| Home |
 |
Biography | |
Books | |
Articles | |
Videos | |
Order | |
Links | |
Contact |
The Reader of Gentlemen’s Mail
Contents, Preface, Two Chapters (3 and 9), A Footnote
�
Table of Contents
A Short Course in Codes and Ciphers
How Yardley Wrote His Best-Seller
1. All-American Boy
2. His Life’s Work
3. A History of American Intelligence Before Yardley
4. A Rival
5. Staffers, Shorthand, and Secret Ink
6. The Executive
7. Morning in New York
8. His Triumph
10. The Busy Suburbanite
11. End of a Dream
12. The Best-Seller
13. The Critics, the Effects
14. Grub Street
15. A Law Aimed at Yardley
16. Hollywood
17. China
18. Canada
19. A Restaurant of His Own
20. Playing Poker
21. The Measure of a Man
Bibliography
Index
(46 photographs in 16 pages)
�
Herbert O. Yardley is the most colorful and controversial figure in American intelligence. He became a cornerstone of it when he gave America its best intelligence – codebreaking -- and then, for an act to which he was driven by desperation, he became an outcast. Yet he has never had a biographer.
He deserves one. Though throughout America’s history individuals had broken codes as occasion demanded, they abandoned the work when the need ended. Yardley institutionalized it. In World War I, Yardley foresaw that the United States needed the information that could come from signals intelligence, established America’s first permanent agency to intercept foreign messages and break codes, and ran it well enough to prove its importance. He endowed his nation with its most trustworthy, high-level, voluminous foreign information. Then, after a secretary of state disbanded his organization on the ground that “Gentlemen do not read each other’s mail,” Yardley, out of work in the Great Depression, with a wife and son to feed, wrote a sensational memoir about his secret work and its successes. Entitled The American Black Chamber, its betrayal of trust rightly drew down the wrath of intelligence and military professionals, who refused thenceforth to have anything to do with him, even during the emergency of World War II. But it awakened thousands to the value of communications intelligence. Yardley owes his significance to what he did; his fame, to what he said.
At his most successful, in his 30s, Yardley, short, gray-eyed, prematurely balding, was quick-witted, capable, and an opportunistic self-promoter, always looking for the big buck. He did personal work on government time. He sold his soul for his book. He exaggerated his successes in his official reports and in his book but was honest in minor personal matters. He was an extremely competent executive, dealing effectively with subordinates and superiors. He told stories well. People liked him. He drank a lot, but held his liquor. He hunted, fished, played championship golf, won at poker. Though as a codebreaker he rose only a little above the average and as a codemaker he invented no new methods, he achieved a solution that, by helping eliminate thousands of tons of warship construction, saved the world millions of dollars and eased international tensions. Later, the revelations and striking style of The American Black Chamber made it an instant classic of intelligence literature. Yardley is a cult figure.
This book describes the arc of Yardley’s life. I have sought to embed that life into the context of its times, to infer its motivations, and to say why it matters. The book shows how Yardley’s boyhood demonstrated the imagination and initiative that enabled him to achieve what he did. It describes the competition between him and a think tank near Chicago to control American codebreaking and later between him and another great cryptanalyst, William F. Friedman. It considers the stories that Yardley was a drunk and a womanizer. It discloses how Yardley obtained the foreign code messages his agency needed despite laws protecting their confidentiality. It details his greatest success: his solution of Japanese codes before the Washington naval conference in 1921-22. It corrects, on the basis of documents, the almost universal belief that the publication of The American Black Chamber led Japan to change its codes and ciphers. It offers a surprising positive evaluation of the effect of that book by the cryptanalyst who led the attack on Japanese codes and ciphers before and during world War II. It debunks the slander that Yardley traitorously sold to Japan his solutions of Japanese messages. It proposes an answer to why America, alone of all the powers, closed its codebreaking agency in 1929 and, by implication, deals with the morality of intelligence
�
Chapter 3 A History of American Intelligence Before Yardley
Ralph Van Deman was in charge of
military intelligence. But neither he, nor Yardley, knew much about the
background of that activity.
It had begun even before the nation came into being. As head of the Continental Army, General George Washington sought information about British activity from spies. He dispatched one of his first only 11 weeks after independence was declared. The mission failed. Nathan Hale was captured while trying to return to American lines, his fame coming not from what he did but from what he said about having only one life to lose for his country. Washington improved as spymaster as the revolution proceeded. By 1779, he was all but running the Culper ring, which operated out of Long Island. One spy, Culper Jr., he directed, was “to remain in the City [of New York], to collect all the useful information he can – to do this he should mix as much as possible among the officers and Refugees, to visit the Coffee Houses, and all public places. He is to pay particular attention to the movements by land and water in and about the city especially.” He warned too of the need for security: dispatches should be delivered only to those assigned to receive them and forwarded “to no one but the Commander-in-Chief.” The ring’s spies encoded their reports, wrote some in invisible ink, and hid them for pickup in a hollow tree trunk -- what would later be called a “dead drop.” They told Washington how many redcoats were stationed where, what warships were anchored in New York harbor, what provisions were entering the town, and the like. He found their reports “intelligent, clear and satisfactory. But though he avidly sought spy information, none of it helped him win any victories. It seemed to provide him with a general picture of the situation – not as dramatic as a victory, but useful, if only in preventing a possible defeat.
His most valuable intelligence came from communications intelligence. America’s first cryptanalyst, James Lovell, a Harvard graduate, a teacher, and a member of the Continental Congress, solved a British dispatch revealing that a Royal Navy fleet planned to relieve Lord Charles Cornwallis, blockaded in Yorktown. Delivered to the French naval force, it scared off the British, ending any hope of rescuing Cornwallis, who had just surrendered. This set the seal of final victory on the American Revolution.
But as useful as this intelligence was, and as enthusiastic and wise as Washington had been in seeking and exploiting his spy information, the new American army did not incorporate a permanent unit for seeking and evaluating intelligence into its organization. Neither did any other army or navy of the time. For though commanders always sought information, usually it arrived too late to be useful, seldom could it be trusted, and almost never did it award them victory. They had little confidence in it. In only one of Edward Creasy’s Fifteen Decisive Battles of the World: from Marathon to Waterloo did foreknowledge of the enemy matter -- the battle of the Metaurus in Italy in 207 B.C., in which an intercepted Carthaginian message enabled the Romans to concentrate and defeat Hasdrubal before his brother Hannibal could reinforce him. The other fourteen were won by strength, brains, and will. Carl von Clausewitz epitomized this historical insignificance of intelligence when he gave it but three scornful paragraphs in Vom Kriege.
This situation did not change during the Civil War. Most information about enemy forces came, as it had for millennia, from scouts on foot and on horse and from larger cavalry reconnaissance troops. Prisoner and civilian interrogations filled in details, which sometimes were even accurate. The more glamorous tools of balloon observation, intercepted signals, and espionage rarely added anything of value. It is true that rebel spy Rose Greenhow learned of the Union decision in 1861 to advance on Manassas, leading to the Confederate victory at First Bull Run. And Yankee spy Elizabeth Van Lew forwarded reams of occasionally correct spy reports to the Federals. Allan Pinkerton, head of a detective agency, set up a Secret Service of the Army for the Union’s Army of the Potomac. He did well in counterespionage, arresting Greenhow, but less well in espionage. The reports from the spies he sent behind Confederate lines led him to overestimate Rebel strength, encouraging Major General George B. McClellan’s tendency to procrastinate. The North lofted tethered balloons in 1861, using them primarily to spot enemy positions and movements. In a typical case, aerial observation told General Charles Stone that the Rebel force across the Potomac that he was surveilling consisted of only four Mississippi infantry regiments and a Virginia battery. The South had no balloons. Both sides frequently tapped one another’s telegraph wires and read one another’s optical signals -- most messages being sent unencrypted. Of those that were encrypted, the South’s polyalphabetic substitutions were often solved by Union telegraphers, while the North’s route transposition system defied Rebel attempts to break it. So all in all, intelligence made but trivial contributions to the battles of the Civil War, on either side.
With the end of that struggle, the army forgot intelligence. Because intelligence units had varied from corps to corps, from department to department, no pattern could suggest itself to the postwar army. Most importantly, intelligence had not mattered during the war, which had been won without it, and no obvious intelligence targets presented themselves after it. So the army did not establish any agency for intelligence. It returned to subduing the Indians.
But military technology had by then begun to advance at a dizzying rate. Muzzle-loading guns were replaced by breech-loaders. Smooth-bore cannon gave way to rifled ones; cannonballs, to explosive shells; wooden ships, to ironclads. Steam had already replaced sails at sea and horses on land. Propellers took the place of paddlewheels and rails, that of roads. The machine-gun intensified firepower. The telegraph facilitated control. The railroad enabled nations to mobilize and deploy their large armies with timetable precision and to supply them in the field. Intelligence was needed more than ever to keep up with these developments. It also gained for the first time a target of great value: war plans. Knowledge of these gave commanders more specifics about enemy mobilization and likely offensives than older, more generalized campaign plans and consequently more time to prepare defenses than ever before. The generals remained skeptical that such intelligence would provide victory, but they conceded that knowledge of a plan might prevent defeat.
As these technological horizons were expanding, so were America’s political and economic ones. California had become a state in 1850; Japan was opened to trade in 1854; China, in 1858. Alaska had been bought in 1867; the transcontinental railroad was completed in 1869. The nation paid increased attention to the Pacific – and then to the world, as the European nations scrambled for colonies in Africa, Asia, even distant Samoa. In 1874, exports permanently exceeded the value of imports.
These currents pushed the American army and navy toward establishing intelligence sections. But for a while countervailing tendencies deflected that trend. The United States was protected physically by oceans and politically by the Monroe Doctrine. The United States didn’t need intelligence. Moreover, collecting information stank of spies and militarism – dishonest, undemocratic concepts. The United States didn’t want intelligence. Conservative officers rejected the concept, in part because it had never been done, in part fearing that new specialists would compete with them for jobs. They wouldn’t use intelligence.
But the pressures of technology and the nation’s growing interaction with the world drove the armed forces finally to see that more and better information was needed. The secretary of the navy established the Office of Intelligence on 23 March 1882. Three years later, the army’s adjutant general, Brigadier General R.C. Drum, acting, it is said, on a suggestion of the secretary of war because foreign information for which he had asked was not available, assigned an officer and a clerk to collect information about foreign militaries. On 12 April 1889, the secretary formally authorized a Military Information Division within the adjutant general’s office, and Congress the following year appropriated $1,500 “for the pay of a clerk attendant on the collection and classification of military intelligence from abroad.” At last the United States had formal intelligence agencies in its armed forces – a permanent institution for the first time.
In three years, the Military Intelligence Division grew large enough to be reorganized into four branches, in part because much of its work dealt with American mobilization plans and instruction. Yet no spies sought secret intelligence about other countries. No American company or government agency seems to have intercepted cablegrams, much less attempted to solve any that were encoded. Since the United States was not at war, no cavalry and infantry patrols fed information to field commanders and thence to the intelligence agency. The only foreign intelligence that it obtained, apart perhaps from maps, came from the new corps of military and naval attachés who had been dispatched to the major European capitals.
Despite the Military Intelligence Division’s organizational suffocation – it lay not under the commanding general of the army but under one of the secretary of war’s bureaux that dealt with logistical, fiscal, and administrative matters -- it quickened for a while under the command of an excellent leader, one of the first in the American army to recognize the importance of intelligence. Colonel Arthur L. Wagner, an outstanding personality who had helped the School of Application for Infantry and Cavalry evolve into the General Service and Staff College, had published in 1893 the first American work on intelligence, The Service of Security and Information. As tensions grew with Spain, he asked for permission to send one of his officers “to examine and report on the military situation” in Cuba. When the Spanish-American War broke out, he dispatched Lieutenant Andrew W. Rowan on his famous “mission to Garcia” – a reconnaissance that brought back maps and other intelligence from the insurgent Cuban. Wagner wangled himself into a cavalry unit, fighting in Santiago, and after the war returned to the staff college before directing the Army War College. He died in 1905, but he and his missionary zeal for intelligence had excited a subordinate.
Lieutenant Ralph Van Deman first walked into the three-room office of the Military Intelligence Division on the main floor of the State-War-Navy building in June 1897. Tall and gaunt, with big ears, he had been born in Ohio in the last year of the Civil War, had graduated from Harvard in 1888, spent a year in its law school and accepted an infantry commission in 1891. The army let him complete a medical degree from Miami University in Ohio. During the Spanish-American War, he had charge of the White House war map. Wagner’s evangelism for intelligence persuaded him of its importance, for when, sent to the Philippines to help fight the Aguinaldo insurrection, he was assigned to convert the Bureau of Insurgent Records into the Military Intelligence Division, he accepted the job not reluctantly, as did most officers assigned to intelligence, but with interest and effectiveness. Van Deman discovered a plot for an attack on Manila, which was thwarted.
In the United States, meanwhile, Congress and the press were exposing the army’s embarrassing lack of planning during the Spanish-American War, with its insufficient and rancid rations, its distribution of Civil War winter uniforms for a July campaign in Cuba, its employment of black powder instead of smokeless, its bungled troop embarcations, its inadequate medical supervision. The outraged nation and the new secretary of war, Elihu Root, demanded a better system. The English writer Spenser Wilkinson’s The Brain of an Army and General Emory Upton’s The Armies of Asia and Europe publicized the idea of the general staff. Prussia had proved its effectiveness when it engineered the speedy and efficient victories over Denmark in 1864, Austria in 1866, and France in 1870-71. Though America’s fear of militarism joined the army’s bureau chiefs and the commanding general in opposing a general staff, an ethos of progress and new concepts of scientific management helped Root convince Congress of the soundness of a general staff. The bill creating one was approved by President Theodore Roosevelt on 14 February 1903.
Where should the intelligence function be put? Britain merged it with operations. Prussia incorporated it into its two war-planning sections, east and west; it formed a separate intelligence element only upon mobilization, reverting to the peacetime amalgam after hostilities ended. France, however, specified intelligence as the 2nd Bureau of its general staff. The Americans copied this numeration, making intelligence the second division of their staff, the later G-2. But in 1908, intelligence was merged with the third division, ending its separate identity. Van Deman, who had held several intelligence posts in the normal rotation of army duties, arrived at that division in July 1915. Then, in 1916, despite World War I in Europe, Congress in effect cut the Washington staff, leaving Van Deman as the only officer there with any experience in intelligence. But neither he nor any of the American observers abroad recognized the tremendous changes that that activity was undergoing during the conflict.
It was expanding enormously, not only in volume, but in value, primarily from a new form of information. This was not spies, long regarded as a chief source. Their information was infrequent and slow; it was subjective; it came from untrained observers and so was often erroneous, and it was suspect – the agent may have reported correctly in the past only to set up a deception. Trench warfare produced volumes of prisoners, but they could tell about little more than the units to which they belonged and the weapons they served. Captured documents and materiel likewise yielded mainly order of battle information and technical detail. Aerial reconnaissance -- by eye or camera and from balloon or airplane – indeed became so good that after 1917 neither the Allies nor the Central Powers dared move troops in daylight hours.
But though photographs may persuade best of all, a new source proved the most valuable. This was radio intelligence. Radio has the great military advantage of being able to communicate quickly, easily, and cheaply – but the great military disadvantage that those communications can also be intercepted quickly, easily, and cheaply. And the intelligence produced from communications is extraordinarily trustworthy: it consists of the very words of the enemy. (Commanders have rarely tried to trick the enemy with fake messages, because they can too easily be mistaken for real, causing damage.) Usually, it is not sporadic but voluminous and continuous. Moreover, the intercepts often consist of plans and orders. As a consequence, communication intelligence – in particular its cryptanalytic branch -- gave skeptical commanders confidence in intelligence for the first time. Though no one in American intelligence then knew it, that source had helped Germany defeat tsarist Russia, paving the way for the Communist revolution. It enabled France – the war’s greatest cryptanalytic power – to block Germany’s supreme offensives on the western front in 1918. It had produced the greatest intelligence coup of all time, the disclosure that German Foreign Minister Arthur Zimmermann was offering Mexico three American states if it would join with Germany in warring on America -- a plot that, when made public, helped push the United States into the war and into world power. So communications intelligence was making intelligence into a significant instrument of war, no longer mistrusted but accepted and even welcomed by admirals, generals, and statesmen. This was the source that Yardley wanted to bring to America. But it was not yet known or accepted there – the Army’s Field Service Regulations never mentioned it among the forms of intelligence, so he faced a struggle against ignorance and inertia.
On 11 April 1917, five days after the United States declared war on Germany, the head of the Army War College proposed that the general staff organize a military intelligence unit. He thought it should be a division separate from his. But the chief of staff directed instead that the war college supervise military intelligence. On 3 May the secretary of war approved. Van Deman, the only man in Washington who knew anything about the subject, was chosen to head the section, and thus became the man to whom Yardley was directed.
His office was in the war college building. This handsome McKim, Mead and White structure, with its striking semicircular entrance entablature, rose in solitary splendor at the end of a parade ground on the shores of the Anacostia River where it joins the Potomac. Yardley rode a trolley to it, bearing with him the idea that would forever change American intelligence.
Van Deman’s face reminded him of that of a beardless Lincoln. He looked tired, but he grew intensely interested as Yardley outlined the need for codebreaking services in Washington to solve foreign diplomatic messages and on the Western Front to break enemy military messages. He knew that “neither the State Department nor the War Department has any real experts on cipher work,” and the possibility of the army’s having its own codebreaking agency in Washington seems to have surprised him, for of the memoranda about setting up an intelligence unit only one had even mentioned “analyzing the enemy’s codes and ciphers” – and that in passing in a list -- and did not propose a separate element for it. Hitt, Mauborgne, and the army’s other significant cryptanalyst, Moorman, had duties elsewhere, so Van Deman had accepted an offer from a private research organization near Chicago to solve ciphers for the government. But Yardley’s conviction of his mission intensified his normal persuasiveness. Though young and inexperienced, he convinced Van Deman that the general staff needed its own cryptanalytic unit.
And the major did something about it. He arranged for Yardley’s release from the State Department and for his commissioning. On 29 June, Yardley became a first lieutenant in the Signal Corps in the National Army – the organization into which draftees and volunteers were enrolled – with serial number O-159744. On 5 July, he was assigned to active duty and, on 15 July, he established and, as its sole officer, took charge of MI-8 – military intelligence, section 8. Thus began America’s first official cryptologic agency. No trumpets blared, no drums rolled, no troops paraded. But, unnoticed though it was, it marked one of the most significant steps in American intelligence.
For an office, Yardley was given a few square feet on a narrow balcony on the north side of the west wing of the War College building overlooking the library stacks. Lit by one of the wing’s semicircular lunettes, it had barely enough space for one desk for Yardley and one for a clerk or two. No walls or partitions set it apart. The floor was a grating. This was the cradle of American cryptology. Rocking it was its father, Herbert Yardley.
He was 28 years old, 5 feet 5 inches tall, 125 pounds. His head was round atop a short neck. His nose was straight and small; his hair was light brown, but his early baldness gave him a high forehead. He walked with short quick steps. He was convincing when he talked, tending to dominate a conversation, and he told stories well. He was bright. He had gained self-confidence and experience in organizing and running things from his presidency of his high school class, his captaincy of the football team, his acting in school plays, his creation of the sandlot baseball game, his leadership of many high school pranks, and his general popularity. He had broken some codes and believed he could crack others. He was ambitious. And now he had, via cryptology, a chance to be, not an underling, but a boss. Sure that he could handle the opportunity, he seized it.
But he had no personnel, no organization, no clout. And so the codebreaking went by default to the think tank outside Chicago.
�
Chapter 9 The Fruits of His Victory
[The victory was Yardley’s solution of Japanese diplomatic codes.]
The war to end all wars had left the world abhorring war. But warship programs begun before or during it continued in Japan, Britain, and the United States. They cost millions. As the postwar recession struck, they squeezed budgets. More and more people wanted to stop them. The war had altered the global constellation of power. So in Great Britain, as the time approached under the Anglo-Japanese pact of 1902 for its reconsideration, sentiment grew for its abandonment. The two countries had originally supported each other against Russia, which menaced both India and the western Pacific, and later against Germany. But both those threats were gone, and Britain had come to worry more about the growing assertiveness of Japan, which challenged Britain’s political and economic position in Asia. It felt increasingly that its interests paralleled those of the United States, despite commercial and naval rivalries.
In March of 1921, the first lord of the Admiralty proposed a face-saving way both to reduce naval expenditures and to dump the Anglo-Japanese treaty: an international conference to settle arms and Pacific questions. The United States eagerly assented. Japan was unhappy. The treaty appeared to screen Japan’s ambitions behind British approval and certified that nation as a great power – the only nonwhite one. And it had just proudly launched the Mutsu, the world’s largest battleship, paid for in part by the pennies of schoolchildren. But it saw no way out. Six other nations were invited – France, Italy, China, Belgium, the Netherlands, and Portugal. The conference was set for November.
Britain wanted it to be held in London, and the incompetent American ambassador at first seconded this idea and then offered Havana. The British foreign secretary later proposed Bar Harbor, Maine, as a venue. Neither Tokyo, nor Paris, nor Rome seem to have been suggested. Eventually, the nations settled on Washington, which would make it the first conference of world powers ever held in the United States. Political considerations decided this; nobody considered that the location would confer an intelligence advantage upon the host. The Congress of Vienna in 1815 had given Austria a great deal of information from betrayers, pilfered documents, and intercepts. Probably France had gained similar benefits during the more recent conference in Versailles. Of course, visiting countries could hire spies, but espionage would be easier for the host, and only America would be able to read the instructions to and reports from each nation's plenipotentiaries. Yardley exploited this advantage as best he could.
President Warren Harding opened the Washington Conference on the Limitation of Armament at 10:30 a.m. on the cold and windy Saturday of 12 November 1921. In the colonnaded Continental Memorial Hall across from the south lawn of the White House, he addressed many of the delegates who, the previous day, had heard him plead with the world to give up war as he solemnly interred America's Unknown Soldier. Then Secretary of State Charles Evans Hughes, dignified, bearded, a former presidential candidate, stepped up to speak. The delegates were expecting a pro-forma welcome. He stunned them. He proposed not only that their governments not build major warships for 10 years, but that they scrap 66 of their capital ships. A British correspondent wrote that Hughes had proposed sinking more tonnage than all the admirals in history together. Hughes began with his own country. Thirty of the scrapped ships would be American. He then compassed the other navies. When he urged canceling the new pride of the Royal Navy, four Hood-class super battle cruisers, Admiral David Beatty, the victor of Jutland and first sea lord, looked “slightly staggered and deeply disturbed.” The first lord of the Admiralty “turned the several colors of the rainbow and behaved as if he were sitting on hot coals.” On the other hand, the Japanese delegation’s impassive faces revealed no emotion but they continued “looking straight ahead” as Hughes’s proposal to reduce Japan’s navy included junking the Mutsu. But at the end of his talk the delegates applauded stormily, and his proposal won the enthusiastic support of the nation.
Under Hughes’ formula, the remaining size of the navies would be based on their existing strength. The United States would have 500,000 tons, the United Kingdom 500,000, and Japan 300,000 -- a ratio of 10:10:6. The two major powers quickly agreed on their parity, but Japan was unhappy. It wanted 350,000 tons. This would make the ratio with the Americans 10:7. Those figures were based not just on national pride but on solid technical reasons, though these seem never to have been mentioned in the discussions. They stemmed from naval experience and the famed gunnery equations of British theoretician Frederick Lanchester.
Naval planners assumed that a fleet lost 10 percent of its effectiveness for each 1,000 miles it sailed from its base. This would be caused by wear and tear, bottom fouling, and enemy attacks. They expected that the U.S. Fleet would steam 5,000 miles from Pearl Harbor to the Philippines. There it would engage the Imperial Japanese Navy, 1,500 miles from Japan. If the Americans sailed with 10 battleships, they would arrive with a strength equivalent to 5. If the Japanese sailed with 6 battleships, they would lose 15% of their power and would arrive with a strength roughly equivalent to 5. The fleets would be of equal strength. If, however, Japan sailed with 7 battleships, it would arrive with a strength of 6, outnumbering the Americans. This advantage was exacerbated by the Lanchester gunnery equations. Lanchester hypothesized that the fleets fired in salvos, that the accuracy of fire was 10%, and that each hit destroyed an enemy gun. Working this out, he concluded that the power of the two forces was not in the ratio of the number of ships but in the ratio of the square of the number of ships. This meant that under the Japanese plan the strength of the two forces would be not just 6:5 but (rounding) 35:25. In other words, instead of only a 20% advantage, the Japanese would have a 44% advantage. The United States could not stand for a 10:7 ratio; Japan insisted upon it. This became the hardest fight of the conference.
Yardley hoped to provide the American negotiators with information about the other nations’ negotiating positions and intentions. He had a good start with Japan. But he did not fare as well with the cryptosystems of the other two major powers. Though Britain had abandoned its two-power standard, in which the Royal Navy would be larger than any two other fleets combined, it could not accept naval inferiority. Yet the United States had called for “a navy second to none.” Foreknowledge of Britain’s intentions would help American diplomats. But Britain was alive to the need for security in its communications. So though the Cipher Bureau solved some codes for trivial messages – such as one urging a Philadelphia Orchestra concert for the delegates – it read no significant cables. The same thing happened with France, which possessed the world’s third largest navy . Work on its cryptosystems, as on Britain’s, started belatedly and without any background, for neither MI-8 nor its successor had studied them. Moreover, French cryptology, with decades of success behind it, was one of the best in the world. So though the Cipher Bureau’s Claus Bogel made a few tentative identifications (as 1272=experts) and discovered that France was using four two-part codes for the conference, it never advanced much beyond that.
But Japan was the main target, and here Yardley had succeeded. The Cipher Bureau had been reading the two main codes, Jo and Jp, since before the opening of the conference. The intercepts reached the Cipher Bureau on 37th Street in bunches of half a dozen to a dozen around six days after transmission, delayed that long because couriers brought them from Washington. At first the typists merely copied them on legal-sized sheets of paper and handed them to Livesey and Ruth Willson for cryptanalysis. Soon two of the typists -- one the beautiful Edna Ramsaier, whose married name was now Hackenburg -- became so proficient and so familiar with the Japanese system that they themselves mentally divided the cryptograms into the two- and four-letter groups and then decoded them almost as fast as they could type. Livesey and Willson recovered unknown groups, translated the Japanese plaintext, and returned this to the typists to be turned into fair copies. These were given to Yardley. Sometimes during the conference the staff worked 24 hours in 12-hour shifts; work until midnight was not uncommon. Once, before Thanksgiving, Hackenburg told Yardley that she could not work on the holiday. Retorted Yardley: “I can't play golf either” -- and she knew she'd have to come in that Thursday.
Most intercepts were solved and translated the day they were received; a few took a day longer; even fewer, more. Originally the solutions in Washington were delivered in weekly batches by 44-year-old State Department officer William L. Hurley, a former newspaperman in his early 40s who had served with the military attaché in London during World War I; a later evaluation gave him a rating of “Fair – as clerk. For secret service work, apparently very good.” Then when the conference began couriers -- among them Tracy Lay, a Foreign Service officer recalled from abroad -- brought them down daily. They reached Washington the day after they were solved and translated. The volume was great. Between half a dozen to a dozen a day arrived, and not a few of the telegrams, such as reports of meetings or of press reaction, ran five or six single-spaced pages, sometimes more. For security, they bore no markings as “For State Department” or “For Military Intelligence Division.” Most went to F.E. – State’s Far Eastern Division—which was deeply involved in the negotiations. Interestingly, few or none went to Hughes. He was too busy meeting with American and foreign delegates, chairing committees, talking to subordinates, studying position papers, and holding daily press conferences to read many of them.
An episode early in the conference convinced officials who may have doubted the veracity of the solutions of their accuracy. On 18 November, Hughes granted an interview to newspaper correspondents on the grounds that the information not be attributed to him. Several days later, the codebreakers furnished State with the Japanese report of the meeting, and the officials could see for themselves that it was right, as well as a clear and fluent translation. As further intercepts came in, American officials chuckled as they read Tokyo's asking whether the American government, then in the depth of Prohibition, would object if liquor were brought to the conference. They observed that the auditor mentality was not confined to the West when they read a Tokyo message beginning, “The expenditure of a year's rent for a building which is to be used for a few months would ordinarily not escape the censure of accountants.” Their moral concern about reading others' messages ebbed when they saw Japan wiring $30,000 for secret service work during the conference. And although, somehow, the head of the Far Eastern Division thought the intercepts were “very difficult to read” and “almost never of real value,” he conceded that they did keep “the members of F.E. currently informed of Japanese feelings.”
But cryptanalytic evidence was not the only information available to the American negotiators. The press, though less authoritative, was quite accurate and sometimes faster, particularly on the key issue of the 10:6-or-10:7 ratio. Thus The New York Times reported on 25 and 26 November that Japan remained resolute on its 10:7 demand. However, on Monday, 28 November, three days after Crown Prince Hirohito had been named regent and was expected to surround himself with moderates, it front-paged a change in Japan's attitude. “Tokio Is Prepared to Yield on Ratio,” the headline ran, over a story that “It is understood that Admiral Kato [Baron Tomosaburo Kato, navy minister and delegation head] received from Tokio today very explicit instructions.... The report was in circulation today that Japan, failing to have her own standard of measurement accepted, would be quite disposed to accept the ‘5-5-3’ arrangement.”
The Cipher Bureau lagged. The next day, it solved a four-day-old Jp cryptogram implying that Japan would not yield. Tokyo was reminding Kato that he himself had once said that “the ratio of 10 to 7 between the American navy and our navy should be the limit. We understand that you will work to maintain this limit without any change.”
But events were outrunning Yardley. While the Cipher Bureau was forwarding this days-old message, The New York Times was front-paging a story that conference opinion believed that Japan’s acquiescence in a 10:6 ratio was only days, if not hours, away. The next day, Wednesday, 30 November, it headlined “Crisis on Naval Ratio Plan Passed; Agreement Likely in Session of Dec. 5.”
Thus evidence that was up-to-date but unverified conflicted with evidence that was authentic but outdated. The American negotiating staff perhaps agonized over the discrepancy as, on Friday, 2 December, Hughes, Kato, and Arthur Balfour, the chief British delegate, discussed the naval ratio. Kato repeated what he had told Balfour the day before: an agreement by America not to fortify Guam and the Philippines and by Japan not to fortify its mandated islands would help in getting Japan to accept the 10:6 ratio. This was the first official hint that the empire might yield. But the meeting ended with Kato saying only that he would ask for further instructions.
That very day, as he was cabling Tokyo, a telegram arrived at the Cipher Bureau. Dated 28 November and consisting of 64 10-letter groups, it began: “Koshi, Washington URGENT 0073 vrxpm dozoorupuh uteletamme fuinofridy ....” It was copied triple-spaced on a legal-sized sheet and its Cipher Bureau serial number, J6204, was penciled in the upper right corner; someone jotted JP above it. A clerk typed out the plaintext between the lines. “ ka too zen ken he ‘ gokuhi ’ kiden kai i 7 4...,” the plaintext began. It was translated later that day. The result showed that Tokyo had softened.
“...we are of your opinion,” the Foreign Ministry was telling Kato, “that it is necessary to avoid any clash with Great Britain and America, particularly America, in regard to the armament limitation question. You will to the utmost maintain a middle attitude and redouble your efforts to carry out our policy. In case of inevitable necessity you will work to establish your second proposal of 10 to 6.5. If, in spite of your utmost efforts, it becomes necessary in view of the situation and in the interests of general policy to fall back on your proposal No. 3 you will endeavor to obtain a wording which will make it clear that we have maintained equilibrium with the American fleet by limiting its power of concentration and maneuver in the Pacific through a guarantee of reducing, or at least maintaining in status quo, the Pacific defenses. No 4 is to be avoided as far as possible.” The codebreakers scrupulously noted that "6.5 [is] reconstructed from a garbled passage.” They did not say what either No. 3 or No. 4 was.
Despite that omission, this solution, when it arrived in Washington on Saturday or Monday, solidified the American delegation's sense that the Japanese would abandon 10:7 and would accept 10:6.
This message, Yardley believed, was "”he most important and far-reaching telegram” that ever passed through the Cipher Bureau's doors. He was right about that. But he failed to look at the press and the negotiations when he added, “It is the first sign of weakness on the ten-to-seven Japanese demands” and he exaggerated when he claimed “This telegram was definitely to determine the respective strength of the fleets of Japan and the United States.” He summarized it correctly:
“It shows that if America presses Japan vigorously, Japan will give up proposal 1, then proposal 2, and that provided the status quo of the Pacific defenses is maintained, she will even accept a ten-to-six naval ratio.
“With this information in its hands, the American Government, if it cared to take advantage of it, could not lose. All it need do was mark time. Stud poker,” he concluded with a metaphor dear to his gambler's heart, “is not a very difficult game after you see your opponent's hole card.” The skeptical head of State's Far Eastern Division concurred. The one intercept that provided help “made it clear that the Japanese would finally give in on the naval ratio .” This, he said, “stiffened Mr. Hughes' attitude.” Thus fortified, Hughes let time work against the Japanese. Intercepts undergirded his strategy. In one message Kato moaned that “...they put pressure on us by arguing that if we do not accept the American plan ..., the whole plan will break down,... It is extraordinarily hard to persist in our proposal.” In another, he warned that some newspaper articles charged that Japan's failure to accept the American proposal would mean that it "would block the success of the conference.” He added that “...even British correspondents in America are as a rule reporting unsympathetically on the attitude of our country regarding the naval question.” And Tokyo itself appeared to be weakening. In a message solved on Saturday, 10 December, the Foreign Ministry confessed that now “many of our own people ... appear to desire the reaching of an immediate agreement through some compromise.” None of this appeared in The New York Times.
Finally, Japan capitulated. Good relations with the United States were more important than the battleship and a half that the greater ratio would have given it – and which it concluded it would not get anyway. At a late afternoon meeting at the State Department on Monday, 12 December, Kato agreed to the 10:6 figure in return for the face-saving agreement he had proposed: that both parties not fortify their possessions in the Pacific. The intelligence provided by the codebreakers had bolstered Hughes's toughness and helped win the day
The Cipher Bureau staffers had worked nights and weekends turning out thousands of intercepts. At Christmas, their government gave them bonuses: a procedure rarely seen among federal employees but not unprecedented among cryptanalysts – those of the absolutist monarchs’ black chambers had been given extra money for solutions. The U.S. Army could pay this money because the Cipher Bureau funds were unvouchered. The bonuses , a little more than a week's salary for each of the 15 persons receiving them, ranged from $37 to $184, with the latter amount going to Yardley. They were accompanied, he said, “by personal regards and assurances that our long hours of drudgery during the Conference were appreciated by those in authority.” And the hard-working typists, codebreakers, and translators had as well the secret satisfaction of having helped their government win a major diplomatic victory and save hundreds of millions of dollars – and the world to enjoy more peace.
�
The story of Yardley’s alleged betrayal was first made public by Ladislas Farago in The Broken Seal (New York: Random House, 1967), 57-58. As a consequence of this report, Fred C. Woodrough Jr., a Japanese linguist who translated messages for American cryptanalysts during World War II, was directed to investigate the matter, using captured Japanese records. He submitted a report on 28 November 1967 to Vice Admiral Rufus L. Taylor, deputy director of central intelligence, agreeing that the betrayal had taken place (photocopy in David Kahn Collection). Both he and Farago base their belief upon a memorandum (in Library of Congress, Microfilm UD Series, Item 52:Reels 29-30:02-03, 05) by the chief of the Japanese Foreign Office telegraph section dated 10 June 1931 – ten days after the publication of The American Black Chamber. The chief states that the Japanese ambassador to the United States, in telegram No. 105 of June 1930 (no day is given), reported that Yardley had sold him a large volume of solved Japanese messages for $7,000. But though at my request two Japanese scholars – Ikuhito Hata, a World War II historian who translated an abridged The Codebreakers into Japanese, and Sadao Asada, a specialist in Japanese naval policy whose Yale University dissertation dealt in large part with the Washington naval conference – have independently searched for telegram 105, both report that it does not exist in the files, though a listing summarizes it as “cryptographic leak.” Hata wrote in a letter of 14 May 2000 to Dr. Edward Drea that the “telegram itself cannot be found.” Asada examined the 600-plus-page Yardley file in the Foreign Ministry archives but said, in a letter of 27 June 1998, that “it contains no world-shaking new discoveries,” deals mainly with “the Foreign Ministry’s reaction to Yardley’s book,” and includes no materials prior to 1 June 1931, the book’s publication date. In a letter of 10 July 1998, he wrote, “The fact that I could not find the dispatch No. 105 or pre-June 1931 telegraphs on the Yardley incident in the Foreign Ministry archives can mean that they were destroyed, given the very delicate nature of the subject. My conclusion is that Japanese archives do not substantiate the story of Yardley’s betrayal, although the Japanese Foreign Ministry leaders believed in it for one reason or another.” Moreover, the files contain no internal memoranda about the proposal (Is it a trick? If it is legitimate, should the decrypts be bought? How much should we pay?), no payment vouchers, and – most significantly -- no documents from Yardley. These would have existed if the deal had gone through.The charge of betrayal rests upon a post-American Black Chamber allegation for which great motivation but no evidence exists. Consequently, I believe that Yardley never sold any documents to the Japanese and that the story was fabricated to denigrate him and save Japanese face. Woodrough says that this hypothesis is “super-intricate” and “outsmarting itself,” but whatever he means, I think that it is simple, coherent, and reasonable. I do not know whether the summary listing was contemporary and without this knowledge cannot agree that it supports the accusation. A revival of the betrayal charge, repeated uncritically in Rhodri Jeffreys-Jones, Cloak and Dollar: A History of American Secret Intelligence (New Haven: Yale University Press, 2002), 110, was disposed of by Louis Kruh in a letter to the editor, Cryptologia, 19 (October 1995), 377-379.
#l
Copyright © 2008 David Kahn. All rights reserved.