The Woman Who Smashed Codes

by Jason Fagone

  The British codebreakers worked at Bletchley Park, a mansion in the countryside outside of London. Bletchley grew from a handful of people in 1938 to thousands by 1945, the bulk of them women, recruits from the Women’s Royal Navy Service who operated the bombes, among other jobs, and were billeted in large country houses.

  The Enigma codebreaking program would come to be known as ULTRA; Enigma decrypts were stamped with the imposing phrase TOP SECRET ULTRA as a reminder to handle them with the utmost care. Later, America would join forces with the British, assembling its own ULTRA factories in Washington and sharing the burden. But early in the war, when Elizebeth and her coast guard unit analyzed their first Enigma machine, ULTRA was a strictly British franchise. There was no one to tell the Americans what to do. They had to invent their own method.

  At first, Elizebeth didn’t know that she was dealing with an Enigma at all. Enigma cryptograms look like lots of others, generic blocks of nonsense letters. In January 1940, coast guard radio monitors began intercepting one to five messages per day with the call signs MAN V NDR and RDA V MAN. Elizebeth wasn’t able to make heads or tails of the first twenty or thirty messages that were intercepted on this circuit. However, after accumulating a greater “depth” of messages, sixty or seventy, she was able to write them one on top of another on a worksheet and see the letters in a new way by gazing down the columns.

  Enigma is poly-alphabetic. It creates a new cipher alphabet with each key press. That’s the beauty of the machine. But if an Enigma user types a number of messages using the same starting position of the rotors, the first letter of each message will use the same alphabet—and the second letter of each message will use the same alphabet, and the third letter. In other words, any individual message is full of alphabets, but if a codebreaker lines up the messages in a tower, each column in the tower is mono-alphabetic—one alphabet:

  1

  2

  3

  4

  5

  6

  7

  D

  X

  J

  X

  L

  H

  N

  . . .

  L

  W

  S

  X

  I

  Y

  F

  . . .

  M

  H

  O

  S

  S

  L

  C

  . . .

  The letters in the first column here, D L M, all use the same alphabet. And the letters in the second column, and so on.

  With only three messages, there isn’t enough information to help the codebreaker. The “depth” is too low. There need to be more floors in the tower. At greater depths, closer to twenty messages and beyond, letter frequencies become visible:

  1

  2

  3

  4

  5

  6

  7

  D

  X

  J

  X

  L

  H

  N

  . . .

  L

  W

  S

  X

  I

  Y

  F

  . . .

  M

  H

  O

  S

  S

  L

  C

  . . .

  M

  A

  P

  A

  C

  T

  Y

  . . .

  F

  P

  W

  S

  G

  S

  C

  . . .

  Y

  Q

  A

  S

  A

  C

  W

  . . .

  N

  S

  H

  W

  U

  F

  C

  . . .

  F

  U

  W

  X

  G

  S

  P

  . . .

  M

  B

  D

  W

  X

  U

  O

  . . .

  O

  P

  O

  D

  Y

  X

  L

  . . .

  A

  J

  Y

  S

  X

  F

  D

  . . .

  M

  W

  S

  X

  E

  C

  C

  . . .

  M appears four times going down column 1. In this column, M might be equivalent to the letter E, the most frequent letter in German as well as English. In other columns, different letters might be equal to E. And now the codebreaker can use tried-and-true methods to fill in plaintext letters and piece together the adversary’s words.

  In this way, the technique of “solving in depth” can take a hard problem and turn it into a simpler problem. The trick is often to get the messages aligned in depth in the first place. If the Enigma user changes the starting position of the rotors from message to message, the floors of the tower have to be staggered to track with the shift in the starting position.

  Figuring out how to align messages in depth is a subtle art. It can be done with clever guesswork and trial-and-error, and it can also be done by applying the principle of the Index of Coincidence, William Friedman’s fundamental insight about the relationships between letters that sit in towers of text. Elizebeth tended to use both approaches in her work, but luckily, in this case, she didn’t need to align the messages, because the senders had made a mistake by using the same starting position for all the messages. The messages were already in depth. Before long, then, the coast guard codebreakers were able to identify frequent letters in the columns and use those letters to piece together the plaintexts for most of the first batch of messages.

  The words seemed to be in German.

  Elizebeth and her colleagues still didn’t know what type of cipher they were dealing with, so now they decided to write down the alphabets for many of the messages they had solved in depth, the ciphertext equivalents of the plain letters, to see if a pattern popped out. They quickly noticed that no letter was ever enciphered as itself: an A never meant A, a B never meant B. This suggested an Enigma.

  They went to the shelf in their coast guard office and picked up their old commercial Enigma machine.

  The codebreakers had already solved most of the messages, but now they wondered if they could solve the machine itself—the wiring. Knowing the wiring makes it easier to solve new messages. Without the wiring, they would have to repeat the laborious process of solving in depth every time the key changed. Their challenge now was to use the text they had recovered, the plain letters and the cipher letters, to work backward toward the unknown machine, almost like a police detective analyzes the spatter pattern of blood at a murder scene, starting with the red evidence and rewinding back to the moment of the crime, deducing from the crusts of blood the speed and angle of the knife.

  Unbeknownst to the coast guard, groups of British and Polish codebreakers working on the Enigma problem had already discovered methods for working backward from the text to the machine. The Poles had done it with an algebraic approach, the mathematics of permutations, and one of the brilliant Bletchley codebreakers, a linguist and scholar of classical literature named Dilly Knox, had relied more on pattern recognition and a kind of alphabetic grid called a “rod square.” But the coast guard didn’t know about these approaches, and so, working in isolation, the codebreakers had to grope toward their own method. They poked and prodded and turned the wheels; they wrote alphabets on sliding strips of paper and moved the strips against one another, think
ing.

  It seemed to Elizebeth that there must be a fixed relationship between the alphabets she had already discovered by solving in depth—the plaintext letters and their cipher equivalents—and the motions of the Enigma’s wheels. To test this hypothesis, she drew a number of diagrams that visualized the relationships between letters at each position of the machine. She wrote new kinds of towers of letters on the worksheets that were more like X-rays than photographs, probing more deeply into the identities at the heart of the Enigma, and immediately she saw clear patterns, hints of order and regularity.

  Certain letters repeated vertically on the page, like LL and HH, and also pairs of letters, like SJ and EM. Elizebeth and her colleagues realized that these letter groups were telling them something about the spacing between pairs of wiring contacts on the Enigma’s rotors. The maps were whispering secrets about the physical intricacies of the machine. Building upon these “remarkable results” over the following days, filling more worksheets to the brim with letters, drawing more towers and analyzing the patterns that appeared, the codebreakers managed to solve the wiring for all three wheels of the unknown Enigma. Then they were able to reveal the full plaintexts of all unsolved messages from the radio circuit.

  The codebreakers now realized two slightly disappointing facts: The plaintexts seemed to contain no Nazi secrets; later the codebreakers learned that the messages had been sent by the neutral Swiss army, which sometimes used Enigmas to communicate in German. Then the coast guard shared the wiring diagram of the Enigma with William’s codebreaking team at the army, in case it might be useful to them, and the army reported back that the diagram corresponded exactly to the wiring of a commercial version of Enigma.

  Elizebeth had hoped that she was mastering a new kind of Enigma entirely. Still, it was a significant achievement. “This recovery of wiring assumed to be unknown was achieved without prior knowledge of any solution or technique and is believed to be the first instance of Enigma wiring recovery in the United States,” her team wrote in a secret technical memo after the war. As far as Elizebeth and her codebreakers could tell, and they were hardly prone to bragging, they were the first Americans to solve an unknown Enigma.

  Until this moment, cipher machines had always been William’s territory, not Elizebeth’s, but her solution of the commercial Enigma showed that she had a similar aptitude for solving machines, and this initial headfirst dive into the pool of Enigma codes would lead her to deeper waters later in the war. Of course, she didn’t know this in early 1940. Demolishing that first Enigma was just work. She was confident enough in her abilities that solving an Enigma seemed like a reasonable and normal thing that she might accomplish with her team on a given week. She didn’t brag or make a big deal. Anyway, there was no time. New puzzles were arriving at the coast guard all the time, new codes to break, along with increasing demands for assistance from outside agencies.

  All along her plaintexts had been circulating through other parts of government. Each time her unit solved a message, a clerk typed the English solution on a fresh sheet of paper, a decrypt, and gave it to the coast guard chief of communications, Vice Admiral Farley, for dissemination. Depending on the content of the decrypt, the vice admiral might send a copy to navy intelligence (OP-20-G), army intelligence (G-2), the State Department, British intelligence, or the FBI. The decrypts were like blood cells in the veins of government, delivering the vital oxygen of raw intelligence, and as different intelligence agencies realized that Elizebeth had tapped into a trove of information about Nazi spies, they inevitably asked the coast guard for more decrypts. In the 1920s, she had complained about government men “appearing on my doorstep,” wanting her to solve puzzles. They were still appearing on her doorstep, but now, instead of relatively anonymous T-men, they were some of the most powerful spymasters in the world.

  J. Edgar Hoover liked to eat dinner at Harvey’s restaurant on Connecticut Avenue, next to the Mayflower Hotel, a five-minute walk from his suite of offices in the Department of Justice headquarters, a gargantuan gray edifice near the National Mall. Harvey’s had separate dining rooms for men and women. The ladies’ dining room was on the second floor, accessible by a separate entrance at street level. The first floor was the gentlemen’s restaurant and bar, with waxed floors and rich leather banquettes. It was one of those places in Washington where men of influence slurped oysters and let their guard down for an hour or two.

  The FBI director’s face was beginning to acquire some of the first creases and pouches that would characterize the eventual marble busts of him. He was forty-five years old, one of the few immutable objects in an ever-changing city. He wore white shirts, double-breasted Brooks Brothers suits, and a hat with a brim that could be turned up or down. His agents wore the same uniform. A pink expanse of forehead separated his bushy black eyebrows from his thinning hair. The large neat desk back at his office, a corner office on the fifth floor of Justice, contained a radio, usually a vase of fresh flowers, and a framed copy of “Penalty of Leadership,” the text of a Cadillac advertisement from 1915. It read in part, “When a man’s work becomes a standard for the whole world, it also becomes a target for the shafts of the envious few.”

  At Harvey’s he usually ordered steak or roast beef and a Caesar salad. He ate at the same table every time, the most secure in the room, almost invisible from the door under a stairway. A reporter once watched Hoover sign twenty autographs during a single dinner at Harvey’s. He liked to sit there with his chief deputy, armed bodyguard, and longtime companion Clyde Tolson. It was a table for four with just two chairs. There was always a bottle of wine waiting for Hoover at his table when he arrived—part of a ritual that he performed here.

  William Friedman dined at Harvey’s on occasion. There were times at the restaurant when the cryptologist sensed motion in his peripheral vision, when a shadow darkened the white cloth. He turned his head and saw Hoover standing there with the bottle of wine. Without saying a word, the director nodded and poured wine into the cryptologist’s glass. He had respected William for years and appreciated his periodic assistance with FBI cases, with the little encrypted notes written by criminal suspects that William would solve in his free time and send back to the bureau.

  Hoover was almost certainly aware of Elizebeth Friedman. But he would not yet have had many chances to cross her path. She wasn’t allowed to eat in the gentlemen’s dining room at Harvey’s. There were a lot of male enclaves like this in the city, inaccessible to her. And Hoover was a chauvinist of the old school. When he first took charge of the bureau in 1922, there had been three female agents. He got rid of them. The next two female agents wouldn’t join the bureau until after his death in 1972. He argued that women weren’t agent material because they couldn’t be taught to shoot guns. Female clerks and secretaries at the bureau had to wear skirts and weren’t allowed to smoke at their desks as the men could. One of his least favorite people was Eleanor Roosevelt, who wrote him a mildly indignant letter after the FBI conducted an intrusive background check on one of her friends. Hoover compiled a secret dossier alleging she was a communist. “When a woman turns professional criminal,” he wrote once, “she is a hundred times more vicious and dangerous than a man.” Women at Hoover’s bureau were only deemed fit for “boring clerical functions,” according to the memoir of one longtime agent. “It was perfectly all right to bullshit ’em and ball ’em: Just don’t tell ’em any secrets.”

  But by 1940 Hoover had gotten himself into a jam serious enough to require the technical assistance of a woman.

  It had long been the FBI’s job to disrupt espionage rings within U.S. borders. Any Nazi spies operating in America were Hoover’s quarry. However, he didn’t seem to be very good at catching them. He had built the bureau’s name on its flashy investigations of jazz-age gangsters, men who enjoyed attention and went out in public with entourages. Counterespionage was another discipline entirely, a matter that required a certain finesse, and the bureau’s first sizable Nazi spy case
, in 1938, had ended in a public-relations disaster.

  That year in New York, the FBI arrested a Chicago man of Austrian parentage, Guenther Rumrich, along with two associates suspected of spying for Nazi Germany. Then an FBI agent named Leon Turrou made the mistake of tipping off Rumrich’s collaborators that an indictment was coming. They panicked and fled the country.

  Newspapers mocked the FBI for letting Nazis slip through its fingertips, and U.S. intelligence agencies that had long resented the FBI found new reason for their scorn. Over the years, Hoover’s insatiable hunger for publicity had caused a lot of bad blood; in the press he repeatedly claimed sole credit for investigations to which other agencies had contributed but were not free to discuss. The head of army G-2, George Strong, one of William Friedman’s superiors, despised Hoover, and the navy OP-20-G chiefs couldn’t stand him, either. Henry Morgenthau at the Treasury hesitated to even speak the director’s name in meetings, and Hoover thought of him as “that Jew in the Treasury.” When British intelligence officers started to arrive in Washington in 1940, hoping to forge links with U.S. agencies, they were shocked by this toxic atmosphere of mistrust and quickly traced the cause to Hoover. “J. Edgar Hoover is a man of great singleness of purpose, and his purpose is the welfare of the Federal Bureau of Investigation,” a group of British operatives later wrote. “It was once remarked of a well-known Oxford scholar that, while he had no enemies, he was hated by all his friends. Something of the same kind would express the feelings towards the FBI of its fellow U.S. agencies.”

  For a man as vain as Hoover, and as publicity-obsessed, and as intensely disliked by rivals in his own government, the bungled Rumrich case represented both a personal black eye and a threat to the FBI’s future authority. Somehow he needed to salvage the bureau’s reputation, to prove that it was capable of catching fascist spies, and in 1939, he proposed a bold plan to do just that.

  Hoover knew that the concept of “hemisphere defense” had become a fixation with Roosevelt and military chiefs: Guarding the United States meant guarding the entire Western Hemisphere from Nazi encroachment. In other words, it wasn’t enough to fortify U.S. defenses. South America must be protected as well. Roosevelt talked about hemisphere defense in speeches, arguing that “no attack is so unlikely or impossible that it may be ignored,” and Secretary of the Navy Frank Knox raised the specter of Nazi planes taking off from South American airfields in the night and dropping bombs on “our own women and children in our teeming seaboard cities.” Seeing an opening, J. Edgar Hoover pressed Roosevelt to dramatically expand the FBI’s jurisdiction. For the sake of “the common defense of the Western Hemisphere,” Hoover argued, the FBI must be allowed to operate beyond U.S. borders. He demanded the authority to send men into South America, “to seek out and identify agents of the Axis operating in all the Americas, to ensure the ultimate safety of the United States.”