The Japanese, quite frankly, were not terribly bright at cryptology. The Germans were very good, but the Poles and the English were better, and the heroic efforts put into breaking the German codes at Bletchley Park played an enormous part in the victory in Europe. The situation in the Pacific was much different. The Allies had already broken several of Japan’s principal ciphers even before the start of hostilities, and continued to do so throughout the war; despite warnings from the Germans (who were unquestionably much brighter about cryptology) about the weakness of their ciphers and seemingly clear knowledge that their ciphers had been, in some cases, compromised, the Japanese continued to use them unchanged. Given this, it’s reasonable to conclude that the Japanese command did not have a good understanding of cryptography. That they apparently failed to make any effort to break the Navajo codetalker codes is, in retrospect, not terribly surprising. I think the answer here is probably found in the Japanese cultural psyche more than anything else.
Yeah, I’ll tell you about my mother…
Maybe I need to start another thread for this (Mods please let me know), but I’d like some of the people contributing here to comment because they seem rather knowledgeable in this area.
My question is, what would a WWII military code have looked/sounded like? I’ve read a little (not much, really) on code systems and encoding, but when I try to picture what a particular code would look like I’m at a bit of a loss. I know that the allies’ codes, in ‘pen and paper’ form might look something like this:
jsyd owmd wifj suwi rivf … etc.
But then how was that transmitted to its destination? By radio, mostly, I guess, and if so, was it just someone speaking into the microphone saying “j” “s” “y” “d” …etc? (I think not because of the difficulty of distinguishing letters like ‘m’ and ‘n’ over the radio), or was it done in Morse code, or some other code?
And for Japanese code, how did they do it? For example, what would a received message have sounded like to a Japanese submarine signals operator? Did they use Morse too? Would it be dashes and dots or a human voice spelling out katakana?
The Japanese almost certainly had the Navajo prisoner of war listen to a recording of the codetalking, not a transcript of it. As I said in a previous post, transcribing a language which you know nothing of is incredibly difficult. A Navajo would probably have gotten almost nothing from a transcript of the codetalking done by even a Japanese with a little training in linguistics. As it stood, the prisoner of war probably could have told the Japanese nothing about the recording except that it was Navajo words, but they were in some weird jumbled-up order that made no sense to him.
Remember, the codetalkers had orders when they designed the code to make it very difficult for the Japanese to understand, even if they could find someone who spoke Navajo. They knew perfectly well that there were Navajo soldiers in the U.S. armed forces who weren’t codetalkers, and there was some chance one of them might be captured. They knew perfectly well there was a small chance that some Japanese citizen (or even some German citizen) might know Navajo well. They were prepared for that. Forget this nonsensical notion that codetalking was some utterly unbreakable code. If the Japanese had had a bunch of Navajo speakers, many recordings of codetalking, and many months to work on it, they probably could have broken it. They didn’t have those things, so they didn’t break it.
I have been doing some research into Navajo codetalking, and I found this article, which is apparently the best summary of the quality of the code:
http://findarticles.com/p/articles/mi_qa3926/is_200010/ai_n8914162/?tag=content;col1
(Notice that it’s split up into 19 webpages.)
There’s also this article with a little additional information:
http://marc.info/?l=cryptography&m=102435507721925&w=2
There are many ways that the Japanese could have figured out that Navajo was the basis for codetalking. As I said, once the Japanese had recorded a few instances of the codetalking and once they had tried to see if it was any of the European and Asian languages that they could find speakers of in Japan, it was an obvious guess that it was an American Indian language, given that the American military had previously used such languages. Given the large number of codetalkers all over the Pacific during the war (420 of them), they knew that it had to be a common language, not one with just a few speakers. Since Navajo had the most speakers, it was an obvious guess.
The article points out many other ways the Japanese could have found out. They never captured a codetalker, but they captured other Marines. They could have figured out by questioning some of them that the Marines speaking on the radios were Navajos. They won’t even have had to torture one of them. Somebody might have mentioned his Navajo Marine buddy during questioning. No copy of the code book could have been stolen since the codetalkers had to memorize the code and bring nothing with them. It’s possible but unlikely that they found out about Navajo codetalking from some espionage.
There was a lot of sloppiness that could have revealed what was going on. The most incredible slip was an article in the tourist magazine Arizona Highways during the war which was a discussion of the ways that the Navajos contributed to the was effort. In this article, it mentioned that Navajo Marines were being trained to use a code in radio transmissions. How such a huge stupid leak happened is hard to tell, but the Japanese probably didn’t find out in this way because it’s unlikely that they spent their time reading American tourist magazines for possible slips.
The most likely way is that the Japanese heard someone mention that they were Navajo during a transmission on the radio. The article talks about clear cases where a codetalker and another American military person got into an argument on the radio because the other person didn’t know about codetalking. The other person, in these arguments, assumed that the Japanese were breaking into their transmissions and would scream that they should get off American transmission frequencies. The Navajo codetalker in these arguments, in anger about being called Japanese, let slip the fact that he was a Navajo.
This is not universally true. There are cultures where it’s considered absurd to try and teach children to speak; they will speak when they learn on their own. Stephen Pinker in The Language Instinct gives the example of a society that laughs at “Western” cultures for trying to teach infants to speak, yet goes to lengths to teach their children to sit up properly. He argues that left alone, children learn to speak just as quickly from casual exposure to language as children who are taught explicitly.
In theory any code can be broken given enough computing power, enough time and a large enough sample size. The Japanese lacked all three.
The Japanese military of WWII was remarkably primative. It discouraged initiative at all levels and was very centralized. So we would have to imagine the Japanese collecting acres of reel-to-reel tapes and FedExing them Tokyo. Once there the samples would have to transliterated into some form of script.
Hundreds of workers would then have to look at the scripts to find patterns. These workers would have to be highly trained, imaginative and familar with the context of the messages and their probable vocabulary.
You can see what a hard nut this would be to crack.
There were things that could have been done. Traffic analysis (who talks to who when, in order to tease out why) comes to mind. But please remember this was a tactical code, used in a small local area on the fly.
Be that as it may, even casual exposure to language doesn’t completely lack context. Were I to go to Japan tomorrow, simply watching and listening to the people around me would quickly teach me that “Konichi wa” and “Konban wa” are some form of greeting, “Sumimasen” is what you say when you bump into someone, etc…
But reading or hearing those words completely out of the blue on the radio, how would I know ?
I haven’t read that book, or know which culture he’s refering to. Here’s a detailed rebuttal of the theory of a language instinct.
So while not all cultures will use specific baby talk or point, it’s still not radio. Children will have 10 years of adults using correct language around them with context to identify it, and adults will react to wrong/ correct language from the children, often by correcting or ignoring wrong language. Very different from the radio situation, and even more advanced than watching TV.
Unless they borrowed them from the Germans, they could not have had reel-to-reel tapes. The technology available to the Japanese of the time were either wire recorders (terrible quality and fragile) or 78 RPM records (better quality but even more fragile).
Code when sent on the radio from one station to another was by morse code. Code by a field combat unit in battle was spoken. And had to be changed regularly. Code books were necessary, and they were captured by the emeny and used. The code talkers used a code that was not written down and any unit with a code talker could communicate to another unit with one. And it was difficulte to break.
Interestingly enough, Navajo wasn’t the only Native American language used by allied armies in World War II; Canadian units sometimes had radio operators who were Ojibways, who used their native language as well. The main reason for it was that it was felt that the enemy wouldn’t understand it at all, since there were so few speakers of it among that generation, thanks to things like the residential school system, where the use of one’s own language was discouraged.
The American armed forces used a number of other American Indian languages during both World War I and World War II. Navajo was just the largest and most thorough of these code systems. The Germans, at least, knew that the Americans were using American Indian languages. A number of German linguists visited American Indian reservations during the period between the wars. It’s not clear how much of this was because of linguistic research (since the Germans knew more about linguistics than anybody else in the world at the time, and, besides, they had always had an interest in American Indians) and how much was a deliberate attempt to know about the languages in case the Americans tried the same thing in World War II.
It’s been a while since I read Pinker’s book, but I would say this an example of a culture having no explicit language teaching rather than an example of a culture that doesn’t teach a child how to speak. Even if they never say to the child “This is an ‘apple’, an apple,” they undoubtedly do more things to teach than a television or radio.