The Emperor's Codes Read online

  THE

  EMPEROR'S

  CODES

  THE

  EMPEROR'S

  CODES

  THE THRILLING STORY OF THE ALLIED

  CODE BREAKERS WHO TURNED

  THE TIDE OF WORLD WAR II

  MICHAEL SMITH

  Copyright © 2000, 2011 by Michael Smith

  All Rights Reserved. No part of this book may be reproduced in any manner without the express written consent of the publisher, except in the case of brief excerpts in critical reviews or articles. All inquiries should be addressed to Arcade Publishing, 307 West 36th Street, 11th Floor, New York, NY 10018.

  Arcade Publishing books may be purchased in bulk at special discounts for sales promotion, corporate gifts, fund-raising, or educational purposes. Special editions can also be created to specifications. For details, contact the Special Sales Department, Arcade Publishing, 307 West 36th Street, 11th Floor, New York, NY 10018 or [email protected].

  Arcade Publishing® is a registered trademark of Skyhorse Publishing, Inc.®, a Delaware corporation.

  Visit our website at www.arcadepub.com.

  10 9 8 7 6 5 4 3 2 1

  Library of Congress Cataloging-in-Publication Data

  Smith, Michael, 1952 May 1- The emperor's codes: the breaking of Japan's secret ciphers / Michael Smith.

  p. cm.

  Original edition has subtitle: Bletchley Park and the breaking of Japan's secret ciphers. Originally published: London : Bantam Press, 2000.

  Includes bibliographical references and index.

  ISBN 978-1-61145-017-0 (pbk. : alk. paper)

  1. World War, 1939-1945--Cryptography. 2. World War, 1939-1945--Electronic intelligence. 3. Great Britain. Government Communications Headquarters--History. 4. Machine ciphers. 5. Cryptography--Japan. 6. Cryptography--United States. 7. Cryptography--Great Britain. 8. ULTRA (Intelligence system) I. Title.

  D810.C88S656 2011

  940.54’8641--dc22

  2011001771

  Printed in the United States of America

  For Ben, Kirsty, Louise, Leila and Levin

  CONTENTS

  Acknowledgements

  Map

  Introduction

  1 Singapore, December 1941

  2 Borrowing the Cables

  3 A Spy Base in the Far East

  4 Diplomatic Secrets

  5 Preparing for War

  6 Purple Magic

  7 Working with the Americans

  8 East and West Winds

  9 The Americans Take the Lead

  10 A Tricky Experiment

  11 Midway: The Battle that Turned the Tide

  12 Friends Fall Out

  13 Breaking the Military Attaché Code

  14 Central Bureau's Big Break

  15 The Yamamoto Shootdown

  16 The Bletchley Park Strippers

  17 Return to Colombo

  18 An Alliance under Threat

  19 Operation Capital

  20 Defeat into Victory

  21 MacArthur Returns

  22 The Atomic Bomb

  Notes

  Bibliography

  Index

  Illustrations follow pages 78, 126, and 222

  ACKNOWLEDGEMENTS

  I am extremely grateful to all those former codebreakers and wireless operators who wrote to me or agreed to be interviewed for this project. Whether or not they are quoted in the book, their contributions played a major part in providing the information and encouragement I needed to complete it. I am particularly grateful to a number of people who went out of their way to help and encourage me. These include Geoff Ballard; Edith Becker; Geoff Day; Joan Dinwoodie; Joe and Barbara Eachus; Robert Hanyok; Margaret Henderson; Phil Jacobsen; Dennis Moore; Joe Richard; Harry G. Rosenbluh; Hugh Skillen; Joe Straczek; Jimmy Thirsk and Dennis Underwood. I am also grateful to Steve Kelley for sharing his expertise in Japanese machine ciphers with me and to Hilary Jarvis for her advice on the Japanese language. Special thanks are due to Hugh Melinsky; Norman Scott; Alan Stripp and Maurice Wiles for allowing me to quote from their own works and for their generous help and advice. I must also thank Ron Bonighton, Director of Australia's Defence Signals Directorate, for kindly sending me a complete set of the Technical Records of the Central Bureau (a uniquely generous example of openness from a signals intelligence organization); Margaret Nave and the Australian War Memorial for allowing me to quote from Eric Nave's original manuscript; the Library and Department of East Asian Studies of the University of Sheffield for permission to quote from the diaries of Malcolm Kennedy; Christine Large and the staff of the Bletchley Park Trust for their assistance and dedication to keeping the memory of the codebreakers alive; and last but certainly not least Ralph Erskine whose willingness to share his unrivalled knowledge of wartime codebreaking operations with me provided a constant lifeline.

  I should also like to thank Charles Moore, Editor of the Daily Telegraph, for generously allowing me to take a sabbatical to write The Emperor's Codes; Sally Gaminara, Katrina Whone, Simon Thorogood and Sheila Lee at Transworld for their painstakingly professional work on this project; my agent Robert Kirby for his constant enthusiasm; and my wife Hayley for her patience and encouragement. The Emperor's Codes

  ‘I believe most experienced cryptanalysts would agree with me that cryptanalysis is much closer to art than to science, and this is what makes the personal factor so important.’

  John Tiltman, Chief Cryptographer at Bletchley Park 1940–5

  THE EMPEROR'S CODES

  INTRODUCTION

  The extraordinary achievements of the British codebreakers based at Bletchley Park in cracking Nazi Germany's ‘unbreakable’ Enigma cipher are now widely known. This initially oddball collection of mathematicians, classicists and musicians performed unexpected miracles in the hastily constructed huts scattered around the grounds of an ugly mock-Tudor mansion in the small Buckinghamshire market town of Bletchley. The location had been selected in part because it was just far enough outside London to escape the threat of German bombs. But perhaps more importantly, it was midway between the ‘glittering spires’ of Oxford and Cambridge, from where most of the leading codebreakers had come.

  The apparently amateurish nature of these early beginnings was epitomized by the actions of the codebreakers’ boss, Admiral Hugh Sinclair. His power over the Government Code and Cipher School, as the codebreaking operation was known, came from his role as ‘C’, the head of MI6. Perhaps surprisingly, this was in fact no secret within the limited confines of what was then described as ‘polite society’. But Sinclair was far better known among his immediate circle of friends for his reputation as a man with a taste for the high life. They christened him Quex after the title character in Arthur Pinero's popular play The Gay Lord Quex, who was described as ‘the wickedest man in London’.

  This passion for the high life, and the fact that he had the personal wealth to fund it, dominated Sinclair's dealings with the codebreakers – an earlier headquarters had been based in the Strand in central London, apparently for no other reason than that it was close to the Savoy Grill, one of his more favoured haunts. Unable to persuade anyone in government to pay for the codebreakers’ new country mansion, he dipped into his own pockets to buy it, immediately moving his favourite chef from the Savoy Grill to Bletchley Park to ensure that the codebreakers were well fed.

  The result was an atmosphere at the highly secret Station X not unlike that of a weekend party at an English country mansion. The initial number of codebreakers in what was known for security reasons as ‘Captain Ridley's Shooting Party’ was little more than 200. They lived in hotels in the surrounding towns and all initially worked either in the mansion itself or in a neighbouring boys’ school th
at had been commandeered for the purpose. When not breaking codes, or if simply seeking somewhere to contemplate their cryptographic puzzles, the codebreakers could wander through the lawned grounds of Bletchley Park, which included a maze, a lake and a number of rose gardens.

  This brief idyll was shattered in part when the chef, unable to cope with the sometimes unusual demands of the more idiosyncratic ‘guests’ at Bletchley, attempted to commit suicide. But other factors were far more crucial in the transformation that was about to take place. As the Phoney War drew to a close and Hitler turned his attention to Western Europe, Britain's back was very much against the wall. The need to break Enigma and to obtain a permanent source of high-grade intelligence on German movements was paramount.

  Bletchley Park was soon swamped by new recruits, some of whom had experience within those British companies which were adopting more efficient business practices pioneered in America. By the time the pre-war British codebreakers such as Dilly Knox made the first British breaks into Enigma, Bletchley Park's operations were already turning into a production line. The grounds increasingly resembled a building site. The maze and rose gardens disappeared to be replaced by prefabricated wooden huts in which worked thousands of people, the vast majority of them women.

  The popular view of Bletchley Park, as an organization manned by brilliant but amateurish eccentrics, is misleading. Certainly early on the organization was dominated by eccentrics such as Knox and Alan Turing, but the incredible progress made by some of them not just with Enigma but also with the development of Colossus, the world's first programmable electronic computer, gives the lie to suggestions of amateurism. The vast majority of those working at Bletchley Park were in fact neither amateurish nor eccentric, although many were indeed brilliant.

  The work they did to unravel the intricate workings of the Enigma cipher machine is estimated to have cut around two years off the length of the war in Europe. But that is only part of the story. The British codebreakers were not just working on German codes and ciphers, they were breaking those of all of Germany's allies in Europe: Italy; Hungary; Romania; Bulgaria; and initially, of course, the Soviet Union, as well as those of neutral countries such as Sweden, Spain and Portugal. The codes and ciphers of a number of non-European countries were also coming under the scrutiny of the British codebreakers, and the most important by far of these was Japan.

  Almost a year before Pearl Harbor, the British and the Americans had agreed in principle to share their codebreaking resources. The two most important items on the agenda were the major machine ciphers of Germany and Japan. Bletchley Park handed over a paper version of the Enigma machine, providing the Americans with the start they needed to read the Wehrmacht's messages. In return, the US Army gave the British its most precious piece of codebreaking equipment. The Purple machine had been built to decipher the messages passing between the Japanese Foreign Ministry, the gaimushoo, and its major embassies abroad.

  Although it was nowhere near as difficult to solve as Enigma, or the later German enciphered teleprinter systems known collectively as Fish, the breaking of the Japanese ‘Type B’ diplomatic cipher machine, by a team led by the American mathematics teacher Frank Rowlett, was one of the great cryptographic achievements of the Second World War. The Americans christened the Type B system Purple and the intelligence derived from it was given the codename Magic, probably because William Friedman, the head of the US Army codebreaking operation, routinely referred to his cryptanalysts as ‘magicians’.

  The amateurish eccentricity of the Bletchley Park operations is not the only myth to have grown up over the years. Enigma had originally been read not by the British but by the Poles. A team of Polish mathematicians, Marian Rejewski, Henryk Zygalski and Jerzy Rozycki, had managed to break it initially in 1933. While this was a brilliant achievement in its own right, it came at a time when German security procedures were very slack and the machines in use were at their most basic. Nevertheless, it provided the springboard that allowed the British to break the far more complicated systems put in place when the Second World War began.

  Quite why the British had not attempted much earlier to break the Enigma machine, which was first introduced within the German armed forces in the late 1920s, is difficult to understand. It may be, as Josh Cooper, one of their number, said, that they really believed that, having had their codes broken easily by the British during the First World War, the Germans would never allow it to happen again. At any event, it was not until shortly before the Second World War broke out that the Government Code and Cipher School made any serious efforts to break the Enigma cipher. During the inter-war years, the British codebreakers had concentrated on what were regarded as the three main threats: Bolshevik Russia, widely perceived as being the biggest danger not just to the Empire but to civilized life itself; the upstart United States, still an uncertain force in world affairs; and the main threat to British colonies in the Far East – Japan.

  This brings us to the other main myth: that the Japanese codes and ciphers were broken by American codebreakers. When Britain and the USA agreed to collaborate on breaking the codes of the Axis powers, a year before Pearl Harbor brought America into the war, the British were determined to keep control of the operations to break the Enigma cipher machine. It made sense that the Americans with their dominance of the war in the Pacific should likewise control the breaking of Japanese codes and ciphers, and a natural division of labour built up.

  At the end of the war, details of how the US Army broke the Purple cipher were swiftly made public, to the horror of the British codebreakers. Almost certainly as a result of the rivalry between the two American services, news also leaked out of how the US Navy had broken the main Japanese naval code JN25, allowing its aircraft to shoot down the Commander of the Imperial Japanese Navy's Combined Fleet, Admiral Yamamoto Isoruku. This publicity led to the lasting impression that the Americans had broken the Japanese codes.

  By contrast, the British clamped down on any mention of the remarkable achievements of their own codebreakers, determined to ensure that they could continue to intercept the communications of other countries with impunity. It was not until the mid-1970s, with the publication of Frederick Winterbotham's book The Ultra Secret, that sketchy details of how Bletchley Park had broken the Enigma ciphers began to leak out. News of the construction by Tommy Flowers, a British Post Office engineer, of the programmable computer Colossus only emerged much later, displacing the claims of ENIAC, an American computer which had until then been regarded as the world's first.

  The official files, still in the possession of Bletchley Park's secretive successor GCHQ, did not begin to filter into the archives of the British Public Record Office at Kew, in south-west London, until the 1990s and those on the British codebreaking efforts against the Japanese were among the last to be released. Only now has evidence begun to emerge of how much work on Japanese codes and ciphers was done by the British codebreakers, not just during the Second World War but long before it even began.

  During the inter-war years, several of the leading members of the Government Code and Cipher School became expert at breaking Japanese codes and ciphers: John Tiltman, an infantry officer who had won the Military Cross in the trenches of the First World War; Eric Nave, a Royal Australian Navy officer seconded to work with the British; and Hugh Foss, one of the true British codebreaking eccentrics, were breaking Japanese military, naval and diplomatic codes and ciphers long before the equivalent American codebreaking operation saw similar success. Nave was reading all of the early Japanese naval codes in the late 1920s; British codebreakers led by Foss were the first to break a Japanese diplomatic machine cipher in 1934; and it was Tiltman who first broke JN25, a few weeks after it was introduced in the summer of 1939.

  The Americans, working independently, later came to make their own breaks into the Japanese codes and ciphers. But they were hampered by their lack of experience; the occasional refusal of the government to accept the necessity of their work (
one inter-war Secretary of State, Henry Stimson, later declared that ‘Gentlemen do not read each other's mail’); and an intense rivalry between US Army and US Navy codebreakers. It was not until Rowlett's breaking of Purple, at a time when all the British machine cipher experts were concentrating their energies on the German Enigma machine, that the American codebreaking operation really came into its own.

  None of this is to deny the dominant role played by the Americans in the interception of Japanese messages. At the start of the war, the British were forced to throw all their efforts into breaking the German ciphers. They had neither the resources nor the incentive to pay as much attention to Japanese codebreaking as it undoubtedly deserved. The main Royal Navy codebreaking establishment in the Far East had been forced by the Japanese threat to move first from Hong Kong to Singapore, then to Colombo and swiftly on to Mombasa, where its very remoteness from the battle crippled its ability to receive the Japanese radio messages on which the code-breakers depended. The Americans, by contrast, had immense resources both in terms of manpower and the mechanical tabulating machines that were indispensable to the operations to break the Japanese codes. The US Navy station at Pearl Harbor in particular was second to none in its efficiency and capability.

  The British understandably assumed that since they were willing to provide the Americans with all the intelligence they gained from breaking the Enigma ciphers, the Americans would do the same when it came to material gathered by breaking the Japanese codes. The US Army codebreakers proved as helpful during the war in the Far East as they had with their provision of the Purple machine. Sadly, the same cannot be said of some of the administrative officers controlling their naval counterparts.

  The refusal of certain sections of the US Navy to share intelligence with the British seriously hampered operations in eastern waters and led the Admiralty to suggest that the fledgling exchange agreement on codebreaking with the Americans should be scrapped. Given that this was to form the basis for the UKUSA Accord which played such a vital role during the Cold War and still links the two countries’ intelligence services, the potentially disastrous effects of the lack of cooperation are not hard to imagine.