What would a Japanese WW2 military coded message have looked (or sounded) like?

Wikipedia isn’t much use for this question. So for an allies WW2 coded transmission, I guess we had morse code or something, and then that was transposed in to alphabet letters, and then the decoding could progress from there. Germany had that too with their enigma machines.

But Japan doesn’t usually use the roman alphabet. What did they use instead? I ask because phonetic transcription could easily be misunderstood in Japanese, I think.

While Japan doesn’t usually use the Roman alphabet, they do still have an alphabetic writing system (multiple systems, actually), with a number of characters comparable to the Roman alphabet. I would assume that there’s some equivalent of Morse code to work with katakana or hiragana.

Thanks for the response. Well here’s the thing. I guess an allied submarine commander, for example, got their orders from an audio Morse code transmission (or something else?). And then they applied their code manipulations to the plain text to get the real message.

If Japan used a similar system, as you imply by mentioning katakana, it would have to be a phonetic based system, which doesn’t work very well in Japanese.

Hey I speak Japanese and have lived in Japan for decades. But I know none of my friends family or workmates would know the answer to this question. Best to ask here maybe.

Here is a sample 1941 message:

Not relevant. First you encode the original message any way you want into the 26-letter alphabet (which the codebreakers have to figure out), but only the ciphertext ultimately gets sent over the wire via Morse or any other code.

Was that sent to a Japanese task force by Japan?

Not that one; it was sent on December 7, but the caption says it is part 1 of 14 of the message ultimately delivered to the US Secretary of State.

Why would it have to be phonetic-based, and why wouldn’t that work well in Japanese?

In European languages or Japanese alike, there would be a few steps: First, you need some standard way of writing the language. This has existed for ages, and is well understood. Both Japan and Europe used phonetic writing systems, without problems, because both have more or less standardized spelling.

Then, if you want secrecy, you need some sort of code system for converting the characters of that writing system to other characters, in a difficult-to-predict way. By the time of World War II, this was usually done by machines, such as the German Enigma (all of the great powers had machines that were qualitatively similar: You adjust some settings for the key you want, type one character a time, and a different character shows up in the output).

Then, you take that string of enciphered characters, and convert that into something easy to transmit over the radio, like Morse code.

At the other end, you go through the reverse process: You turn the Morse code back into language characters, then (if it’s encrypted) type the language characters into a code machine to get decrypted language characters, and then you read the resulting writing.

According to this

Type B Cipher Machine - Wikipedia

was implemented by first transliterating the message to the 26 character Roman alphabet and then using a variant of the German enigma machine for the encoding.

Although Japanese is not generally written in Roman characters, there are well-established systems for doing so, referred to as Romanji. Some these were in use before World War 2, and there was even a movement to replace the existing Chinese character based system, although it did not catch on. But I think that knowledge of Romanji was common enough in that time that using a Roman based system was not an issue.

There’s 48 katakana characters so you need a Morse (or other) code for all of them. Then after that, even after encoding them, and decoding them, Japanese isn’t readily understandable without kanji. And kanji can’t be transmitted in ww2 level technology, to the best of my knowledge.

I assume there is also a codebook involved at some point? For example, on such and such a day a ‘battleship’ might be an ‘ice cube’, on another a ‘cement mixer’? We need to have a look at the original sources to see a full example of the naval codes in use.

The guy receiving the message is responsible for decrypting it and typing it (or writing it) up using kanji and other Japanese alphabets, before handing it to the skipper or ambassador, as the case may be.

ETA classic American fuck-up in 1944:

I think this is the part I’m not getting. I had understood that katakana was considered informal, or possibly childish, but that it carried the full meaning. And while Japanese can be written in romanji, isn’t that a 1:1 correspondence with the katakana, with most katakana characters corresponding to one Roman consonant followed by a vowel? If romanji can carry the full meaning, then, why couldn’t katakana?

You are correct, katakana or hiragana don’t carry the full meaning. They just make the same sound as the intended word.

Japanese has a lot of homophones. So seeing the kana can be like hearing someone making a certain sound but not knowing if they are saying too, to, or two without seeing it written down (which is what the kanji are like in this analogy.)

Genji Monogotari was written in hiragana.

For instance (and I know almost no Japanese, but I knew this one) the word pronounced “kami” means god. And hair. And paper. And seasoning.

True, but the same situation arises when listening to spoken speech. We were told so often by our native Japanese teachers that it became a joke: “it’s clear from context”.

That would be “Purple” – the diplomatic code of which I know little.

Japanese Navy messages used “JN-25” which was all digital. Words and phrases were changed into 5-digit groups using a code book and these groups were then encrypted.

Actually as syllabaries, katakana and hiragana perfectly understandable by themselves while kanji will often have tiny “helper” furigana by them to make them more parseable. Books for young children are printed using hiragana as they have not yet memorized a lot of kanji. The translation of The Hobbit was printed this way and looks a bit odd to adult eyes.

But wouldn’t homophones be equally problematic using romanji, or any other writing system?

Sorry, my mistake!

Here is a sample JN-25 message:

You can also see a sample (reconstructed) codebook page; e.g., “battleship” is “31143”. See Mathematical Horizons 24, vol 3, “Cracking the Japanese JN-25 Cipher”