The Emperor's Codes

by Michael Smith

  The telegrams handed over by the cable companies were to provide the mainstay of the codebreakers’ work. Although GC&CS came under the control of the Admiralty, the interest in naval intercepts had ended with the war. The codebreakers’ main work was deciphering diplomatic messages, Nobby Clarke said. ‘To give some idea of the change: in the early days of 1920, the strongest section of the GC&CS was the United States section.’ The other most important priorities were the diplomatic ciphers of France, the Soviet Union and Japan, a former wartime ally but now increasingly regarded as a major threat to Britain's colonies in the Far East.

  Fortunately, Maine was not content with obtaining the cables passing through London; he also made arrangements with Cable and Wireless to receive those available at other points on the international network, in particular Malta, a move that was to prove crucial to the interception of Japanese diplomatic telegrams. The Cable and Wireless repeater station on the island handled all the traffic between Europe and the Far East, including the telegrams from Tokyo to the Japanese embassies in London, Paris, Rome and Berlin. On Maine's instructions, Cable and Wireless had copies of all the telegrams sent to London for ‘accountancy purposes’.

  The GC&CS Japanese expert was Ernest Hobart-Hampden, a former senior official at the British Embassy in Tokyo and co-editor of the leading English–Japanese dictionary. So easy were the Japanese codes that ‘the cryptographic task was for the first ten years almost non-existent’, Denniston said. ‘For the language, which was the main difficulty, we were lucky enough to have recruited Hobart-Hampden, just retired from thirty years’ service in the East. For a long time, he was virtually alone. But with his knowledge of the habits of the Japanese, he soon acquired an uncanny skill in never missing the important.’

  The first big test of Hobart-Hampden's uncanny skill came with the 1921 Washington Conference of the nine major powers. The Japanese had emerged from the First World War as the third largest naval power behind Britain and America and were determined to expand their influence in the Far East, particularly in China. Britain and America were determined to curb these ambitions and planned to use the conference to limit the ratio of their own naval ships to ten for every six Japanese vessels.

  Both GC&CS and its US equivalent, the Black Chamber, were able to monitor the Japanese attitude to the Washington Conference, giving them a valuable advantage in the negotiations. During the first few weeks of the conference, the Japanese stuck out for a much tougher deal than the 10:6 ratio wanted by the British and Americans. But on 28 November 1921, with the talks reaching their conclusion, the codebreakers intercepted a message from Tokyo to the Japanese delegation. ‘We think it necessary to avoid any conflict with England or the United States,’ it said. ‘If it becomes absolutely necessary, make it clear that this is our intention in agreeing a ratio of 10:6.’ From that point on, the US and British delegations could hold out for the required ratio safe in the knowledge that the Japanese would eventually agree to it.

  Herbert Yardley, the head of the US Black Chamber, later published his memoirs, revealing the success of the American code-breakers covering the conference. But the full scale of the information provided by GC&CS to the British negotiators, which no doubt included intelligence on the position of the Americans as well as that of the Japanese, is still shrouded in secrecy. ‘No-one will ever tell how much accurate and reliable information was made available to our Foreign Office and service departments during those critical years,’ said Denniston. However, he made it clear that Hobart-Hampden's contribution was critical. ‘Throughout the period down to 1931, no big conference was held in Washington, London or Geneva in which he did not contribute all the views of the Japanese Government and of their too verbose representatives.’

  Although virtually all the messages decrypted by GC&CS were diplomatic, the Foreign Office still appeared happy to allow the Admiralty to continue to control, and perhaps more importantly to fund, the cost of the codebreaking operations, Clarke recalled.

  Things went on in a fairly peaceful way until in March 1922 a curious thing happened. Lord Curzon, at that time Foreign Secretary, had an interview with the French Ambassador in London at which he expressed certain views which did not coincide with the views of his colleagues in the Cabinet; at any rate he impressed on the Ambassador the desirability of keeping them secret.

  The Ambassador duly reported the interview to the French Foreign Minister, as was his duty. His dispatch went by wireless, was intercepted by our stations and decoded by us. The decode was duly circulated to the usual recipients, the War Office and others. Curzon was furious. His reaction was immediate and he at once persuaded the First Lord that he might just as well hand GC&CS to the control of the Foreign Office, the very department which had refused to have anything to do with it three years before. The change made little difference, except that our new heads had only a vague idea of our possibilities.

  A year later, Admiral Sinclair was made ‘Chief’ of Britain's Secret Service, which was also controlled by the Foreign Office, and was again placed in charge of GC&CS. Sinclair was a man of considerable means who was fond of fast cars and even faster living. But the Treasury's demands for a post-war peace dividend had forced him to run MI6 on a shoestring, recruiting former army officers because they already had pensions and paying them Christmas bonuses out of his own pocket. ‘I liked him very much,’ said Clarke. ‘He was an extremely able and shrewd man, and during the next seventeen years I got to like him more and more. He trusted his subordinates and would take one's word as to the advisability of a certain course and give one the utmost help.’

  With Sinclair's support, Clarke spent a great deal of time trying to persuade the Royal Navy that more attention should be given to intercepting foreign naval messages. ‘There was no naval cryptography done and the Admiralty wireless stations were all engaged in diplomatic traffic. I tried to get the Fleet interested in interception as I thought it would be useful one day. A large quantity of useful material could be got if sufficient interest in the work could be aroused.’ Clarke gave lectures on the exploits of Room 40 and eventually managed to build up interest in codebreaking with a number of enthusiastic officers working in their spare time to produce a steady stream of French and Russian naval messages.

  ‘A greater problem presented itself in respect of Japanese, which it was recognized quite early on was of great importance,’ said Clarke. ‘The first difficulty was that the Japanese Morse code consists of about twice as many signs as the English.’

  The advent of the telegraph had brought problems for the Japanese, whose written language was based on pictorial characters or ideographs, called kanji, and around seventy phonetic symbols called kana. The sound of words depicted by kanji can be represented using kana. However, this can lead to ambiguity since the Japanese language has a large number of different words which, while having distinctive written forms, sound the same (much like principal and principle in English). So a system of transliteration known as romaji developed which allowed the kana syllables to be spelled out in Roman letters. The Japanese created their own Morse code which contained all the kana syllables plus the romaji letters and was totally different from the standard international system that the Royal Naval wireless intercept operators were used to taking, Clarke recalled.

  Ordinary wireless operators were faced with a number of Morse signs which were new to them and early intercepts from HM ships in Eastern waters (which of course took a long time to reach London) were quite unintelligible. Operators had to be trained to take this down in some form or other and then the messages so received had to be transformed so as to represent the actual Japanese messages. The initial stages were very difficult and many an obstacle had to be surmounted, but directly it had been proved that progress was possible the authorities at last recognized the importance of our work.

  Clarke was appointed to head a small naval section within GC&CS. The Admiralty sent the best officer from its most recent three-year Japanese-language co
urse, Paymaster Lieutenant-Commander Harry Shaw, to assist Hobart-Hampden with the telegrams sent by Japanese naval attachés, but there was a desperate shortage of Japanese interpreters in the navy and it was difficult to find suitably qualified officers to work on other naval messages.

  The Royal Navy set about increasing the incentive for officers to learn Japanese, but with the course requiring a minimum of two years to be spent in Tokyo, this was bound to be a slow task. ‘Owing to the extreme difficulty of the language, we have very few officers who could be relied on to read an intercepted “en clair” Japanese message,’ the Naval Intelligence Department complained. ‘This state of affairs might be disastrous in the event of a war in the Far East.’

  Clarke discussed the problem with Harold Parlett, co-author with Hobart-Hampden of the leading English– Japanese dictionary and the latter's replacement as Japanese Counsellor at the Tokyo Embassy. Parlett, who was the head of the board which tested the Royal Navy interpreters, told him of a brilliant young Australian linguist who had received even higher marks than Shaw and whose knowledge of Japanese was ‘on a higher plane of practical utility than is usually obtainable in two years’.

  Paymaster Lieutenant Eric Nave had joined the Royal Australian Navy in 1917 at the age of 18 and hired a private tutor to teach him Japanese after hearing that proficiency in the language would entitle him to an extra five shillings a day. He did so well in the official language test that, in February 1921, he was sent to Japan to study as an interpreter. After two years’ intensive study, Nave's language skills were tested at the British Embassy. ‘My examiner was Sir Harold Parlett, an accomplished scholar and to me a respected figure,’ Nave said. ‘He was after all the Japanese Counsellor, the senior post at the embassy for one with language knowledge. He came from the Consular Corps, a small band, all specialists in Japanese who spent their service in Japan, Korea, or Manila, with an occasional stint at the Foreign Office in London. I was quite delighted when given my examination result, a pass with over ninety per cent, the highest ever achieved by any service language officer.’

  On his return to Australia, Nave was posted to HMAS Sydney, the RAN's flagship, where he began setting up an operation to intercept Japanese messages. ‘I would suggest that all ships in the Fleet be instructed to intercept as many messages in Japanese as possible, these messages to be sent to me for decoding,’ he said in a memo to his superiors in October 1924. ‘Telegraphic messages in Japanese are more difficult than in English, as the Japanese ideographs are not easily abbreviated. After reading plain-language messages, I propose attempting to decode ordinary Japanese economic code messages, with a view to later breaking down the ciphers.’

  It is not clear if Clarke was aware that Nave was already working on intercepted Japanese messages or was simply going on the recommendation of Parlett and Shaw, but he seems to have decided that the Australian was just the man he needed to help him break the Japanese naval codes. Denniston saw the fact that Nave was Australian as an insurmountable difficulty, Clarke recalled: ‘So I short-circuited him and went to Admiral Sinclair, who listened patiently and went off straight to the Admiralty whence I was rung up two hours later to ask what was the exact type of officer I wanted.’

  The RAN received a message from the Admiralty asking them for the loan of Nave's services.

  My appointment was to be in HMS Hawkins, the flagship of the Commander-in-Chief, China Squadron, to act as a Japanese interpreter. The Naval Board had asked for further information and whether it was for duty on Admiral's staff. In a few weeks, we were told: ‘Admiralty pressing for appointment of Nave for interpreter's duties on Commander-in-Chief's Staff. Naval Board consider Admiralty wishes should be met and propose embarkation in Mishima maru 26 June.’ The die was cast, but I could have no idea of the effect of this appointment on my career and indeed my whole life.

  Nave arrived in Shanghai in July 1925. ‘The admiral on whose staff I was to serve was Sir Edwin Sinclair Alexander-Sinclair, a dour character from the north of Scotland. He had received advice from Admiralty, London, that “Lieutenant Nave is not to be employed in the Admiral's Office nor on ship duties, details will be forwarded by safe hand bag.” This was intriguing and when they arrived the Admiralty instructions were noteworthy for their complete lack of information and directions.’ The orders did in fact lay out precisely what the Admiralty wanted, albeit with a firm instruction that there were to be no additional funds to pay for it.

  Their Lordships have had under review the question of interception of foreign wireless telegraphy messages – a subject to which they attach the greatest importance. The extent to which this method of obtaining intelligence can be utilized in war largely depends on a plentiful supply of naval cipher messages in peacetime. Owing to the necessity of economy, it is not possible to provide any special organization for this purpose on the China Station. The material must, therefore, be provided by HM ships under your command. To enable this to be done, ships should be specially detailed for the duty when opportunities arise. Up to the present, only a small number of naval cipher messages have been received and a great many more are required.

  Their Lordships are of the opinion that the difficulty of passing intercepts from such places as Shanghai to Singapore rules out the latter as a deciphering centre. Nor does Hong Kong appear to be suitable for this purpose, for owing to its distance from Japan, it would have to be supplied from other areas and its liability to attack is an additional objection. The use of a ship as a combined intercepting and deciphering centre appears to offer the best solution.

  The orders included a stern security warning: ‘Their Lordships attach great importance to secrecy in this matter, for although the Japanese may be doing the same thing, they may not credit us with an equal degree of intelligence.’ In order to restrict the number of people who knew about the interception of foreign messages it would henceforth be known as ‘Procedure Y’ and all reports or correspondence on the subject were to be addressed direct to the Director of Naval Intelligence.

  There was, however, no indication in the Admiralty orders as to how the messages were to be intercepted or the codes broken. Nave said, ‘I had to organize this interception myself. Success or failure depended on me alone.’

  He was allocated a wireless-trained petty officer, Gordon Flintham, to help organize the interception operations and together they began to unravel the mysteries of the Japanese Morse system. ‘When we started we could not even read Japanese Morse,’ said Nave. Eventually they intercepted a practice message in which the operator had run through the entire Japanese Morse code symbol by symbol. ‘This was our start. Instructions could now be issued to all ships on the China Station to intercept Japanese naval traffic and forward the messages to the flagship.’

  Nave used the intercepted Japanese naval messages to build a picture of the make-up of the Japanese fleets and began attempting to break the tasogare, the basic naval reporting code used by the Japanese to announce the sailings of individual ships.

  We started with watch on Tokyo radio and found it changed from the Commercial Call Sign ‘JJC’ to the Naval Service Call Sign ‘AB’ at hourly intervals. [Station AB was the main Tokyo naval control which was communicating with all the major Japanese naval bases.] This immediately led to identifying the main naval bases of Yokosuka, Kure and Sasebo and to the lesser bases under their umbrella. Yokosuka repeated signals at times for Chichijima, Kure for Maizuru, Sasebo for Bako, etc. The addressees of messages were spelled out at the beginning. This could be called a self-evident code with ‘da’ for daijin – the Minister of Marine; ‘shichi’ for shireichookan, or Commander-in-Chief. Ships themselves were abbreviated and could readily be identified. This step led me to identifying many call signs of commands, battleships and other vessels. Moreover, it had further value in showing a great deal of the naval organization and the extent of authority of the main naval bases. I had excellent co-operation from wireless operators, particularly those in gunboats in isolated ports. The
volume of traffic kept me working at my desk six to seven hours each day, often Sundays as well, extracting and collating a vast amount of information. I certainly had no shortage of material for my regular reports to London.

  All the information Nave managed to produce, together with any messages he was unable to break, were to be sent back by bag to the Admiralty in London which passed them straight on to the code-breakers. ‘From then onwards,’ Denniston wrote, ‘there was a flow of traffic by bag to London where the various codes were segregated and broken as far as possible and a return flow of officers with skeleton books to carry on the work locally.’

  When the messages began arriving back in the naval section, Harry Shaw was moved across to take charge of the London end of breaking the Japanese naval reporting code. ‘A system of rotation arose,’ Denniston recalled. ‘Officers still on the active list came to us for two years and then joined the China Squadron in a ship where there were facilities for local interception. Thus a first start was made on Japanese naval traffic.’

  The British codebreakers were assisted by the Japanese belief that their language was impregnable to others. Communications security was lax and many messages were sent in clear rather than encoded. When the receiving station had trouble deciphering a message, the other operator often provided assistance. Even where the Japanese stuck strictly to the codes, their insistence on using flowery preambles was a great help to the codebreakers. Messages to a superior tended to begin with the predictable phrase: ‘I have the honour to report to your excellency that’, providing Nave with an easy ‘crib’ of what was likely to be in the first part of the message and a way into the code itself.