Question about WWII Cryptography

Silly question: I was reading Cryptonomicon and they seemed to be breaking codes in English - that is all the cyphers they received were English, even the Japanese ones.

I presume this is because they were transmitted via Morse code and Morse code is for transmitting English.

My question is: did the Germans and Japanese really send their cyphers around in English and not Japanese/German? I assume that it would both be possible and preferable to make the cyphers in their own languages.

German ciphers were sent in German, but I don’t know about the Japanese.

And the Japanese used Japanese.

Alan Stripp’s Codebreaker in the Far East (1989; Oxford, 1995) covers such matters in detail. Though I’m slightly inclined to doubt that Stephenson has read it.

Okay, how the heck do you use character-based languages (like Japanese) with Morse code?!

You don’t transmit in Morse code, because Morse code offers no security whatsoever. I don’t know exactly how Japanase cryptography works, but English is also a character-based language; it’s just that our alphabet is considerably smaller.

My father’s job in the U.S. Army in WWII was to monitor the traffic between the Japanese embassy in Moscow and Tokyo. He didn’t decode the transmissions, which were primarily in the Purple code, but the decoded transmissions he received were in Japanese, which he spoke.

Japanese cryptography used romanized Japanese.

I have just finished reading The Emperor’s Codes by Michael Smith , which is a very history of the breaking and reading of many Japanese codes and cyphers by the Allies. Taken from this book is a brief explanation of Japanese Morse code:-

The advent of the telegraph had bought problems for the Japanese , whose written language was based on pictorial characters or ideographs, called kanji , and around 70 phonetic symbols called kana… A system of transliteration known as* romaji** was developed which allowed kana syllables to be spelled out in Roman letters. The Japanese created their own Morse code which contained all the** kana** syllables plus the romaji letters and was totally different to the standard system.*
I recommend this book as it is both informative and entertaining.

Its not just smaller, its based on phonemes (sounds) while Japanese is based on, well, characters or ‘words’.

And the Germans transmitted everything in Morse. They just encrypted it first. What else would they have used? There weren’t many secure digital datastreams back then…

Not quite. Their higher security Schluesselzusatz system - known to Bletchley as Fish - transmitted five character International Teleprinter Code Baudot symbols, in some cases automatically at speeds far faster than a human could key.

Colossus ( the world’s first programmable computer) was developed by Bletchley Park and the Post Office to decrypt these transmissions.

Wouldn’t that be why they used the cyphers in the first place?

I figured they’d the cyphers to turn “Attack burning ships on fire” to “AJRHC AHREF AHJRH DHRLD FHTLD FUEUD” and then transmit the cypher in Morse code?

From a theoretical cryptographic standpoint, there’s no distinction. A message is made up of a sequence of characters from an alphabet, and it doesn’t matter if there are 52 of them or hundreds of thousands.

From a practical cryptographical standpoint, the distinction isn’t as big as you might think. Take the size of your alphabet, take the next biggest power of 2, and that’s how many bits you need to represent your characters. From there, it’s all the same.

Point taken.