Reconstructing an enemy codebook is called “book-breaking,” and Ann Caracristi, working with the linguists, turned out to be stunningly good at it. By summer 1943, they knew that 1113 stood for shibucho, 1292 for taichoo, 1405 for butai, and 3957 for bukkan. They dug out the code groups for Hiroshima, Singapore, Kupang, and Tokyo.
The address codes carried a bounty of operational military information. The work that Ann Caracristi and Wilma Berryman were doing enabled U.S. military intelligence to construct what is called the order of battle: an accounting of the strength, equipment, kind, location, and disposition of Japanese Army troops. They were able to pinpoint where the enemy was quartered and headed. Soon “MacArthur’s headquarters had as good a picture of the Japanese military set-up as he had of his own,” as Solomon Kullback put it. As it grew in importance, the address unit grew in number. In addition to Ann and Wilma, it included women with names like Olga Brod, Bessie E. Grubb, Edna Kate Hale, Mildred W. Lewis, Esther A. Sweeney, Lottie E. Miller, Bessie D. Wall, Violet E. Bennett, Goldie M. Purks, Mabel J. Pugh—and, yes, a jewel, Ruby C. Jones.
“That outfit was 100 percent female,” Solomon Kullback said later. “There were only girls working on that address system.” There were two military men assigned to the unit—Reuben Weiss and Mort Barrow—and the pair used to circulate, joking and stoking morale, showing up in the middle of the night to raise the spirits of the graveyard shift and jolly the women along. There were “only two officers,” Kullback recollected, “and at least a hundred girls.”
Wilma’s team worked alongside a unit called “traffic analysis,” also mostly female. It was the job of traffic analysis to follow fluctuations in Japanese Army message traffic, without worrying about the actual message content. External information helped in developing the order of battle. If a flurry of radio messages began going back and forth to a new location, this meant somebody was on the move. Before long, Wilma would find a representative of G-2—military intelligence—standing behind her daily, urging her on and telling her which addresses to concentrate on. Arlington Hall began to produce a daily order-of-battle summary. Every morning at five o’clock there was a “black book session” discussing the report that would be compiled each evening and taken to the Pentagon.
For the women code breakers, it was exhilarating. The same address code was appended to all the different Japanese Army systems, so soon enough every kind of message—air force, administrative, water transport, ground troops—was passing through their hands. “I think that’s one of the things that made it so much fun. We saw everything, everything had to come through us,” Wilma later said. “You had cryptanalysis, you had traffic analysis, and you had order of battle, and you could hardly ask for a better job. You were sitting right on top of the world and you were following the Japs all over hell’s half-acre, all the time. You weren’t just reading something about, ‘Send me three pounds of sugar and ten pounds of rice.’ Or something like that. I thought it was the most exciting job in the place.”
Staying current was a never-ending task, though. As with the Navy, the Japanese Army routinely changed both its books and its methods. There was a grim period when the radiomen began using one method to encipher addresses that had an odd number of code groups, and another method to encipher those with an even number. The address unit seemed on the verge of collapse.
“We were in an awful pickle, because it was war,” Wilma Berryman later recalled. “They had been getting such beautiful order of battle, and all of a sudden, just like that, it was wiped out. And so we had to do something about it.” Wilma and Ann began working on traffic that had an even number of group counts, and they were on the verge of being able to read it when Solomon Kullback came by and said he would help. Their boss told them to gather up half of the messages and bring them into his office.
“I’ll take care of the evens, you take care of the odds,” he told them.
As grave as the situation was, the two women thought this was extremely funny. Given all the work they’d just done on the even-numbered code groups, their boss was getting by far the easier job. They had to dash away so they wouldn’t lose their composure. “Annie and I just ran, it was so funny,” said Wilma. For years after that, all Wilma had to say was, “You take care of the odds, I’ll take care of the evens,” and Ann would crack up.
Nothing could shake loose Ann Caracristi, who, even in this elite group, was in a class by herself. “My capabilities compared with Ann’s were nothing,” Wilma said. Wilma put Ann in charge of a small research team that included a mathematician named Anne Solomon and a male Harvard graduate, Ben Hazard, who had a physical condition that disqualified him from service. Naturally, they nicknamed him Hap. They were young but formidable. As the Americans began to retake Pacific islands, the Japanese had a hard time distributing new codebooks to isolated outposts. Sometimes they could sneak the codebooks in by submarine, but often the marooned cryptographers had to devise a jury-rigged solution, using old books in a new way. The Japanese would make squares, in which they would take additives from the old book, and—rather than adding them—run them vertically and horizontally to construct a table. Squares are hard to break for a number of reasons, among them the fact that there is no mathematical relationship between the underlying code group and the enciphered group that gets transmitted, because no addition has been done.
But group counts and sum checks do create patterns, and Ann’s scanty training included some preparation for breaking squares. The first time she encountered this kind of enemy innovation, she sat puzzling over it. Solomon Kullback was instantly at her side, urging her on as the solution came into view. “It was fascinating, actually, to work in the world, become aware of it, and find you can really do something useful, without being a mathematician or linguist,” Ann recalled many years later.