It was late when the small band of Americans arrived at Bletchley Park. As they walked up to the front door of the old mansion house, the dark of a January winter was underscored by the blackout curtains that eliminated every last chink of light from the windows. The journey of the four men – known as the Sinkov mission, after their leader Abraham Sinkov – had been stormy, dangerous and utterly secret. To reach their destination they had braved bullets and bureaucratic mistrust, but their arrival marked the opening courtship of what would become the longest-standing, most powerful intelligence alliance in history. If there is one area in which the so-called ‘special relationship’ has always been most intimate, it is in communications intelligence and code-breaking.
America was not yet in the war when the group arrived in early 1941. Britain had stood alone and vulnerable, fearing invasion and defeat. It needed help, and so the previous summer the first tentative discussions had taken place with the US. As plans for the visit progressed, Churchill and many senior intelligence figures had real reservations about opening up to a country that had not yet committed to the fight.1 The British were interested in the Americans’ progress against Japanese codes, but there was deep nervousness about revealing too much about their own progress against the Germans for fear of the secret escaping. In November, the chief of MI6 told Churchill it might be awkward only to talk about Japan and not open up about Germany. It would be obvious ‘we have something to hide’, he explained. Churchill agreed they should show a little leg. However, one British official added a rather telling note to a memo: ‘What will they think if they find we have been reading their own stuff?’2
A decade earlier, in 1930, Abraham Sinkov, the leader of the expedition to Bletchley, had been given a memorable induction into the secret world of American code-breaking by William Friedman, his new boss. Friedman, the flamboyant son of a Hungarian-Russian-Jewish immigrant, enjoyed his work so much that dinner guests would find the menu in code.3 In 1930 he had just taken charge of a tiny unit called the Signal Intelligence Service within the US Army. His first step had been to double its size by bringing in three young recruits – Abraham Sinkov and Solomon Kullback, close friends and classmates from Brooklyn, and Frank Rowlett from Virginia – all mathematicians who would play a key role in the coming decades.
On a humid June day in 1930 Friedman, in a natty blue suit, asked the three new recruits to follow him, making it clear they were going to be let into something truly secret. Decades later they would be able to recall that moment in forensic detail, memories captured in now declassified files. Friedman took them down the stairs to the second floor of the Munitions building. Friedman swung left into a deserted corridor and stopped outside Room 2742. From his inside coat pocket he took out a small card and began to work a combination lock on the front of a steel door. The bolt swung open. Behind it was another steel door. This time, Friedman extracted from his coat pocket a key to unlock the inner door. Inside the room was pitch-black. There were no windows and foul air spilt out. He then produced a small box of matches and lit one so he could find a pull-cord for the ceiling light. It revealed a room twenty-five-feet square jammed with filing cabinets packed so close the drawers had barely enough room to open. It was the dustiest room the men had ever seen. Friedman turned to them and said in solemn and imposing manner: ‘Welcome, gentlemen, to the secret archives of the American Black Chamber.’4 Today, deep inside the heart of the NSA, sits its most secret room where the records of the codes it has broken lie. It is still called the Black Chamber.
‘King Solomon’s mines could have offered no greater treasures for us,’ Rowlett thought as the new recruits were shown an archive of solved and unsolved codes dating back years. ‘We lost all track of time.’ Friedman explained that the room contained all the working files of a secret unit which had operated in New York until it was closed a few months earlier and which had succeeded in breaking many diplomatic codes. He explained that the original Black Chamber had been the creation of Herbert O. Yardley, a colourful poker-playing character, who had started working as a code clerk for the State Department just before the First World War. He said he cracked a message from the President in two hours on his night shift. This led to a role with America’s military intelligence branch reading all the traffic he could get hold of. He would later say he learnt from British colleagues of their ‘long and dark history backed by a ruthless and intelligent espionage,’ and believed America needed to match that capacity if it wanted to be a great power. After the war, Yardley convinced superiors at the War Department and State Department to fund an organisation operating under a commercial front (the ‘Code Compiling Company’ in New York). This spied on the diplomatic traffic going in and out of Washington of around three dozen countries, thanks to messages provided by cable companies.5
‘Gentlemen do not read others’ mail’ was Secretary of State Henry Stimson’s sniffy verdict in 1929 when he learnt about the work. That had led to the closure of the Black Chamber. But the military simply snaffled all the files and created its own new team under Friedman to protect its own codes and prepare to attack those of an enemy in time of war. Closing the door behind Sinkov, Rowlett and Kullback, Friedman explained it was now his new team’s job to catalogue the dust-laden files and understand what they contained. Secrecy was vital – not least from the State Department.
The team of code-breakers began work. Resources were limited with the Great Depression under way, meaning they had to pay for their own pencils and paper from the dime store or write on the back of old weather reports. They would sit in the Munitions building in Washington with no air conditioning during the hot summers. The fans needed to keep them cool meant the team had to hold down the pieces of paper they were working on to stop them blowing away. The first traffic they focused on was known as ‘Rum-Runner’. Prohibition, which banned the sale and transportation of alcohol, had led criminal gangs to employ retired navy men to help smuggle liquor by boat from Canada, Mexico or the Caribbean. They were organising by radio, using codes. Friedman’s wife worked in the Coast Guard and passed on messages for his team to crack. The team also had what they called a ‘nut’ file full of people who had written to the government saying they had invented the perfect cryptographic system for keeping secrets and they were willing to sell it for $1 million, but they regretted that if their own government was not interested they might have to go to another country. The team asked for samples and normally solved the codes within minutes.6
At first Japanese messages were intercepted by an industrious army colonel in San Francisco. He rigged up an alarm clock with a clothes-pin to act as a time switch to tape-record the radio traffic sent to Tokyo at a regular time. The tapes would then be airmailed back to the code-breakers to work on. ‘What we were doing was theoretically illegal,’ Solomon Kullback reckoned, and the team discussed whether they would ever end up on trial. Friedman told them it was in the national interest and if they had any qualms to get out. Eventually the group received an opinion from the Attorney General legitimising the act after the event, which was kept in the office desk. Until war led to a sounder footing, this seems to have been the rather thin legal foundation for asking telegraph companies to photograph traffic.7 Everything was a little informal. Rowlett himself never had any kind of security check before the war.
But the secrecy of the Black Chamber did not last. Whereas in Britain loose-lipped politicians revealed the code-breakers’ work after the ARCOS raid, in America it was one of their own. The demise of the Black Chamber had left Yardley out of a job. Broke, angry at his treatment and believing both he and the art of code-breaking deserved recognition, Yardley decided to tell his story. His book The American Black Chamber appeared in June 1931, with excerpts running in the papers. The book included a picture of a reconstructed 10,000-word British Foreign Office code book which the Black Chamber had worked on and made it clear that London’s diplomatic traffic was not immune from American interest, just as Britain had been targeting the US.8 It also revealed that America had exploited the discovery of Japan’s fall-back position in negotiations about disarmament in the First World War. Japan realised it was vulnerable and began to change all its codes (Rowlett later thought this actually helped, as it made the team up their game). Even countries not directly mentioned in the book began to tighten their systems. There was panic on a scale not matched until another insider revealed secrets three-quarters of a century later. And in Yardley’s case, this was a man who knew everything. Congress passed the first legislation criminalising the revelation of secret code information. The military tried to track Yardley down, fearing he had taken classified material with him. When three officers turned up on his doorstep he denied possessing anything and began writing spy thrillers (The Blonde Countess) before going to China, where he was paid $10,000 a year to help target Japanese communications.
The team in Washington eventually began breaking back into Japanese systems. But in March 1939 an unreadable message was intercepted between Warsaw and Tokyo. The new system was codenamed Purple. It was Japan’s highest-level diplomatic cipher machine and it stumped American code-breakers for a year and a half until it was broken in September 1940. Friedman had suffered under the pressure. He had a breakdown and was briefly hospitalised just as the trip to Britain was organised. So it was Sinkov who left for Bletchley carrying an incredibly precious cargo – their reconstruction of the Purple machine along with a hoard of related documents.
Together with Sinkov and an army colleague, Leo Rosen, were two men from the navy – Robert Weeks and Prescott Currier. The problem for America’s code-breakers was that there was not one but two code-breaking teams and they competed fiercely and destructively. Bureaucratic turf wars have been fought back and forth across the US intelligence community for decades. Before the war, bitter rivalry between the army and the navy meant they each duplicated the other’s work and jealously guarded what they produced. In a sign of just how absurd the fights could be, early in the war US naval analysts worked on Japanese diplomatic codes on odd days of the month, the army on even days.9
The four men left Annapolis on a cold January day. They spent hours on a small boat alongside the British battleship King George V waiting for the new British Ambassador to Washington to disembark so they could begin their mission.10 The ship headed for the Orkney Islands, just off the north coast of Scotland. By 7 February it arrived at Scapa Flow. The plan was for two flying boats to take them down, but the weight of the thirty crates they had brought – adding up to nearly two tons – meant they simply could not fit everything through the hatch of the flying boats. So they were left stranded. ‘It was raining,’ recalled Prescott Currier. ‘It was kind of unpleasant.’ Neptune, a battered, barely afloat cruiser, was diverted up. Captained by the colourful, red-cape-wearing and poetry-reading Rory O’Connor, it got halfway down the coast before it passed a convoy tailed by a German naval reconnaissance plane. ‘We all knew what this meant,’ Currier later recalled. Two German dive-bombers appeared. ‘I was trying to eat some soup but my mouth was so dry I couldn’t swallow it,’ Currier said. ‘I’ve never been so scared in my life. We heard a bomb landing on one side and one on the other side and the ship would bounce out of the water and back down again. Then I heard something that sounded like someone dragging chains along the deck.’ The chain sound came from German guns strafing the deck, half a dozen times in all. ‘God, I thought, all that gear is stacked up on deck unprotected. Absolutely unprotected. And I was so scared I couldn’t go up and do anything about it anyway. Well, that’s the Purple analogue gone,’ thought Currier. After the attack he and the three other Americans went up on deck. It was strewn with spent bullets. To their relief, the Germans had been using copper-jacketed explosive bullets which never penetrated the crates. It was a lucky escape for the Purple machine (although a year later the Neptune would strike a mine off the Libyan coast, killing all but one of the 767 men on board). At Sheerness the Americans were met by a delegation from Bletchley. ‘And there was Brigadier Tiltman standing in his full regimental [uniform] with his legs spread apart and his hands behind his back.’
Waiting at Bletchley behind the blackout curtains for his guests’ arrival, the Director, Alastair Denniston, had given strict instructions to his assistant. ‘There are going to be four Americans who are coming to see me at 12 o’clock tonight. I require you to come in with the sherry. You are not to tell anybody who they are or what they will be doing.’11 When they entered his study at the front of the house, the Americans were met by the senior staff of Bletchley, stood in a semicircle, who were introduced in turn. Denniston’s assistant, who had never seen an American before, ‘except in the films’, poured out the sherry (whisky was hard to come by). The British had been ordered to help but were nervous.
The American gift of the Purple machine was an ice-breaker though, ‘a tremendous gesture,’ recalled Tiltman. ‘Somebody had to make the first step and the Americans made it.’ Tiltman was a pivotal figure in the Anglo-American relationship, seen by the Americans as the embodiment of British eccentric brilliance but without falling into their other caricature of officious condescension.12 He had worked on Japanese codes and shared what he knew. But what of Enigma? Tiltman’s superiors feared a leak could endanger the nation. Many senior figures in Britain were not even allowed to know the truth – deceived instead by the idea that the rich intelligence was coming from a German spy codenamed Boniface. Tiltman thought full exchange meant just that and went to the head of MI6, who agreed that the Americans could be told if they limited the knowledge to an agreed list of people.13 And so, right at the end of their visit, the Americans were let into the great secret. They were shown the bombes and told most (but not all) of the details of the break into Naval Enigma. It is easy to forget quite how remarkable this was. A year before America actually joined the war, the two countries were sharing their most precious secrets. The visit was a success – professionally and personally. The Americans were taught the game of rounders, which they thought of as baseball played with a broomstick. Sinkov could recall being put up in a fancy country house where the butler looked down on him for failing to use a butter knife correctly. There were reminders that this was a country at war. They visited a nightclub one night to see the band playing American music. The next night a bomb hit the club, killing all the musicians.14
The team returned after an exhausting, rough crossing and arrived bedraggled back in America. Their destitute state may have been one of the sources of gossip in Washington that they had come back empty-handed – after all, they had given the British a Japanese machine but had not brought back a German one. Sinkov’s notes in America’s National Archives make it clear he did not put down on paper everything he knew when he returned, and the fact that the team had to keep secret what they had witnessed may also have fuelled the speculation of those not in the know that America had somehow got the raw end of the deal and was being given the ‘runaround’.15 There was tension in the early days. By November 1941, the US Navy – less keen on working with Britain – was making a fuss. The US Navy and Army would both end up developing independent relations with Britain. ‘British officials regarded negotiations with the Americans as a little like dealing with the former colonies after the American Revolution – disorganized and frustrating at times, but they could still play one off against another to achieve objectives,’ an official NSA history notes.16 The American army-navy rivalry was believed by some to have contributed to the disaster of December 1941, when the Japanese launched a devastating surprise attack on Pearl Harbor which the code-breakers had failed to predict. It would draw the US into the war, force the services to work more closely together and also lead to a long-standing desire for communications intelligence to provide warning of ‘strategic surprise’. A division of labour was also agreed, with the US leading on Japanese systems and Britain supporting, while the roles would be reversed when it came to Germany.
America sent a stream of about 100 experts to Bletchley who were integrated into the work of each hut rather than operating as a distinct team. Already some of the cultural differences between the US and the UK were evident, which would persist to the present. The Americans noted how at their new headquarters, Arlington Hall, everyone would wear their military uniform if they had one. The atmosphere was formal. That was not the case at Bletchley. To make the point, one American liked telling audiences the story of two wounded personnel assigned to Bletchley to help out who, while convalescing, looked at the way people dressed and acted and believed they had been sent to a mental institution. The Americans also seemed an exotic bunch to those at Bletchley (including a number of the young women, some of whom found their manner infuriating; others thought it more appealing).
America offered industrial as well as intellectual muscle. Alan Turing visited the US in November 1942. After some initial problems getting through immigration at Ellis Island in New York, he saw the way in which the US was adapting his designs. He was not entirely convinced by their understanding of all the concepts and expressed it in the rather superior British attitude often found at the time. ‘I am persuaded that one cannot very well trust these people where a matter of judgement in cryptography is concerned,’ he wrote in his official report before adding, ‘I think we can make quite a lot of use of their machinery.’17
Britain had been struggling to produce enough bombes; America’s industrial war machine would not. The US had decided, when Britain had not been able to read German Naval Enigma in 1942, that it would develop its own machines, especially because it sensed Britain was not sharing everything and it too was suffering losses in the Atlantic. The US took what Turing had designed at Bletchley and industrialised it on a scale that mirrored the imbalance in resources between the two Allies. The National Cash Register Company (NCR) in Dayton, Ohio, built bombes at a pace that both sides knew foretold a shifting balance of power in the future. By the second half of 1943, the US produced seventy-five bombes – more than Britain managed throughout the whole war. These would process requests sent from Britain in an hour.18 The Americans were happy to take on some of this work as they understood that it gave them the chance to master a process that Britain had developed. They wanted to be in a position to be independent.19 The work at NCR and with companies like Kodak and IBM displayed the ability of the growing American private sector to re-engineer products to help the war effort. America had been at the forefront of experimenting with machines in the 1930s thanks to pioneers like Vannevar Bush, including support in code-breaking, but the focus on building bombes diverted manpower and expertise from the dream of building a proto-computer like Colossus.20
The Americans who made the pilgrimage to Bletchley during the war were particularly taken with the work on traffic analysis which logged data about German communications. An elaborate cross-indexed system of five-by-eight-inch cards was used, including punch-card machines. America had been using these machines from the 1930s, but Bletchley was more systematic in its application. This was used to build up knowledge about the communications pattern of existing known targets. But it had other uses. If you had a systematic understanding of what you already knew, it was also easier to spot something that was different. This might be, for example, a new type of signal or communications device – a target which, once discovered, might need monitoring. And as had been learnt in the First World War with the Zeppelins, even if you could not break the code of a system, you could glean useful intelligence. Based purely on the externals of the traffic in the Second World War, Britain was able to collate an order of battle for the Germans and scramble planes to intercept German bombers when a certain pattern of signals was detected. The latter process was performed by people called ‘computors’ who studied German air force radio call signs. They were able to report in near real-time the take-off and course of German fighters (and also tell the difference between fighters and bombers, which early radar struggled with). Their accounts in the US National Archives show how the Americans reported home that Britain might even have underestimated the value of this systematic use of data. The US became determined not just to replicate but to improve on the system, something British visitors would acknowledge by the end of the war. To sort this sea of data, a liaison unit from IBM was installed inside Arlington Hall to work out how best to apply the technology. This understanding of the hidden secrets that could be unlocked from raw data was a sign of things to come.21
Arthur Levenson, whose talent for maths was in part expressed through an amazing ability with baseball statistics, was one of those visitors who would build on his Bletchley experience to become a leading figure in American code-breaking. He always remembered the freezing outdoor toilets, as well as working with one colleague with a refined Cambridge accent and another from Glasgow. While he was made to feel very welcome, he never had a clue as to what either was saying. Levenson worked in Hut 6 developing the menus to be programmed into bombes breaking Enigma.22 There was one decrypt, he would always remember, in which Rommel asked why every supply ship was being sunk by a British destroyer. ‘That can’t be coincidence. That damn machine,’ Rommel had said. The team held their breath but the Germans shied away from facing up to the possibility that Enigma was broken and having to upgrade their entire communications system in the middle of a war. Another message Levenson decoded was a 70,000-character account by Rommel of his inspection of defences along Normandy before D-Day which mapped out the entire Western defences. This located a German Panzer division just close to where they were going to drop the 101st Airborne. ‘They would have been torn to pieces. So they moved it, based on intelligence . . . That – I’m sure – saved lots of young men in the 101st Airborne,’ he explained in a now declassified account.
With the invasion a success, Levenson was in Paris on VE Day heading towards Salzburg as part of a special team belonging to TICOM (Target Intelligence Committee). These teams, which drew on Bletchley veterans, fanned out across Europe hunting the secrets of Nazi Germany in a race with the Soviets. The full story of their work remains secret, but both the US and UK knew that getting hold of Nazi people and equipment would help them discover which of their own codes had been broken and also find out what progress had been made against Soviet codes, which they could then take advantage of.
Levenson went to Austria looking for an original Tunny machine, something which no one at Bletchley had actually seen. ‘Occasionally you’d get a guy and then we told him, if he doesn’t want to answer questions we’re turning him over to the Russians. And they talked,’ he remembered. His team drove an entire German communications train of six or seven vehicles packed with encryption equipment to Britain. Also retrieved were the Germans who demonstrated how it was used (some Germans would end up working with the NSA). Howard Campaigne, another Bletchley veteran, found the revelations from interviewing the captives were instructive. ‘We found that the Germans were well aware of the way the Enigma could be broken, but they had concluded that it would take a whole building full of equipment to do it. And that’s what we had. A building full of equipment. Which they hadn’t pictured as really feasible.’23
That’s what Britain had constructed – a building full of equipment at Bletchley Park. Howard Campaigne would go on to lead America’s quest to master machines in the years to come. But the more immediate question at the end of the war was: what to do with the rooms of equipment and the huge spy agencies that ran them?24 The seamless transition from Bletchley’s huts into GCHQ and the NSA was not as smooth as sometimes portrayed. There was fear, especially in the US, that tightened budgets would mean that all but the most cursory code-breaking capability would have to be abandoned, as happened after previous wars. But an important difference was that the most senior military and political leaders knew what the code-breakers had achieved.25 The goodwill would buy time in the difficult years ahead.
Britain knew that breaking Tunny was an immense achievement that might offer hope against future systems. The US meanwhile had never managed its own independent attacks on Tunny during the war. In August 1945, according to declassified US documents, Britain made a remarkable proposal that has never been talked about since. It offered to give one of its Colossus machines to the US and ‘hinted’ more would follow. To Britain’s shock, the offer was rejected. ‘They found it difficult to understand why and so did some Americans,’ a recently declassified NSA report notes.26 The reason was that the US was determined to build its own version. It did not want to be tethered to Britain but to be independent. Refusing Colossus was part of a broader strategy to be friends with Britain but also to ensure that America would lead and not follow. This was evident in wider relations between the two countries, including across the intelligence field. There were even some in the US signals intelligence world who believed the best way for them to stay afloat was to break the alliance with the UK. A number of ‘influential men in the intelligence community’ suggested ties with Britain be cut. The theory was that if policymakers knew they had to depend on the US code-breakers alone for the type of intelligence the alliance had provided, then there would be less chance of their budgets being slashed. But some early British successes against Russian cipher machines and their ability to intercept and process non-Morse transmissions like Tunny proved critical in persuading the Americans that continued co-operation was worthwhile.27
An Iron Curtain was rising across the centre of Europe, Stalin’s Soviet Union erecting barbed wire and border patrols to keep citizens in and spies out almost as soon as Nazi Germany was defeated. During the war, the US and Britain had both been – independently and secretly from the other – targeting the communications of their Soviet ally. Now, Stalin’s Soviet Union was steadily moving to take control of Eastern Europe, crushing democratic parties and dissent. It would come to be seen as the Cold War, but there was every chance at the time it might turn hot. The two Western Allies realised co-operation was going to be vital. The British, who had almost immediately turned their radio receiving stations towards Russian traffic when the war ended and had more history with Russia, were initially ahead. ‘The British seemed to be reading almost everything; the Americans virtually nothing,’ a declassified American study noted. ‘The British provided much of the cryptanalytic expertise, the Americans most of the processing capability,’ an NSA history recalls. Work against the Soviets was given the cover term ‘Bourbon’.28
Secret it may be, but code-breaking is also a team activity – as the work against Tunny showed. Two heads – and two countries – were always better than one. And so what might have been a brief affair solidified into a marriage, albeit one based on a formal written agreement rather than pure emotion. Recently released documents have for the first time revealed the nature of that marriage contract: a May 1943 agreement was expanded and eventually christened UKUSA. Those who negotiated the marriage had often worked closely together at Bletchley Park and would go on to rise to senior positions on both sides of the Atlantic, providing a deeply personal bond which extended beyond the formal provisions of the treaties.29
The two partners promised honesty and openness with each other. This meant they agreed to divide up the world in terms of what they intercepted and to unrestricted sharing of all raw traffic and technical work in order to eliminate duplication – with exemptions only when specifically requested by one party and agreed by the other.30 The aim was marriage, but with a bit of space to pursue their own interests so they could still respond to their own national requirements.
They promised fidelity – a clause said there would be no deals or sharing with any third parties unless one told the other. The British Dominions of Canada, Australia and New Zealand were eventually included, but only after Britain allayed concerns over leaky security and penetration by the Soviets. They also promised not to exploit the relationship for money. A clause outlines a blanket prohibition against passing material to any ministry, agency or individual who might use it for commercial competition or economic gain or advantage.31
They promised secrecy. The fear was that if other countries understood the vulnerabilities that were being exploited, they might move to more secure systems. ‘The value of Communication Intelligence in war and peace cannot be over-estimated; conservation of the source is of supreme importance . . . The time limit for the safeguarding of Communication Intelligence never expires.’32 It was agreed that no one who knew the secrets should be put in a position where they could be captured and subject to interrogation.33
They promised commitment – for better or for worse. And they made preparations for the worst. Not divorce but war, with detailed plans for how they could cope (GCHQ would grow to 9,000, with two-thirds from Britain and the rest from America).34 The British operation had moved to Cheltenham, a location chosen to be away from London since it would be a prime target for bombs. Some thought that even this was not far enough away and that the whole operation should be moved to Canada. But in 1947 a private visit to Cheltenham led to word of some old Ministry of Pensions buildings which had become vacant. Someone had a look and reported back that there were good communication lines from when the Americans had used the area as a wartime base and the town seemed a nice place to live. And so GCHQ, as it became known, found itself in the West Country.
Meanwhile in the US, the army and the navy had eventually begun to co-operate – motivated partly by a fear of budget cuts but also by the need to agree joint collaboration with the UK. The disaster of Pearl Harbor had increased pressure for a more centralised system, a push driven further after failings in the Korean War. This led in 1952 to the creation of the NSA. A military officer would be in charge but with a civilian deputy. Its home at Fort Meade was the third largest government building after the Pentagon and State Department. A pneumatic tube system could carry papers at twenty-five feet per second and handle 800 message tubes per hour.
Joined in matrimony, British and American code-breakers expected to be inside all Eastern Bloc intelligence and diplomatic communications, and even inside the top Soviet machine that carried the highest level of communications as the Cold War began. Engineers were building analogues of the Russian ciphers as they had done for Enigma and Purple and building the equivalent of bombes to focus on solutions to specific machines.35 Everything looked set for a repeat of wartime success against Germany. Soon, they thought, they would be in Moscow’s mind, as they had been in that of Hitler’s Berlin. They were wrong. And it would be that failure that would shape the history of computers and of spies.