Navy Cryptology: The Early Days
"All nations have learned the lessons of the World War and will probably make even greater efforts to intercept and read enemy messages in the future than were made in the past…. In a war between nations of approximately equal strength, Radio Intelligence could easily become the decisive factor." -- Laurance Safford, speech, Naval Postgraduate School, 1925
During World War I, the U.S. Navy engaged in "radio intelligence" only in a limited way. Navy personnel attempted, unsuccessfully it appears, to track German submarines by direction finding. Primarily, the Office of Naval Intelligence (ONI) cooperated with MI-8, the cryptanalytic office in the Army's Military Intelligence Division. At the end of the First World War, the office was staffed by 13 officers, 14 male enlisted personnel, and 34 "yeomanettes," a term used at that time for enlisted women.
It is reported that this office solved a Japanese diplomatic code before the end of the war, but the project was abandoned because of a lack of traffic and a lack of linguists. No further information is available about this early effort. The office was quickly downsized at the end of the war.
Just before the end of World War I, ONI established a Research Desk in the Code and Signal Section to engage in cryptanalysis. At the end of the war, because cryptanalysis was seen as an important adjunct to preparing secure systems for one's own use, the Research Desk was kept in existence and placed under the Director of Naval Communications (DNC). Under an agreement between ONI and DNC, the Research Desk was responsible for intercept and cryptanalysis, while ONI would handle translation, evaluation, and distribution.
In 1918, the Code and Signal Section of ONI acquired spaces in the sixth wing of the Navy Headquarters Building on Constitution Avenue in the District of Columbia. This would remain its home until February 1943, when the Navy acquired the Mt. Vernon Academy on Nebraska Avenue, and moved the Naval Communications Annex there.
In 1922, the office was given the designator OP-20-G, which it would carry until shortly after World War II.
In the post World War I period, the Research Desk was authorized one officer, two enlisted men, and four civilians. The officer billet was filled in 1924 with the assignment of Lieutenant Laurance F. Safford. A few months later, Miss Agnes Meyer -- who had been among the "yeomanettes" during the war -- became one of the civilian employees.
Agnes Meyer (married name Driscoll) became one of the key figures in Navy cryptanalysis. In addition to her experience during the war, she had some training with Edward Hebern's private code company, and, it is believed, spent some time with Yardley's "Black Chamber." Edward Hebern made cryptographic machines for sale to businesses, although the machines were not secure enough to interest the government. The time spent with Hebern, however, gave Ms. Meyer invaluable experience with the new concept of machine ciphers. Because officers were required to rotate out of OP-20-G every few years for sea duty, Ms. Meyer provided continuity on targets. She also trained many of the important Navy figures in cryptology.
Safford, who eventually rose to the rank of captain, is regarded by many as the "father" of Navy cryptology. Throughout the 1920s he strongly advocated more effort for Communications Intelligence, took steps to acquire intercept for analysis, and conducted training for good cryptanalytic prospects.
Safford recruited promising cryptanalysts by putting puzzles in the Navy's monthly Communications Bulletin, beginning in mid 1924. Over the years, he recruited many who sent in successful solutions.
The first pupil to come to Safford for training, in 1924, was Ensign Joseph N. Wenger. A more formal class in all aspects of communications intelligence and communications security was begun the following year; among the students in that group of five was LT Joseph J. Rochefort. Both Wenger and Rochefort went on to make significant contributions to American cryptology.
It should be noted that the first formal class, which ran from October 1925 to January 1926, was comprised of four Navy officers and one from the Marine Corps. The Marine Officer, however, never received a COMINT assignment.
Among other actions, Safford used a secret ONI fund to purchase five typewriters; he had them modified and sent to the Asiatic Fleet to facilitate copying Japanese kana characters by Navy radiomen. Over time, about 40 units were purchased, and eventually, this machine came to be called the RIP-5. Some of these machines were shared later with the counterpart organization in the Army.
When Safford rotated out for sea duty in 1926, he was replaced on the Research Desk by LT Joseph Rochefort. In turn, Rochefort was replaced in September 1927 by Lt. Bern Anderson. Safford returned from sea duty to the Research Desk in August 1929.
The Chief of Naval Operations recognized the need for radio intelligence, and, in a letter dated May 13, 1929, indicated his intention to establish a radio intelligence office with the Asiatic Fleet and to organize cryptanalytic units afloat. He noted that Navy war plans were being modified to include major cryptologic units in Washington and Hawaii, and the establishment of a special cryptographic system strictly for COMINT.
The increased emphasis on COMINT operations paid off early. Larger numbers of intercept operators enabled the U.S. fleet to copy a considerable volume of radio traffic from the Japanese fleet's 1930 Grand Maneuvers. The traffic was forwarded to Washington, where it was worked on by LCDR Safford, LT. Wenger, and Mrs. Driscoll. These messages revealed Japan's battle plan against the United States, Japanese fleet mobilization procedures, and Japanese plans for defense of the western Pacific. The surprised Americans also learned that the Japanese had an excellent grasp of American war plans for the Pacific.
In 1933, Lt. Joseph Wenger compiled all the American intercept of that year's Japanese fleet maneuvers, and, through traffic analysis over a period of six months, produced a report on the composition of the Japanese fleet. The results of three years of cryptanalysis on the same messages eventually proved Wenger's list correct.
(Often, in the period before war broke out, traffic analysis was the only way to track movement and other activity by the Japanese fleets. Traffic analysis also was of considerable assistance to the cryptanalysts. Once war broke out, the knowledge gained over time in the pre-war period about radio patterns and callsigns proved valuable in working Japanese traffic.)
In 1930, the Director of Naval Communications, Captain Stanley C. Hooper sent OP-20-G a memorandum suggesting the use of tabulating machines for "mechanical labor" such as counting frequency distributions and repetitions. LCDR Safford and Ms. Driscoll discussed the idea and dismissed it as unfeasible due to the unpredictable nature of ciphers.
Captain Hooper, however, did not let the matter drop and arranged for Lt. Wenger to attend demonstrations of modern office equipment. Among the machines demonstrated, Wenger saw possibilities in a Hollerith sorter distributed by the Tabulating Machine Company. He, in turn, persuaded Safford and Meyer to get a demonstration, and, after further discussions, OP-20-G got funds to rent a punch machine and a sorter.
Wenger himself did not get to use it. His replacement, Lt. Thomas Dyer, became the first U.S. cryptanalytic machine operator.
The Navy itself was small, and Navy cryptology began with a very small organization -- even by 1941, OP-20-G had only about 60 persons plus small field contingents. But it had backing at the highest levels, and by the outbreak of the Second World War, the organization had become a respected component of Navy operations.
Joseph Wenger at the end of his career became Vice Director of the Armed Forces Security Agency, NSA's predecessor, and was promoted to Rear Admiral, the first Navy cryptologic officer to achieve flag rank.
Incidentally, the Tabulating Machine Company later became IBM Corporation.
- Doug Adams, "The Early Years," NCVA (Paducah, Ky.: Turner Publishing Company, 1996), pp. 8-10.
- Captain J. S. Holtwick, Jr., USN (Ret), Naval Security Group History to World War II, June 1971, SHR-355, CCH Files.
- Carl Jensen, "Traffic Analysis," NCVA (Paducah, Ky.: Turner Publishing Company, 1996), pp. 26-27.
- James McIntire & R. D. Howell, Sr., "U.S. Marine Corps in COMINT," NCVA (Paducah, Ky.: Turner Publishing Company, 1996), pp. 46-50.
- Albert Pelletier, "Cryptography -- Target Japan," NCVA (Paducah, Ky.: Turner Publishing Company, 1996), pp. 27-32.
- J. N. Wenger, U.S. Navy Communication Intelligence Organization, Liaison and Collaboration 1941-1945, 8 October 1945, SRH-197, CCH Files.