Encrypting Chinese, pre-computers

December 5, 2005 9:23 AM

A friend of mine and I were wondering today how encryption was done in character-based languages like Chinese and Japanese before the advent of computerization. Whereas all kind of substitution and transposition ciphers come to mind immediately for letter based languages like Latin, it isn't obvious how you could use them for a language with many thousands of distinct characters.
posted by sindark to writing & language (9 comments total)

In Japanese, you can break down the kanji into kana, which represent syllables and are used much like a phonetic alphabet. (In publications aimed at children, the kana is usually written small under the kanji.) Then you could perform the same sort of ciphers you do with Latin character sets.

Wikipedia says that this is exactly what was done--simple substitution ciphers with syllables.

I'll bet you could do the same thing with Chinese dialects, but I'm less familiar with those languages, and there isn't a nice Wikipedia page for cryptography in China. ;)
posted by trevyn at 9:42 AM on December 5, 2005

Chinese also has a phonetic alphabet, used mostly for children's books. But I don't remember the name, and have no idea how old it is.
posted by scottreynen at 9:57 AM on December 5, 2005

Bopomofo is the name, and Wikipedia calls it "traditional", which I'm guessing means it's older than computers. Now whether or not it was actually used for encryption, I have no idea.
posted by scottreynen at 10:03 AM on December 5, 2005

You might be interested in The Emperor's Codes, about the breaking of the Japanese codes during WW2.

One of my university professors had been a member of the British team who helped to break the Japanese codes. He told me that he had been sent on a Naval Intelligence training course, given three months to learn Japanese from scratch, and told that if he failed the final exam, he would be assigned to the Atlantic convoys (one of the most unpleasant and dangerous jobs in the Navy). He passed.
posted by verstegan at 10:06 AM on December 5, 2005

Chinese ciphers.
posted by languagehat at 10:46 AM on December 5, 2005

The difference between "code" and "cypher" is that a cypher takes the characters and changes them in a systematic way, but a code is where you pre-establish alternative meanings for words and phrases, so that I can write to you and say "take your shirts to the dry-cleaner" and you understand it as "fly a plane into the WTC".

I assume they just used codes, not cyphers.
posted by AmbroseChapel at 12:30 PM on December 5, 2005

before chinese characters were introduced in japan, they used a system of writing that was derived from the shapes of knots. the ninja supposedly adapted these characters after they fell into obscurity, because it enabled them to leave hidden messages in ropes.
posted by mshrike at 12:38 PM on December 5, 2005

@AmbroseChapel

I have seen 'code' and 'cipher' defined in a number of different ways. Most commonly, I've heard people say that 'codes' are not meant to conceal, but just to organize, like ASCII, and that 'ciphers' are meant to conceal information. In any case, you understand the meaning of the question.
posted by sindark at 12:56 PM on December 5, 2005

Perhaps my earlier comment was too laconic. This section gives some history, this one describes the technique.
posted by languagehat at 5:17 PM on December 5, 2005