Codes have been around for as long as civilization, maybe longer. Virtually as soon as humans developed the ability to speak and write, somebody somewhere felt the desire to say something to somebody else that could not be understood by others. The point of a coded message is to engage in intimate, often urgent communication with another person and to exclude others from reading or listening in. It is a system designed to enable communication and to prevent it.
Both aspects are important. A good code must be simple enough to be readily used by those privy to the system but tough enough that it can’t be easily cracked by those who are not. Julius Caesar developed a cipher in which each letter was replaced by a letter three spaces ahead in the alphabet (A would be changed to D, B to E, and so forth), which met the ease-of-use requirement but did not satisfy the “toughness” standard. Mary, Queen of Scots, used coded missives to communicate with the faction that supported her claim to the English throne, which—unfortunately for her—were read by her cousin Elizabeth and led to her beheading. In medieval Europe, with its shifting alliances and palace intrigues, coded letters were an accepted convention, and so were quiet attempts to slice open diplomatic pouches and read them. Monks used codes, as did Charlemagne, the Inquisitor of Malta, the Vatican (enthusiastically and often), Islamic scholars, clandestine lovers. So did Egyptian rulers and Arab philosophers. The European Renaissance—with its flowering of printing and literature and a coming-together of mathematical and linguistic learning—led to a number of new cryptographic systems. Armchair philosophers amused themselves pursuing the “perfect cipher,” fooling around with clever tables and boxes that provided ways to replace or redistribute the letters in a message, which could be sent as gibberish and reassembled at the other end. Some of these clever tables were not broken for centuries; trying to solve them became a Holmes-and-Moriarty contest among thinkers around the globe.
Many Renaissance cryptographers perceived (as counterparts in the Middle East had done before them) that the alphabet itself has underlying mathematical properties: six vowels, twenty consonants, some of these much more frequently used than others. Cipher systems often are created by juxtaposing or “sliding” two or more alphabets against each other so that B in one alphabet lines up with, say, L in the other, and C with M, D with N, et cetera. One of these, the Vigenère square, developed in the 1500s and named for the French diplomat Blaise de Vigenère, achieves this by creating a twenty-six-by-twenty-six-square letter table, in which twenty-six alphabets are stacked on top of one another, each alphabet beginning with a different letter, with a keyword telling which alphabet to select for each letter to be changed. America’s great innovator Thomas Jefferson dabbled in secret systems, inventing a cipher wheel that could transform one letter into a new one; it was discovered in his personal papers more than a hundred years after his death.
During the American Revolutionary War, many code systems were employed not only by diplomats and statesmen but also by spies and traitors. Sometimes, of course, diplomats and statesmen were the spies and traitors. Jefferson and Ben Franklin at times used coded language, as did Benedict Arnold. During the American Civil War, the military began to experiment with codes and ciphers. Union commanders sent messages by means of a soldier working a little handheld disk. The disk changed each letter to a new one, and the new one would be transmitted by a soldier waving a big flag in the direction of the intended recipient, in a signaling cipher called wigwag. Confederates used a cipher so complicated that they became confused by it. Sometimes they also would intercept Union messages, publish them in the newspaper, and invite readers to submit solutions.
Elizebeth Smith and William Friedman were fascinated by this history—and by each other. They took long bicycle rides, swam in the Roman pool, drove through the countryside in a Stutz Bearcat. Theirs was an early example of what sociologists nowadays call homogamy, which is marriage between equals. Elizebeth found William to be sleek and sophisticated; he found her vibrant, smart, and dynamic. The two married in May 1917, less than a year after they met, and Elizebeth moved into the windmill.
Despite the newly married pair’s skepticism about Gallup, the soirees continued, though the topic expanded beyond Bacon to encompass codes and ciphers and their solution. From time to time, Fabyan liked to summon a University of Chicago English literature professor, an amateur cryptanalyst, John Manly, and pit him against Elizebeth Friedman. Cryptanalysis at that time was a parlor game. The operation gained a certain cachet: Movie stars sometimes came to tour Riverbank. The Friedmans developed genuine expertise, attracted freelance assignments, and won acclaim when they decoded a batch of correspondence that helped expose a conspiracy between Hindu separatists and German agents trying to foment revolution against the British in India. Elizebeth was disappointed that William was the one who got to testify in the trials (during which one defendant was shot dead); though they had done the work as a team, she had to stay behind because, as she put it, “someone had to oil the machinery at Riverbank.”
War, though, began to transform the tenor of the operation. As unlikely as the setting was, Fabyan’s estate incubated the first serious cryptanalytic efforts of the U.S military. In 1916, Fabyan began to sense that America’s involvement in world war was in the offing. In his voluble self-promotion he had devoted much energy to cultivating ties with important people in Washington, often taking William Friedman with him on visits. Domineering and massive, Fabyan was friends with another massive man, Joseph Mauborgne, an Army officer and radio enthusiast who had solved the notorious Playfair cipher—a table in which pairs of letters are substituted for other pairs, used as the primary cipher system of the British Army—during a six-month trip on a steamer. Now, thanks to Mauborgne’s influence, the U.S. Army was taking a greater interest in code making as well as code breaking.