6The Alerts Agreement in Action
The Tripartite Intelligence Alerts Agreement had been agreed with only a vague sense of how it might work in practice. The most pressing issue after the agreement was finalized was to set up a communications system by which the three governments could, in fact, alert each other to indications of war. But substantive questions about the whole indications program were still being posed throughout the rest of 1957.
For instance, both the Canadians and British had received and read a copy of US NIE 11-3-57, “Probable Intelligence Warning of Soviet Attack on the US.” The document raised eyebrows in Ottawa and London. It revealed that the Americans understood their own military and intelligence activities could trigger Soviet reactions that would look like indications of war.
In a vaguely phrased paragraph, the NIE noted that, in a crisis situation, the volume of intelligence reports could be expected to increase dramatically. Furthermore, in crises there “is also an increase in the number of reports from sources of known reliability, some of which sources come into play as a result of a crisis situation.” This, the NIE explained, might include “[p]hotographic and electronic reconnaissance over Soviet controlled territory.” It might also involve “[a]gents held in reserve for such a situation and equipped with special means of communications [that] could be activated.”1
The reconnaissance was clearly a reference to overflights by the Americans’ new high-altitude aircraft, the U-2. Increased activities of this type, and “in particular air penetrations, could have the effect of increasing tensions or even of provoking Soviet attack.” Given the possible upshot of these actions, they would “probably require policy decisions.”2
The director of Canada’s JIB noted that the measures referred to in the NIE “would be subject to very strict security protection, both because of the nature of the job they are intended to perform and because of the political sensitivity surrounding their use.” As a result, the information would be “controlled by code words and . . . a very limited number of specially indoctrinated personnel would be involved.”3 These would be among the closest guarded secrets in the American intelligence community.
While Bowen had no doubt about the “willingness” of the Americans to “discuss such things in times of great crisis,” he did question “the functioning of the machinery at such times if entirely new decisions have to be taken.” As a result, Bowen suggested that the chair of the JIC work with Allen Dulles to ensure that “the staffs concerned will have the necessary authority to transmit to us information from these sources during times when the risks of war appear to be very great.”4
Crean agreed and went even further, believing that “it would be most desirable if we were told of such operations prior to their taking place, especially since they might be of a provocative nature.”5 Here the Canadian concerns echoed the point in NIE 11-3-57 about the effect of US actions on the Soviet Union and the possibility of American operations resulting in Soviet responses that could be interpreted as indicators of preparation for war. The Canadians would pay close attention to this issue going forward, seeking to ensure that they were aware of the activities of “friendly forces,” especially Strategic Air Command.6
Communications Issues
While the Tripartite Intelligence Alerts Agreement was completed in 1957, the three parties still needed a communications system by which they could meet their obligations under the agreement to exchange information. And although the idea for such an agreement had come at the behest of the British and the Canadians, the Americans were always one step ahead of their allies in implementing the agreement.
In October 1957, Dean and the British were “slightly embarrassed” when an American team showed up in London ready for technical discussions regarding the communications network required to operate the alert system. The British were simply not ready. Neither were the Canadians. The Canadians were invited to attend the discussion with the visiting Americans but had no one in London prepared to attend. Nor could they get someone from Ottawa to London in time.7
Despite being caught unready in October, one month later Dean was eager to gain quick agreement from all three parties on cipher equipment. He wrote a letter to Dulles and Crean suggesting that the primary link between the three parties be a recently laid transatlantic cable. The cable connection could be backed-up by a radio system. In normal times, Dean proposed, this cable channel would be used for ordinary diplomatic and intelligence telegrams, but in a crisis the system could be equipped with a “special switching device, with alarm facilities” that would clear the line for “exclusive Alert use” upon activation by Washington, Ottawa or London.8
Crean did not like this idea. It meant that any communication lines that were used for the transmission of information in regular (that is, non-crisis) time, would be taken over by the alert system in an emergency. This was unacceptable because the cable already carried “highly important intelligence which had not yet been dealt with by directorates.”9 Switching to exclusive use for intelligence alerts would, at the same time, cut down the overall flow of intelligence.
For the same reason, the Canadians also resisted using the existent Hydra system (although Hydra would, ultimately, be the main conduit for the TIAA communications system). “Hydra” was the name for the transatlantic communication system that had been built at Camp X in Oshawa, Ontario, in 1942 as a training facility for British and American covert agents. It had served as the main communications hub for British Security Coordination during the war and was the centre through which Ultra intelligence was shared between the three governments.10 In the postwar world Hydra had been maintained as an important tripartite communications link. The Canadian JIC, Crean pointed out, remained “reliant on Hydra to obtain intelligence items rapidly from overseas.” The Communications Branch of the National Research Council, Canada’s signals intelligence agency (later Communications Security Establishment, CSE) used Hydra, too. If, in a crisis, Hydra were “given over” to Alert messages, “the individual items of intelligence which may be equally vital to us will not be able to pass if there is a radio blackout.”11 The Canadians preferred a system that would not be online all the time, for it would either lay idle or have to be used for something else.
In February 1958, under pressure from Canada’s partners to agree to the British communications plan, Crean wrote to Dean and Dulles describing a set of problems “which I suspect arise only in Ottawa” and as a result “do not appear to have been taken into consideration when formulating your proposal.”12 The Canadian concern, ultimately, was that if any of the existing systems for transatlantic communications went “online” as the channel for alert communications, it could not be used for other purposes and would then limit the number of communications channels on which Ottawa received information.
Crean listed the four main Canadian requirements for transatlantic communications: They were, first, the “rapid and continuous passage of raw and finished codeword intelligence” between GCHQ and CBRNC. Second, the rapid exchange of assessments between JICs via the JIC liaison officers in Ottawa and London. Third, the rapid exchange of diplomatic reports from the High Commissioner in London back to Canada. Fourth, and new, was the rapid exchange of assessments on tripartite basis.13
To meet all these requirements, Canada had a tape relay centre at the Ottawa Wireless Station in Leitrim (the Canadians referred to the station simply as “Leitrim”) equipped with a number of radio channels to the UK. The Canadians and British had also recently added a duplex cable circuit which strengthened the Hydra network. Canada used this system to exchange information with both the NSA and GCHQ, and to exchange British-Canadian diplomatic traffic. Canada had no other transatlantic circuits to achieve its four requirements (while the British and Americans did — they had a direct cable link between GCHQ and NSA).
If the transatlantic link was to go “on-line” for alert traffic, then GCHQ-CBNRC traffic, as well as communications between the Department of External Affairs and Canada House in London would be limited to exchange over radio, which might not work in a radio blackout. This was a drawback because the Canadian JIC relied on these exchanges for its intelligence. While Washington and London would still be able to exchange information via their transatlantic cable facilities, it would effectively leave Ottawa in the dark. No doubt the “scheme you propose looks quite satisfactory to you” in London or Washington, Crean appealed, but it would “leave us in a very awkward, and I believe unacceptable position.”14
Crean urged his counterparts to “to take another hard look” at his suggestion of an “off-line system” that used the Canadian-developed Rockex cipher machine.15 Alerts traffic would bear a “special top priority designation” that would see it passed ahead of all other traffic on the Canada-UK cable. The Rockex system would be slower, but everyone who used these systems knew that the biggest temporal challenge was not transmission of messages but assembling the necessary officials to meet in each capital after a message had been received. (The Canadians also pushed for Rockex instead of the American Sigtot equipment because Rockex was available in Canada and Canadians would not need spare parts.)16
Despite some grumbling in London, Dean said that the British were likely to accept the idea of off-line ciphers, and Tracey Barnes, the CIA representative in London, said the US was likely to accept too.17 Dean responded formally to say that he agreed with the Canadian suggestion and that, he “on balance agree[s] with you that the delay of a few minutes in passing telegrams by the off-line system is acceptable.”18
Crean, on the advice of Drake at CBNRC, also concluded that an off-line connection between CIA and Leitrim was the best answer for the US-Canadian link.19 Lieutenant-Colonel Paul E. Amyot, the deputy director of signals in the Canadian Army, suggested that there be established a tie-in line from Leitrim to the JIR, and avoid working through CBNRC. A line terminating at Leitrim would also make it easier to resume communications “should circumstances suddenly force a change of venue upon J.I.C.”20
In 1958, the “so-called direct line” to Ottawa from the CIA actually travelled via NSA to CBNRC and then on to the Joint Indications Room, requiring re-encipherment at both ends.21 The Canadians and Americans met to discuss the establishment of a truly direct line from CIA to Leitrim. At the meeting, the Americans noted such a line would cost US$1,000 a month. They did not want cost to “hamper or delay” the establishment of an effective system, but they asked the Canadians to consider two separate options: A line from the US embassy in Washington to the JIR or Leitrim, or a line from CIA to the Canadian embassy in Washington. The Canadians pressed for their original plan — the direct line from CIA to Leitrim. While they did not let on to the Americans, they privately concluded that the higher cost was “well justified since in the long term we could look forward to a fair volume of current intelligence passing from the CIA direct to Ottawa.”22
The Canadians also worried that the staff at both embassies was too small to handle the traffic, and the “trouble with any tie-in-and-switch arrangement is that in a crisis somebody might forget to turn the switch on.”23 On top of the risk of such human error, the embassy’s intelligence and diplomatic traffic would have to compete, and working through the US embassy would also give the American embassy “control of the line” which “in certain circumstances” might be undesirable. “Economy at this stage would be false economy” and so if the Americans quibbled about the cost, the Canadians in Washington were to say that Canada would consider sharing the cost of the line, perhaps by paying for the portion of the line running from Leitrim to the border.24
In April 1958, representatives of all three parties met in Washington in an attempt to finally resolve all communication matters, including message procedures, tape procedures and supplies, cryptographic systems, and other requirements. The Americans agreed to use the standard operating procedures already employed by the British and Canadians, and a regular type of message heading and classification, with the goal of ensuring faster relay within capitals.
The allies agreed that an “ALE” prefix atop a message would serve as the symbol that the message was calling for, or responding to, an alert. (ALE, of course, being the first three letters of “alert”.) Each state would have its own call-sign (LON and OTT for London and Ottawa, respectively, and WAS would be used by the Americans from either Washington, DC, or their alternate location outside the capital) indicating who had sent the message (i.e. ALE-LON, ALE-OTT, ALE-WAS for alert messages from London, Ottawa, or Washington, respectively). They would later establish a pattern whereby non-Alert messages could be sent, with a different prefix, such as JICOTT, JICLON, and CIAWAS. A message with prefix JICOTT, then, was a message from the Canadian Joint Intelligence Committee but of lesser importance than an ALE message.
The Canadians and British already had Rockex equipment, and the British would install three Rockex systems at each of the US sites and provide maintenance until US technicians, trained in Ottawa, were ready to staff their own systems.25
While the British agreed that these technical discussions must come first, they knew that next would come a “common doctrine for bringing into use of the system, its operational use, and so on,” and began working on those issues.26 After the communications issues were largely settled, plans were made for a tripartite meeting in June 1958. A month before, in May, CIA representatives met with Canadian and British JICLOs to discuss the implications of the alert system for the liaison officers, to consider “joint indications lists and a common philosophy of indications intelligence” and plans for a test of the communications system.27 They achieved little and much was deferred until the June meeting.
How and When to Invoke the Agreement
The Canadians learned that Dean was under pressure from his government to achieve substantial progress at the June meeting.28 The pressure, undoubtedly, stemmed from ongoing negotiations between Washington and London regarding an agreement on nuclear retaliation procedures that would govern the launch of US weapons in the UK.
In April 1958, Dean travelled to Washington and met with Robert Murphy to study “how procedures of the two Governments might [be] concerted for reaching a decision to respond to a Soviet attack by committing nuclear retaliatory forces to the attack from the United Kingdom.”29 In annex B of the agreement listing “Procedures Preceding Attack by United States Retaliatory Forces from the United Kingdom,” the US side noted that on “receipt by NIC” of strategic warning indicating “an enemy is likely to launch an attack . . . the intelligence information and the evaluation thereof will have been passed to the Joint Intelligence Committee (London) and the Joint Intelligence Committee (Ottawa) pursuant to the Tripartite Alert procedure agreed to among the Governments of the United Kingdom, Canada and the United States.”30 The TIAA and nuclear retaliation procedures were flip sides of the same coin. The British, in their own internal communications in 1959, described the “Tripartite Alerts Agreement” as one of the agreements “which relate to the mechanics and procedures for using nuclear weapons.”31
Given the practical and driving need to establish a system to support the Murphy-Dean agreement, the Canadians let Dean’s people take the lead in drawing up a working paper that they would respond to, rather than drawing up their own.32 Thus the British set the pace for the first major discussion of the tripartite system.
The British distributed their paper for use of the “Tripartite Alerts Communication System” in May. The paper suggested the system be used “at all times” for transmitting intelligence information between the three parties, including the British Red Book (a weekly intelligence survey), CIA comments on the Red Book, the BJSM weekly telegram to the UK minister of Defence, the US Watch Committee report, and so on. Dean expected this would keep the channel in “good working order” and the “operators practiced.”33 As for when the system was to be “brought into use for the main purpose”34 — exchange of intelligence regarding a threat to the NATO area — the British made a series of suggestions: If any capital changed its own state of alert, and if this change was connected with “indications of Soviet bloc warlike intention,” they should send an ALE message to its tripartite partners. Also, if what the British called their “special heads of section” called for a meeting to consider intelligence related to the NATO area, others would be alerted of the meeting and its subject.35 A first message informing the partners should be sent and then followed, as soon as possible, with a more detailed description of intelligence available and “asking for information or opinions.”36
At the June meeting, Amory of the CIA thought the British plans for passing so much information over the alert communications system was “excessive.” He thought it should be restricted only to the US Watch Report and the Canadian and UK equivalents. Any discussion or comments could pass through normal channels.37
The meeting also demonstrated the inconsistencies in just what each party meant by the word “alert.” The British paper had simply referred to “alerts,” but the US officials pointed out they must have meant “intelligence alerts” to differentiate from military or operational alerts. The three states did not have the same system of intelligence alerts, and the Canadians and Americans did not have well-defined alert stages. They would both seek to conform to the British system as far as possible.38 It was left unsettled, but the plan was to work for a “common nomenclature for stages of alert . . . to ensure that each party knows, for each stage of alert, exactly what it implies for the other two in terms of organisation and state of readiness.”39
Amory proposed a Command Post Exercise–type test of the system in the fall to test delivery of “an agreed intelligence assessment to the policy levels of government.” He suggested preparing fake intelligence.40 The British and Canadians both tabled indicator lists (the American list having been recently approved by the IAC) that would form the basis of a common indicator list.41
To move forward, however, all the allies would need to have functioning indications centres. Here, the British did not have their act together. The Canadians were much further ahead, and of course the Americans already had a well-established centre. Before considering major debates over how to use the system that occurred in 1958, the next section examines the state of Canadian and British Joint Indications Rooms that year.
Building a Canadian Indications Room, Part 2
The Canadian Joint Indications Room and communications systems in Ottawa all experienced growing pains in 1958, but with the pain came growth. After the tripartite communications meeting in April, the Canadians created new lines of communication and procedure within Ottawa, all designed to ensure the connection of the JIR to all necessary facilities.42
From April 1958 on, the JIC was hard at work establishing a new set of procedures, entitled “Indications Intelligence: Communications and Watch Procedures” (variously altered, and a sign of the adjustments made, as JIC 278(58); JIC 278/1(58); JIC 278/2(58); JIC 278/3(58)).43 A study of the speed by which messages were received in Ottawa from abroad indicated some serious delays in intelligence traffic: EMERGENCY messages from Europe and the Far East were sent from the “originating outstations” in fifteen or twenty minutes but then, once received, took a “rather incongruous” two hours “to progress the 100 feet” to the Joint Indications Room. One analysis concluded, perhaps obviously, that “as the indications net develops tripartite-wise the problem of maintaining speedy and efficient communications to facilitate discussion will be of utmost importance.”44 Staffing issues were considered, and the twenty-four-hour watch that had once been in place seems to have been cancelled and then resumed.
The British, however, were having even more trouble. Pat Black, the Canadian JICLO in London (JICLO(L)), wrote that “there is still a good deal of vagueness in L[on]d[o]n as to how a JIR should be run and the merits of having it manned by trained personnel on a twenty-four-hour basis.” Part of the problem was no doubt the reorganization of the London JIC’s relationship with the Cabinet Office and Chiefs of Staff Committee.45 Because of a crisis in the Middle East in the summer of 1958 (see below), the British did set up an ad hoc Joint Indications Room that could handle tripartite alerts communications. But once the ad hoc system was closed after the crisis, the British would not be ready to participate in an ongoing tripartite alert system. 46
London aimed for a start of October 1, 1958, but was unlikely to meet it due to administrative difficulties, lack of personnel, and because the terminal equipment for the new JIR had not been established.47 At one point it seemed like it might not be ready until 1959 but was in fact working by November 1958 (see below).48 In addition to the real administrative problems in Whitehall, some in London thought the whole idea of alerts was “just another American ‘fad.’”49 They were perhaps oblivious to the origins of the idea, in which the British had pushed so hard.
It may seem peculiar that the tripartite powers were seeking a system of indications intelligence at a time when the development of Soviet intercontinental ballistic missiles (ICBM) seemed to further limit the warning time that would be available ahead of a war. In spring 1958, Ivor Bowen wrote a lengthy paper called “The Continuing Need for Indications Intelligence.”50 He pointed out that while indications intelligence could hardly guarantee warning of a Soviet ICBM attack, “this lack of certainty is nothing new.”51 Indications intelligence had never been able to guarantee warning against manned bombers, either. And still, proponents of indications intelligence had recognized that launching a major war was such a complex and difficult process that there were many opportunities to compromise surprise.
Even more important, and almost contradictorily, while the Soviet ability to launch ICBMs was coming nearer, global war seemed less likely than it had at the start of the decade. A “global war is at present conceivable,” Bowen wrote, “as a result of an accident or of a miscalculation by either side, involving an initial attack mounted in great haste.” Hasty preparations would compromise operational security, perhaps resulting in more indications of war. And thus, even in the missile age, there were many scenarios in which a continued indications intelligence effort was necessary and would be so even after 1965, or whenever the Soviets would choose to rely only on ballistic missiles. While it was thought to be no longer possible to wage war early enough to prevent atomic weapons reaching North America, indications intelligence could still “give invaluable warning” and help North American authorities make “survival” decisions.52
The Canadians Want to Go Global
Still, despite the obvious importance of the system to all parties, there remained significant work to be done, and confusion to alleviate, after the June 1958 tripartite meeting. The Canadians left the meeting unclear on three things: the procedures for using the system, the agreed indicator list, and the stages of alert. The JIC (Ottawa) had, at first, “no stages of intelligence alerts and no indicator list.”53
The matter of coordinating intelligence alert stages was unnecessarily complicated. The Canadians adapted the British system of alerts (pre-alert, stage 1, stage 2), while the British simultaneously adopted the new Canadian system of alerts (stages 1, 2, and 3).54 The two sides finally agreed to both revert to the original British suggestion, beginning with a “pre-alert” stage, followed by stage 2, and then the most serious alert being stage 1. They then moved in 1959 back to the original Canadian system of stages 1, 2, and 3. There were also continued efforts, both nationally and on a tripartite basis, to establish a joint indications list.55 But just when to invoke the system remained a crucial sticking point.
The Canadians spent the summer of 1958 working on a major paper, JIC 1103/1(58), “Tripartite Alerts System,” to guide discussion of the use of the new system. It was to be presented at the next major tripartite conference scheduled for August.56
A draft of JIC 1103 makes it clear that Canadians were seeking a nearly global application of the tripartite alerts system. The documents warned that “the likelihood of hostilities in the Middle or even the Far East should be considered as immediately threatening the NATO area,” because crises in these areas “through a chain of events” could develop into a major NATO crisis “in days or even hours.” And not only did the Canadians want to expand the geography of the agreement, but also for the alert system to cover a range of types of potential conflict, from “one where our armed forces go in with guns blazing, as in Korea” to a non-contested peacekeeping operation like that of the United Nations Emergency Force (UNEF) at the time of the Suez Crisis. On both issues the Canadians were influenced by their “experience in indications matters” during the summer of 1958 when events in both Lebanon and Jordan had escalated into major crises (see below).57
Another section of the draft paper also sought to ensure fast and total communication of all indications intelligence gathered around the world. In essence, the Canadians argued that the “agreed tripartite indicators list” must not only be used in the NIC and JIR, but “also at all United States, United Kingdom, and Canadian intelligence collections points around the world.” The Canadians suggested that an “agreed codeword, such as MAYHEM,” be used and that all MAYHEM signals “will be recognized by all as dealing with tripartite indicators wherever they may originate. 58 There is a hint here of the early Canadian thinking that any indicators recorded on an index card in Ottawa would be copied to the United States. (Later, all three states would push their indicator lists to field collectors, especially attachés, but the indicators collected were never transmitted through the Tripartite Alert System in raw form.)
When the Canadians showed Amory the substance of JIC 1103/1(58) he was clearly alarmed. The proposals in the paper, he said, “go considerably beyond what I believe to be the understanding of the IAC as to the intent of the basic Tripartite Alerts Agreement.” As a result, he (and the CIA) would not discuss the paper unless it was taken first to the IAC. But he warned of his own “personal belief” that chances for the paper at the IAC were “dubious at best” and “would almost certainly confuse and delay effective implementation” of the tripartite alert system, “which is already long overdue.” Seeking to soften the blow, Amory told the Canadians there would likely be “ample opportunity for negotiations to improve the system” later, so maybe they could raise the issue then, without running the risk of holding up implementation.59
The British Embassy officials concerned with intelligence learned of the Canadian paper but had not seen it. They requested guidance from the JIC (London), and so JIC (London) asked the Canadians for a copy of the paper from JIC (Ottawa). There is no record of the Canadians passing the paper to London, though they likely did. JIC (Ottawa) concluded that the British would be “influenced by Mr. Amory’s reaction” and assumed the British would support Amory’s suggestions that “we delay formal introduction of our more far-reaching proposals until the tripartite alerts system was a going concern within the present context.” The JIC instructed the JICLO(W) to withdraw major portions of the document from consideration at the upcoming tripartite meeting.60
Hamilton Southam, who had replaced Crean as chair of the JIC, made the decision to withdraw. He recognized that the “new wine of our thinking was too strong for the old bottle of the working group.” Yet, he still clung to Amory’s suggestion that the issues could be taken up later. “We are all convinced here [in Ottawa],” he wrote, that “we know what we are doing, and that eventually we shall be able to rally our American and British friends to our views.”61
Setting the Indicator List
The major tripartite meeting scheduled for August was delayed until October 1958, perhaps because of events in the Middle East. But the results of the October meeting, a “Report of Tripartite Working Group on Tripartite Alert System,” is a crucial document that set up the key elements of the system. The report included an annex with an agreed list of “Critical Sino-Soviet Bloc Actions.”62 This was a shortened list of crucial indicators. The list, based on an IAC paper with input from JIC (Ottawa) and JIC (London), would serve as the bedrock indicators list for the system and for indications analysts in all three capitals.
The thirty-five indicators identified as “Critical Sino-Soviet Bloc Actions” ranged under nine headings: General, Missiles, Air Forces, Submarines, Naval and Merchant Ships, Ground Forces, Air and Civil Defense, and Logistics. Some indicators were quite straightforward, including “Sino-Soviet Bloc declarations of war or acts of war against the US, UK, Canada or any other NATO member or forces abroad.” Others were more nuanced, including the initiation of security procedures, evidence that attack orders were being passed to submarines, or the arrival of specialist units — “especially interrogation and medical units”63 — in forward areas. For the most part, the indicators were related to fairly obvious preparations, movements, and reinforcements of military units.
The group recommended the list be reviewed semi-annually to reflect new “awareness of changing Sino-Soviet capabilities and operating methods,” especially as more information about missiles became available.64 Going forward, there would be meetings to review the indicator list every March when UK intelligence officials travelled to Washington for an annual Standing Group meeting that occurred at the same time.65 Later on, the semi-annual meeting occurred less regularly.
The report also contained a chart summarizing the operating procedures of the indications centres in each country. Contrary to both British and Canadian ideas that the system should be used to exchange a whole range of intelligence, the tripartite meeting agreed that the “tripower alert system is to be limited to the transmission of critical intelligence and whatever additional information may be necessary to make proper assessment of the critical intelligence.” Critical intelligence was defined as “information indicating a situation or pertaining to a situation threatening the NATO area which affects the security interest of the US, UK, or Canada to such an extent that it may require the immediate attention of the heads of the three governments.”66
After the important October meeting, there followed a series of smaller and more specialized meetings to discuss watch procedures and clarify prefixes and serial numbers of messages. The British pushed to use the network to share their weekly indications report, ostensibly to keep the system in good working order.67 On New Year’s Eve the British proposed the code word “DRUMSTICK” to give protection to the agreement. The Canadians agreed, and the US agreed in principle, although they needed to refer to their own list of code words to ensure DRUMSTICK was available.68 Why the British suggested DRUMSTICK is unclear. Perhaps the beat of indications intelligence accompanies the march to war. The code word seems not to have been used going forward.
The Crises of 1958
The years immediately following the establishment of the Tripartite Intelligence Alerts Agreement and its communications network were studded with crises and war scares. Given the delays in London mentioned above, it appears that ALE messages were not sent until late 1958. The crises in the Middle East in the summer of 1958 and the Taiwan Straits in 1958, then, came after the signing of the agreement but before the communications system and procedures were fully established. The crises in Berlin that came and went from 1958 and into the early 1960s, were, however, the subject of ALE messages.
On July 14, 1958, the government of Iraq was toppled in a coup. Lebanon was already in the midst of a civil war, and the Lebanese leader, fearing a similar threat to his regime, asked the Americans to intervene. On July 15, the United States landed marines in Lebanon. The Jordanian king made a request for similar assistance, and two days later, British troops landed in Jordan.
The crises in Iraq, Lebanon, and Jordan appeared to be localized. And yet, behind the scenes, the United States took steps with potentially global implications. On July 15, the United States Strategic Air Command was placed on “improved readiness” to act in case the Soviets intervened, or, perhaps more accurately, as a threat to ensure the Soviets did not intervene. In support of the SAC alert, North American Air Defense Command (NORAD) forces were placed on alert, too. Although NORAD was a bilateral US-Canadian command, the United States did not consult with Ottawa.69
On July 15, the Canadian JIC enabled the Joint Indication Room’s twenty-four-hour watch. The JIC also requested a Daily Indications Report be prepared, and the report was to include “the movement of friendly forces.” There is some evidence that messages dealing with indications intelligence in this period were delayed.70
The rapid start to these July crises, then, seemed to confirm for the Canadians that crises outside of the NATO area had the potential to escalate into a great power war. It also emphasized the importance of establishing a swift and reliable system for exchanging information with both Washington and London.
Throughout the crisis, the JICLO(L), Pat Black, liaised regularly with his British counterparts. He also held regular meetings with the Canadian DAI, DMI, DNI and JIB liaison officers in London to make sure that any information of importance had been passed to Ottawa.71 In Washington, the USIB set up an Ad Hoc Working Group for this crisis, as well as for crises in the Taiwan Straits and Berlin. In all cases, Philip Uren, the JICLO(W) had been informed of the work of these groups.72
The First ALE Message
By late November 1958, the TIAA communications system was working. In the first month of the Tripartite Alert Communications System, there were “nil” ALE messages and twelve “routine” messages (JICLON, JICOTT, and CIAWAS), exchanging weekly indications reports and discussing the functioning of the system.73
It appears that the first ALE message was sent in 1958, by the Canadians. On December 11, 1958, as a result of “an apparently deteriorating situation in the Middle East,” the JIC (Ottawa) called an Intelligence Alert Stage 1. Four days later, on December 15, JIC (Ottawa) dispatched the first ALE-OTT message to London and Washington asking for an assessment.74 The alert set off a series of debates and discussion, not over whether war was imminent, but whether the planned procedures worked and whether the Canadians had been right to call an alert in the first place.
The day after the alert, the United States Intelligence Board (which had replaced the IAC) met and held “considerable discussion” as to the “correctness of [the Canadian] use of the alerts system.” Sheldon of the CIA thought it was fine, while Cabell was neutral but furious that the United States had not been able to respond as quickly as it had promised. In a follow-up discussion between the Canadian and British JICLOs in Washington, Uren and Paul Jones, and a CIA official from Sheldon’s office, Jones thought the use was incorrect because a crisis in the Middle East did not pose a “direct threat to the NATO area.” Uren, for his part, argued the Canadian case that this interpretation was “legalistic and would hamper the effective use of the system.”75 He pointed out that critical situations in the Middle East had previously been used to increase the state of readiness in both NATO forces and NORAD. Uren’s point, clearly, was that everyone recognized events in the Middle East might “constitute the beginning of a chain of events” that could threaten NATO.76
Both the JIC (London) and its representative in Washington insisted that the ALE prefix use by JIC (Ottawa) was “incorrect.” “This prefix should not be used by any of the national authorities for an alert outside the NTO [sic] area and that if this is done it extends the tripartite alerts agreement beyond the area agreed by the three governments concerned.” If the Canadians wished to call an alert in Ottawa, and inform the other two states via a JICOTT message they were doing so, that was their prerogative.77
The Canadians disagreed with the British, complaining that the UK authorities were “placing an unnecessarily narrow interpretation on the agreement.” Did the British really think a war between the United Arab Republic (UAR) and Israel would not affect the NATO area? “This seems to us to reveal not only shortsightedness but a rather short memory, bearing in mind the Soviet notes at the time of the Suez Crisis,” the Canadians concluded.78
The Canadian told their allies that Ottawa’s use of the alerts system was “warranted” and suggested that the Alerts System allow for two stages of alert for various ranges of alert.79 The USIB rejected this idea, because it “regards the use of the circuit itself as an indication of a need for high level concern.”80 If the message was “not sufficiently important to involve the Chairman of the USA Watch and to invoke the Agreement, then it would be the opinion of the USIB that it did not [repeat] not warrant the use of the circuit.” The Americans were concerned about the system being filled with “working level traffic” that would “degrade its value as an alert mechanism.”81
The Canadians rejected this “overly mechanistic” understanding of the system “as a ‘mechanism’” and the “suggestion that the circuit itself is important, rather than the messages passed on it.”82 The tension remained, then, between the function of the alert system as a part of British and American nuclear release procedures, and a Cold War in which both analysts and policymakers had long seen a connection between non-European crises as possible precursors to superpower confrontation. But minds can change. In a matter of months, the British would go from being critical of the Canadian view to championing it.
Ahead of the March 1959 review of indicator lists in Washington, the JIC (London) requested a special meeting of the tripartite alerts working group. The British representative, Antony Duff of the Foreign Office (representing the UK Joint Intelligence Staff), planned to present “views on the use of the tripartite intelligence alerts system for the exchange of intelligence on marginal or developing situations which do not present an ‘immediate threat to the NATO area.’” The British brief was essentially a restatement of the Canadian position that consultations on non-NATO areas, including the Middle East, were warranted.83 The precise reasons for this British about-face are unclear, but it is indicative of a return to early British thinking that had envisioned the ultimate expansion of the agreement (see Chapter 5).
When the British presented their proposals in March 1959, it was obvious to the attendees they were “primarily concerned with providing for consultation on critical areas outside of the NATO area but not repeat not with modifying the levels at which consultation might take place.” The British were now pressing for an optional system whereby any of the three states could “ask for the views of the others on a critical situation anywhere in the world.”84 The parties would not use ALE, and would not formally invoke the agreement, but they would use the alerts system to communicate.
The hypothetical offered by the British and recorded by the Canadians was “a critical situation in the Middle East in which the UK might contemplate some form of military action or intervention and might wish to have the views of the US (and by implication, not repeat not, by assertion, those of Canada).” The Canadians said little in the meeting since they agreed with the British position. The Americans, however, gave the idea a “very negative reception,” with Sheldon explaining, at length, the difficulty of responding to such requests. The “US could only respond to a request on the system by giving a fully coordinated community view, with all the staff work which that implied.” Again, the size and difficulty of managing the US bureaucracy posed a challenge to coordination with others. Despite the US opposition, Sheldon admitted there was “some merit” to the proposal and suggested Dean write to Dulles and John Starnes (who had replaced Southam as chair of the JIC (Ottawa)).85
Starnes, who did not attend this meeting but met with Amory a week later, had learned the Americans “probably could be persuaded to extend the tripartite alerts agreement to geographical areas outside the NATO area.” They would be reluctant to automatically apply the alerts systems to crises over the Quemoy Islands in the Taiwan Straits, but this could be circumvented if all accepted the principle that “each national authority is free to decide when a developing situation, whether or not it is strictly in the NATO area, warrants the calling of an intelligence alert.”86 Starnes wrote to Dean on March 21 to suggest that the British follow up on their proposals made in the March meeting with a letter to Dulles and Starnes himself as chair of the JIC (Ottawa).
Only the day before, however, Dulles himself raised the matter in a meeting between Dean and Dulles in Washington before Starnes’ message reached Dean. Dulles’ proposal was to formally amend the agreement itself “to provide for it coming into force automatically in the event of any Sino-Soviet bloc aggressive action” whether affecting the NATO area or not.87
While this was a “step forward,” Dean told Starnes that it does not “altogether meet either our wishes or yours.”88 The idea pleased Starnes, who recognized the agreement would then come into effect for situations like the Formosa Straits, which, during a recent crisis, the US had “tended to regard as their business alone.” Still, Starnes did not like the limitation of attention to the Sino-Soviet bloc, for the “the actions of other states can be equally if not more dangerous. What if Iraq were to take some sudden drastic step such as the seizure of Kuwait?” Starnes admitted this was “perhaps an exaggerated example, but serves to illustrate the point.”89
Following his conversation with Dulles, Dean wrote him to formally propose amending the agreement and gave examples such as a threatened Soviet move against Iran, the threat of a Viet Minh invasion of South Vietnam, or “serious recrudescence of Communist aggression in the Taiwan Straits.” Dean added to the letter “another thought,” that of making more use of the “special communications channel which has been established to serve the Tripartite Alerts Agreement.”90 This, Dean told the Canadians, was something they were anxious to achieve.91
The Canadian JIC considered the possible amendment. Bowen thought the idea “should be strongly supported by the JIC.” Still, there were very few aggressive actions taken by any states outside the Sino-Soviet bloc that would pose a serious threat to Canada. The one Bowen could imagine was “Arab action against Israel, which would involve Canadian forces around the perimeter of Israel.” So while Canada would welcome the British effort, JIC (Ottawa) would not “have a case to render strong support.”92
Some members of JIC (Ottawa), and especially Bowen, wanted Starnes to push the Americans on another issue: those withdrawn passages from JIC 1103/1(58), especially the parts calling for “for the automatic and expeditious passage of all relevant intelligence information,” including from field posts.93 Amory, however, again poured cold water on the issue, and Starnes did not bring it up with Dulles at their next meeting.94
Allen Dulles wrote to Dean to explain that the USIB agreed to an amendment to broaden the terms of the agreement. However, he stressed the American position that “traffic passed in this connection be strictly limited to situations of the highest priority in order to avoid diverting this communication channel to uses which might become detrimental to the objective for which it was originally established.”95 Final wording was proposed by Dulles at the end of August 1959.96
Starnes passed the new draft agreement to the chairman of the CSC, General Charles Foulkes. He explained that the JIC supported the change not only because it corresponded with the scope the Canadians wanted for the system, but “also because it is so clearly advantageous from a Canadian point of view in that the Canadian authorities will be brought quickly into the picture on critical situations which may develop outside the NATO area.” Foulkes took the agreement to Minister of National Defence George Pearkes; the wording of the revised agreement then went to the Secretary of State for External Affairs and Prime Minister Diefenbaker, who agreed on October 5, 1959, to an exchange of letters that month.97
The Berlin Crises, and the Broken Cables
The Berlin Crises of 1958 and 1959 were particularly important for establishing the patterns of habit of the British Joint Indications Room and, as a result, the Canadian liaison with the UK JIR. Pat Black used his experience in the summer of 1958 to establish a practice of meetings with Brits and Canadians to ensure that Canadians were getting every piece of intelligence possible. This included, for instance, a plan to drop in daily on the JIR at a certain intelligence alert level, and especially to seek daily reports from the headquarters of the British Army on the Rhine.98
In both Washington and London, the Canadian liaison officers were working to maintain their relationships with American and British counterparts, respectively. It is important to recognize that these relationships continued to be the mainstay of indications intelligence exchange.
And the communications system remained in regular use for the exchange of reviews and reports. During the Berlin Crisis, for instance, the British sent their heads of section and JIC assessments to Ottawa and Washington every Tuesday and Thursday.99
In January and February 1959, there were seven ALE messages exchanged between the three powers (along with twenty-eight exercise or test ALE messages, and thirty-three routine JICOTT, JICLON, or CIAWAS messages).100 But the seven true ALE messages, all sent in late February, were not about the Berlin Crisis. They were about transatlantic cables that had been severed.
On February 25, 1959, the Canadians sent an emergency ALE message to their allies, with a follow-up message the next day. London and Washington both sent back replies. A transatlantic cable belonging to the American Telegraph and Telephone Company (later AT&T), and four Western Union cables had each been broken.
After the first cable was disrupted, American Telegraph and Telephone charted an aircraft to reconnoitre the cables. The aircraft was observing the Soviet trawler MV Novorossisk at the time the fifth cable was broken. The damage had been done roughly 125 miles (200 kilometres) east of Newfoundland, where the cables were 200 fathoms (roughly 365 metres) below the ocean. It seemed at first that the trawler might have cut all five cables during a fifty-mile (80-kilometre) run from north to south.
There was initial confusion as to whether the cables had been cut cleanly, which would have been very difficult at that depth — or whether they had been broken by a trawler dragging its tackle against the cable. It was not terribly unusual for cables to break, and indeed there had been previous times when multiple cables had been damaged in the same short period. But in those cases, multiple trawlers had been in the area at once.
The Canadians and Americans each dispatched a destroyer to intercept and board the trawler.101 Later, in July, a staff member from the Privy Council Office attended a Joint Intelligence Committee meeting in Ottawa, bringing with him two sections of cables to demonstrate the “difference between a cable which had been broken through tension and one which had been cut.”102 The tenor of the discussions suggest that at least one of the cables had been cut cleanly, but this is not certain given the records released to date.
Future declassifications may shed light on how the cables were severed. But the effect of the incident was dramatic. As the UK JIC realized immediately, “if the Russians cut all the cables and simultaneously jammed the radio circuits, we should have no communications with North America.”103 In addition to the obvious threat posed to the cables themselves, the US “Argus” stratospheric nuclear test explosions had confirmed that nuclear explosions would interfere with radio communications, which were the back-up in case of disrupted cables. The US, UK, and Canada could have all the agreements in place they wanted, but if there were no practical means to communicate, the agreements were worthless.
The three allies decided that the February cable cuts were not indicators of imminent war, even though transatlantic cable-cutting was an item on the indicator list. The UK JIC considered Novorossisk’s journey might have been an “experiment” undertaken by the Soviets to see how difficult it was to cut cables in preparation for a “future planned operation.”104
The British urged a study of transatlantic communications, and the study was launched.105 An early cursory review by the Canadians included, ominously, that “our present communications facilities are completely vulnerable to Soviet interdiction.”106 Separately, US officials also noted the vulnerability of the transatlantic cables and began exploring the possibility of using circuits of a “USAF Wide-Band Tropospheric Forward Scatter System” — a system that did not require a physical cable link — to connect the three capitals as part of the broader study of possible Soviet interdiction of transatlantic communications.107 Within a few months of coming online, the alerts communication system was unreliable and obsolete, and the allies would begin looking for new systems to implement the agreement.
The Canadian Indications Procedure
In the time since the tripartite agreement was signed in 1957, the Canadians had continued to develop and improve their own indicator lists and supporting documentation, including JIC 312/2(59), February 5, 1959, “Probable Enemy Activities Prior to the Outbreak of War.” In February 1959, while considering the lists, it occurred to Bowen that JIC (Ottawa) had not taken steps to encourage Canadian representatives in missions abroad to report indications intelligence. Obviously, the missions in Moscow, Warsaw, and Prague, if properly organized and briefed on what was occurring in the Joint Indications Room, could be on the “continual lookout” for information listed as “Critical Intelligence Indicators,” and also be in a position to respond to requests from the JIR.108
As of Bowen’s writing, suitable short lists of indicators, and even the stages of intelligence alert adopted by the JIC (Ottawa), existed only as “parts of very highly classified documents relating to the Tripartite Intelligence Alert Agreement.”109 They were thus inaccessible for most field collectors, not because the lists themselves were important but because they disclosed the sensitive agreement. Bowen recommended that these lists be stripped from the larger tripartite packages and issued as separate JIC reports with as low a security classification as possible for distribution in the field.110
The Canadians coordinated with the British, who were also trying to bring Foreign Office posts “into the indications picture.”111 Just a month later, the Americans issued a General Indicator List to all diplomatic missions, and included a thorough explanation of how the lists were used by the NIC.112 (The Americans had sent out a similar message in 1955 after the establishment of the NIC.)
Taking the example of their allies, the Canadians distributed a list of key indicators, JIC 312/2(59), to diplomats posted abroad. It offered a list of “major actions and developments which it is believed may occur prior to the Soviet initiation of major hostilities against the West, and particularly against North America.”113 They also distributed a separate report of the list of Critical Intelligence Indicators, intended to serve as a guide for intelligence reporting officers in the field. Coincidentally harkening back to the late 1940s origins of indications intelligence, the document was to provide “a check list.” The list of alert stages was also distributed, and arrangements were made to advise the field when an alert had been declared. In an alert, officers were to review their lists and report anything which might assist assessment in Ottawa.114
On September 28, 1960, the JIC agreed to the “Joint Indication Room: Standing Orders,” JIC 378/1 (60). This document, which served the purpose indicated by its title, began with an overview of the Tripartite Intelligence Alerts Agreement. It laid out the process and procedures by which Canada would receive and communicate intelligence alert messages with Washington and London. The document would be superseded by updates in 1963 and in 1966: these documents are the best overviews of how the TIAA system was supposed to work. They confirm that Canada’s indications intelligence system was built as part of an allied partnership.
In 1962, the JIC planned to update JIC 378/1(60). At first, they hesitated to undertake significant study of the issue until the end of a major NATO exercise, FALLEX 62. The exercise was designed to test all NATO governments’ ability to wage and survive nuclear war. (FALLEX 62 did not go well for Canadian intelligence officials. They found out during the exercise that there was no room for the JIC at Canada’s National Emergency Headquarters.)115 Efforts to update the indications procedures were delayed again in October 1962 by a crisis with higher stakes than FALLEX 62.
The Cuban Missile Crisis
In early September 1962 the Joint Intelligence Committee approved a new paper titled “Intelligence Warning of Military Attack on North America,” JIC 443/2(62). The purpose of the paper was to assess the length of warning time available in the event of a Soviet attack on North America at any point in the next decade.
JIC 443/2(62) stated plainly that the allies could expect no intelligence with “direct knowledge” of a Soviet decision to go to war, but, as per the fundamental premise of the entire intelligence indications effort, they should expect to receive “evidence of preparations.” The main Soviet activity that would provide such a warning would be the “deployment or readying of Soviet nuclear forces for a massive attack on this continent.” But the main focus of the paper was on the launch of heavy bombers, and this “would provide little or no intelligence warning.”116
The same was true for Soviet deployments of missile submarines off the Atlantic or Pacific coasts of North America. As time passed, and the Soviets fielded ballistic missiles that were “permanently deployed and ready,” even “radically new sources of intelligence, such as reconnaissance satellites” would be challenged to offer timely warning of attack.117
The “clandestine introduction” of nuclear weapons into North America — here the Canadians were referring to early Cold War fears of small weapons smuggled into North America — would also be difficult to detect, even as smaller bombers were capable of offering greater and greater yield.118
Overall, the paper offered a grim assessment of the likelihood of intelligence warning. The authors were more correct than the JIC could have realized at the time. In July 1962 the Soviet premier, Nikita Khrushchev, and the Cuban leader, Fidel Castro, had agreed that the Soviet Union would deploy Soviet nuclear missiles to Cuba.
At the time the Canadians were writing the JIC 443/2(62), and noting challenges of both receiving warning about the launch of ballistic missiles and detecting smuggled weapons, the Soviets were clandestinely introducing medium-range ballistic missiles (MRBM) onto the island, about 150 kilometres from the United States.
Canadian intelligence officials did not learn about the missiles in Cuba via the Tripartite Intelligence Alerts Agreement. They learned about them at lunch.
On October 18, 1962, Director of Central Intelligence (and head of the CIA) John McCone invited Ivor Bowen and J. J. McCardle, the chair of the JIC (Ottawa) along with Geoffrey Cook, the JICLO(W), to lunch at his house. The Canadians were in Washington, along with British, Australian, and New Zealander colleagues for a CIA conference on the “impact of a changing world on the conduct of intelligence.”119 But for lunch, it was just the Canadians, McCone, and several other CIA people, including Deputy Director for Intelligence Ray Cline.
It was McCardle who first turned the conversation to Cuba. McCardle explained that Canada was going to refuse Soviet aircraft the right to overfly Canada enroute to Cuba.120 The Canadians, along with the other allies, were concerned by the increase in Soviet military aid being sent to Cuba. Indeed, over the summer the Canadian embassy in Havana had been reporting on the influx of Soviet personnel and material on the island.121 But the Canadians had no idea that the aid included nuclear missiles.
McCone told the Canadians that ballistic missile sites had been identified in Cuba.122 Photographic intelligence — that is, from U-2 spy planes — along with “other intelligence media” had convinced the United States that the USSR had installed about 40 offensive ballistic missiles in Cuba which “directly threaten the Security of U.S.A.”123 This was Canada’s first indication of what would become the Cuban Missile Crisis.124
At the end of the lunch, McCone asked the Canadians not to share the information of the missile sites with their government in Ottawa. The Americans had made other arrangements for this purpose: President John F. Kennedy has asked Livingston Merchant, an American diplomat who had been the ambassador to Canada until earlier in the year, to visit Ottawa and inform the Prime Minister in person. McCardle made clear he could not keep this information from his government. Upon returning to Ottawa, he reported to Norman Robertson and Robert Bryce, who passed the information to the Prime Minister. When Merchant arrived in Ottawa, Diefenbaker already knew about the missiles.
This exchange between the CIA officials and the Canadians harkens back to the Canadian discussion as to why an alerts agreement and network was important: that the Americans, in a crisis, might be so preoccupied that they would not inform the Canadians of intelligence information without a formal system in place. All available evidence suggests that, at least in the earliest days of the Cuban Missile Crisis, it was personal and working relationships and habit — lunch with the DCI — that got the intelligence to Canadians first. The Kennedy administration seems not to have used the alert system established by the TIAA to share information early in the crisis, and instead relied on a personal emissary, Merchant, who did not arrive in Ottawa until October 22, 1962.
In the afternoon of October 22, at 4:00 p.m., McCardle, as chair of the JIC, directed that a duty officer maintain a “continuous watch” in the Joint Indications Room (JIR). This was the day Merchant arrived, and McCardle knew Kennedy was going to make an announcement of a “serious nature” that night.125 That evening, Kennedy announced that the United States would quarantine any military equipment being sent to Cuba by ship and demanded the Soviets withdraw the missiles. This dramatically increased the possibility of a direct conflict between US and Soviet forces.
The JIC started meeting regularly the next day, on October 23. The committee reviewed a “Cuban Situation Report” that would become, over the next week and a half, a daily document with the new name “Special Intelligence Report.” This daily report developed over the crisis to contain information about the states of allied intelligence and readiness, Soviet motives in placing the missiles in Cuba, and the US and USSR positions at the United Nations. The committee also agreed on October 23 to declare a “Stage 2” intelligence alert, but it is not clear whether the alert took effect on the 23rd or the 24th.126 While the alert level is not definitive, it does suggest that Canada had not received ALE messages from the US or UK, nor had Ottawa sent such a message by this date.
The minutes of the discussion in the JIC on October 23 is instructive.127 The director of air intelligence was tasked with investigating “the matter of providing air service to Alert” — that is, the Alert Wireless Station in what was then Alert, Northwest Territories (now Alert, Nunavut). The place name for Alert, the northernmost continuously inhabited settlement in the world, is simply a coincidence. But this discussion point suggests that Canada was gathering signals intelligence on the Soviet Union from Alert with an eye to gaining indications intelligence.
At the same time, the communication system that supported the TIAA had recently been tested, and so the JIC agreed not to send any further test messages, presumably to keep the line clear and reduce any confusion in case of a true alert message. The DNI was tasked with investigating and providing “ways and means by which knowledge of cable breaks would be made rapidly available to the JIC,” and the Air Force was to notify the JIC of any “communications interruptions considered to be due to deliberate action.” The JIC agreed that if there were indicators of “sabotage” or “communication interruptions,” they would pass messages to London and Washington.128 There was no mention of any ALE messages having been sent, but the Canadians were clearly trying to keep the line clear and ensure it remained in working order.
The JIC was also concerned with ensuring that Canada was aware of the states of operational readiness of US forces. This stemmed from the Canadian understanding that US actions could result in Soviet responses that would appear as indicators of war. Yet it also seems likely that the Canadians were monitoring the state of US readiness to ensure the Government of Canada had the fullest possible knowledge of American actions.
The JIC continued to meet daily, adjusting its meeting times to the early work hours. This allowed the committee time to prepare briefs for others, including the Chiefs of Staff Committee and daily briefings to the Cabinet Defence Committee. These briefs were usually given by McCardle as chair. Diefenbaker himself was secretly pre-briefed to allow him to ask knowledgeable questions in front of his colleagues in the CDC.129
The JIC efforts during the Cuban crisis were, fundamentally, an indications effort. Indications liaison officers from the various service branches met daily to prepare information for the JIC meeting and background material for the prime minister who spoke daily in the House.130
On October 24, 1962, Canadian intelligence officials noted that “both sides” — the Americans and the Soviets — “have taken significant precautionary military measures.” Still, the “intelligence available does not permit a judgment of Soviet intentions.”131 These precautionary measures might have included indications that the USSR was preparing to move bombers to Arctic bases.132
On October 26, 1962, the JIC declared an Intelligence Alert Stage 1. The motive for the decision — whether the result of a Canadian assessment or receipt of an ALE message — is unclear. Later that evening, Kennedy received a private message from Khrushchev backing down to American demands.
October 27, 1962, was perhaps the most critical day of the crisis. Khrushchev sent a second message to Kennedy, seemingly backtracking on a first message he had sent to end the crisis. That day, the US Watch Committee noted that five of the MRBM sites in Cuba appeared to be fully operational.133 US Navy Growler planes were shot at over Cuba, and a US U-2 aircraft was shot down and the pilot killed. War was imminent.
In a letter in support of a later departmental oral history project, Malcolm N. Bow, the special assistant to Secretary of State for External Affairs Howard Green, recalled how he shuttled “intelligence information” from Green to Diefenbaker on the evening of the 27th:
At the most critical hour of the missile crisis on the evening of October 27, I was summoned to the Greens’ residence in the Roxborough Apartments. My instructions were to deliver a sealed envelope to the Prime Minister at his residence, to be sure that he read the contents immediately and telephoned his reaction to the Minister. Subsequently, I learned that the intelligence information I delivered concerned a US intention of bombing the Cuban missile sites that evening and that the Soviet assurances of withdrawal which forestalled hostilities were received only 25 minutes before the air strike deadline.134
It is unclear just what this “intelligence information” was, or its source. Earlier in the week, the phrase “intelligence information” had referred to McCone’s meeting with Bowen, McCardle, and Cook. It is possible, but not certain, that the intelligence information Bow is referring to was received via the TIAA network.
The next day, on October 28, 1962, the JIC noted that Radio Moscow was announcing the Soviet Union’s willingness to “retire” the offensive weapons in Cuba. The JIC declared an Intelligence Alert Stage 2, a lesser stage of alert. Over the coming days, it became evident the Soviet Union was, in fact, dismantling some of the missile sites in Cuba.
In the famed thirteen days of October 1962, the world was on the brink of nuclear war. Despite the enormous number of declassified American intelligence records related to the crisis, it remains unclear whether the Tripartite Intelligence Alerts Agreement was invoked during the Cuban Missile Crisis. On balance, it appears the agreement was not invoked.
The crisis, one in which the decision for escalation rested with the United States, was not the type of event that the TIAA had been established to identify. It was, however, in keeping with the Imminence of War papers developed nearly a decade before. War had nearly come not because of a major Soviet offensive, but rather a Soviet miscalculation about how the United States would react to Soviet policy. Canadian procedures and processes had shied away from considering situations in which the United States would choose war — an intelligence challenge with too many political implications for Canada to meet squarely.