William F. Friedman

William Frederick Friedman (September 24, 1891-November 12, 1969) was a US Army cryptologist, running the research division of Army's Signals Intelligence Service through the 1930s and its followon services right into the 1950s. He personally cracked the Japanese Purple code after noticing a seemingly unimportant pattern in the cyphertext it generated, thereby leading directly to the destruction of the IJN's fleet at Midway Island. He is considered by many to be one of the greatest cryptologists of all time, and his application of statistical methods to code breaking to be one of the greatest advances in the field. He is also responsible for much of the language used in the field, introducing terms such as cryptography and cryptanalysis.

Friedman was born in Russia, the son of postal the worker who migrated to Pittsburgh in 1892. He studied at the Michigan Agricultural College in East Lansing and received a scholarship at Cornell University to work on genetics. Meanwhile George Fabyan, who ran a research lab to study any project that caught his fancy, decided to set up his own genetics project and was referred to Friedman. Friedman joined Fabyan's Riverbank Laboratories[?] outside of Chicago in September 1915. As head of the Department of Genetics one of the projects he ran was to the effects of moonlight on crop growth, and so he experimented with the planting of wheat during various phases of the moon.

Another of Fabyan's pet projects was funding Elizabeth Wells Gallup's research into Francis Bacon's coded messages he hid in various texts when he ran Elizabeth I's printing press. Gallup was convinced that Bacon wrote many, if not all, of William Shakespeare's works, based on the fact that many of Shakespeare's works also included hidden messages. Friedman had become something of an expert photographer while working on his other projects, and was asked to travel to England on several occasions to help her photograph historical manuscripts during her research. It was at this point that he became facinated with cryptology, while he courted Elizebeth Smith, Mrs. Gallup's assistant and accomplished cryptologist. They married, and soon after he became the director for the Department of Codes and Ciphers as well as the Department of Genetics at Riverbank.

With the US's entry into World War I, Fabyan offered the services of his Department of Codes and Ciphers to the government. They had no existing federal department for this kind of work (although both the Army and Navy had such organizations), and soon Riverbank became the first unofficial cryptologic organization in the US. During this period the Friedmans cracked a code being used by German-funded Hindu radicals in the US who were planning on shipping arms to India in order to gain independence. Based on the format of the messages they realized that the code was based on a dictionary of some sort, a common encryption technique. They soon managed to decode most messages, but it was not until long after the case had come to trial that the book itself was discovered, a German-English dictionary published in 1880.

The government decided to set up their own official code breaking service, and sent Army officers to Riverbank for training by Friedman. For instruction Friedman turned out a series of technical monographs, completing seven by early 1918. He then enlisted in the Army, and was sent to France to be the personal code-breaker for General John Pershing. He returned to the US in 1920 and published an eighth monograph, which is considered to be the most important single publication in modern cryptology.

In 1921 he joined the government's American Black Chamber[?] where he was placed in charge of researching new codes and ways to break them, and in 1922 promoted to head the Research and Development Division. The Black Chamber was dissolved in 1929, and Friedman moved to the Army's Signals Intelligence Service in a similar capacity.

During the 1920s a series of new codes produced by machines started to be built, based largely on typewriter mechanicals attached to basic electrical circuitry – batteries, switches and lights. One of the first of these was the Hebern Rotor Machine designed in the US in 1915 by Edward Hebern. The system was so secure and simple to use that he invested heavily in producing the machines, feeling that all companies would want them. But the company went bankrupt when the war ended, and Hebern eventually landed in prison for (unreasonable) claims of stock manipulation in his own company.

Friedman realized that this sort of rotor machine was going to become popular, and devoted time to cracking Hebern's design. Over a period of years he discovered a number of problems that would be common to any rotor machine built with a few specific design features, namely having the rotors turn once with every keypress, and having the fast rotor (the one that turns with every keypress) at either end of the rotor stack. In this case the output generated by the machines will have strings of 26 letters that are a simple substitution cipher, and by collecting enough cyphertext and applying a standard statistical method known as the kappa test[?], he showed that it was possible, although very difficult, to crack any such machine.

Friedman then used his understanding of the rotor machines to develop several of his own that were immune to his own attacks. Nine designs were eventually developed, six of them remaining secret to this day. The culimation of various earlier designs resulted in the SIGABA, which became the US's standard encryption system during World War II, and was far more secure than anything used by any other nation during the war. In fact it appears to remain unbreakable even today, except by brute-force methods on computers.

In 1939 the Japanese introduced a new code system for their embassies, replacing an earlier system SIS referred to as Red. The new code, referred to as Purple, proved to be impossible to crack. Friedman was convinced it was a mechanical system, and set about attacking it, spending several sleepless months studying the cyphertext trying to discover its patterns. Then one day, he had it. The rotors were not based on a 26-position disk like the Herbern machine or Enigma, but used telephone stepper switches used in automated telephone exchanges and thus had strings of 10 characters with the "same code".

By the end of 1940 his team at the SIS had constructed an exact duplicate of the Purple machine, even though they had never seen one. With an understanding of Purple and duplicate machines of their own to use, the SIS was then able to decrypt an increasing amount of the Japanese traffic. One such intercept was the the orders on December 7th 1941 to break off negotiations with the US, a clear indication of war only hours prior to their attack on Pearl Harbor. The decryption of Purple was particularily important in the Battle of Midway, when the US was aware of Isoroku Yamamoto's plans long before the battle actually occurred, thereby tipping the Pacific war in their favour.

After solving the Purple machine, Friedman was admitted to hospital in 1941 for a nervous breakdown, the result of months of frenzied effort. After his recovery, he served as Director of Communications Research for the SIS for the rest of the war. Friedman visited the British code-breaking operations at the Government Code and Cipher School at Bletchley Park in 1941. He exchanged information on his techniques for attacking Purple for the British information on how they had broken the Enigma. From that point onward, the Americans and British high commands were able to obtain detailed information on all major German and Japanese military and naval operations before they occurred.

Following the war, Friedman remained in intelligence. In 1949 he was made head of the code division of the newly-formed Armed Forces Security Agency (AFSA), and in 1952 become the chief cryptologist for National Security Agency (NSA) when it was formed to take over from the AFSA.

He retired in 1956 and turned his attention, with his wife, to the problem that had originally brought them together: examining Bacon's codes. In 1957 they authored The Shakespearean Ciphers Examined, in which they demonstrated that Gallup's work was sadly mistaken. His health began to fail in the late 1960s, and he died in 1969.

Elizebeth Friedman was also heavily involved in cryptography, although typically on the civilian side of things. During the 1920s she gained some fame for repeatedly breaking the codes being used by "rum runners" bringing alcohol into the US during Prohibition, and in 1927 was hired by the US Coast Guard to help them with their policing operations. By 1930 she had cracked over 12,000 messages for the Coast Guard, Bureau of Customs, Bureau of Narcotics, Bureau of Prohibition, Bureau of Internal Revenue, and the Department of Justice.

In 1934 she was involved in a particularily odd case, in which a Canadian-registered ship, the I'm Alone, was sunk after being chased into international waters from the US. She decoded several messages that demonstrated the ship was actually paid for by an American, and thereby ostensibly US-owned. The result was to expand the law regarding police chases, allowing a ship involved in illegal activity to be followed into international waters, and thereby extracting the US from an embarasing political scandal.

During World War II she moved to the OSS and was one of their chief cryptologists. She was involved in a particularily famous case in which a husband and wife team were sending coded messages to the Japanese, writen on dolls that the wife sold as a thriving mail-order business. Velvalee Dickinson became known as "The Doll Woman" when the case was broken to the press. Elizebeth retired after her husband's death in 1969 and lived on until 1980.

External links:

• http://www.nsa.gov/honor/w_friedman
• http://www.sans.org/rr/history/friedman.php