACE: "Keeping a human in the loop" is impossible. Sure, you can keep a human formally in the loop, but all he will be doing is watching screens go berserk. Under such conditions, human beings operate according to pre-programmed responses and standard procedures, so it is irrelevant whether it's a human being or a piece of software. If you only have a minute, a human being performs no better than a computer. It's only when you've got time to sit and think carefully that you can perform in that kind of judgmental role. We won't have that much time.
CANDIS: What measures can alleviate the dangers we are facing?
WALLACE: First, let's agree to get rid of fast-attack systems and short flight time systems. That's possible. The technology of discovering close-in submarines is good enough-not for targetting, but the Americans could be sure the Soviet subs were not too close for the agreement. They could pull the Pershing IIs out. Those things are tradable and if they really wanted to do it, they could. They still have so much overkill, it's just silly to hang onto them.
Also Blair advocates measures to harden C3I [Command, Control, Communication and intelligence systems] so they can't be decapitated.
Above all, keep secure command systems, so the other side will know that they can't save themselves by chopping off your head. Those are the first prudent measures to take.
Next, upgrade the level of communication by bringing third and fourth parties into the loop. For example, why can't we have hardened communications to the U.N. Secretary General? All he can do in a crisis now is pick up the phone, and that may not work. In fact, there is some question now as to whether the hotline is really secure and hardened. That could be improved a lot. It's fifteen-year old communications technology. That's pretty old.
These are only minor things. They address only the immediate problem and they treat symptoms, not fundamental causes. The fundamental thing is that nuclear disarmament is a prerequisite for survival. There's no question about it.
We need both a short term and a long term strategy, and we mustn't conceive of them as mutually exclusive. The short term strategy is vital because you can't go to the public and say, "I have the answer to accidental nuclear war... Everyone is going to get rid of all their nuclear weapons. We can't wait that long. That's going to take a while. One the other hand, we can't say, "It's enough that we take these minor measures and let the military go on playing with their toys just the way they were before." We have to handle the worst danger and then plan follow-up steps. Unfortunately, that requires some political accommodation between the U.S. and the Soviet Union. The survival of humanity requires a deal between the superpowers. It goes against the grain, but it's a fact. That's why the mind-set of today's administration in Washington is so deadly.
CANDIS: Is there a danger that hardening the Command and Control systems-making them survivable-might encourage nuclear war-fighting strategies?
WALLACE: If you assume that you'd use whatever weapons you have left over to knock out the other side's remaining silos. But I doubt that assumption. That wouldn't be the mind-set after a nuclear exchange. I don't think that the cold-blooded rationality of the strategic planners would prevail after 100 million people or so have been killed. I don't know what would prevail, but I have the feeling that those plans would break down pretty damn quickly.
What a secure C3I does is to allow a political or military official to say, "Do you really think that's the big one coming in there on the screen? Well, we can wait. If we are wrong, there are the backup command centres in the submarines, but if it does turn out to be a false alarm, we will have saved the world by waiting." They will be able to say that and make it plausible.
We're faced with a bizarre situation. We need to free ourselves from ideological blinders to look at our options. Even a little safety is better than nothing. I wouldn't say, "either we abolish nuclear weapons or I don't want to hear about it." We have to work for midrange goals under the present circumstances. So I say that, on balance, hardened C3 I is more stabilizing than destabilizing in a crisis.
CANDIS: How does Star Wars fit into the threat of accidental nuclear war? If you listen to some of the proponents, they claim it will not increase the risk but reduce it.
WALLACE: That's because they're talking about one missile being fired and then it can be knocked down. I don't see any plausible scenarios for one missile being fired. And with Star Wars, the other side's missiles would fire during their boost phase, so the warning time will be seconds instead of minutes. And then there's certainly no way to leave a human in the loop. They have to fire automatically. Sooner or later something will go wrong and they are going to fire at the wrong thing.
Also, an inherent part of SDI is anti-satellite warfare. Now suppose two adversaries are staring at each other nose-to-nose in a crisis. And suddenly one side blinds the other's satellites, which are the only way they can tell when an attack starts. That side is going to assume that the blinding of his satellites is the immediate precursor to an all-out attack. What else could he assume? So he, in turn, will want to shoot down the other side's missiles, and to do that he'll want to blind their satellites and keep their attack from being effective.
So taking this defensive measure will instantly be interpreted by the other as a precursor to an attack. It might all be automated. Most of the SDI research that's likely to be successful-at least for the short term-will be anti-satellite things. It's not hard to knock down satellites. They're just floating up there and they're easy to hit. But to develop that capability is incredibly destabilizing.
CANDIS: You organized a conference on the risk of accidental nuclear war at the University of British Columbia. Tell me about it.
WALLACE: I thought: The trouble is, we're dealing with a problem that's at the interface of disciplines. The strategists know something about how nuclear command strategies work, but they don't understand the computers. Crisis behavior people don't necessarily understand the nuts and bolts of strategy. Computer scientists are not experts in politics. We need to put all these people together in one room.
Besides the interdisciplinary perspectives, you get international perspectives. Americans, Europeans, and Eastern Europeans look at this problem in very different ways.
It took over a year to organize the conference. We had everyone from mathematicians and computer scientists to people who were intimately familiar with their country's command and control systems. We had social scientists who were familiar with deterrence crisis behavior and people who were familiar with human reliability studies. We had people familiar with every possible body of knowledge that would impinge on the problem, and we put it all together. And it was pretty darned scary. The consensus was that the danger is very real. Definite measures have to be taken. It would be worthwhile to institutionalize the study of this, which could be done in several ways. We could have centres for the study of crisis decision-making. There have been proposals to strike a committee of the Forty Nation Disarmament talks in Geneva to study the issue.
I was disappointed that neither the American nor Canadian military were interested in discussing the issue. Both were invited to send representatives. We offered that they could speak off the record. The offer was refused.
The Soviets did send a representative. I was extremely surprised. Usually, getting the Soviets to send somebody is like pulling teeth. This was just the opposite. They sent a General Officer and frankly, I have never seen a Soviet representative talk that freely at a conference.
CANDIS: Is that a reflection that he is a very senior man? Or because there's a new broom in the Soviet Union? Or maybe it's an indication that they're scared.
WALLACE: I think they are scared of this issue. They're worried about their own system or the Americans' system breaking down. In a crisis they'd face a situation with such great uncertainties that they'd have to risk everything to an American strike or else pre-empt. They're scared of being put into that corner and want to prevent it.
CANDIS: Reagan has declared SALT II dead. Is it in the American interest to abandon SALT?
WALLACE: One wouldn't have thought so. It's simple military calculation. You don't even have to love arms control to make this argument. The Soviets have scrupulously observed all the major ceilings on the numbers of missiles. Their violations involve peripheral things. The only real violation that they have made is the Krasnoyarsk radar station. I can't really blame them for that because it's an anti-aircraft radar and it's known that it's not supposed to be there. However, it plugs a huge gap in their air defences and it's hard to imagine with SALT or no SMI that they'd have let that gap sit there.
Now what will happen if both sides say, "Let's abandon the ceilings." What could they both do? The U.S. could build more Trident subs. But from laying down the keel to getting one of those things commissioned is over five years. It takes a long time and I don't know if they could speed it up very much. They could speed up the MX program but we're talking of a rate of production that is only 12 per year right now. It's going to take a long time before that makes a substantial dent in the number of missiles.
The Soviet Union has several very large heavy missiles already deployed-the 55 18 and SS 19 in particular. They also have models of both of these for their new Typhoon submarine. They can quickly add lots of these older, but very powerful sea-based missiles.
The major point of SALT that motivated even the hawks to support it was their desire to restrict Soviet heavy MIRV'd missiles. That was the alpha and omega of American interest in SALT. That's what SALT has done and that part of SALT has been successful.
Now they're saying, "Let's jettison that." Pull the limits off the Soviet heavy MIRV'd missiles which do threaten the Americans' deterrence. Their reason for doing this is to add a few cruise missiles or keep deployed a couple of old, utterly tattered Poseidon class boats from the mid-sixties. It doesn't make any sense militarily.
In addition, the Americans are "between a rock and a hard place" in terms of launch capacity now. Their program is in deep trouble. They picked a lousy time to challenge the Russians to a rocket-building race. They picked the worst possible time because the Soviets are going great guns.
CANDIS: What are your predictions about arms control agreements during the remainder of the Reagan presidency?
WALLACE: I think there's an ideological camp in that administration that rules out any arms control agreement with the Soviet Union-however peripheral. So I don't see a shift until 1988. I suppose all kinds of people like Steinbruner, Carter, Bracken, and Blair are all sitting in the wings, waiting for a more sensible administration in Washington. These are all very bright people-by no means all doves-who don't want an accidental nuclear war. Right now, the people in charge are totally opposed to any agreement. So I don't see what can be done until that changes.
CANDIS: Do you see any hopeful candidates for the next presidential election?
WALLACE: I've come to respect the Governor of New York, Mario Cuomo. He got up in front of a convention of police chiefs and argued against capital punishment. I figure that anyone with the guts to do that adheres to his own set of values, and I started off liking him right there. He's taken up the anti-nuclear cause strongly. I think he has the qualities of an exceptional president.