105. Memorandum From K. Wayne Smith of the National Security Council Staff to the President’s Assistant for National Security Affairs (Kissinger)1


  • SALT Work Before Helsinki

This memorandum outlines the work I believe should be undertaken by the Verification Panel and the Working Group prior to the reopening of the talks in Helsinki on November 2, explains the steps I have taken at the Working Group level to get it done, and recommends a schedule for the Verification Panel Principals to follow.

The Work that Needs to be Done

I start from the premise that we are not going to undertake a major revision of the U.S. position for the opening of the Helsinki round. We have had no real reaction from the Soviets to our latest proposal and surely should wait for such a reaction before we start making major changes.

Both ACDA and OSD have underway very comprehensive studies of the whole range of SALT issues,2 including such questions as what modifications we could accept in the numbers we have proposed and a variety of other U.S. fallbacks. I believe it would be a mistake for the Verification Panel to initiate a similar study on an inter-agency basis, because it would imply that the current position is not a firm one, and, no matter how firmly we said that a comprehensive review was intended only to prepare for contingencies, would tend to re-open all the old conflicts.

Moreover, I doubt if such an exercise would help much substantively. We don’t know very much about what changes in our position the Soviets may seek and trying to anticipate all possible Soviet attitudes would simply provide a vehicle for agencies to re-argue their old views.

[Page 345]

As I see it then, the basic analytic job to be done for Helsinki II is to sharpen our analysis and understanding of the current position.

The major subjects where more work is needed are:


The collateral constraints. It doesn’t require much speculation about Soviet views to recognize that these will be a major subject of discussion. As preparation, we need to know which of these it is most important to insist on. This involves knowing both what each one buys us in terms of verification and what each one costs in terms of limiting U.S. actions. We also need to know much more precisely what specific actions would be covered by the collateral constraints. The need for more precise statement of the collaterals was demonstrated late in the Vienna talks in connection with the ban on “externally observable modification” of existing silos, where some members of the Delegation gave off-the-cuff explications which, if accepted as the final and official definition of the ban, could seriously undercut its value for verification.

The Working Group has underway a study of these constraints. This study, which is a good start but in a somewhat dormant state now, should be completed, taking into account Soviet questions and views expressed at Vienna. Once this is done, it can serve as a basis for drafting more detailed explanations of the constraints for the guidance of the Delegation, as well as possible later re-assessment of our position in the bargaining process.


The accidental attack problem. We need more fundamental work, not only on the technical details—which is already going on—but on the basic problem of “accidental” war. I think there is some danger that we are letting ourselves be driven by the negotiating context and the Soviet and U.S. proposals and counterproposals instead of understanding the basic problem—how a nuclear war might start by accident—and deriving our position from that.

The current draft of the “technical” accidents paper3 does considerably better with the technical details than the earlier draft. In addition, a special study is being made of improving the direct U.S.–USSR communications link. We need to keep this work focused on the central questions: What events have the greatest potential for setting off an “accidental” response and what are the practical problems in [Page 346] assessing such events and communicating with the other side about them?

A specific problem, however, is to make sure that the technical work is connected to the policy problems involved, and that we take advantage of this opportunity to look critically at how our command and control procedure could break down in the face of a limited nuclear attack. There is some tendency in some of the work done so far to have the attitude, “If these regulations are followed, there will be no wrong reaction, and the regulations must be followed, because not to follow them would be a breach of the regulations.”

Fortunately, some people in DDR&E are interested in addressing this problem in a somewhat broader context, and they should be enlisted in the work being done for the Panel.

The “form of the agreement” problem. This, as explained in my earlier memo to you recommending that you initiate a study of this subject (Tab A)4 involves not only the particular documents and words to be used in an actual agreement, but the substantive effect of choices of different approaches. In addition, this is the context in which to look at the procedures we would want to set up under the agreement for dealing with new situations and possible violations.
Other details. There are a number of other parts of the U.S. proposal which require further definition for us to understand precisely what it is we are asking for. Among these questions are:
  • —Destruction and dismantling procedures;
  • —The duration of the agreement; (We have used five years as a possibility, but we should be sure that is a sensible number. I believe, for example, that Mr. Nitze thinks it is too short and a somewhat longer period would be wiser.)
  • —Invitational inspections; (What good would these inspections do? When would we use them? If they are purely voluntary so that the Soviets could refuse without there being a violation, why make an issue of asking for the right to ask for them—which we would have in any case?)
  • —The U.S. radar equivalent to the Hen Houses;
  • —Forms and procedures for pre-announcement (e.g., of deployment of new types of SAMs, of ABM tests), and consultations (e.g., on new non-ABM radars, new types of bombers);
  • —Transitional arrangements to reach agreed levels;
  • —Procedures and implementation of non-interference with national means of verification.

[Page 347]

These seem to me to be the major areas of continued uncertainty in our position (in addition to those covered by the work on collateral constraints, accidents, and the form of agreement).

We need a relatively brief paper addressed to each of these subjects, identifying the major choices and the strategic, verification and negotiating implications of each approach, as the basis for preparing detailed guidance amplifying the present instructions.

Some of them are relatively straightforward (e.g., procedures for dismantling ABM radars), and all that is required is setting down in some agreed and authoritative form what is meant. Others we know will prove controversial, e.g., equivalent U.S. radar system, and some of the ones we think should be simple may prove controversial.

What I Have Done

On Thursday, September 17, I held a meeting of the Working Group where I laid out a schedule of work for Helsinki II. At this meeting, I:

  • —Directed that agency comments on the existing papers (i.e., those on collateral constraints, accidental attacks, and the paper on provocative attacks) be provided to the drafting agency by Tuesday, September 22 and that the drafting agencies have revised drafts completed by Friday, September 25.
  • —Directed certain of the agencies to prepare short papers on the other issues mentioned above (i.e., destruction and dismantling procedures, etc.). These papers are to be completed by Thursday, September 24.

A copy of my directive to the Working Group is attached for your information. (Tab B)5

Suggested Schedule

I suggest that a Verification Panel Principals meeting be set for about October 16 (that is, two weeks before the talks resume) to consider the papers now being done by the Working Group. That will allow sufficient time for referencing items of major policy importance for the President and for drafting any detailed guidance that is necessary.

I do not believe that an earlier meeting of the Principals is necessary nor would it be productive. The work mentioned is well in hand but it isn’t far enough along to be considered in a high-level discussion. Accordingly, I think there is nothing for you to do at this time. Around the first of October I will prepare a memorandum scheduling the Verification Panel meeting for October 16 if that still appears to be the appropriate date.

  1. Source: National Archives, Nixon Presidential Materials, NSC Files, Box 879, SALT, SALT talks (Vienna), Vol. XII, July 20–September 1970. Top Secret. Sent for information.
  2. On October 14 Smith forwarded to Kissinger a paper entitled “SALT Stand-Still Arrangement”; a copy is ibid., SALT talks (Helsinki), Vol. XIII, October–December 1970. A summary of OSD studies prepared as a memorandum for distribution by B.T. Plymale of the OSD SALT Support Group for a scheduled October 27 Verification Panel meeting is in Washington National Records Center, RG 330, OSD Files: FRC 330–76–076, Box 12, USSR, 388.3.
  3. Not found. The summary of the OSD studies, cited in footnote 2 above, described the “technical” aspect of the provocative attack problem as follows: “Issue of the mutual capability of detecting the origin of an attack and exchange of information in event of an actual attack. OSD Position. SALT should cautiously pursue the Technical problem alone. Allies should be briefed on the U.S. position immediately. Guidance to the Delegation should be explicit in that any basis for the assertion that we have an understanding with the Soviets on provocative attacks is completely false.”
  4. Attached but not printed. See footnote 1, Document 104.
  5. Attached but not printed.