Participation by the U.S. Navy in World War 1 was too limited for any degree of SIGINT development, but interest was stimulated by successful British efforts. During that war, the Navy relied on the British Admiralty for intelligence information in Europe.

In the spring of 1920, however, naval intelligence financed a series of break-ins by the FBI at the Japanese consulate in New York. These “black bag” jobs were repeated several times in the 1920s and were intended to acquire evidence of subversive activities in the United States. The unexpected result was the acquisition of the complete Japanese fleet code book, photographed during several visits to the consulate.

With this code in hand, the Navy in 1924 established an intelligence unit within the Office of Naval Communications and named it OP-20-G, or the “Research Desk.” OP-20-G had, as its initial task, to exploit the principal code of the Imperial Navy – the cryptanalysts called it the “Red Code,” after the color of the original books.  (The Japanese used this code until early 1931.)

Possession of the Red Code precipitated the establishment of a network of intercept stations in the Pacific to acquire radio traffic for the cryptanalysts – the first station was at Shanghai in 1924. Prior to this time, Washington received only a handful of Japanese messages at irregular intervals sent by naval communicators or radio hams on their own initiative. The only U.S. Naval intercept units in 1924 were at the U.S. consulate in Shanghai and some limited intercept capability aboard the USS Huron (a). The unit at Shanghai was begun with self-trained general communications service operators copying Japanese traffic in their spare time.

In Washington, Lieutenant Laurance F. Safford, head of OP-20-G, established the primary course in cryptanalysis for the Navy. The first solution of Japanese cipher, using the Red Book, was accomplished in 1926 by naval cryptanalysts. By 1927, the Navy had established stations in Shanghai, Beijing, and Hawaii. Other stations were organized at Guam in 1929 and in the Philippines in 1930.

Part 2 of 6 Laurance Saffordx
Lieutenant Laurance F. Safford, head of OP-20-G

The development of the Navy COMINT organization now moved forward rapidly. Lieutenant Commander Ellis M. Zacharias, the officer-in-charge at Shanghai in 1926, used the USS McCormick to test the feasibility of regularly intercepting traffic from afloat during that year. In 1927, Zacharias, aware that major Japanese maneuvers were in progress, convinced the CNO and Commander-in-Chief Asiatic Fleet (CINCAF) to use the USS Marblehead and ground-based stations to intercept Japanese fleet communications. He returned to Shanghai in late 1927 and wrote the first comprehensive intelligence report on the Combined Japanese Fleet in action.

The first phase of the 1927 maneuvers extended from 10 to 20 October. The USS Cincinnati, cruising off the southern tip of Korea, intercepted tactical signals and general service traffic from a position about 400 miles from the maneuver area. Most of the traffic was encoded but the submarine traffic was in plain language. Sixty tactical calls used by the Japanese were recorded by the Cincinnati. According to Zacharias, “by means of the address in the messages, (it was) possible to identify the majority of (the tactical calls)” – but only after the maneuvers were completed.

During the second phase, the USS Marblehead, steaming between Shanghai and Nagasaki and then to Kobe via the inland sea, intercepted Japanese fleet communications from 20 to 30 October. The Navy quickly realized that the ship’s operators had no knowledge of Japanese code or procedures and therefore could not identify and record callsigns. Direction Finding (DF) was atrocious (a 10 degree error was routine) and the receiving sets were not calibrated. The Navy did, however, copy about 100 pages of encoded tactical traffic for the cryptanalysts.

Zacharias, a Japanese linguist, observed that:

tactical calls (are) allocated alphabetically and numerically according to types. Addresses always precede (the) body of messages (and) appear to be parts of the names of ship or stations to which addressed. Various departments (can be identified) by KA (Commanding Officer) (and) CHI (Commandant). (We) identified the (RED – 3rd Fleet) flagship address (IN) since this ship tuned frequencies for the rest of the ships.  Later (this was) corroborated by work of JN with units. Japanese used too much calling through whole lots of tactical calls to see if (there was) any traffic. When a tactical call didn’t answer, the Japanese switched to international calls and signatures which helped identify tactical calls.

The final report from CINCAF for the 1927 maneuvers concluded:

1. Intercept from the ships was valuable in that a tremendous amount of traffic was obtained which would assist cryptanalysis.

2. The Japanese radio operators were undisciplined regarding radio procedures.

3. Due to central control being exercised from Tokyo, radio silence was impossible.

4. The Japanese system of codes, tactical calls, and addresses would not allow for radical changes without disrupting their communications.

5. The U.S. Navy had no effective DF to track vessels.

6. The Japanese did not use HF communications.

Zacharias recommended to the CNO and CINCAF that the U.S. Navy (1) cover future Japanese maneuvers, (2) train operators to intercept Japanese communications (a), and (3) recruit linguists and cryptanalysts to assist in identifying the traffic. At this time, identification and analysis of traffic was accomplished in a very limited fashion by either the operator, the linguist, or the cryptanalyst – there was no consideration given to a separate SIGINT discipline for traffic analysis. For the Navy at least, that was not to come until the maneuvers of 1930.

Source: NSA

(a) Station HYPO note: Recommendation #2 was made one year prior to the establishment of On-The-Roof training located in Washington D.C. at the Navy Main Building.

(b) USS Huron is featured image in Shanghai, China, December 19, 1924