When I read Patrick Marnham’s War in the Shadows this spring I noted, but did not pay enough attention to, a remarkable passage where he describes the activities of the wartime TWIST committee. This group was an offshoot from the London Controlling Station, commanded by Colonel Bevan, and its role was to act as an inter-departmental committee for deception. Among its members (according to Marnham) were T. A. (‘Tar’) Robertson of B1A in MI5, responsible for handling ‘double agents’, Frank Foley of MI6 (SIS), and Anthony Blunt of MI5, and the document exploited by Marnham indicated that Foley’s duty was ‘the transmission of disinformation to the enemy through double agents of the Secret Intelligence Service abroad’, and that regular meetings were held with operational agencies such as SIS and SOE.
Now I see several extraordinary aspects of this disclosure. The casual reader might look to authorised histories of intelligence in World War II to learn more about this clearly important and influential committee. No entry for TWIST appears in the Index of Christopher Andrew’s Defend the Realm: likewise it is absent in Keith Jeffery’s Secret History of MI6. Michael Howard’s Volume 5 of British Intelligence in the Second Word War carefully avoids any coverage of the committee: for a moment, Volume 4, by Henry Hinsley and C. A. G. Simkins, (Security and Counter-Intelligence) raises hopes with an Index entry for ‘Twist’ – but it turns out to refer to a double agent with the cryptonym Twist who offered his services in Istanbul. William Mackenzie’s Secret History of SOE is silent on the issue: neither M. R. D. Foot’s SOE or SOE in France has anything to say on the TWIST committee that presumably managed the organisation’s double agents in France. J. D. Masterman barely recognises the existence of MI6-controlled agents in The XX System, and certainly does not identify the TWIST committee, the meetings of which he attended.
Only in Roger Hesketh’s Fortitude: The D-Day Deception Campaign, written in 1945, but not published until 2000, posthumously, can one find an authoritative statement about TWIST. Hesketh explains that the committee was set up, soon after General Frederick Morgan’s appointment as COSSAC, Chief of Staff to the Supreme Allied Commander (designate), in April 1943, to translate the requirements of Morgan’s planning staff for the COCKADE deception campaign. A Section known as Ops (B), under Lieutenant-Colonel John Jervis Read, was formed to deal with deception, and Hesketh was given the task of managing ‘controlled leakage’. Hesketh himself presented the requirements to the TWIST committee. Bevan dissolved the committee in January 1944, replacing the arrangements with direct co-operation with B1A of MI5. Nothing more was said about its weekly meetings: when COSSAC was merged into SHAEF, the gatherings of TWIST ‘no longer provided an ideal solution’.
So where did Marnham derive his valuable information? The answer is: documents purloined by one of the NKVD’s agents, Anthony Blunt. Marnham gives his source as Triplex: Secrets from the Cambridge Spies, edited by Nigel West and Oleg Tsarev, published by the Yale University Press in 2009, pp. 274-278. And indeed, one can find there a comprehensive description of the Committee, its members, and its activities. When I read this book, many years ago, I can see from my annotations that I was focusing on other matters not concerned with double-cross, While it is worth noting that the transcriptions, as they appear, are translations back into English from the versions in Russian, there is no doubt in my mind, from the level of detail, that the documents are authentic, and that they shed some remarkable new light on a topic that the British authorities have tried to conceal.
A delicious irony is evident here: the UK authorities managed to conceal from the British public the facts about the short existence almost seventy years ago of a committee that presumably abetted one of the great successes of British counter-intelligence – the Double Cross Operation. Yet it was revealed by one of the greatest espionage successes of the twentieth century – the Soviet infiltration of British government institutions. Fortunately, no Soviet colluder shared the information gained with any Nazi intelligence officer – at a time when the Soviets and the Germans were stated to be holding tentative peace talks in Sweden. And, for all the attention paid to the Freedom of Information Act in the UK, the details of TWIST became available only through the casual release of records of a highly secretive and hostile foreign intelligence organisation.
The facts must have been concealed because they were devastatingly embarrassing. The primary ‘double-agent’ management of that period (1943) was the attempt to use the SOE air traffic coordinator Henri Déricourt as a vehicle to pass deceptive messages concerning the timing of an assault on the Northern French coast to the Sicherheitsdienst in Paris, with whom Déricourt admitted he had been in contact. SOE was not represented on the Committee, however. Major Foley, working for Claude Dansey, represented SOE’s ‘interests’, and indeed SIS was responsible for managing German spies who had been detected and reportedly ‘turned’. Thus the charter of the TWIST committee, and its representation, explain the statement made by Nigel West (at which I expressed my surprise a couple of months ago) that SOE’s ‘double agents’ were managed by MI6.
Yet Dansey and his crew had no idea what it meant to ‘turn’ an ideologically committed enemy agent (if that indeed is what Déricourt was). The eventual outcome of that process was the destruction of the PROSPER network in France. It is small wonder that MI6 and MI5 (whose emissaries Masterman and Robertson must have looked on in amazement) wanted to eradicate any traces of the committee from the historical record. Moreover, it is astonishing that neither Hambro (as head of SOE), nor his deputy, Gubbins (director of operations) nor the Minister responsible (Lord Selborne) ever challenged or questioned this extraordinary set-up. Why did they not demand direct representation on the Committee? In the decade since the Triplex papers shed light on Hesketh’s disclosures, no one in authority appears to have picked up the dramatic anomalies of the revelations on the TWIST committee and incorporated them into an ‘official’ history.
I describe this phenomenon as a way of introduction to a transcript of some recent exchanges on the demise of PROSPER that I have recently enjoyed with Robert Marshall, the author of All The King’s Men, as well as some observations by Patrick Marnham (the author of War in the Shadows [WITS]), my reading of which initiated this whole investigation. All the King’s Men (subtitled The Truth Behind SOE’s Greatest Wartime Disaster, and identified hereafter as ATKM) was published in 1988, and Mr. Marshall retains a very active memory of his interviews with some of the participants during that decade.
Close watchers of coldspur may recall that Marshall originally posted the following comments on July 21, in response to my piece Claude Dansey’s Mischief:
I might just add a small postscript to these pieces;
While I’m delighted that Markham follows fairly closely the outline of events as set out in ATKM (1988), both he and Suttill and a great many others, assume Dansey recruited Dericourt as a way of executing STARKEY [the part of the COCKADE plan that framed an amphibious landing in northern France] and smashing SOE.
Déricourt was recruited in September 1942, before STARKEY had been devised. He was recruited because of his links to a senior figure in the Paris SD. Dansey wanted eyes and ears inside the SD. To do that role, Déricourt had to have real intelligence to pass on to the SD and Dansey calculated that anything Déricourt knew about SOE operations was expendable.
By the time COCKADE – STARKEY had been developed, it was clear that Déricourt’s relationship had provided the SD with so much information on the PROSPER (and other) network(s) that it was about to collapse.
There was a clear understanding within the Twenty Committee that doomed SOE networks should be exploited to promote a deception plan. It’s in the XX papers.
So, the deliberate exploitation of the PROSPER network was an opportunity that arose; it was not planned from the outset.
The running of that operation was typical of Dansey, who saw opportunities, played outside the rules and cared little for the consequences.
Déricourt’s role became redundant when his contact within the SD, Karl Boemelburg, was reassigned and sent down to Vichy.
I thanked Marshall publicly as follows:
Thanks so much for your posting, Robert. Good to hear from the author of ATKM.
I have read much more since my original report (your book, Suttill’s, Marnham’s biography of Moulin) as well as re-read Cave-Brown, Hesketh, Mackenzie and Howard.
One thing that struck me was that the PWE/SOE deception plan for STARKEY was not delivered until July 18, 1943, i.e. three weeks after Prosper had been arrested. Thus, if Dansey was engineering the sacrifice of the network by agitating Suttill, it was all being done unofficially and prematurely. Why does Howard, in his coverage of the ISSB, never mention Déricourt’s role as a double-agent? And, if the decision was made to exploit ‘doomed’ networks (by the LCS?), how did the committee know which were ‘doomed’, and did they care how they had been destroyed?
Your further thoughts welcomed.
I then invited Robert Marshall to communicate with me further off-line, if he found that method easier. (I have made minor editing changes in the following transcripts of the messages, for reasons of accuracy, clarity, continuity, and focus. In one or two places I have inserted parenthesized unitalicised comments to correct or clarify.) Accordingly Marshall wrote to me on July 22:
It took me a while to find your comments, but happy to offer some feedback. I never thought I’d be back into this material again, but it does appear to be of perennial interest.
I have most of the COCKADE papers, and it’s clear that the STARKEY element is being formulated before the official plan was circulated, as you say – July 18th. However, that’s by-the-by.
From interviews I conducted back in 1987, the evidence about what was going on is fairly overwhelming. Before conducting these interviews, I decided that I would not explain to any of the contributors much of the background. I didn’t want anything they said to be influenced by a ‘context’ which I may have inadvertently provided. This had consequences, both good and bad. For example, Robert Cecil (former FO official) was happy to mouth-off about Dansey, whom he had disliked intensely, until he saw the program and realised how his remarks were used; at which point he dashed off a bitter complaint to the BBC. Everyone who spoke to me had no sense of the whole story, only their specific chapter.
Some critics of ATKM built exaggerated straw dogs to make it easier to attack the book’s central premise; saying that I claimed the PROSPER collapse was Dansey’s/ MI6’s plan to destroy SOE, and that COCKADE had been devised as part of that heinous plot / conspiracy. These attacks came from those with deeply vested interests: e.g. Overton Fuller’s relationship with Dericourt, Suttill’s with his father etc. But ATKM makes no such claims. The PROSPER collapse was simply a consequence, albeit a foreseeable one, but not the ultimate aim.
I’ll give you some examples of un-contextualised sources:
In Overton Fuller’s interviews with Déricourt, conducted around 1957/8, he describes his visit to London during Easter of ’43, during which he reported to a senior officer of ‘another organisation’ that PROSPER was deeply penetrated. This suggests that Déricourt himself may not have appreciated the likely consequences of his ‘mission’, and in any case, agents were usually briefed on a need-to-know basis. No context. I was criticised for quoting Déricourt himself, but my point was that in the 1950s Overton Fuller and most other historians knew nothing about the separation between SOE and MI6 and absolutely nothing about COCKADE. Déricourt’s remarks, at that time, are both naive and revealing.
The XX Committee’s decision to exploit a ‘fatally penetrated PROSPER’, was seen as just another vector through which to prosecute COCKADE.
Suttill was recalled to London in June and from interviews with a number of his former colleagues, it’s clear that he returned with the date of the invasion. Not the anticipated July or August, but September. Suttill seemed focused, but stressed; as you would be if you knew you had to hold out for another 3 months. Everyone I interviewed said they fully expected an invasion in the summer of ’43. Who briefed Suttill that it would be September? Who can say? Buckmaster told me Suttill had been summoned to the War Office during his June visit, after which he, Suttill, confided the September date. So it seems Buckmaster was also exploited.
Harry Sporborg told me that a number of people at his level were made aware of deception plans which would exploit SOE networks. He added that Buckmaster would never have been briefed.
The arrest of Suttill and wrapping up of the networks was forced upon Boemelburg, who hadn’t wanted to move until the very last minute. When was that last minute? Had he already been given the September date by Déricourt? We can only speculate. (It was the arrests of Cullioli and others that forced B to move on PROSPER.)
Déricourt’s flying visit to London, after Suttill’s arrest, has also been much disputed. And yet Foot describes it in his book, having spoken to people with whom Déricourt stayed. Foot also revealed that no one at Baker Street knew of that visit. Rémy Clement told me the story of Henri suddenly climbing into a Lysander with two other passengers: it was a moment seared into his memory, because they had a big Hudson reception to do a few days later. Foot and Hugh Verity told me that Déricourt returned to France on an RAF flight, ‘for another organisation’, which RAF records confirm was MI6. Verity also confirmed that “no one could get onto an MI6 flight without MI6 approval.” Verity wasn’t aware of the significance of that remark, however Foot clearly would have been.
In short, I am certain Dansey simply saw SOE and its operations as expendable. He wouldn’t have wasted effort plotting its downfall; he simply didn’t care about it. “Sabotage won’t win the war, intelligence will” he once said. Déricourt was just one of many opportunists whom Dansey scooped up and used in his war against the SD, especially after Venlo.
Had Déricourt already been recruited by Boemelburg before he left for England in August ’42? He certainly knew Boemelburg (as did Bodington) from before the war, but I really searched hard and found no evidence. On the contrary. Boemelburg didn’t run agents abroad, he was purely counter-intelligence.
Between the fall of France and August ’42, D was certainly active in the black-market, and as a pilot he may have done a bit of courier work for ready cash. But until he met Doulet and learned of the opportunity to get to England, I really don’t think he would have been of interest to anyone. I found no evidence that D had any contact with a German intelligence service that ran agents abroad. In any case, he would have had to have been trained beforehand and provided with means to contact his controller. It didn’t happen. Just because he’d jagged a ‘ticket’ to England didn’t mean he would end up in a position that would have been useful to German intelligence. Going to England was a risk D took, knowing that he could easily be interned for the duration. The only asset he had to wave at the Brits was Boemelburg. There is evidence from Buckmaster and D’s SOE file that he made mention of his German contacts.
Which brings us to MI5’s role. The MI5 files make it clear they were deeply suspicious of D and his colleague Doulet. The two pilots were separated a few days after their arrival, and more than two months later D’s name comes to the attention of SOE. SOE couldn’t believe their luck, put him through a streamlined training course and sent him into France asap. MI5 still hadn’t given him a clearance, and continued to have their doubts months into D’s operation. I have looked hard at the LCS and COCKADE papers, but I saw no evidence that TAR Robertson was aware of D. I suppose he may have been aware that Dansey was running him, but I doubt Dansey would have shared that with anyone. No one knew much about his operations, even those who worked alongside him.
MI5 finally gave Doulet a clean bill of health, but their suspicions about D persisted. Which takes us into the realm of how D was protected. Bodington was definitely tasked with that; he spoke on D’s behalf time and again, and the entire trip to Paris to report on the PROSPER collapse was an exercise in casting suspicions away from D. As I pointed out earlier, D’s ‘mission’ came to an end because Boemelburg was reassigned to political duties in Vichy. Dansey would have preferred to keep D in place right up to D-Day, but he didn’t have the kind of relationship with Kieffer, Götz, or any other SD officers that he’d enjoyed with B. D’s value had diminished.
Once again, Sporborg confided that after they recalled D and investigated the claims against him, he and John Senter concluded that D must be working for MI6. He added that they knew of others inside SOE (Bodington) who were working for MI6.
I hope that’s of interest.
P.S. I hadn’t meant to cast doubt on Markham’s discovery of T.A. Robertson’s handwritten reference to Déricourt, only that I hadn’t found any evidence of MI5, the LCS or XX Committee ever referring to him. I still maintain that Dansey kept all his agent’s [agents’?] details to himself.
After a few days, and having consulted Patrick Marnham, I replied as follows:
Thanks again for your very thorough commentary. I am enthusiastic about moving the debate further, but I must bring up the matter of methodology. In my researches, I have been very careful to try to ascribe agencies and dates to every action or reaction encountered: thus I avoid the passive voice, and question undated assertions. I also treat very carefully oral testimony, as my investigations into the behaviour of MI5 officers (e.g. Dick White), and eminent persons (e.g. Isaiah Berlin) have taught me that they cannot be implicitly trusted. Any volunteered statement by a participant in the intelligence business has to be challenged in the expectation that the person is trying to frame his or her legacy to look more respectable than it probably was. (Would you agree with that approach?) So, having cleared the air, I respond to your various points in sequence.
By the way, I did send your email to Patrick Marnham, and he has given me some preliminary observations, which I reflect in my comments. He is very pleased that you have joined the discussion.
1. The COCKADE plan: I don’t think that the question of an early STARKEY venture is ‘by-the-by’. Of the traditional historians, Hesketh (and he alone, apart from Thaddeus Holt, it seems) refers to the TWIST committee of the LCS (which he led) that formulated deception plans before the official COCKADE project. (Marnham gets his information on TWIST from files purloined by the NKVD!) No doubt the use of the ‘double-agent’ Déricourt was part of this initiative, and it was undertaken outside the deliberations of the XX Committee, which, so far as I can ascertain, never formally recognized the use of Déricourt in their proceedings. I would imagine that Bevan’s decision (in January 1944, when COSSAC was merged into SHAEF) to close down the TWIST committee, and have Ops (B) work with B1A directly, might have derived partly from the disastrous experience with Déricourt. I note that Howard has nothing to say about the TWIST committee, which is unpardonable.
2. The details on the TWIST committee, reproduced in Triplex by Nigel West and Oleg Tsarev (pp 274-278), should cause the history of COCKADE and deception to be re-written. Apart from showing that Churchill’s attempts to conceal the lack of invasion plans in 1943 to be utterly bankrupt (since Anthony Blunt was a member of the committee), the lack of reference to it by Howard and Masterman, and only cursory attention by Hesketh, must indicate that its proceedings and decisions were a subject of great embarrassment. If Robertson and Masterman were aware of the attempt to use Déricourt as a controlled agent, they were derelict in their duty by not calling him out. If they were not aware, then Foley (the SIS representative on the Committee) and his boss (presumably Dansey) somehow managed to conceal their stratagems. The fact that Foley’s duty was ‘the transmission of disinformation to the enemy through double agents of the Secret Intelligence Service abroad’ suggests the former. Marnham is of the opinion that Dansey was beyond control, since he had all sorts of activities underway that official channels never detected. That makes sense, and newcomers like Robertson and Masterman would probably not have had the gall to challenge him.
3. In that regard, Masterman’s comments on STARKEY, and COSSAC’s role in planning deception, are, I believe, very relevant – and provocative. He states, erroneously, that COSSAC was formed by General Morgan only in June 1943, while other sources indicate that Morgan was appointed on April 26. Morgan then had to assemble his staff: Howard writes that the first COCKADE plan was not forged until June 5. (Thus Marnham’s claim that COCKADE was ‘under way’ by May 1943 cannot be strictly true, although various initiatives deriving from the original blueprint in January may have been given the ‘COCKADE’ moniker.) Masterman then suggests that the Controlling Officer of Deception (Bevan), who authorized operations in outline, was then able to leave the detailed planning to COSSAC, and then, later, to Ops. B. SHAEF. It would appear that Masterman was attempting to bury the activities of the TWIST committee. Indeed, he also writes that ‘the agents survived with undiminished prestige ready for the more important operations of the future’, clearly overlooking the role and demise of Dansey’s ‘controlled agent’ Déricourt. Hesketh (note 3, p 29) indicates that technical control of agents remained with MI5, but that was an equivocal statement, and was clearly not completely true if Dansey was running a cowboy operation through Suttill because of his access to TWIST, presumably, and his knowledge of evolving STARKEY objectives.
4. I agree that Suttill grossly distorts and undervalues your thesis concerning Dansey (I cannot speak on Overton Fuller, since her books are practically inaccessible for me). I agree that it makes no sense to suggest that COCKADE was devised as part of Dansey’s plot, as it had its origins in much more serious war bodies, starting with Churchill’s rather capricious call for deception and ‘insinuation’ on April 18 (Howard, p 74). Dansey presumably felt free to insert his own inspired projects. But the statement that ‘the PROSPER collapse was simply a consequence, albeit a foreseeable one, but not the ultimate aim’ raises further questions. What caused the PROSPER collapse? Careless practices (meetings, poor wireless usage) or betrayal (from insiders or from Déricourt)? If it was ‘foreseeable’, who foresaw it, when? If it was predictable at some stage, why did no one do anything to salvage it, or close it down? And, if Dansey was conducting deception plans that accelerated the demise of PROSPER before the official COCKADE plan was approved, and PROSPER was destroyed before COCKADE was approved, at what level were they authorized? The network’s downfall has to be described in completely different terms.
5. The critical events seem to have taken place in mid-May, when Suttill returned to London. I learned from Marnham that in those fateful few days, Sporborg, Gubbins & Boyle were told by SIS that PROSPER was penetrated, Buckmaster was told that PROSPER would be sacrificed for strategic deception [but see correction later], and Suttill was told that French landings might take place sooner. Marnham has since told me that he derived those accounts from you, in ATKM, so we are both keen to know whether you have any supporting documentation beyond what Sporborg told you. The only explanation for such happenings would seem to be that Dansey heard about the incipient STARKEY plans, well before the TWIST committee had finalised and submitted them for approval, and used his insider information to apply the coup de grâce to the PROSPER network.
6. It occurs to me that the instructions given by the Chiefs-of-Staff to SOE on March 20 were ambiguous, conflicting, and rather naive, and that Dansey may have been able to exploit them. They clearly set France as a lower priority than the Balkans, and insisted that SOE should yield to SIS priorities there. Hambro, in his response, made requests for more air support based on the dangerous idea that shipment of materials was somehow useful in order to maintain the morale of French resistance forces (even though they would explicitly not be able to use them until 1944). In any case, the Chiefs of Staff declined Hambro’s request on July 27, stressing that SOE operations in the Balkans should be supported at the expense of supplies to the resistance groups in western Europe. Somehow, the extensive parachuting of supplies that took place early that summer must have taken place without SOE’s fully liaising with the Chiefs. Mackenzie wrote that ‘up to June 1943 the whole Suttill circuit had received 254 containers of stores, and in ten days in June it beat all records by receiving 190 more containers’. How does that tally with the lower priority given to France in March, and who authorized the shipments?
7. Yet, while Dansey informed Hambro that PROSPER had been penetrated, he surely did not disclose to them how it had been undermined. The plan to ship arms to the PROSPER circuit as a key deception tactic, however could not have been achieved without the approval and co-operation of the Chiefs of Staff and Bomber Command (because of the need for planes). It appears to me that the whole military command was in over its head, not understanding the implications of providing weaponry (as opposed to ingredients for sabotage) prematurely to a civilian army under severe stress. Did Hambro lose his job over this, I wonder? The conventional accounts imply that he was fired for withholding information on Yugoslavia from Selborne, or disagreeing with policy in the Middle East, but it could well be that he was forced to walk the plank because he promoted the shipment of weaponry campaign before the COCKADE plan had been developed fully, and without performing due diligence on how the PROSPER network had been penetrated. I notice that SOE was not represented on the TWIST committee. As Marnham points out, SOE was allowed access to information about deception operations only on a ‘need to know’ basis, which means that SIS represented SOE’s interests entirely. Why did Dalton and Selborne put up with this?
8. Your comments on Déricourt’s testimony (‘both naive and revealing’) are very apt, I think. But can we trust what he said to Overton Fuller about his trip to London at Easter 1943, where he claimed he told a senior officer ‘of another organisation’ that PROSPER was deeply penetrated? That was either an obvious feint (‘see how attentive I was in informing my bosses what was happening’), or an indication that he was following Dansey’s instructions, and, having ensured that PROSPER was indeed deeply penetrated – by himself – was reporting ‘Mission Accomplished’ to his real controller.
9. Where does the expression ‘fatally penetrated PROSPER’ occur? And when did the XX Committee decide to exploit it? If the Committee did indeed consider exploiting it, it would imply that they knew how it had been penetrated, and would have had to use Déricourt to pass on deceptive messages, but there is no record (so far as I know) of their doing such, and they would never have approved of Déricourt as a controlled agent. They knew in April 1943 (via Masterman and Robertson) that Déricourt was being used as an agent, but it was not under their control. Moreover, since PROSPER was effectively destroyed by June 1943, it could never have played a role in the final COCKADE plans. So the claim that PROSPER was ‘just another vector through which to prosecute COCKADE’ does not make sense to me. (Is this perhaps just an example of confusion over terminology?) In July, Portal had received the report from SIS (then unseen by SOE, although the events of May indicate the SOE chiefs were informed orally) that described the serious penetration of SOE’s French networks: it insisted that ‘resistance groups are at their lowest ebb and cannot be counted on as a serious factor unless they are re-built on a smaller and sounder basis’. So why would the XX Committee see potential in PROSPER?
10. I have always interpreted ‘September’ as occurring within ‘Summer 1943’. As Marnham has pointed out to me, the important point is surely that Suttill was mis-led about landings in 1943. The notional date of the invasion was always September 22, as Howard records, and it was broadly known, even to a select number of US officers. Why do you write ‘anticipated July or August’? Who expected that, and where does it appear? Can you trust what Buckmaster, who knew in mid-May that PROSPER would be sacrificed for strategic deception, said? [But see later correction to this misunderstanding.] Likewise, did Sporborg really know what was happening, or was he covering for Buckmaster? And, if Sporborg and others were informed about deception plans exploiting SOE networks [plural], who told him, when? Did he object?
11. (Marnham’s comments on the arrest of Suttill being forced upon Boemelburg) “I don’t agree with this, for reasons I have given in the book. We know from French and German records that Boemelburg in 1943 was under great pressure from Berlin to produce results a.s.a.p. Furthermore I don’t think Henri Déricourt would have been given a ‘D-Day date’ in 1943. At that stage, HD’s role was to undermine and deliver PROSPER. The same phoney invasion scenario had already been fed to both Major Suttill and Jean Moulin. Once arrested there was a reasonable chance that one or both would encourage German fears of imminent landings.”
12. Your summarization of Dansey’s perspective, and its ‘expendable’ operations strikes me as half-true. If he truly thought that intelligence (and not sabotage) would win the war, he would have regarded SOE’s interference in wireless-work, recruiting resisters, drawing Gestapo attention, etc. as a possibly serious threat to his own intelligence-gathering, so I don’t think it is correct to state that he didn’t ‘care’ about it. And the Chiefs of Staff supported that view. But for him aggressively to help ruin a resource that COSSAC would eventually be relying on suggests to me that he was out of control, and that LCS & TWIST were unaware of the circumstances in which Déricourt had been recruited. ‘Scooping up opportunists’ (and who were the others?) in a perceived war against the SD (by whom he had already been outmanoeuvred) does not indicate the behaviour of a seasoned intelligence professional. And the whole plan of using resistance fighters in a dummy deception exercise, when their goodwill would be required later, was a disaster, as Cave-Brown and others recognized.
13. Yes, it is difficult tracing the date when Déricourt officially became a SD asset. But he was given a Vertrauensmann number – BOE.48, as well as a lot of money. The point is that a red flag should have been waved immediately it was known that he had been in contact with the SD. I do not think the point about his German contacts ‘not running agents abroad’ is relevant: he was not intended to have a controller in the UK, but was expected to return to France after short visits to England. The original journey was indeed risky, because he could (and probably should) have ended up in Camp020, but after Dansey and Bodington had whisked him away, he was safe, even despite MI5 questioning. Marnham adds: “I don’t agree about HD’s motives in coming to England. HD was an experienced pilot. He was bound to be useful to British intelligence, and Boemelburg knew that. HD needed no additional SD training, just the courage and cynicism to play both ends against the middle. He was of course a pawn. In the event, his British controller was cleverer than his German controller.” I agree otherwise with what you write about MI5 and ‘TAR’ Robertson. As we have seen, TAR probably learned about Déricourt through the TWIST committee, but had no further influence at that time.
14. Marnham reports that Boemelburg was assigned to help out police duties in Lyon as early as 1942 (in The Death of Jean Moulin, p 167), so I don’t see how Déricourt’s mission came to an end because of Boemelburg’s transfer. Boemelburg must have been moving between Lyon and Paris. I had at first suspected that Dansey was rumbled in the August-September time-frame – especially when Gubbins took over from Hambro in September, and the loss of PROSPER was explained to him by Buckmaster and Bodington, but that may have been premature. Marnham’s view is that Déricourt was not fully ‘rumbled’ until the late autumn of 1943 (November 18), when Henri Frager provided MI5 with the necessary ammunition. Frager had warned Bodington in August, but Bodington naturally rejected Frager’s input, and surely buried Frager’s suspicions when he briefed Gubbins. But there was no way that Dansey could have persevered with Déricourt after a post mortem on PROSPER was carried out: the dissolution of TWIST and the new relationship with B1A probably resulted from the lessons they learned. SOE had been under fire for most of 1943, and David Stafford writes that operational control of SOE activities was formally handed over to COSSAC in October 1943, shortly after Gubbins took over.
15. (Marnham, on Bodington’s trip to Paris) “No. Bodington, and very probably Dansey, WERE trying to keep Déricourt in France until D-Day.. . .”
I imagine you have concluded by now that my theory is that Hambro was taken in by the penetration story without understanding the Déricourt role, and that he convinced LCS and the Chiefs of Staff that the sacrifice of the PROSPER network was a worthwhile tactic to help the incipient STARKEY project, all through the elusive TWIST committee. (Yet I do not understand the mechanisms of how this happened, given SOE’s absence from TWIST.) But Hambro lost his job over it, unlike Dansey.
Patrick Marnham and I look forward to your comments. Are you comfortable with my eventually posting our exchange on coldspur?
I was very gratified to receive Marshall’s thorough response to my points soon afterwards:
Many thanks. Yes, I take your point and agree.
The HD project didn’t lead me to luminaries like Dick White or Berlin. Most of my contributors were minor characters with a very specific and limited connections with HD. But again, no one I spoke to knew where my research was going. I did speak with TA Robertson, a couple of times I think, and he was very circumspect about names. Never mentioned HD, but did confirm that the exploitation of penetrated SOE networks was standard practise.
The only people I was wary about interviewing were Vera Atkins and Maurice Buckmaster. Atkins had the most extraordinary gift of recall. She was excellent on who did what and when, but all she knew about HD was what had been in his file and that was all BS. Buckmaster was very different. He had been made to look foolish in the past and was guarded. Like Vera he was very ‘old-school’ and loathed to speak ill of anyone. However, he did provide his own explanations for what had occurred in ’43, which included Prosper being told there would be an invasion that year, however I sensed he knew he’d been deceived many times, by subordinates and superiors, and everything he said felt tentative, rather than based on certain knowledge. Sporborg had gone back to a career in banking after SOE and had no particular axe to grind with anyone. He was very sanguine about that period.
1. COCKADE and DÉRICOURT.
If I recall, Cockade had its origins as a deception plan to shroud the 1943 invasion. When that was postponed, it was decided to continue working on Cockade, partly for the practise and for political reasons It’s certainly possible that Dansey used HD to promote aspects of Cockade, and perhaps his name might have come up in XX discussions; but I just can’t imagine Dansey bandying his name about.
2. I agree completely. Dansey also sat on the XX Committee.
4. HD had been a cold gamble for Dansey. He couldn’t make SOE recruit him, he just had to hope that they would. He couldn’t make SOE send him into northern France; likewise. Nor could he control precisely what information or how much of it HD handed to Boemelburg. He probably gave him solid guidelines, but once HD was in play he was his own man; making his own decisions. He was, after all, a self-serving opportunist out to line his own pockets – and Dansey certainly could pick them. I think my point is that once HD was in play, from Dansey’s perspective back in Broadway Buildings, the collapse of PROSPER (or some other network) was highly likely.
I believe HD reported PROSPER being penetrated during his Easter (May) visit to London.
I don’t claim HD was entirely responsible, but he contributed mightily. We know Boemelburg was given every single agent that came in or out, that he saw all the mail generated by Prosper’s people that had been given to HD to send on to London, and that he knew where all the key officers in the network lived. Yes, the network was too big, there were aspects of its security that were slack, and some radios were being played but we can’t underestimate how much HD’s info added to the SD’s knowledge.
N.B. Once HD was in play, there wasn’t much Dansey could do, or would do to alter any consequences.
5. It’s news to me that Buckmaster had been told Prosper would be sacrificed. Sporborg was absolutely clear that no one at ‘Country’ level (and especially not Buckmaster) was briefed about deception plans.
I’m also very clear about what Suttill was told; and that was to expect the invasion in September. The entire raison d’être for Prosper was a ’43 invasion and throughout the spring/summer his and the Scientist network were gearing up; stocking up on food and supplies for the invading armies as well as arms and munitions. After his return in July, a number of his former colleagues said he seemed desperately worried, anxious and stressed about security. I doubt he would have shared the date of the proposed invasion with many except his closest colleagues, however Jacques Bureau and one other said Suttill told them September. That surprised me because, as I’ve said before, in 1987 no one knew about Cockade. Re: Dansey – I’m certain he had a very clear picture of Prosper’s situation, and would have briefed XX that it was penetrated and doomed, and should therefore be exploited. Not sure how he could have initiated its coup de grâce.
6. I’m not aware of the Balkans being given priority over France. However, the arms drops to Prosper through spring and summer are pretty dramatic. (p292 ATKM) They ramp up exponentially from January, reaching a peak in August – and then drop off in September. A number of Prosper’s lieutenants complained that all the activity (setting up receptions, then transporting and hiding the materiel) was putting people at risk. Some field groups were out every single week.
7. I agree with all of this, but have no answer to your final question.
8. I concluded that the most reliable of Overton Fuller’s accounts of her conversations with HD are the earliest, because so little was known about any wider context – at that time. However, as the years rolled by OF became obsessed with HD, especially as info emerged about HD’s links with MI6. She became determined to try and prove otherwise. Why? I have no idea. She published countless books about him, re-hashing the same material over and over, reinterpreting it every time. She even published a 300 page horoscope of him. Barmy? You decide.
It’s possible HD may have reported ‘mission accomplished’, except that I don’t think HD’s purpose was to sink Prosper. I believe Dansey had much bigger plans for HD, beyond September ’43.
9. I see it this way. In May HD reports to Dansey of his conversations with Boemelburg, and from this Dansey concludes that Prosper is ‘fatally penetrated’. I accept that HD possibly was given false information to pass on to B, but there is no doubt in my mind that Prosper was also briefed to expect a September invasion. The rationale being that, information extracted under interrogation was more likely to be believed by the enemy.
An aside: In all HD’s accounts he refers to something called ‘British Intelligence’. It’s my view that he saw it all (SOE and MI6) as one, large amorphous organisation. He certainly didn’t perceive two separate, competing organisations.
10. See my 5. Prior to his July trip to London, Prosper would have assumed a summer invasion (May to August); I think he was genuinely horrified when he learned that it was scheduled as late as September.
11. I met and spoke with Knocken (B’s Chief in Paris) and Kopkow; the Head of Counter Intelligence in Berlin. The pressure on B was to pin down the precise date of what was universally assumed to be a ’43 invasion. I have no view on whether HD had been given a date for the putative invasion, however I disagree with your contention regarding HD’s objective. He role was to provide regular, detailed intelligence on SD’s operations, plans and objectives. He was a living breathing listening device inside the Paris SD.
12. Most of the information on Dansey’s views about sabotage vs intelligence come from the period when SOE was set up; hiving off Section D (and its people) from MI6 etc. There was a great deal expressed at the time about how this would stuff up MI6 operations and MI6 fought tooth and nail to prevent it. When it became clear Churchill wouldn’t budge MI6 was forced to live with the situation. However, it insisted that during SOE’s first year they couldn’t undertake any operations without MI6’s clearance first. ‘No bangs without FO approval’. As time rolled by, the two organisations found ways to work along side each other, but Dansey never altered his view that they were ‘a bunch of amateurs’ or the ‘boys from Baker Street’.
HD fell into Dansey’s lap because Dansey also controlled MI9, the escape service that brought HD to Britain. Dansey’s views about foreign agents has been quoted many times; that there was no point throwing them in prison when you can use them. He would have assessed HD pretty accurately and his one singular asset was his friendship with Boemelburg. How to exploit that? Put him back into France, but keep him isolated from his (Dansey’s) own networks in France. Get HD into SOE, and let them send him into France – a gamble but worth taking. If HD was going to be attractive to B, he had to have info to hand over, so what could they afford to give away? Anything SOE was doing.
There is a great deal of material in the XX papers about fears that exploiting resistance groups for deception would damage their loyalty to the Allied cause. But there are also assurances that this could be overcome, because the absence of an invasion would be seen as a fortune of war, not a deliberate deception.
Happy to continue, if there are further concerns.
Soon after this, I wrote to Marshall and Marnham, as follows:
Thank you so much for moving this debate further, Robert. I am not sure what the outcome will be, but I sense that, between the three of us, we shall be able to provide a more authoritative version of what went on with the PROSPER network. Whether that narrative appears as a new edition of Patrick’s book, or a new publication by you, Robert, or a bulletin on coldspur, or perhaps even a joint article in one of the intelligence journals (although I think multi-authored articles are a difficult feat to pull off) I do not know, but I am energized by these exchanges, although I regard myself very much as a junior partner alongside you both in the study of SOE and the French networks.
In any case, I had a few responses and questions arising from Robert’s latest message:
* I agree that the minor characters are generally much more reliable in their testimony than the big wheels.
* Surely, since Dansey had installed his Z-men (Nelson, Bodington, etc.) in SOE, he could rely on them to recruit people like HD?
* I do not see a source for the claim in WITS that Buckmaster was told in May that PROSPER would be sacrificed. [Can you help, Patrick?] (In a separate exchange, Marnham and I concluded that I had misinterpreted an arguably ambiguous passage in WITS. Buckmaster had NOT been informed of the deception plan.)
* You state that the entire raison d’être for PROSPER was a 1943 invasion, Robert. Who made this decision? If it was the Chiefs of Staff, did they review PROSPER’s status in mid-April, when the 1943 landings were abandoned? And, if not, why not?
* David Stafford provides the March 20 instructions to SOE as an Appendix (Document 7) in Britain and European Resistance 1940-1946, and sources it as CAB 80/68. He analyses Hambro’s response to it in his text. I believe the dissonance between Hambro’s interactions with the Chiefs of Staff, and Gubbins’s and the TWIST Committee’s interactions with COSSAC and the LCS, is significant. Massive shipments to France were taking place when the Chiefs of Staff were minimizing France’s role, and Hambro was stressing the dampening-down of guerrilla activity there. Your chart on p 292, Robert, shows enormous increases in arms shipments directly after the instructions by the Chiefs of Staff to downplay activity in France (March 20). How could that happen?
* I notice that you do not mention Hambro in ATKM, Robert. On page 153, where you write that Gubbins, Sporborg and Boyle were informed of the penetration of PROSPER, you do not list Hambro. Why should the chief not have been informed? Was it being concealed from him? And Patrick gives him only brief attention in WITS, echoing the view that he was very much an absentee landlord. Yet Stafford shows that he was active on SOE business during this period. How should he be drawn into the story?
* You write, Robert, that HD saw SOE and MI6 as ‘one large, amorphous organisation’. Yet in your earlier message, you indicated that Déricourt over Easter 1943 told Overton Fuller that he reported the penetration of PROSPER to a senior officer of ‘another organisation’. Why would he have said that?
* You state, Robert, that Dansey believed that ‘there was no point throwing them [foreign agents] in prison when you can use them’. But that attitude is so reckless, and so far from MI5’s policy, that it must be revisited. MI5 was petrified that a loose-lipped controlled agent could blow the whole scheme, and it was very protective of the contacts of even those it truly believed it had converted ideologically (e.g. TATE). When SUMMER tried to escape, they had to isolate him lest he pass on what he knew, and when a senior German officer, imprisoned on the Isle of Man, took part in a high-level swap with the Germans, MI5 again was very concerned that he might have learned about the XX system from fellow-prisoners. Through TWIST, Masterman and Robertson must have known what was happening, but Masterman (and other authorised historians) must have whitewashed the whole business. I cannot agree that Dansey’s gambles with HD could have been justified to anybody.
* In summary, it seems important to try to pinpoint exactly when COCKADE first saw the light of day. Patrick has referred to deception plans going back to January (p 92), but it is hard for me to imagine that any unit had at that time considered implementing any such activity. From my master Chronology, I have compiled a list of the sequence of events in a separate document (attached).
My interpretation of this sequence would be:
i) Churchill was much more concerned about keeping Stalin supportive (and discouraging him from pursuing a deal with the Germans) than he was about deceiving the Germans over any 1943 assault.
ii) It seems to me that no serious plans for COCKADE-like deception plans would have been considered before April 13, when Churchill gave up. A closer reading of the archive would indicate that any deception plans up till then were in support of a real 1943 landing, not an attempt to emphasise the probability of a fictitious one.
iii) From April 18, COSSAC’s instructions are clear: to prepare camouflage and deception plans to keep the Germans fully engaged in France. But who took the decision to sacrifice PROSPER? The Chiefs of Staff? Morgan? Bevan? Hambro? Gubbins? TWIST? On August 1, the JISC declared that SOE had been less than frank about the situation in France, suggesting that some sort of inquest had taken place, and that the SOE was taking the rap.
iv) Dansey was no doubt a contributing factor, since, while explaining to TWIST that PROSPER had been compromised, he did not explain how that had happened. As we know, SOE was not represented on TWIST, but was it (through Gubbins) acting independently and cavalierly, or was it following instructions? The ability of SOE to gain aircraft and material support suggests the latter. Gubbins’s biographers stress his regular and constant negotiations with the Chiefs of Staff. Yet SOE was not officially told that the invasion was off until July 22.
v) If Hambro was not closely involved with the status of PROSPER (or even did not know what was going on), one would have to ascribe full responsibility to Gubbins. But, when Hambro was fired, Gubbins took over. That would suggest to me that Selborne was ignorant of what was really going on.
vi) It seems to me that, once the 1943 attack was officially cancelled, and certainly when Stalin was formally notified of it (June 2), someone in authority should have stepped back and said: “Why are we sacrificing this valuable asset when we know that the fact of the French Resistance being armed will not alone convince the Germans that we are planning an attack this summer, and we stand to lose material, lives, and French resistance goodwill?” Someone at the Bevan/Morgan/Gubbins level should have pulled the plug and dismantled the network. Remember, Stanley had resigned as LCS chief because he would not use resistance forces for deception purposes.
vii) The fact that all the authorised and semi-official historians (Hinsley, Howard, Hesketh, Masterman, Foot, Seaman, even Trevor-Roper – who attended LCS meetings) have tried to bury any details about TWIST suggests to me that this was the seat of the problem, and the source of a major embarrassment.
Marshall responded on August 3:
I’m in the middle of something at the moment, so will just send a quick response, and a more considered one later in the week.
Not sure Bodington was ever a Z man. He applied to join the Service and was turned down twice. Of course Dansey clearly had him doing things, but like a lot of Dansey’s people, on an informal basis. Nothing on the books.
The thing to look closely at, is who came across to SOE when it was originally set up. Clearly Dansey wanted people embedded in Baker Street who would keep him informed. I’m not certain for sure, but I suspect Bodington was used on an ad-hoc basis, along with people like Andre Simon et al. Nelson? Probably.
The FO’s ‘SOE Adviser’ with whom I had a number of lengthy meetings with told me that was Prosper’s role. Certainly every single Prosper veteran I spoke to ‘knew’ their purpose was to support the invasion – expected in the summer of ’43.
I can’t explain why arms shipments should be in contradiction of orders from the Chief of Staff. These figures either came from the SOE Adviser, or papers in what was then the PRO. In 1985/6 they looked to me like Cockade in operation.
I didn’t intentionally ignore Hambro, he just didn’t figure in my research. I can’t recall, but he may have been interviewed. I’ll check.
I will go back and check the precise quote in OF’s original book, and get back to you.
However, in a sworn statement HD made after his arrest in 1946, this is what he said:
“I was transferred to SOE, a unit specially concerned with sabotage. This service, like all Allied services at the time, was controlled by SIS. I entered into an additional commitment, through Andre Simon, about the secrecy of my work.”
In a revised version of the statement made in 1947, the reference to SIS is removed. (I have HD’s trial papers.)
I’ll get onto the rest later this week.
At this time, I also had some email discussions with Patrick Marnham, primarily about Moulin’s allegiances when he was betrayed, and who his probable persecutors might have been. I had studied Marnham’s Death of Jean Moulin quite carefully, and had to state that I was not convinced that (as Marnham believed) Moulin had become, by the time of his last return to France, a proper republican and utterly loyal to de Gaulle. Yet I had to defer to Marnham’s deep exploration into this subject: he has agreed to return, before long, to my points about Moulin’s frequent expression of communist sympathies. This issue has tangential relevance to the PROSPER betrayal, of course, because part of Marnham’s thesis in WITS is that Moulin was a parallel sacrificial agent of Dansey’s.
On August 13, Marshall sent the following message:
I’ll try and pick up where I left off:
Dansey and Double Agents: I’m not certain of the precise dates but around 1917 to early 20s Dansey was based in the US where, amongst other things, he was involved in training-up a nascent US intelligence service. He is on record as recommending that when a foreign agent was picked up, that you should make every effort to turn them before deciding to throw them in goal. They were potentially far more valuable in play, than behind bars. Of course, if they can’t be turned or you couldn’t trust them, then off to the chokey.
I recall TA Robertson telling me that by 1943, they had every single German agent in the UK under their control.
I see Dansey as an opportunist (not unlike HD), who scooped up people and put them in play, with safety measures in place, then waited to see what they delivered. I think it’s hard to stitch together a grand plan or scheme and besides, I just don’t believe the world works that way. The war, like all eras, was a constantly changing stream of events, disasters, opportunities, catastrophes, luck – good and bad. Success had a lot to do with how well you overcame obstacles and seized opportunities. With someone like HD, he would have wanted him back in France next to Boemelburg, but nowhere near any of his own networks or agents. Getting him into SOE was the way to achieve that.
Looking back at old transcripts, I note that back in ’83 or ’84 MRD Foot concurred that it was likely that HD’s name came to SOE via MI6. We never found the paperwork (no surprise) but examining the pathway in the SOE records Foot concluded that it probably came from 6. He also speculated that HD’s citation for a DSO may also have come from 6. Again, no proof. I never spoke to Cohen.
When researching the book the ‘SOE Adviser’; told me he could find no evidence of how HD was introduced to SOE. He also said he couldn’t trace who had put up his DSO which, by the way, was never gazetted. Now this was odd, because just about every other individual he gave me info on, he could trace how the name or recommendation had come to SOE. Likewise the SOE’s very first DSO? Not to know who proposed or drew up the citation?
Foot told me that he had asked Cohen if HD had been run by him. Cohen said not. Not much that can be drawn from that.
However, by the time I was writing the book (after the BBC program), Foot was much less forthcoming and was highly critical of the book. Read into that what you will. He did, after all, heavily revise his SOE in France.
I am ploughing through JOF’s original book (grief) and will forward a selection of quotes from that. What makes her account such hard work is the ‘dance’ she reports going on between her and HD. She is not dispassionate nor objective, and she rambles – at times it’s almost like stream of consciousness. However, I will pick out what there is and forward over the weekend.
That is where things stand. Mr Marshall and I have started a rewarding investigation into the events of summer 1942, when both Bodington and Déricourt were making their way from the Marseille region to Gibraltar, one in a felucca, the other in a trawler. I am keen to ascertain whether there was any possibility of a meeting between the two, and how much MI6 in London knew about Déricourt when they surprisingly approved his passage on the MI9 escape line. I shall resume that story, and others, later, when Patrick Marnham returns from his holiday/vacation/leave in France at the end of August. In conclusion, I believe there are several fascinating conundrums still to be sorted out:
* The exact status of Déricourt when he was recruited in England in September 1942 by MI6/SOE. (I am developing a profile that compares his situation with other ‘controlled agents’ who were processed by MI5.)
* The allegiance of Moulin when he was arrested, and exactly why and how he was betrayed. Patrick Marnham and I have some discussions ahead.
* The linkage between the betrayal of Suttill and that of Moulin. Were they really both sacrificed as part of the deception campaign?
* The curious way in which Charles Hambro (the head of SOE) has been left out of most of the accounts of the early COCKADE proceedings, with questions as to who was calling the shots at this time.
* The puzzling facts about SOE’s intensification of weapons drops to France in April 1943, after the Chiefs of Staff had clearly ordered them to reduce France’s priority below that of the Balkans.
* The activities of the TWIST Committee, and how they have been comprehensively stifled by all the official and authorised historians.
* The strange acquiescence by Dalton, Selborne, Hambro and Gubbins over MI6’s takeover of SOE ‘double agents’, and the non-representation of SOE on the TWIST Committee.
* The failure of SOE to pull the plug on the PROSPER network and withdraw once they knew that it had been infiltrated and betrayed.
* The achievement for which the KCMG was conferred upon Claude Dansey on June 2, 1943.
Any readers who can shed light on any of these matters are encouraged to contact me at firstname.lastname@example.org.