In 1938, the Japanese Navy changed its operational code, replacing a system that had been in use since 1931. The U.S. Navy terminology for the superseded system had been the BLUE CODE, based on the color binding used for the American recoveries to the system; the new code was designed BLACK. The BLACK CODE introduced the use of an additive book to superencipher messages.
The BLACK CODE did not last long. The Japanese Navy introduced two new enciphered codes for general purposes in 1939; American cryptanalysts dubbed one of the systems the "Flag Officers Code" and the other, JN-25. Both systems were worked by Navy cryptanalysts, although the U.S. Navy got little intercept in the Flag Officers Code and the effort against it was abandoned in December 1941.
The Flag Officers Code was never solved by the Americans. JN-25, however, became one of the most widely used Japanese Navy systems -- and, eventually, a critical source of intelligence for the Allies.
JN-25 consisted of a codebook with approximately 27,500 entries and an additive book for superenciphering the codebook values. The additive book consisted of 300 pages, each page containing 100 random five-digit groups. It should be noted that this additive book for JN-25 was not a one-time pad: the five-digit groups were re-used, as needed.
In studying JN-25, U.S. cryptanalysts had to collate large numbers of Japanese messages over time. Their first goal was to recover the indicator in each message which showed where in the additive book numbers were taken; then recover and strip away the additives themselves to get down to the codebook values; and, finally, recover the meanings in plaintext Japanese of the underlying codebook values, which would allow messages to be read, at least in part.
Navy cryptanalysts used IBM machines to correlate and compare values in JN-25-based communications. This allowed them to identify depths, which were the key to recovering the numerals used in messages. This method also enabled them to identify a "garble check" the Japanese used to ensure each message had been copied and decoded correctly. This was of prime value in helping to strip away the additive numbers and get to the underlying codebook values.
Once the additive had been stripped away, U.S. cryptanalysts took advantage of stereotyped messages to recover some basic text. Commonly repeated texts included vessel port entry requests, daily position data for convoys, and medical reports.
Decryption of JN-25 was facilitated by information from plain language Japanese messages with ship movement information. American cryptanalysts also found the Japanese Naval Officer's Register helpful for recovering code groups.
A second version of this system, known to U.S. Navy cryptanalysts as JN-25B, was introduced on December 1, 1940. Six months later, the Japanese Navy replaced the additive book. The additive book was replaced in August and again on 4 December 1941, three days before the Japanese attack on American bases in Hawaii.
These rapid changes in the codebook and its additive required that U.S. cryptanalysts begin again with each change -- virtually at the beginning -- to attack the system. It is estimated that prior to the change of the additive book in August 1941, the cryptanalysts had recovered only 2,000 code groups in JN-25 -- about 4% of the codebook -- and these were mostly numerals and stereotyped phrases.
During the Currier-Sinkov mission to Great Britain [see tomorrow's installment], as U.S.-UK cooperation began, the British authorized their cryptologic unit at Singapore to begin exchanges about JN-25. It turned out the Americans and the British had had roughly equal success in recovering JN-25 codebook values -- but the recoveries were in different parts of the codebook, so, at a stroke, each doubled their number of recoveries!
Even with this bonanza, however, there were too few recoveries to begin reading the underlying text.
One factor limiting progress was the limited staffpower available for cryptanalysis. Another factor was priorities: since America's political leaders were avid readers of the PURPLE diplomatic decrypts, the major effort in Washington was to process the diplomatic system.
The Army, by the way, was not making comparable efforts against Japanese Army communications because of a lack of traffic. While the Japanese Navy was deployed throughout the Pacific and was forced to communicate by radio over long distances, the Japanese Army's overseas deployment in the late 1930s was largely in China; their long-distance communications were less in volume than the Navy's, and in areas less vulnerable to U.S. intercept.
Solution awaited more traffic and more CA personnel.
Sources:
- Robert L. Benson, A History of U.S. Communications Intelligence During World War II: Policy and Administration (Center for Cryptologic History, 1997), pp. 20-21.
- Stephen Budiansky, Battle of Wits: the Complete Story of Codebreaking in World War II (New York: The Free Press, 2000)
- Frederick D. Parker, Pearl Harbor Revisited: United States Communications Intelligence, 1924-1941 (Center for Cryptologic History, 1994).
- Captain Duane L. Whitlock, U.S. Navy, Retired, "The Silent War Against the Japanese Navy."