Bill Sibley was one of the interpreters translating the Japanese Military Attachés’ messages.
I was recruited from Balliol (I was a classicist, at the end of my first year) for the second Japanese-language course at Bedford, which began in September 1942. We were summoned to a five-minute one-to-one interview with John Tiltman, having been pre-selected by the Master of Balliol, A. D. Lindsay. I then went to Bletchley in the spring of 1943 and was set to work as a translator on the Japanese military attaché code until the end of the war, apart from a period of a few months when I was seconded to work on the Japanese naval attaché cipher.
Our work began after the real cryptographers had done their work and identified where in relation to the double substitution cipher ‘keys’ individual messages were located. The texts on which we worked were provided for us by ‘key-breakers’ who were not trained in the language, and whose task was to ‘break’ the keys used in the messages, relying on acquired familiarity with the frequencies of the bigrams in which the messages were composed before being enciphered.
We lived an introverted existence, insulated from the real world. Our masters did occasionally send us words of encouragement, but I can't recall that at our level we were ever told of any examples of our work having produced any positive results. Nobody knew what was going on in the rest of the place. It was a funny life, very funny, particularly the secrecy, and the oddity of some of the people. There was one famous professor of English who used to read about three detective novels a day. He used to walk around the grounds reading them.
In early May the Japanese Naval Attaché in Berlin made his own tour of the German defences. Sent in the newly broken Coral JNA20 cipher, his report was easily read by the Allies. It was more authoritative than that of Oshima, whose pro-German tendencies led him to accept unquestioningly what he was told. Rommel, who had been appointed to lead the main force resisting an invasion, intended ‘to destroy the enemy near the coast, most of all on the beaches, without allowing them to penetrate any considerable distance inland,’ the Naval Attaché said. ‘As defence against airborne operations, he plans to cut communications between seaborne and airborne troops and to destroy them individually.’
The report gave detailed appraisals of the German dispositions and intentions and, worryingly for the Allies, said Normandy was regarded as a prime target and was being reinforced. This trend was confirmed by a report from Oshima of a meeting with Hitler at which the Führer had told him that the British were expected to establish an initial bridgehead in Normandy before launching the main front against the Pas de Calais.
It was this information that enabled the British to use the so-called double-cross agents to refine Fortitude, the D-Day deception operation, to weaken the defences in Normandy by reinforcing Hitler's suspicions of a second front. MI5 had turned the majority of Nazi agents who were sent to Britain – the rest were jailed or hanged – and was using them to feed back disinformation to the Abwehr, German military intelligence.
The double-cross agents built up the impression that the Normandy landings would be a feint. The real assault was to be mounted on the Pas de Calais by the 1st United States Army Group, a completely fictitious organization created with false reports from the double agents, dummy invasion craft planted in east coast ports and mobile wireless vehicles travelling around south-east England broadcasting messages from a number of different locations to fool the German radio interception units.
But although the information relating to German preparations for the Allied invasion of Europe was the most high-profile material obtained from the breaking of Coral, the Japanese Naval Attaché also gave far greater detail of German technical developments than Oshima or Ito. He had close contact with the chief of the German Naval Operations Division and through him the Allies heard the first details of the new, faster U-boats under development for the German Navy just a few weeks after Coral was broken. He also provided the precise specifications of the new Arado 234 and Messerschmitt ME-163 jet aircraft that were being introduced for the Luftwaffe.
Professor R. V. Jones was in charge of British scientific intelligence during the war and received much of his information from Bletchley Park.
The Japanese attachés even sometimes had with them men who were specifically termed ‘Scientific Intelligence Officers’ to assist them in gathering information about their ally. We watched their activities with interest, and could say, for example, that the Germans had supplied early forms of German radars, listening receivers for submarines and guided anti-shipping bombs, although we noted a reluctance to let the Japanese have the latest models. We also learned details of Japanese developments in airborne radar as they revealed them to the Germans. We were thus able to throw a useful sidelight on Japan, and one of my officers was posted for a time to Lord Mountbatten's command in South-east Asia.
Evidence of Japanese scientific abilities was also collected by a small Japanese commercial section co-located with the diplomatic sections in London's Berkeley Street and headed after his return from Kilindini by Captain Harry Shaw. Encoded and plain-text messages between Japanese companies and their offices abroad were intercepted at a number of interception sites around the world, including Edward Twining's Mauritius operation and the Canadian naval base at Point Grey, Vancouver.
While co-operation between the Bletchley Park and OP-20-G sections dealing with the Japanese naval attaché material remained good, there were, even at this late stage of the war, still clear signs of tension over the sharing of information on JN25. The new BRUSA network linking Colombo with Hawaii and Melbourne came into force on 27 June 1944, three weeks after the D-Day invasion. But even as it did, Colonel Tim O'Connor, who had replaced Geoffrey Stevens as Chief UK Liaison Officer in Washington, was complaining to Frank Birch that he had difficulty persuading the Americans to agree to ‘any new departure which will mean the US helping Great Britain to take a greater share in the Pacific War’.
The response from Birch revealed just how close the British had come to ditching co-operation with the Americans over the US Navy's reluctance to co-operate, a decision that would have had immense implications for the future of the close exchange of intelligence between the two countries. The US attitude was ‘no doubt the real reason for the general impression we get that whereas our co-operation is whole-hearted – almost too naïvely so – the Americans are co-operative with considerable reservations,’ Birch told O'Connor. He added:
OP-20-G's reluctance at times to go out of their way to meet our requests is due, to quote your own letter, to OP-20-G's feeling ‘that they lead the Allies in the Naval cryptographic effort in the Japanese war’ and that ‘GC&CS should begin by filling in the gaps which OP-20-G has been unable to fill in and only begin to duplicate OP-20-G's work when there is labour and coverage to spare’.
American caginess in the past has constituted a grave threat to our relations with them because Admiralty very categorically refused to be dependent on USA for Japanese intelligence. The lack of US intelligence supply to CinC Eastern Fleet led the British to consider ditching the Americans on the Japanese side. That could have been done; it still could be done. The test is the BRUSA Circuit. If the Americans play ball and circulate all their available intelligence, as the British will theirs, all is well.
In the end, after some initial caution by the US sites and a number of communication problems, the BRUSA Circuit did begin to work. But even then, while most of the material produced in Hawaii was swapped with Anderson, barely two-fifths of that emanating in Washington found its way to Colombo. Nevertheless, this was a great improvement on the previous situation. As a result, the productivity of the Royal Navy codebreakers improved markedly and the Americans began to see more reason to co-operate. Japanese suspicions that the Allies might be able to read JN25, heightened by injudicious use of signals intelligence by some US commanders, had led to security improvements that made it increasingly difficult to break.
In his tirade against OP-20-G's rel
uctance to co-operate fully with the British on naval codes, Birch pointed out that the gaps in the Allied ability to break the Japanese naval codes were ‘appallingly wide and, as far as JN25 is concerned, widening dangerously’. The tensions were beginning to show on both sides to the point where, in a moment rich in irony, Fabian complained to Admiral Somerville that material was being ‘deliberately held back’ from him by the Colombo codebreakers.
During the summer Bletchley Park began to throw an increasing amount of resources into the JN25 problem. Hugh Alexander was appointed to head naval section IIJ, the department attempting to break the main Japanese naval systems JN11 and JN25. His transfer to Hut 7 was a measure of the priority which Bletchley Park now attached to improving its record against the Japanese naval codes.
By late summer Bletchley Park was so concerned by the situation that Tiltman had also been called in to look at the problem. ‘The loss of JN25 over so long a period has deprived us of about 60 per cent of strategic intelligence and Naval Air Order of Battle is becoming rapidly out of date,’ one BP codebreaker complained. Recent signals from Colombo seemed to indicate that the codebreakers would soon be back on top of the problem, he said, but co-operation with Washington was apparently still not all that could be expected. ‘We are suffering considerably from not knowing at present how current is information in OP-20-G because of the paucity of intelligence from their end of BRUSA.’
Alexander flew to Washington in August in order to try to co-ordinate a joint attack and found that OP-20-G was struggling as much as Bletchley Park, although clearly less willing to admit it. But there was now a recognition in Washington that they needed the British contribution. During a visit to Colombo, Admiral Redman finally agreed to a direct link between the British outstation and OP-20-G. Bruce Keith was in little doubt as to the motive behind this. ‘The large increase in Anderson's output was the main reason for the Americans reversing their previous refusal to accord direct communication,’ he said. Colombo was now able both to intercept and to decode very many intercepts that were simply not available elsewhere.
In October 1944 Travis flew to Washington to put his signature to an updated agreement that, while maintaining the OP-20-G control over Japanese naval signals intelligence, at last offered genuine cooperation between the two sides.
19
OPERATION CAPITAL
The Japanese, who were determined to cut the Burma Road and isolate Chiang Kai-shek's Chinese Nationalist forces from supplies, had overwhelmed the British forces in Burma by the end of May 1942. Despite a limited British counter-offensive and the highly publicized but somewhat overrated efforts of Orde Wingate's Chindits, they were still very much in control of the country as 1943 came to a close.
Allied plans for a major campaign in Burma designed to divert Japanese forces that might otherwise be directed against MacArthur's drive northwards towards the Philippines were delayed by political arguments between the British and American commanders. The abrasive US general ‘Vinegar Joe’ Stilwell, deputy to Lord Mountbatten, the Supreme Allied Commander South-east Asia, refused point-blank to countenance any campaign which had as its ultimate aim the retaking of other British colonies, principally Malaya.
Eventually they agreed a more limited offensive under the command of the British general Bill Slim aimed purely at driving the Japanese out of Burma. Operation Capital was to begin in the dry season of early 1944. The Chindits would penetrate deep inside Burma to capture Indaw ahead of the advancing Allied forces. Stilwell's Chinese Nationalist troops were to invade from the north, aiming to capture the strategically important town of Myitkyina. Stilwell was to be supported by a Chindit-style US special forces unit which became known as Merrill's Marauders after its commander, Brigadier-General Frank D. Merrill. Meanwhile, the British and Indian forces of 15 Indian Corps were to advance down the Arakan region along Burma's western coast, capturing the ports of Maungdaw and Akyab. The main thrust would then come through central Burma, with the rest of Slim's 14th Army pushing across the Chindwin River towards Indaw and then south towards Mandalay.
The primary signals intelligence support for these operations was to come from the British via Bletchley Park and, more particularly, its outposts in India controlled from Delhi by the Wireless Experimental Centre, which, alongside its codebreaking operation, had eighty-eight radio sets intercepting Japanese communications twenty-four hours a day at various sites across India.
The 1943 Bletchley Park conference, which allocated responsibility for breaking the different Japanese high-grade military codes to various Allied signals intelligence centres, had given Arlington Hall overall control of breaking the General Army Administrative Code 7890. But the latter was nevertheless exploited by C Branch, the codebreaking division of the WEC, which had a staff of around two hundred.
Michael Kerry, a young RAF pilot-officer, was trained in Japanese and codebreaking in the relaxed and rarefied atmosphere at the School of Oriental and African Studies and Bletchley Park before being sent out to the WEC. It was something of a culture shock.
The trouble with the WEC was that it was much more militaristic than BP. It was such a contrast. Four of us went out in June 1944, all of us RAF officers. We were equipped for the Boer War with drain-pipe shorts and pith helmets. So we bought some new khaki shorts, which were fine, and some new hats which as it later turned out were not. We arrived at the WEC and were shown into Peter Marr-Johnson's office. He said he didn't like the RAF and he liked them even less improperly dressed, and he told us to get out. We puzzled over this and eventually realized that the new tan hats we had bought were in fact ‘Bombay bowlers’ as worn by Indian civil servants.
Despite being in the RAF, Kerry was put to work on high-level army codes.
I was sat opposite Wilfrid Noyce, the mountaineer, who was a very nice chap but had a very battered face. He had fallen off a number of mountains and must have had India-rubber bones. A great character in the section was Hugh Lloyd-Jones who became Regius Professor of Greek at Oxford. We worked on army messages, mainly using the 7890 code and carrying daily reports of strengths of things like weapons, rations, sickness. It was quite hard at times working with indistinct photographs of captured codebooks in a temperature of 117 degrees. There was no air-conditioning and an Indian woman was employed purely to sit by a great big woven straw door, called a khaskhas tatti, throwing handfuls of water over it every few seconds to keep the temperature down.
Unlike the Royal Navy, which had shown no qualms about dispatching Wrens to the far corners of the Empire, both the army and the RAF had been extremely protective of their female services and there were very few ATS and WAAF personnel posted abroad. So in complete contrast to Bletchley Park there were not many women around, Kerry recalled.
A friend of mine who took his sex life very seriously pursued an Indian servicewoman but that was it and I can't remember very much drinking or alcohol at all. When we weren't working on anything current we worked normal daytime hours and at six o'clock you went off to your little bungalow, you each had a bearer, and he served you with tea. Once a week, every Wednesday, we had a curry lunch and took little horse carts or ghari into Delhi to have tea.
By late 1943 the BRUSA agreement had ensured complete cooperation on military codes between the British and the US Army. The WEC had a direct link to Arlington Hall via its resident US Liaison Officer and indirect links via a relatively large Signal Intelligence Service analytical and intercept unit, Station 8, also based in New Delhi.
It also had direct dedicated signals intelligence chatlines to Bletchley Park and to Central Bureau in Brisbane, over which it could discuss recoveries, as well as direct contact with its two main outstations, the Western Wireless Signal Centre in Bangalore, where, owing to seasonal difficulties with reception in Delhi, much of the interception was now taking place, and the Eastern Wireless Signal Centre at Barrackpore, near Calcutta. It was the intelligence and codebreaking operation in Barrackpore, known for cover reasons as Intelligen
ce School ‘C’, which provided the immediate signals intelligence needs of Slim's 14th Army, the main force involved in the invasion of Burma, with the assistance of a number of mobile Special Wireless Sections.
Air intelligence operations were covered by the Tactical Air Intelligence Section (TAIS) based at Comilla, the headquarters of both 14th Army and the 3rd Tactical Air Force, and by a number of RAF wireless units. The section had its own intelligence and code-breaking operation quite separate from Delhi and Bangalore which provided detailed information on the traffic of Japanese aircraft into and out of airfields in Burma. The 3rd Tactical Air Force detailed a squadron of USAAF Lightning long-range fighters to react to TAIS intelligence and the US pilots enjoyed ‘spectacular’ success during a series of snap sweeps of enemy territory, shooting down 135 enemy aircraft and destroying a number of others on the ground, thereby ensuring Allied air superiority at a crucial point in the offensive.
One of the codebreakers sent to the Wireless Experimental Centre at Anand Parbat to keep track of the heavily depleted Japanese Army Air Force was Alan Stripp, a young Intelligence Corps subaltern who had worked on army air codes in the Bletchley Park military section. Stripp recalled his arrival at the WEC.
At the bottom of the hill, further away from the road, was the lower camp where we lived in rows of standard-issue Public Works Department bungalows. In each were ten or so rooms opening off the long front veranda with ‘bathrooms’ behind; these were cubicles with a canvas bath and a ‘thunderbox’. Majors and above had two rooms. In hot weather we often moved our beds on to the veranda, never forgetting the obligatory mosquito-net. The rule was to wear long sleeves and long trousers as an anti-malarial precaution after twilight, though most people preferred shorts and short sleeves during the day.
The Emperor's Codes Page 28
