Christopher Andrew and the Strange Case of Roger Hollis

[This text has been edited by the author for online publication]

When Christopher Andrew’s Defence of the Realm: The Authorized History of MI5 was published in late 2009, the Australian’s European correspondent, Peter Wilson, hailed it with a short piece headed “‘Mentally ill’ spycatcher more dangerous than KGB”. The spycatcher in question was Peter Wright, whose book, Spycatcher, the British government made strenuous efforts to suppress, only to be foiled in the Australian courts by the young Malcolm Turnbull. The book was published in Australia in 1987 (not 1985, as Wilson reported) and became a best-seller. Wilson’s piece was an eyecatcher, because it concentrated exclusively on the question of Wright’s work and, in particular, his famous claim that Roger Hollis, Director General of MI5 between 1956 and 1965 (and before that Deputy Director General 1953–56), had been a Soviet mole—a claim the authorised historian of MI5 dismisses with contempt. 

Wilson’s article caught my eye, because I had only recently finished reading Chapman Pincher’s new book, Treachery (Random House, 2009), which argues in great detail that Roger Hollis was, indeed, a Soviet mole. I knew that Pincher had been making this argument for nearly thirty years, since Their Trade is Treachery (1981) and Too Secret Too Long (1984). If the authorised historian dismissed Peter Wright’s claim as due to mental illness, what of Pincher? 

I looked forward to reading the authorised history in order, chiefly, to discover how Christopher Andrew demonstrated both the mental illness of Peter Wright and the baselessness of the charges against Roger Hollis. What I found was profoundly puzzling: the authorised historian completely fails to address the indictment of Hollis. More precisely, he attacks the conclusion of Wright and Pincher that Hollis was a Soviet agent, but he fails to address their arguments as they are exhaustively presented by Pincher in his new book. Andrew repeatedly states that suspicion of Hollis was the obsession of “a small but determined group of conspiracy theorists”, but nowhere does he so much as mention the specific claims on the basis of which Hollis was and is suspected; to say nothing of demonstrating that they amount to a groundless conspiracy theory. 

I’ll go over the claims in more detail shortly, but let me first register the puzzlement I felt in reading Defence of the Realm. Pincher’s case against Hollis rests on numerous lines of evidence and he does not always follow through in his analysis as rigorously as he might have done, but consider just a few of the chief claims he makes. When GRU cipher clerk Igor Gouzenko defected in Ottawa, in September 1945, according to Pincher, he declared that there was a GRU mole at a high level in MI5 by the codename Elli. This was contested by Gordon Brook-Shepherd more than twenty years ago. He claimed—in The Storm Birds: Soviet Post-War Defectors (1988)—that Gouzenko had said no such thing in 1945; that it was only in 1952, in the wake of the flight of Burgess and Maclean and the suspicion of Philby, that Gouzenko mentioned such a mole and, even then, did not offer a codename. The codename, he says, surfaced only in 1980, when Chapman Pincher and Peter Wright were collaborating on Their Trade is Treachery. Indeed, Brook-Shepherd argued, the codename Elli was surely improbable, because it had been the codename of Kathleen Wilshire, another GRU spy, whom Gouzenko had exposed as working in the British High Commission in Ottawa in the early 1940s. Surely, he reasoned, Moscow would not have run two spies with the same codename at the same time? 

But note well: Christopher Andrew, with his unprecedented access to the MI5 archives, accepts the claim that Gouzenko mentioned a mole inside MI5 codenamed Elli in 1945. Indeed, he goes so far as to cite, from the MI5 archives, a report that Hollis had implied, shortly after Gouzenko’s description of the mole in MI5 as Elli, that Anthony Blunt was Elli. Blunt was, of course, a KGB spy, not a GRU one, and was not Elli. He had three codenames: Fred, Tony and Blunden (the last two of which look surprisingly transparent in retrospect). Andrew uses this as a basis for declaring it “sadly ironic” that Hollis himself should be suspected of having been Elli. He then argues that Elli was actually Leo Long, a known Soviet spy. But Long, also, was KGB. His codename, it turns out, was Ralph; and he worked in MI-14, on the German Order of Battle, for the War Office. He was not, therefore, a GRU spy and he did not work at a senior level in MI5. So the authorised historian has a problem: by his own account there was an Elli in MI5, but it was not Blunt and it was not Long. Who was it? Brook-Shepherd may have asserted that there was no Elli in MI5, but the authorised historian says there was. In any case, we now know, from the Vassiliev notebooks, that on 24 November 1945, the Soviet Commissar for State Security, Vsevelod Merkulov, sent a personal message to Stalin and Beria confirming that Gouzenko had betrayed the existence of ‘the GRU agent inside British intelligence, Elli’. There is, therefore, an inescapable and unanswered question as to who this Elli was. 

Next, consider that Hollis mixed in communist circles at Oxford in the 1920s and then covered up his links with the likes of Claud Cockburn, after joining MI5 in 1938, although Cockburn was listed at MI5 as a “dangerous communist” by then; that Hollis spent nine years in China (1927 to 1936) mixing in communist circles again in Shanghai and Beijing that included Agnes Smedley and Rewi Alley, as well as the important GRU agents Arthur Ewert and Richard Sorge; that he returned to England via Moscow in both 1934 and 1936, though ever afterwards and even at the time obscuring both the fact of those visits and their duration; that among those whom he may have met, through Ewert or Sorge, in China was Ursula Hamburger nee Kuczynski, who went on to become a GRU agent by the codename of Sonia; that she was decorated multiple times by the Soviet government and cited, after her death in 2002, by Vladimir Putin himself, as a super agent of Russian military intelligence; that Sonia was based outside Oxford very close to the MI5 wartime headquarters at Blenheim Palace, where, according to Pincher, Hollis worked during the years 1941–45; that during those years, by her own later account, she felt she had a protective hand inside MI5; that Hollis at that time headed F Division of MI5—Soviet counter-espionage—while Kim Philby headed the corresponding section in MI6; and that reports by the Radio Security Service in those years of illegal radio transmissions from the Oxford area, later found to be Sonia’s, were invariably returned to the RSS by both Hollis and Philby marked “No Further Action”. 

Against this background, one of the first things I did when I got hold of Defence of the Realm was turn to the index to check what Andrew had to say about these matters. What I found astonished me: there is no entry under Shanghai; no entry under Ewert or Sorge; no entry under Sonia; no entry under Kuczynski (or Hamburger); no entry under Radio Security Service; no entry under Cockburn. There is no discussion anywhere in the authorised history of any of these matters. Andrew does state (at page 231) that the counter-espionage officers stayed in London at a building in St James Street, during the war, but he does not make any reference in this context to Hollis or to Pincher. Hollis’s background before joining MI5 is passed over by the authorised historian in a single paragraph, which reads as follows, at page 136 of the mighty 1000-page tome:

‘Roger Hollis joined the Security Service in June 1938 … He came from a prominent Anglican family; both his father and elder brother were bishops. Hollis left Worcester College Oxford without taking his degree to begin a business career in the Far East. In 1937 [sic] he was advised to return to Britain for health reasons, and contacted the Security Service to say that he had been told there was a vacancy for “someone with a practical knowledge of the Far East”. Hollis struck his initial Service interviewer as “A rather nice quiet young man whose only qualifications were a knowledge of the northern Chinese language and Chinese and Japanese commercial industry. Might perhaps be given a job.” Kell [the founding Director General of MI5 and still its head after almost thirty years in 1938] decided to recruit him.’

It’s curious that he gives 1937 as the year of Hollis’s return, not 1936, since according to Pincher, Hollis was in England in 1937 and had meetings with a number of communist or fellow-travelling figures as well as making a mysterious visit to Paris, which is where the GRU ran its European operations. Given that vetting was not introduced until 1952, we might not be altogether surprised that MI5 would recruit a nice young man from an impeccable Anglican background with so few questions in 1938. But to pass over Hollis’s background so insouciantly in 2009 beggars belief. It might be explained on the basis that the authorised historian is entirely convinced that Hollis was indeed a nice young man who went on to serve his country faithfully. But to explain is not to excuse. Given the grave allegations made against Hollis, it was surely incumbent on Andrew to do some kind of retrospective vetting and establish the bona fides of the deceased. He did nothing that can even remotely be so described. 

Before going into this matter and its implications in more detail, I should make clear that Christopher Andrew is very well qualified to have written the authorised history of MI5. He has been at work on history and diplomacy for more than forty years and on the history of Soviet espionage in the West and in the world at large for at least thirty of those years. His book Secret Service: The Making of the British Intelligence Community was published by Heinemann in 1985. (Ironically, it was Heinemann that was also to pick up Wright’s book shortly afterwards.) His collaboration with KGB defector Oleg Gordievsky in the late 1980s produced KGB: The Inside Story of its Foreign Operations from Lenin to Gorbachev, in 1990. But he had the even greater good fortune, from the early 1990s, to work with KGB defector Vassili Mitrokhin on the masses of secret KGB files that the latter had smuggled out of the collapsed Soviet Union in 1992. This issued in two bulky tomes: The Mitrokhin Archive: The KGB in Europe and the West (Allen Lane, 1999) and The Mitrokhin Archive II: The KGB and the World (Allen Lane, 2005). This publishing record enabled him to be granted unprecedented scholarly access to the archives of MI5 and a contract to write its official, authorised history. 

Of course, he has been far from alone in writing about Soviet espionage in the light of revelations from the archives of the former communist superpower. The collapse of the Soviet Union has been followed by something of an avalanche of books on the subject. Some of the more notable have been John Costello and Oleg Tsarev, Deadly Illusions: The First Book from the KGB Archives (1993), which had the intriguing cover comment, “The KGB secrets the British government doesn’t want you to read”; Sam Tanenhaus, Whittaker Chambers (1997); Nigel West and Oleg Tsarev, The Crown Jewels: The British Secrets at the Heart of the KGB Archives (1998); Nigel West, Venona: The Greatest Secret of the Cold War (1999); Lauren Kessler, Clever Girl: Elizabeth Bentley, the Spy Who Ushered in the McCarthy Era (2003); Tennent H. Bagley, Spy Wars: Moles, Mysteries and Deadly Games (2007) and John Earl Haynes, Harvey Klehr and Alexander Vassiliev, Spies: The Rise and Fall of the KGB in America (2009). Such books have done much to fill in the picture of Soviet espionage operations during the Cold War. They provide a background against which to read Andrew’s official history of MI5. 

The single most important thing to take from this background is that controversy over who was or was not a Soviet spy has a long and tangled history; hence the curious and catchy cover blurb on Costello and Tsarev’s book in 1993. Apart from the famous British cases, there is that of Alger Hiss in the US State Department. Haynes, Klehr and Vassiliev conclude that the declassified Venona decryptions and the Russian archives finally prove that Hiss was, indeed, a Soviet spy, but Hiss’s supporters in “liberal” circles in the United States have continued to argue that the evidence is circumstantial and the case unproven. Those wishing to acquaint themselves with the evidence should read Sam Tanenhaus’s summation of the case against Hiss at pages 515–20 of Whittaker Chambers. But in all this history, nothing is more evident than the fact that, over many decades, MI5 (and MI6 and Whitehall) were both very poor at detecting Soviet moles and spies in the UK (at least until after Hollis retired in late 1965) and extremely reluctant to admit it publicly when they did find them. You would never guess this from reading the authorised history, but the details are very telling and must be considered as a rather substantial caveat emptor when we are invited to swallow an official or authorized pronouncement on the subject. 

These disquieting realities were notoriously the case with the Cambridge Five: Philby, Maclean, Burgess, Blunt and Cairncross. That all five were recruited without any kind of vetting in the 1930s was only the start. They then worked within the British foreign policy, intelligence and security system for many years completely undetected and would quite possibly have remained undetected had it not been for the Venona decryptions in the United States, followed by the instability and indiscretions of Burgess and the flight of Burgess and Maclean in 1951. Even as it was, Philby was not definitively exposed until 1963, when he also fled to the Soviet Union; while Blunt and Cairncross were quietly pushed into the shadows by the British government until it became impossible to deny their treachery. Even then, no adequate accounting was offered of the nature of their betrayal of their country and its allies, to say nothing of the breathtaking scale of it. 

The scale of Soviet espionage in Britain—and in the West more generally—has been obscured over the years by a combination of official reluctance to admit its extent and left-wing denunciation of any steps in this direction as “McCarthyism”, “Red-baiting” and “witch-hunting”. The official reluctance stemmed from two sources: a determination to protect the sensitive intelligence on which any admissions would have to be based, and embarrassment at the manner in which its “firewalls” had been breached by the other side. The single most notable body of sensitive intelligence was the Venona program, which had enabled US intelligence operatives in the late 1940s to intercept thousands of KGB and GRU messages to and from Moscow between 1940 and 1948, of which more than 2000 were decrypted over a period of many years The interception program was disrupted when Kim Philby alerted Moscow to the fact that its codes had been cracked; whereupon they were changed. Until the Soviet Union collapsed and the KGB’s archives were partially opened, Venona was the best source available for monitoring what the Soviet espionage services had been up to in the 1940s. The intercepts that were decoded revealed the existence of Soviet espionage on a disturbing scale. 

It was Venona that pointed to Maclean, as well as to the espionage being conducted in the United States by such well-placed moles as Alger Hiss (for the GRU) in the US State Department, and Nathan Silvermaster and his KGB ring, including Harry Dexter White, in the US Treasury and Lachlan Currie in the White House. The Silvermaster ring, as Nigel West remarks in his history of the Venona program, had “tentacles in every part of Washington’s political life”. Yet all of them escaped prosecution—except Hiss’s 1949 prosecutions for perjury—because of the sensitive intelligence on which proof of their guilt was based. Hiss denied everything; White died of a heart attack after denying everything Elizabeth Bentley had said about him; while Silvermaster and his wife, also a spy, escaped prosecution by the simple expedient of taking the Fifth Amendment before Congress. Silvermaster and the key members of his ring were quietly moved to other jobs and the Left was, therefore, able to get away for decades with assertions that these people were the victims of a witch hunt. 

It never seemed to trouble those who cried witch hunt during the Cold War that in the Soviet Union itself at that time such people, had the boot been on the other foot, would have been tortured into making a confession and then executed. We now know that hundreds of thousands of people were executed during Stalin’s Great Terror on the grounds that they were “potential” fifth columnists in the event of a future war, by virtue of their class, family or ethnic background. This has always been the moral puzzle of “romantic” Western communists or self-styled “progressives”: that they would fiercely denounce any and all counter-intelligence work within the liberal democracies as paranoid or nastily repressive, while either apologising for or simply denying the reality of mass murder and brutally massive repression inside the communist world, especially in Stalin’s Soviet Union in the 1930s when so many of them signed up for the cause of “revolution”. 

Only after the Cold War was over was the Venona material made available for public consumption. But at the same time, the KGB opened its archives a crack; and Vassili Mitrokhin smuggled out thousands of pages that had not been declassified, which threw further light on the scale of Soviet espionage all the way back to the 1920s. Only then did it emerge that, in addition to the Cambridge spy ring, KGB spy rings had been recruited at Oxford University and the University of London at the same time, in the 1930s. The first reports out of the archives were able to show that these rings had existed and to cite the codenames of their members, but not their actual names. The Russians boasted that they had had spies who were “household names”, but would not reveal those names, as this would be bad tradecraft. It emerged, years later, that the head of the Oxford ring had been Arthur Wynn, codenamed Scott, who had been a senior civil servant for many years until his retirement in 1971. 

Christopher Hill, the famous Marxist historian of the English Civil War and the Levellers of the seventeenth century, it turns out, was also a Soviet spy in the 1940s. When confronted with the evidence in his last years, he confessed but pleaded that the truth not be made public until after he had died. He had concealed his membership of the Communist Party in order to get jobs in military intelligence (1940–43) and the Foreign Office (1943–45), and there he worked for the Soviet Union. Chapman Pincher missed the update on the identity of Scott, in 2009, since Scott’s identity remained unknown when Treachery went to press, and speculated that he may have been Christopher Hill. But he had, at least, picked up the scoop on Hill himself being a Soviet spy. And, as he remarks, “His employment is a further testimony of security incompetence by both the Foreign Office and MI5, because his membership of the Communist Party alone should have barred him.” At a time when vigilance against Nazi spies was both high and extremely effective, the watch on Soviet spies was so lax and ineffective as almost to defy belief. Roger Hollis was right at the centre of where that vigilance was supposed to have been exercised. 

Hill, as it happens, formed a partnership of pro-Soviet influence with Peter Smollett (whose real, Austrian name was Smolka), the head of the Russia desk in the Ministry of Information, who was himself a Soviet spy and a friend of Kim Philby. Smollett, it seems, persuaded a number of publishers to reject the manuscript of George Orwell’s Animal Farm before he was exposed, at which point he fled to the Soviet bloc. Hill himself, meanwhile, had written a kind of “Orwellian” work of pro-Soviet propaganda while working at the Foreign Office. It was a book titled The Soviets and Ourselves: Two Commonwealths, which Smollett helped him to get published after the war under the pseudonym K.E. Holme. The book sang the praises of Lenin as a kind of genius; asserted that Soviet citizens all enjoyed the freedom to vote; and described the Great Terror as a non-violent social movement akin to the Chartist movement in nineteenth-century England. This same man then went on to become a tenured historian, the author of influential books on the English revolution and Master of Balliol College at Oxford University. He died in comfortable retirement in February 2003. 

The scale of what all these spies accomplished for the Soviet Union was such that, even after a lifetime spent exposing and excoriating it, Chapman Pincher felt moved to astonishment and fury in writing Treachery in the past few years. The KGB archives, he discovered, reveal that Guy Burgess supplied 4600 documents to the KGB. Anthony Blunt supplied 1771 documents, among them, Pincher relates, “American secrets of the greatest sensitivity, including plans for the D-Day landings”. Donald Maclean supplied 4593 documents, which filled forty-five boxes in the KGB archives, each box containing 300 pages. John Cairncross supplied 5800 documents and Kim Philby 914; though Philby’s were often long and detailed and were supplemented by a good deal else in letters and oral reports in his frequent meetings with his KGB controller. As Pincher summarised the work of the Cambridge spies: 

‘The total take from these five spies alone—17,526 classified documents—is a measure of the number of their meetings with Russians and emphasizes the enormousness of their treachery, which was far worse than ever suspected, and more may yet be revealed. The argument by some of Hollis’s supporters that the traitors he missed did no great damage is blown away by these disclosures.’ 

One of the strangest aspects of Christopher Andrew’s authorised history of MI5, quite apart from its treatment of Hollis himself, is that it provides little more than a conventional account of the history of the Cambridge Five, no account at all of the Oxford or London rings (there is no entry in its index, for example, under either the codename Scott or the names Wynn, Arthur or Hill, Christopher) and does not even provide, as Pincher does, a quantification of the documents supplied to Moscow Centre by the Cambridge spies, to say nothing of their nature or the internal damage assessment done at MI5 after the spies were uncovered. Most troubling of all is the fact that the authorised historian does not appear even to have been curious as to why MI5 was so effective against the Nazis and so utterly ineffective at the same time against the communists. 

It must be allowed, of course, that in attempting to write a history of MI5 from its foundation almost down to the present, Andrew faced a formidable task of synthesis and had, inevitably, to make choices about what to include and what to leave out. One would have thought, however, that the abysmal failures of the organisation to prevent or detect communist penetration in the Stalin years and afterwards might have been included; if only because counter-espionage was the very raison d’être of MI5. What kind of authorised history is it that judiciously omits a reckoning with these dismal failures of forty to eighty years ago? 

In his acerbic remarks about Peter Wright to Peter Wilson last year, the author of Defence of the Realm stated that Wright had been guilty of fabricating and distorting evidence “to fit his conspiracy theories”; that he “tended to select a solution, then tailor the evidence to fit it”; and that “His standard manoeuvre when worsted in argument was taking refuge in mystery—‘If you knew what I know.’ This was later shown to be a dishonest charade.” Given that Andrew himself simply omits almost everything that points to MI5 incompetence or the very possibility of its Director General at the height of the Cold War having been a Soviet mole, one is left wondering whether this critique of Wright is not the pot calling the kettle black. The fact that Andrew had supposedly unfettered access to the MI5 archives ought to have allowed him to set out at least some of these sensitive matters chapter and verse. Not only does he not do this, but when he cites the archives—which he does many times—he provides no indication whatsoever of what specific file or document in the archives he has drawn upon. The endnotes, in every case, simply read “Security Service Archives”. Not only would this set considerable difficulties in the path of anyone with similar access trying to check his citations; it also means that we haven’t the slightest idea what kind of document he is drawing upon in making any given claim. 

An annoying and significant example of this, in his defence of Hollis, is his claim (at page 282) that Hollis had, in fact, a “remarkable insight into Soviet intelligence penetration”, as witness his alleged suspicion of Anthony Blunt. What are his sources for this idea? Of all people, Kim Philby and Anthony Blunt. Philby, he says, “later recalled that ‘Hollis was always vaguely unhappy about him’” (citation: Security Service Archives). And:

‘After Blunt eventually confessed to working as a Soviet agent, he told his interrogators, Peter Wright and Arthur Martin [both paid-up members of his “small but determined group of conspiracy theorists”], “I believe Hollis disliked me—I believe he slightly suspected me.” Blunt recalled one particularly dramatic example of Hollis’s suspicions. After Gouzenko had revealed the existence of an unidentified Soviet agent codenamed Elli, Hollis turned to Blunt and said, “Isn’t that so, Elli?”’ [citation: Security Service Archives]

The authorised historian comments, “It was sadly ironic that Wright and Martin, the most damaging conspiracy theorists in the history of the Security Service, should later persuade themselves that the unidentified Soviet agent was Hollis himself.” 

What calibre of historical analysis is this? To begin with, we are not given any clear idea of what kind of record in the archives has yielded these two little gems, or when precisely they are dated; but both would seem to be records of conversations without any other verification. We are not even told whether anyone other than Blunt witnessed this supposed remark by Hollis. Nor are we told to whom Philby made his observation, or in what context. In each case, we are asked to take at face value the alleged remark by an exposed Soviet spy who had made a career of lying to British intelligence and who, by all accounts, continued to lie every which way as long as they lived, that they thought Hollis had been suspicious of Blunt. Does it not occur to the historian to ask whether they may have been lying? 

Let’s ask the rather obvious question: Had Hollis been a Soviet spy, what would you expect two very well placed Soviet spies, under interrogation, if asked about Hollis’s relations with Blunt, to say about him? That he never suspected Blunt and appeared oblivious, also, to what Philby was up to, even though, while head of Soviet counter-espionage in MI5, he had regular dealings with each of them? It would have been far better, surely, to say that he seemed to have his nose ahead of the pack, thus diverting attention from him. Oddly enough, that is also what the authorised historian has now done. 

In short, whatever its other merits as a basic chronological history of the existence of MI5, Defence of the Realm fails entirely in what one might have thought would be the prime task of a history of MI5 in the wake of the Cold War: to fully explore and openly explain the catastrophic failure of MI5 with regard to Soviet espionage in Britain between the early 1930s and the mid-1960s, coinciding with Hollis’s career. Given his rather vehement emphasis on the mental capacities of Peter Wright and the innocence of Hollis, in public statements and in thinly documented claims in his book, it is more than a little unsettling to find that Christopher Andrew does not do anything of the kind. This renders the massive book he has written of very much diminished stature and significance. Chapman Pincher’s book, though far from authorised, is a great deal more thorough, forthright and revealing in this respect. And, whatever Andrew may say about Peter Wright, there can be no credible claim that Pincher is mentally ill or a mere conspiracy theorist. Indeed, he is more restrained than the authorised historian in actually drawing conclusions. Whereas Andrew, after a fatally flawed argument that totally omits to address the key problems in the case, declares Hollis innocent; Pincher, after amassing a great deal of circumstantial evidence that Hollis was a GRU mole throughout his career at MI5, concludes that we do not know for sure whether he was, because we do not have access to the GRU archives and Moscow is not telling. However, the Merkulov message to Stalin and Beria regarding Elli, combined with the demonstrable fact that Elli cannot have been either Blunt or Long, constitutes a very considerable problem for those, like the authorized historian, who wish to dismiss the charges against Hollis. 

Pincher draws attention, moreover, to a number of developments in Russia in this decade which bear on the question of whether Stalin had a highly-placed mole in MI5 in the 1940s. These developments centre on the manner in which Moscow got a copy of the Quebec Agreement, a highly secret two-page agreement between Churchill and Roosevelt, only sixteen days after they had signed it. The Agreement stated that the two great Anglo-Saxon powers would collaborate to produce an atomic weapon and that neither would communicate any information about this to any third party without joint consent. The document was very, very closely held, as testified by several documents in the British National Archives, showing that even in June 1949 Clement Attlee was concerned that it might be revealed to a US Senate committee and then leak; while in 1951, Harry Truman was still agreeing with Attlee that the document should remain secret. Leslie Groves remarks, in his memoir Racing for the Bomb, that “For several years after the war the existence of the agreement was not known either to the American Congress or to the British Parliament.” Little did Truman, Attlee or Groves know that it had been sent directly to the GRU in Moscow by Sonia from her illicit radio near Oxford on September 4, 1943. This was the kind of work, kept secret for as long as she lived, for which Sonia was decorated as a super agent of Russian military intelligence by Vladimir Putin. 

But Sonia was a courier, not a mole inside the British government. Moreover, she worked for the GRU, not the KGB, so her source was not one of the Cambridge Five, or the Oxford ring or the London ring. Can it have been Elli who supplied her with the details of the Quebec Agreement? Pincher points out that Sonia received a summary of the Agreement from a GRU mole, not a KGB one. If that mole was not Elli, then there are at least two still unidentified GRU moles from that period. He further remarks, “the private papers of Vyacheslav Molotov, Stalin’s most trusted deputy”, which were only released in 2006, reveal that on October 15, 1943, a GRU spy in Britain had revealed “details of the early plans for Operation Overlord, the Anglo-American assault on Normandy”. The GRU had reported to Molotov that “our source of trusted credibility” had supplied plans and maps of the operation and the intelligence that it would not be attempted before the spring of 1944. Pincher does not claim that Elli was the source of these plans, but he notes that they were passed to the GRU quite separately from Blunt supplying them to the KGB and comments: 

‘Sonia’s historic achievement is tantalizing proof that she had some prime, high-level British source whose name has still been withheld by the Russian authorities. When her [Quebec Agreement] coup was made public in 2002, in a GRU-sponsored book, The GRU and the Atomic Bomb [by Vladimir Lota], GRU colonel general Alexander Pavlov, who vouched for its authenticity in a foreword, was at pains to point out that “the time has not yet arrived when still unsuspected or unproven wartime sources can safely be named”.’ 

Pincher’s judgment is that Hollis, as Elli, while he may not have been Sonia’s source for the Quebec Agreement or the Overlord plans, may have made it possible for her to pass on these things and a great deal else without being detected. Hollis was not then at the top of MI5, as he was in the 1950s and 1960s, but head of Soviet counter-espionage, so he would not likely have had access to so closely held a document as the Quebec Agreement. This prompts the thought that the source of the betrayal was even more highly placed, not in MI5, but in Churchill’s immediate circle of advisers or the British chiefs of staff. How did such a source, described by the GRU to Molotov as “our source of trusted credibility” get a summary of the Quebec Agreement and the plans for Operation Overlord to Sonia? Pincher draws a circle around Hollis by arguing that it was Hollis’s job to prevent such massive breaches of security; that Sonia was the channel for the despatches to Moscow and that Hollis marked all Radio Security Service expressions of concern about the illicit radio transmissions outside Oxford, “No Further Action”. Perhaps he was, at the very least, her protective hand inside MI5. 

In Chapter 63 of his book, “The Penkovsky Problem”, Pincher tackles what has long been one of the key objections to the case against Hollis: the claim that, had he been a GRU agent, he would have warned Moscow right from the start, in 1960, that the GRU’s Oleg Penkovsky had turned and was spying for Britain. Hollis was, by then Director General of MI5 and insisted on being told the name, as distinct from the codename, of the spy inside the GRU. Hollis’s supporters, as Pincher comments, have maintained for many years, that “all the evidence against him, however suggestive, can safely be dismissed, because if he had been a spy, he would immediately have warned Moscow and Penkovsky would never have been allowed out of Russia a second time”. Yet he was allowed out and continued bringing a high volume of intelligence to MI6. It turns out, however, that the KGB had known of Penkovsky’s treason at least sixteen months before he was arrested on October 22, 1962, and that he had deliberately been allowed to continue his spying, under surveillance, in order both to trap him and to protect the mole in the West who had betrayed him. It was the KGB chief Vladimir Semichastny who revealed to the GRU chief Ivan Serov, in April 1962, that Penkovsky was a spy for MI6. Penkovsky was given a show trial and executed in May 1963. 

Here, as with the Quebec Agreement, however, Pincher has a technical problem of which he seems unaware. If Hollis, putatively a GRU mole, was the one who revealed the treason of Penkovsky, why would he have revealed it to the KGB and not to the GRU? Why would Serov have had to find out from Semichastny, many months later, that he had a high-level British spy in his organisation? The fact that the KGB had discovered him so early, by their own account through a foreign source that they wished to protect, as former KGB deputy chief of foreign intelligence Vitaly Pavlov stated in his 1996 memoir Operation Snow, does suggest a mole, but it isn’t clear that that mole was Hollis. Is it possible that Hollis, in order to lay a smokescreen, passed the information to Moscow via the KGB, while withholding it from his GRU control in London? One would not have thought so. 

Significantly, Tennent Bagley, in Spy Wars (2007) argues that the blowing of Penkovsky pointed to a “deep penetration of CIA or MI6—a mole knowing what only a handful of our officers knew”. He makes no mention, in this context, of MI5 or of Hollis. Christopher Andrew, on the other hand, true to form, makes no mention even of Penkovsky, whether to use his fate as proof of Hollis’s innocence or even as a dramatic episode in the history of the case. In short, Penkovsky’s fate would appear to demonstrate neither Hollis’s guilt nor his innocence. Possibly he played his hand very quietly in this matter, given how senior he was by then. We have no direct evidence. By Bagley’s account, the mole would appear to have been someone else—who has never been discovered. 

The Quebec Agreement and Penkovsky cases demonstrate that the circumstantial evidence against Hollis needs to be analysed carefully if we are to form a judicious opinion as to whether or not Hollis was a GRU spy. That Christopher Andrew addresses none of this evidence, even to make the kind of points I have just made, makes his authorised version of the matter all but useless. Pincher, on the other hand, deserves credit for doggedly trying to pull the complex and disturbing story together, even if at times he fails to dot the i’s and cross the t’s. Among the pieces of the puzzle, which he lists at the end of his book as “The Scroll of Anomalies”, there are a number which it would, to say the least, have been helpful for the authorised historian to have addressed. Seven particularly stand out as casting a dark shadow over Hollis’s career. 

First, the positioning of Sonia close to MI5’s wartime headquarters and Hollis’s dismissal of Radio Security Service concerns about what turn out to have been her transmissions to Moscow. Second, Hollis’s omission of all members of the Kuczynski family from a list requested by the FBI of German communists living in Britain, although Sonia, her sister, her brother and her father had already been described in MI5 files by then as dangerous communists and were all living in Britain. Third, Hollis cleared Klaus Fuchs six times, despite warnings from others that Fuchs could be a Soviet spy, enabling him to pass the basic technical secrets of the atomic bomb to Russia. Fourth, Hollis prompted the promotion of Philby in MI6 and worked closely with him for many years. Fifth, throughout Hollis’s career, MI5 secured no Russian defectors of consequence. After he retired, evidence of penetration of MI5 ceased and important defectors were secured. Sixth, though one would never guess this from reading Defence of the Realm, Hollis derided Gouzenko as a source of information (something Philby also did in such cases) and never seriously investigated the Elli claim. According to Pincher, though most of MI5’s papers on the Gouzenko case have been released, Hollis’s reports on it have been withheld—if they still exist—after sixty years. Seventh, Hollis did all he could to ensure that Blunt, Long, Cairncross and other exposed Soviet agents would never be prosecuted and that their legal immunity would be extended to anybody else they named as agents or assistants. 

Pincher supplies other details which can only add to the suggestion that something was seriously awry in Hollis’s vicinity right up to the end of his twenty-seven years in MI5. At a time when Hollis knew he was being investigated, in 1964, a second edition of Handbook for Spies, an MI5 production, appeared for no apparent reason. It had a new introduction which contained deliberate false statements regarding the then whereabouts of Sonia and her sister Brigitte, which would have discouraged anyone from trying to find and interview them. When the Fluency committee, which was investigating Hollis at that time, put together a chart of MI5 operations that had failed and of Soviet operations that had succeeded, Hollis’s fingerprints were all over both. Finally, the number of proven Soviet spies and active agents who operated for years under Hollis’s nose, when his chief task was detecting and neutralising such people, was more than thirty, excluding actual Soviet intelligence personnel working in the Soviet embassy. And the number of previously unidentified Soviet agents that are being revealed as the Soviet archives open is steadily increasing. 

All this, surely, demonstrates beyond reasonable doubt that there is a case to answer with regard to Hollis. Christopher Andrew treads a well-worn path in failing utterly to address that case, much less refute it once and for all. He declares that official inquiries have long since cleared Hollis, but provides not a scintilla of evidence as to how this was done, despite his vaunted access to Security Service Archives and other official documents. This is, frankly, an insult to serious readers of the history he has written, which, in this regard, deserves to be consigned quickly to the waste paper bin. 

That this does not mean Peter Wright was correct in any or all of his claims in Spycatcher should go without saying. That Chapman Pincher could still be in error he himself allows. But one is left with the uncomfortable sense that the self-protective instincts of both MI5 and Whitehall, which can be traced back over decades in these matters, have once again resulted in a bowdlerised and misleading history, which does nothing to lay to rest the ghosts of the Cold War, but leaves them roaming about restlessly. That is both a great pity from a purely historical point of view and a grave disservice to all those professional intelligence officers who worked thanklessly and, for the most part, anonymously for the West during that long struggle. 

All of this ought to be of more than passing interest to Australians, because in our own provincial way we were also a playground of undetected and unpunished Soviet spies during the Cold War and we do not have anything like an adequate history of these matters as yet. It remains of interest that when, in 1948, evidence of Soviet penetration of the Australian Department of External Affairs led to serious security concerns in Washington and London, MI5 sent Roger Hollis to Canberra, as its authority on protective security and on communism, to help investigate the problem and set up what became ASIO. It was MI5 Director General Percy Sillitoe who instructed Hollis to undertake this work. Hollis, as Christopher Andrew phrases it, provided “detailed advice on the charter, organization and senior personnel of ASIO”. 

Pincher then observes that up to 1948 the KGB traffic to and from Canberra had used a code system that had been deciphered, “but shortly after Hollis had returned to Britain in 1948, it was changed and decipherment was never possible again”. As Christopher Andrew himself records, Hollis urged that none of the dominions undertake their own study of the Soviet intelligence service, since that would be “redundant” and could be left to London. Hollis was closely involved with Canberra’s efforts to trace spies revealed by Venona, but arranged that Canberra would not embark on any new line of enquiry without first checking with MI5. How very useful for Moscow, if Hollis was Stalin’s man in MI5. 

One of the great mysteries of the two fat volumes on the Mitrokhin archive that Christopher Andrew authored between 1992 and 2005 is that they tell us next to nothing about KGB (to say nothing of GRU) operations in Australia. Nor has any other account about those operations surfaced, whether based on the Mitrokhin archive or any other archival sources. To the best of my knowledge, in the case of the Mitrokhin documents this is because the files on Australia were suppressed in 1992 at the request of the Keating government, and have never been released for scholarly or journalistic analysis. This speaks to a self-protectiveness on the part of the Australian Labor Party and ASIO which is every bit as dubious as that in Britain. 

The irony of such dishonesty about failures and treason, of course, is that it feeds the very kind of conspiracy theory that Christopher Andrew derides and bemoans. The unresolved and strange case of Roger Hollis is a symptom of this problem. Defence of the Realm had an unprecedented opportunity to address the problem candidly and do some good. It is deeply regrettable and indeed unacceptable that it failed to take that opportunity. The very least that should be done in Australia itself is now to insist on the declassification, without redaction, of the Mitrokhin documents on Australia, going all the way back to the 1930s.


Paul Monk is a frequent contributor to Quadrant. His recent books Sonnets to a Promiscuous Beauty and The West in a Nutshell are published by and available from the Canberra publisher Barrallier Books (www.barrallierbooks.com).


Leave a Reply