The inequities of this curiously one-sided ‘co-operation agreement’ did not end there. The Americans also insisted that the naval section of Nave's unit be totally absorbed by Fabian's FRUMEL; that Nave, an Australian, be sent ‘back to Britain’; and that the British ‘abandon naval cryptanalysis at Kilindini and retain there only an exploitation unit which will read traffic from recoveries supplied by other units and supply to these other units any code or other recoveries obtained in the course of this reading’.
Washington would assume responsibility for sending ‘pertinent naval information’ to GC&CS which could then pass it on to Somerville and Kilindini. The agreement made it clear that this would include all Japanese naval code and cipher key recoveries plus ‘radio intelligence from Japanese naval communications indicating major strategic moves in any area and any details bearing upon operations in the Indian Ocean Area’. GC&CS would also be sent all intercepted Japanese material by pouch. ‘The British will withdraw from active cryptanalytical work in the Pacific Area but will continue to intercept and read Japanese traffic at Kilindini. The British plan however to maintain a research and intelligence unit at GC&CS so as not to lose touch with the Japanese problem. With regard to German communications, the British accede to US desires to attack the submarine and naval problems.’
It was clear that the US Navy at least had a very different idea from the British of what constituted co-operation. Birch was distinctly unhappy with the results of the negotiations. FRUMEL did take over the naval interception section in Melbourne, including the former FECB codebreaker Lieutenant-Commander Alan Merry. But the signal informing Shaw of what had been agreed was a much watered-down version, describing Kilindini's new role as ‘to act mainly as book recovery and exploitation centre for C-in-C Eastern Fleet and any special problems such as JN167 which they can undertake’.
This left open a loophole through which Kilindini's operations would continue much as before, but with the added assistance of American recoveries, albeit through Bletchley Park rather than Melbourne. But even in its watered-down version it was a demoralizing blow to the British codebreakers who had hoped to be reintegrated into the US codebreaking network as an equal partner in the attempt to break JN25.
The dismissive, almost arrogant approach adopted by Washington towards the British codebreakers is difficult to understand given that without their greater experience in breaking JN25 it would have taken the Americans very much longer to get on top of the Japanese code. That the agreement came at the same time as the British were breaking JN40 and JN167 only added insult to injury. ‘Unfortunately, it was difficult to get full confidence and co-ordination because the United States codebreakers had the idea that they were ahead of their British colleagues and were consequently very restrained in their cooperation,’ one British codebreaker said.
Brigadier W. A. Jolly, Royal Marines, a member of British naval intelligence who visited Melbourne during a tour of the Pacific theatre, put it more succinctly. ‘The most notable feature was the inability of the Americans to appreciate the full meaning of the word “co-operation”,’ he said. ‘The atmosphere was “What is yours is mine, and what is mine is my own.” ’
Throughout 1942 and into 1943 British attacks on the Imperial Japanese Navy's codes and ciphers remained hampered by the lack of co-operation from the Americans and the emphasis at Bletchley Park on winning the war in Europe. While the latter was simply a reflection of the Allies’ agreed policy of ‘Europe first’, the former had to be sorted out if the British codebreakers at Kilindini were to make any headway. More importantly, the limitations on the amount of signals intelligence being passed to the Admiralty for Somerville's Eastern Fleet under the agreement with the Americans was causing serious concern both in London and Mombasa. The British were supposed to receive all strategic intelligence on Japanese naval activity plus details of any Japanese movements in the Indian Ocean area, or more specifically west of 110 degrees longitude. In practice, the Americans showed little interest in sending anything. The cooperation still remained very one-sided. ‘Granted they had the leadership, we were bound to help them where they needed help,’ said Birch. ‘By and large as time went by and they grew bigger and bigger they needed less and less help and would therefore be less and less obliging and more obstructive.’
The problem threatened to rip the fledgling intelligence cooperation arrangements between Britain and America apart since the Admiralty was rapidly coming to the conclusion that if the Americans were not prepared to share their intelligence, the British would be better off going it alone.
‘The lack of US intelligence supply to C-in-C Eastern Fleet led the British to consider ditching the Americans on the Japanese side,’ said Birch. ‘Admiralty was not willing to be dependent on such small scraps as US were willing to provide and the only alternative to sharing all available intelligence between the two countries was for this country to build up independently an organization big enough to provide, without American help, as much intelligence as could be got with American help.’
The British set about building up their Japanese codebreaking efforts both at Bletchley and Kilindini. Shortly after arriving to replace Harry Shaw as the head of Kilindini in February 1943, Bruce Keith concluded that the problem lay in the inherent shortcomings of the agreement with OP-20-G. His report reflected the disappointment felt both at Bletchley Park and Kilindini at the lack of American co-operation and the resultant rows. The British were too dependent on the Americans taking an interest in matters that were essentially only of interest to the British, he said. Admiral Somerville was right: the only answer was to exchange liaison officers. British liaison officers inside OP-20-G would be able to pick out the messages that the Commander-in-Chief of the Eastern Fleet was interested in, said Keith.
This agreement depends entirely on the goodwill of the higher US authorities, and the efficiency of the US Japanese interpreters working on the cipher. Messages may contain information vital to the Commander-in-Chief [Eastern Fleet] but whether he gets this information or not is entirely beyond his control. It is not much good coming to an agreement in so vital a matter if you have no means of checking up that it is being efficiently carried out.
An American interpreter picking up a message mazy see that it affects the Commander-in-Chief, Eastern Fleet's area, but may say to himself that it is not important and anyway there are too many unknown groups to get anything out of it, so he passes on to the next message. If one of our interpreters were working in this room then this message would be passed to him and he, realizing its importance, would make every effort to find meanings for the missing groups and to extract as much information as possible. The US intelligence officer going through the translator's efforts may or may not pick out a message as being of interest to the Commander-in-Chief. If we had one of our people in this room, he would have a far better idea about this, and if there was a partly decipherable message which looked as if it might be of considerable interest, he would take the trouble to go back to the translator and ask for further elucidation. A US officer might not take this trouble.
This was the logical solution to the British problem. The difficulty was that it was just that, a British problem. The poor reception at Kilindini meant it could only intercept messages sent by the main stations using powerful transmitters and these were all easily accessible to both Melbourne and Hawaii. There was nothing that OP-20-G needed on the Japanese side from the British and as a result it had no incentive to co-operate. Its refusal to allow Birch access to its intelligence section emphasized the difficulty of implementing Keith's proposals. ‘Bouncer’ Burnett, who despite his poor relations with some of his British colleagues appeared to get on well with Commander Joseph Wenger, the head of OP-20-G, was keen to act as a Royal Navy liaison officer in OP-20-G. But Travis refused point-blank to contemplate Burnett's staying in Washington. ‘Colonel Tiltman informed me that, although there was nothing personal in it, Commander Travis would in no circumstances agree to my a
ppointment for liaison with US in Washington,’ Burnett recalled. The official reason at least was that his ability at Japanese was too precious to be spared.
Burnett returned from his visits to Washington and London convinced that Bletchley Park was deliberately preventing Kilindini from having a direct link to Washington in order to bolster its own position. ‘The Kilindini unit will suffer from malnutrition while BP grows fat on the pick of the food,’ he said. This erroneous view, which was undoubtedly widely held among members of the East African station, remained a constant irritant in the relationship between Kilindini and Bletchley. It was reinforced by the fact that it often took six days for Washington's recoveries to reach Kilindini via Bletchley Park. But Burnett's claims were rebuffed by Commodore Edmund Rushbrooke, a former member of the Far East Combined Bureau who had now taken over from Godfrey as Director of Naval Intelligence. Rushbrooke pointed out that Bletchley Park had repeatedly asked for most of what Burnett was demanding but it had always been blocked by the Americans.
Several other senior naval intelligence officers had been dispatched to the various US Navy codebreaking operations and to Kilindini to report on what could be done. They included Captain H. R. M. Laird, Admiral Somerville's chief intelligence officer, who visited Melbourne to try to find some way of improving relations. His mission was not a success. ‘Fabian was friendly,’ said Laird. ‘But when it came to producing a practical solution of our mutual problem he raised every sort of objection.’
13
BREAKING THE MILITARY ATTACHÉ CODE
In the wake of the victory at Midway, both MacArthur and King urged that the Allies should build on the momentum with an offensive aimed at recapturing a number of the Pacific islands taken by the Japanese. MacArthur wanted a direct army-led thrust at Rabaul, the main Japanese base on the island of New Britain; King believed the US Marine Corps should be thrown into a series of island-hopping operations. The agreed compromise involved a pincer movement with US Marines moving north-west through the Solomon Islands towards New Britain, while MacArthur's joint US–Australian force seized control of New Guinea as a stepping stone for the assault on Rabaul.
The phased operation began on 7 August when the 1st US Marine Division landed on Guadalcanal, swiftly overcoming the small Japanese garrison and taking the neighbouring islands of Tulagi, Gavutu and Tanambogo. But the Japanese, anxious to prevent the Allies from gaining the upper hand, poured troops on to the island, with its navy making regular night-time forays down the ‘Slot’, the channel between the islands to the north-west, in a move dubbed the ‘Tokyo Express’.
Despite John Tiltman's pre-war work on Japanese military codes and ciphers, the Allies were unable to read any of the high-level operational army messages during the bloody battle for Guadalcanal, and although low-level plain text was available, along with traffic analysis, much of the sigint assistance during this period came from decoded Japanese Navy JN25 messages. One high-level code was, however, available to the Allied commanders. Tiltman's roles as Chief Cryptographer, head of the Bletchley Park military section and troubleshooter on liaison with the US codebreaking operations kept him very busy. But throughout the first half of 1942, he had spent much of his spare time attacking a new code used by Japanese military attachés based in Japan's embassies abroad (the JMA code as it was dubbed at Bletchley Park). The codebreakers knew on past experience that this was likely to provide a wealth of useful intelligence on Japanese activities, filling in the gaps caused by loss of continuity on the high-grade army systems.
‘About the end of 1941, I had to rescue the material from a party of French cryptanalysts who had remained with us after the fall of France,’ said Tiltman. ‘I had given it to them as a task when I and my research section were fully occupied with German problems and they had diagnosed it wrongly as a combined substitution and transposition system and had got the intercepts in a hopeless tangle.’
Tiltman took up the problem himself using messages sent back to Tokyo by the Military Attaché in Santiago and discovered that the JMA code was a digraph code in which the basic kana syllables stood for themselves and other two-letter groups stood for certain words or phrases commonly used in military communications: for example, AB stood for ‘west’ and AV for ‘message continued’. The two-letter groups were then set out in a square grid in adjacent squares, sometimes horizontally and sometimes diagonally, and the letters were read off vertically to form the basis for the encrypted text. They were then enciphered using a pre-arranged ‘literal additive’, a series of letters that would be notionally ‘added’ to the letters taken out of the grid on the basis of a pattern laid down in advance on a separate table. Reading off the enciphered letter along the relevant horizontal line and the ‘additive’ letter down the appropriate vertical column would produce a superenciphered letter which would be transmitted by the operator.
‘By the time I first went to Washington in March 1942, I and my section had partially recovered the indicating system and had diagnosed the cipher as a literal additive system with indicators which gave the starting and ending points for messages,’ Tiltman recalled. ‘It became clear that the normal practice was to tail successive messages rigorously through the additive tables: i.e., to start reciphering each message with the additive group following the last group of the preceding message.’
After returning from Washington Tiltman set to work on a large number of JMA messages emanating ‘from some unplaced station in the middle of Europe’. The cipher clerk in this as yet unidentified Japanese embassy had used the additive table again and again, giving a large ‘depth’ for Tiltman to work on. ‘It was clear from the indicators that the sender had tailed right round his additive table five times and it was this depth that I set myself to resolve.’
Tiltman found similarities were particularly strong in the first five groups of each message, where the same letters appeared frequently in the same position in different messages. ‘From this it could be deduced that the system was a true additive system, addition being in normal cyclic alphabetic order,’ he said. The solution took a lot of work, but with a depth of five on the cipher additive, he eventually managed to break the system. One of the first JMA messages deciphered revealed the Japanese intentions to construct a ‘Burma Railroad’. It was not until several months later that it became clear from another Japanese military attaché decrypt that British prisoners-of-war would be used as slave labour to build the railway.
Tiltman set up a small Japanese military section at Bletchley Park in June 1942. It comprised a codebreaking sub-section and a traffic analysis team but its main purpose was to handle the JMA material, using some of the graduates of the Bedford Japanese course. These included Maurice Wiles, who was told he must join the army and then transfer to the reserves to ensure that he was not called up for another unit.
Three of us, Mervyn Jones, myself and Roland Oliver, were sent to the army recruitment office in Bedford where we were interviewed by this caricature of an old-style army colonel. He asked me which school I had gone to and when I said Tonbridge, he told me he knew the school secretary. ‘We were fellow brigade commanders in Lucknow,’ he said. He then asked me if I played cricket. ‘Yes,’ I said. He beamed. ‘Were you in the first XI?’ ‘Yes,’ I said. He beamed again. ‘Did you bowl?’ ‘Yes.’ He beamed again. ‘Did you bowl left-handed?’ It was at that moment that I fell from grace. He then interviewed Mervyn Jones who had been at Trinity, Cambridge, and had a starred first in Classics. ‘Got your school certificate, eh?’ he said to him. ‘Well done. Well done.’
Tiltman explained to the new recruits that although he had broken the cipher, his knowledge of Japanese was not good enough for him to take it any further forward. That was up to them, he said. It was initially an intimidating task, Wiles said.
We were pretty ill-equipped. Knowledge of the language was essential to the task assigned to us, but it was no straightforward matter of translation. As yet there were no texts on which to exercise our newly acquir
ed translation skills. None of the thousand or so characters that we had painstakingly learned were there on the page before us. Something much more was needed, for which we had no specialist training – an approach to problem-solving that our initial interviewers no doubt hoped had been ingrained in us by our interest in chess and crosswords.
But eventually the messages began to flow, producing a good deal of intelligence and, on occasions, amusement. ‘Foreign names were normally spelled out in kana,’ Wiles recalled. ‘Until one got used to them, they were not always easy to recognize: CHI-YA-A-CHI-RU does not obviously spell “Churchill” to the untrained eye. But once the principle was mastered, they offered plenty of entertainment as well as a quick guide to the subject matter.’
Thereafter, the codebreakers were able to read the messages of the Japanese military attachés without any problems, aided by the fact that each time the code changed, the difficulties of distributing code-books dictated that some were forced to respond to messages from Tokyo in the new book using the old broken one. The messages provided a wealth of useful material on the movements and existence of Japanese military units. Much of the more valuable intelligence came from the so-called Tokyo Circular, which was sent to all military attachés around the world and contained a rundown of Japanese activity on all fronts.
The Emperor's Codes Page 19
