able man would ignore the possibility, but no one can say that this probability is a certainty.

Then he continues:

Between a risk and a certainty, a sane man does not hesitate.

Let me summarize the basis of C. P. Snow's argument.

If one believes that unless a particular arms control measure is adopted the world will automatically go into thermonuclear war, and that this thermonuclear war will result in an end of history, then one does not have to bother thinking any more. He can say it is quite clear that this arms control measure is better than the present policy, because it cannot be worse; therefore, it must be better.

However, one can also have the view that just because one is in serious trouble with the current policies, it does not mean that any other policy is a better one.

Let me again, with the permission of the committee, tell another story I often use to illustrate this point.

This is a fairly ancient joke today, and I would guess many here have heard it, but it illustrates the point better than any other anecdote I know. The story concerns a mouse that goes to work for Aerospace Corp. For the benefit of the other interests of this committee, I should mention that it is a very low-salaried mouse.

The mouse is describing his new line of work to his country cousin. It is an experimental mouse. It is shot out into outer space; it goes through the ionosphere and the Van Allen Belt. It is cosmic-rayed; it is weightless; it falls back into the earth's atmosphere; it falls into the water; they fish him out and take him back to the laboratory where they pull out his kidneys and look at them; pull out his entrails and look at them and put them back in and sew him up and start the whole process over again. In every way it is a miserable job. Well, this mouse is describing to his country cousin his new line of work. His country cousin is totally appalled and says, "Why don't you go back to your previous line of research?" And the mouse says, "Back to cancer research?" [Laughter.]

Well, this is a major point. If you can convince people that the current system is cancer research then it is clear that any other line of work is better.

I happen to think the current system is bad, but I think it is more like the Aerospace "job" than cancer research, and some of the suggestions that are made are more like cancer research than Aerospace.

I am not, in other words, an apologist for the current system. I do not think the current system is sensible. I do not think it will work indefinitely, but I do believe that it is useful to fix it up, to repair it, to make it safer and to hedge in various ways against its breakdown while we are looking for a better system or while a better system evolves on its own. This policy may not work, but I do not know of a better one. I am very conscious that there are some systems which are suggested which seem to me, to put it mildly, decidedly worse than the current system. Even a system that evolved as a result of a war or very intense crisis might be better than some of these alternatives. One of the major reasons that some people refuse to believe in the posibility of survival is that they want to convince themselves that nothing can be worse than the current situation and they can then stop thinking.

POSSIBLE ENEMY ATTACKS

I would like to close my testimony this morning with a discussion of the different kinds of wars that can occur. I believe that unless one understands there is a range of possible wars, a range of possible situations, one cannot fully appreciate the potential effectiveness of the different kinds of civil defense programs which might be recommended. This very crucial point has been ignored in most discussions. Let me start with the targeting objectives that an attacker might have. This next chart gives five alternative target systems that the attacker might aim for. (See fig. K-6.)

1

2

Countervalue.

FIGURE K-6.-FIVE POSSIBLE ATTACKS

Counterforce+countervalue.

4 Counterforce+bonus.

5 Counterforce+avoidance.

The first system is the so-called countervalue system. The attacker is mad at the defender and wants to destroy that which the defender values most, irrespective of whether it helps the attacker to achieve his positive objectives. Now, it is easy for most people to believe an attacker would want to do this. They visualize themselves as a defender and think of an attacker as being angry at them, and therefore, he is going to try to hurt them.

We value most, people and property, in that order. Therefore, if we believe that the attacker is out to hurt us we may study countervalue attacks and indeed such attacks could occur. We wish to study this kind of attack because it may happen.

The next kind of attack that could be studied might be called the counterforce plus the countervalue. In this attack the attacker distributes his attack between the things which might hurt him, the defender's strategic forces, and the things which the defender values, his cities, his people, and his property.

The next kind of attack might be called the modern attack. The attacker says to himself, "This other man's cities can't really hurt me. He is not going to manufacture anything in them that is important, he is not going to draft any soldiers, there is no morale problem especially since there are not going to be any elections between the time this war starts and the time it ends. Adding up all of this, why should I bother hitting cities?"

So the attacker then goes for the other side's strategic force. If he goes for them in a straightforward way, ignoring the cities completely we could call it a counterforce attack. This is a reasonable pattern of attack which might well occur.

The next attack might be called counterforce plus bonus. It is basically motivated by the same considerations as the counterforce attack, except that the attacker says to himself, "I would still like to destroy as much of the other side's people and property as possible. I will, therefore, change or compromise my target system. I will change my designated ground zeros, I will use somewhat different yields, I will ground-burst instead of air-burst, even though it may be militarily most effective to air-burst. I will do all this in order to increase the bonus damage." Such an attack would result in quite different casualties than the straightforward counterforce attack even though the primary objectives are much the same.

The last attack pattern on the chart is counterforce plus avoidance. In this attack, the attacker says to himself:

I am not mad at the enemy's civilians. There is no reason why I should kill them. I only want to protect myself. I want to destroy the other side's forces, but to the extent that I can, I will avoid his people and his property.

He might do this for two separate reasons. The first is sheer morality. Nobody wants to kill 100 million people without a good reason for doing so. One might not even want to kill 100 million people even if he thought he had very good reasons.

The second reason for the avoidance tactic is that the attacker can use the enemy's people and property as hostages. He can say to the other side, "I haven't killed your people. If you start killing my people I will kill your people in reprisal."

In other words, one can try to make deterrence work after the war has started.

CONTROLLED AND UNCONTROLLED WARS

Now, Americans, by and large, find this very hard to believe. It is difficult for them to believe that once a war starts they would be deterred from any action against the enemy by fear of further reprisal by the enemy. They have a feeling that the war must be all-out and uncontrolled.

This is, in two separate ways, a very naive point of view. It is naive because it is not sensible, and it is naive because it may not be true. Even if one tries to be uncontrolled, one may find himself being threatened so persuasively by the enemy that he will control himself at the last moment. I have several illustrative examples in my book of preand postattack coercion and I am told that I have made the possibility, under some circumstances quite persuasive.

One reason why we Americans and much of the West do not fully understand these possibilities is that we have been bemused by the examples of World War I and World War II. These were, indeed, two of the most all-out wars in history. In these wars there was very little attempt to negotiate during the war. The attitude was to destroy the enemy and then, having destroyed him, to dictate a peace. One did not negotiate during the course of the war unless one was either clearly victorious or clearly defeated. While these wars are prototypes for most people, actually they were rather extraordinary wars.

If one goes over the history of warfare, one finds very few periods in which the World War I and World War II ideas of how a war should be fought held sway. The more classical way a war is fought is to fight for some definite objectives, generally limited objectives. In particular, one does not try to destroy the other side, but to attain some desirable objective for oneself or to prevent the enemy from attaining some objective that he wishes. Countries tended to limit their actions, fighting, pressures, and reprisal to be in some sense consistent with the objectives they were trying to attain.

Modern technology is such that the ability to fight uncontrolled is greater than it has ever been in history.

However, even though it is easier today to fight uncontrolled wars, it is also more disastrous; the sanctions against fighting uncontrolled wars are much greater than has been true in the past.

We found this out in Korea. Before Korea was fought, very few Americans would believe that we would limit ourselves the way we limited ourselves in Korea.

After Korea we learned that just like anybody else we can be scared. We can be cautious; we can be responsible.

One way to phrase this is to observe that Americans are no tougher than, say, the Japanese or the Germans, and these people surrendered rather than fight to the last man. Similarly, we may be restrained by sufficiently large threats. This restraining can happen after the attack as well as before the attack.

I suspect that the main reason that Americans find it difficult to believe a war can be fought rationally or reasonably is that, by and large, in our country we do not give force any rational or reasonable role. We feel that only a violator, a criminal, a desperado, an insane man, or a sick person uses force, and, therefore, when we find somebody using force he is not only our enemy, he is an enemy of humanity, and should be exterminated or locked up or treated, not negotiated with. We then go all out in our attempt to destroy or control him.

This is, I am afraid, a somewhat naive view of the world. Force has been around for many years. It has been used by both good, bad, and indifferent people. It has been used rationally and reasonably as well as irrationally and unreasonably. In particular, it is perfectly possible for us or the Soviets to use force in a rational and reasonable fashion at least, in the sense, that we do not use it in a wildly irrational or wildly unreasonable fashion.

This is true even if it is in some sense unreasonable to settle disputes by the use of force. Having unreasonably decided to use force, one might still decide to use the force in a reasonable fashion.

To the extent that the other side tries to keep one's people as hostages, he can threaten even after the war has started. He can say, "If you destroy my cities I will destroy your cities, but for the time being I am not destroying your cities."

To the extent that he destroys your cities in his first attack, he cannot destroy these hostages in a later attack.

This is one of the ways in which civil defense can decrease deter

If one puts people in fallout shelters in the middle of cities, one has improved the enemy's capability to make them hostages. If they were not in the shelters, they would have been killed. To the extent that he can hope to limit your response by the use of such blackmail tactics, to that extent he may be more willing to go to war. In addition, just thinking through the civil defense problems is likely to make the decisionmakers of a country think of a war as an experience rather than as an end of history. This is the proper psychological preparation for postattack blackmail. Therefore, it is quite possible that the suggested civil defense program might actually decrease our deterrence against some kinds of attacks.

This is not an overwhelming argument against civil defense. On the one hand, the civil defense does not decrease our deterrence very much; while on the other, it increases markedly our ability to survive a war if a war occurs, and is fought by what might be called rational methods. And as I will discuss later, it may also increase our ability to deter other kinds of attacks or provocations.

Let me be specific about some of the kinds of wars that can occur. In addition to the differing target systems we or the enemy might

adopt, there are many different situations in which a war could start. I have listed eight situations on the next chart. All of them can be characterized as being lower priority situations than the ones normally studied. Yet all have the characteristic of being sufficiently important so that they must be considered. (See fig. K-7.)

FIGURE K-7.-SPECIAL COUNTERFORCE SITUATIONS

I once had a conversation with a senior military officer in which I was discussing one of these problems, and he said, "That strikes me as being sort of a fifth-priority mission." I said, "That is exactly right. It is fifth priority and, therefore, up at the top."

A fifth-priority problem can be a very important thing in international affairs. One of the real difficulties a country like the United States, or perhaps any country has, is paying sufficient attention to relatively low priority but absolutely important missions.

If we adopted the usual attitude toward second priority in our private affairs, we would say that the highest priority objective is milk for the baby, and let us do that and nothing else.

But we do not live that way. We live complicated lives, and there are many things we must take care of. The international situation today is such that an objective which is fifth priority can still, if inadequately prepared for, kill you.

I repeat, one can die as a result of a fifth-priority failure as well as the result of a first-priority failure. We cannot be simple minded and say, "Let us take care of the first-priority mission and nothing else," even if it is intellectually easy to have this attitude.

Let me now discuss some of these lower priority wars which can occur and in which civil defense may perform a lot better than would be expected if it were evaluated only in the worst case. These are still probable enough and important enough so that if we designed a particular civil defense program only for one of these situations and it was useless for all the others, one might still be able to justify the expenditures for such specialized civil defense.

CONTROLLED WARS

The first one, the controlled wars, I have already discussed. These are wars in which one side or the other attempts to use force in a rational and discriminating way.

This controlled war notion is directly opposed to what I have called the spasm war. In a spasm war each side is trying to get rid of all of its weapons as fast as it can in sort of an orgiastic spasm of destruction.

The controlled war includes withholding tactics, an adequate command and control capability, the use of intrawar deterrence, bargain