Gordon Welchman

by Joel Greenberg

  A deeper understanding of the problem and of the objectives of the proposed development of overview teams can be derived from the story of intelligence in World War Two. This story was suppressed for a long time, but in the last few years a good deal has been revealed to the public, and there is now no excuse for neglecting this important aspect of military history. It should be taken into account both in our overall planning for the future and in the direction of our R&D projects. We need, both in peacetime and in battle, to know as much as possible about a potential enemy’s method of fighting, his capabilities, and his intentions. We also need to prevent him from discovering what we do not wish him to know about our peacetime defence planning and wartime objectives.

  The second document, dated December 1982, was entitled ‘The Neglected Claims of Intelligence in Today’s Defence Spending’. It was based on his last piece of work for MITRE and supported, he felt, by a number of recent books on intelligence in the Second World War. Their authors had all pointed out the importance of interplay between forms of intelligence derived at different times and from different sources. Welchman felt strongly that lessons from actual combat were being neglected by the people who controlled research and development projects.

  On 21 August 1982 Welchman had been guest of honour at the annual convention of the American Cryptogram Association. David Kahn, one of the first authors to write about the history of cryptography, had met him in New York and driven him to the convention. One of the keynote speakers was his old BP colleague Bill Bundy, who was effusive in his praise of Welchman’s contribution to the success of Hut 6 specifically and BP in general. Kahn introduced Welchman to James Bamford, an author and journalist who was about to publish a book about the NSA. Not surprisingly, the NSA had been trying to prevent Bamford from issuing The Puzzle Palace, because it regarded the book as an exposé of its organization. Welchman thought that Bamford’s battle with the NSA could prove helpful to him. In fact, it may have been the cause of his problems!

  Several years earlier, Bamford had become interested in a clandestine NSA operation to read all international telegrams being sent out of the USA. After repeated attempts to find out more information about this programme, codenamed Shamrock, he eventually received a lengthy document which described it in some detail. The NSA subsequently tried to get Bamford to return the document on the basis that it had been accidentally declassified. Existing American law offered him some protection and he was able to provide details of Operation Shamrock in his book.

  Welchman was not the only veteran of Allied cryptographic work to fall foul of the NSA. At the end of 1958, members of the NSA visited the home of William Friedman and confiscated a number of items from his personal cryptographic collection. The years of clandestine work had taken their toll on Friedman and in early 1963 he was admitted to a psychiatric hospital unit suffering from depression. Later in the year, disgusted by the way he had been treated by the NSA, he decided to bequeath the remainder of his remarkable collection of documents, dealing with almost every aspect of cryptography, to the George C. Marshall Research Foundation in Lexington, Virginia. According to Bamford, he managed to gain access to Friedman’s papers despite attempts by NSA officials to prevent researchers from doing so. The NSA subsequently tried to prosecute Bamford after the publication of his book.

  As computer security specialist Bruce Schneier explained in his review of Bamford’s second book about the NSA, Body of Secrets:

  ‘The Puzzle Palace’ was a landmark book, and widely read in circles that knew something of the NSA. Inside the NSA itself, where the agency’s secrecy prevents its employees from knowing much about their own history, it was a best seller.10 The book was a history of American intelligence from 1917 and was both shocking and pedestrian. Operations like Shamrock were exposed for the first time, but Bamford also spent a lot of pages simply explaining how the NSA was organized. Nobody knew anything, so it was all interesting.11

  In early October 1982, Bamford sent Welchman a copy of The Puzzle Palace. Bamford had been helped in his battle with the NSA by Mark Lynch, an attorney for the American Civil Liberties Union (ACLU) specializing in national security matters. Welchman spoke to Lynch initially in November, and sent him his paper ‘The Story of My Story’. Lynch agreed to take on his case and offer advice on the best way to proceed. His initial view was that Welchman was unlikely to be in trouble as long as he simply ignored the NSA and used his own judgement about what he said in public. The NSA strategy seemed to be to delay making any decision on the matter while pressuring McGraw-Hill to desist from publishing further editions of The Hut Six Story.

  Ronald Lewin had been appalled at Welchman’s treatment by the intelligence services. He had read The Puzzle Palace and found it both ironic and infuriating that a spill-the-beans book of this kind could be published while Welchman was being penalised so heavily for so little. In September 1982 he decided to write an old friend about the situation. His old friend, William Casey, happened to be Director of the CIA.12

  PERSONAL

  I enclose a sad letter from one old friend about the sad situation of another old friend. No doubt some aspects of the case are already familiar to you, but anyway the letter speaks for itself.

  I would only add two points. First, I heard just the other day from the highest possible source that it was Welchman’s drive, dedication, and brilliant cryptanalytical flair which, during the first winter of the war, led to the very first breakthrough into the German Enigma cipher. Please do not think it is impertinent of me if I say that I find it shameful that a man of total loyalty and integrity, and of massive achievements, should come to be in his present situation after having made so enormous a contribution to the security and the survival of both our nations.

  Secondly, I have it on the personal authority of Welchman himself that at no stage was he ever given a briefing which laid down what he might or might not write about. The book itself is very familiar to me as I recently reviewed it in glowing terms for The Times, and I am bound to say that as one who is constantly alert in security matters I find nothing in the latter portion of the book which could conceivably be held to be a dangerous disclosure of classified material.

  Please excuse my making use of our friendship to bring this business to your attention. I can only say that if you were in Gordon’s position I should be fighting to the death on your behalf, as I intend to do in his case until some satisfactory solution or explanation emerges – and at least until I can feel that for what seems to be formalities he is no longer in danger of having to stand in the bread line.

  From August to November, Welchman had regularly sought the advice of his lawyer, John Stevens. While he had at last realized that he could not achieve anything through MITRE, he still wondered if there was some other way of passing on what he had learned. At seventy-six years of age he was growing tired, and felt that Lewin’s comment to Bill Casey on the danger of ‘having to stand in the bread line’ was close to the reality of his situation.

  When Welchman received his copy of Bamford’s book, he noticed Frank Rowlett’s name in the acknowledgements and decided to contact him. Thirty-seven years earlier, they had worked together when Welchman had visited the US in his role as Assistant Director for Mechanization. Rowlett had been third in command at the NSA before he retired in 1962. During a long telephone conversation, Welchman told him that he was considering a direct approach to the US Government Accountability Office (GAO) whose Office of General Counsel issued legal decisions, opinions, and reports on bid protests, appropriations matters, and other issues of federal law. Rowlett advised against such an approach. Mark Lynch also felt that this would be a mistake as ‘it is just another government agency scrapping with the rest of them’. He offered to speak to a friend in the GAO, and suggested instead that Welchman approach Senator Edward Kennedy and Congressman Mavroules. To help with such an approach, Welchman wrote a document titled ‘Appeal of an Author’, which he sent to Lynch along with his ‘The Urgent Ne
ed for Development Overview Teams Reporting to Congress’. Rowlett, however, did not think that Senator Kennedy would be much help. His view was that the sort of problem that Welchman was posing was not one which could be effectively resolved in a political arena. Rowlett did offer to prepare a statement regarding the importance of Welchman’s contribution to the vital intelligence produced from the Second World War German Enigma intercepts. This statement, dated 17 January 1983, coming as it did, from one of the USA’s leading cryptanalysts during the war, gave a good idea of how high a regard he was held in and went on to say:

  In my position in the Army Security Agency, I had full access to the technical information on the processing of the Enigma intercepts conducted at Bletchley Park, including the development of the procedures and equipments used for their exploitation. It was in this connection that I met Gordon Welchman, one of the key individuals involved in the British effort on the Enigma intercepts.

  Gordon Welchman’s imaginative foresight, coupled with an exceptional administrative capability and an extraordinary comprehension of the intricacies of the Enigma cipher machine, enabled him to make an outstanding contribution to the successful and timely exploitation of the Enigma intercepts. He deserves credit for a major contribution to the design of a special cryptanalytic device employed in the recovery of the various daily keys used in the Enigma intercepts and for developing organizational and administrative procedures for the expeditious processing of the intercepted messages. Because of the paramount need for secrecy, as well as the technical complexity of the procedures and equipments employed, the full impact of Gordon Welchman’s contributions could be appreciated only by the relatively small number of those of us in the U.S. and U.K. who were involved in the technical management of our respective parts of the combined effort.

  Welchman pressed on and wrote to the mayor of Newburyport, who apparently had read The Hut Six Story. His heart doctor, Dr Leary, was a close friend of the mayor, who in turn was a friend of Kennedy. Leary was convinced that the stress of the persecution was affecting Welchman’s health. He thought he had made a strong case and, with high hopes, he submitted his documents with covering letters, to the two senators and one congressman from Massachusetts. He received neither an acknowledgement nor a response. Though this was hard to understand and disappointing, he eventually realized that politicians, like the senior managers at MITRE, had to worry about their sponsors too. What he was proposing would be unpopular with many senior members of the military establishment and with the weapons industry. It would be hard to explain in their constituencies and they would prefer more straightforward issues.

  Meanwhile, life had to go on. The Welchmans were having financial problems as Teeny had been laid off at the local hospital where she had worked for the previous seven years. Welchman made enquires about possible royalty payments and the chance of a second printing of his book. He documented the changes he had in mind for his agent John Cushman’s consideration. Ralph Bennett, a Cambridge historian, had contacted him for help on a new book and, in May 1983, he had a visit from Whitfield Diffie, one of the pioneers of public-key cryptography. However, his ever fragile body gave out once again and on 10 June he had a left hip total replacement operation.

  His spirits were lifted the following month. Rowlett told him that, in his view, he had made a greater contribution in the Second World War than Turing. He then received a letter from Marcus Shudforth, a Sidney Sussex alumnus and contemporary of John Jeffreys at Brentwood School and Cambridge. Following a report in the Sidney Sussex College Annual, Shudforth had read The Hut Six Story. He had recently had lunch with the senior maths tutor at Brentford School, who had never heard of Jeffreys. Shudforth asked Welchman to autograph a copy of his book and send it to the school in Jeffreys’s memory, which he duly did. It was somewhat appropriate that the school’s motto was ‘Here in the past, may the present, find the means to fight the battles of the future.’ The dedication in the book, which is held as a showpiece in the school library, reads

  In Memory of John Jeffreys, a

  Most valued friend and colleague.

  Perhaps the story of what he did

  Will be an encouragement to

  Present and future generations

  Of mathematicians at Brentwood.

  Gordon Welchman

  Newburyport, Mass., U.S.A

  October 1983

  The events of the previous year had clarified his thinking and he had finally realized that Peter Calvocoressi’s publishing instincts had been right when he had advised against trying to write two books in one. As he said in a letter to Robin Denniston, dated 17 January 1984

  The failure of my book, The Hut Six Story, has also been a major frustration. Having shut down their General Books Division, McGraw-Hill wanted to suppress it, and I gather that Penguin/Allen Lane made no real attempt at promotion. But it has at last dawned on me that it was a mistake to include Part Four, which is concerned with military problems of today.

  It was an understandable mistake. At the time I wanted to justify my revelations of Hut Six methodology by showing that they contain valuable lessons for today. But I was drawing on a wide range of historical evidence of which the Hut Six experience was only a small part. Moreover my studies since the book was finished have provided new and better material, and my thinking has matured a great deal. Already Part Four seems obsolete, whereas the rest of the book is not.

  Thinking on these lines I wondered whether the remainder of the book could stand on its own if Part Four was cut out. It could! Only a few minor modifications would be needed, and the result would be a far more attractive book that would retain its interest.

  Returning to the ideas of a new version of The Hut Six Story, I could try to get permission to add two appendices. I can now see the relationship between the Polish Bomba and the British Bombe. And the story of the Stephenson/Bayly development of teletype encryption should be told.13

  Welchman also wrote again to Sir William Stephenson on 10 January 1984 telling him about Hodges’s book and drawing parallels between himself and Turing. Stephenson offered to do anything he could to help. Welchman proposed the creation of a foundation along the lines of the ‘Stephenson-Donovan Foundation for the Advancement of Military Communications, Intelligence and Security’. It would be the ideal sponsor for his Development Overview Team idea. Stephenson cabled back on 12 February suggesting that Welchman write to William Casey. He also said ‘as Travis would say to Bayly, get cracking’.

  Buoyed by the good news from Penguin that The Hut Six Story would be published in paperback in the UK on 31 May, Welchman wrote to Casey on 23 February, as suggested by Stephenson. Casey replied on 10 March, saying that he was familiar with Welchman’s thoughts on battlefield management and he had read The Hut Six Story as well as Lewin’s letter in 1982. Casey said ‘I will get the views of some people here and think about the people who might participate in the meeting and carry on whatever evaluations might be decided upon.’ He requested copies of Welchman’s two papers, which Welchman duly sent to him on 19 March. Once more Welchman reached a dead end; he never heard from Casey again.

  The year 1984 brought the sad news that his agent John Cushman had died and Jane Wilson was picking up his work at JCA. He also met Linda Malvern who had written a front-page article in the Sunday Times about James Bamford and his trouble with the NSA over The Puzzle Palace. She had been shocked by Welchman’s treatment by the NSA and the Department of Defense but Mark Lynch advised against any kind of public protest on Welchman’s behalf.

  Following a stay in hospital with a bad bout of flu he wrote to Penguin’s Head of Publicity, Dotti Irving, pointing out that they had put the wrong machine on the cover of the paperback version of the book and that the cover information completely ignored Part Four However, in July, Irving told him that they had sold 4,400 copies into bookshops and that a good review by David Darby in Time Out, had particularly praised his Part Four.

  Throughout his battles with t
he NSA and his repeated attempts to get his views about battlefield management on to the American political agenda, one other project drove Welchman on. He had learned some of the details of the pre-war work by the Poles on the Enigma machine too late to include them in his book. He had suggested to David Kahn that he write a paper for Cryptologia in January 1983 to put the record straight about the Polish contribution. His proposed title, ‘From Polish Bomba to British Bombe’, delighted Kahn, but it appears that pressure from the intelligence services, prevented the project from going ahead. By the middle of 1984, Robin Denniston was encouraging him to press on with it. Welchman had already written to former colleagues whom he thought could help, including Harold Fletcher, Peter Twinn, Dennis Babbage, Reg Parker and Jean Howard. He had also written to Tadeusz Lisicki, with whom he had had an exchange of letters after the publication of The Hut Six Story. Apart from relating an accurate account of the Polish contribution, he was also determined to give appropriate credit to the achievements of Dilly Knox and Alastair Denniston.

  The Polish historian Władysław Kozaczuk had claimed in his book Enigma in 1979 that none of the Polish ideas had been thought of by the British and that ‘virtually all major cryptological techniques that the British used to break Enigma during the Second World War had been thought of and used by the Poles earlier’.14 The revised (1984) English version had been particularly critical of Welchman. This was no doubt written to counter Hinsley’s account in the Official History.

  The early part of the story was also incorrectly told in other sources. In a letter to Denniston, in July 1984 Welchman said: