Appendix 1: A brief account of Allied code-breaking in WWII

It's beyond the scope of this paper to describe the inter-service rivalries that affected the Army Air Force units in World War II performing duties which other units felt were their prerogative - like construction or intelligence gathering - but they were very real. For example, the Army Signal Corps' official history of the war doesn't discuss AAF radio intelligence efforts with more than a mere mention despite filling three heavy, green volumes with other information related to equipment specifications, unit actions, etc. Similarly, the Navy and Army often seemed to operate more closely with the corresponding services of other countries than they did with each other: Americans shared only limited intelligence inter-service, while the U.S. Army's cooperation with Australian and British soldiers in the south west Pacific was outstanding.

To provide some background, I should explain that the Japanese and German coded messages which were intercepted by radio, were usually divided into a few major groups. Messages known as PURPLE were encoded with a relatively weak code used only by the Japanese diplomatic corps, while ULTRA messages were from the military high command and which had been encoded with a German code machine called Enigma. It was vitally important that the enemy never discover that the Allies had broken the Enigma machine's codes: each time in the past that the Nazis had suspected it had happened they made the machine more complex, and it was feared that if it got any tougher the Allies might not crack it again!

The information that was extracted from messages encoded with an Enigma machine was called ULTRA intelligence or simply "ULTRA"; the pool of people who knew that ULTRA intelligence even existed - much less knew the specifics that those messages revealed - was tiny. Besides Prime Minister Churchill and President Roosevelt, and the few officers in the highest levels of the chain of command who had to decide how to use the information, the men and women who intercepted and analyzed these messages were about the only ones who ever saw them or even knew that they existed. The word "ULTRA" was never used to describe these facts, and in many cases the information was mixed in with intelligence from other sources in order to hide its origin. In fact, some of the highest-ranking members of General MacArthur's staff weren't trusted to know about ULTRA, and so they were making battle plans involving thousands of men and planes and ships in relative ignorance.

Not all messages could be understood, but every intercepted message could be useful in several ways. One technique that found value in even still-encrypted messages was traffic analysis: "traffic" refers to the messages going back and forth across the Pacific Ocean. Traffic analysis tries to draw meaning out of the patterns of these messages, regardless of their content. For example, an increase of radio transmissions in a particular area might indicate the build-up of enemy troops preceding an invasion, or might suggest the presence of a high command post (given that higher levels of the Japanese command sent out more messages, and in different codes, than did lower-level units), thus identifying potential targets or providing warning of enemy intentions.

Of course messages that were readable were the most valuable, even when the exact names and places weren't identifiable. One man at Navy intelligence headquarters in Hawaii kept a huge card catalog of every Japanese unit and officer that had been mentioned in a decoded message in order to try to identify troop movements later. The patterns in messages - when they were sent, who they were sent to, who they were sent by - could sometimes be worked out by persistence and reason, but sometimes trickery was required. In one famous case a fake message was broadcast by the Americans about the water treatment plant breaking down on the U.S.-held island of Midway: when the Japanese reported back to their headquarters about the opportunity to take advantage of the American weakness, the Allies were able to confirm their guess about the island's Japanese code name!

Intercepted messages couldn't always be decoded and acted on immediately. However, they were all data for statistical analysis of tough codes, for example, or as confirmation of American guesses at the meaning of a Japanese code word. Consider the hypothetical case of an American analyst who believes that he knows the Japanese code name for a specific ship, and who finds out that it had been sunk the previous day. He could watch for a message reporting the sinking of a ship with that code name: this would not only confirm the meaning of that ship's code name, but possibly also the meaning of code names for Japanese Navy units involved, the names of nearby islands, and other facts. The analysts had to make a lot of assumptions and outright guesses, and the smallest hint or corroboration was used.

Another, less glamorous part of radio intelligence-gathering was finding out where the enemy had his radios, since this indicated where his troops or observers were. Tracking the Japanese high frequency radio transmitters was called HFDF (High-Frequency Direction-Finding) and pronounced "huff-duff." While this couldn't tell the Allies where the enemy was headed in the future, it was the surest way to determine where they were at any given time. HFDF figured prominently in traffic analysis, since the content and timing of messages was more valuable when analysts knew where the signals came from.


| Home |
| Introduction | Main | Timeline | Maps | Gallery | Decorations & Medals |
| Appendix 1 | Appendix 2 |
| Works Consulted | Agencies Consulted | Individuals Consulted | Websites |