The U.S. Nuclear Presence in Western Europe, 1954-1962, Part II

In the midst of the ongoing deployments of nuclear weapons to NATO Europe, John Y. Millar wrote to Richard Finn about his plan to visit stockpile sites. What piqued his interest in the visit, which had the full support of the State Department, was a question raised by the Joint Committee on Atomic Energy: whether the United States actually had “exclusive custody” of the nuclear weapons that had been assigned to NATO allies. Thus, Millar wanted to find out “what the actual arrangements are for maintaining custody and control at various stockpile sites.” For example, what were the arrangements once a “weapon had been hung on a foreign aircraft”? He asked Finn to discuss the plan with his military contacts so he could line things up” by the time Millar arrived in early July. Millar enclosed a list of the countries (UK, Germany, Italy, and Turkey) and the weapons systems that would be relevant to his trip.

By the time John Millar arrived in Paris, SACEUR General Lauris Norstad had found out about his itinerary and strongly objected to it. Believing that the proposed review of custody arrangements raised questions about his judgement and challenged his authority, Norstad refused to let Millar visit the sites. Merchant sent Norstad’s political adviser, Raymond Thurston, a letter explaining the Department’s interest in the stockpile custody arrangements: “we must be able to make our own decisions as to the appropriateness and legality of the projects within the framework of United States law and policy.”

Thurston’s long response argued that the Millar trip had been hastily arranged and that the State Department could get the desired information from the Pentagon. According to Thurston, when it was only “the military who can really know from day to day what is really going on in this particular endeavor, I find it difficult to see why the Department should feel impelled to accept this inappropriate burden.” He further suggested that U.S. allies would be puzzled by a visit to U.S. support sites for their NATO forces.

In his memorandum to Assistant Secretary for European Affairs Livingston Merchant, Ivan White refuted Thurston’s arguments and implied that it was better not to reply to the letter to avoid further controversy.

After the denial of John Millar’s proposed inspection of stockpile sites, a visit to Washington by General Norstad provided an opportunity to raise State Department interest in custody matters. Phillip Farley advised Under Secretary for Political Affairs Merchant to use a meeting with Norstad to persuade him that the Department needed to be knowledgeable about nuclear weapons and custody arrangement so that its officials could best defend them against congressional critics. Consistent with this, Farley wanted Norstad’s consent for State Department officials to visit stockpile sites with DOD officers on the staff of AEC Military Liaison General Herbert Loper.

According to Farley, the State Department’s statutory responsibility for nuclear negotiations and its policy interest in preventing nuclear proliferation and ensuring security for U.S. military assets led it to pay close “attention to the effective implementation of the provisions for the retention of U..S. custody and control” of weapons deployed overseas for the NATO stockpile program. So far, State had been confident that the Defense Department could prevent unauthorized access or use mainly because the weapons were stored in igloos “under exclusive U.S. custody” and there they would remain unless hostilities broke out. That, however, had changed with the development of new tactical nuclear weapons that might not be separately stored because it would be inconsistent with their “operational efficiency.”

Such changes were especially evident “in cases such as Genie air-to-air weapon, the Lulu anti-submarine weapons, the Davy Crockett, some of the shorter-range surface-to-air missiles and air-to-surface Missiles, and probably mobile IRBM's.” Thus, to achieve wide dispersal, quick reaction times, and optimum maintenance procedures, it was “necessary … to affix weapons on foreign aircraft kept on alert status or to incorporate (i.e., to “mate”) the weapons into missile delivery systems at the launch sites.” With foreign personnel near the weapons this makes custody more difficult, especially in the case of the Genie and Lulu weapons that must leave the ground (and U.S. custody) when hostilities are near.

Farley reviewed the State-Defense discussions of the Genie missile and the problems raised by the Joint Committee on Atomic Energy’s objections to the custody arrangements. With such disputes likely to reoccur, “it is essential that we have a better understanding of this complicated field.” Therefore, “we hope you will be able to reassure General Norstad that the Department’s interest in this field in no way reflects our doubt about his competence and sincerity in affording proper protection for U.S. nuclear weapons.”

Farley asked Merchant to convey to Norstad that the State Department agreed that security arrangements for nuclear weapons was a matter for the responsible commanders in the field. Nevertheless, if the Department was “to defend our position that we still maintain adequate custody and control despite the newer deployment techniques we must be able to speak with knowledge.” The extent to which Merchant succeeded in persuading Norstad remains to be learned.

State Department interests in the deployments continued into the new Kennedy administration. One of the Department’s lawyers, J.H. Pender. reviewed the recent attempts to secure information on custody arrangements for the deployments in Europe. The Defense Department had begun to provide some details, but key Pentagon officials believed that the State Department had no responsibility for “actual procedures” at nuclear sites. Nevertheless, Pender believed that the Department needed to know more because members of the JCAE, who had “first-hand knowledge of actual practices at the deployment sites,” were likely to question State Department representatives on “what they think of the domestic legality and general effectiveness of custodial control arrangements; whether the Department had endorsed them from either standpoint; and whether, in effect, the Department regards itself as having any ‘civilian control’ responsibility in this area.”

Believing that the Department had a civilian control responsibility, Pender suggested that it “exercise general supervision of the program to the point of periodically visiting SHAPE installations to satisfy ourselves that the custody and control procedure are legitimate and effective.“

Pender also recommended that the matter be raised with the President. State Department legal adviser Abram Chayes sent the memorandum to the secretary of state but Rusk did not make a decision on it [see document 21]. It is not clear what exactly was decided, but by April 1962 Pender was one of the State Department observers to an AEC-Defense stockpile program survey (see document 29), for which he wrote a full report.

Congress had the most traction with problems of nuclear custody and control and the Joint Committee on Atomic Energy had a major policy impact. During late 1960, Rep Chet Holifield [D-CA] and other members of the JCAE, accompanied by staffers from Atomic Energy Commission laboratories, inspected 15 nuclear weapons installations in eight NATO countries. The trip report, often referred to as the “Holifield Report,” has been available for years in excised form and efforts to prompt the release of more information through appeals have been unsuccessful. Indeed, some of the excisions are in the silly secrets category such as the deletions of references to Italy and Turkey in the context of discussion of Jupiter missile deployments in those countries. In any event, a State Department summary is in the declassified record [See document 7] and it can be used to decode some of the excisions.

The report had a significant impact on the incoming Kennedy administration because it raised so many troubling questions. One of the Committee’s foremost concerns was the political stability of countries such as Italy and Turkey where Jupiter IRBMs were deployed. Other and related concerns included the vulnerability of weapons systems to seizure by a “psychotic” from the host country’s military or by a colonel’s coup, the danger of accidental detonation or unauthorized use, security of weapons design information, arrangements for destruction of weapon if necessary, and sufficiency of trained personnel in the event of an accident. Also of concern was what the Committee saw as virtually “fictional” U.S. custody arrangements, whereby the nuclear weapons were in the virtual control of the host country, such as West German fighter-bombers on “Quick Reaction Alert.”

Concerning those fighter bombers, one of the scientists, Harold Agnew, who accompanied the JCAE members to Europe, vividly recalled the situation at an unnamed West German base where German bomber pilots had virtual control of the nuclear weapons. Concerned about the possibility of misguided or accidental nuclear use, Agnew spoke to a young U.S. soldier who had responsibility for guarding the weapons and asked him what he would do if the German pilots came “running out and they're gonna take off and no one has told you that it's all right.“ The sentry was uncertain, so Agnew advised him to disable the weapons with his gun: “shoot at those things and don't worry about it."

The Committee did not have an overarching proposal to fix the problems that it had identified. Instead it suggested a range of options, including: 1) a complete U.S. system of U.S. possession and custody; 2) separating nuclear components from delivery systems and placing them under U.S. possession; 3) continuing “current fictional custody arrangements, involving elements of joint possession and control;” 4) explicit joint possession arrangements for alert weapons; 5) “transfer of control of U.S. nuclear weapons to independent NATO task force arrangement;” or 6) transfer of nuclear weapons to individual NATO countries, with different arrangements with different countries depending on geography and political stability, among other considerations.

Among the specific recommendations the JCAE made was to strengthen custody and control was the development of electronic means to arm or disarm weapons that were on quick-reaction alert delivery systems. That would mean solving such problems as finding ways to control weapons that were already deployed and creating new devices for ‘improved weapons systems.” This recommendation was one of the roots for what became known as “permissive action links.”

This is the full version of Holifield’s letter to Kennedy, with references to Jupiter missile deployments in Italy and Turkey.

The State Department’s review of the JCAE report gives a sense of some of the issues obscured by the excisions, such as IRBMs in Italy and Turkey. The staffer who prepared the appraisal characterized the JCAE’s report as a “high-level forceful and continuing foraging expedition on the most sensitive and critical arrangements of foreign policy.” While he found that the JCAE did not have consistent political, strategic, or ideological views, he did find the “apparent assumption that only with the U.K. can we have reliable land-based weapons and nuclear weapons cooperation.” Thus, Holifield and his colleagues were especially critical of deployments of land-based missiles on the European continent because of their concern about political instability in Italy and elsewhere

According to CNO Arleigh Burke’s account of this meeting, Kennedy told the Chiefs that he was reading the Holifeld report and “wanted to make sure we all read it.” Kennedy said he was concerned about the control of nuclear weapons and “wanted to know what our ideas were on nuclear weapons control and whether our weapons control actually conformed to law.” Suggesting that he was still learning about presidential power, Kennedy also asked: “What is the President's authority? What should he do? Should he change our procedures in any way? He wanted to discuss this thoroughly at the next meeting.”

After an extended discussion of guerilla war issues, Kennedy raised more questions about nuclear weapons, including whether the Chiefs were “satisfied with the nuclear weapons system.” He asked the Chiefs for a briefing on “local control of nuclear weapons and dispersal of nuclear weapons.” In particular, he wanted to know if a Communist attack across the East-West German border with conventional weapons could be halted if it was with “such force that the Europeans were being forced back.” He continued: “would they [presumably NATO forces and/or France] not shoot their nuclear weapons? How could we control it? Do we have positive control?”

It is not clear if or when the discussion of nuclear weapons that Kennedy asked for happened. With his diminished confidence in the Chiefs after the Bay of Pigs fiasco in April 1961, Kennedy turned elsewhere for military advice.

In light of the controversial issues raised by the Holifield report, U.S. dispersals to NATO allies came to an end. According to a Gilpatric memorandum from September 1961, on 11 April 1961 he recommended to the President that the U.S. halt dispersals of nuclear weapons to “non-U.S. forces” although the rest of the nuclear weapons dispersal plan that had been approved at the end of the Eisenhower administration could proceed. Gilpatric further noted that those dispersals would halt until a decision could be made based on a pending policy review. Kennedy sent Defense a letter approving Gilpatric’s recommendation on 20 May 1961. That would leave the deployments to NATO allies at 500 weapons [See Documents 23 and 24 for more details]

At this time, Washington and London were still negotiating a stockpile agreement that would cover U.S. nuclear weapons assigned to British NATO forces in other areas or for the support of NATO units in the United Kingdom. The British provided a new draft for consideration by U.S. negotiators with language that they considered “more general” so that future developments would not require renegotiation of the agreement. One problem that the British brought up was that the agreement could not be called an “agreement” because under their law treaties and international agreements had to be registered with the United Nations in New York. To get around the problem, they suggested that the agreement be characterized as a memorandum of understanding that would be the subject of an exchange of letters. As U.S. negotiators did not want to see the agreement published either, they agreed to characterize it as an “understanding” [See Document 14]. The language of the British draft remains classified but presumably it was along the lines of the all-purpose “umbrella” agreement discussed earlier in 1960.

Assistant Secretary of State for European Affairs Foy D. Kohler provided Secretary of State Dean Rusk with background on the ongoing negotiations on atomic cooperation and atomic stockpile agreements. On the former, negotiations with Belgium and France (probably on the atomic stockpile for French forces in West Germany) were ongoing. Also, underway was the negotiation to update an earlier atomic information agreement with NATO.

The stockpile agreement with Italy had been delayed because the Italian Government wanted the United States to bear the financial costs that other host countries have been willing to accept. The U.S. was formulating a response to the Italian position. The other pending stockpile agreement was with the United Kingdom, which would cover U.S. nuclear support for British NATO forces, including the “NATO forces of third countries stationed in the UK.” For example, in 1960 the U.S. had provided the British Army of the Rhine with Honest John missiles but nuclear weapons could not be provided for those units without the stockpile agreement.

Owing to the military importance of these agreements, Rusk gave his approval to expediting their negotiation.

The ongoing U.S.-Italy atomic stockpile negotiations faced a new complication when Italian officials proposed that the agreement include language formalizing a role for Italy in nuclear use decisions. During a conversation with Embassy official Ernest Siracusa, Paola Pansa, the Chief of the NATO Section at the Foreign Office, cited the “two-key” arrangement controlling the use of Jupiter missiles but observed that “no such concurrence covers the presence here of other weapons for which the United States controls both the warheads and the delivery systems.” Noting that all Italian dual-use delivery systems have a “two-key” control procedure, Pansa believed that the “situation should be formalized for all weapons so that Italy would control over the use of any such weapon on Italian territory.”

Pansa said that it was not yet a formal proposal but was “receiving serious consideration,” adding that “he would not be surprised if it became" a “formal proposal.” To that, Siracusa said he saw “no need for such an arrangement” because the “real basis of the two-key system is to comply with U.S. law regarding warhead custody and use.” That system was “practiced with respect to all nuclear-capable Italian-held delivery systems, without being spelled out in an agreement.” Also protecting Italian interests was that nuclear delivery systems would be used only in accordance with the SACEUR’s instructions.

Siracusa tried to close the case by arguing that because none of the other stockpile agreement had consent language the “lack of need for such a provision is strongly suggested.” That argument, however, would find no traction in Rome.[5]

The matter of an Italian role in nuclear use decisions raised in the 29 June 1961 airgram and other messages from the Rome embassy created “serious concerns” in Washington “over possible implications of adding any language to stockpile agreement itself which injects bilateral US-GOI understanding on employment of weapons.” The State and Defense departments saw such language as a troubling precedent that “could lead to requests for similar most favored nation treatment if it became known to other NATO countries.” Moreover, it could “easily lead to requests for physical operational control procedures to make such understandings effective, which could have adverse effects on ability of NATO forces to respond in emergency.”

The State Department suggested that the Embassy could reassure the Italians that they have a “real voice in use of weapons through NATO procedures.” Accordingly, the agreement could include such language as the weapons will be “employed in accordance with procedures established by SACEUR which will be in accord with approved NATO plans and policies.” That, however, did not settle the matter.

Washington and London finalized their stockpile agreement in the midst of the Berlin Crisis. The language remains classified, but it probably provided the flexibility, that both sides sought, for the deployments of naval nuclear weapons and weapons assigned to British-controlled delivery systems in West Germany.

U.S. Ambassador to Italy George Frederick Reinhardt asked the Department to supply instructions authorizing him to complete the stockpile negotiations by including a “separate ‘consent for use’ agreement” and providing for consultations on numbers of weapons deployed. With respect to the latter, the Italian Government had “strongly argued that it must know extent to which Italy being used for storage weapons this type and believes this is a matter which no sovereign government could or should ignore.” That the U.S. had taken a strong position against this had aroused “suspicions that we may plan store in Italy weapons beyond needs for Italian defense.”

Reinhardt further noted that Rome “has been very cooperative” on atomic weapons issues by permitting “entry weapons thus far on basis informal arrangements” and had “acquiesced in our having nuclear weapons at Aviano without even informal approval.” On consent for use, Reinhardt agreed with General Norstad that “such a request cannot legitimately be denied.”

Deputy Assistant Secretary of State for European Affairs William R. Tyler informed Rusk that the stockpile negotiations with Italy had reached a “critical stage.” Completing the talks was necessary for the sake of U.S. relations with Italy but also to improve the “readiness of NATO forces in Italy during this period of crisis over Berlin.”

The State Department had to coordinate with the Defense Department a position on two major issues. The “most important is the Italian request for an arrangement which would provide for obtaining Italian Government consent prior to the use of any nuclear weapons in Italy.” The second matter was meeting the “Italian desire for greater assurance that the Italian Government is fully consulted in regard to decisions on numbers of nuclear weapons” deployed in Italy. A lesser problem was cost issues.

On use decisions and consultations on numbers of weapons, Tyler advised Rusk that “we should promptly and gracefully indicate to the Italians that we are prepared to meet their desires on these two points.”

Tyler also recommended that Rusk sign a letter to Secretary of Defense McNamara that presented the Department’s views on these issues. According to the letter, assurances on consent for use of nuclear weapons “cannot be refused to a host country which requests it and which attaches political importance to it.” The U.S. already had similar provisions in agreements with the British and the French and also with respect to the Jupiter missiles deployed in Italy. Further, it would be “most unfortunate if we were to persist in positions [against the Italian requests} which might undermine the mutual basis of trust which has existed in the atomic weapons field with Italy up to now.”

In this update on the Italian negotiations Allen James informed an official at the Bureau of European Affairs that in McNamara’s response to Rusk, he wrote that the Department preferred to avoid an agreement with the Italian government on nuclear use decisions until NATO had had more discussions on the issue. Nevertheless, in their instructions Defense and State Department officials gave Ambassador Reinhardt considerable latitude in reaching an agreement. Thus, if the U.S. could not persuade the Foreign Minister to delay consideration of the issue until NATO had discussed nuclear use issues, if he “still wishes separate understanding on use [of] nuclear weapons in Italy, Ambassador may state US prepared to meet GOI wishes.” Furthermore, the message authorized Reinhardt to approve, which he would do, language concerning “agreement of the two governments [to] be given” in light of “circumstances at the time.”

With the completion of the U.S.-Italy atomic stockpile negotiations, including agreement on the consent language, the final agreements were signed in Rome by the U.S. chargé, Outerbridge Horsey, and Italian Foreign Minister Antonio Segni. The agreements covered such issues as storage site funding, procedures for use, and responsibilities for maintenance and security.

A separate note covered decisions to use the weapons, which had been one of the most challenging negotiating issues. The two governments agreed that use decisions required that “agreement of the two Governments would be given in the light of circumstances at the time and having regard to the undertakings they have assumed in Article 5 of the North Atlantic Treaty.” In addition, the U.S. would arrange with Italian military authorities procedures for keeping them informed of the “numbers of weapons scheduled for location in Italy in furtherance of NATO plans.”

Whether other stockpile agreements were modified along the lines of the U.S.-Italy agreement remains to be learned. Later in the decade, the West Germans successfully negotiated an agreement with Washington concerning limited or “selective” use of weapons stored in West Germany, but it was a head of state understanding, not specifically related to the stockpile system.

For several years, the U.S. and Canada had been discussing bilateral arrangements to cover the various nuclear deployments that involved the two, not only weapons assigned to Canadian forces in NATO Europe, but also naval nuclear weapons and weapons for air defense purposes. From Washington’s perspective in the early 1960s, improving difficult relations with Canada depended in part on a resolution of the stockpile negotiations, but the anti-nuclear sensitivities of Conservative Prime Minister John Diefenbaker and Canadian public opinion had to be taken into account. While the U.S. had accepted the prime minister’s demand that nuclear weapons not be stored in Canada for the BOMARC air defense missile, Armstrong argued that in exchange for that concession Ottawa should accept storage of U.S. strategic weapons at U.S. air bases in Canada.

On the point of strategic weapons, Armstrong was slightly confused because since 1950 Canada had allowed temporary/transient storage of strategic bombs at Goose Bay Air Base, which became home to the Strategic Air Command’s 95^th Wing. In 1951 Canada authorized the construction of special bunkers for the weapons, but storage remained temporary only.[6]

The released record of the top secret talks in Ottawa with Air Chief Marshall Frank Miller and Assistant Secretary for External Affairs Ross Campbell about the stockpile agreement is a prime example of U.S. government over-classification. The fascinating published Canadian record on U.S.-Canadian nuclear relations includes documents on major issues that have been excised or exempted in this group of documents.

To participate in the talks with Miller and Campbell, the Defense Department sent William B. Lang to work with two senior embassy officials, Ivan B. White and Rufus Z. Smith. The opening of the discussions focused on access by Canadian forces to U.S. nuclear weapons stored in Europe, such as in West Germany. Like the other stockpile agreement with NATO allies, the U.S. proposal was a basic “umbrella agreement” that would cover a variety of delivery systems. After the U.S. presented the Canadians with a draft agreement, they “appeared to be agreeably surprised at [its] relative simplicity.” When Miller observed that “there must be a mass” of additional agreements to be worked out Lang responded that all that would be needed were “technical level arrangements, fleshing out certain details on matters covered in principle by the government-level agreement.”

For all practical purposes, the State Department exempted the discussion of NORAD and Canadian forces in its entirety and the follow-up conversation on nuclear weapons for Canadian air defense is heavily excised. The declassified Canadian record includes a detailed account of what the Canadians proposed: keeping the nuclear delivery systems in Canada but the nuclear warheads in the U.S. where they could be retrieved by Canadian air forces in an emergency. The proposal came from the Cabinet Committee as a way for Canada to retain its “formal non-nuclear” status while trying to cooperate with the U.S. on air defenses.

White said that Washington would consider the proposal and discussions continued on various options, from ferrying the warheads to keeping a vital nuclear component in the United States, until such time as it needed to be deployed to Canada. In any event, the U.S. rejected the Canadian proposals in early January 1963 on a variety of grounds: “Aside from tactical disadvantages and serious practicable difficulties, such as the RCAF going at the wrong time in the wrong direction, legal advisers within the United States Government have confirmed the previous view of the impossibility of the United States Government releasing control of the weapons for free flight, including return to Canadian bases, until the military emergency was far advanced. At such a stage of emergency, the usefulness of RCAF aircraft would be greatly degraded at a critical time for effective defense purposes since they would not be in a normally prescribed defense location.”[7]

The U.S. response symbolized the stalemate in the stockpile negotiations but matters worsened when Prime Minister Diefenbaker disclosed the top secret talks publicly in a speech to the Canadian Parliament and misleadingly suggested that Washington considered Canada’s proposals on storage for air defense nuclear weapons to be feasible. The U.S. then took the unprecedented step of critiquing a prime minister’s speech in a press release. Those, among other issues, precipitated a near-crisis in relations with Washington, with the U.S. tacitly supporting Diefenbaker’s Liberal opponent Lester Pearson, who had become willing to accept storage of nuclear weapons on Canadian soil. The controversy ended with elections that elevated Pearson to prime minister and the conclusion of a nuclear agreement during 1963.[8]

With the State Department role still unresolved, Pender briefed Chayes on two Defense Department papers, one by Assistant Director of Defense Research and Engineering Marvin Stern on electronic devices to control weapons, and one by the Joint Chiefs of Staff on the custody problem. To prevent the problem of accidental detonation, which Stern saw as a “finite probability,” he proposed mechanical and electronic “interlock” to keep weapons in an “untriggered” mode. An electronic interlock could prevent premature or unauthorized detonation, but it was years away from development. A mechanical interlock could prevent unauthorized use and contribute to greater safety, but it would not prevent premature nuclear use by commanders in the field.

While Stern believed that the interlocks would safeguard custody and control, Pender was not wholly persuaded because the devices could “not preclude seizure of, or unauthorized access to, the weapon and consequently may not resolve our custody dispute with the Joint Committee even though the device would probably remove the primary motive for seizure.”

The gist of the JCS paper was that everything concerning custody, control, and security of nuclear weapons was satisfactory: it “reflect[ed] suspicion of anything which could make it possible to circumvent the established chain of command.” Consistent with that, Pender wrote that he was not surprised to “find the JCS memorandum concluding that all is well with the atomic stockpile program and there is no need for any changes, even such as that posed for consideration by Dr. Stern.” Pender was most disconcerted by the “feeling one gets from its general tenor that units in the field equipped with the weapons are going to use them [in a conflict situation] when and how they see fit ... without awaiting a decision by the President.”

Pender saw “nothing in either paper … which would warrant a change in our previous position that we should insist on having a voice in at least certain of these matters before action is taken.”

President Kennedy remained concerned about the control of nuclear weapons, especially as the Berlin Crisis intensified. On 31 August 1961, he sent Secretary of Defense McNamara a memorandum asking that General Earle Partridge’s National Command and Control Task Force “give urgent attention to measures to strengthen control over nuclear weapons in NATO Europe during the Berlin crisis.”

The Partridge report found that arrangements to safeguard weapons under U.S. control were adequate but recommended steps to prevent seizure of weapons assigned to non-U.S. forces. While Partridge found that nuclear weapons deployments authorized by General Norstad were “appropriate to the situation,” Rusk advised Kennedy that the Partridge Task Force had not focused on the problem of weapons security, and that the Defense Department should be encouraged to look into this issue further.

As a security measure, commanders of units armed with nuclear weapons “have been directed not to fire (or to make final preparations for firing) nuclear weapons without specific authority emanating from CINCEUR. The report also noted that “permissive links” would strengthen effectiveness of controls against unauthorized use, and that the AEC and the Department of Defense needed to give priority to developing and producing such devices.

Rusk provided the text of a letter that President Kennedy could send to Secretary McNamara. It included language supporting vigorous efforts to encourage the strengthening of “US custodial safeguards against unauthorized use of nuclear weapons by non-US NATO units,” and the hastening of the “development and production of the ‘permissive link’ by according it priority in both the DOD and AEC.”

This massively excised report to President Kennedy was released in 2004; it is possible that a new review would produce somewhat more information, but the numbers of weapons and the names of the countries to which they would be assigned would probably remain classified.

Gilpatric reported to Kennedy on the ongoing policy review of security and safety for dispersed nuclear weapons as well as on the policy importance of continuing the dispersal program to NATO countries. To improve the security of the weapons, “corrective action” and the program for PALs was proceeding on an “urgent basis.” In addition, a review of nuclear sharing policy had concluded that it “would be extremely disruptive to alliance cohesion if we were not to withhold the nuclear weapons” that are needed to make “fully effective” the delivery systems that the U.S. has committed to provide to the allies. Only a full discussion among the allies could lead to a change in the policy, but the United States did not want to prejudice the outcome of such talks by providing more weapons than were needed to make good on U.S. commitments or by unilaterally bringing the dispersals to an end. Consistent with those considerations, Gilpatric asked for authority to begin a program of dispersals that would meet “immediate operational requirements.”

Whether Kennedy approved all of the details of Gilpatric’s request is unclear, but he soon authorized the dispersal of 1,580 weapons, as of 1 July 1962, in the air strike, battlefield, and air defense categories. Perhaps on Gilpatric’s recommendation, Kennedy further decided that two-stage weapons (H-bombs) would not be placed on Quick Reaction Alert aircraft

With President Kennedy’s decision to resume the dispersal of nuclear weapons in NATO Europe, Defense Department and White House officials gave a briefing to JCAE members on the dispersal plan. As noted earlier, as early 1961 the U.S. had 500 air strike and battlefield weapons assigned to non-U.S. NATO forces and Kenned approved the assignment of 1580 more weapons as of 1 July. Further deployments would depend on “future decisions on NATO strategy.”

Kennedy had given top priority to installing permissive action links on Jupiter missiles based in Italy and Turkey. According to AEC Chairman Glen Seaborg installation on the Jupiters would be completed by the end of the year, with air strike weapons, Sergeant, and Pershing missiles next in line for PALs.

Deputy Secretary of Defense Gilpatric explained that the reasons for the dispersal were both political (“to avoid damage to the alliance”) and military (to prevent “degradation of military capabilities).” The members of the Joint Committee were skeptical of the plan, with Senator Henry Jackson (D-WA) arguing “that the Europeans would use [it] as an excuse for not building up their conventional forces.” He further argued that it “was not dispersal but rather proliferation.” McGeorge Bundy thought otherwise, later saying that “we must try to hold NATO in a single nuclear position or risk seeing it disintegrate into a series of national nuclear capabilities.”

Presidential aide General Maxwell Taylor had visited two AEC laboratories working on PALs: Livermore in California and Sandia in New Mexico. Scientists at Sandia had developed a combination lock that could be used for Jupiter missile warheads and other weapons. For intruders seeking to disassemble the weapon, it could delay their access by several hours, but would not prevent tampering or sabotage.

According to Taylor, scientists at Livermore were developing a far more “sophisticated long-range solution” that could safeguard against misuse and tampering. “It takes into account many contingencies, an ally seizing a weapon, the action of a psychotic attempt to fire one, a saboteur bent on firing the weapon or rendering it inert, the theft of a complete weapon.” Using a special communications network to ensure that the code becomes available in time, the Livermore PAL would take several years to develop but could only be used on new weapons.

So that the scientists at the labs can work with “maximum effectiveness,” they need “more precise guidance.” Thus, they need to know what performance is expected of the PALs, “the time available to produce them, the weapons systems to be protected, and the point in the chain of command where the permissive code is to be held in custody.” Kennedy would soon provide guidance that met some of those concerns.

In this follow-up to the White House meeting a few weeks earlier, senior members of the JCAE raised questions about commitments made by the Eisenhower administration, nuclear weapons on quick-reaction alert aircraft, front-line dispersals and the impact of the deployments on the Soviet Union. They welcomed Kennedy’s interests in permissive links but cautioned that the Eisenhower administration had not committed the U.S. to the “planned proliferation of hundreds more weapons,” which may be unnecessary in light of the anticipated Polaris deployments and “may in fact cause additional reflex actions on the part of the Soviets of an extremely dangerous portent.”

The Committee members raised doubts about decisions to “beef up” the force of QRA aircraft but endorsed Kennedy’s decision to excuse two-stage weapons from the non-U.S. fighter bomber force. Believing that nuclear weapons on the front-line were vulnerable in the event of a “sudden onslaught or probing action,” they recommend pulling them back further to ensure that front-line commanders did not make unilateral decisions to fire the weapons “if they were overrun by a superior conventional force.

The JCAE also raised questions about the possibility of nuclear proliferation in the Soviet bloc, the “questionable” principle of uniform treatment of NATO members, the need for public disclosure of some of the NATO arrangements, such as QRA aircraft, and their concern whether the United States had legally “lost possession” of nuclear weapons deployed on non-U.S. NATO fighter-bombers.

In this National Security Action Memorandum, President Kennedy directed the agency chiefs to make a “commitment to procure appropriate devices for all nuclear weapons, now dispersed and to be dispersed to NATO commands, for both non-U. S. and U.S. forces.” This was Option V in the attached paper by science adviser Jerome Wiesner, which also included weapons “based in the U.K. and assigned to the naval attack aircraft on carriers based in European waters.” Thus, the PALs would prevent unauthorized use by non-U.S. forces and U.S. forces, including actions by an “individual psychotic,” but also against premature use in periods of “high tension or actual military combat.”

In his paper, Wiesner called for the undertaking of a “vigorous program … to develop an improved electronic lock which would be incorporated directly in the electronic package associated with all future weapons so that the option of a permissive link would always exist.” He believed that AEC equipment that was then under development could be “used as the basis £or a crash program’ with “initial production and installation” beginning in the immediate future. The proposed equipment would involve an “electro-mechanical lock which would have to receive a preset numerical code in order to make the weapon operable.”

He listed five alternative programs for the application of permissive links, including option I limited to the warheads on Jupiter missiles deployed in Italy and Turkey and the weapons assigned to non-U.S. quick-reaction alert aircraft in NATO Europe. That option was ruled out perhaps because it was considered to be inequitable.

One of the battlefield nuclear weapons that the U.S. had deployed in West Germany was the Davy Crockett: a short-range recoilless gun, for use by infantry, that carried a weapon with an explosive yield of ten to twenty tons TNT equivalent. While Defense Department officials had thought that PALs could be applied to the Davy Crocketts, it turned out that the locks were too heavy to be used on these relatively small weapons.

Because of the “dangers” associated with Davy Crockett deployments, Alexis Johnson provided Rusk with a letter that could be sent to Secretary of Defense McNamara. In the letter, Rusk cited the “risk of an unauthorized firing” should a soldier “armed with a Davy Crockett ... fire off his weapon, without orders to do so, in self-defense as the fog of combat swirled around him.” With more of the weapons deployed, the risk would increase.

Rusk also pointed to the possible negative impact of Davy Crockett deployments on NATO. In McNamara’s recent speech at the NATO meeting in Athens, he spoke depreciatingly of the “possibility of limited and useful employment

of tactical nuclear weapons and emphasiz[ed] the possibility of more than transient non-nuclear combat in Europe.” The Davy Crockett deployments could “degrade the credibility” of McNamara’s emphasis on conventional defense of NATO Europe, especially when Germans were asking about the weapon.

Whether Rusk’s argument had an impact on the Defense Department planning, e.g., by moving the Davy Crocketts further from the frontlines, the weapons remained part of U.S. army forces in Germany until 1967.

State Department lawyer John Pender remained involved in policy on nuclear weapons custody and eventually had an opportunity to see at first hand the procedures at U.S. nuclear sites in NATO Europe. He was one of two State Department observers with an AEC-Defense survey at the sites during April 1962.

Pender found that the situation—virtually fictitious U.S. custody and control—was close to the one described in the 1961 JCAE report. At Air Force sites, he wrote that U.S. “custodial personnel are … not guardians of the weapons; they do not control access to the fenced areas within which such weapons are positioned.” The custodial personnel “have only accountability responsibilities and are expected to rely upon local allied personnel to deal with any and all intruders.” Moreover, allied military personnel “could easily help themselves to the mated weapons whenever they felt it necessary to use them, without United States consent; there is no over-all considered or effective United States plan in being at the moment to prevent this eventuality.”

Referring to a statement made by General Herbert Loper to the JCAE in 1958 that “we hold in our possession and custody” the nuclear component, Pender observed that he found it “difficult, if not impossible, to reconcile” current arrangements with the “original impressions as to how United States forces would retain custody and control of stockpile weapons.”

On the weapons assigned to the Jupiter missiles deployed in Italy and Turkey, Pender argued that U.S. custody of the warheads was “essentially symbolic” because it was “impossible to evacuate or destroy the warheads” without the cooperation of the Italian or Turkish military. It was essentially a “joint custody” arrangement, not the “exclusive” control that was characterized in a U.S. aide memoire to the Soviet Union.

Before permissive action links became a reality, Pender believed that the Defense Department needed to make some interim arrangements to provide for certain control of the weapons. The U.S. would want “the Air Force custodians to be genuine guards not just symbols; that the aircraft should have some device under United States exclusive control preventing takeoff; that plans and temporary devices for quick destruction of Jupiter [launch control] consoles and the aircraft be worked out urgently.” Whether any of these proposals received further consideration remains to be learned.

Joint Chiefs of Staff Chairman Lyman Lemnitzer testified during a JCAE executive session, although the committee members did most of the talking. Permissive action links were one of the topics as was the role of conventional and tactical nuclear forces in NATO. Holifield provided some background on the Committee’s role in advancing the PAL concept but noted that “our Committee has become concerned lest there should be a premature decision to install devices in all weapons systems prior to obtaining operational experience.” Lemnitzer had a similar concern about moving “too rapidly to install the permissive links … with respect to weapons strictly in U.S. custody.” He observed that “we can be disarmed, in effect, by having a device that is hurried through and does not permit the commander to employ the weapons effectively.” This was a tacit criticism of NSAM 160, but Leminitzer did not take it any further.

As they had in their report, the Committee members remained concerned about

custody and control issues, but they were more optimistic about the situation than Pender. Commenting on the AEC-DOD survey of stockpile sites, in which a JCAE staffer was involved, Holifield saw “great improvement in the security of nuclear weapons assigned to the NATO forces.” One of the improvements was the “greater attention …to the selection and training of American custodial and maintenance personnel including consideration of emotional stability and security background checks.”

On 8 September 1962, members of the Joint Staff gave Bundy a briefing on NATO nuclear capabilities. Bundy asked to read the script, which Lemnitzer provided. The briefing included SACEUR and CINCEUR organization, roles, and missions, with details on U.S. and non-U.S. combat forces, delivery systems and nuclear weapons, but much of the detail on numbers of weapons and which countries would have access to them and at which bases in an emergency are excised. The only country names that are declassified are England and West Germany. Also excised are details of “threat lists” (principle targets) and the relationship between the SIOP and SACEUR.

The briefing concluded with a review of “problem areas,” including permissive action links. The Defense Department had submitted for President Kennedy’s approval a schedule for installing the devices, with work to be completed by the summer of 1964, at an estimated cost of 70 million dollars. Other topics were “defense data” presentations to top officials in NATO countries and “limitations on two-stage weapons.”

When Secretary of Defense McNamara and top aides visited NATO nuclear sites, Assistant Secretary of Defense for International Security Affairs Paul H. Nitze reported their disquieting findings. His account did not mention PALs but what McNamara’s party saw probably confirmed in their minds the importance of tightening up control of nuclear weapons that had been dispersed to allied forces.

A West German Honest John missile unit had a nuclear warhead and the two U.S. custodians who, according to Nitze’s deputy Henry Rowen, looked “rather lonely.” They “kept the secrets of their trade in what seemed to be a wooden safe.” Another problem was the Davy Crockett, about which “McNamara was told that it would take so long to get the order to fire that the screening force would be over-run before the Davy Crockett could be used.” On Mace missiles, which targeted airfields in Eastern Europe, Rowen concluded that they “were the most dangerous delivery systems now in Europe, both because they could be fired so readily and because their vulnerability would create great pressure to fire them in a period of tension or limited hostilities.”

A striking example of the lack of direct U.S. control that had troubled the JCAE were the nuclear-armed quick reaction alert aircraft operated by West German pilots. “The assumption that the German pilots do not know how to arm these warheads turns out to be fictional; on request, one of the pilots showed the US visitors how this was done.”

In this meeting of the North Atlantic Council, the participants returned to the question of nuclear weapons use consultations. In response to concerns expressed by allies, the U.S. had made a statement to the NAC in April 1961 designed to assure them that nuclear weapons “would be used under circumstances and rules endorsed by Council and that there is absolute certainty that if there is nuclear attack in Europe, US will respond with all weapons at its command.” NATO Secretary-General Dirk Stikker did not see that as the final word, however, and at a meeting on 9 March 1962 he presented the gist of a paper that would be officially presented at the May 1962 NATO meeting in Athens. It covered U.S. statements and assurances about the provision of nuclear weapons to the alliance, the delivery of nuclear information to NATO at all levels (individual members, regional committees, and as a whole), and guidelines for consultations on nuclear weapons use. Those topics would be the subject of continued discussion at the NAC before the Athens meeting.

During the discussion, Ambassador Thomas Finletter observed that the “use of nuclear weapons was [a] dreadful matter and we would want to use them only if need were very clear.” Nevertheless, he recognized that the European countries were “exposed to more immediate losses of their territory than other members of alliance,” although all alliance members were under the nuclear threat.

Noting Rusk’s statement that the U.S. should try to “preserve vital interests, including integrity of our forces and territorial interests without resort to nuclear weapons,” the German representative, probably Ambassador Wilhelm Grewe, argued that by the time that NATO realized that it could not withstand a conventional attack, it “would already be too late.” Territorial integrity would already have been jeopardized. For example, “Munich and Hamburg, situated close to the Iron Curtain would already be gone.” He proposed that nuclear weapons could be used in instances of “threats” to “territorial integrity or integrity of forces.”

Observing that the U.S. “intended not to take action when it was too late,” Finletter conceded that it was a “very difficult problem.” Grewe agreed that it was a “vital point and must be explored,” given the “exposed situation of countries such as Turkey and West Germany.” The Athens Guidelines would include the language suggested by the West Germans on integrity of territory and forces.

A briefing by State Department officials to members of the JCAE staff covered the substance of the NATO meeting in Athens. One of the topics was a “general guarantee” that the United States would provide NATO with ”adequate nuclear capability.” No numbers were discussed but there was another part of the assurance: that “Soviet targets threatening Europe would be adequately covered.” This was an implicit reference to U.S. plans for SIOP targeting of Soviet intermediate and medium-range ballistic missiles that were deployed in the Western U.S.S.R.

Another assurance was that Washington would provide NATO with more nuclear information so that allies “know more about our nuclear arsenal and planning,” as a contribution to their “education.” While the State Department briefs suggested that better knowledge of nuclear weapons could help persuade the allies to undertake a buildup of conventional forces, James Ramsy of the Committee staff was doubtful that adding to the NATO stockpile would “encourage the conventional buildup.” In reply, Russell Fessenden observed that by providing the allies with “general information on numbers of weapons” that would be available, they would realize how “complete” NATO’s nuclear defense was, which would “reinforce our case for the conventional buildup.”

The guidelines on nuclear weapons use was the final element of the “package” agreed to at Athens. According to Fessenden, they spelled out “the circumstances under which nuclear weapons might be used and the rules for consultation.” Further, they served a “dual purpose of assuring those countries which fear there might be an unwarranted delay on use,” speaking to the concerns raised by the West Germans, “and those which fear that they would not be consulted.”

Whether Fessenden quoted from the guidelines or not, they included these central points: 1) an “unmistakable” nuclear attack by the Soviets would be answered with a proportionate nuclear response, although opportunities for consultation would be “extremely limited,” 2) a Soviet “full-scale conventional attack” signaling the “opening of general hostilities,” would be met with a nuclear response “if necessary,” but time for consultations would be available, and 3) in a smaller-scale Soviet attack that “threatened the integrity of the forces and the territory attacked,” but which could not be repelled with “existing conventional forces,” the NATO Council could decide whether to use nuclear weapons. The last set of circumstances related to a contingency raised by the West Germans, although a quick decision would need to be reached in the event of the dire scenario (seizure of Hamburg or Munich) that had been described during the March 1962 NAC meeting.

In addition, as Fessenden noted, the guidelines included a U.S. undertaking to consult with the NAC “on the use of nuclear weapons anywhere in the world if time permitted.” In general comments on the guidelines, Fessenden observed that they struck a “good balance between our readiness to consult and our retaining freedom of action to respond quickly in an emergency.” Because NATO enabled members to respond collectively or individually to defense themselves, the United States “must always be a party to the decision to use nuclear weapons.”

After the Athens meeting, U.S. defense officials pondered how much nuclear weapons information they would share with NATO allies. With Deputy Secretary Gilpatric slated to meet West German Defense Minister Franz-Joseph Strauss, the chief of the Military Assistance Advisory Group in Bonn anticipated that Strauss would ask about the “availability of nuclear weapons held by US for FRG.” Sending the first cut of a briefing paper for Gilpatric’s use, the chief’s approach was to tell Strauss how he could figure out an approximate number for himself, not exactly in the spirit of the Athens Guidelines. He began with the point that at the Athens NATO meeting Secretary of Defense McNamara had told the allies that the U.S. had deployed 5,000 tactical nuclear weapons to Western Europe.[9] The MAAG chief observed that in most circumstances, the “division of existing nuclear devices would be in proportion to the delivery weapons available,” but it was not so cut and dried that there would be a nuclear weapon for each delivery system: “effective utilization will call for employment in relation to the threat.” Therefore, FRG forces would have “warhead requirements in relation to their percentage of the total of delivery weapons available in Europe” and “would receive at least a like proportion of the related nuclear devices available.”

Recognizing that other U.S. officials might have provided the West Germans with information on the stockpile, the MAAG chief asked for any particulars to determine whether it would be possible to be “more forthcoming” in the talks with Strauss. What Gilpatric actually said to Strauss about the stockpile, if anything, probably remains classified.