SUGG: --from Stansfield Turner, who has sort of given us a sense for the very real dangers of an accidental event with nuclear weapons. Talk about this sort of notion of launch on warning, hair trigger mentality. What is the mental state of the Russian nuclear forces, and the United States nuclear forces, as they sort of reflect this hair trigger notion?

BLAIR: Well both Russia and the United States adopted the concept of launching missile forces before they're destroyed on the ground by incoming missiles that's called launch on warning. It's the main option in the Russian war plan today, still, and it's a major option in the U.S. war plan. And to implement the idea of prompt launch, as it's sometimes called, Russia and the United States still operate their missile forces as though the Cold War never ended, and as though we may have to fire them at any moment, en masse.

So we have 2,500 or so strategic weapons on missiles on both sides, poised for immediate firing. The equivalent of some 100,000 Hiroshima bombs that could be lobbed around the planet in a matter of a few minutes, if the order was given right now, today, from the Pentagon, or its counterpart in Moscow.

The mentality is that although our political relations have improved, both sides are keeping their powder dry, we're hedging against some reversal in our relations that might lead to a nuclear confrontation. That's the official logic and rationale for maintaining nuclear weapons on hair trigger alert.

So, a nuclear weapon on hair trigger alert today means that there are missiles in silos that will fire immediately if they receive a couple of short, coded computer instructions from the launch crews. The missiles are fully fueled, they're armed, they are ready to go; they just need a target command and a launch command, and they're out of the silos.

So literally, within less than one minute, missiles technically could be launched by either side, by the launch crews.

SUGG: It seems with a doctrine of launch on warning, that is going to put an important premium on the reliability of the warning itself. Is our warning system in tact, and is the Russian warning system up to the task of, you know, identifying and, describe the state of the Russian warning system, if that's what they, if that's what will determine when they're going to launch.

BLAIR: Well, launch on warning is inherently dangerous, regardless of the performance of warning systems on either side, because it requires that human beings assess an attack, brief the President, and the President reach a decision for retaliation in a matter of just a few minutes. That's an enormous amount of stress and strain on just the human component of the command and control and early warning network.

The commander in a Russian early warning systems, and of the U.S. warning system are allowed only about three minutes to determine whether their countries are under attack, from the time that the censors first report the attack.

It's an enormous amount of strain, even at that level, and it's compounded through the whole chain of command. So, even with hardware and computers working perfectly, it's an inherently dangerous posture for the two sides to maintain.

But compounding that problem is the deterioration of the Russian early warning network of satellites and ground-based radars that are increasingly decrepit. They're not being properly maintained, even the people who operate them are not as proficient, their morale is low, there are enormous problems, on the technical, physical side of Russian early warning, compounding these dangers.

And so we're looking on top of everything else, there's the potential for the Y2K, the millennium bug glitch to cause serious problems in the Russian early warning network. And the Russian government has acknowledged this, and they're scrambling to try to fix the problem. But they'll be lucky to fix Y2K by the end of the year.

So there are just a host of human and technical and organizational problems afflicting Russian early warning, and nuclear command and control, that means that this inherently dangerous posture of quick launch is becoming more dangerous every day.

And the solution to this problem, frankly, is not just to fix early warning, although it would help if the Russians send some officers over to Colorado to sit in our early warning center, to monitor missile launches through our eyes.

But that's not the solution to what is an increasingly dangerous posture on the Russian side, and that is to stand down the arsenals. Take them off alert, so that they can't be fired at a moment's notice.

SUGG: All right, let's reset and address that directly.

BLAIR: There's the Russian early warning system, to assess whether Russia is under attack in two to three minutes. And then, this option allows only two to three minutes for the leadership of Russia to decide whether to retaliate.

So it's an enormously compressed time frame, that is inherently dangerous. It's becoming more dangerous, because Russia's early warning network is deteriorating, and is more likely to produce false warning of an attack.

Compounded by the millennium bug, or Y2K problem, which could generate false indications of an attack. And the context for this is Russia's growing reliance on nuclear weapons, and on their early first use, because of the deterioration, decay of their non-nuclear, or conventional forces, which are very weak.

And Russia, therefore, relies more on nuclear weapons and on their quick use, and because their long range strategic nuclear weapons are also in a state of decay, Russia is in worse shape today than it has been in decades, in terms of its ability to maintain secure, invulnerable forces. They can't keep submarines at sea, because they're broken down. There may be only one or two, or even none at sea at any given time. They're all stuck in port, needing maintenance. The list goes on and on.

So, most of their, virtually all of their nuclear forces are sitting in vulnerable positions, and therefore, the Russian general staff, the highest level of the Russian military, worries that it could not ride out an attack, and that it must fire quickly, these vulnerable forces.

And on paper, the general staff could and does conclude that Russia is in a position to use or lose nuclear weapons today, and that it relies more than ever on using or losing nuclear weapons, at least more than at any time since the early 1960's.

SUGG: So this mix of Russia's new sense of vulnerability and concern, the deterioration of their command and control and their early warning systems, and the condition of the personnel that are in charge of all these things, given the troubles in Russia right now, they inspired you to begin looking at this concept of de-alerting. Talk about how de-alerting would address some of those things. What is it, basically, and how does it address some of these modern features of the U.S.-Russia nuclear dynamic?

BLAIR: Well, the current path of nuclear arms control doesn't address the problem of a failure of control over nuclear weapons. It only addresses the size of the arsenal. How many weapons are allowed on each side. And no matter how many weapons each side has, they continue to maintain a large fraction of those forces on hair trigger alert.

If we got down to ten weapons on each side, there probably would be four or five still on a hair trigger, ready to be fired at a moment's notice, and still susceptible to scenarios of accidental or unauthorized, inadvertent launch. Mistaken or unauthorized launch. Clearly, the old path of arms control doesn't address this problem of a disintegrating country with a failing nuclear command and control with early warning network.

So it's a safety problem, of the first order. And it won't be solved by just reducing the size of the arsenals, unless we got to zero, which is not in the cards for the immediate future, to say the least.

So, an alternative way of enhancing this operational security and safety of these weapons needs to be invented, and the obvious solution would be for our countries to stand down these arsenals, take them off alert, so that they physically couldn't be fired at a moment's notice.

We did this in 1991, with our bomber force. We took all of U.S. long range, heavy bombers off alert, on the order of President Bush, which meant that the bombers on strip alert, ready to take off in fifteen minutes, with bombs on board, that those bombs were taken off and put in storage.

There's no way that you could launch a bomber today, under normal circumstances, with a warhead on board, either to launch it accidentally or illicitly. And so, and it would take at least twelve hours before any bombs could be loaded back on to our bomber force. So we have a margin of twelve hours. Not a lot of time, but it's better than a margin of seconds or minutes under this hair trigger posture that our two countries maintain.

So the idea is to extend the time it takes to launch nuclear forces, particularly missiles, from the current period of a few minutes, to a period of hours, days, weeks, months, eventually even years. Which can be done. It's feasible, technically, to reconfigure the arsenals on both sides so that they could not be fired quickly. The obvious first step–

SUGG: First let's take a breath. I want to get right to specifically this question. You talk about-- recalling that our audience are people sort of familiar with working on their car and working in the kitchen. What are some methods of de-alerting that you see as viable, obtainable, you know, and with respect to say, you've talked about bombers, but with respect to, say, ballistic missiles and submarines?

BLAIR: Okay. Well, the biggest challenge is not bombers, but missiles, either missiles in silos or on submarines. And for missiles in silos, in the ground, on land in our countries, probably the best solution is simply to take the nuclear weapons off those missiles, and put the bombs in storage, so that it would take probably roughly a day for either country to re-mate, remarry the bomb to the missile. Which would be a substantial amount of time.

Remember that the coup of August of 1991 in Russia, in Moscow, lasted for only 24 hours. There was a chance during that coup that illegitimate people would seize the reins of nuclear control, and actually order the use of nuclear weapons. That coup lasted 24 hours.

Under my scheme of removing the bomb from the missile and putting the bomb in storage, even if the coup plotters ordered the missiles to be readied for launch, by the time the missiles had been readied for launch, the coup plotters would have been defeated.

In other words, you buy time by taking a warhead or a bomb off the delivery system, usually a missile, and you can buy hours, and you can buy days, weeks or months, depending on where you store the bomb, how ready the missile is to receive the bomb, etcetera.

So there's not going to be any quick launch of missiles on false warning. If the bomb's not on the missiles, if it takes a day to put the bomb back on the missiles, there can't be any unauthorized launch by demoralized, disaffected nuclear units out in Siberia who are finally fed up with their situation.

They couldn't possibly fire a missile without authority if it doesn't have the bomb on it. And against a whole range of scenarios, of unauthorized or accidental launch of missiles, we have bought a very substantial margin of safety that currently is razor-thin; it's measured in minutes and seconds.

So, take the bombs off the missiles on land. There are other steps that could be taken, too, for missiles on land. In 1991, the first thing we did when President Bush de-alerted 500 missiles in silos in the ground in the Western United States was to take the launch keys away from the launch crews. And then he had maintenance teams go inside of every silo, each of which held a missile, and put pins in that would block the launch circuit, so that no one could get an electrical launch command to any of those missiles, from any source.

So it would take maintenance crews many hours to go back inside just a single silo, to remove those pins, to put that missile back on alert.

You could remove other vital components from missiles on land, like the guidance sets, the brains of the missile used in flight to guide the missile to its target. That could actually be removed and even stored inside the silo.

The Russians have proposed two other solutions. They're a little reluctant to take warheads off missiles, in part because they don't have a lot of storage space for a lot of warheads. And we're talking about thousands of warheads here. They've proposed that they remove the batteries from the top end of the missile, that provide power to the guidance systems on board the missile that guide the missile to its target, so if that battery is removed, the missile can't fly.

They've also proposed that they, and we, disable the launch tube with the cover on it, in such a way that the cover can't come up. So that a missile, if it were launched, would just crash against the cover of the silo.

And there are ways of disabling the missile silo so that the cover could not be blown off or raised quickly.

SUGG: It sounds like you and your colleagues have thought through some very good ways to buy time to assess a possible miscalculation, or an event. Have you also thought about how we would verify each other's adherence to a de-alerting kind of an environment?

BLAIR: Of course. Verification is an important aspect of the de-alerting proposal, although it is more or less important, depending on the step that's taken, depending on how far we go to stand down these arsenals. Clearly, if we stood down, took off alert all of our weapons, all of our missiles, then verification is more important than if we only took off alert, say, ten percent of our weapons.

Of course, I advocate zero alert, going to zero, because even one weapon on hair trigger alert poses a danger, a safety hazard of the first order.

So, as we look at deep de-alerting, we have to take verification seriously, and there are a range of ways of doing that. For submarines, for example, I would simply reduce the number of boats at sea, and you can readily monitor that fact from space, by looking down and counting the number of submarines in port. You know that whatever you haven't found in port is at sea.

There are other measures involving submarines, where I would put them out of range of their targets, so that it would take, say, days for the submarines to get back within launch range.

And that relocation of submarines could be verified if the submarine, say, surfaced every couple of days, and reported its location, (inaudible), through coded signals, to monitors who could determine that, indeed, that submarine is located no longer in the North Atlantic, ready to hit Russia in ten or fifteen minutes, but say, in the Southern Hemisphere, below the Equator, where it would need to transit for days to get back within launch range.

Well, that could be monitored. I would take, for submarines, I would take, instead of taking warheads off right away under this proposal. I would take other vital parts of the missile out of the submarine, and actually store it on board the submarine, and put seals over the holes where this component goes back in, so that certainly at the end of a patrol, which normally lasts two or three months, for a U.S. submarine, a Russian inspector could go look at the missiles and say indeed, these seals, special seals are in tact, which means that that missile never had the vital components on it during its patrol at sea. Or these seals on the missiles could periodically, through special codes, send signals through buoys that are released from the submarine, to satellites, to monitors that would verify the fact that indeed, that submarine missile has not been armed, and cannot be quickly fired. So some of the verification means are very simple. Looking down from space by satellite, counting the number of submarines in port.

Others are more technically sophisticated, such as the use of high tech seals on systems that would report through high tech communications links the status of that weapon. There are just a wide range of possible methods of verification. A lot of on site inspection would be involved in some cases.

SUGG: Speak directly to the sort of scenario I have in my mind's eye, which is a missile field somewhere in Russia, where presumably warheads have been de-alerted. How would one verify that kind of a situation?

BLAIR: Well, if warheads have been taken off missiles in Russia, then, we are already allowed a fairly substantial number of on-site inspections.

Under current arms control treaties, to go in, to check how many warheads are on missiles, under the current treaties. Well we can use these same kinds of inspections to go verify that in fact there are no warheads on the missiles.

And understand that taking warheads off and putting them back on missiles involves a lot of heavy equipment, it is a big operation that takes a lot of time, in Russia and in the United States, and much of that activity can be observed from space, which, coupled with the challenge, the ability to demand an immediate inspection of some suspicious location, coupled with other means of cooperation, we should be in good shape on verification.

Not all of the problems of verification have been solved, but if, I'm convinced that if, not just myself and a handful of other people working on this proposal, but the U.S. government, the Russian government would put their minds to it very seriously, that they could readily invent a very adequate means of verification.

SUGG: As a last sort of summary thing, can you walk us through, you've talked about sort of de-alerting, partial de-alerting, etcetera. Can you walk us through, say, the next ten years, where de-alerting is being instituted as part of our arms control policy? And talk about the sort of step-wise set of things that would get us to a truly safe world. And perhaps at some point, you know, Year Two, or Month Six, introduce this notion of outright de-coupling. Is that something you could do?

BLAIR: I could try.

SUGG: Okay. I'm sure you can do it. I just wondered, the de-coupling doctrine is something that you're comfortable sort of expanding into, or do you want to just--

BLAIR: The de-coupling doctrine, referring to--?

SUGG: I mean that notion of, not just sort of standing down, but really just giving these things, the warheads, totally apart, you know, if you want.

BLAIR: Totally apart from the missiles?

SUGG: Missile system. You don't have to; I'm really just manifesting my own sort of superficial knowledge of--

BLAIR: No--

SUGG: I'm looking at, what does a program look like for the next ten years, just taking us through the steps and the evolution, you know, to get to some place out there where perhaps there may be no nuclear danger.

BLAIR: Well, there are two programs (tape cuts out, then comes back in) - just to finish - I just want to say that the inertia of the Cold War is still very strong, and for example, from 1995 to today, the United States increased the number of targets, Russian targets in its strategic war plan by 20 percent, from approximately 2,500 targets, to 3,000. So that gives you an idea of how kind of cut off our military establishment is from the idea of de-alerting. They're still operating with a Cold War mentality. They still strongly believe that deterrence is the key problem in our relations with Russia. I don't. I think that problem is easily satisfied, at far lower levels of weapons, on far lower levels of alert. I think safety is the key problem, and the Pentagon is not ready to embrace that idea. Once they do understand that it's safety that's the threat, the hazard that Russia poses an operational safety problem, much more than they pose a problem of intentional, deliberate attack, then I think our military will come around to de-alerting more energetically than they have to date.

SUGG:Okay. That was very good. Talk, if you're surfing the web and doing some research on this, opponents of de-alerting say, well what happens when you get in a situation where, you know, you do have a serious military threat, and it takes too long to reconstitute your forces?

BLAIR: Well how much time do you need today, to reconstitute forces? A, do you need to be able to reconstitute them in hours, minutes, months? It depends on how quickly the other side would re-alert its forces. And if they have been thoroughly de-alerted, the configuration of their forces is such that they take a lot of time to put them back into a launch-ready configuration. Let's say that that's weeks. Then, we would need to be able to re-alert in minutes, or hours. Days would be ample. But what most of the critics overlook is that, even weapons that are not alert, and can't be fired quickly can still be secure and invulnerable. For example, submarines at sea. They may be days out of launch, away from their launch position, but they're still invulnerable. And does it really matter if it would take those submarines, say, two days or even a week to get back in launch position? To fire in retaliation to an attack that was launched earlier by an adversary that was able to re-alert quicker? I think that the threat of retaliation, even if it is implemented hours, days, or even weeks after an attack, that that threat provides sufficient deterrence. Many in the U.S. military believe that for deterrence to work, you have to be able to launch an attack in retaliation immediately upon indications of an incoming attack, that if we waited for hours, then deterrence fails. I mean, the opponent wouldn't be deterred any longer, if they knew that they were going to get hit, hours or days later, instead of minutes later. I think that's not valid. I think that's nonsense. I think that any rational state will be deterred from launching an attack, if it knows that retaliation is virtually certain. Even if that retaliation occurred days or weeks later. To boil it down, we still would want to have secure forces that could not be suddenly destroyed by an adversary that re-alerted faster, or launched a sneak attack because it cheated on the agreement, and was able to defeat our means of verification. We would still want to have a secure, core, invulnerable force that could be reconstituted, even though it might take a substantial period of time, and could be then launch-ready, at some point.

SUGG: Does de-alerting fit neatly into existing arms control tracks, Start II, III, etcetera? I mean, is it something that could be grafted on to this traditional preoccupation with levels, and delivery systems?

BLAIR: Yeah, de-alerting goes hand in hand with traditional arms reductions. When weapons are reduced, and banned by traditional arms control agreements, one could de-alert them and deactivate them as the first stage in their elimination. And in fact that's written into some of the agreements. De-alerting is part of the current track. But unfortunately, those agreements are to be carried out over the next ten years and longer. We're looking at decades, for implementation of Start II, and Start III. And I'm looking at a problem here that's immediate; a danger that's serious, immediate, and growing, and we don't have the luxury of grafting de-alerting on to the sluggish process of reducing the size of the arsenals. So we need to have a fast track on de-alerting to address this immediate problem.