The Woman Who Smashed Codes

by Jason Fagone

  The Zimmermann Telegram, as it came to be known, was indisputable proof of a German plot against America, “clear as a knife in the back and near as next door,” as the historian Barbara Tuchman has put it. Residents of Texas were particularly displeased to learn that the Kaiser was trying to give them to Mexico, but outrage against Germany was general across the States. The telegram sped up history. It pushed America into war with Germany, whether America was ready for war or not.

  It was not.

  And this is how the telegram changed the destinies of Elizebeth Smith and William Friedman: as American codebreakers, they happened to possess an extraordinarily rare and suddenly indispensable set of skills.

  Elizebeth got word in January 1917 that her mother, Sopha, long ill from cancer, was on the verge of death, and Elizebeth should come to Indiana to say goodbye. She packed a bag and rode the train back to Huntington and her childhood home. Her father was there, and her sister, Edna. The two sisters consoled each other as physicians prowled through the old house, murmuring about a growth. Sopha was in a lot of pain and vomited violently. A doctor turned her on her belly and spread iodine across her back. He used cocaine to numb a particular spot and tapped a metal rod into the skin, removing what Elizebeth felt was a horrifying quantity of pinkish fluid.

  She had brought some cipher materials with her, hoping to get work done. “My book-bag lies here unopened,” she wrote to William at Riverbank. “I try to make myself work, but I cannot. I sit a moment, then spend the hours pacing back and forth from Mother’s bed, in the vain hope that there is something I can do. It is so awful—Billy Boy—to look on the face of death like that—the beckoning face—Do you know it makes me think a lot about posterity, and responsibility, and all that?” She wasn’t sure what to call William in these letters, or to call herself in relation, so she mostly kept things platonic, signing her letters “yours, Elsbeth,” and thanking William for being “one of the truest friends I’ve ever had,” although she did admit that she missed William’s “rocking,” his comforting way with a rocking chair, and in one letter she slipped in something stronger: “I love you / Elsbeth.”

  When Sopha died, in February 1917, Edna stayed behind to arrange the funeral, while Elizebeth returned to Riverbank, seized by a new impatience. She had no desire to spend any more time on the Bacon ciphers. Life was too short to waste on fruitless quests. When she reunited with William, he said he felt the same. They both agreed they had to remove themselves from the project. The question was how.

  Confronting Mrs. Gallup seemed a little cruel. She had worked too many years in a single direction to admit her compass was broken. She had treated both of them with kindness. They tried talking to Fabyan instead. On a few occasions the two youngsters buttonholed him and tried to get him to listen. Mrs. Gallup’s theory was unsound, they said. Fabyan’s money might be better spent on other projects. He shouted them down, as they expected. Fabyan said he wasn’t paying them to question the theory, only to persuade the academy that it was correct.

  But by now, even if he didn’t want to admit it, a new scheme was diverting Fabyan’s attention from the literary ciphers. Shakespeare, Bacon, Mrs. Gallup, old books, dead men—it paled in urgency to the world of the living.

  For months now Fabyan had been advertising his patriotism and his willingness to place Riverbank at the disposal of the flag. He had ordered his groundskeepers to expand the network of model trenches next to the Lodge, and after months of digging by a team of mud-spattered workers the trenches reached a total length of three miles, enough to be useful for the Fox Valley Guards to conduct infantry drills complete with live mortar rounds. And Fabyan had told officials in Washington that if they needed help with codebreaking, Riverbank stood ready to serve.

  “Gentlemen,” he wrote to Washington on March 15, 1917, “I offer anything I have to the government, and if you care to have any of your local men call on me, and see the work that is being done, I should be very glad to show it to them.” He described his interest in old ciphers, especially the biliteral cipher of Francis Bacon, and added: “To avoid any possibility of being considered a crank, or a theorist, I respectfully call your attention to the fact that I was the business partner of the late Cornelius N. Bliss, formerly Secretary of the Interior, whom most of the older men in Washington remember with a great deal of respect and admiration.”

  Military officials were of course reluctant to give any power or responsibility to a fake colonel in Illinois, but they had little choice but to accept Fabyan’s offer. They were desperate for codebreakers because of the way radio and wireless technology was changing the art of war.

  In earlier conflicts, codebreakers had mattered less; fewer military and diplomatic messages were encrypted because the messages were harder to intercept. If you wanted to steal an enemy message, you had to capture a messenger on horseback, or open an envelope at a post office, or install a tap on a telegraph wire. But with radio, all it took to intercept a message was an antenna. The air was suddenly full of messages in Morse code, dots and dashes that registered as audible pings and whines. You could pluck them out of the sky. So to protect their secrets, armies had begun encrypting their wireless messages before sending them over the wireless in Morse.

  This simple fact transformed codebreakers from disreputable freaks into potential superheroes, wizards with power over life itself. Now the air was full of encrypted information of enormous tactical significance and the utmost stakes. The routes of ships at sea. Troop movements on the ground. Airplane sightings. Diplomatic negotiations and gossip. Reports of spies. Thousands upon thousands of puzzles zipping through the atmosphere, any one of which, if decrypted, might win or lose a battle, wipe out a regiment, sink a ship. In this new world, a competent codebreaker was suddenly a person of the highest military value—a savior, a warrior, a destroyer of worlds.

  And yet, as Elizebeth would later write, “There were possibly three or at most four persons” in the whole United States who knew the slightest thing about codes and ciphers. She was one of them, William another.

  The government lacked the capacity to reliably intercept foreign messages, much less break the codes and read them. The CIA didn’t exist in 1917. There was no NSA, and the FBI was a crumb of its future self, a nine-year-old organization known as the Bureau of Investigation, which fielded only three hundred agents, on a total budget of less than half a million dollars. There simply was no intelligence community as we think of it today. The Department of Defense was called the War Department then, which operated the army, and though the War Department did contain an intelligence-gathering section, the Military Intelligence Division (MID), it was tiny and underfunded. On the day Congress declared war, April 6, 1917, the MID employed just seventeen officers. The officer in charge of the MID, Major Ralph van Deman, considered the government’s ignorance of codes and ciphers an “emergency.”

  So, in the second week of April, the War Department dispatched an emissary to Riverbank, an army colonel named Joseph Mauborgne, to check out the place and report back on its suitability.

  Mauborgne was one of the three or four people in America who knew something about codebreaking. In 1912, while stationed at the Army Signal Corps School in Kansas, a bare-bones airfield and laboratory to probe radio technology, Mauborgne had made history by figuring out how to send a radio signal from a plane to the ground for the first time, and in 1914 he became the first American to break the British army’s field cipher, known as the Playfair Cipher, based on a table of letters arranged in a five-by-five grid.

  When Mauborgne arrived at Riverbank, Fabyan greeted him with the usual overwhelming gusto and brought him to the second floor of the laboratory building, declaring with a flourish that the Riverbank Department of Ciphers was open for business. The office appeared busier and more crowded than it had ever been. In anticipation of the army man’s visit, Fabyan had gone out and hired a dozen clerks, stenographers, and translators fluent in German and Spanish, to provide support for Eli
zebeth and William. Fabyan hoped the two young people would be able to lead the effort, to break codes for the government, while Mrs. Gallup continued her long labors on the Bacon ciphers. Superficially, the office looked like the picture in Fabyan’s imagination, the pitch he had sold to Washington. It looked like a codebreaking agency on the prairie.

  There in the new Department of Ciphers, Elizebeth and William introduced themselves to Mauborgne. They clicked with him immediately. He was thirty-six and big—big body, big voice, big brain, with perfectly round, black glasses. He was the only man they had ever seen stand eye to eye with Fabyan and not seem intimidated. Mauborgne liked Elizebeth and William, too. He saw a spark in the pair of young codebreakers. (He would later call them “the two greatest people I have ever known.”) They had little formal training but were bright and eager. Fabyan made him wary—a mess of a man, lunging wildly from promise to promise—but it was undeniable that Riverbank had excellent security from a military standpoint. Aside from the virtue of being in the middle of nowhere, safe from enemy attack, it was protected by the lighthouse, and Fabyan also had the Fox Valley Guards nearby—his own private army. If all else failed and the Germans invaded, Fabyan said he would open the bear and wolf cages in the garden and sic the beasts on the intruders.

  On April 11, Mauborgne informed his commanders that Riverbank was ready. He urged the army and also the Justice Department “to take immediate advantage of Colonel Fabyan’s offer to decipher captured messages,” owing to “the mass of data” in his private library of cipher books, the security of his compound, and the quality of his employees, “a force of eight or ten cipher experts who spend their time delving into the works of antiquity, discovering historical facts hidden away.”

  After reading Mauborgne’s enthusiastic report, Van Deman of the MID wrote to Fabyan with gratitude, thanking him for “your exceedingly kind and patriotic offer of assistance,” and soon encrypted messages started arriving at Riverbank from Washington. They came in the mail and by telegram, sent by different parts of the government: the War Department, the navy, the Department of State, the Department of Justice. The messages had been intercepted by covert means, mostly from various telegraph and cable offices across the country.

  Fabyan had gotten his wish: for the foreseeable future, Riverbank would become ground zero for military codebreaking in America, a de facto government agency. He had drafted Elizebeth and William into the war, assuming they would be able to handle what was coming. But when they looked at the messages, the fresh piles of gobbledygook spilling from the mail sacks onto their desks, they weren’t sure that he was right.

  A woman and a man are sitting side by side in a busy room. People come and go and the door opens and closes and there is the sound of typewriter keys smacking ink into pages. Outside the window, hawks fly and cows moo and a bear scratches himself in an iron cage and a parrot sings and a river runs and there are also monkeys in diapers for some reason.

  The two people, Elizebeth and William, notice nothing except what is in front of them on a slab of desk. They are looking down at a sheet of paper. All of their intensity is shining down at the paper, a bright beam of desire to understand the text that is typed there.

  It looks like nothing. It is clearly not written in the biliteral cipher of Francis Bacon that they are familiar with. It is something else, a new level of mystery. They must understand it. But they don’t know what they are looking at.

  BGVKX

  TLXWB

  SHSFW

  KWGRI

  KZTZG

  RKZFE

  YDIWT

  KOFOB

  GUHGD

  SFVRE

  UIUQX

  HSLDS

  OHSRM

  HTWKY

  VHUIK

  BJDUH

  VSART

  BGVNG

  VBAFO

  AZOXG

  PQPMJ

  DRODW

  RCNML

  MTMXL

  SSVAR

  A hiss of symbols, a raw block of babble. A cryptogram. Someone wrote and sent it for a purpose, and someone else intercepted it, and now it is here on your desk. These letters contain meaning. How to unlock it without knowing the key?

  The basic task of codebreaking might seem impossible if you think about how many different ways a message might be encrypted. Each human language has its own quirks and curiosities. Then, within each language, a cryptographer can choose from among dozens of varieties of locks—codes and ciphers. And each lock will accommodate only one of a vast array of possible keys.

  For instance, one of the simplest kinds of ciphers, called a mono-alphabetic substitution cipher, or MASC, swaps out one set of letters for another. Perhaps A=B, B=C, and C=D, or perhaps A=X, B=G, and C=K—or any other map between the 26 letters of the plaintext alphabet and 26 different letters in the ciphertext. A MASC is a very basic method of encrypting a message. But even here, there are 403,291,461,126,605,635,584,000,000 potential alphabets: 403 septillion. A thousand computers, each testing one million alphabets per second, would take more than a billion years to exhaust the possibilities.

  And yet anyone who has ever solved a cryptogram on a newspaper puzzle page has conquered the 403 septillion possibilities, because, of course, there are shortcuts, ways of taming the task by grabbing on to certain patterns in the text.

  This is the essence of codebreaking, finding patterns, and because it’s such a basic human function, codebreakers have always emerged from unexpected places. They pop up from strange corners. Codebreakers tend to be oddballs, outsiders. The most important trait is not pure math skill but a deeper ability to pay attention. Monks, librarians, linguists, pianists and flutists, diplomats, scribes, postal clerks, astrologers, alchemists, players of games, lotharios, revolutionaries in coffee shops, kings and queens: these are the ones who built the field across the centuries and pushed the boundaries forward, stubborn individuals with a lot of time to sit and think and not give up.

  Most were men who did not believe women intellectually or morally capable of breaking codes; some were women who took advantage of this prejudice to steal secrets in the shadows. One of the more cunning and effective codebreakers of the seventeenth century was a Belgian countess named Alexandrine, who upon the death of her husband in 1628 took over the management of an influential post office, the Chamber of the Thurn and Taxis, which routed mail all throughout Europe. The countess had a taste for espionage and transformed the Chamber into a brazen spy organization, employing a team of agents, scribes, forgers, and codebreakers who melted the wax seals of letters, copied their contents, broke any codes, and resealed the letters. This was an early example of what the French would later call a cabinet noir, or black chamber, a secret spy room in a post office. The countess’s male contemporaries were slow to discover her true occupation because they couldn’t imagine that a woman was capable of such deceptions. “What if this countess does not merely open our letters but is also capable of deciphering their contents?” one diplomat wrote in panic to another. “God knows what she is capable of doing to us!”

  The two most prominent codebreakers in America when Elizebeth and William started were a married couple, Parker Hitt and Genevieve Hitt. Parker was a tall, dashing Texan with weathered skin, an army infantry commander in his thirties who had gotten interested in cryptology after volunteering to fight in the Spanish-American War and trying his hand with messages intercepted from the Mexican army. His wife, Genevieve, a proper southern girl, had scandalized her family by falling in love with a man they saw as a cowboy. She also studied cryptology, eventually becoming chief of the code operation for the War Department’s Southern Division, based in San Antonio. “This is a man’s size job,” she wrote to her mother-in-law, “but I seem to be getting away with it, and I am going to see it through. . . . I am getting a great deal out of it, discipline, concentration (for it takes concentration, and a lot of it, to do this work, with machines pounding away on every side of you and two or three
men talking at once).” Parker supported Genevieve and was proud of her: “Good work, old girl,” he wrote to her in one letter.

  Parker was the only American to have written a serious book about cryptology. Aimed at army units with no prior training, Manual for the Solution of Military Ciphers showed how to set up a quick-and-dirty deciphering office in the field with five or six soldiers, some radio equipment to intercept enemy signals, and a day or two of study. Hitt went over the basics of military cryptography and explained, accurately, that the methods of the world’s armies had not changed much in hundreds of years. Just like there are millions of chicken recipes in the world but only several basic methods to cook the bird (roasting, frying, poaching, boiling), there are countless ciphers but only a handful of common types. Then he laid out some basic steps for solving a cryptogram written in cipher. Today a computer could do any of these steps in picoseconds, but in 1917 it all had to be done by hand, with a pencil and paper.

  The first step was usually very simple: count the letters in the cryptogram. In English, the most frequently used letter is E, the most frequent two-letter group is TH, and the most frequent three-letter group is THE. So if you count the letters in the ciphertext and the most common letter is B, it might stand for E, and if the most common three-letter group is NXB, it might stand for THE.

  You can count other things in a cryptogram, like the total number of vowels and consonants, and how often particular letters or groups of letters appear before or after other letters. All of these counts give hints to the hidden structure. A frequency count can also reveal if the plaintext was written in English, German, French, Spanish, or some other language, because the frequency of letters in a language is like a unique signature. The most common six letters in German, starting with the most common, are E, N, I, R, T, S. In French, E, A, N, R, S, I. In Spanish, E, A, O, R, S, N.