Up at Commilla Slim was getting all he needed to know from London, Washington and Australia. ULTRA certainly provided the information on which the Battle of Imphal and Kohima was cunningly fought by General Slim and Admiral Mountbatten, the battle which was the turning point of the campaign.
Perhaps the cryptographers’ greatest naval triumph in the Pacific was the way in which Admiral Nimitz used it so skilfully at the Battle of Midway, the last great sea battle, where the big ships never saw each other and where the battle was entirely fought by aircraft.i.
While the expanding might of Hitler’s Germany had been causing growing unease in Europe, a parallel process had been taking place in the Far East. The Japanese Empire had, in less than a century, developed from a feudal backwater, which foreigners were not even allowed to visit, into an industrial and military giant that dealt on equal terms with the leading nations of the world.
In World War I, Japan had sided with the Allies and its armed forces had taken part with British troops in capturing the German colony of Tsingtao on the China coast. Japan was affectionately dubbed ‘the Britain of the East’, but it had not joined World War I out of idealism. Its government and emperor were heavily influenced by militarists who had a single-minded desire to expand the country’s possessions on the Asian mainland.
In 1921, the Japanese signed a naval treaty with America and Britain, consenting to restrictions that would give them a smaller navy than either. The story of how this agreement was reached has become one of the epic tales of codebreaking. Colonel Herbert Yardley had headed America’s cryptographic organization during World War I and afterwards he retained his interest in this field by working on diplomatic codes. He was particularly concerned with those relating to countries with which the US had any sort of rivalry. A history of cryptanalysis described what happened:
Japanese is a unique language. While it may be expressed in ideograms or brush-stroke characters, like Chinese, it may also be written in phonetic symbols like our alphabet, whereby each symbol corresponds to a syllable, not a sound. Thus a relatively short code book can be used to represent everything in the Japanese language except proper names.
Yardley and his associates attacked the Japanese code with spectacular success, the more remarkable since Yardley could not speak the language. On November 28, 1921, while the Washington Naval Conference was in progress, a communication in cipher from the Japanese Government to Prince Tokugawa, its representative at the negotiations, was intercepted and turned over to Yardley. It turned out to be one of the most important secret communications in world history, and Yardley succeeded in unscrambling it.
The decoded cable revealed that if pressed hard enough the Japanese would agree reluctantly to build only three battleships to every five constructed by the United States and England. This information was all the stubborn American and British negotiators needed to put Japan in a subordinate position.
Seven years later, in the spirit of world-wide disarmament and treaties by which countries promised not to make war, the American Government decided to scrap its cryptographic department. Colonel Yardley found himself out of a job. He turned his talents to writing a book, The American Black Chamber, in which he told the whole story of America’s success with cryptanalysis, including the background of the Washington Naval Conference. This book possibly badly damaged the United States and indirectly could have cost many American lives. Publication so outraged the Japanese that the Government in Tokyo fell, and a wave of anti-American feeling swept the country. The Japanese militarists declared that they wanted no further part in treaties limiting the size of their navy, and they began to lay the keels of the ships that fought against the United States in the Second World War.
In 1931, Japanese troops invaded Manchuria, thus beginning a conflict – the opening salvo of World War II – that would not end until 1945. Korea became a Japanese territory, and was administered with terrible cruelty. Throughout the 1930s, Japan was apparently occupied with consolidating its hold; in reality, its military leaders were preparing for a further series of conquests. Japan was poor in raw materials, particularly oil, so set in train the action needed to seize the resources of neighbouring lands.
This process was prepared by subterfuge and followed by military force. All over East Asia – in Hong Kong, Singapore, Malaya, Java – Japanese began to proliferate as shopkeepers and businessmen. Many owned photographic studios, which gave them a perfect justification for taking pictures of strategic features such as harbours and bridges. They formed a highly developed spy network that prepared the ground for the armies that would conquer these lands within a few years, and they communicated with Tokyo through code that made use of innocuous commercial terminology.
Japan also wished to force a showdown with the United States which, with its large navy and strategic island outposts, dominated the Pacific. During the interwar years the Japanese, like the Germans and Italians, began to use Enigma. The codes were being read by US Intelligence some time before the two countries were at war, though it took many years to break their cipher. The Japanese government had decided to use Enigma as early as 1930 and had acquired an early model of the machine. This was largely used for diplomatic purposes, but was later adopted (in a more developed version) by the armed forces. American Intelligence broke the Japanese cipher in 1940 and began to share its knowledge with the British. The number of ciphers, however, was so large that success could never be universal. Regarding one type of enemy cipher, Alan Stripp commented:
There can be no simple answer to the question ‘Why were the Japanese army codes not broken?’ First, several of them were. Second, there were probably as many as twenty.
Given that US eavesdroppers were able to read Japanese diplomatic ciphers for more than a year before the attack on Pearl Harbor, it may be wondered why America was taken so completely by surprise when, on 7 December 1941, Japanese fighter bombers appeared out of the early morning sky and sank much of the Pacific Fleet. The answer proved to be a matter of tragic human circumstance: an encoded message announcing that the attack was imminent had reached US Intelligence, but everyone in authority, who could have taken the important decisions and alerted the navy, was away for the weekend. The facts did not become widely known for some time, even though there was an official enquiry, because they would have made the officials involved seem incompetent. More importantly, they would have shed too much light on the way in which information from Japan was being gathered.
This matter, however, did not go away. In the 1944 presidential campaign one of the most effective means by which the Republican candidate, Thomas E Dewey, could attack his opponent Franklin Roosevelt would have been by bringing up the issue of unpreparedness over Pearl Harbor. But Dewey received a letter from the Chief of Staff, General George C Marshall. Marked ‘TOP SECRET FOR MR DEWEY’S EYES ONLY’, the letter disclosed the extent to which America’s conduct of the war was dictated by the deciphering of enemy messages and stressed the danger that could result from making this knowledge public. Requesting that Dewey abandon the issue as a campaigning tactic (he did, and it may have cost him the presidency), Marshall’s letter was a useful summary of the US war effort in the Pacific and the specific events that were affected by knowledge of the Japanese cipher. It also emphasized the fact that the European and Pacific theatres were interrelated: just as the Americans and British shared information through coded messages, so did the Axis powers. The letter said:
The most vital evidence in the Pearl Harbor matter consists of our intercepts of the Japanese diplomatic communications. Over a period of years our cryptograph people analysed the character of the machine the Japanese are using for encoding their diplomatic messages. Based on this, a corresponding machine was built by us which deciphers their messages.
Therefore, we possessed a wealth of information regarding their moves in the Pacific which in turn was furnished the State Department, but which unfortunately made no reference whatever to intentions toward Hawaii until the last message before December 7th, which did not reach our hands until the following day, December 8th.
Now the point to the present dilemma is that we have gone ahead with this business of deciphering their codes until we possess other codes, German as well as Japanese, but our main basis of information regarding Hitler’s intentions in Europe is obtained from Baron Oshima’s messages to Berlin reporting his interviews with Hitler and other officials to the Japanese Government. These are still in the codes involved in the Pearl Harbor events.
To explain further the critical nature of this set-up which would be wiped out almost in an instant if the least suspicion were aroused regarding it, the Battle of the Coral Sea was based on deciphered messages and therefore our few ships were in the right place at the right time. Further, we were able to concentrate our limited forces to meet their advances on Midway when otherwise we almost certainly would have been some 3,000 miles out of place.
We had full information of the strength of their forces in that advance and also of the smaller force directed against the Aleutians which finally landed troops on Attu and Kiska.
Operations in the Pacific are largely guided by the information we obtain of Japanese deployments. We know their strength in various garrisons, the rations and other stores available to them, and, what is of vast importance, we check their fleet movements and the movements of their convoys.
The heavy losses reported from time to time which they sustain by reason of our submarine action largely results from the fact that we know the sailing dates of their convoys and can notify our submarines to lie in wait at the proper point.
The current raids by Admiral Halsey’s carrier forces on Japanese shipping in Manila Bay and elsewhere were largely based on timing of the known movements of Japanese convoys, two of which were caught, as anticipated, in his destructive attacks.
The Roberts Report on Pearl Harbor had to have withdrawn from it all reference to this highly secret matter, therefore in portions it necessarily appeared incomplete. The same reason which dictated that course is even more important today because our sources have been greatly elaborated.
As a further example of the delicacy of the situation, some of the OSS, without telling us, instituted a secret search of the Japanese Embassy offices in Portugal. As a result the entire military attaché code all over the world was changed, and though this occurred over a year ago, we have not yet been able to break the new code and have thus lost this invaluable source of information, particularly regarding the European situation.
The conduct of General Eisenhower’s campaign and of all operations in the Pacific are closely related in conception and timing to the information we secretly obtain through these interrupted codes. They contribute greatly to the victory and tremendously to the saving of American lives, both in the conduct of current operations and in looking toward the early termination of the war.
This overview of American successes was, of course, written at a later stage in the war. In the meantime, the Japanese themselves were having plenty of victories. Having chosen the moment to begin the war on their own terms in December 1941, Japanese armies swept like a prairie fire through the surrounding territories – through China to Hong Kong, into Burma, Thailand and Malaya and down the archipelago to fall upon Singapore, then across the sea to the Philippines, the Dutch East Indies and New Guinea. They were as successful at sea as on land, sinking HMS Prince of Wales and Repulse off Malaya and the carrier Hermes off Ceylon (now Sri Lanka), and obliterating a combined Allied fleet at the Battle of the Java Sea.
Their victories were largely easy. The forces that opposed them were too small and ill-equipped to offer serious or protracted resistance, and local populations, initially at least, often welcomed the ousting of their colonial overlords by fellow Asians. Only in Burma and New Guinea was their expansion to be checked by Allied and local troops.
In the winter of 1941–2 the Japanese seemed unstoppable. Within months they had conquered a huge swathe of territory, wiped out the opposing armies, displaced and interned the colonial rulers, and set up their characteristically cruel form of occupation. Though they called their empire the ‘Greater East Asia Co-prosperity Sphere’, there was no attempt at sharing prosperity, or anything else. They ruthlessly exploited the natives and resources of conquered lands. Their sights were set on India and Australia.
At the insistence of Churchill, the Allies had given priority to the defeat of Hitler. The Pacific conflict was lost, for the time being. The Allies sought to keep Japan out of India, Burma and Ceylon (now Sri Lanka) but could not yet go on the offensive.
Both in Britain and in the United States, training of personnel for the Pacific war was being carried out as a matter of urgency. Cryptographers were needed and one of the training schools for them was opened at Vint Hill Farms in Virginia. This establishment, like Bletchley, was so secret that its inmates were forbidden to talk about it. It was also disguised to look as if it were a real farm. Irene Brion, a young woman who enlisted in the US Army as a cryptographer for assignment to the Pacific theatre, was swiftly processed through it in 1942. She outlined the structure of the course:
Classes began the following Monday on the principle of self-instruction via a training manual, with each person working at her own speed. We were graded, and anything lower than eighty-five had to be rewritten. The classes lasted for eight hours, except for lunch hour and time off for retreat, and were given six days a week.
We began with cryptography, which involved learning to encipher and decipher messages in code and cipher. Mastering a number of systems was only the beginning; we then had to learn to analyze them, because we were being trained as cryptanalysts. A cryptanalyst had the military occupational specialty (MOS) number 808, which was defined in a current army manual as follows:
‘Decodes and deciphers enemy messages and cryptograms without aid of the device or key used in preparing them. Using deductive reasoning and employing knowledge of the various cryptographic codes, analyses messages and determines key to code. May supervise others in cryptanalysis. Must have cryptographic clearance. Must have training in cryptanalysis and be familiar with all types of cryptographic systems and their variations in military communications. Must possess initiative, patience, and marked deductive ability. Should have some mathematical training and be familiar with at least one foreign language.’
We practiced on messages used by the Union army at Gettysburg, because a similar encoding procedure was used by the Japanese in one of their shipping codes. We worked with only one machine, the M-209, a medium-level cryptographic system.
In addition, we had a daily class in Hepburn Kana, the system devised for converting the Japanese characters into syllables – a necessity for transmitting the language in Morse code. We learned a lot of Japanese military vocabulary and the grammatical structures we’d be likely to encounter in their messages. We ‘assumed’ (a word used frequently by cryptanalysts, and jokingly among ourselves) that we’d be going to the Pacific, but we dared tell no one.
Finally, the last classes were over. We said our goodbyes to Sergeant Nelson, our ‘keeper’, and to Tokashi Kajihara, the nisei (Japanese American) who taught us Japanese. The next morning, September 16th, we boarded trucks and left Vint Hill behind us. From there we’d be going to Fort Oglethorpe for overseas training.
Academics were needed as well as clerical staff, and in Britain, too, recruiting became urgent. Once again universities were trawled, but while there were many undergraduates studying European languages, there were virtually no Japanese speakers. The search therefore targeted those with an aptitude for the Classical tongues, on the basis that they had already demonstrated skill in mastering difficult languages. Alan Stripp was at Cambridge in the middle of the war and had already been accepted by the Air Force. One day, however, his destiny took a sudden turn:
In October 1942 I went up to Christ’s College, Cambridge, for a year because I was too young to be called up for the RAF. One day in the spring of 1943 I went to my tutor to have my Greek composition marked. He sat in his comfortable chair like a well-fed pussy cat, and said, ‘Oh, by the way, before we begin, would you like to learn Japanese? They are very short of people who can translate it.’ I nearly fell through the chair, but managed to say, ‘What would I have to do?’ He replied, ‘You’ll have to have an interview. If you’re accepted you’ll train for six months in this country and then probably go out to the Far East in the army. I thought this sounded interesting and replied, ‘I’m prepared to try anything once.’ So I had an interview with Colonel Tiltman from the War Office department MI8. He asked me if I played chess, and I said I didn’t; or did crossword puzzles, and I said I did; or could read music, and I said ‘Reasonably;’ and other questions which at the time seemed odd, but which later I understood. He said that I would hear from him, and I told him about the Air Force offer, but he said ‘Don’t bother about that.’ So I started my summer holiday.
Half way through, on a day when I was sunbathing in the garden, a telegram arrived telling me to report to the Inter-Services Special Intelligence School at Bedford on the last day of August. So began the hardest six months’ work of my life. There were thirty of us in the class, who had all done Greek and Latin at Oxford and Cambridge and so they thought we could tackle Japanese.
Stripp described the training given to candidates and the particular difficulties that this language involved:
On the first day we were issued with two Japanese–English dictionaries, one in English letters and one for characters. The Japanese do not normally write with letters as Europeans do, but with pictures called ‘characters’. Every character can be pronounced in two ways, one Chinese (where the characters were invented) and one Japanese. A Japanese book starts at the back because they write from right to left and from top to bottom. If you are going to send a message by radio you cannot use characters and so you have to use what the Japanese call ‘kana’; these are fifty syllables, each of which has a Morse equivalent, and they can be written down in English letters. [On the subject of Japanese code books:] If they wanted to send a message containing ‘air special wireless unit’ (kookuu tokushu musentai), all they sent was ‘0700.’ The person at the other end looked up the number in his book. We, of course, did not have the book (at least not normally) and so we had to break the code without it.
The Bedford School was a remarkable, and very English, piece of improvisation. It was the invention of Colonel John Tiltman, an expert code-breaker who had been working at Bletchley Park since 1939. After Pearl Harbor he was asked to recruit men who knew written Japanese and found about a dozen, mostly at the School of Oriental and African Studies (part of the University of London). The School said it might be possible to train people in two years but not less. That was no good for Tiltman who was in a hurry, and he turned to Captain Oswald Tuck, a retired naval officer who had studied Japanese on his own initiative and was eventually appointed Assistant to the Naval Attaché in Tokyo. On 21st December 1941 Tiltman asked Tuck to conduct the first six-month course for twenty-two men and one woman, most of whom were classical scholars from Oxford and Cambridge. Tuck, who had never taught anyone Japanese, commented in his diary ‘the idea sounded impossible but was worth trying’. He tried, provided all the text books himself, five in number, and was brilliantly successful – at the age of sixty-five.
We finished on the 9th February 1944 when we were told that twelve of us were going to be sent to Australia as quickly as possible because the need was urgent. Apparently Britain and America between them could find fewer than forty people who were competent in the language. My name was among the twelve.
Radio operators who were destined for the Pacific often began their training in Britain but continued it en route, bringing their skill to a pitch of perfection during a lengthy stopover in the Middle East. RC Grindlay attended one of the signals schools in the region. His was in the Biblical setting of Mount Carmel in Palestine:
The training now started in earnest and mainly consisted of sending and receiving Morse code. Most of us were at the standard of 8/10 words per minute. It was pointed out to us that this was far too slow and that something in excess of 20 w.p.m. would be required for operational work.
During the next six weeks we had Morse code send and receive lessons for at least six hours a day and most of us got up to the high twenties. A lot of emphasis was placed on reading through interference (i.e. static and other stations); this certainly proved somewhat difficult but practice makes perfect – or nearly.
There were establishments all over the East, as well as in England and the US, for dealing with Japanese codes. Stripp listed them:
Decoding of Japanese traffic was carried out at Bletchley and at Arlington Hall in Virginia, where the ‘really intractable problems’ were dealt with. Then there were large US & British units at Delhi, as well as other, smaller units elsewhere in India. There was a Sigint base at Anderson in Ceylon, one in Mauritius and one in Mombasa. They worked with US bases in Guam, Leyte etc. Chain of intercept stations that ran ‘in a great arc from Melbourne, Canberra, Brisbane and Darwin through Ceylon to Calcutta and South China’.
He himself was posted to Brisbane, a city that would become familiar to a great many more codebreakers before the war ended. He discussed the work in which he was engaged, and outlined a number of important differences between the Japanese and German approaches to cryptography:
My section was ‘Naval Air’ (the Japanese did not have a separate air force, but their army and navy each controlled its own). Our task was to decode and translate messages picked up by wireless units which listened to Japanese aeroplanes, flying anywhere from Tokyo to Singapore, and to their bases, and sent messages to us in Brisbane. These told us a great deal about what the Japanese were doing and were intending to do, and [we] passed vital information immediately to the staff of General MacArthur who controlled all the American and Australian forces in the South West Pacific Area.
The Japanese approach to codes was quite unlike that of the Germans. Most of the important German messages for their three Services used the Enigma machine, with variations. The Germans were quite sure that it was unbreakable and unbroken, and remained so throughout the war. The Japanese, on the other hand, used a great variety of codes and ciphers. At any one time during the war at least fifty-five different systems were being used for army, navy and diplomatic messages, of which twenty-four naval and twenty-one army have been identified. This is all the more remarkable considering their love of tidy organisation.
There is a technical difference between a code and a cipher, though this is often blurred in many books: a code is a method of using groups of two or more letters or numbers to stand for words or phrases of a message; a cipher … is a method of using letters or numbers to stand for the letters of a message, one at a time, or to change their order. All the work I did was with a code book which used a four-kana group to stand for a word or number, and a two-kana group to stand for a phrase, like ‘estimated time of arrival’.
In all Japanese military, naval, air and diplomatic traffic, codes heavily outnumbered ciphers, a fact which is often obscured, partly because some writers do not distinguish clearly or correctly between codes and ciphers, and partly because the process of further concealing a code text is called re-ciphering. Japanese codes were generally based on numerals rather than on letters – though they used both – and were mostly reciphered for extra security. Letter codes were reserved for lower grade signals.
Captain Nave received messages from several wireless units. This was useful because if one version of a message contained gaps or faults another version might put these right. The greater the quantity of material available the better was the chance of breaking the code.
There were two great helps in this process. The first was being familiar with the shape of the message. Many messages were routine and followed the same pattern; for example, weather reports gave the place and time of origin, the general weather, the temperature, the amount of cloud, the wind direction and speed and perhaps the further outlook. In addition, the uncoded messages or ‘traffic’ from or to the aircraft might very well give a clue to one or several groups in a coded message.
The second were ‘cribs’, something which ought never to happen with well-trained wireless operators. A plane is flying, say, to Balikpapan, and sends its estimated time of arrival. The base replies, ‘Sorry we do not have your code book 24; please send in 23.’ The operator repeats the identical message in code 23, which the interceptors have already largely broken. So it is easy to read off all the groups in the new code 24. Code books were changed regularly and it was not always easy to supply copies of the new books on time to the more distant outposts. In January 1943 two New Zealand corvettes rammed a Japanese submarine off Guadalcanal and forced it to beach on an outlying reef. It was found to be carrying 200,000 ‘code books’ (a term which may include other crypto materials), including, apparently, JN25, one of the most important naval codes. The Japanese changed some of it but left JN25 unchanged.
Stripp found that he and his colleagues, especially with enemy code books at their disposal, could think themselves into the mindset of their opponents and that fascination with his work could take over his off-duty hours:
Each word in our messages was made up of four kana syllables, like KO FU TE SI which might stand for ‘enemy,’ together with some two syllable groups, like HI TO which might stand for ‘expected time of arrival is …’. Every evening I sent to Captain Nave a list of things I thought I had broken together with their meanings, and every morning I received back from him a list confirming them if they were right, correcting them if they were wrong, and giving me additional ones.
Just to make matters worse the Japanese changed this particular code book about every twelve days, but our wireless operators were very canny. The most experienced of them knew the touch of the Japanese operators so well that they could tell which plane was transmitting even if it had just changed its call sign.
After I had been there three months I was sent an actual code book which had been captured (I suppose from a crashed plane), and I spent a lot of my ‘spare’ time translating it into English. It gave one more of the feel of a Japanese message when one was familiar with the book from which it came.
As for all cryptanalysts, nothing made more interesting reading than captured enemy code books. Stripp was particularly delighted by:
… a wonderful find at Sio, west of Sattelberg, of some steel boxes in a water-filled pit. These contained many high-grade code books, which meant for the next months instant translation of all main-line messages with top-level information.
And went on to illustrate:
Here is a typical actual message:
SE NO SU sendan convoy
ME U TE no of
MU KU NO ichi position
TU RO HI R Rabaul
YA U TI kichi base
TO MO TI 32 32
TU RI NE 6 6
MO SI HE 50 50
YA SU KE shinro course
MU KA NU 355 355
TI HO SO sokuryoku speed
YA HO NA 8 8
Translation: Position of convoy: bearing 326 degrees, 50 miles, from Rabaul base; course 355 degrees, speed 8 knots.
Despite the difficulties of dealing with a language, a culture and a mentality that were entirely different to those of Europeans, Americans or Australians, the codebreakers had quickly managed to get the measure of their enemy.
In Australia, as in Britain and America, the field of cryptography was one in which women could make a valuable contribution. Those who were engaged for the work went through a process of selection and training similar to that experienced by their counterparts elsewhere. Patricia Penrose was a native of Melbourne, and was recruited through a family friend to work at a military camp within the city. She found herself engaged in highly secret work only a short bus-ride from home. She remembered the formalities of selection and induction. The women were advised by an officer:
… to forget what we saw in signals and never to even mention any Army matters when off shift. Then individually we were taken aside and, holding a Bible, took an oath not to divulge any information contained in any signal, nor to divulge the method or workings of any codes or ciphers that we learned. The punishment for doing so, under National Security Regulations, was a long prison sentence. Each of us signed our names to this as well as swearing to it.
Training then began in earnest:
Lieut Gange then asked us individually to spell out different words and military terms which were often used in signals. Our lectures and lessons were terribly interesting. Naturally, we could not take notes. We learned that certain codes and ciphers need a key, and for security reasons it was necessary to change this key every twenty-four hours. This was the minimum time it took enemy cryptographers to ‘crack’ even our highest grade ciphers. They had specially-trained personnel, just as the British had theirs at MI5 in England. By this time the enemy picked up several hundred or more enciphered radio messages in a day, so they had plenty of material on which to work. As there are certain consonants and vowels used time and again, the figures and groups which represented these were repeated more often.
She detailed the methods used to encrypt messages:
The recurrence of these groups of figures gave the enemy cryptographers their opportunity to ‘break’ enciphered messages and piece together the key for the day. The cipher books we were using were those used by the British in the 1914–18 war, so the whole of our security rested on the changing of the key each day. As short messages are easier to crack than lengthy ones, it was bad security for an officer to originate a message of only a few words. However, should we ever have such a message to encipher, we had to use padding. This was to follow a sequence but have absolutely nothing to do with the war, and should be so far removed from the original message that it couldn’t possibly be deciphered and placed in with the original context. The War Office example was ‘Mary had a little lamb’ – message on – message off – ‘whose fleece was white as snow’. Most signals originating in clear commenced: ‘To - - ’, followed by ‘From - -’, then sometimes ‘For attention General So-and-so’. This meant that most signals leaving our headquarters gave as their origin: ‘From Landforce Melbourne’, and about 65% were addressed to ‘Landops Brisbane’. The remaining 35% were addressed to Alice Springs, Darwin and other areas, which meant a constant recurring, year in, year out, of the same addresses. So it was essential to secrete this somewhere in the main context of the message. The person deciphering it was responsible for leaving out the padding, and replacing the address and message in correct order. When enciphering a message you would pencil-mark the original or clear version, showing where you inserted the address and padding. So your actual message would read something like this: ‘Oh, I do like to be beside the seaside. Message begins Gen. Blamey and Gen. MacArthur to visit bases in South West Pacific Area (S.W.P.A.) Estimated time of departure 0900 hours 25 June repeat 0900 hours 25 June. Brackets on. To Landops Brisbane from Landforce Melbourne. Time of origin 0300 hours 24 June repeat 0300 hours 24 June. Brackets off. To check conditions morale and inspect camps. Make necessary arrangements. Carlyon message ends. How I do like to be beside the sea.’
The method of enclosing padding and address should never become stereotyped. As radio conditions were not always clear for reception and therefore a group of figures could be incorrectly received in Morse, this could leave you with one indecipherable word in a message. If the word was obviously ‘and’ or such like you could hazard a guess and mark with a query, but the missing word could be vital so, for this reason and to prevent delay in sending back to the office of origin for a repeat, figures and dates always had to be repeated, also NOT and MUST. These were high grade ciphers we were learning, the lower grade being used in the field, between commands and units, such as a Transposition Cipher whereby the first or second word in each sentence was transposed into another, or a Substitution Cipher where certain letters received in a message were substituted with quite different letters. This could be made a little harder with a double substitution, but such low grade ciphers had practically no security as far as the enemy was concerned.
Despite being separated by vast distances from scenes of fighting, the women had to follow rigorous security procedures regarding their code books:
Signals were to be enciphered and deciphered wherever possible behind locked doors. When not in use, the cipher books had to be locked in a safe. In the face of an enemy advancing or the possibility of a bombing attack, the cipher books were the cipher operator’s first responsibility. He must guard them at all times with his life, risking all to burn them if necessary.
All these regulations referred to males, as to this time there had been few females handling these books. It was also laid down that officers only were to use them. However this was now to be relaxed and cipher operators could be NCOs to the level of sergeant. Should you take out a message commencing NODECO, short for No Decode, you were to call the officer on duty to decipher the remainder. This word was later changed to TOP SECRET. The priority of dealing with messages being MOST IMMEDIATE, IMMEDIATE, IMPORTANT and URGENT in that order, these being labelled accordingly with different brightly coloured tags.
Thousands of miles away, on the west coast of the USA, yet another group was training in 1942 for the Pacific war and they, too, would become involved in the transmission of coded signals. Their job, like that of Patricia Penrose, would be to send rather than decipher them. They were American Indians of the Navajo tribe, and the notion of using their own language, impenetrable to outsiders, for sending military messages was one of the war’s great communications breakthroughs. They were to invent terms for a multitude of words that were otherwise difficult to disguise:
| Terms for ships: | ||
| Battleship | Lo-tso | Whale |
| Aircraft carrier | Tsidi-ney-ye hi | Bird carrier |
| Submarine | Besh-lo | Iron fish |
| Destroyer | Ca-lo | Shark |
| Transport | Dineh-nay-ye-hi | Man carrier |
| Cruiser | Lo-tso-yazzie | Small whale |
| General vocabulary: | ||
| Airdrome | Nilchi-began | Air house |
| Armored | Besh-ye-ha-da-di-the | Iron protected |
| Artillery | Be-al-doh-tso-lani | Many big guns |
| Commanding officerh | Hash-kay-gi-na-ta | War chief |
| Convoy | Tkal-kah-o-nel | Moving on water |
| Hospital | A-zey-tah-ba-hogan | Place of medicine |
| Report | Who-neh | Got words |
| Route | Gah-bih-tkeen | Rabbit trail |
The notion of a code based on the language of an American Indian tribe was originated by Philip Johnston, an engineer employed by the city of Los Angeles, shortly after Pearl Harbor. Johnston, the son of a missionary, had grown up on a Navajo reservation and was fluent in the language. He was aware that security was seriously compromised by the lack of an unbreakable code and by the skill of Japanese decrypters, whose knowledge not only of English but of American slang, military jargon and even profanity enabled them to read and understand messages without difficulty. Large numbers of Japanese had been educated before the war at high schools and colleges in the United States. They therefore spoke English with authentic American accents, and knew as much about the country’s culture as most natives. Johnston visited Camp Elliott, near San Diego, and asked the Area Signal Officer, Lieutenant Colonel JE Jones:
‘Colonel, what would you think of a device that would assure you of complete secrecy when you send or receive messages on the battlefield?’ Jones answered:
‘In all the history of warfare, that has never been done. No code, no cipher is completely secure from enemy interception. We change our codes frequently for this reason.’
The colonel was persuaded to test the idea, though there was some initial scepticism. There were one or two precedents. In the previous war, the 141st US Infantry had used a small group of Choctaws to send orders by field telephone. The Canadian Army had also used Native Americans for signalling, but the difficulty had been that there were no words in their vocabulary for many of the terms necessary in military communication. Johnston’s idea was that such terms could be created and Navajo trained in their use. Other Native American languages would have been equally suitable; it was simply his personal knowledge of this tribe that led him to suggest using theirs. One Navajo recruit, Keith Little, explained the advantage of using his tribal language:
It was a whole lot faster. The ordinary, conventional military code, in comparison … if you match the two together, the Navajo code always beats them, for the simple reason that conventional code has to be changed, updated, maybe every day, two days, maybe every week.
The conventional code uses scrambled numbers and letters and it is scrambled – an ordinary message is written on a pad or a paper and given to an expert to scramble, to code. So when he’s done he gives it to the radio operator and the radio operator sends it, and it is written down at the other end the way it is sent. And from there it goes to another expert and he unscrambles it. So there’s time lost.
If a Navajo code-talker is really fluent with his code system he is given a message, it’s sent, he looks at it, calls the receiver on the other side: ‘Here’s the message,’ and the receiver will say, ‘Go ahead and send it,’ or something like that, and he encodes them as he is talking and the guy over there is decoding it as he’s saying it. Maybe it comes through a minute or two minutes. Ordinary messages are never more than three minutes, and when you’re doing that you’re saving lives.
San Diego is one of the most important bases of the United States Marine Corps, and it may have been for that reason that Johnston approached this branch of the services. He was not to know that, in the war that had just begun for America, the Marines would not be deployed in the European theatre of operations at all. They would see extensive action in the Pacific, but would not face Hitler. As a result, the Navajo communications specialists recruited at his urging would be used only against the Japanese.
The scheme called for the enlistment of 200 Navajo into the Marines, though in the event only 30 were authorized as a cautious beginning. Recruiters visited schools on the Navajo reservations to find young men who showed the requisite intelligence. The tribe’s population was not extensive; it was in the region of 50,000, spread throughout Arizona, New Mexico and California, and within this community there was only a small proportion with the level of high school education necessary to send or receive complex and vital messages. In addition to mental ability, candidates had to be able to meet the physical demands of the Corps, by tradition the toughest and most highly disciplined part of the United States Armed Forces. Once a sufficient number had been recruited, they would have to undergo intensive, exhaustive training to prepare them for the fierce fighting of the Pacific theatre.
The first intake formed the 382nd platoon, USMC, and passed through ‘Boot camp’ at San Diego in 1942. Their background made them unfamiliar with any concept of ‘spit-and-polish’ military discipline, yet they adapted surprisingly well – and quickly – and were highly praised by their commanding officer for their cleanliness, order and skill in weapons-handling. They were described as ‘excellent general duty Marines’ and there was no question of any lowering of standards to add these valuable signallers to the Corps.
Once their general military training had been completed, they received further instruction as signallers, being taught semaphore and Morse. Only after this did they begin their important work: the developing of the code. After much labour they created a list of 211 frequently used military terms, as well as an alphabet for the spelling out of other words. The letters were mostly associated with the names of animals and plants; for instance, the code for D (moasi in Navajo) was the word for ‘deer’, that for Q (ca-yeilth) was the word for ‘quiver’, and Y (tsah-as-zih) the word for ‘yucca’.
Keith Little remembered that keeping track of all these terms could be a challenge:
That’s the only way, you memorize it and you write it down. And you keep that in your mind because … there’s about over four hundred words by this time. The initial group made about 211 military words and about 26 military letters …
Once the initial batch of recruits had been posted to active service in Guadalcanal, the success of the experiment began to be appreciated. Commanders in the field asked to be sent more Navajo signallers, and the Marine Corps arranged for the recruitment of another 200 men. The training programme became established, with a succession of classes passing through Camp Elliott, and Philip Johnston, though in his forties, was able to enlist in the corps as the course instructor. As the Pacific war expanded, involving ever larger numbers of men, the objective became the recruiting of 25 Navajo a month, with a view to making 82 available to each Marine division. This number could not be procured, and divisions had to make do with only seven or eight. Johnson suggested establishing a training centre in the Pacific at which both recent recruits and men already serving could attend courses and formulate additions to the code, since military terminology – such as the names for new types of tank and aircraft – was continually evolving. The centre was established in Hawaii and the code was extended, the number of words virtually doubling. Further terms were also added (the term for Britain was ‘bounded by water’ and for Germany ‘iron hat’).
The code had the advantage that, despite the lengthy and complex-sounding Navajo words, it made communication faster than any other system used by American forces. This speed was matched by the efficiency instilled in training. One graduate of the course, who took part in the fighting on Guadalcanal, later remembered some of the essentials that he and his comrades had been taught:
One thing we learned in school was not to be on the air longer than was absolutely necessary. We had to be careful not to repeat words in a sentence – that is, if the message had to go through more than once, we tried to say it differently every time. We were also told not to use the same word too often in a sentence, and that we had to be accurate the first time! If a message has to be corrected or repeated too often, you are giving the enemy a better chance to locate you. On Guadalcanal we had to move our equipment in a hurry because the Japs started to shell the very spot where we were operating.
The Navajo were expected not only to ‘soldier’ as regular Marines but to be available at all times to send or receive signals. If a radio operator was sent a message prefaced by the words ‘New Mexico’ or ‘Arizona’, it meant that the nearest available Navajo had to be brought in to listen. In addition to radio and telephone communications, they had in theory to be able to work as ‘runners’, carrying messages by hand from one part of a battlefield to another. This need to dodge bullets during combat was one major reason why the Navajo recruits had to be fit. The work could be highly dangerous, and cost the lives of several of them. In practice, they were too valuable to be put at unnecessary risk. One who took part in several of the amphibious assaults on Japanese-held islands remembered that:
Men in the first wave on the beach during an invasion barely have a chance to survive; the second wave has a little better chance, and chances are fair for the men in the third wave. On Tarawa, they put me in the third wave to give me a better chance for survival. I was in the front lines a lot, but they would keep us where the action was only a half day – not longer than a full day – then send us back, keeping us involved, but trying to protect us in that they didn’t want anything to happen to us. We couldn’t be replaced like some of the men with general skills.
Though they did not think of themselves as in any sense an elite, at least one of their instructors (himself a member of the tribe) tried to instil this feeling in his students. He sought to ensure that they would keep their knowledge to themselves even under threat of death:
I wanted those code talkers to guard their secrets with their lives. I thought of the idea of comparing them with the Japanese suicide pilots and the Nazi elite guards. If they were captured, they should guard the code with their lives! I would ask, ‘Would you refuse to give away the secret of the code if you had a samurai sword at your throat? If the enemy would ask, “What is your word for A?” would you tell them? You begin to bleed; you begin to feel your own blood trickling down, with the cutting a little deeper. You WOULD lay down your life before you would tell, wouldn’t you?’
I would look them square in the eye and ask them to answer yes or no. I would say, ‘Are you devoted that much? Are you willing to die for this secret method?’
As an instructor in the field I ramrodded that. If, when I asked a question, a man would nod his head, I would say, ‘No, I don’t see a thing in your eye.’ If it wasn’t definite enough to suit me I would say ‘I didn’t quite hear you!’ Then he’d say with emphasis ‘Yes, sir!’ I’d stay with it until I heard it loud and clear, and then I knew I could trust that man – just like the suicide pilots would give their lives when they made that last dive. Like the Nazi elite corps – no ‘if’ or ‘maybe’ about the way they did things, I wanted absolutely strict discipline – military discipline.
Fortunately this level of commitment did not prove necessary, for none of the code talkers ended up in Japanese captivity. Their presence in the front line, however, gave them occasional trouble with their own side. The physical resemblance between American Indians and Japanese was sufficiently close that, in the dark or in the confusion of battle, individual Navajo were arrested as Japanese spies. Their American uniforms and identification, as well as their knowledge of the language and culture, were merely seen as effective camouflage, and in one or two instances they were in danger of being shot until vouched for by their officers. One man remembered:
I had been on Guadalcanal for some time and was hungry for something like orange juice. The army usually had some, and a transport had just come in close to where we were waiting on the beach to leave the island. I walked over to the army supplies and started digging for orange juice when somebody put an iron in my back. I thought whoever it was was just kidding and kept on digging. He finally said, ‘Get out of there, you damn Jap!’ The sergeant standing there said, ‘He has Marine Corps identification and speaks good English.’ The man with the gun said, ‘I don’t care if he graduated from Ohio State. We’re going to shoot him.’ I finally mumbled, ‘I’m from right down there,’ but I couldn’t see the direction because there was too much sweat running down into my eyes. Finally, they took me back to my outfit. I had 15 men around me, and the sergeant of the guard had a 0.45 cocked against my back all the way and I had my hands up all the way. When we got to the beach, they asked, ‘Is this your man?’ and of course got the answer, ‘Yeah, that’s our man. Hey – are you guys serious?’ ‘You’re damn right we’re serious,’ they said, ‘If you guys don’t make a positive identification we’re going to take him back.’
Finally the lieutenant came around and said, ‘What’s the matter?’ ‘Well, we caught this man over there in our yard and we think he’s a Jap. If you don’t identify him we’re going to take him back over there and shoot him.’
After that they gave me a bodyguard. If I went to the head, he would come along behind me – this tall white man; I’d go swimming and he’d go swimming. They made him stick right by me.
A total of 420 Navajo communications specialists served in the Marine Corps between 1942 and 1945. One Marine officer, Major Howard Conner, declared: ‘Were it not for the Navajos, the Marines would never have taken Iwo Jima.’26 The Japanese, who had broken other American codes without difficulty, never succeeded in interpreting their messages. In the view of a signals officer, Colonel Norman Goertz, the enemy’s difficulty in understanding the language, compounded by a lack of technical equipment and time for analysis, contributed to the Allied victory. He wrote:
My guess is that while the Japanese may have monitored their transmissions at one time or another, because the Navajos were used during periods of tactical engagement, the Japanese did not have either the time or facilities to identify the transmissions properly. Having had personal observation of the Japanese electronic equipment through the Iwo Jima campaign, I can speculate that, in the field, they did not have enough equipment to monitor our channels and record the transmissions for analysis.
The Navajo were so successful that it is only to be wondered why the other branches of the American government and armed forces did not make use of similar schemes, or why the Marines’ experiment (which ended with the cessation of hostilities) was not continued into the postwar era. Their contribution had been vital and Keith Little reflected with satisfaction on the difference he and his colleagues had made:
I have a real feeling that I was able to be a marine, and not knowing what I was going to do. All I was going to do was maybe carry a rifle or a machine-gun and shoot, but I ended up not using my weapon. My weapon was my language, and that language probably saved countless lives.
Japan’s territorial expansion had reached its limit in the spring of 1942, when it established a defensive perimeter that ran from Burma to northern New Guinea. This very success, however, carried the seeds of disaster, for an empire of this size meant stretching manpower and resources and could not be easily defended. At the same time, the United States had begun to hit back. The ‘Doolittle Raid’ on Tokyo on 18 April 1942, in which bombers had for the first time taken off from an aircraft carrier, had made it possible, also for the first time, to take the war into the enemy’s camp.
The next month Japan received two further blows and from these it would not recover, for they would end its dominance of the region and begin the steady decline in its fortunes.
In April, Admiral Chester Nimitz, the American Commander-in-Chief Pacific, received intelligence from intercepted radio traffic that the Japanese fleet was on its way to capture Port Moresby in New Guinea – clearly a stepping-stone for an attack on Australia. Knowing exactly where to find the enemy ships, and knowing also the number of vessels and aircraft he could expect to encounter, Nimitz ambushed the Japanese in what became known as the Battle of the Coral Sea. This was the first naval engagement in history to be fought between two opponents whose ships were not even within sight of each other, for the actual combat took place between fighter aircraft launched from their respective flight decks. The advantage of surprise did not bring outright victory, for one American carrier was lost and another disabled. Nevertheless, the enemy was prevented from reaching its objective.
If the Coral Sea was the ‘left hook’ delivered by the US Navy, the Battle of Midway a few weeks later in June 1942 was the ‘right hook’. This time the defeat of the Japanese was decisive. They lost four aircraft carriers, and with them any ability to dominate the Pacific. Once again success proved costly for the Americans, but the price was worth paying. After Midway, Japan was unable to threaten Hawaii, or the United States, again.
This battle was also a result of overheard Japanese radio traffic. Hugh Melinsky, who was a British cryptanalyst in the Pacific, explained that:
Intercepted signals showed that a great Japanese fleet would move and try to capture Midway Island, known as the sentry for Hawaii itself. The instructions in the signal were ‘to provoke action with the main American fleet and to destroy it piecemeal’, and finally it disclosed that an attack on the Aleutian Islands would be made which was intended to draw off the Americans to the north. And leave the way open to Midway. As a result, Nimitz was able to avoid the trap of allowing his ships to be drawn off to the north. The Battle of Midway was to be the turning point in the Pacific.
The Midway Islands are north-west of Hawaii and are part of the same island chain. In 1942 they were America’s westernmost outpost in the Pacific. That they might be a Japanese objective had been the conclusion of the US Navy’s expert codebreaker, Lieutenant Commander Joseph Rochefort. An eccentric in the best traditions of Bletchley Park, Rochefort shared with the British ‘boffins’ an instinct for getting inside the mind of the enemy. He had been instrumental in breaking Japan’s JN25 code, and it was he who had disclosed the Imperial Navy’s plan to attack Port Moresby.
Rochefort was aware that Japanese code referred to places by groups of two letters, of which the first was A. AH was Pearl Harbor; AG was the site of a refuelling rendezvous for seaplanes. He decided that AF, a location that was featuring in messages with increasing frequency, was Midway. To put this notion to the test, it was arranged that a false message would be broadcast in ‘clear’ to say that the water-distillation plant at the naval base on Midway was out of action. Within 48 hours a Japanese signal was intercepted which announced that fresh water supplies at AF had run out.
The purpose of targeting this lonely American outpost would be to draw their Pacific Fleet into battle. Despite the pounding it had taken at Pearl Harbor, the US was still a formidable presence in the region, for by chance its aircraft carriers had not been in port on the morning of the attack. Admiral Yamamoto, the Japanese naval commander, sought to divide his enemy by staging a diversionary attack far to the north, and then destroying them in two actions. He had significant superiority in numbers, and had his plan succeeded it would have left America’s west coast undefended. In the event, his opponents knew his intentions as well as his own subordinates did. Between 3 and 5 June 1942, the US lost one carrier and one destroyer. Japan lost four carriers, two cruisers and three destroyers. American air power was intact. Japan’s was wiped out.
Though it would seldom exert such important influence on events, signals intelligence in the Pacific provided (as it did in the other theatres) a steady flow of highly useful information that often gave the Allies the edge, though – as always in the cryptanalysts’ war – there were costly setbacks caused by an inability to decipher changes in codes. Hugh Melinksy recalled:
In early 1942 1 Wireless Unit (Royal Australian Air Force) intercepted a message referring to the Japanese building an airstrip on an unknown island in the Solomons, spelled in Kana GA DA RU KA NA RU. This turned out to be Guadalcanal. (The Japanese language has no ‘L’ sound because they cannot hear any difference between ‘R’ and ‘L’.) The US naval staff decided that it must be recaptured before the airstrip was completed, and so US Marines landed there on 7th August. A ferocious struggle continued until December with heavy losses on both sides. The Allies lost four cruisers on 8th August partly because they were unable to decode a naval message about two Japanese cruiser divisions proceeding to Guadalcanal. The message was in JN25 code which changed periodically and it took us time to catch up with the alterations.
The codebreakers’ greatest coup, other than the discovery of plans for the attack on Midway, was the ambush of Admiral Yamamoto. Eavesdroppers picked up the news that the admiral – to Allied minds perhaps the most notorious Japanese arch-villain – was to make a tour of inspection by air. There was considerable debate as to whether he could be shot down without compromising the source of this information, and only after lengthy debate did the attack go ahead. Alan Stripp told the story:
In the first half of April 1943 Yamamoto ordered large scale attacks against Guadalcanal, Port Moresby and Milne Bay, and followed these up with a personal visit to his forward naval bases in the Bougainville-Solomons area. The Rabaul commander sent a message on the 13th giving the admiral’s travel arrangements and this was picked up by 1 W.U. at Townsville, and also by the Americans in Honolulu. It was decoded almost entirely, and read:
On 18th April C in C will visit RXZ [Ballale-Buin], R.. and RXP in accordance with the following schedule:
Depart RR [Rabaul] at 0600 in medium attack plane escorted by 6 fighters.
Arrive RXZ at 0800. Proceed by minesweeper to R.. arriving at 0840. (Have minesweeper ready). Depart R.. at 0945 and arrive RXZ at 1030.
Depart RXZ at 1100 in medium attack plane. Arrive at RXP at 1110.
Depart RXP at 1400 by medium attack plane and arrive RR 1540.
In case of bad weather the trip will be postponed one day.
The Americans weighed the risk that to use this information would be to blow the gaff and perhaps gain them the assassination of one man for the loss of great quantities of tactical and strategic intelligence. The heavy personal item in their balance sheet was that they could not forgive him for the totally unexpected, unprovoked and devastating attack on Pearl Harbor, which was known to be his personal idea. It was also reckoned that his loss would depress Japanese morale, and that he was a ruthlessly efficient naval commander whom it would be hard to replace adequately.
On 18 April 1943 a flight of P38s, with long-range fuel tanks attached, set off from an airfield on Guadalcanal. At 0745 that morning they sighted their quarry: two bombers with an escort of fighters. Both bombers were shot down. Yamamoto’s plane went down in jungle and the Admiral was killed. His chief of staff, Admiral Ugacki, was aboard the other bomber, which crashed in the sea, and he survived. The Japanese waited three days before announcing Yamamoto’s death. To the immense relief of the Allies, the JN25 code was not changed. The operation was thus a complete success.
Throughout the Pacific islands, administrative staff followed the fighting troops, setting up camps and ‘offices’ (usually in tents or dug-outs) to house the various support services upon which the armies depended. Alan Stripp described his British unit’s arrangements:
The operational part of the unit consisted of three large lorries with a canvas awning stretched between two of them. Under this I and four or five others worked at tables, with the wireless sets in one lorry and the Intelligence Officer in the other. The third lorry contained generators for providing the electricity. Airmen brought me messages which the operators had picked up listening to Japanese aeroplanes, and I did my best to decode and translate them. The other traffic consisted of all the other signals sent by the planes and their bases, and this was carefully examined and collected.
Irene Brion served with the American WAC in the recaptured territory of Hollandia in the Dutch East Indies. Her working environment was somewhat less primitive, but was still a challenge. She outlined the nature of her work:
The WAC area was on a partially levelled-off section of mountainside above the lake. The entrance to the burlap and wire-fenced compound faced the mess and rec halls, behind which a path led about fifteen feet to the tent area below. There were eighty tents, and two sets of showers and latrines. We were allowed to pick a tent. We decided to call our home ‘Secret trash,’ a term referring to waste papers that we used in our signal intelligence work.
We were ordered to report for duty in the advance echelon of SIS (Signal Intelligence Unit). SIS was a branch of Central Bureau (CB), an international bureau composed mainly of American and Australian signal intelligence units based in Brisbane, Australia. SIS, under the operational control of GHQ, was directed by Col. Abraham Sinkov. He was one of the three original disciples of the ‘father’ of SIS, the brilliant William S. Friedman, who was appointed by the army to found and organize the service in 1929.
Our office, one of three Quonset huts near the GHQ buildings, was ideally situated for work that required a quiet, relaxing atmosphere.
At that time, SIS, which included girls who had been trained before our arrival, was working on decoding one of the Japanese shipping codes. We were told that, in the haste of their departure from Midway, the Japanese hadn’t burned the codebooks or thoroughly stirred the ashes, setting the stage for a valuable entry to be made into their code systems. [During training], we’d been thoroughly warned about the necessity of destroying all papers and scrap pieces that we’d been working on during the course of a shift, by burning them and carefully stirring the ashes. Despite the remote location of GHQ, this procedure was carried out every night, using a large oil drum that stood outside our hut.
It was our job to recover new values from messages to fill in gaps in the codebooks, then pass them, along with any pertinent information from the actual messages, to our superiors in signal intelligence. We’d been so well trained at Vint Hill [training camp in Virginia] that the transition was smooth despite the fact that we received little if any direction or supervision from our superiors. After preliminary explanations, we sat down at tables, spread out our papers, and went to work.
She outlined the structure of SIS:
We knew little about the organization of SIS, but it worked something like this. Signal radio intelligence would intercept messages that the Japanese transmitted in an enciphered four-figure code and send them along to traffic analysis, where call signals and dates of transmission would be identified and analyzed. Somewhere along the line, the key words used to encipher the four-figure code would be determined, and the messages would be passed along to our cryptanalytic section. At that time the Japanese were using exotic names for key words, which they changed every month.
Each of us, or each pair if two were working together, would copy on a large sheet of paper a number of messages that used the same key word. Decoding with known values, we’d run through the messages until they yielded a sufficient amount of plain text to make sense. From the text of a given message we could often assume values for the unknown code words and test them against the texts of other messages to assure validity. We’d work until we either completed the messages or determined a sufficient amount of text to extract information and add values to the codebook.
The messages usually followed a set pattern – the call sign of the transmitting station, sets of numbers informing the receiver which code to use, and other necessary information. The text of the message ignored one of the great sins for the transmitter – the necessity of avoiding repetitive phrases. Typically, they began with the name of the ship, the time of departure, the nature of the cargo, and the expected time of arrival. Most textual material was boring to us but often informative to our superiors in G-2, who were eager to get information concerning frequency, concentration, and types of supply movements. However, I remember a message that told of a ship carrying an interesting cargo – women. Were they geishas? Whoever they were, they didn’t rate the status of passengers.
I loved the work. I would lose all track of time pursuing the meaning of the messages, and it was exciting when things began to fall into place. I’d take a breather and pass a note to Bell or Gemmy, usually containing a cartoon, a pun on a Japanese word, or a line from a pop song we’d been trying to recall. They did likewise. Despite those frustrating times when we struggled to come up with something that made sense, we all took great satisfaction in doing the job we were trained to do and making a tangible contribution to the war effort.
And, as might be expected, it wasn’t always easy to stay healthy:
Our chief problem was ‘jungle rot,’ a term that covered a wide range of fungal infections. Sooner or later most of us got it, usually the mild variety that was treated with gentian violet. I got some between my fingers, which cleared up quickly, but Marge Wilhelm and Edith McCann’s had cases that required hospitalisation. Marge broke out in sores all over her legs. Marge’s was the most serious – the type that attacks the optic nerve and causes blindness – and she had to be shipped home. We learned later that she lost 90 per cent of her vision.
Patricia Penrose worked at home in Australia, but one of her tasks was to deal with messages from clandestine groups in the jungles of New Guinea. In a sense she felt she shared the danger of those at the other end of the radio link. On one occasion particularly she held her breath. She remembered:
Scattered throughout these islands were groups or parties of men attached to ‘Z’ Force. Each party being known by an individual name such as ‘Lady Bird,’ ‘Lizard’ etc. They were made up of Dutchmen, Englishmen, Filipinos, Japanese, Americans and Australians. All hand-picked personnel, many with a local knowledge of the islands and conditions, they were highly trained, equipped and surreptitiously landed at their destinations. It was necessary for us to learn American, Dutch and more English ciphers as groups from all these countries and Australia were combined in ‘Z’ Special Unit.
The job of the ‘Z’ Force parties was to radio back information about enemy movements as well as raiding and sabotaging enemy supplies, equipment or landing strips. Their work was hazardous, hiding out in caves, living on dehydrated rations and forced to work under cover. Not only were they in danger from Japanese troops, but they also risked betrayal by hostile local inhabitants. Working under these conditions it is little wonder that when we received their cryptograms they were often corrupt.
Rarely could we send for a repeat, so this would mean spending hours working out possible alternatives or solutions. I remember one signal sent from a party up a tree, their last line read: ‘Japs are all around. Imperative you maintain strict radio silence.’ Failure to do so on our part would have meant their easy detection and certain death. After several worrying days we were relieved to hear they were comparatively safe again.
Her own experience – working in an office, where rows of similarly uniformed young women sat at tables encrypting and decrypting – was more typical of the work of codebreakers, whether in Melbourne, Brisbane or Bletchley.
The Cipher Office was arranged with an officer at the head of the room. Next to him was a Warrant Officer who was Section Leader. The remainder of the cipher staff sat in twos, facing each other across tables.
Day shift commenced at 8 a.m. and finished at 4.30, evening shift concluded at 11 p.m. and night shift went until 8 a.m. The bulk of the traffic was dealt with during the day so two girls were rostered for this shift, if there was a big carry-over of work to the evening shift one would stay on duty a while to help the girl on evening shift. The night shift wasn’t so busy, but it was very necessary to deal with incoming messages which could be vital. Shifts were rotated every six days at the end of which you had a 24 hour stand-down.37
… It was our job to faithfully encipher the cryptograms on the Type X machine. These signals were in turn passed to Staff Duties for retransmission to MI5 in London, where expert cryptographers worked at breaking enemy ciphers. [One of her colleagues] stayed at it for eight months before the strain became too great. Carrying out such exacting work on a loud and noisy machine in a confined space left her mentally and physically exhausted.
She remembered how the constant need for security could also become stressful:
Arriving at the Cipher Office one morning I was given an unusual assignment. I had to take a book of key figure groups to a certain printer. He was to make a number of copies and, when the job was completed, I was to break the type, then return with the original and the type. I had strict instructions to keep the printer and his press under observation at all times. This was alright in theory but the man found it necessary to leave me on several occasions and obviously there were places I just couldn’t follow him. So I sat with my eyes fixed on the press. As the hours went by, I became aware of a pressing necessity to leave also, but the call of duty proved greater than the call of nature and I crossed my legs and sat it out. Surely type was never broken so quickly as by my hands that day.
Some of her colleagues suffered even when asleep, as the pressure of their work invaded their subconscious:
Early in the new year of 1944 the strain of high pressure shift work and camp conditions started to take its toll on the girls. Some of the morse operators became ‘morse-happy,’ often sitting up in bed, ‘dit-dahing’ in their sleep. Throughout each section girls were suffering with nervous disorders and a general lowering of resistance.
The job was not without its lighter moments, for domestic news as well as tactical and logistic messages were sent through their Morse keys. They often had to encrypt and pass on to men serving overseas the news that their wives had given birth. To save time, a formula was used by the operators for this purpose:
They set up a pre-arranged signal for officers whose wives were expecting babies. This was the quickest and most direct method of notifying the fathers once they left the mainland. Should the baby be a girl, we sent: ‘Bicycle arrived today.’ If it was a boy, we added ‘complete with tools’.
The war was influenced by units like those Alan Stripp described because they pinpointed targets for Allied air forces. In spring 1944, said Stripp:
The Japanese were withdrawing their main air bases westwards. Intercepts up to late March indicated that there were at least 310 aircraft parked on the Hollandia airfields, and on the 29th March the interceptors picked up a message from the air commander, Lt. Gen. Teramoto, directing the Hollandia base commander to evacuate his planes the next morning. The message was quickly enciphered again and flashed down to Brisbane. The next day, 150 [Allied] planes struck at midday destroying 25 planes; on the following day 138 attacked and destroyed a further 138; and on 3rd April a third raid by 310 aircraft completed the destruction of what planes remained. Aerial photographs revealed 288 wrecked aircraft littering the runways. The way was paved for the invasion of Hollandia.
The quality of specific intelligence enabled the Allies to repeat one of their most successful operations. Having done away with Admiral Yamamoto, they dispatched his successor in the same way:
Wireless Units kept MacArthur’s staff informed of enemy air plans, locations of airfields, movements of aircraft and types, details of bombing and reconnaissance missions, fuel stocks, air cover over convoys, and weather reports. On 3rd April (1944) a message reported that a flying-boat which had left Palau on 31st March failed to arrive at destination in the Philippines. Admiral Koga, C-in-C of the Combined Fleet lost in electric storm. His death was kept secret for more than a month.
Stripp described how the targets they indicated could be economic:
Many successful sinkings were achieved in the second half of 1944, especially with those convoys plying from Japan to Singapore via Manila, and picking up vital oil from Borneo on the return trip. So valuable was this fuel that the Japanese even used battleships to escort their tankers.
They could also be what proved to be the first of the Kamikaze squadrons:
An intercept of 3rd November 1944 disclosed the formation of a ‘special attack unit’ at Clark Field (Philippines) of twenty-two bombers and fighters, the first to be devoted to suicide bombing.
As Japan’s resources were reduced by destruction or capture, its armies’ desperation grew. Stripp recalled an act of deception that in turn called for the usual counter-deception – the pretence that an action was the result of chance encounter rather than codebreaking:
By June 1945 the shortage of shipping in the East Indies area was so acute that the Japanese had decided to use a hospital ship, the Tachibana Maru ‘for purposes other than those prescribed by the Geneva Convention.’ The ship was renamed the Hirose Maru and bore floodlit crosses twenty feet high. It was to carry 1500 troops and 150 tons of stores and munitions from Tual in the Kri Islands to Surabaya in Java. In case the vessel was stopped and searched, the troops were issued with hospital clothing, daily sickness reports were kept, and the regimental colours were sent by air so as not to give the game away.
The messages arranging this deception, sent in May, were read, and the vessel was intercepted by two destroyers north of Timor. The search party found that cases marked ‘medical supplies’ contained ammunition, and the only wound among the patients was confessed to being the result of a packing-case being dropped on a thumb. In order to conceal the source of the information leading to the interception of the ship, its sighting had to be attributed to air reconnaissance, and – as always in such instances – a flight was especially arranged so that the enemy would not become suspicious.
By the spring of 1945, it was clear that Japan was as doomed as Germany. Its forces were in retreat everywhere and the ‘Greater East Asia Co-prosperity Sphere’ was almost all in Allied hands. Major defeats throughout Asia – at Midway, Kohima and Imphal, Leyte Gulf – left no doubt that its military power was spent. Japan itself was being devastated by relentless intensive bombing (the country’s predominantly wooden housing meant that damage to property was greater than in Europe) as Anglo-American aircraft roamed virtually unchallenged. Japanese planes, frequently grounded by lack of fuel, were increasingly used by suicide pilots in a desperate attempt to trade the lives of a few untrained fliers for enemy ships. The militarist government, aware that invasion would come within months, had no intention of giving up, and intended to contest every landing. Like their allies in Europe, they were driven to fanaticism by the knowledge that they could expect no mercy. Here too there would be no negotiated surrender.
It seemed to many among the Allies that the biggest clash of the entire war was still to take place and the build-up for the invasion of Japan, which it was estimated could cost the lives of a million soldiers, had already begun. In the event, the atomic bomb was to make this operation unnecessary, but in the meantime the European war had ended at the beginning of May 1945 and vast resources were suddenly available for the Pacific. These included the cryptanalysts of the SLUs, who had followed Allied armies from North Africa to Italy and, in history’s longest continuous advance, from Normandy to Berlin. While Bletchley stood down its German speakers, Japanese became the language of greatest importance. Japan was far from defeated and the conflict in the Far East was thought likely to last a further one to three years. It would be necessary to set up there the same infrastructure that had existed in Europe. Little by little, as their duties dwindled and they became available, the specialists were sent to Australia to consolidate an efficient organization before being spread through the myriad islands and archipelagos that made up the Pacific theatre.
When veterans of the European and Middle Eastern theatres arrived in Brisbane to prepare for what was expected to be a long and costly campaign against the Japanese home islands, they were struck immediately by the slow and easy-going pace at which their counterparts worked. While they found Australians charmingly hospitable and enjoyed what promised to be an active social life, they privately despaired of injecting a sense of urgency into the operations of the SLU. Another potential difficulty was the complete American dominance of the Pacific theatre, which meant that the codebreakers were working directly for the reputedly anti-British General Douglas MacArthur. It promised to be a challenging time. Squadron Leader SF Burley remembered:
I am afraid my first impressions on arrival were grim and I despaired for there seemed to be such a lack of enthusiasm. I was only interested in setting up all the arrangements for getting Ultra signals going and I had the impression that there was no hurry. Perhaps my only consolation was that the first echelon of the SLU would soon arrive and then we could get moving. At that time I had reluctantly come to the conclusion that as far as the Australian services and civilians were concerned the war was a long way away, which in fact was very true. It was a strange feeling to be located somewhere where after the hustle and bustle at home, where the order of the day was ‘to keep moving, there was a war on,’ now to find myself just waiting around. We had ideal offices, a whole floor at the Australian Mutual Providence Building in Queen Street, Brisbane. This had been part of General MacArthur’s headquarters. My office was next to the American Intelligence and the whole lay-out was satisfactory. General MacArthur’s attitude was very much anti-British and he really did not like the idea of us being around but despite this our relations with American units to whom Ultra was given were good and the material much appreciated.
Some experienced the discomforts of working in the jungle as well as in a Brisbane office block. The primitive shelters in which they sought privacy were all too familiar to veterans of the Middle East, Italy and Normandy, though the climate was more savage:
We were requested to set up Command Posts in Morotai, Lae and subsequently Labuan. These were under the command of Squadron Leader Parsons who had been in North Africa and Italy. Here we had problems for we had tented army headquarters and it was difficult to get what I wanted. The general idea it would seem in the minds of the Australian advance headquarters was that the SLU should be considered as part of the signal section which was totally unacceptable, but we were able to organise a separate area. Even so it was still pretty primitive and I found it necessary to brief all personnel on the security position and its problems and to make sure that a) the area allotted to the SLU was out of bounds to all other personnel, and b) in order to achieve this the Type X tents etc. should be enclosed by a further surround of hanging tent material to keep out anyone who became too nosey. It achieved its object and there were no incidents. Apart from all these problems the heat and the swamps didn’t help to make life too easy. Transport found difficulty in the ever-present mud and so to overcome this problem palm trees were cut down by the army and laid to make a road.
Jack Poole was another old hand who expected to be in the Far East for a lengthy spell. He too worked in the Australian Mutual Providence Building in Brisbane, and felt that his unit was just getting into its stride when they found themselves redundant. For him, the war ended in spectacular, if convivial, circumstances:
This conspicuous semi-skyscraper in the centre of the city had been part of General MacArthur’s headquarters. The SLU offices were on the 8th floor and commanded a magnificent panoramic view from the window of the cipher room. While the SLU staff already in occupation did most of the work those of us who arrived as reinforcements lost no time in acclimatising ourselves to this new field of Ultra. After the activities in the European campaign we found the set-up rather different but basically, of course, the system was the same. It was not many days before the remainder of the party with whom we had left the United Kingdom joined us and we were slotted into the duty-rosters of the A.M.P. Building. Our paramount task was to ‘get the feel’ of the type of material handled in the Pacific area in readiness for future operations. Life was relatively undemanding.53
… Things were moving behind the scenes that were calculated to affect us but in the meantime we enjoyed our work in the A.M.P. Building. Then came the day we were called together and Squadron Leader Yeudall explained that Flying Officer Reynolds was to take an SLU up to Manila. The list for the plane trip was read out and I found myself at the head of one of the parties. We packed our kit-bags in readiness for the signal to depart for the Philippines and decided to have a last fling in Brisbane. I and a couple of sergeants accepted an invitation to a dance at the RAF Recruits’ Centre.
And then suddenly the world in which the codebreakers had lived for five long and eventful years came to an end. After the dropping of the atomic bomb on Hiroshima on 6 August 1945, the Japanese stopped fighting a few days later and would officially surrender on 2 September:
As the nostalgia flowed there was a commotion on the dance floor. People stopped dancing and began to dash towards the hall door. An airman rushed into the Officers’ Mess. ‘It’s all over!’ he yelled exuberantly, ‘The Japs have surrendered.’ The atom bombs had obviously brought the Far Eastern hostilities to a precipitate end. But the officers at the dance were cautious and telephoned the Brisbane Courier Mail for confirmation of the news. Yes, said the news room, it was no rumour!
The hectic celebrations continued next day and there was pandemonium in the centre of the city. For days the revelry was sustained and SLU boys played a conspicuous part, when not on duty, in the various events that were arranged. While all the shouting and the tumult was dying down, the SLU team was unwinding. The Manila trip was, of course, called off. Although I was glad that the war in the Far East was over I must admit to a feeling of slight disappointment in not getting up to the Philippines before the end came. It would have meant that I should have been at the scene of both the surrenders in World War II.
The swiftness of the Japanese capitulation took everyone by surprise and men who had barely unpacked, or had at least hardly acclimatized, found themselves ordered home for demobilization. There would be no further glory, or achievement, for the SLUs after all. Josh Reynolds ended his recollections with this fine tribute to all those who served in the Special Liaison Units:
A small contingent of SLU marched in the victory parade in Brisbane. I think little has been said or even known about the work of the SLU organization in the Pacific. That it must, as in Europe, have proved invaluable was evident from the insistence of the main Australian and some American headquarters to have Ultra and to have it quickly. There is also no doubt our boys did a marvellous and most demanding job often under the most difficult conditions.
Squadron Leader SF Burley also remembered the final days of his service. With little time in which to adjust to changed conditions, his unit had to begin the process of closing down their office. He too paid tribute to his colleagues:
The role of the SLUs was now coming to a close for although we were well advanced with the arrangements for echelons to be provided to forward elements of both air force and army commands ahead of those already provided, the atom bomb was dropped and the war in the Far East had come to its end.
There was of course the formal closing down of stations and the mammoth task of collecting signals and other materials etc. for return to the UK. It was arranged that Squadron Leader Parsons was to be responsible for returning these by H.M. ship and having metal banded boxes made to take them. Parsons carried cover instructions that no one, repeat no one, could have access to the boxes and if there were any questions he should refer them to Group Captain Winterbotham. There were none asked and in due course the boxes arrived back. Here perhaps it was good to remember that our final signal to Bletchley Park saying we were closing down was not strictly in official parlance and we were delighted to receive back an acknowledgement in the same light-hearted tone – ‘Well done oh true and faithful servants, thanks a lot,’ or some such words. We at least appreciated that Bletchley Park had a sense of humour.
And Josh Reynolds concluded:
May I sum up the vital part of my life in the SLU. A unique allied inter-service operation by few people, mostly enthusiastic amateurs with relatively low ranks who distributed vital, accurate information on enemy intentions to Allied political and military leaders and decision-makers on an almost worldwide basis. It was a quick, economical service which operated for five and a half years, 24 hours a day, 7 days a week, and all this time there was no breach of security. Security both technical and operational of the whole operation was paramount. The SLUs were never found wanting.