Code breaking in the Pacific War

"Tell Jimmy to get on his horse."

In 1922 the major naval nations met to negotiate limits to the sizes of navies, and Britain hoped, agree to eliminate submarines. But France and others wanted to keep submarines.

To help our negotiators know what the Japanese were planning, New York City detectives, FBI agents, and Naval intelligence agents broke into the Japanese Consulate in New York City and photographed code books. A retired missionary to Japan was hired to translate these into English. Our negotiators knew the instructions radioed to Japanese negotiators before negotiating sessions.

As the codes became more complex, the Navy was able to continue breaking them. In the 1930s the Japanese switched to a machine code, and code breakers were able to hand build a machine that read the code. The code breakers were highly skilled people, some civilian, some Navy, some Army. One of the most skilled was a remarkable genius, Mrs. Agnes Meyer Driscoll.

IBM punch card machines sorted through and organize large volumes of information. At Pearl Harbor in 1942 3-million punch cards per month went through the machines to analyze the 5-digit Japanese code groups. Commander Joe Rochefort was in charge, a code genius with great memory, who knew the nuances of the Japanese language. He and his group worked almost round the clock. They were known as FRUPac — Fleet Radio Unit Pacific. Midway through the war the name was changed as it moved to larger quarters and grew to over a thousand people, but often still was called FRUPac.

The Japanese developed these codes by 1941:

Diplomatic code that in World War II was used by Japanese diplomats in Germany to report to Tokyo, which gave us insight into Nazi strategic plans in the European theater, but no tactical information. The summer of 1945 we read messages from Tokyo to their embassy in Moscow, instructing their ambassador to Russia to help end the war, and his replies. Broken messages were called MAGIC.
The flag officers code was the most difficult and never broken, but little used by the Japanese Navy.
The major naval code was in its 25th iteration at the time of Pearl Harbor, and Pearl Harbor code breakers on December 7 were working to break it, because the "keys" had just been changed twice. By February before the Battle of the Coral Sea they could read enough snatches to help Admiral Nimitz. Broken messages were called ULTRA.
The "maru" code was easy to read and was used by harbor masters to tell Tokyo and destination harbors of ship movements. Tokyo radioed ships and convoys the "noon positions" where they must be.

In addition there was a code of the Pacific to give locations that we needed. While FRUPac broke information about the upcoming Battle of Midway in the spring of 1942, locations were called AF, AK, etc. Cmdr. Rochefort and his code breakers figured out what some meant. In January 1943 the Japanese submarine I-1 was attacked by 2 corvettes near Guadalcanal. First the Kiwi blanketed the larger I-1 with gunfire, then rammed it 3 times. It sank in very shallow water. Australian and American swimmers quickly dove through it, retrieving documents and codes, including the coded grid of the Pacific. Now our crypto analysts knew the codes for all locations.

From the 1920s the Navy sent top graduates from the Naval Academy to Tokyo to serve for two years as assistant naval attaches to master the Japanese language, and many of these then went into intelligence. For example, at the Battle of the Philippine Sea, June 1944, one of these, Lieut. (jg) Charles A. Sims, translated the orders the Japanese air coordinator radioed to his bombers, so our Hellcat pilots were vectored to the best altitude and position over where the enemy bombers had been told to go ready to dive to the attack.

Using the noon positions and other messages in the maru code, and the coded Pacific grid, the code breakers gave information by secure line to Admiral Lockwood, Commander Submarines, who with his staff vectored the most appropriate submarines to where they could expect targets. Interceptions were planned at night when submarines could use radar on the surface. At the end of the war 231 sea going Japanese merchant ships remained afloat.

"Tell Jimmy to get on his horse"
American leaders, who knew we read Japanese codes, were very sensitive to the possibility that Japan might be breaking our codes. The super secret operation under Col. Jimmy Doolittle to bomb Tokyo with Air Corps B-25 bombers from the carrier Hornet involved many groups. Arrangements were made by a Navy and an Air Corps officer personally without any radio messages. When all was ready, the message alerting all involved to start read, "Tell Jimmy to get on his horse." Coded messages to submarines often included the latest jazz terms and slang.

Japan and the U.S. used radio deception. For example while our only three carriers rushed from the South Pacific to Hawaii for the Battle of Midway, the Tangier, near where a Jap scout plane had seen the carriers, made radio transmissions that a task force of carriers would normally make; Japanese monitoring and radio direction fixes would indicate the carriers remained in the South Pacific. The Japanese used similar deception in the weeks before December 7, 1941.

Several books tell the history of this code breaking. A good summary is in Silent Victory by Clay Blair, Jr., in his opening chapters. Much more detail is in many other books such as And I Was There by Adm. Edwin Layton.