The Woman Who Smashed Codes

by Jason Fagone

  She had never been the boss of men before and she worried that her new employees might not accept her authority as a woman, but this concern proved unwarranted, with one exception, the all-time top scorer on the math exam, a young mathematics Ph.D. from Columbia University who was supposed to be a prodigy. He refused to complete Elizebeth’s exercises and babbled nonsensically about an “indecipherable cipher”; she decided “he did not comprehend the English language” and replaced him with a candidate who showed a more practical cast of mind. Elizebeth’s first two successful hires were Hyman Hurwitz, a twenty-one-year-old electrical engineer from Dorchester, Massachusetts, who liked to tinker with radios, and thirty-one-year-old Vernon Cooley, from Kalamazoo, Michigan, who had taught schoolchildren there and also worked for a time in the factory of the Kalamazoo Paper Company. Both were given the title of Assistant Cryptographic Clerk. A third young codebreaker-in-training soon joined them, Robert Gordon, who was twenty-three and hailed from Waco, Texas.

  The three men, “able, agreeable, and cooperative,” treated Elizebeth with respect, and soon they all settled into a productive routine, learning each other’s quirks and strengths, dividing up tasks as a team. Asking if Elizebeth experienced sexism was like asking if Marie Curie did. Cryptology was a young field. It hadn’t yet sorted itself into rigid roles by gender. And Elizebeth was exceptional. She deployed overwhelming mental firepower against the pull of gravity. She was Cryptanalyst-in-Charge.

  By the end of 1932, when Americans elected Franklin Delano Roosevelt to his first term as president, Elizebeth’s team at the coast guard was pound for pound the best radio intelligence organization in America. They knew how to extract information from clandestine radio networks, map the hidden structure of the transmitters, and hunt the people using them. This set of skills would later make Elizebeth an important figure in the quest to destroy the clandestine networks of Nazi spies. Her fight against smugglers was like target practice for the coming fight against fascism.

  Americans weren’t yet afraid of Nazis. They were too worried about scraping up money for food. By 1933 the Great Depression had put 15 million people out of work. Fabyan told William in a letter that the mayor of Aurora, Illinois, had shut down the town for five days to stop runs on the bank. “The world is a mess, and everything is going topsy-turvy,” Fabyan said.

  Elizebeth took her daughter to hear FDR’s inaugural speech on March 4, 1933, a viciously cold morning. They walked to the Capitol from their house. “The only thing we have to fear is fear itself,” FDR said, focusing on his plans to restart the economy. He did not mention Adolf Hitler or the Nazi Party, who had taken power that January, deploying mobs of men in brown uniforms and swastika armbands to crush dissent. The international press covered him like a normal leader. Many Germans did not think he would really do the things he had said he would do.

  After FDR’s speech people stayed outside in the freezing wind, cheeks pink as raw beef, waiting to see the parade. Elizebeth and Barbara handed out League of Women Voters fliers to the spectators. Elizebeth was pleased to see her daughter playing the role of “ardent worker.”

  Eighteen days later, in Germany, in a vacant gunpowder factory northwest of Munich, the Nazis opened the first concentration camp, Dachau. The occasion was announced by Heinrich Himmler, Reichsführer of the SS, at a press conference.

  One month after that in Washington, at the end of April 1933, Elizebeth prepared to leave the city on her latest assignment for the T-men. “I pack my bag,” she wrote, “and hug my children a good-by which is to last for a week or a month or longer, I know not, and board a train with a prayer that the new fields will be not impossible of conquest.” She kissed William. He asked her to be safe and not take unnecessary risks. If she had known what awaited her in New Orleans, she might never have gone.

  “Please state your name,” America’s top Prohibition official said. He had solemn eyes and spoke in a slow drawl.

  “Elizebeth Smith Friedman.”

  “What is your occupation?”

  “I am a cryptanalyst.”

  “And what are the duties of a cryptanalyst?”

  “A cryptanalyst is a person who analyzes and reads secret communications without the knowledge of the system used.”

  It was May 2, 1933, and Elizebeth was speaking in a witness box in a federal courtroom, about to play her part in a giant conspiracy case against twenty-three suspected agents of the syndicate she had been tracking for years, the Consolidated Exporters Corporation. They had recently expanded into the Gulf of Mexico, directing a fleet of eight rum-running ships from a pirate radio station in New Orleans. T-men intercepted at least thirty-two coded radio messages and mailed them to Elizebeth, and her solutions exposed the ribs of the scheme: the names of the rum ships (Concord, Corozal, Fisher Lassie, Rosita, Mavis Barbara); their system of sneaking crates of liquor into lonely bayou towns on small boats called luggers and then unloading the crates onto freight trains, covered in sawdust. The government considered it “the greatest rum-running conspiracy since Prohibition,” and now Elizebeth had been summoned to this federal courtroom in New Orleans to explain her methods to a judge and jury.

  She wore a pink dress and a hat with a flower pinned to its brim. Directly in front of her was a stack of yellow papers with her message solutions. She looked out at the courtroom, its wooden pews crammed with spectators and journalists. The defendants sat together like a sports team on a sideline, in suits; the previous day in court they had been quietly switching seats to make it more difficult for witnesses to identify any one of them. The accused ringleader was Albert Morrison, a rawboned man in his sixties with white hair. He went by the aliases Charles Cosgrove, M. Ryder, A. A. Brown, Harry Hale, J. J. Jones, B. M. McGregor, and “Mr. Burk.” The government believed that at least three of the other defendants—Nathan Goldberg, Al Hartman, and Harry Doe—were Chicago associates of Al Capone.

  The government had spent $500,000 and more than two years on the investigation, marking this as a case they could not afford to lose, which is why the lead prosecutor today was the chief of the Prohibition Bureau himself, Amos Walter Wright Woodcock, a methodical former army colonel. He had argued for years that the only effective way to enforce Prohibition was to bring “a steady attack” against major crime syndicates and leave the small-time moonshiners alone. Big fish, not little fish. Consolidated was the whale.

  Woodcock looked at Elizebeth in the witness box. “Have you a message which we will identify as 6:07 P.M., April 8, 1931?”

  She shuffled her papers and located the sheet containing this message, which began, in code, QUIDS, ABGAH, FLASH, SLATE, FABLE, SHOOT, BOWSKY. She was about to read the solution aloud when a defense attorney objected, claiming that her testimony was “incompetent, irrelevant, and immaterial.” Eight other defense attorneys leapt to their feet, joining the objection. One was Edwin Grace of New Orleans, who handled Capone’s appeals in federal courts. Grace was joined at the defense table by Walter J. Gex Sr., the patriarch of an influential family from the nearby Gulf city of Bay St. Louis, Mississippi.

  “I believe I am asked my opinion of the reading of this message?” Elizebeth said, turning to the judge. “This is not a matter of opinion. There are very few people in the United States, not many it is true, who understand the principle of this science. Any other experts in the United States would find, after proper study, the exact readings I have given these.”

  “I ask all that be excluded,” Gex said. “I think it is very improper.”

  Woodcock continued to walk her through the messages, and Elizebeth read her plaintext translations to the court, one at a time, interrupted by additional objections from the defense, accusations that she must have gotten her information from federal agents and not from codebreaking (“That certainly is some information somebody gave to that lady”). The syndicate had used a system of enciphered code—words that stood for letters in a cipher that stood for letters in a code. To solve it, Elizebeth had to rewind each step.
A message that looked like

  GD (HX) gm ga HX (GD) R gm OB BT HR CK 25 BT BERGS SUB SMOKE CAN CLUB BETEL BGIRASS CULEX CORA STOP MORAL SIBYL SEDGE SASH (?) CONCOR WITTY FLECK SLING SMART SMOKE FLEET SMALL SMACK SLOPE SLOPE BT SA back to the word SLDGE its SEDGE instead of SLDGE HW

  became in plaintext

  SUBSTITUTE FIFTY CANADIAN CLUB BALANCE BLUE GRASS FOR COROZAL STOP REPEAT TUESDAY WIRE CONCORD GO TO LATITUDE 29.50 LONGITUDE 87.44

  When it was time for cross-examination, Gex rose from the defense table.

  “How shall I address you,” he said. “Madam or Miss?”

  “I am Mrs. Friedman.”

  “Before you could properly translate these symbols somebody had to tell you it was symbols in reference to the liquor transportation?”

  “Oh no,” Elizebeth replied, innocently. “I might receive symbols related to murder or narcotics.”

  “The same symbols these gentlemen used to mean what you say, whiskey, beer, position, could not have been made up by people in code for transportation of women from Europe?”

  No, Elizebeth responded, “not with the meaning given them here.”

  “I move that all of the testimony of this lady be stricken out,” said another lawyer, Maxwell Slade. The judge overruled him.

  When there were no further questions from the defense, Elizebeth reached for her handbag, stepped out of the witness box, exited the courthouse into the damp spring air, and returned to Washington. Four days after her testimony, the jury convicted five of the syndicate’s ringleaders, including Albert Morrison, who received two years in prison. Woodcock gave credit to Elizebeth, telling her superiors that she “made an unusual impression on the jury.”

  And not just the jury. Reporters covering the trial were taken with Elizebeth, describing her in stories as “a pretty government scrypt-analyst or ‘code-reader,’ ” “a pretty middle aged woman,” “a pretty young woman with a filly pink dress,” and “a pretty little woman who protects the United States.” This was her first sustained encounter with the press and it left a sour taste. She was still young enough to resent being called a middle-aged woman and thought the kinder phrases were badly written. The newspapers wrote about her again the following year when she returned to New Orleans to testify in the convicts’ appeal. Facing off against Edwin Grace, Capone’s attorney, Elizebeth grew impatient with his attacks on the validity of her science and told the judge she could settle the issue quickly if she had a blackboard. A bailiff found a blackboard in storage and wheeled it into the court. Elizebeth stood with a piece of chalk and diagrammed the rum ring’s codes on the board until the jurors were nodding their heads and Grace was muttering that this was highly irregular. “CLASS IN CRYPTOLOGY,” one newspaper blared the next day. The headline made her queasy. She had not signed up to be a public figure. She hoped the attention would die down. To her horror, it was only just beginning.

  William went dark as his wife ignited and lit up the sky. The army kept his projects locked behind thick metal doors. He wasn’t allowed to talk about his work, and she knew enough not to ask. “He never put into words and never asked me to put into words,” Elizebeth said. The two cryptologists were so connected, so attuned to the slightest crease of an eyebrow or curl of a lip, that their faces had a way of becoming mirrors. A grim look on his face was instantly reflected in hers. Because neither wanted to cause pain in the other, the most painless course of action was often to make their faces like masks and not speak about forbidden things, though despite these efforts at self-control Elizebeth could tell when something was troubling her husband: “Many times there was a certain grim look that came around his mouth.”

  What she didn’t know at the time, and didn’t learn until after the war, is that the army had asked William to break a series of ciphers used by Japanese officials to encrypt their diplomatic communications, an enormous undertaking that would come to define his career and consume the next decade of his life.

  He wasn’t doing it alone. Like Elizebeth he was starting to build his own codebreaking unit, hiring junior cryptanalysts and molding them to his needs and his vision. The army had never agreed to such an expansion before, but the plates of government codebreaking had shifted due to a fateful decision made by the newly appointed secretary of state, Henry Stimson. A former artillery officer in the Great War, Stimson thought the idea of reading other nation’s messages in peacetime was immoral, and upon learning that the State Department was paying codebreakers in New York to read the mail of foreign diplomats—Yardley and his American black chamber—the secretary was appalled. “Gentlemen do not read each other’s mail,” he was supposed to have said. Whatever the reason, Stimson decided to pull Yardley’s funding in 1929, and the chamber was forced to shut down.

  Because Yardley had developed some expertise with Japanese ciphers, the closing of his bureau left America with no ability to break new ciphers developed by Japan, a nation with a growing military and ambitions of empire. Worried about being left in the dark, the army turned to William, and in 1930 he launched a new army codebreaking unit that would later become the nucleus of the National Security Agency. William called his new organization the Signal Intelligence Service, or SIS.

  The first three people he hired for the new unit were male mathematicians in their early twenties: Abraham Sinkov, Frank Rowlett, and Solomon Kullback. He gave them adjacent desks on the third floor of the Munitions Building, in a vault behind a thick steel door, and began to train them with dusty books and sample problems. William handed Rowlett a book by a German cryptographer named Kasiski. How is your German? he asked Rowlett. Rowlett said it was not good. William suggested he study the Kasiski text anyway. Now here was a book by an Austrian military cryptographer named Figl, and here were some of Friedman’s own works, the “Riverbank Publications,” written with Mrs. Friedman during the war. “Do you speak French?” Friedman asked Rowlett, who was sorry to say that he did not.

  One morning, when William felt the three young men were ready, he popped his head in the vault and said they should come with him immediately. They followed the boss down a staircase and through a long corridor, where he took a left turn into an area that seemed deserted except for a steel door with a combination lock. Friedman reached into his coat pocket, removed a card, examined a series of numbers on it, and manipulated the lock until the bolt swung free. The door opened; behind it was a second steel door with a keyhole. Friedman fished a key from his jacket and jiggled it in the lock. The inner door opened into a completely dark room. The men waited outside as the boss disappeared into the dark and struck a match. A smell of smoke drifted out of the vault. Friedman found the dangling cord of a lightbulb and pulled, revealing a windowless room full of filing cabinets. He switched on three wall-mounted fans, turned to his young employees, and said, with gravity, “Welcome, gentlemen, to the secret archives of the American Black Chamber.”

  Yardley’s files. The cabinets from his bolted bureau. Friedman had obtained them on the theory the files would prove useful to future army codebreaking projects, particularly regarding Japan.

  Over the next fifteen years, William and his three young colleagues, joined by others, would drive themselves to the brink of collapse while working as codemakers and codebreakers alike, struggling to solve Japanese cipher machines of unknown design and applying what they learned about the weaknesses of the foreign machines to design new kinds of secure machines for America. At heart it was all the same business, the business of dominating secrets.

  William first taught his deputies about cipher machines by challenging them to solve different machines that he himself had mastered years before (the Kryha, the Hebern), offering gentle hints when the young codebreakers got stuck. Then he set them to work on their first mysterious and yet-unconquered Japanese machine, which began transmitting in 1930: Angooki Taipu A, meaning “type A cipher machine” in romanji, the romanized form of Japanese that was used for transmitting messages. The name Angooki Taipu A was the codebreakers’ sole
piece of information about the machine. No one in the Western Hemisphere had ever laid eyes on it. There wasn’t so much as a drawing of one, much less a physical copy. Internally, William and his colleagues referred to it by a nickname, “Red.” Later, in 1938, the Japanese replaced Red with a more sophisticated machine, a significant upgrade: Angooki Taipu B. This one the SIS codebreakers called “Purple.”

  Red and Purple were used by Japanese diplomats in Nazi Germany to communicate with the Japanese government back home, which meant that solving the machines would give the SIS a most intimate view of strategic thinking in both Japan and Germany. To attack each system—first Red, then Purple—the American codebreakers needed to build their own bootleg versions of the Japanese machines, reverse-engineering them based on nothing but educated guesses from analyzing the garbled messages they produced. It was a task akin to building a watch if you have never seen a watch before, simply by listening to an audio recording of the ticking and clicking of its gears.

  Even as he tried to infer the shape of foreign machines, William was building an American machine of his own, to protect American communications: the Converter M-134, a multirotor device of innovative design. He filed a patent in hopes of selling it one day on the commercial market. The patent was granted but held secret for many years, essentially nullfiying his rights. This would become an enduring frustration for William. As hard as he tried to earn money on the side, security concerns always got in the way.

  Thanks to powerful brainstorms from Frank Rowlett, the M-134 eventually evolved into the SIGABA, an Egyptian Sphinx of a cipher machine with fifteen rotors and an ingenious mechanism that Rowlett called a “stepping maze.” Up to four rotors might turn at the same time, with a single key press, and the rotors could be inserted in reverse direction. The army and navy distributed 10,060 SIGABA machines across every theater of the Second World War. President Roosevelt used SIGABAs to communicate from his Hyde Park home and when he traveled on the presidential train. The SIGABA was like an American Enigma machine or Purple machine, only inviolate. No enemy codebreaker, whether German, Italian, or Japanese, would ever manage to break it, despite strenous efforts; the Nazis ultimately stopped intercepting SIGABA messages altogether, since they could not be read. The machine ensured “the absolute security of army and navy high command and high echelon communications,” William later wrote with pride, and “contributed materially to the successful outcome of the war.”