Here at least was something they could speak of freely. Once Alan said at lunch, ‘Shannon wants to feed not just data to a Brain, but cultural things! He wants to play music to it!’ And there was another occasion in the executive mess, when Alan was holding forth on the possibilities of a ‘thinking machine’. His high-pitched voice already stood out above the general murmur of well-behaved junior executives grooming themselves for promotion within the Bell corporation. Then he was suddenly heard to say: ‘No, I’m not interested in developing a powerful brain. All I’m after is just a mediocre brain, something like the President of the American Telephone and Telegraph Company.’ The room was paralysed, while Alan nonchalantly continued to explain how he imagined feeding in facts on prices of commodities and stock, and asking the machine the question ‘Do I buy or sell?’ All afternoon the phone was ringing in his laboratory, with people asking who on earth it was.
On 2 February 1943, the German surrender at Stalingrad had marked the turning of the tide. But while the eastern front was turned by sheer brute force, the western powers had the space and time for developments in which force was not the only element. The intricacy and sophistication of their cryptanalysis was the most extreme example, but this was not the only sphere in which machinery was taking the war out of the old world of duty and self-sacrifice. In November 1942 the ground had been cleared at Los Alamos, and by March 1943 the first scientists were already moving in. The atomic bomb they planned to make would not release greater energy than the raids of 1943 already deployed against Germany. But it would make thousands of bombers redundant, thus mechanising the discipline and coordination of air offensives. The Manhattan Project would still depend upon an aircraft pilot – but then he, too, was being automated at Peenemünde, where the long predicted ‘monster cannons’ were on the way. The V-weapons would lack sufficient accuracy – but such problems of guidance were also being attacked at Germany’s back by the new techniques of proximity fuses, automatic celestial navigation, and automatic fire control. People easily understood powerful guns, fast ships, impenetrable tanks, which extended human limbs. By now the secret of radar was out, and it could be understood how its manifold applications extended human eyesight to the longer wavelengths of the electromagnetic spectrum. But rapidly developing, and not only at Bletchley and Washington, was a new kind of machinery, a new kind of science, in which it was not the physics and chemistry that mattered, but the logical structure of information, communication, and control.
This development was not confined to warfare. In Dublin, Schrödinger was lecturing under the title ‘What is Life?’, and advancing the conjecture that the information defining a living organism must somehow be encoded in molecular patterns. In Chicago, two neurologists had read Computable Numbers, and were publishing10 an idea that connected the definition of the logical machine with the actual physiology of brains. They had applied Boolean algebra to the properties of nerve-cells. As Hilbert died at Göttingen on 14 February 1943, a new kind of applied logic was taking shape. Against the distant thunder of the east, there were the first glimpses of a post-war science. This first half-serious, half-joking talk of ‘thinking machines’ reflected both the immensely wider horizon that the war had opened to science, and the fact that an end at last seemed possible.
By 4 March Alan had completed a report on his suggestions regarding the RCA speech cipher, and had studied in great detail all the speech systems with which they were working. The head of the section had expressed concern that Alan might invent something which would create a tangle over patent rights, but Alan pooh-poohed this, saying that he wanted Bell Telephone to have anything that he thought of. ‘Hands across the sea’, he said. But what idea could possibly compare with those he had already handed across the sea, ideas far too important for any patent office to know existed? From 5 to 12 March he had to spend another week in Washington at the request of the Navy, to look after this side of his mission again. It was another critical point for the U-boat Enigma, for on 10 March the codebook for the short weather report signals, on which the December breakthrough had been based, was withdrawn. But the three months of successful decryption had allowed analysts to develop alternative methods in time, finding in particular that other ‘short signals’, allowed to remain in force, were also enciphered with the fourth wheel in its ‘neutral’ position. Again the German force threw away its advantage, and with more than sixty Bombes at Bletchley now, the Allied dependence upon this special trick was lessening. The change of 10 March was overcome in just nine days.
Returning to Bell Labs, Alan worked for a few more days on the RCA cipher. He wanted them to keep him informed on their progress after he had returned to England, and indicated two possible means of communication: either through Friedman, or through Professor Bayly, a Canadian engineer attached to British Security Coordination. At a quarter past four on 16 March, he had a telephone call from BSC, telling him it was time to embark. He stopped work and left the West Street building within half an hour. His ship11 was not a Queen this time, but the 26,000 ton troop transporter Empress of Scotland. This British vessel could sustain 191/2 knots, while packed with 3867 enlisted men, 471 officers – and just one civilian.
After a week’s delay, the Empress of Scotland left New York harbour on the night of 23 March. She steamed due east into mid-Atlantic, and then swung up to the north. Only one of the thousands carried into the midst of the battle knew the precarious system upon which so much depended – but that knowledge made no difference now. For a week Alan was ordinary again, having to take the risks and trust the authorities like everyone else. The danger was real enough: on 14 March the similar Empress of Canada had been sighted and sunk. Briefly he was on the receiving end of the system, and briefly he was relieved of responsibility for it.
In a sense he had stood upon the burning deck since 1939; but the sentiments of Casabianca, those of doing a patriotic duty contrary to inclination, were very remote from the spirit of Alan Turing’s war. He was doing what he had chosen to do, and was expressing himself, not subjecting himself. His mind continued to work away, fascinated by the problems, even during this voyage home. While briefly sharing in the helplessness, confinement and danger of the war, he spent his time studying a twenty-five-cent handbook on electronics, the RCA Radio Tube Manual, and invented a new way of enciphering speech.
If he dreamt about fighting in the War between the States, it was the reverse of the truth. He was committed to the Yankee side, and had seen no fighting. His struggle had all been behind the lines. But that was not quite the whole story. Once in conversation with his friend Fred Clayton, the question arose as to how scientists could have continued to work for Germany. With both personal honesty and political realism, Alan pointed out that in scientific research it was inevitable that one became absorbed in the work, and did not think of the implications. It was, in this respect, a Looking Glass war, with the B. Dienst analysts no less fascinated by their work.* It could be an entrancing dream world, without connection with the issues of the war. But Fred did get him to admit that Germany raised other questions.
For Alan Turing’s generation, the First World War had been a War between the States, meaningless as Tweedledum and Tweedledee. The mirror symmetry of nationalism had disgusted Russell and Einstein, Hardy and Eddington, who saw only human beings with labels pinned upon them, destroying each other. They longed to jump out of the system of La Grande Illusion, and in 1933 the new generation had voiced this longing openly. But Russell and Einstein came round to support of this war, the war for the Anti-War, the war that could be imagined not as the ‘national war’ but the world Civil War, a crusade against slavery. That it was primarily a war between two tyrannies; that it had massively reinforced national governments; that it had made mass slaughter respectable again; that it had militarised the advanced economies – these did not countervail. Against this enemy anything could be justified. In 1933 they had reviled the arms manufacturers above all others. But they were all arms manufa
cturers now.
There had been British atrocities in an Ireland now so obdurately neutral, but not with filing systems, medical experiments, and industrial cyanide. At Bletchley they had already deciphered some of the figures that Germans did not know, or want to know. That sheer explicit single-mindedness, in following ideas to a logical conclusion, was what lay beyond the grasp of English minds. But that Nazi definiteness had helped to stimulate the scientific consciousness without which the western Allies, at least, would have been helpless.
This dimension of the war went without saying, and did not need talking about. Yet in Alan Turing’s case there was a sharp irony,* in that Himmler had sneered at British Intelligence for making use of homosexuals, and had specifically directed that in Germany useful talents could not exempt those so identified from the general rule. Few indeed could have appreciated that irony, and fewer still have believed that this strange civilian on the Empress of Scotland was playing as much of a part as any in bringing Himmler to his own poison.
In 1939 Forster13 had expressed the numbing conviction that to defeat fascism it would be necessary to become fascist. It had not happened like that, and in many ways the channels of communication had been opened up. Yet far more subtly, the logic of the game was reflecting something inhuman into what was called democracy: not just in the bombing raids alone, but in a deep internal way. As the Allied war turned from defence to offence, from innocence to experience, from thinking to doing, an undefinable naiveté was going with the wind. The very success and efficiency of its scientific solutions was bringing this about. In 1940 there had been a feeling, quite illusory perhaps, of individual contact with the course of events. But now even a Churchill was dwarfed by the scale and complexity of operations. In the 1930s it had seemed that there were simple choices to be made between good and evil. But after 1943, as the Allies prepared to join the Russians in biting on the Nazi apple, nothing would be simple again. Nothing could even be properly known.
In the cold dawn of 31 March, a British escort was waiting for the Empress of Scotland in the Western Approaches. The danger was passed, no U-boat having sighted the ship, and the odd civilian returned safely to his country. For three years now he had helped to stem the tide by thinking, and they had built a colossal machine around his brain. But they could not fight the war by knowing about it. Intelligence was not enough; it had to be embodied in a savage world. Nor would its engineer escape that general rule.
*Curiously, they had not found this by any means an obvious idea, although it was just like the base-10 modular addition used in one-time pad ciphers. They had invented it afresh.
† They had also independently invented a form of pulse-coded modulation.
*The British did not get their way regarding the location of the London terminal. In April the X-system was installed in the American headquarters and only later was a line run to Churchill’s war room.
* At least one of Scholz’s students was working directly against him.
* German security policy was more advanced than the British. In a letter12 of 9 October 1942, Himmler replied to a memorandum from the Consultant Physician to the Reichsicherheitshauptamt (Supreme State Security Office) on the subject of die Homosexualität in der Spionage und Sabotage. ‘I grant you … that the British have found some rather promising (passender) material for their purposes here,’ he wrote, but decreed that there was no question of a remission in the vigorous prosecution of homosexuality for the sake of gaining recruits, in view of the risk of homosexual vice rampaging unpunished amongst the Volk, and whole sections of the youth being seduced. Anyway, he said, if one of these degenerates and crooks (Pathologen und Gauner) were set on betraying his country, he would do so whether punished according to Paragraph 175 or not. Prosecution, in 1942, meant consignment as a ‘pink triangle’ prisoner to a concentration camp. The doctors were sharply rebuked by Himmler on 23 June 1943 for their suggestion of retraining (Erziehungsversuche an anormalen Menschen) as a waste of effort at a time when Germany struggled for its existence, and because the outcome of such efforts was so dubious (höchst zweifelhaft).
Part Two
THE PHYSICAL
5
Running Up
One’s-self I sing, a simple separate person,
Yet utter the word Democratic, the word En-Masse.
Of physiology from top to toe I sing,
Not physiognomy alone nor brain alone is worthy for the Muse,
I say the Form complete is worthier far,
The Female equally with the Male I sing.
Of Life immense in passion, pulse and power,
Cheerful, for freest action form’d under the laws divine,
The Modern Man I sing.
The surrender at Stalingrad had marked the beginning of the end for Germany. The war had turned. Yet in the south and west there was little evidence of progress for the Allies. The African war dragged on, the Luftwaffe still mounted raids on Britain. And the ports were sheltering the survivors of what had been the most damaging convoy battle of the war, fought in mid-Atlantic while Alan had waited in New York.
When Churchill and Roosevelt had conferred at Casablanca, they had good reason to suppose that, with Atlantic U-boat Enigma restored, the sinkings could be kept down to the level of late 1941. In January they were. But in February they had doubled, nearly back to 1942 levels. And then in March they were the worst of the war: ninety-five ships, amounting to three-quarters of a million tons. Massed U-boats had been able to sink twenty-two out of the 125 ships that had set out in convoy on the eastbound Atlantic passage that month. There was a reason for the deteriorating Allied control of events, one scarcely credible. It was not just that the convoys had sailed during the nine days’ blackout caused by the change to the U-boat weather report system. It was that all the time, and to an ever-increasing degree, the convoy routeing cipher, among others, was being broken by the B. Dienst.
Convoy SC. 122 had started out on 5 March, HX.229 on 8 March, and the smaller and more fortunate HX.229A the next day. On 12 March, SC. 122 was re-routed to the north to avoid what was thought to be the position of a U-boat line, the Raubgraf. This signal was intercepted and deciphered. On 13 March the Raubgraf attacked a westward-bound convoy, thus openly betraying its position; SC.122 and HX.229 were both diverted again. Both diversion signals were intercepted and deciphered within four hours. The Raubgraf group could not catch up with SC.122 but 300 miles to the east, the forty-strong Stürmer and Dränger lines were sent to intercept them. There was ill luck on the German side – they were confused as to which convoy was which – but good luck too, for one of the Raubgraf happened to sight HX.229 by chance, and beckoned the others on. In London they could see the two convoys moving into the midst of the U-boat lines – but it was too late to do anything but to have them fight it out. On 17 March, they were surrounded by U-boats, and over the three days that followed twenty-two vessels were sunk, for the loss of one U-boat. Chance had played its part in this particular action, but underlying these and other current engagements lay the systematic failure of Allied communications.
In London and Washington, the first suspicion that this was so had been aroused in February 1943, when it was noticed that three U-boat line diversions were ordered within thirty minutes to operate successfully against a convoy on the 18th. Clear proof came only in mid-May when three doubly-enciphered Enigma messages showed evidence of the decipherment of particular Allied transmissions. Identifiable Enigma information had gone since 1941 into one-time-pad messages, and so had not been directly compromised. But it was implicit in the daily U-boat Situation Report, which by February 1943 was being decrypted. Yet again, the German authorities imputed Allied knowledge to a combination of airborne radar and the treachery of their officers. In a futile gesture, they reduced the number of people allowed to know about U-boat traffic. Again and again, only an a priori faith in the machine prevented them from seeing the truth. The Allies had very nearly given their own game away.
It was a dismal story, not perhaps one of individuals, but of the system. Neither in London nor in Washington was there a section in a position to do the very difficult detailed work of sorting out what the German command must have known, from what it could have known. The cryptanalysts were not given access to Allied dispatches – of which, in any case, there was no complete record. At the OIC they were still understaffed, underequipped, and under great strain with the convoy battles.
The cryptographic and operational authorities were working to standards which to Hut 8 eyes would seem criminally negligent. For one thing, the convoy routeing cipher, introduced as a joint Anglo-American system, was in fact an old British book cipher which the B. Dienst were able to recognise. Although in December 1942 a ‘recipherment of indicators’ had caused a setback for the B. Dienst, every kind of mistake was still being made. According to the American post-mortem:1
USN-British Naval Communications were so complex, and often repetitious, that no-one seemed to know how many times a thing might not be sent and by whom – and in what systems. It is possible that the question of cipher compromise might have been settled earlier than May had the Combined Communications system been less obscure and had there been closer cooperation between the British and the US in such matters.
Alan Turing: The Enigma The Centenary Edition Page 41