When we came to presenting the defense budget and program to the White House each year, which we did in November, uh, we dealt with the Bureau of the Budget, Dave Bell and his people, with the president's science advisor, Jerry Wiesner and his, his people, and the National Security Advisor, McGeorge Bundy and Carl Kaysen. So, um, we were dealing with all of them, and we would send over memoranda explaining the program, and then there'd be quite a few meetings where we'd follow up in detail and, and explain it. And they would scrutinize and analyze and, and challenge, and develop issues to bring to the president, exercising their staff responsibility to the president.

But it was Carl Kaysen who...

I recall Carl Kaysen as, as playing a prominent role in that. Carl uh, put in a lot of effort to become knowledgeable about the defense program, and was uh, a, astute about looking for holes in our case.

At some point you begain talking about, after Ann Arbor, about assured destruction. You didn't talk about counter-force or city avoidance anymore. What did assured destruction mean?

Well... the tendancy had been in the past to use what you might call, uh, Latin euphemisms for things. So in doing analyses about the outcomes of nuclear wars, people wrote and spoke about urban-industrial fatalities, which got abreviated, U.I.Fat. And uh, you know, it was a way of kind of detracting your gaze from the, from the horror of what you were talking about. McNamara uh, preferred to call a spade a spade. I think he, he had a preference for plain Anglo-Saxon language to get people to think about what they were talking about. And uh, in that context, as I recall, he was the one who introduced the term, "assured destruction," say this is really the capability we were trying to have, is a capability that we could assure that if they attacked us, we could destroy their society, as a deterrent. He did, in various documents, make the distinction between uh, assured destruction as a basis for calculating requirements, on the one hand, and a policy for how the weapons would be used if a war broke out. Those were two, uh, different things.

Was it a procurement criteria?

Yes. Assured destruction was a criterion for determining the size of our forces. We wanted to have enough forces that they could survive a Soviet attack, strike back, and destroy their society.

So the war plans didn't change when he started talking about...

No. No.

Did, as McNamara began talking about assured destruction or assured destruction and damage limitation, did this indicate a-- that he was moving away from efforts to refine the strategic arsenal? To consider options... was he leaning towards a deterrence only posture?

Of course, uh, Mr. McNamara ought to speak for himself on this.

My interpretation of what was going on was that the talk about selective use of nuclear weapons, of, of city-avoiding strategies and all that, uh, was proving to be counter-productive. That it was stirring up uh, responses and ideas and debates that were detracting attention from our main idea, which was to make nuclear war unlikely by having survivable retaliatory forces, and by having adequate conventional forces so that we would not have to be the first to use nuclear weapons. I want to emphasize, that was the, the concept of overriding importance. How you would fight a nuclear war, if one arose, I think McNamara increasingly came to feel, uh, was something to which nobody had any sensible or rational answer. And so it was not something worth uh, pursuing in great detail. But when he talked about city-avoiding strategies and all that, uh, it evoked at least a couple of counter-productive reactions. One was, some people started saying, "Yes, sir, that's right! You can fight and win a nuclear war, therefore, for example, we don't have to build up these conventional forces." And we just felt that was terribly wrong. Another group were saying, "That is uh, very destabilizing if you're talking about uh, confining your attacks to Soviet forces, then that looks to us like a first-strike strategy, uh, and that's going to destabilize the balance of terror, and that's wrong." And that school of thought, sometimes known as the minimum-deterrence school, argued in effect, we should limit and confine our uh, uh, attack on the Soviet Union to their cities. My reaction to that is uh, like Albert Wohlstetter's, who said once, "Not even Genghis Khan deliberately avoided attacks on enemy military forces in order that he could save the fury of his attack to attack innocent civilian populations." Uh, that just seemed, uh, wrong also. In fact, I think that a ... uh, if in retaliation we're going to strike back at what the uh, Soviet government values, we would appropriately strike back at their military forces, which I think they probably see as much more important to them than their own people, who they have been willing to sacrifice at times, uh, in the past. But, uh, so all of this talk about, about um, the ways you might fight a nuclear war, while leading nowhere in terms of some deeper insight or better plans or better forces, was detracting attention from the issues of overriding importance that, that is how do we make nuclear war less likely?

For a while, assured destruction and damage limitation were sort of linked as a perhaps not a strategy but a... objectives in determining force requirements and so on. And then damage limitation sort of went by the boards. Did McNamara give up on damage limitation?

In the early 1960s, in studying uh, how much is enough, in offensive and defensive forces, we were looking at the twin objectives of assured destruction. That is, the capability to survive attack and strike back in retaliation and damage limitation, that is, to recognize that one of the purposes served by our offensive-defensive forces was to limit the damage that uh, could be done to us in a nuclear war. As I recall, those objectives remained in the analysis and the basis for force planning. But in the mid-1960s, uh, we faced the question of whether to deploy the anti-missile missile. Uh, this was a proposition to spend what probably would have come to forty billion dollars or more in an attempt to defend the United States from Soviet ballistic missile attacks. Um, we analyzed that issue very thoroughly. And basically, what we found was that if you could safely assume that the Russians would do nothing in response to our 40 billion dollar weapons system, and kind of leave their missiles fat, dumb and happy, uh, with single warheads, no penetration aids and the like uh, then, indeed, this anti-missile missile could probably save tens of millions of lives. The problem with it was that in all probability, the Russians would do what we were doing. In order to be sure that we could defeat and overcome any Soviet anti-missile missile, and of course, they were developing an anti-missile missile also, we were doing things like developing and deploying multiple warheads and penetration aids. For example, we had something called resonant dipoles, packets of little wires that uh, that would resonate on the frequency of the enemy's radar. And then when, when this gets out into space, you would uh, set off a charge and go poof, and these would fill out-- you know, spread out all over space, and space would be filled with things that looked to the enemy radar like uh, like a uh, uh a warhead. And so the, the other side's defensive system couldn't tell which was the real warhead until uh, uh, all these objects came back down into the atmosphere. Well, when we went through the whole analysis, we found that if we spent the 40 billion dollars and the Russians spent something like 5 billion dollars with the kinds of things that we were sure were technically feasible, the net effect would be that we'd accomplished nothing. And McNamara, I think correctly, reached the conclusion that uh, that was a, a futile waste of, of money, and we shouldn't go ahead with, with uh, the anti-missile missile. And, if you like, that sort of cast a pall on the whole objective of damage limitation as just-- just that it, it didn't look like it was going to be feasible.