The vulnerability of Japanese naval codes and ciphers was crucial to the conduct of the Pacific War of World War II, and had an important influence on foreign relations between Japan and the west in the years leading up to the war as well. Every Japanese code was eventually broken, and the intelligence gathered made possible such operations as the victorious American ambush of the Japanese Navy at Midway in 1942 (by breaking code JN-25b) and the shooting down of Japanese admiral Isoroku Yamamoto a year later in Operation Vengeance.

The Imperial Japanese Navy (IJN) used many codes and ciphers. All of these cryptosystems were known differently by different organizations; the names listed below are those given by Western cryptanalytic operations.

Red code

The Red Book code was an IJN code book system used in World War I and after. It was called "Red Book" because the American photographs made of it were bound in red covers. It should not be confused with the RED cipher used by the diplomatic corps.

This code consisted of two books. The first contained the code itself; the second contained an additive cipher which was applied to the codes before transmission, with the starting point for the latter being embedded in the transmitted message. A copy of the code book was obtained in a "black bag" operation on the luggage of a Japanese naval attaché in 1923; after three years of work Agnes Driscoll was able to break the additive portion of the code.

Knowledge of the Red Book code helped crack the similarly constructed Blue Book code. but Vice Admiral Katsuo Abe, a Japanese representative to the Axis Tripartite Military Commission, passed considerable information about German deployments in CORAL, intelligence "essential for Allied military decision making in the European Theater."

Jade

A cipher machine used by the Imperial Japanese Navy from late 1942 to 1944 and similar to CORAL.

Dockyard codes

A succession of codes used to communicate between Japanese naval installations. These were comparatively easily broken by British codebreakers in Singapore and are believed to have been the source of early indications of imminent naval war preparations.

JN-11

The Fleet Auxiliary System, derived from the JN-40 merchant-shipping code. Important for information on troop convoys and orders of battle.

JN-20

An inter-island cipher that provided valuable intelligence, especially when periodic changes to JN-25 temporarily blacked out U.S. decryption. JN-20 exploitation produced the "AF is short of water" message that established the main target of the Japanese Fleet, leading to a decisive U.S. victory at the Battle of Midway in 1942.

JN-25

JN-25 is the name given by code-breakers to the main, most secure, communications scheme used by the IJN during World War II. Named as the 25th Japanese Navy system identified, it was initially called AN-1 as a "research project" rather than a "current decryption" job. The project required reconstructing the meaning of thirty thousand code groups and piecing together thirty thousand random additives.

Introduced from 1 June 1939 to replace Blue (and the most recent descendant of the Red code) it was an enciphered code, producing five-numeral groups for transmission. New code books and super-enciphering books were introduced from time to time, each new version requiring a more or less fresh cryptanalytic attack. John Tiltman with some help from Alan Turing at Government Communications Security Bureau (GCSB) had broken JN25 by 1941. They knew that it was a five-digit code with a codebook to translate words into five digits and there was a second "additive" book that the sender used to add to the original numbers. Knowing this did not enable them read a message.

By April 1942 code-breakers could read "about one in five words"; traffic analysis was far more useful. Tiltman had devised a (slow; neither easy nor quick) method of breaking it and had noted that all the numbers in the codebook were divisible by three. "Breaking" rather than "solving" a code involves learning enough code words and indicators so that a message can be read. JN-25 was significantly changed on 1 December 1940 (JN25a) and again on 4 December 1941 (JN25b) just before the attack on Pearl Harbor.

British, Australian, Dutch and American cryptanalysts co-operated on breaking JN-25 well before the Pearl Harbor attack but because the Japanese Navy was not engaged in significant battle operations before then, there was little traffic available to use as raw material. IJN discussions and orders could generally travel by routes more secure than broadcast, such as courier or direct delivery by an IJN vessel. Publicly available accounts differ but the most credible agree that the JN-25 version in use before December 1941 was not more than perhaps 10 per cent broken at the time of the attack, and that primarily in stripping away its super-encipherment. JN-25 traffic increased immensely with the outbreak of naval warfare at the end of 1941 and provided the cryptographic "depth" needed to substantially break the existing and subsequent versions of JN-25.

The American effort was directed from Washington, D.C. by the U.S. Navy's signals intelligence command, OP-20-G; the Navy Combat Intelligence Unit (Station HYPO, at Pearl Harbor, also known as COM 14) was commanded by Commander Joseph Rochefort. broken in May 1940. 28 May 1941, when the whale factory ship Nisshin Maru No. 2 (1937) visited San Francisco, U.S. Customs Service Agent George Muller and Commander R. P. McCullough of the U.S. Navy's 12th Naval District (responsible for the area) boarded her and seized her codebooks, without informing Office of Naval Intelligence (ONI). Copies were made, clumsily, and the originals returned. The Japanese quickly realized JN-39 was compromised, and replaced it with JN-40.

JN-40

JN-40 was originally believed to be a code super-enciphered with a numerical additive, in the same way as JN-25. However, in September 1942, an error by the Japanese gave clues to John MacInnes and Brian Townend, codebreakers at the British FECB, Kilindini. It was a fractionating transposition cipher based on a substitution table of 100 groups of two figures each followed by a columnar transposition. By November 1942, they were able to read all previous traffic and break each message as they received it. Enemy shipping, including troop convoys, was thus trackable, exposing it to Allied attack. Over the next two weeks they broke two more systems, the "previously impenetrable" JN167 and JN152.

Sources

  • Bletchley Park - Japanese Codes
  • Video interview with one of Rochefort's JN-25 codebreaking team
  • Detailed explanation of the structure of JN-25