341. Letter From Killian to John Foster Dulles1
In your letter of July 3, you asked whether the President’s Science Advisory Committee could explore in a preliminary way some of the general facets of the Surprise Attack problem, with particular reference [Typeset Page 1389] to its scientific and technical aspects. Since receiving your letter, I have brought together a group of the Committee and we have engaged in a two-day preliminary discussion as to how the study you propose is to be undertaken. This discussion has resulted in a progress report containing several conclusions which I wish to present to you before we proceed with further study.
We would emphasize first that the Surprise Attack problem is wholly different in kind and in magnitude from the problem of devising an inspection system to monitor a nuclear test cessation agreement. It cannot be dealt with in such specific technical terms, therefore, as we have dealt with the current problems under discussion in Geneva. We also conclude that the examination of existing proposals for inspection zones will require a more general analysis, and in this general analysis some of the following problems will be important.
Our early warning lines and our radar detection systems are designed to provide us with a few hours’ warning of approaching hostile aircraft. This time might be extended greatly by aerial inspection of the Soviet territory or by appropriately-stationed observers within the Soviet Union. The problems involved in thus extending our warning time may be more military and political than technical. The use of aerial inspection or ground observers clearly involve difficult military estimates of the result of any specific proposal about limited aerial inspection zones or limited numbers or mobility of ground observers.
In the event that we enter into discussions with the Soviets on the problem of Surprise Attack, it seems highly likely that they will raise the problem of the training flights by our Strategic Air Command bombers, particularly when these bombers are carrying nuclear weapons. If the U.S. delegates are to discuss such a topic, it would seem of great importance that we analyze in advance the present Fail-Safe technique and its possible modifications, the effects of limiting the numbers of aircraft in flight at any time, and of restrictions on their flight paths. These, too, are predominantly military problems.
[Facsimile Page 2]We must also expect that the proposed discussions with the Soviets will not be limited to present weapons such as manned aircraft, but will encompass also long-range missiles. Their era is rapidly approaching, and consideration of them in the proposed discussions would greatly increase the significance and value of these discussions.
The size of the Soviet satellites leaves no doubt about the early achievement by them of missiles of essentially unlimited range when carrying warheads of thermonuclear size. They can be located anywhere in the USSR and still reach targets in the United States. Since this is true, it is clear that in the missile era geographically limited aerial or ground observer zones in our respective countries can provide no substantial [Typeset Page 1390] reduction of the threat of an attack without warning, although they may have other attractive features.
It now appears that unlimited aerial surveillance alone, no matter how elaborately reinforced by scientific devices, can increase warning time only to the limit of the duration of flight of missiles, i.e., to 30 minutes for ICBM’s and to 15 minutes for IRBM’s. Any further increase in warning time would require quite different kinds of measures and any discussions with the Russians would doubtless have to examine these measures. Examples of these additional means include: (a) Devising of intelligence indicators for providing strategic intelligence; (b) Ground surveillance adequate to cope with dispersed or even mobile missile launchers; (c) Controls on the design, deployment, and numbers of offensive weapons.
Any further increase in warning must be based on the observation, not of hostile acts, but of preparations for hostilities. Examples of the sorts of preparatory acts that might give indication of intention to launch a surprise attack include; (a) Forward deployment of long-range bombers; (b) Fueling and positioning of missiles in preparation for firing. It is an inherent limitation of observations of such action that they cannot give certain indication of the intent to launch hostilities. At best, the observable acts are clear indications of a state of heightened readiness on the part of the observed power. It may turn out, therefore, that no reliable system can be devised to provide dependable advance warning of a surprise attack except in conjunction with agreed limitations on weapons numbers or deployment. Given such agreement, the observation or detection of deployments or weapons production in violation of the agreement might furnish the kind of indicator that would be a reliable warning of attack.
[Facsimile Page 3]With such considerations as these before us, we early came to the conclusion that discussions with the Soviets cannot long be confined to the strictly technical aspects of surveillance. They will inevitably touch upon the problem of weapons controls, controls of deployment, etc. While there are technical aspects to almost every part of the Surprise Attack problem, it is also clear that the subject of the proposed meeting with the Soviets is of such a nature that technical questions are in inextricably intertwined with political and military considerations.
We suggest that any group making preparatory studies for the conference should recognize the impossibility of separating technical questions from non-technical questions. If the Science Advisory Committee, therefore, is to be effective in conducting the study you request, we must join with competent representatives of the State Department, the Department of Defense, the Central Intelligence Agency and other government agencies.
[Typeset Page 1391]If it is agreed that we should assemble such a representative group, we are prepared to do so, if the President approves. An Alternative arrangement would be for the Department of Defense to take responsibility for organizing it.
Before we proceed further, therefore, I wanted to place before you these conclusions to make clear our own conviction that a discussion of Surprise Attack involves many elements and that it cannot be studied in a limited technical manner. We believe, further, that the planning for the proposed discussions with the Russians must be undertaken with the assumption that those discussions will certainly lead into non-technical areas. We understand, finally, that the function of this study is not to recommend positions to be taken by the Government, but solely to make an analysis and list the alternatives which might be useful to you in reaching positions and making plans.
Yours sincerely,
- Source: Problem of surprise attack not limited to technical or scientific issues. Secret. 3 pp. Eisenhower Library, Records of the Office of the Special Assistant for Science and Technology, Disarmament, Surprise Attack.↩