The Third Eye

Autonomous signals intelligence has put Canada at the heart of Five Eyes sharing — writes Maria A. Robson

CANADA FOUNDED ITS first intelligence agency, the Communications Branch of the National Research Council, in 1946. The word “Security” was added in 1973 and since 1975 it has been known as the Communications Security Establishment (CSE). In the nationalist mood following the Second World War, some policymakers assumed that the development of autonomous signals intelligence would allow the country to go its own way and assert independence from Great Britain. As it happened, this expectation fit nicely into a colony-to-nation narrative. However, declassified wartime archives and the records of postwar signals intelligence (SIGINT) negotiations tell a different story. It turns out that a Canadian autonomous signals intelligence capability was designed to strengthen national credibility by solidifying Canada’s value to the British and the Americans. In fact, the paramount motivation for Canadian SIGINT was always to have a seat at the table in the nascent intelligence-sharing alliance known today as the Five Eyes. Canada developed CSE and tailored its products specifically to be part of that all-important club.

Canada has always been a net beneficiary in intelligence sharing: The quantity of intelligence we receive exceeds what we provide. Nonetheless, the value of our SIGINT has been sufficient to warrant inclusion in the world’s most-powerful intelligence partnership to emerge from the Second World War — the United States, United Kingdom, Canada, Australia, and New Zealand. The name of “Five Eyes” derives from the “For XXX Eyes Only” designation on intelligence products, in which XXX might stand for any or all of the five members. While economics tells us that all trading arrangements involve partners identifying their comparative advantage (and there is evidence that the U.K., a powerful intelligence player, has done the same, often tailoring its collection activities to meet American needs) the declassified documents reveal how for Canada, alliance-tending was paramount in the Canadian thrust for SIGINT. In many highly visible ways, during the war and in postwar negotiations, Canada distanced itself from the Commonwealth as it deepened its bilateral relationship with the U.S.A. Even so, British influence continued to shape Canadian intelligence, and the Commonwealth bond persisted — perhaps closer to the national interest than ever before — through Five Eyes.

Origins of SIGINT

Before the Second World War, Canada left almost all foreign intelligence collection to its partners and allies. Canadians maintained a modest intelligence presence in what was called, cryptically enough, the Examination Unit (XU), operating under British command. The British and Americans’ cryptographic units recognized the potential of Canadian contributions: early assessments show that Canadian commercial SIGINT provided good value in British eyes. By contrast, New Zealand was admonished to improve its choice of targets, and South Africa to be more timely in producing intercepts, while Southern Rhodesia was simply told to stop.¹

CANADA'S EXAMINATION UNIT flourished in response to British and American demand, with its role being to fill gaps in Allied intelligence collection. Canadian SIGINT included intercepts of Vichy French communications and low-grade agent traffic, including picking up the communications of German and Japanese operatives in North America. The Examination Unit added Japanese translation capabilities in August 1941, with one official quietly reporting: “The Unit is producing results of high value to our Intelligence Services and we are now on the point of substantially increasing its output as we have added some Japanese and Spanish translators to the staff.”² Once the Japanese unleashed the Pacific War and the United States joined the allies in December, Canada’s targeting of Japanese communications became an even more valued contribution to American intelligence. Great Britain’s ability to collect SIGINT on Japanese targets was constrained and British cryptanalysis liaison files reveal the Canadians’ shrewd decision to specialize in areas no one else was covering. One GC&CS official wrote to another: “Suggested at beginning of August that Canada might do Jap. Commercial as they wanted to do work not duplicated elsewhere.”³  Thanks to geography and adaptability, Canada was well-placed to do so.

In Ottawa, the Department of External Affairs negotiated a guaranteed intelligence exchange, in which Britain shared knowledge of value to Canadian military efforts in Europe, such as information about the Abwehr, German military intelligence.⁴ The Royal Canadian Navy and the Canadian Army in Europe were also both engaged in Ultra, the secret and immensely successful British cryptanalytic program at Bletchley Park which was run by the Government Code and Cypher School (GC&CS), later known as GCHQ or Government Communications Headquarters. Canadians worked at Bletchley Park, and the Royal Canadian Navy was involved in routing Allied convoys safely away from the U-boats, using Ultra decrypts to anticipate their movements.⁵ From the British point of view, the integration of Canadian SIGINT into British efforts, the dependence of Canada’s fledgling capabilities on their guidance and expertise, and the close alignment of each country’s priorities, combined to make the Dominion a reliable partner.

The Americans, too, relied on XU Pacific intercepts to supplement their own and found the Canadian product useful, even though it was, as one Canadian scholar put it, “only 1/100 the size of its American equivalent.”⁶ Canada was obviously a junior partner given the Americans’ vast superiority in resources. Indeed, with increasing cross-border SIGINT cooperation, 1943 correspondence in British archives reflects Canadian concerns about the risk of becoming merely an “appendage to the U.S.”⁷ Still, the vital point was the relationship of exchange. Canada could not expect to receive anything for nothing.

Colonel W.W. Murray, the Director of Military Intelligence, later reflected on Canada’s contributions to intelligence-sharing:

When our contribution was nil, we received nothing from either Bletchley or Washington. When, in agreement with them, our contribution became substantial, we received ample return — a seat in their counsels and a regular budget of valuable Intelligence. If we contribute to the pool, we shall draw something from it in the form of finished products; if we fail to contribute, we shall receive nothing.”⁸

Ultimately, Canadian SIGINT efforts were channelled to provide a maximum-value product for Allied priorities, which in turn served the Canadian national interest.

‘Younger brother’

During the war Canadian SIGINT operations were not autonomous. They depended on British and American expertise and technology. The British controlled Canada’s critical Hydra communication station at Camp X on the shore of Lake Ontario near Port Hope. British codebreakers headed the Examination Unit, which at first was headed by an American, Herbert Yardley. Canada’s partners saw Yardley’s leadership as a disaster. Infamous for having already revealed details about American and British SIGINT in his 1931 memoirs, The American Black Chamber — one of the earliest modern intelligence exposés — Yardley presented what should have been an obvious risk to wartime SIGINT.

Unfortunately, as documented by historian Wesley Wark, the Canadians appointed by the NRC to quickly set up a prospective Canadian SIGINT unit were convinced by a glowing report from the U.S. Army Signals Corps, and by Yardley’s own assurances of how swiftly he could stand up a Canadian operation. Mistakenly assuming that the broader American intelligence community had faith in Yardley, and without seeking American or British official approval, they expedited him into a six-month trial role without a background check. The FBI was not thrilled. GC&CS also reacted strongly, withholding intelligence sharing from the Canadians lest it be leaked. Lester Pearson, an assistant under secretary involved in intelligence matters, fearful of a Canadian exposé if Yardley were to be dismissed on bad terms, negotiated for him to finish the six-month trial period.⁹ In December 1941, Yardley was replaced by the more reliable Oliver Strachey, a Bletchley veteran.

Canadian cryptanalysis definitely had growing up to do. British archives reveal officers referring in 1941 to “unreliable” decrypts, with the British continuing to guide the Canadians, who were given the opportunity to take cryptographic training courses in the U.K. Capt. Humphrey R. Sandwith described certain Canadian capabilities in July 1942 as “in embryo. They need a great deal of advice and assistance.”¹⁰

Fortunately, in this growing-up process, the Canadians had a kind of nursery chum in the Americans. Their SIGINT, as Canadian cryptanalysts later recalled in a postwar review, had been “almost equally beholden to the U.K. for a start in many systems.”¹¹ The Americans were, therefore, “that much the more willing to give Canada every assistance” in developing their own signals intelligence.¹²

Acutely aware of their junior status, the Canadians were occasionally frustrated. After visiting Ottawa in September 1942, Capt. Geoffrey Stevens, GC&CS liaison to the U.S. Army SIGINT agency, observed: “Canadians feel they are being treated rather as a younger brother.”¹³ The following year, GC&CS officials privately observed that Canadian liaison was “tricky,” stating that the Canadians felt “they had have made a big contribution to the war effort, which has not brought them enough recognition.”¹⁴ Examination Unit personnel frequently focused on whether they were making a unique contribution, and whether British intelligence gave them enough credit. It is difficult to gauge how much of this feeling was justified — after all, “a big contribution” for the Canadians was not necessarily big in the context of British or American resources — and how much derived from trying to make sure that Canada was not left out.

Strengthening ties

By 1942-43 the Americans and British were laying the groundwork for intelligence cooperation at a depth that the world had never seen before. The 1942 Holden Agreement established intelligence sharing between the U.S. Navy and the GC&CS.¹⁵ The BRUSA Agreement of 1943 deepened Anglo-American and Commonwealth intelligence cooperation. Historian Bradley Smith describes BRUSA as “one of the most remarkable acts of trust and cooperation ever achieved by two great powers” but notes that BRUSA also reflected American political opportunism and tough bargaining,”¹⁶ foreshadowing aggressive U.S. postwar intelligence negotiations.

Under BRUSA, the British were tasked with monitoring German and Italian communications, and the Americans with Japanese. Canada leveraged its advantages to contribute to both, tailoring its effort to Allied needs. In one case, the XU collected Japanese military traffic in China,¹⁷ which was of no interest to anyone in Ottawa but could be traded for useful information gathered by the Allies. Naturally, as the Americans’ war effort and output ramped up, Canada’s focus shifted in their direction. This was manifested in direct SIGINT liaison, and in the XU recalibrating its collection targets to meet U.S. needs.

The Examination Unit presents something of an unsung Canadian success story. An internal postwar assessment judged that Canadian SIGINT grew “[f]rom practically nothing in 1941” to a point in 1945 where Canadians “pulled our own weight.”¹⁸ These achievements did not go unnoticed.

‘Independence’

After the war, Canadian policymakers had to decide: should they continue to authorize foreign intelligence gathering in peacetime? Or was it time to shut down covert collection? And at what cost to Canadian influence and autonomy? The Canadian Joint Intelligence Committee was determined not to revert to prewar “dependence.”¹⁹ Others at the Department of National Defence emphasized SIGINT’s wartime value, and even suggested that relying on Britain for intelligence would cost as much as running our own agency. In the view of Lt. Gen. Charles Foulkes, Chief of the General Staff, Canada’s growth “as a World Power depends on an independent appraisal of World Affairs.”²⁰

On the flip side were policymakers at the DEA such as Hume Wrong and Norman Robertson, who doubted the usefulness of peacetime intelligence collection, preferring information gathering via traditional diplomatic means. Martin Rudner has contrasted the Canadian military’s “ambitious” postwar ideas with the “moralistic,” resource-constrained External Affairs officers who put their faith in the diplomatic method.²¹

International developments, and in particular the growing Soviet threat, nudged Canadian policymakers towards setting up an intelligence agency. Canadian strategic assessments from 1944 onward identified the U.S.S.R. as the primary threat to North American security.²² That threat was underscored by the 1945 Gouzenko Affair, in which papers smuggled out of the Soviet embassy by junior clerk Igor Gouzenko revealed aggressive espionage against the West that involved Canadian agents. Although the Affair itself does not appear to have directly influenced the decision to develop autonomous Canadian intelligence,²³ it was a reminder that American secrets might be targeted by spies in Canada, making this country a possible weak link — and thus pushed Canadian policymakers back towards international security cooperation.

AN INTELLIGENCE AGENCY needs a specialty, and in apportioning limited resources Canada now had to choose between SIGINT and HUMINT. Human intelligence, that is to say, espionage based on the collection and often high risk work of field agents and networks, had strong negative connotations in Canadian circles. In the wave of post-1945 idealism, HUMINT was seen as inconsistent with Canada’s desired international image. On the other hand, SIGINT’s stock was now high. Canada’s intelligence leaders felt good about their record, and a continued focus on SIGINT was widely recognized as the country’s best option. Specializing in SIGINT would allow Canada to be highly specialized, leveraging its comparative advantage. Ultimately, the government decided to collect foreign intelligence by intercepting communications rather than creating a foreign espionage agency like MI6.

Having decided to maximize Canada’s value to its top customer, the U.S.A., the Canadians resisted a British scheme called the “Travis Plan” after Sir Edward Travis, head of GC&CS, for a joint Canadian-British SIGINT effort. In a system comparable to the early 20^th century “tin pot navy” (with Canada providing ships for the Royal Navy) the Travis Plan would have had Canada providing raw decrypts for the British to analyze. Although the External Affairs brain trust had been sceptical of investing in intelligence, they were if anything more sceptical of Canadian intelligence being controlled by Britain. ²⁴ Thus while other Commonwealth countries remained under the London-centric model, Canada’s SIGINT relations mirrored a broader trend towards autonomy and closer ties to the United States.

This was no more than Travis had himself foreseen when he predicted to MI6 chief Sir Stewart Menzies in 1943 that the British could “not afford to refuse the Canadians at this point. If we do they might turn to the U.S.”²⁵ It could not be said that the British had failed to reach out. Canadian officials, however, evidently felt less concern about gravitating away from British tutelage, only to fall into the American orbit. While the Canadians could not shake off strong British cultural and structural influences in their intelligence operations – and would benefit from both for years to come – British involvement moved from overt control to softer influence as the Canadians formally distanced themselves from the London-centric Commonwealth model.

‘Communications Branch’

These postwar tendencies, then, led to the creation in 1946 of the Communications Branch of the National Research Council, a deliberately benign name chosen for an autonomous peacetime SIGINT agency with a dual mandate of collection and communications security. As noted above, “security” was not added to the name until 1973. The military also set up new intercept stations. Canadian SIGINT specialists took control of Hydra in 1947,²⁶ in order, in the words of then-High Commissioner to London Norman Robertson, to “make an acceptable contribution” to Allied intelligence sharing.²⁷

Though the agency was independent, British influence remained strong. Canadian encipherment, an embryonic skill-set, relied on British machines. The Communications Security Board, which directed SIGINT-related matters, closely mirrored the British structure. British influence continued to somewhat overshadow the emerging culture of Canadian SIGINT, with much of the CBNRC leadership coming from GC&CS and MI6.²⁸

Far from representing a break with the past, the organization that would become the CSE was founded “to continue the work they were doing during the war with as little disruption as possible to the collaboration that had developed between Canada, the United States and the United Kingdom on the sharing of Signals Intelligence,”²⁹ as the CSE website says today (emphasis added). Since 1946 Canadian SIGINT has been dedicated to maintaining relations and exchange with allies.

A Third Eye

The Five Eyes grew out of wartime intelligence-sharing discussions, starting with the 1942 Holden Agreement and culminating in 1946 with the initial  pact that became known as UKUSA. The agreement was refined over the next few years. Canada joined and by 1955 Australia and New Zealand were also in.³⁰ When the British and Americans negotiated UKUSA, Canada was a point of contention. American negotiators treated Canada differently from other Commonwealth countries. U.S. Army and especially Navy representatives advocated for direct Washington-Ottawa intelligence-sharing, helping the Canadians to escape subordination to Britain. At a 1945 meeting, Brig. Gen. W. Preston Corderman noted that the Canadians had “manifested a desire to make arrangements with us without consulting the British.” ³¹ Commodore Thomas B. Inglis, a supporter of bilateral American-Canadian SIGINT collaboration,³² indicated that the U.S. Navy would not accept the London SIGINT Board (LSIB) negotiating on Canada’s behalf until the Canadians agreed, which he was sure they would not.³³ Cdre Inglis also emphasized that American SIGINT authorities needed “freedom of action” to work with “any Dominion [that] does not recognize the authority of the London SIGINT Board.”³⁴ Under American and Canadian pressure, the British finally exempted Canada from their negotiating protocol. SIGINT was therefore one of the first areas in which Canada officially broke from British control, asserting its independence while quietly continuing to leverage British expertise.

In the ensuing Five Eyes relationship, each country was tasked with monitoring a specific region of the world. Canada was well-situated to provide intelligence on the Soviet Union, its European satellite states, and South America. In the early Cold War, Canadian SIGINT targeted Arctic-based Soviet research stations and naval, military, and strategic rocket force communications, critical for preparing North American defences against Soviet aggression.³⁵

POSTWAR CANADIAN policymakers pushed for independence in foreign affairs, but in building its first autonomous foreign intelligence agency, Canada was actually cementing its niche in allied intelligence sharing. Canadian military intelligence argued for independent operations, and declassified documents confirm how aggressively the Americans were pushing for that outcome. The British were already open to Canada controlling its communications security, but were less keen on Canada getting into independent foreign intelligence collection.

The result was what American representatives in the UKUSA negotiations wanted: namely, that Canada should leverage its comparative advantage in SIGINT and liaise with them directly. The groundwork was laid during the war.

In the end, the importance of Canadian “independence” from the British should not be overstated. While officially independent, SIGINT operations relied heavily on British expertise, technology, and methods, and Canada stayed relatively aligned with the British intelligence model and British security needs in and out of the Five Eyes.³⁶ However, on the surface, Canada presented an image of autonomy and shifted from the London-centric model of intelligence-sharing towards direct intelligence exchange with the Americans. While Canada’s external relations were increasingly independent of the U.K., Canada sought with some success to avoid becoming merely “an appendage to the U.S.”³⁷

Reciprocity was key for Canadian decision-making. Defence officials and policymakers emphasized that Canada must contribute to the pool to benefit from it. From the 1930s onward, Canadian signals intelligence capabilities were influenced by, in Robertson’s words, making “an acceptable contribution.”³⁸ The declassified files show the Canadian desire to move from the status of “younger brother,” even as the U.S. Army and Navy pushed for Canada to conduct its own signals intelligence and to work directly with Washington. As would soon be the case with the United Nations and the North Atlantic Treaty Organization, the overarching goal was a seat at the table. The newly-minted Communications Branch ensured Canada’s place at the heart of the all-important Five Eyes, a position that continues today.

Notes

1. “Appendix A,” Meeting in London on 28 July 1937, Examination Unit, WWII—9 Pt. 1: XU Correspondence, LAC RG24, Vol. 29164.
2. Letter from Pearson to Massey, 22 August 1941. Examination Unit, WWII—9 Pt. 1: XU Correspondence loc. cit.
3. “Dossiers of Canadian Liaison. Notes on Sigint cooperation with Canada during World War II,” 9 September 1942, TNA HW50/08, 4.
4. Wark, “Cryptographic innocence,” 654.
5. Beesly, “Ultra and the Battle of the Atlantic”, 1977, released by the NSA.
6. D.B. Robinson (ed.), History of the Examination Unit, LAC RG24 Vol. 29166 File WWII-31, 47.
7. K.J. Maidment, “Dossiers of Canadian Liaison. Notes on Sigint cooperation with Canada during World War II,” 21 Aug 1943, TNA HW50/08, 6.
8. Skaarup, Out of Darkness—Light.
9. Wark, “Cryptologic Innocence”, 646-652.
10. “Dossiers,” Y Committee Papers: 24 Jul. 1942, TNA HW50/08.
11. Robinson (ed.), History of the Examination Unit, LAC RG24 Vol. 29166, 46.
12. Ibid.
13. Geoffrey Stevens to Alistair Denniston, “Dossiers of Canadian Liaison. Notes on Sigint cooperation with Canada during World War II,” Sept 2, 1942, TNA HW50/08, 4.
14. O’Connor to Commander Edward Travis, “Notes on Sigint cooperation with Canada during World War II,” 1943, TNA HW50/08, 6.
15. Erskine, “Holden Agreement on Naval Sigint,” 187.
16. Smith, “Sharing Ultra in World War II,” 70.
17. E.g., “Notes on Discussions Held during Dr. Robinson’s Visit to Signal Security Agency, May 17-20, 1945”, in Examination Unit. WWII—22: XU—Liaison with Washington, LAC RG24, Vol. 29166. These discussions included the setting of Canadian signals intelligence collection targets, including targeting communications in France and Northern China.
18. Ibid.
19. Cooper, CFIS: A Foreign Intelligence Service for Canada, 14.
20. Jensen, Cautious Beginnings, 124-7.
21. Rudner, “Historical Evolution,” 68.
22. Eayrs, In Defence of Canada, Vol. III, 320.
23. Rudner, “CSEC from Cold War to globalization,” 100.
24. Jensen, Cautious Beginnings, 123.
25. Travis to Menzies, 16 Feb 1943, TNA HW50/08.
26. Department of National Defence, “Memorandum: Hydra Communications,” Canada Declassified, online.
27. Jensen, Cautious Beginnings, 161.
28. Robinson, “CSE Facilities,” Lux Ex Umbra, 26 June 2011, online.
29. (Emphasis added.) Communications Security Establishment Canada, “The Beginning: The Communications Branch of the National Research Council,” 9 January 2018, online.
30. See NSA, UKUSA Agreement: “Principles of UKUSA Collaboration with Commonwealth Countries Other Than the UK; Appendix J 13 Feb 1961,” esp. ‘At this time, only Canada, Australia and New Zealand will be regarded as UKUSA-collaborating Commonwealth countries.’ Annex J1 lays out ‘UKUSA arrangements affecting Australia and New Zealand.’
31. “Joint Meeting of ANCIB and ANCICC”, 29 Oct 1945, NSA and GCHQ, UKUSA Agreement, TNA HW80/01.
32. Wark, “The road to CANUSA.”
33. “Joint Meeting of ANCIB and ANCICC”, 29 Oct 1945, TNA HW80/01.
34. Ibid.
35. Rudner, “Contemporary threats,” 148, and Wark, “Favourable geography”.
36. Wark, “Road to CANUSA,” 29.
37. Ibid.
38. Jensen, Cautious Beginnings, 161.

Reproduced from the Autumn-Winter 2020 edition of THE DORCHESTER REVIEW.  Maria A. Robson is a Ph.D. candidate in Political Science at Northeastern University in Boston. She holds a Master’s in Military, Security, and Strategic Studies from the University of Calgary, and a B.A. in International Relations, Economics, and History from the University of Toronto. A longer version of this paper appears in the Journal of Intelligence and National Security 35:7.