Combined Fleet Decoded: The Secret History of American Intelligence and the Japanese Navy in World War II

Chapter One

"Your Message, Affirmative"

Fed Rogers sat, annoyed, at the foot of the table. Lunch had not been served and, as mess treasurer for officers of the U.S. battleship Virginia, Rogers was responsible. Rogers called the steward and addressed him in a foreign tongue, demanding that the meal be produced. The steward happened to be a Japanese national, the language of their conversation Japanese. Their exchange was possible only because Lieutenant Fred F. Rogers was a Japanese-language officer, the only one in the whole U.S. Navy that fall of 1913. At one time there had been another officer fluent in Japanese but that man had resigned, and the Navy had not seen fit to replace him.

    Witness to the exchange in the Virginia wardroom was Ellis Zacharias, then a twenty-three-year-old ensign on his first extended-duty assignment. Fascinated with the idea of learning Japanese, and with the whole concept of a language officer, Zacharias became close to Rogers, thriving on his tales of life in Japan and of the difficulty of mastering the language.

    In 1917 Zacharias moved on to the engineering department of a survey ship, and in World War I he served as engineering officer of a light cruiser, then as gunnery officer of a bigger cruiser in the Atlantic. After the war Zacharias taught engineering at Annapolis; he accompanied the midshipmen on their summer cruise, which had reached Honolulu when Zach, as he was called, received a cable that changed his life.

    Long before, Zach had asked Fred Rogers how to go aboutbecoming a Japanese-language officer. The word had been that there was no chance; the Navy had no intention of expanding its cadre of Japanese linguists. Now, in 1920, the Navy planned to send two new men to Japan for language training—and Rogers was in Washington, on the Japan desk of the Office of Naval Intelligence, which made the personnel selections. Commander Rogers's cable to Zacharias asked if he was still interested in going to Japan as a language officer.

    "Your message, affirmative," Ellis Zacharias replied.

    So began an oriental adventure for the thirty-year-old Floridian Zacharias, an experience repeated by sixty-five other young Americans over the next two decades. Officially the language officers were assigned to the naval attaché's office at the U.S. embassy, but except in special instances like that of the Tokyo earthquake, their duty focused simply upon learning Japanese. Zacharias himself rose to become deputy director of naval intelligence and it was he, under the press of war requirements, who established a regular routine for the language officers. During those earlier years things were very much more catch-as-catch-can.

    The informality of the programs did not prevent learning, however. The accounts of Zacharias himself and other early language officers such as Arthur McCollum, Henri Smith-Hutton, and William J. Sebald describe a remarkably similar experience. Many roomed in the same houses, used the same teachers, followed the same program. Periodically they submitted to language examinations given by State Department officials. Through much of the interwar period the diplomat administering the tests was Eugene H. Dooman, who by 1941 had risen to embassy counselor, a rank that made him one of the most senior American officials dealing with Far Eastern affairs.

    In the beginning there were very few students. Zacharias and Hartwell C. Davis became the first of the new breed. Until 1927 the Navy sent over one more man each year; then it assigned three language officers annually. Gradually the program evolved into a three-year course with examinations every six months. The slow pace of assignments was partly accounted for by the lack of need for numbers of Japanese linguists during the early years, but also by the Navy's limited budgets and difficulty finding good candidates. By 1927 policy was to have some officers in training at all times. The pace increased with the perception of a Far Eastern crisis after the Manchurian Incident of 1931, and 1932-1935 became the peak for language-officer assignments, although as many as eight men in various stages of learning would be sent home when hostilities clearly impended.

    "Quaint" might be the best adjective for the life of officers in the early language program. Bill Sebald, for example, got interested in becoming a Japanese linguist while serving aboard the battleship Texas at Guantánamo Bay, where he met Ellis Zacharias and felt the pull of foreign places. A friend at ONI put in a good word, and Sebald got orders to Japan in the summer of 1925. He roomed with three other linguists in a house in Karuizawa. David Wells Roberts, whose mother had gone to Japan with him and kept up the house at Karuizawa, helped show Sebald around. The house was rented from Dr. Benninghoff, a missionary teacher at Waseda University. Arthur McCollum ended up marrying Benninghoft's daughter; both had been born in Japan. Sebald met his future wife in Japan too. Officers would arrive in the country, be presented to Japanese officials in dress uniform with cocked hats—the naval equivalent of tie and tails—then venture all over the country, surfacing occasionally for a diplomatic function or a work assignment, One student who arrived in 1926 was Henri ("Hank") Smith-Hutton, whose rapid progress astounded Dooman and others, Sebald, Smith-Hutton, and McCollum, who roomed for a time with Eugene Dooman, have all left oral histories describing their training.

    Each officer had a fund from which he employed tutors, usually two, and bought texts. The most useful Japanese grammars were one in English by a German named Lange, and another by the Englishman Basil Hall Chamberlain, professor of philosophy at Imperial University, Most popular tutor was a Professor Naganuma who, Sebald tells us, preferred the "direct" method, in which students memorized so-called "type" sentences. Naganuma illustrated his points by repeating the sentences and he and the students, working two hours a day at their homes, would sometimes go over the same ones for days on end. Naganuma would begin with conversation, gradually ease the language students into grammar, and start them reading the katakana and hiragana written characters. Written Japanese using the romaji system with Western-style roman letters would be introduced only after the students were familiar with the native Japanese styles. Professor Naganuma's methods proved so successful that they were copied by Berlitz, and he was in such demand that he was obliged to hire several assistants. Naganuma ended up teaching Army language officers and diplomats too, and worked for the Americans right through 1941. After the war he opened a language school for Americans from the occupation forces. Naganuma died in 1973, still revered by many former students.

    There were several teachers in addition to Naganuma. William Sebald recalls as many as six Japanese who made their living by teaching for the United States Navy. At times Sebald worked with three tutors at once, though he felt that two was an ideal number and one instructor, if good, could be ample. Sebald worked on his Japanese six days a week and thought he was doing well learning as many as ten written characters in a day. He and the other students had allowances of $50 a month to pay their tutors.

    In addition to learning the language, indeed precisely because they were learning it, the students received occasional assignments from the American naval attaché at the embassy. One early job was to translate navigational instructions in the Japan Pilot, a project undertaken by Bill Sebald and Hank Smith-Hutton. Another typical work assignment, handed out in March 1927, called for papers reviewing evidence on various subjects. That time Sebald was to study the Japanese shipbuilding industry, and Smith-Hutton the chemical industry; Franz B. Melandy was to compare the gun power of the U.S. and Japanese battle fleets, and David Roberts to review the Japanese steel industry. The officers worked from naval attaché files, military attaché reports, trade journals, Japanese government releases, consulate-general files, and conversations with knowledgeable individuals. These sources, in fact, were the basic stock-in-trade for the naval attachés on all subjects all the time.

    The essential value of the language-officer program was to provide both naval attachés and the Office of Naval Intelligence with a pool of trained linguists who could use original source materials in a difficult language. As naval officers these men knew the technical meanings and importance of things; as linguists they learned of corresponding things Japanese. Washington never imagined how important its language-officer program would turn out to be.

It should not be surprising that in the early 1920s the United States saw only limited need for Japanese-language officers. In the late twentieth century it is difficult to recollect the mood of that era; almost no one is now left who was a direct observer; the events of subsequent years have crowded in to such an extent that those days seem shrouded in the mists of time. To the degree that we recall anything, it is the social history of the 1920s that commands attention—the "Roaring Twenties," a brassy, beaming era when anything seemed possible and everything appeared to be in place. In fact, international relations followed broader trends, and the world, which had just put the global 1914-1918 war behind it, dared to dream of universal peace. In international relations the 1920s became the decade of accommodation, of the League of Nations, of the Kellogg-Briand Pact outlawing war: the newsreel films with their jerky footage captured the images of diplomats shaking hands over the latest diplomatic agreement. Above all the 1920s were the age of naval arms control, and that fact was of specific importance to both the United States and Japan.

    In the period prior to World War I there had been competition in naval armaments as Great Powers each strove to acquire the strongest, most technologically advanced battle fleet. So fierce became competition for navies that people spoke of a "race." The experience brought to language the terms "naval race" and "arms race." The Anglo-German race to build dreadnought battleships is the best-remembered aspect of this competition, but France, Italy, the United States, and Japan were also participants. After the war, when Germany no longer figured as a significant naval power, the stage seemed set for a race between the United States and Great Britain, while Japan also engaged in a naval building program that put it in a strong third-place position. Not only was this competition for armaments costly, but analysts increasingly agreed that the Anglo-German naval race had figured as one of the direct causes of World War I. It seemed worthwhile to head off similar postwar competition.

    Complex political factors also helped create conditions for accommodation. Most important was the revulsion everywhere after the incredible carnage of World War I. Many felt it necessary to avoid future war at all costs. and if naval racing was leading to war, then such competition had to be stopped. In the United States the administration of Warren G. Harding found it politically expedient to host a grand negotiation, capitalizing on the rise of isolationist sentiment in this country. There were also outstanding differences with Japan over, for example, Japanese encroachments in China during the war years, which had created a perceived threat to the American policy of an "Open Door" in China. Political leaders in Japan saw a grand accommodation in the Pacific, which included naval arms limits, as a way to restrain the political power and continental ambitions of the Japanese Army. Japanese Navy commanders felt naval limits might be acceptable to prevent the Army from increasing its influence. In Britain political factors also applied. In all three countries many considered the cost of a naval arms race prohibitive; this created powerful inducements to accommodation.

    As a result of these varied pressures, in the summer of 1921 the United States proposed an international conference in Washington to consider arms limitations plus Far Eastern issues. Following some delay the grand negotiation opened in November. On the second day of the conference American secretary of state Charles Evans Hughes made public a plan that became the basis for agreement. The plan acquired immense popularity because it provided for wholesale scrapping of major warships: battleships, considered the most important combatant vessels, or "capital ships," of that day. As finally agreed upon, the "battleship burning" encompassed some sixty-eight warships aggregating 1,861,643 tons.

    The American proposal also introduced the concept of naval "ratios," by which the tonnage of each naval power's capital ships would be set at a fixed proportion of the tonnage of the strongest ones'. There would be specific rules governing when ships could be replaced and what characteristics new battleships could have.

    The powers clearly anticipated a proposal embodying some of these elements. In Tokyo, for example, authorities delayed accepting their invitation to the Washington conference while an Imperial Navy expert group compiled technical studies comparing aggregate tonnage of Japanese warships in service and building with those other navies had or planned. In 1907 the Japanese Navy designated the United States its hypothetical enemy for budget purposes. From that time Japanese naval thinking considered that the fleet needed to equal 70 percent of the U.S. Navy to be adequate for national security. But the expert group formed to consider arms limits in advance of the Washington Arms Limitation Conference discovered that the Imperial Navy had never attained that strength. The figures showed that at the end of 1921 Japanese tonnage would be just 52 percent of American. By 1925, when the Americans completed a massive building program, Japan would have increased its naval tonnage only slightly. In 1927, when the Imperial Navy finished building its own program of eight battleships and eight battle cruisers, the fleet would still aggregate 200,000 tons less than the American.

    Thus technical studies in Tokyo suggested that Japan could not meet its minimum standard for national security against the United States, much less build a fleet equal to the American. That realization implied a need to prevent the United States from completing its own program. Far from a luxury, naval arms limitation suddenly seemed imperative. Similar calculations were made in Washington and in London.

    In his speech to assembled delegates at the Washington Arms Limitation Conference, Secretary of State Hughes set down quotas the powers might be allowed: For every five British battleships there could be five American and three Japanese—or, stated as a ratio, 5:5:3. The Japanese held out for their 70 percent formula, which meant a ratio of 5:5:3.5, usually rendered as 10:10:7. Western powers resisted this concession.

    About two weeks into the conference the Americans enjoyed a demonstration of good intelligence work. It happened that the War Department had a small codebreaking unit under Herbert O. Yardley, which had labored on the Japanese codes for a long time. By the summer of 1921 Yardley's so-called Black Chamber had mastered the Japanese diplomatic code and was reading Tokyo's communications with its Washington embassy. During the conference the Black Chamber read the Japanese cables instructing its delegation. On November 28 Tokyo informed delegates that the Foreign Ministry agreed with their advice to avoid any open break with the British and Americans on naval limits even if this meant making concessions on strength. Yardley's Black Chamber decoded this cable, which revealed for the first time that Tokyo might abandon its insistence on a 10:7 ratio. The message mentioned ratios of 10:6.5 and 10:6 as possible fallback positions, the latter in conjunction with agreement not to fortify islands in the Pacific Ocean and to maintain the status quo in other respects.

    There would be long negotiations before details were arranged, but the basic compromise that created the Washington treaty system was prefigured in that cable. The Japanese insisted on some changes, such as substituting the battleship Mutsu, completed and commissioned during the conference and partly paid for by fund-raising among Japanese schoolchildren, for another vessel Japan was supposed to retain under the agreement. Japan opposed the British desire to outlaw submarines in the agreement, while several powers got the British to give up their effort to restrict battleships to no more than 30,000 tons. The final naval treaty, signed in Washington on February 6, 1922, permitted battleships to be replaced after twenty years, replacement vessels to be not over 35,000 tons' displacement and to carry no guns of larger than 16-inch caliber. There was also a Four Power Treaty, signed December 13, 1921, intended to preserve peace in the Pacific, and a Nine Power Treaty, also signed in February 1922, concerning extraterritorial rights and duties in China.

    By February, in support of the U.S. negotiators at the conference, Herbert Yardley's Black Chamber deciphered more than 5,000 foreign communications. Yardley and most of his people were wrecks from nervous exhaustion. The boss took to bed and was then ordered to Arizona by anxious doctors. In June, when Yardley returned to Washington, it seemed a new world, one in which naval arms competition had been banished and the Pacific subjected to a security regime. Small wonder the Navy did not see great need for Japanese-language officers.

    Unfortunately, the seeds that led to a far different situation were already in the wind. One Japanese concern not dealt with at the conference had been desire to redress discrimination in America against Japanese citizens and persons of Japanese descent. Not only were no measures taken to address these concerns, but court cases in the mid-1920s and an immigration law Congress passed in 1924 greatly inflamed the situation. California discrimination laws were held constitutional, Japanese nationals were denied American citizenship. These developments confirmed Japanese fears that their country was not going to be accepted into the club of developed nations.

    Narrow military considerations added to prospects for rising tension. Regardless of Japan's lack of resources, money, and shipbuilding capacity to compete in a naval race, many Imperial Navy officers considered agreement on a 5:5:3 ratio an excessive, even unpatriotic act. Nor was this feeling confined to the officer corps; a hint of the state of opinion among Japanese seamen lies in the sudden popularity of a brand of cigarettes called "5:5:5," the name of which was held to signify the yearning not just for a higher relative ratio but for full equality. More seriously, Admiral Kato Kanji, chief of the naval expert group at the Washington Conference, a man who had argued against defining Japanese security in terms of any particular ratio, returned from Washington to voice strong complaints. Around Kato coalesced a cohort of like-minded officers eventually called the fleet faction. Those advocating arms limitation became known as the treaty faction. This cleavage within the officer corps became a central feature of Japanese naval politics for the next two decades.

(Continues...)

---

Excerpted from COMBINED FLEET DECODED by John Prados. Copyright © 1995 by John Prados. Excerpted by permission. All rights reserved.