78. Letter From the President of the California Institute of Technology (Brown) to the President’s Assistant for National Security Affairs (Kissinger)1

Dear Henry,

Thank you for seeing me earlier today, and for the chance to discuss SALT. I will try to convey briefly my reactions to what has happened so far and suggestions about where we go from here.

[Page 268]
I am convinced that the Soviets see advantages in reaching some agreement by the end of July, and I see some for us. But it is clear that neither U.S. approach comes close to being acceptable to them, and their ‘basic principles’ while still not so clearly defined, are very far indeed from being acceptable to us. We do not yet know whether they are flexible enough to accept an arrangement we could live with. I submit that they don’t yet know that about us, either—I suppose this is what the Soviets mean when they question our ‘seriousness.’ Our next problem is how we and they find out whether agreement is possible.
Principal objectives and concerns of the Soviets are:
To keep U.S. from getting a country-wide ABM. Safeguard is their greatest concern, because they believe it could quickly be made thick. The possibility of our deploying it is our strongest bargaining point, so long as the Congress is willing to appropriate money for it at some level (e.g. a compromise for FY ‘71 which started the third site but delayed the rest would serve to keep the negotiating value of Safeguard).
They are genuinely concerned about our forward-based aircraft and would like to get them removed. But by now they should realize that this is not in the cards; we would not jeopardize even the current poor state of the NATO alliance so much even for an otherwise acceptable SALT agreement.
They are much less eager to stop MIRV/MRV where it is, because they are behind. Clearly there is strong Soviet military pressure to have it; this is responsible for their position to allow its development. They want also to ban production and deployment without inspection, relying on ‘trust,’ and noting (uniquely for any of the ‘principles’ in their list) that world opinion is against deployment. Their obvious annoyance at our MIRV positions may stem from a feeling that we played them a nasty trick by offering to ban MRVs (which they didn’t expect) while coupling it with on-site inspection (which they thought we had agreed to drop) in one approach, while allowing MIRVs in the other, with the existing U.S. lead preserved though not frozen.
They continue to want numerical equality of an unspecified sort. Clearly, what they would prefer is to catch up to equality wherever they are behind while freezing an advantage wherever they have it (except for the NCA defense, where we could catch up—if Congress is willing to accept another change in ABM system); well, who wouldn’t? We would like to do the reverse. More precisely, they may hope to fix numbers and allow qualitative improvement in principle, but count on anti-military sentiment to stop it in the U.S. in practice.

Where do we go from here? We should try to pin them down more closely, using as a framework the agreement between Smith and Semenov to take up the work program. [Page 269]

Beyond pinning them down, we can stick, and they can too, spending the rest of the Vienna phase in debate. I am concerned that in this case they, and we, may become increasingly suspicious and turn away without finding out whether there is a mutually advantageous and mutually acceptable agreement.
We can adopt fall-backs or new packages. Though this would get new reactions from them, it might well encourage them to believe we are more eager than they and that they should therefore ask for more and wait us out. In fact I think they have as much to gain as we in security, economy, and political benefit, and we should not give them signals encouraging them to believe otherwise. At the same time, I judge that the leadership in Moscow is in fact more disunited on this issue than is the USG, and therefore more likely to react to our initiatives than to come up with new ideas of their own.
We can try to find out where they will give, and how far, and in exchange for what. The ideal would be to do so without letting them know that about us, so that the President could make final choices freely on the basis of full information. But in fact the delegation will have to convey some information about U.S. attitudes on the negotiability of individual elements and about their tradeoffs. The kind of statement-question which needs to be exchanged informally and privately is,—“We can’t possibly accept A, but perhaps we could work out something involving B and C. I don’t know whether my government would be willing, but if you think yours might, I’ll pursue it at home.”

But to say even this much one needs to have one or more possibly acceptable outcomes in mind. These would be referents with respect to which the delegation would be trying to get Soviet reactions without revealing what these conceptual outcomes were until one was chosen later as the basis of a specific proposal. By probing Soviet attitudes toward characteristics of these outcomes, a picture of one or more possible Soviet minimum positions may emerge. They would probably get a corresponding picture of ours, but we would not have set our positions forth explicitly, much less made a choice among them.

I favor this third approach. It requires, however, construction of such ‘possibly acceptable outcomes.’


I think this construction should take place in Washington, not only in Vienna by the delegation. Let me give some examples, with comments, and without endorsing any of them.

First, there is a separable item. This is forward aircraft plus MR/IRBMs. We want the first unlimited, the second limited. They want the reverse. Possibly acceptable outcomes: both excluded from the agreement; both tacitly limited to present numbers (no new IRBM silos to prevent new SS–11s in IRBM holes). This can be attached to any of the other examples.
Option D without reductions; this is, option B. This allows MIRVs, and an NCA defense. By not reducing, it makes the forces more unstable than D, but less so than without any agreement. The Soviets must be negative about a reductions item which hits only their principal strategic force. This agreement, though less complex than D, would still be comprehensive and unlikely to be achieved this summer.
Option C without on-site inspection. This meets the Soviet stated desire to ban MIRV. By forbidding testing of MRV this approach gives us some assurance against their deployment. On-site inspection would provide more deterrent, but in practice the Soviets could store replacement front-ends nearby, deploy singlets and replace them on short notice. The SAM upgrade verification loses more by omitting on-site inspection, and this question is critical to a U.S. judgment on whether C without inspection is in fact an acceptable outcome. It would be most valuable to know whether the Soviets would accept a MRV test ban and tough collateral constraints if we would forego on-site inspection. To find out may require that we ask them, being as careful as possible not to commit ourselves, but recognizing that asking the question indicates some degree of U.S. interest and risks charges of bad faith if we later decide not to follow up a Soviet expression of interest.

This outcome would also probably take longer than summer to negotiate—and we would be negotiating a removal of some U.S. MIRVs which by that time have been deployed.


A simpler and more limited agreement, covering perhaps 4 a) above, NCA ABM, and an offensive force limit. The last item is a vital part of such a package, because the second is our big trading card and we must get a lot for it. Examples: i) Soviets to complete forces under construction up to a total of 1710 SLBM/ICBM (as in Options C, D). No mobile land-based missiles. No mixing. ii) Soviets can complete all presently under construction but phase out SS–7/8 (would give total of about 1800 SLBM/ICBM), or not phase out SS–7/8 (would give about 1900). This Soviet advantage would then compensate for the U.S. bomber number advantage. In (i) Soviets might demand U.S. bomber phasedown; we could respond by asking for selection of other than SS–9s for completion to the number 1710.

There are many other variations.

I believe that 4 a) combined with 4 d) is likely to prove the most promising line. But the Washington national security apparatus should explore all of the above, and others, to provide a framework for interrogation of the Soviets. And we must also try to work the negotiations so that we are not stuck with the onus of having drawn back from a MIRV ban (Option C) or reductions (Option D) and forced a narrow agreement (though what a great advance it would be) while the Soviets wanted something ‘comprehensive’(their ‘basic principles’).
[Page 271]

By virtue of my own past and present experience, I remain highly conscious of the difference between giving advice and having responsibility. Take the above thoughts for what you think they are worth.

I appreciate very much your willingness to convey to the President the statement signed by 1000 Caltech people. In turn I will continue to remind our academic community that, whatever criticisms of Government policy may be warranted, there is no other entity in the American body politic (or elsewhere in America) which can provide coherent leadership if the President’s ability to lead is destroyed.


Harold Brown
  1. Source: National Archives, Nixon Presidential Materials, NSC Files, Box 808, Name Files, Brown, Harold. Secret. Sent for the attention of Sonnenfeldt. There is no indication that Kissinger saw the letter, but a handwritten notation on the letter reads, “OBE per Sonnenfeldt.” Brown was a former Secretary of the Air Force.