The debate over the feasibility of national missile defense has been likened to a theological controversy -- both sides arguing with an almost religious fervor. Skeptics accuse missile-defense proponents of a blind faith in America's technological prowess, and point to the many inherent scientific and technical obstacles to building an effective defense against long-range ballistic missiles. The proponents, in turn, accuse skeptics of a kind of scientific arrogance, even ideological short-sightedness. They argue that progress has been slow but steady, that they see no technological challenges that cannot be met, and that the horrific prospect of a missile attack on a U.S. city, unlikely as it may be, makes missile defense a moral necessity.

In "Missile Wars," FRONTLINE interviews several of the top scientific, military, and policy experts on both sides of the debate. Below are extended excerpts from our interviews with Philip Coyle, former director of the Pentagon's office of operational test and evaluation; Richard Garwin, a physicist who helped to build America's hydrogen bomb; Lt. Gen. Ronald Kadish, director of the Pentagon's Missile Defense Agency; Richard Perle, chairman of the Defense Policy Board; Steven Weinberg, a Nobel Prize-winning physicist and science professor at the University of Texas; and Paul Wolfowitz, the deputy secretary of defense in President George W. Bush's administration.

+ Countermeasures: Decoys and "Bomblets"
+ Boost Phase: Pros and Cons
+ Terminal Phase: Pros and Cons
+ The "Hit to Kill" Tests
+ Faith or Reason?

Assistant secretary of defense and director of operational test and evaluation at the Pentagon from 1994 to 2001, he was a consultant to FRONTLINE on "Missile Wars."

Let's talk about midcourse and the problems of the decoys. What is the description of midcourse?

The midcourse of an intercontinental ballistic missile is the period of flight after the missiles have gotten outside the atmosphere on the way up and they haven't yet gone back into the atmosphere on the way down. It lasts maybe 20-30 minutes. It's the time when the enemy missile [and] all the debris that came along with it, from the rocket stages pushing it up there and any decoys or countermeasures that they might have put in there, chaff, things like that ... is spreading out and is floating along together.

Out in space, a feather moves just as fast as a lead brick, and it's hard for us to think about that, but it does. So you can have a balloon that looks just like a re-entry vehicle floating along with a re-entry vehicle which is much heavier, much more dense, will have different aerodynamic properties when it finally gets close to the United States. But out in space, it's very hard to tell them apart.

What kind of problem does that create?

The problem with decoys and countermeasures is that your sensors, radar sensors on the ground and the infrared sensors on the interceptors, have to be able to tell these objects apart. They have to be able to tell the warhead ... from all the other debris, things that are blinking and moving around, and which may look very similar to that re-entry vehicle.

I talked to someone yesterday, a non-scientist, who said there was no law of physics of which he was aware that would prevent the U.S. from solving this problem of discriminating against decoys.

Well, the way these infrared sensors work is they are picking out a cold object -- namely, the enemy's re-entry vehicle. Out in space, it's very cold. ... The background of space is also very cold. So one [countermeasure] scheme that is very challenging ... [is if a decoy] is basically refrigerated and shrouded, so that it looks to be the same temperature as the real re-entry vehicle.

Now indeed this would be difficult to build. But anybody who's smart enough to build an ICBM with all the guidance systems and everything else could probably do that also. ...

Chairman of the Defense Policy Board, an influential group of advisers to the Pentagon, he served as assistant secretary of defense in the Reagan administration from 1981 to 1987.

What about the problem of discriminating decoys, which many scientists, physicists, even Nobel Prize winners, say is just impossible?

Of course it's not impossible. By definition, decoys have properties different from real warheads. If they're the same size, weight and shape, then they really don't add very much. You might as well just tap in another warhead. So to be effective, decoys have to multiply the number of objects that would have to be detected by the defensive system, which means they have to be much lighter. Frequently things like balloons are contemplated for this purpose, because they are in fact much lighter and you can have large numbers of them. ...

Technology permits you to make the decoys look and behave more and more like warheads. Behavior is important, because a balloon re-entering the atmosphere does not behave the same as a warhead with a blade of material re-entering. So if you wait long enough, the decoys will all disappear. The question is, how much time do you need for your interception? But there's a cat-and-mouse game between making the decoys look more and more like real warheads, and developing the sensing technologies that can distinguish between real warheads and things that are made to look like them.

I'm betting on the side of the sensing technology, because it's harnessed to computing technology. The ability to very rapidly take a lot of measures and integrate them with high-speed, real-time computing is going to make it very difficult to produce effective decoys. And it's going to be especially difficult for primitive first-generation missile systems to incorporate very sophisticated decoys.

Sometimes I'm amused that the same scientists who say it's going to be very difficult for the United States, with all of its enormous technological capacity, to build an effective defense because the North Koreans, who are suffering malnutrition in the millions, are going to be able to develop brilliant decoys -- technologically sophisticated, beautifully manufactured, high reliability decoys. There's something wrong somewhere. ...

Do you ever visit the home of the young couple who have just spent every penny to build the house and they're sitting on packing crates because they couldn't do everything -- they couldn't build the house and furnish it at the same time? It's tremendously demanding to have sophisticated decoys. I'm not worried that we will lose the decoy business.

Winner of the Nobel Prize in physics in 1979, he is a professor of science at the University of Texas at Austin.

We've been trying to develop methods of discriminating decoys in warheads for many years. I worked on that in the mid-1960s. It was hard then, and it's just as hard now -- in some ways harder, because the planned defense system of the administration would intercept missiles above the earth's atmosphere, where the atmosphere doesn't provide any screening that separates decoys from missiles. ...

In the plan that's at the center of the [Clinton and Bush] missile defense program [ground-based midcourse defense], we would send up rockets that would strike the warhead while it's still coasting above the earth's atmosphere. The problem is that when the warhead is deployed by the rocket coming up out of the earth's atmosphere from our adversary's launch site, it can be accompanied by other things. It can be accompanied, for example, by balloons that reflect radar, that are visible to infrared sensors onboard the anti-missile missile. ...

Galileo pointed out that the rate at which a body falls doesn't depend on its mass. That isn't really true in the earth's atmosphere, as you can easily see: if you drop a feather and a bullet at the same time, the bullet will strike the ground faster, because the feather is supported by the earth's atmosphere. But above the atmosphere, it is true. A feather and a bullet will travel at exactly the same speed. It's also true of a warhead and a decoy. So their trajectory gives you no indication of which one is the dangerous one that has a nuclear weapon in it, and which is [a decoy].

There are ways of trying to distinguish them. They might look differently to a telescope or an infrared sensor. But it's not hard at all to make the decoy look just like a warhead. Or you can make the warhead look like a decoy by putting it inside the same kind of balloon that you're using as a decoy.

We won't know in advance. We can't know in advance just what will be the strategy adopted by the attacker. So I don't see how we can defend against it. ...

Someone [I spoke to] was skeptical that a rogue state, for example, could develop a sophisticated decoy system.

... I would say it's a lot easier to develop a decoy system than to develop the intercontinental ballistic missile itself. I would think that any country that could develop the missile could develop quite a decoy system. It doesn't have to be terribly sophisticated.

It would have to be terribly sophisticated, I think, if we were trying to intercept these missiles near the end of their flight, when they re-enter the atmosphere. There it's harder to make a decoy, because lightweight decoys like balloons simply don't travel the same way that warheads do once they're in the atmosphere. But above the earth's atmosphere, it's not that hard.

[What are "bomblets"?]

There is another way of defeating an anti-missile defense that was suggested, for example, or at least discussed in the Rumsfeld report a few years ago. That is, instead of carrying a single nuclear warhead or perhaps two or three nuclear warheads, the rocket would launch hundreds of little biological warfare bomblets each containing, for example, anthrax spores.

Because there are hundreds of them, they wouldn't need to have decoys. There would be far too many targets for any missile system to defend us against. Also, because there are large numbers of them, these could spread the biological warfare agents over a large urban area. It's been estimated that that sort of attack striking the city could kill perhaps 100,000 people, where a nuclear weapon would kill 50,000. And the kind of missile defense that's being contemplated in the Clinton/Bush national missile defense program would have no effectiveness against that at all. ...

How would we defend against bomblets of anthrax?

I'm afraid that it's not possible to design a defense against every conceivable threat that you can think of. ... Von Clausewitz said, "He who tries to defend himself against everything defends himself against nothing." ...

A physicist and the director of science and technology at the Council on Foreign Relations.

I should mention the other penetration aid or countermeasure, which is rarely addressed by the missile defense fans, and that is we fear greatly a biological warfare attack on the United States. ... So if you fear the [biological warfare] payloads such as Iraq had loaded into its short-range missiles, [and] that could be used by North Korea or one of the other proliferant states, then you ought to ask, how are they going to do it?

Well, it's peculiarly effective for them to send over, not a one-ton or 500-pound bunch of anthrax germs. ... [It would be] much more effective in killing people if they would break this up into 100 or several hundred little bomblets, each of which would have its own heat shield and doesn't get warm on the inside as it comes through the atmosphere and it explosively disseminates the anthrax. Unfortunately, the United States perfected these things and published them in the 1960s, not as ballistic missile re-entry vehicles, but simply as bomblets which would be dropped from airplanes. But you put one of these things in one of these heat shields and you have a very effective counterpopulation weapon.

Now what does this do? As the missile gets up to speed and comes through the atmosphere, as soon as the propulsion stops, these 100 bomblets are liberated -- just kicked out with a few feet-per-second speed, not a throw, just a gentle push. Twenty minutes later, by the time they re-enter the atmosphere, they cover a whole city much more effectively than a single so-called unitary warhead -- a clump of anthrax, hundreds of pounds landing in one space, one place. These bomblets come to the ground. They explode, tried and true, and there you are at nose level, the bugs that you want to disseminate. ...

The midcourse or any terminal ballistic missile defense system is totally impotent against these bomblets. ...

Director of the Pentagon's Missile Defense Agency since June 1999.

If you're beaten by countermeasures, you won't be beaten all the time in the process. It's a war-fighting kind of mentality. ... I think most of us would say, if we got a shot, we ought to take it. Hopefully it will be effective enough, and we're going to work as much as we can, as hard as we can, with as best talent as we have in this country, to make sure we solve the problem and make it effective.

In the end, it's a tough problem. ... But we're making good progress. We've got a long road ahead, but there's nothing right now that I see that says we should stop because it's too hard to do. ...

There's no military system that's perfect, and there will be no defensive system that's absolutely perfect that isn't susceptible to some kind of countermeasure. ...

Our response to this countermeasure issue is multifaceted in a way. ... What we're trying to do with our system is shoot it in the boost phase, shoot it in the midcourse phase, and shoot it, if we can, in the terminal phase. [It's] a layered defense with multiple shots. The system we're talking about now, and we're testing rather vigorously, is this midcourse phase that's susceptible to countermeasures. ...

We're going to do the best we can to overcome countermeasures in the midcourse phase, using the system that we have. But then we're going to add boost phase and other measures to handle and complicate any adversary's countermeasure issue. It's a problem we worry about all the time. It is not something that I think is overwhelming to the system at this point. ...

Winner of the Nobel Prize in physics in 1979, he is a professor of science at the University of Texas at Austin.

There is a possible missile defense system that doesn't have most of the disadvantages of the midcourse interception system that's at the heart of the Clinton and Bush missile defense programs. It's called boost-phase intercept. ...

When a rocket rises above the atmosphere but it's still firing its rockets -- and before it has a chance to deploy warheads or biological warfare bomblets or decoys or anything, it's still just one rocket boosting up -- if you can attack it at that moment, then decoys are irrelevant. ... It doesn't have a chance to deploy hundreds of bomblets which we can't intercept. So it's technologically very attractive to be able to hit a rocket at the end of its boost phase. ...

But it also has a technological disadvantage, because the boost phase only lasts a short period. In order to be able to hit the rocket during that short time, you have to be close to it. You have to be within a few hundred miles if you're going to hit it with another rocket of your own, or within a few hundred miles if you're going to try to zap it with a laser beam from an airplane, for example.

So a boost-phase intercept system will only defend you against attacks from a specific adversary. For example we could defend ourselves against attacks from North Korea by having ship-borne boost phase intercept systems in the Sea of Japan. ...

But the point is that the defense would only work against that country toward which it's aimed. A system based on the Sea of Japan would have no capability against missiles launched from the heart of Russia or China. That technological disadvantage is a political and a strategic advantage, because the boost-phase intercept system does not threaten the deterrent capability of countries like Russia and China, who we are trying to convince our missile defense is not aimed at. The best way to convince them that our system is not intended to work against them is to make it incapable of working against them. That's what the boost-phase intercept system would do. ...

We don't know whether it can be developed. ... [But] it seems to me if certain technical problems could be solved, [it's] the one kind of missile defense system that might make sense at some future time. ...

This technology doesn't exist yet? ...

Certainly the laser boost-phase intercept system does not yet exist, even on the drawing boards, as something that we could imagine building. I think research should continue on this. It's premature to imagine testing any such system. ...

Assistant secretary of defense and director of operational test and evaluation at the Pentagon from 1994 to 2001.

Some people think that the missile defenses we are building will work against a large country like China or Russia in the boost phase -- that when the enemy missile is still ascending, still boosting up into outer space, we can shoot it down during that period.

The advantage of shooting it down during that period is that you blow up the whole missile, decoys and everything. So you don't have to worry about the decoys later on, as you would in the midcourse. However, to do that, to shoot down an enemy missile in the boost phase, you've got to get close and you've got to be very fast. You don't want to be in a tail chase, because you probably can't catch up. So it takes very fast missiles, and you have to be close.

Well, for a country the size of China, the country is simply too big. You just can't get close enough without invading China first. ... Same would be true for a country like Russia. So there are people in Congress and people, I believe, in the administration, who think that the money they're spending, among other things, will produce a missile defense against Russia or China in the boost phase. And it won't. ...

The same is true for countries that are not quite as big as China or Russia. For example, Iraq is a pretty large country. So is Iran. And there are geopolitical problems in placing interceptors close to those countries, anyway.

Chairman of the Defense Policy Board, an influential group of advisers to the Pentagon, he served as assistant secretary of defense in the Reagan administration from 1981 to 1987.

It may well be that the best way to solve [the decoy problem] is to destroy a missile that has decoys as well as warheads on it before it is able to separate the two ... after it's been launched, in the boost phase. In principle, a relatively slow-moving missile in the ascent, with a thermal signature that is easy to detect, is the most vulnerable ballistic-missile system.

How long does the boost phase last?

It varies. One of the countermeasures against a boost-phase system is something that's known as a quick-burn system, in which you minimize the amount of time that that very hot plume -- thousands of degrees -- is visible, because we have sensors that will pinpoint it. But it lasts long enough for us to believe we can track the missile with great accuracy.

Some people have expressed concern that the boost phase would not allow enough time ... for the president or the National Security Council to be consulted.

I think it depends on the scenario. If it's an event completely out of the blue -- no crisis, no confrontation, suddenly there is an object in space that we detect and identify as a ballistic missile headed toward us -- would there be time for the National Security Council to meet, for the president to listen to what his advisors had to say, to prepare a position paper? Of course not. But neither would there be any enormous risk involved in destroying that missile. If it turned out to be a dreadful accident, and the missile that was destroyed was in fact not intended as a weapon against the United States, I'm sure we could cover the compensation and apologize. ...

Assistant secretary of defense and director of operational test and evaluation at the Pentagon from 1994 to 2001.

One of the things that's very difficult to do is to protect the United States in the so-called terminal phase, when the enemy missile has made it all the way across the oceans and is now re-entering the atmosphere over the United States. That's a very difficult job.

The advantage of trying to shoot down a missile in the terminal phase is that you use the atmosphere to strip out the decoys. In effect, they sort of float away from the re-entry vehicle and so you don't have to worry about the decoys. But you've only got maybe 30 seconds ... during which you can take advantage of that effect ... and the defended area of terminal interceptors is very small. So you would need hundreds of them all across the United States. ...

Speaker of the House of Representatives from 1995 to 1999, he is a member of the Defense Policy Board, an influential group of advisers to the Pentagon.

What about terminal phase?

It's the hardest problem. It's the least desirable. ... Again, there are two ways to measure cost. One is against alternative budget, and the other is against human casualties and the cost of recovering a city. If you could develop a reasonable terminal phase, you might do it. You certainly want to do it at the theater level for troops and ... for airfields, for ports. But I would personally like to never see missiles get that close. I mean, I think that's really dangerous. I'd much rather kill them as far away -- kill them over there -- than have them kill us over here. The farther over there I can get them, the safer I feel.

A physicist and the director of science and technology at the Council on Foreign Relations.

Discuss the technical challenges in the terminal phase. ...

Some potential adversaries have been testing GPS-guided ballistic missiles. So it would come in ... within feet of their intended target, and those would be very effective. But those can be countered by a terminal interceptor -- by a truly short-range terminal interceptor ... which would go out maybe only one kilometer, and a radar which would wait and really do nothing until it saw warheads attacking the point that you wanted to defend. ...

Now the problem with defending the entire U.S. territory or our cities against nuclear-armed ICBMs, or ICBMs armed with biological weapons in the form of bomblets, [is that] the bomblets cannot be defended against at all, because there would be hundreds of them in a payload. You would need hundreds of interceptors to go after them individually. ...

Now we worry about North Korea sending one or two or four nuclear weapons against the United States, not thousands which will destroy the whole country. So it makes very little difference to North Korea if Washington and New York are defended by terminal missile defense systems. They could pick St. Louis or San Francisco or Los Angeles or Chicago, and they would be just as effective in destroying American people as if we were totally undefended.

So you have to defend, in a terminal defense, the entire country. It's far easier, instead of defending the entire nation, to [intercept ICBMs from] North Korea, which is the size of the state of Mississippi ... before they reached a speed that would carry them to the United States. ...

Winner of the Nobel Prize in physics in 1979, he is a professor of science at the University of Texas at Austin.

These aren't realistic operational tests. You know, when the Army wants to test its strategic doctrines, wants to test its equipment, wants to test the abilities of its officers, they have maneuvers. In the maneuvers, the blue team is not told in advance exactly what the red team is going to do. The tests so far have been like a maneuver in which the blue team is told exactly what the red team's plans are. That's not a serious test of a system. ...

Director of the Pentagon's Missile Defense Agency since June 1999.

How did you feel in July 2000 when that test failed right before the Clinton administration's deployment decision?

... Well, I guess we ought to review where we were. We had the first test that was successful ... in [October] 1999. Then we had a test ... that following January, and it failed. We had a glitch in the cooling system, as I recall. We fixed that, and we came up to the next test [in July 2000] that, at the time, became the linchpin in the deployment decision for what we were calling at the time the National Missile Defense System. We had an unfortunate event where the kill vehicle never [separated from] the rocket it was on. ... Then we went on to actually fix those issues and have had three successive intercepts in the program [since then]. So we are 4 for 6.

I don't like to have a situation where we are not able to accomplish our objectives. So anything in a test program that prevents us from accomplishing our objectives is something that we try to avoid if at all possible. But this business is about success and failure. I know this sounds trite to a lot of people, but we learn as much from our successes as we do from our failures. When we find something wrong and fix it, we have pretty good confidence it won't happen again, or at least we try to minimize that chance. ...

Assistant secretary of defense and director of operational test and evaluation at the Pentagon from 1994 to 2001.

Before you went to [the] Kwajalein [test site, in July 2000], you wrote a memo within the Pentagon to the director of acquisitions ... raising some of your concerns about ... these tests. ...

Well, in all early developmental tests, there will be artificialities. There will be, as the testers say, "test limitations." There's nothing wrong with that. It's almost unavoidable early in a developmental test program. What was different this time [July 2000] was that these tests were leading to a deployment decision, perhaps, by the president, years before the program would ever be ready for realistic operational tests such as most military equipment has to go through. So it was important, I felt, that people understand the limitations of these early tests in the context of a deployment decision. ...

You were worried ... [that] some people in the Congress and the Pentagon didn't understand what was going on with the tests?

Yes. These are very complex tests. They cost close to $100 million apiece. There's all kinds of equipment, some specifically just for the purpose of the test ... [and] all of these complexities are hard for people to keep up with and understand. I was concerned about that. ...

Was there also a feeling, both within the program and within the Pentagon, that there was political pressure to make this happen, and make it happen quickly?

I think then, and today, there are people both in the Pentagon and in the Congress who would like to see a national missile defense system successfully developed, built and fielded. If we knew how to do it, especially if we knew how to do it at some reasonable cost, I think most Americans would feel the same way. So then as now, yes, there was pressure to move this program forward as fast as possible. ...

[Later, you] issued a report. What did that report say?

It was about a month later. The report that I and my staff wrote described all of the tests that had taken place so far, including the most recent two, both of which had failed. [We] described what was realistic and what was not realistic about those tests, all of the tests; described all of the things that the program still had to do, which were many; and basically concluded that, in our view, the program was not ready for deployment, and wouldn't be for many, many years.

Were you critical of the structure of the tests?

We found fault with some aspects of the tests. For example, the intercepts are conducted fairly close to Kwajalein, relatively speaking. In a real battle, a real war, you wouldn't wait until the enemy missile nearly had reached the United States. If you could, you'd reach out and get them way out in the distance.

So we felt that eventually there needed to be tests where intercepts were conducted at much greater ranges from Kwajalein than anything that had been attempted at that point or since. We also felt that there were other artificialities in the test, artificialities which were perfectly understandable for such early developmental tests as these were, but which were not acceptable if you were thinking about deployment. ...

Critics have charged that these tests have been fudged in some way. Is that true?

... I don't think anybody has ever cheated. But there are artificialities in these tests, and critics have pointed out that they certainly look like cheating. For example, the target, the surrogate enemy target, the re-entry vehicle has a sort of a beeper on it, a beacon, that tells you where it is. It also has a global positioning system on it. Obviously an enemy wouldn't tell you with a beeper, "Here I am."

In an early test, there's nothing wrong with something like that. It's a surrogate for a forward-based early warning radar that we didn't have in the tests at that point, and probably won't for many years. So for an early test, it was not a bad thing to have. But to a critic, it looks very artificial, and it's something that I know the program will seek to eliminate as the years go by. ...

Describe for me the difference between developmental and operational tests.

Developmental tests focus on ... achieving certain detailed technical specifications such as speed or maneuverability or something like that. Operational tests are tests that are very real-world. In operational tests, the equipment is operated by soldiers or sailors or airmen or Marines, not by contractors, as they might be in a developmental test. The tests may have to be conducted at all times of the day and night, in bad weather, in dust and dirt and mud and rain and snow and all of the things that can happen in battle. ...

So in an early developmental test, for example, you might have prior information about when the target's going to be launched, what the trajectory is, what the re-entry vehicle looks like that you're going after with your interceptor. You might have prior information about all those things that you might never have in a real battle.

Chairman of the Defense Policy Board, an influential group of advisers to the Pentagon, he served as assistant secretary of defense in the Reagan administration from 1981 to 1987.

There have been mixed test results. The tests themselves have been criticized.

There are always mixed test results early in a program. I don't know of any program that doesn't have its failures. The important thing about failures is that you want them to be as illuminating as possible. You want to learn from them. If you have a failure in which a component breaks and that's the source of the failure, that's very good news, because you can go back and fix that component. The failures that are unexplained can be much more troublesome. You may have to have several of those before you begin to see the source of the failure.

But it is perfectly normal to have failures. Here there's a problem with the quality of the reporting on the technology. A failure is unquestionably a story; the full implications of failure are not always understood. It's frequently very good news that you've unearthed a fragility in the system that can then be fixed. You'd be far worse off if that failure didn't occur, and you built an entire system, only to discover that the failure was in fact inherent and would show up with some regularity in a statistical sense.

So of course there'll be failures. Is there reason to believe that the obstacles that need to be overcome are insurmountable? That somehow we would have to defy the laws of physics to move an object fast enough or to capture enough light to pinpoint an image? I have seen no convincing argument that there are physical phenomena with which we can't cope. ...

Deputy secretary of defense in President George W. Bush's administration, he is the former dean of the School of Advanced International Studies at Johns Hopkins University.

[We] haven't heard people say that they're very confident that they can erect even a limited defense in [the five to 10-year] timeframe that would have a high probability of stopping a ballistic missile attack on the United States.

We're still in an experimental development phase of the technology. ... But look at it from the other end. We've made extraordinary progress. The Scud missile has a re-entry velocity, I believe, in the neighborhood of 2,000 to 3,000 miles per hour -- only two to three thousand miles per hour. That's hitting a bullet with a bullet. After 10 years of work, we can now hit a bullet if it's traveling only a mere 3,000 miles an hour.

What we're really aiming at [with midcourse defense] is hitting a bullet that's traveling between 10,000 and 15,000 miles per hour. But we've done that successfully already now in [four] tests. They're tests. They're somewhat artificial. The conditions are somewhat arranged. But we're making really quite amazing progress.

I'm not sure on what basis John Kennedy said, in 1961, that we could put a man on the moon in 10 years, but I must say, in a way, what we need to recapture is some of the pioneering spirit of America of 30, 40 years ago. We put ballistic missile submarines -- one of the marvels of modern military technology -- in the water within six years of a decision to do so. And, frankly, we didn't do it by deciding on day one that it could be done. We decided on day one we were going to do it.

I think we do the same thing here with missile defense. If this country puts its mind to it, we can succeed. ...

There were some famous German physicists, including some Nobel Prize-winners, who said it wasn't possible to build an atom bomb. I'm a little skeptical of people who aren't engineers, trying to solve a problem, saying it's impossible to solve a problem. History is just littered with problems that were solved that were supposed to be impossible.

Winner of the Nobel Prize in physics in 1979, he is a professor of science at the University of Texas at Austin.

... One often hears it asked, why, if the United States could put men on the moon, can't it develop a defense system that will defend our country against attacking missiles? On the face of it, it's a reasonable question. But there are pretty great differences between going to the moon and defending the country.

The moon is a fixed target. We know where it is. With absolute precision, we know where the moon will be at any future time. It's not going to evade us. If we send rockets to the moon, there will be no effort on the part of the moon to prevent them from reaching there. They will just go there. It's completely predictable. When we had developed the rockets that went to the moon, we knew they would get there.

Unfortunately with missile defense, we face human beings on the other side. Whatever we do to defend ourselves, they can take another step. They always have the last move, and [can] deploy decoys or other penetration aids that will defeat our defense. ... Dealing with a human adversary is considerably more tricky. ...

Someone has said to me that, in this whole debate and beyond, the technology is lagging behind the political fervor. Do you think that that's a fair statement?

There have been technological advances. It's really quite an achievement to be able to send a missile up that can hit an incoming missile in space above our atmosphere in midcourse. But the technology isn't there for a credible defense, for a defense that will be able to deal with an adversary that sends decoys along with the warhead. We're not remotely yet in the position of testing against that kind of credible offense. ...

I think the rocket technology is there, or will be there within a reasonable amount of time. But the organization of a defense including the development of sensors, computer capability that would deal with a reasonably competent attack, is just not there. I don't know that it ever will be, because as we improve our technology, the kind of thing the adversary might do can also be improved. ...

We won't know what the offense will be. We may know the nature of the nuclear warhead. But we will not know what other things are sent along with the warhead in the form of decoys that will exhaust our defenses. I don't see how that problem can ever be solved.

Speaker of the House of Representatives from 1995 to 1999, he is a member of the Defense Policy Board, an influential group of advisers to the Pentagon.

I would urge every skeptic to read a wonderful book called The Invention That Changed the World, which is the history of the radar labs at MIT. ... [The book] takes you from the initial discovery of radar through winning the Second World War. At any point in that time, a skeptic could have said, "I don't believe we'll be able to do the next thing." But the next time you use your microwave oven, which was originally called a radar range, OK, just remember not only did we invent it, we made it so commonplace that it warmed up your coffee.

I look at people who doubt our ability to create this, and I think, what century are you living in? For the last 250 years, humans have been increasingly good at inventing science and technology that accomplished things. Arthur C. Clarke once said, "If a famous scientist tells you something can be done, he's almost certainly right. If he tells you something can't be done, you don't have a clue," because the record of famous scientists being wrong is so constant.

Assistant secretary of defense and director of operational test and evaluation at the Pentagon from 1994 to 2001.

What do you say to people who essentially say, "American science and ingenuity has always accomplished what it wants to accomplish?"

Technology has been the show stopper for national missile defense for all of the 30 or 40 years that it's been on the American political scene. The particular technology that we've tried to use at each phase has changed over the years, but it's always been very difficult. ...

I think we'll be able to learn how to shoot down short-range missiles like Scuds overseas in places like the Persian Gulf. It's much more difficult when the missiles are long-range missiles and when they are launched with these huge rockets that can also send up decoys and countermeasures and chaff and everything else along with the real target. Much more difficult then. ...

Shooting down an ICBM with a hit-to-kill missile is like hitting a hole-in-one when the hole on the green is going 15,000 miles an hour. Doing it when there are decoys is like hitting a hole in one when the hole's going 15,000 miles an hour and the green is covered with a bunch of other holes that look just like the one you're supposed to hit. ...

The missile defense programs that the Department of Defense is pursuing today are the most difficult thing they've ever tried to do. ... Some people compare it with the Manhattan Project to develop the atom bomb, but in many ways that was much more focused and more tractable technically than parts of missile defense are. And of course, the Manhattan Project was carried out under the exigencies of a world war. ... [It] was about building a single thing, an atom bomb.

Missile defense is about defending, as Secretary Rumsfeld has said, all ranges of threats -- short range, medium range, long range, intercontinental ballistic missiles; defending the United States, our friends and allies all around the world in all kinds of battlefield situations that we might be able to imagine -- with land-based, sea-based, aircraft-based and space-based systems. So when you put all that together, it's a very diverse set of programs and very complicated. ...

Director of the Pentagon's Missile Defense Agency since June 1999.

We've talked to scientists, some of whom have worked on the problem of countermeasures for the Pentagon, who just flat out say that the problem [of decoys and countermeasures] is never going to be solved. ...

... If my memory of history serves me well, it was Lord Calvin, a very eminent scientist at the time, who said that man will never fly. So I respect the opinions of the people who say that because they worked on those types of issues. But over time, we have good ways of solving those types of problems. ...

I would not be so arrogant at this point in time to believe that this problem is insoluble or soluble for that matter. But it will be a give-and-take. It will be an approach that we're going to make a centerpiece of the program over time, that once you build it, you've got to make sure it's viable against all the things somebody's going to try to beat you with. It's no different than any other weapons system that we've ever built.

At this point, I would say the evidence is such that it is a tough problem to deal with countermeasures in the midcourse. That's why we're building a multi-layered approach. Shoot it in the boost phase, midcourse, potentially in the terminal phase, and you greatly complicate the adversary's problem. I believe that we have an approach to this problem that is viable. And with all due respect to the experts, we have our own experts. ...

home · the threat · the technology · the strategy · introduction · map
timeline · producer's chat · interviews · quiz: missile test · discussion
tapes & transcripts · press reaction · credits · privacy policy
FRONTLINE home · wgbh · pbsi

photograph © 1996 CORBIS; original image courtesy of NASA/CORBIS
web site copyright WGBH educational foundation