by Jason Fagone
These two campuses soon evolved into an American version of Bletchley Park, deeply secret compounds where workers solved puzzles behind barbed wire and never spoke about what they did. Many were women. It’s where the machine era of cryptology began, the era of brute force, women operating machines the size of rooms, American bombes and some of the first IBM punch-card computers.
The women of Arlington Hall and the Naval Annex were mostly WACs and WAVES, members of the army and navy auxiliary programs designed to patch the wartime shortage of male labor. They lived together in barracks and apartments. Hundreds had been trained in secret cryptology courses offered at the Seven Sisters colleges, the likes of Bryn Mawr and Vassar and Mount Holyoke, the professors relying on exercises and concepts first pioneered by William and Elizebeth Friedman. Inside the high-security buildings encircled by barbed wire and guarded by U.S. Marines some of the women sat at long rows of desks, smoking and drinking coffee, and identifying cribs to feed into the bombes, while others operated the bombes that ticked and whirred as they explored the keyspaces of distant Enigma machines. The buildings were hot and unventilated. An Arlington Hall codebreaker named Martha Waller recalled that in the summer, it was often 90 degrees indoors at 8 A.M., and because of the wartime nylon shortage the women couldn’t wear nylon stockings, so “we rejoiced in going bare. . . . Sitting quietly at a desk, one could feel drops of sweat rolling down one’s legs.”
Less than a year from now the navy would force Elizebeth and the Coast Guard Cryptanalytic Unit to move to the Naval Annex, but even then she would never work inside the large, hot rooms with the rows of young women at desks. The unit remained separate, a small elite team doing its own thing. From time to time she and her colleagues would take advantage of the Naval Annex’s technology to make progress on clandestine circuits, relying on IBM punch-card machines to perform statistical analyses that saved time in solving certain ciphers, but the initial assaults on the puzzles emerged from their brains alone. Elizebeth was among the last of the paper-and-pencil heroes. And in the summer of 1942, as the U.S. and Japanese navies clashed in the Pacific and the Nazis ordered French Jews to wear the Star of David, she was taking on the most fiendish challenge of her career, for the biggest stakes.
Exactly as she had feared, the Nazi spies had changed their codes after the March police raids in Brazil. As a direct result of the FBI’s roundup, “Germany was unmistakably informed that the systems had been solved,” wrote the uniformed commander of her unit, Lieutenant Jones, and “the inevitable consequence was that systems on all clandestine circuits were almost immediately thereafter completely changed.” Within two or three weeks the Nazis were back online in multiple locations across South America, and from there the radio network expanded, adding nodes in new cities and countries.
Elizebeth, Lieutenant Jones, and their coast guard teammates watched in frustration as new circuits lit up throughout the summer and fall of 1942—two, then five, then fifteen—each using a different and yet-unbroken code. It was as if the FBI had tried to destroy an approaching asteroid with a single huge bomb but instead just blasted the rock into dozens of sentient fragments able to regenerate and spread wreckage over a wider swath of earth.
The codebreakers were hardly the only Americans troubled by the FBI’s actions in South America. Intelligence chiefs at the army and navy couldn’t believe it either. “Unfortunately, the matter got out of hand, and it became public knowledge that the ciphers used by the espionage agents in that territory were being read by our government,” wrote Joseph Wenger, head of the navy’s OP-20-G, in an internal memo. “It might be much more valuable to the military services to obtain the information flowing through clandestine stations than to close them up.” The British were also taken aback—they had never trusted J. Edgar Hoover in the first place—and as more newspaper stories appeared in Brazil about the court cases of the arrested Nazi spies, local British officials had the stories translated and exchanged secret telegrams about the unfortunateness of it all. “You may care to read the attached rough translation of the story as it appeared in the Brazilian Press,” one diplomat in Brazil wrote to an official at MI5, Britain’s counterintelligence agency. “Rather shattering, I’m afraid.”
The whole fiasco triggered four months of jurisdictional squabbles between the different intelligence services, conducted by memo and conference. The army, navy, and British complained about the FBI; the FBI pushed back. These fights resulted in a series of awkward compromises. The army and navy forced the State Department to promise that no further clandestine stations would be seized without their approval, and all parties recognized that the coast guard had the authority and the expertise to monitor clandestine circuits in the Western Hemisphere. But no one could stop the FBI from doing counterespionage in South America. Hoover had clear authority from the president.
And so, unable to remove his power on paper, the other agencies simply started to freeze him out in secret, routing information so that it flowed around the FBI as much as possible. British relations with the coast guard “grew steadily closer and more informal,” according to the BSC history. “It was understood by both sides that no information received from the coast guard was to be divulged to the FBI.” People literally whispered secrets to one another whenever an FBI representative was in the room; the British observed that “valuable items of intelligence were imparted, hurriedly and sotto voce, either before the meetings began or after they had adjourned.”
In April, the navy ordered the coast guard to stop disseminating clandestine decrypts itself and provide the decrypts to OP-20-G instead, for tighter control. That month, representatives of the army, navy, coast guard, British Security Co-ordination, and Canadian intelligence met at a weeklong conference in Washington to talk shop about radio and spies. It was chaired by Commander Wenger of OP-20-G. The FBI was not invited. Elizebeth was. On the day she explained the coast guard’s approach, April 8, there were seven male naval officers in the room, three male army officers, four male Canadians, five male British, three male coast guard personnel, and her—listed at the top of her cohort in the meeting minutes:
U.S. Coast Guard
Mrs. Friedman
Lt. Comdr. Polio
Lt. Comdr. Peterson
Mr. Bishop
Her Cryptanalytic Unit had grown since Pearl Harbor, but it was still fairly small, with fewer than twenty cryptanalysts, translators, and clerks. They now worked as a team to fix the mess the FBI had caused: to break the new codes on the multiplying circuits and wrangle the chaos into some kind of order.
Until the arrests in March 1942, most spies had been using book ciphers or single transposition systems (a common, Scrabble-type cipher) that were similar in style and relatively easy to break. After March, the spies switched to weirder, harder stuff: running-key systems, double-transposition systems, poly-alphabetic substitution with columnar transposition. They started hopping on the radio at different times of the day in an attempt to avoid interception. They mixed and matched cipher methods in unpredictable ways. One of the new procedures relied on “rail-fencing”—a more sophisticated application of the same principle that Elizebeth and William once used for writing love notes, only now, instead of the plaintext being JE T’ADORE MON MAR or I LOVE YOU VERY MUCH, it was a snippet of German from a Nazi spy in France: HERZLICHE WEIHNACHTSGRÜßE UND WÜNSCHE ZUM NEUE JAHR: “Warm Christmas greetings and wishes for the new year . . . ”
As soon as Elizebeth and Lieutenant Jones could get a handle on one circuit and break the codes, a new one came online. It was a cryptanalyst’s nightmare. Here is a partial list of the circuits the unit was monitoring by the end of 1942:
3-G
Hamburg—Valparaíso (Chile)
3-J
Hamburg—South America
4-C
Lisbon (Portugal)—Lourenco Marques
(Mozambique)
4-D
Madrid—West Africa
4-F
Hamburg—Lisbon
(Portugal)
4-G
Stuttgart—Libya
4-H
Hamburg—Unknown
4-I
Hamburg—Bordeaux (France)
4-L
Hamburg—Gijon (northwest Spain)
4-M
Hamburg—Spain
4-N
Hamburg—Unknown
4-O
Berlin—Madrid
4-P
Hamburg—Madrid
4-Q
Hamburg—Tangier (Morocco)
4-R
Hamburg—Vigo (northwest Spain)
4-S
Berlin—Tetuan (Morocco)
4-T
Berlin—Teheran (Iran)
5-D
Hamburg—The Crimea (USSR)
This was a planet’s worth of radio signals, a one-of-a-kind view of the earth as it convulsed with war, borders blurring, power shifting hands. The Nazis had invaded Crimea, in the Soviet Union. Iran was partially occupied by British and Soviet troops. Morocco was controlled by the Nazi-collaborationist Vichy government of France. Portugal was neutral and fiercely contested by both sides. Libya had been conquered, for the time being, by Italian and German troops. The Nazis had spies and saboteurs in all these places, sending back information over clandestine radio transmitters, and Nazi diplomats and even military officers sometimes borrowed the transmitters to speak with Hamburg and Berlin in times of transition and stress. The clandestine network, though built for espionage, was really just another communications channel, a way for Nazis of all sorts to share information in a fluid situation. Nikola Tesla predicted in 1926 that “when wireless is perfectly applied the whole earth will be converted into a huge brain.” The clandestine network was the Nazi brain, fragmentary but already encircling the earth, and adding new synapses at a fearsome clip.
The information flickering through the brain wasn’t necessarily accurate—some circuits seemed to contain little but “a miscellany of military information partly factual and partly distorted,” wrote Lieutenant Jones in a 1944 memo, “evidently pieced together from barroom conversations with merchant seamen”—but even the garbage circuits were useful to monitor. “In all cases,” Jones went on, “the reading of these circuits provided assurance that if any serious leak of vital information did occur we would learn of it almost before it became known in Germany and would be in a position to provide immediate safeguards.” In other words, the Nazi brain served as a kind of early-warning system for the Allies. Beyond that, the solved messages provided a wealth of information about the Nazi grasp of American military capabilities, because when Germany asked the spies for information about U.S. ballistics or antiaircraft guns, they were revealing that this was information they lacked.
Elizebeth raced to stay on top of the shifting codes, the proliferating patterns. Her worksheets grew weird, beautiful. She filled the grid squares with letters and numbers that made different geometric shapes when you stepped back and looked at the worksheet from a distance. Some of the shapes were parallelograms, some looked like stairs, others like labyrinths. She pulled mischievous letters from the sky and sorted them on the page. The invisible world was all out of whack, misaligned, and she had this set of tricks to knock it back into order.
Once Elizebeth broke the code on a circuit, she solved every message that she came across, no matter how trivial or personal. Sinister messages, personal messages, sober messages that spoke of bombs and guns and ships and submarines, messages that just seemed bizarre. (Berlin to South America: “Can you procure details about the process of making explosives from cacao?” Cacao is the main ingredient of chocolate.) She learned intimate details of the spies’ lives. Berlin informed a spy in Iceland that his wife, Erika, had given birth to a healthy baby girl, Jutta. Several times a family member of a spy was allowed to transmit a personal message from Berlin. One of the wives shared Elizebeth’s first name, although she spelled it in the traditional way. “My dear JOHNY: My heartiest congratulations on your birthday and thousands of loving greetings and kisses from your Elizabeth.”
Elizebeth Friedman solved and disseminated these messages like all the others. Such were her weapons against fascism: pencils, puzzles, circuits, names, dates, places, check marks, handwritten notes affixed to typed pages with staplers, stacks of solved messages rising with the hours and days and weeks.
And she stepped lightly as she worked. Her name did not appear on the documents of the Cryptanalytic Unit. She wasn’t the top commander anymore, so she didn’t write official memos to other parts of the intelligence community (Lieutenant Jones did that), and she didn’t meet directly with FBI officials (Jones did that, too). Jones’s name was the name typed on memos about clandestine radio traffic that circulated within the navy’s OP-20-G. And although Elizebeth was not shy behind closed doors, sometimes quarreling with Jones about the direction of their work, disagreeing about which puzzles were more urgent or less urgent to tackle (she found his judgment clouded sometimes by careerism, a hunger for promotion that was irritating), she didn’t mind being anonymous on the page. Her experiences as a cryptologic celebrity in the 1930s had convinced her that in this secret world, attention was a kind of poison. She said after the war that she considered herself “one of the workers” in the Cryptanalytic Unit—she was certainly being paid like a worker, not a leader, earning $4,200 a year, equal to $63,000 today, as a P-5 civil servant, a middle classification of government employee—and this combination of Elizebeth’s tendency to minimize her own contributions and Jones’s role as commander is one reason that her role in the war would go undiscovered for so many years.
Still, unavoidably, she left fingerprints in the records, little traces of her labor. Her initials, ESF, were typed at the bottom of some coast guard decrypts. Her handwriting appeared on many notes stapled to the decrypts, preceded by a code name, “GI-A,” that meant “analyst” and was shared by others in the unit, and sometimes she wrote directly on the decrypts themselves, a characteristic burst of red or blue colored pencil next to a particularly crucial name or phrase, scribbled in the distinctive slant of Elizebeth’s hand that had not changed since she was a college student writing in her diary.
At the end of the day, the unit was hers, not Jones’s. She was the one who first proposed its creation in 1930. She staffed it. She trained the cryptanalysts. She guided its work over the years, building it into a powerhouse. This gave her an informal authority. People who had been paying attention knew that she was the beating heart of coast guard codebreaking. In December 1942, two British intelligence liasons met with Elizebeth in Washington to talk about closer coordination between Bletchley Park and the Coast Guard Cryptanalytic Unit. One of the liasons, Major G. G. Stevens, described the meeting in a “Most Secret” cable to his superiors in Britain and suggested that they arrange for a coast guard representative to visit Bletchley. “For this one would like to see Mrs. Friedman go,” the British official wrote, “but probably at this end”—i.e., the Washington end—“it would be considered more suitable to send Lieutenant-Commander Jones, the official head of the section.”
Given the tension of her job, the pressures of extreme secrecy, and the bleary afternoons stacking alphabets, Elizebeth was glad for her friendship with F. J. M. “Chubby” Stratton, the British astronomer and radio expert who looked like Santa Claus. He was a calming presence. He had a habit of appearing at Elizebeth’s desk at unexpected times. She would be at her desk, nose buried in some problem, and look up to see him there, smiling. Without ever revealing anything about himself, he had a way of making you feel like you had known him forever, and that everything would be OK. And, of course, he was brilliant—a bit of a tinkerer. He loved the challenge of locating hidden radio stations. He had invented a device he called a “snifter,” a little piece of electronics that fit in the pocket and allowed a person on foot to covertly pinpoint the exact location of a pirate radio transmitter within a building, accurate to the floor. He managed a staff of direction-finding experts
who cooperated with the FCC to narrow down the location of pirate signals.
Through terrific effort, teamwork, and long hours, the coast guard codebreakers and their partners finally regained mastery over the clandestine circuits after an unpleasant period of blindness. By winter 1942, Elizebeth’s eyes had adjusted. In the dark, lights out, she was watching them now, one letter of the alphabet at a time. She had the Nazi brain in a jar on her desk, alive and glistening, electrodes running out to her pencil. The unit had conquered every new circuit, every new code that came online that year.
Except one.
The circuit exchanged its first messages on October 10, 1942. The messages seemed to resist solution. She wondered if it might be an Enigma circuit, the messages encrypted by an Enigma machine of some kind.
She called it Circuit 3-N.
Presumably the messages on Circuit 3-N were sensitive enough to require a stronger-than-usual code, so she guessed that the messages must be important. She was more correct than she knew: in the end, the fate of the Invisible War would turn on the sinister frequencies of Circuit 3-N.
For now the anxiety of an unsolved puzzle was more than enough to motivate her. Here were some messages she couldn’t read and she wanted to read them. The code simply had to be smashed.
At weekly radio intelligence meetings with her British and FCC colleagues, she talked about Circuit 3-N, and all the agencies chipped in with clues. The FCC determined that one of the stations was located in Europe and the other in South America, and the British confirmed the FCC’s bearings.
New intercepts from Circuit 3-N arrived in Elizebeth’s office each week. By December 1942 she had accumulated twenty-eight encrypted messages. A cursory analysis showed telltale signatures of an Enigma machine.