Code Girls, page 10
“If the Japanese Navy had changed the code-book along with the cipher keys on 1 December 1941, there is no telling how badly the war in the Pacific would have gone,” said Laurance Safford.
As crushing as Pearl Harbor was, it was thanks in large part to Driscoll’s decades-long detective work—and to the example Elizebeth Friedman set for other women—that America did not enter the Second World War quite as blind as it might have seemed.
CHAPTER THREE
The Most Difficult Problem
September 1940
Poland was occupied and had been for a year. Czechoslovakia had put up little or no resistance to being partitioned. The Nazi war machine had overrun Norway and Denmark, defeated Belgium and so many others, and proceeded to march into Paris, where Nazi officers were drinking café au lait and popping champagne corks in the finest restaurants on the Champs-Élysées. England was holding out, but barely, pummeled by German air raids and braced for an anticipated sea invasion launched from the shores of occupied France. Japan was shouldering its way through China and around the Pacific, chasing resources and seeking to establish a “new order” in which Asian nations would be rescued from Western domination and dominated, instead, by Japan. And here—in crowded U.S. Army offices in downtown Washington—a young civilian woman was patiently standing, waiting for a group of men to stop talking and notice her. She had something urgent she needed to tell them, but, shy and reluctant to interrupt, she waited for a pause.
The office was nothing to brag about—just a few rooms tucked away in a Washington eyesore known as the Munitions Building, erected in 1918 as a headquarters for the War Department. The Munitions Building and the U.S. Navy headquarters were side by side, as it happened, both constructed as “temporary” wartime structures during World War I and both still in service even now that the Great War was long over, together dominating the part of the capital city between Foggy Bottom and the National Mall. The twin buildings had concrete facades and a series of thin wings that stretched backward, perpendicular to the facade. Working in them was like working in a multistory warehouse.
On the first floor, the wings of the buildings had long narrow corridors and doors through which bicycle messengers often burst, so people had to be careful not to be knocked down by bag-wielding boys on bikes. On upper floors, wooden desks lined open workrooms, and large windows admitted the Washington air, for better or worse, depending on the season and humidity level. Technically the whole edifice was known as the Main Navy and Munitions Building. Washington’s physical wartime expansion had not—yet—been effected; the Pentagon did not yet exist, nor had Arlington Hall been requisitioned, and in 1940 these two squat ugly buildings housed, in effect, the country’s entire military brain trust.
On an upper floor, occupying the back of one wing of the Munitions Building, a handful of rooms had been allocated to the U.S. Army’s code-breaking operations, which had grown considerably in the past year but were still modest enough to be contained in such a small space. One room held a bank of office machines used for tasks like sorting and collating; others held locked file cabinets with intercepted messages. In most of the others, people sat quietly at tables, heads down, sometimes smoking or drinking coffee, working with pencils on lined or crosshatched paper. Apart from the machines, the offices consisted of the usual government-issue assortment of scuffed desks, battered cabinets, and rotary phones. Somewhat by chance, the U.S. Navy’s code-breaking force was located in an adjacent wing. While capable of cooperation, the rival code-breaking units were still marked by infighting, paranoia, and personality clashes. Unlike the Navy’s code breakers, the Army team was mostly civilian, a hodgepodge of mathematicians, ex-schoolteachers, linguists, and clerical workers. While the Navy tackled Japanese naval ciphers, the Army code breakers were attempting to penetrate systems used by military and diplomatic officials in Italy, Germany, Japan, and Mexico.
It was clear enough that sooner or later the United States would formally enter what was shaping up to be a second world war, and on this Friday afternoon, in the heat of late summer, the atmosphere in the rooms was thick with urgency.
At the center of the Army’s operation was William Friedman. Originally hired to ensure that the U.S. Army developed codes that were more sophisticated and secure than wigwag, Friedman had learned to break codes better than almost anybody in the world. He had hired most of the people in this office. The people laboring at the tables revered him. Friedman’s Army superiors sometimes called him Bill, but the people who worked for him always called him Mr. Friedman. Sometimes in private they called him “Uncle Willie,” but none would have dared to do so in his presence. He was sensitive, easily offended. Meticulous in his work habits, he was good about entrusting important jobs to capable others but did not give compliments easily. An excellent tennis player and ballroom dancer, he had a thin mustache and liked to wear a bow tie and two-tone shoes, and he was fanatic about precision; he hated it when people used “repeat” as a noun and insisted they say “repetition.”
Friedman, now in his late forties, was a legend among the still-small global community of people involved in the making and breaking of codes and ciphers. After leaving Riverbank he had assembled one of the few known libraries on the topic. The literature included Cours de Cryptographie by General Marcel Givierge of France; Manuale di Crittografia by General Luigi Sacco of Italy; and Elements of Cryptography by a French captain, Roger Baudouin, that had been smuggled out of France just before Paris fell. Friedman had written many of the most important treatises himself—top secret monographs including Elements of Cryptanalysis, The Principles of Indirect Symmetry of Position in Secondary Alphabets and Their Application in the Solution of Polyalphabetic Substitution Ciphers, and his masterwork: The Index of Coincidence and Its Applications in Cryptanalysis. He also had written training manuals that were treated as kinds of bibles.
Over the course of a decade Friedman had assembled a tiny band of acolytes. In 1930, shortly after the abrupt closure of Herbert Yardley’s office, his bosses had given him funds to hire three young mathematicians: Frank Rowlett, a southerner who was teaching in Rocky Mount, Virginia, best known for its production of moonshine; and Abraham Sinkov and Solomon Kullback, friends who had attended high school and City College of New York together. Friedman wanted his staff to be young, because he knew it would take them years to master their discipline. Together with John Hurt, a Virginian who knew Japanese and could translate deciphered Japanese messages into English, the men had spent nearly a decade studying Friedman’s methods of “attacking” codes and ciphers. Funding was never abundant, and in the worst of the Depression, they sometimes had to supply their own penny pencils and bring in scrap paper from home.
As the staff expanded, Friedman had done something else: He had begun hiring women. There were several reasons for his willingness to do so. One was plain availability. In the 1930s—well before the war started—Roosevelt’s New Deal had begun drawing women workers to Washington, where the expanding federal government proved more of an equal opportunity employer than the private sector. Discrimination existed in government hiring, to be sure, but for a woman, the advantage of applying for a federal job was that it entailed taking a standard civil service examination. Women took the same exam men did. Federal agencies were given access to test scores and made job offers accordingly. The 1920 census found that nearly 40 percent of employed people in Washington were female.
Equally important was that William Friedman was the sort of man who liked working with intelligent women, as evidenced by his own marriage to Elizebeth, now employed at the Coast Guard as both a code breaker and what might today be called a communications security consultant. At the moment, the Coast Guard’s mission was enforcing “neutrality,” which meant that Elizebeth’s unit was deluged with message traffic from all sorts of ships plying the Atlantic. A valuable utility player for other agencies as well, Elizebeth also designed the code-making unit for the Coordinator of Information—the nation’s new spy service—soon to be renamed the Office of Strategic Services.
Elizebeth’s example was even more valuable to her husband than William Friedman knew. In October 1939, following the outbreak of war in Europe, the Army had given him the funds to further enlarge his cryptanalytic staff. One of these early hires was a woman named Wilma Berryman, who had been attracted to the field thanks to Elizebeth’s renown. Berryman hailed from Beech Bottom, West Virginia, and graduated with a math degree from Bethany College. Though she had been trained to teach high school math, during the Depression the only job she could find was teaching first grade to a classroom of forty-five children. When her husband took a job in Washington, Berryman went to work in the payroll office of the department store Woodward & Lothrop—near the pocketbook shelves—then at the Census Bureau, and at a succession of other agencies. But when she read in the Washington Evening Star about Elizebeth Friedman’s exploits (the article also mentioned William, but it was the wife’s example that struck her) it awakened something in her. She began to envision another future.
Asking around, Wilma Berryman found out that the U.S. Navy had developed a correspondence course, to train its own officers but also to permit hobbyists and other potential civilian applicants to homeschool themselves in breaking codes. The purpose was not so much to teach the subject as to locate talent and winnow out those who had none. Wilma Berryman spent several years tackling the course, sending away for lessons, completing the exercises on her own time, and sending them back. The two code-breaking operations were porous enough that her answers found their way to William Friedman, who was always scrutinizing civil service rolls and whatever other sources he could gain access to, evaluating test scores and looking for the right abilities. Upon being hired, Wilma Berryman was put on the Italian desk—which is to say, she was given a beginner’s textbook on the Italian language and plunged willy-nilly into the secret communications of the fascist government of Italy. Every morning, one of Berryman’s colleagues liked to come over to her table and ask her, “How’s Benito doing this morning?”
That was how the Friedman operation worked: It was a teach-yourself kind of place. Newcomers would spend the morning studying training manuals, puzzling out the answers to questions like “What four things were thought by Captain Hitt to be essential to cryptanalytic success?” (perseverance, careful methods of analysis, intuition, and luck) and “What two places in every message lend themselves more readily to successful attack by the assumption of words than do any other places?” (the beginning and the end). They spent afternoons attacking actual codes.
Wilma Berryman loved it.
So did Delia Ann Taylor, a tall, brainy midwesterner who graduated from Sweet Briar College in Virginia and had a master’s degree from Smith College. Working near her was Mary Louise Prather, the daughter of a genteel family that had fallen on hard times. It was Prather’s job to work the office machines—sorter, reproducer, tabulator, keypunch—that Friedman, a master of winkling things out of a closefisted government bureaucracy, had persuaded his bosses to buy him. While Prather’s might seem like a menial job—running office machines was considered women’s work—these were not conventional machines but had been modified to assist with sorting enemy messages.
Prather also filed the intercepts, which themselves were a kind of contraband. Since the United States was not, strictly speaking, at war, collecting radio and cable transmissions of foreign diplomats was not, strictly speaking, legal. The Communications Act of 1934 imposed “severe penalties for interception of diplomatic traffic,” as one memo noted, but the code breakers had decided to overlook that. Friedman’s Army superior, the now major general Joseph Mauborgne, felt the law could be ignored. Even so, intercepts were hard to come by: The Army did not yet have many clandestine radio intercept stations of its own, so they got some messages from the Navy and some from friendly cable companies who handed them over under the table. Prather kept a careful log of every last one.
And there was twenty-seven-year-old Genevieve Marie Grotjan, hired as a “junior cryptanalyst” in October 1939 for a salary of $2,000 per year. It was Grotjan who was standing waiting for the men in the Munitions Building to notice her. A native of Buffalo, New York, Grotjan had been a brilliant all-around student at Buffalo’s Bennett High School, where she delivered the salutatorian’s address in the customary Latin. She received a Regents scholarship to attend the University of Buffalo, where she majored in math and belonged to the International Relations Club. Graduating summa cum laude in 1938, she won a math prize, received a teaching assistantship to do graduate work, and aspired to teach college math. Like so many women of her day, however, Grotjan was unable to find a university math department willing to hire her. So she came to Washington and was hired as a statistical clerk at an obscure agency called the Railroad Retirement Board, where it was her happy task—she enjoyed it—to calculate pensions. When she took a math exam to secure a routine pay raise, her score attracted Friedman’s attention. She received a call from the Signal Intelligence Service and was asked if she would like a job in the “code section.” Grotjan didn’t know what any of that meant, but she said yes.
Many of the code breakers were social with one another, but Grotjan was not one of these. Shy and introverted, she favored rimless eyeglasses, high-collared blouses, and a pragmatic hairstyle that consisted of tight blond pin curls crimped in a halo around her forehead. She rented a room in a boardinghouse at 1439 Euclid Street, a modest section of northwest Washington.
After less than a year on the job, however, Grotjan was shaping up to be one of the team’s most promising code breakers. She was known for her thoroughness, powers of observation, and attention to detail. Humble and reticent, she possessed the pure soul of one who lives for numbers, oblivious to office politics and rivalries, which did exist. Based on the aptitude she demonstrated, she was assigned to the most pressing problem Friedman’s office had undertaken: the cipher system used by Japanese diplomats around the world. It was a completely different system from that of their military counterparts: While the Imperial Japanese Navy often used laborious pen-and-paper systems, which involved a lot of adding and figuring, Japanese diplomats favored the newer machine-generated ciphers. The small team that Grotjan belonged to was trying to do something that almost certainly had never been done: reconstruct an unknown machine without having seen it or even a piece of it—not so much as a blueprint or drawing. They were attempting to penetrate the machine’s inner workings by scrutinizing its pilfered output, sitting at tables looking at strings of random-seeming letters.
There were many challenges to this task, chief among them that the Japanese machine had been produced in an environment to which they did not have access. This was an era when governments and businesses alike were turning to cipher machines to keep their messages secure from intermediary parties (anybody from a Morse operator to an actual spy) and inventors were always designing new machines to enable them to do that. Friedman’s office kept its own “nut file,” recording the outlandish systems that hobbyists tried to sell them. Usually, the inventor wanted either a job or a million dollars and threatened to sell to the Russians or Germany if the U.S. government didn’t bite. Friedman, like his Navy rival Agnes Driscoll, was a master at finding the weaknesses of these machines, and his acolytes often could break a nut’s systems in a matter of hours.
But some machines on the Western market were first-rate. One of the best was called Enigma. Envisioned as a tool for bankers, the Enigma, invented by a German engineer and marketed by a German company in the 1920s, had been adapted for military use by the Nazis. In 1933, Hitler ordered it taken off the commercial market, however, so his military could have sole access. Many military machines, like Enigma, were small and light and sturdy, not much bigger than typewriters. Enigma in particular was a durable, portable, battery-powered device that could be lugged around and used during battle, or welded to the command center of a submarine, where it had one job and one job only: to change each letter of a message to a different letter.
This new Japanese machine had the same letter-changing mission, but nobody knew how it worked. No Westerner had laid eyes on it, or even a facsimile or prototype. The machine was not as mobile as the Enigma—not mobile at all. Unlike Enigma, it ran on electrical current and needed an outlet it could be plugged into. Only the most important Japanese embassies were given access to it—those in Washington, Berlin, London, Paris, Moscow, Rome, Geneva, Brussels, Peking, and a few other major cities.
A machine ciphering system worked well for diplomats. In the 1930s, phone calls were costly and vulnerable to being tapped. The foreign office in Tokyo often needed to send the same message to all its far-flung ambassadors, and rather than pick up a phone and make the same call, over and over, it was easier to craft a message and hand it to a clerk, who would write it out in Romaji, a phonetic version of Japanese that used roman letters to spell out the syllables, as in ma-ru, for merchant ship. The clerk would run “maru” through the machine, producing a new stream of letters—say, “biyo”—that could be cabled. The enciphering mechanism could be set in different positions, according to a key, or setting. The machine could be used in either enciphering or deciphering mode, so the diplomats could use their own machines to restore the message to its original meaning. They also could use it to write back to Tokyo.
The Japanese diplomats were, of course, discussing their country’s war plans. They also were meeting with Hitler, Mussolini, and other key Axis leaders. If the Americans could uncover the machine’s workings, they would have access to a priceless stream of insight, gossip, and strategy, involving not only Japanese intentions but those of every tyrant in Europe.
