Magic and the breaking of Purple

"Magic" was the American umbrella name for intelligence from Japanese diplomatic messages that the US Army and Navy intercepted and decrypted before and during the Second World War. Its most famous component was the solution of the Japanese Foreign Office's top cipher machine, which American cryptanalysts called PURPLE. Strictly, Magic referred to the diplomatic material; the breaks into Japanese naval codes that set up the Battle of Midway were a separate effort, though the names are often blurred⁸.

A new machine in Tokyo

American codebreakers had been reading Japanese diplomatic traffic since the 1920s, including an earlier Foreign Office machine they called RED⁷. In the latter part of 1938 they intercepted messages arranging travel for a Japanese communications expert who was to install a new "Type B" cipher machine at major embassies. A message intercepted on 19 February 1939 gave the machine's start date as 20 February 1939¹. The Japanese designation was the 97-shiki obun injiki, roughly "Type 97 typewriter for European characters"^6,7.

The Army's Signal Intelligence Service (SIS) in Washington attacked it. The new machine kept one weakness of RED: it split the alphabet into a group of six letters and a group of twenty, and the "sixes" were enciphered more simply. SIS recovered the sixes within months, but the twenties resisted all attempts^1,7. William F. Friedman, the service's chief cryptanalyst, was assigned to supervise the effort in August 1939¹.

The September 1940 break

Friedman's own report, written on 14 October 1940 and later released through the National Archives, gives the clearest account of the solution. The team's key technique was to find "homologs," messages sent on different days with the same indicator that could be converted to a common base. This produced six messages in indicator 59173¹. On 20 September 1940, studying those messages, the team spotted repeated sequences. Friedman recorded the moment precisely:

This, on September 20, 1940, at about 2:00 P.M., was the very first indication that a successful attack might be possible.

¹

By 27 September, the day Germany, Italy and Japan announced the Tripartite Pact, the team had handed in translations of the first two solved messages¹. They also worked out that the machine used telephone-exchange stepping switches rather than rotors, and they built an equivalent from standard parts; Friedman credited Leo Rosen with the idea of using electrical telephone switches¹. Friedman named the joint leaders as Frank Rowlett and Robert Ferner, and listed Genevieve Grotjan, Albert Small and Samuel Snyder among the junior cryptanalysts¹. Later accounts single out Grotjan as the one who first saw the pattern⁷. Friedman himself insisted that "the name of no one person can be selected as deserving of the major portion of credit for this achievement"¹.

No complete Japanese machine was ever captured. Fragments recovered from the Japanese embassy in Berlin in 1945 showed that the Americans had guessed the switching mechanism almost exactly⁷. Analog machines went to the Navy, to the Philippines and, in early 1941, to Britain, where they were traded for information on British work against German ciphers⁷. See Ultra and the breaking of Enigma.

How Magic was handled

Before Pearl Harbor the Army and the Navy's codebreaking unit, OP-20-G, split the decryption and translation of diplomatic traffic, the Army taking messages of even dates and the Navy odd ones². Distribution was tiny: the Secretaries of War and the Navy, the Army Chief of Staff and the Chief of Naval Operations, the heads of war plans and intelligence in each service, the State Department, and the President through his naval aide². Couriers generally showed the translations and took them back⁸. The congressional committee that investigated Pearl Harbor observed that the secret was guarded so closely that "the fact the Japanese codes had been broken was regarded as of more importance than the information obtained"³. It also recorded that the Army did not send Magic to field commanders, and that there was no record system showing who had seen which message^2,3.

The Magic intelligence was preeminently important and the necessity for keeping it confidential cannot be overemphasized.

³

Pearl Harbor

Magic showed Washington that Japanese diplomats were working to deadlines, that Tokyo regarded the negotiations as all but finished after late November 1941, and, on 7 December, that the 14-part reply breaking off talks was to be handed to Secretary of State Cordell Hull at 1 p.m. Washington time^2,6. Intercepts also included Tokyo's instruction of 24 September 1941 telling the Honolulu consulate to report ships in Pearl Harbor by sub-area, which critics later called the "bomb plot" message². General George Marshall's warning on the morning of 7 December went by commercial telegraph rather than by telephone².

The war and after

Through the war Magic's richest vein was the reporting of General Hiroshi Ōshima, Japan's ambassador in Berlin, whose long PURPLE dispatches described conversations with Hitler and German defenses in France^7,8. In September 1944 Marshall wrote privately to the Republican candidate Thomas Dewey to keep the codebreaking out of the presidential campaign⁸; the committee cited those letters for "a discussion of Magic and its great significance"². The committee also urged Congress to look at the law on leaks, noting that wartime violations involving Magic could not be prosecuted without giving the secret to the enemy³.

The congressional hearings of 1945–1946 made Magic public. In 1977 the Department of Defense released a multi-volume collection of the 1941 intercepts, The "Magic" Background of Pearl Harbor⁵. The work of SIS and OP-20-G passed after the war to the agencies that became the National Security Agency (NSA). For how the British and Americans protected similar material against Germany, see Ultra and the breaking of Enigma and Signals intelligence.

Read next

December 1932 – May 1945 Ultra and the breaking of Enigma How Polish mathematicians first broke the German Enigma machine, and how Britain's codebreakers at Bletchley Park turned that start into Ultra, the most closely guarded intelligence source of the Second World War. 1915–1970s Cipher machines From Hebern's rotor to Enigma, Purple, Lorenz and SIGABA, the electromechanical machines that carried the secrets of the Second World War, and the codebreakers who read most of them. 1914–1990s Signals intelligence Intelligence from intercepted communications and electronic emissions, from Room 40 and the Zimmermann telegram to Ultra, Venona, the Berlin tunnel and NSA's domestic programs exposed by the Church Committee. United States · 1952 National Security Agency (NSA) America's signals intelligence and communications security agency, created by a secret Truman directive in 1952 out of the Armed Forces Security Agency, and exposed to public scrutiny by the Church Committee in 1975. 1944 – 1 May 1946 TICOM The 1945 Anglo-American hunt for Germany's codebreakers, their machines and files, which showed how far the Germans had read Allied traffic and handed the West a working Soviet intercept system.