Breaking Purple: How US Codebreakers Rebuilt Japan's Cipher Machine Blind
transcript
show notes
Imagine reverse engineering a complex machine you have never seen, using only the scrambled text it produces. That is what American cryptanalysts did with the Japanese Type B cipher machine, codenamed Purple, which protected Japan's highest-level diplomatic traffic before and during World War II. This episode traces the origins of Purple back to the humiliation of the 1922 Washington Naval Treaty, when the American Black Chamber read Japanese diplomatic messages, and to the flawed Red machine of 1931 whose weakness the Japanese Navy hid from the Foreign Ministry.
The episode explains how Purple used telephone-exchange stepping switches rather than rotors, why it split the alphabet into a weakly protected group of six letters and a heavily scrambled group of twenty, and how that design flaw, likely rooted in telegraph tariffs, gave William Friedman's Signal Intelligence Service its foothold. It follows the breakthroughs by Genevieve Grotjan and Francis Raven, Leo Rosen's construction of the Purple Analog from commercial switches, and the flood of Magic intelligence that followed, from the 14-part message before Pearl Harbor to Ambassador Hiroshi Oshima's detailed reports from Berlin that became the Allies' best window into Hitler's plans.
- Why bureaucratic silos between the Japanese Army, Navy and Foreign Office, and reliance on mathematician Teiji Takagi, allowed Purple to inherit the compromised design of Red
- How cribs such as I have the honor to inform your excellency, and messages sent in both Red and Purple, gave codebreakers a Rosetta Stone for mapping the machine's wiring
- The September 1940 discovery of the machine's mechanical period from the intervals between repeating letters, and the ten-day key cycle that exposed the rest
- How Soviet codebreakers also read Purple, confirming Japan would not attack Siberia and freeing divisions that helped save Moscow in December 1941
- Why Japan rejected an explicit German warning in April 1941 that its ciphers were being read, and what that hubris suggests about trust in modern encryption facing quantum computing