I will just add that I have heard people speaking what was ostenibly English and I did not understand a word.
Those interested might check out “The Code Book” by Simon Singh (ISBN 0-385-49532-3). Chapter 5 deals with the subject of decoding a language, and in particular discusses the use in WWII of Navajo and code-within-a-code.
There is a book about this (fiction) - I am trying to remember the name, the hero of the story spoke I think either three or four cradle languages (Russian, French and English?) in addition to Japanese picked up later in life.
I believe he was locked up and thrown a Basque? dictionary, and learned the language entirely from that… book may have been called Shibumi?
Anecdote: My wife speaks Cantonese, I speak a very very small amount of Mandarin. She watches Cantonese serials extensively (TVB), so far I have not been able to pick up ANY of the language this way.
My problem is that people can be standing around the room talking, and although I can get certain parts of what they are talking about from the context, (eg: know they talking about food) because of grammar etc I don’t know which word is which. eg: they could be saying, “this food is really fantastic” but they could also be saying " the chef in this restaurant is very talented" - both mean the same thing, but don’t share any common words. I would need some sort of base or starting point in the language to fill in and learn more words I don’t understand…but just MHO - and I have no talent for languages anyway.
I would bet also, that as is being mentioned that maybe they could also use tortoise as slang for slow, or around here a “mountain tortoise” is also a bumpkin…so that is three possible translations for the same word, makin glife very very hard…
and in Mandarin a “turtle” can means a cuckold, and a “turtle’s egg” can mean a bastard, and a “turtle’s head” can mean a glans penis . . . on the other hand, sometimes a turtle is simply a turtle.
No not at all. The point was purely for field communications that could not be feasibly coded and de-coded using regular methods at the time. High-level and strategic codes used conventional encryption techniques.
Previous thread: Navajo code talkers.
That’s a neat trick, considering Basque is a language isolate. Meaning it has no relation whatsoever with Russian, French, English, Japanese, nor the various languages that led to them. Basque stands alone, it’s badass that way.
I’d like to add that further even if you have a written form, you need to know the phonetics. Take the word Navajo. Sound it out in English phonetics and see what you get. Now pretend you don’t know the real pronunciation and see what can be learned about “Nava-Joe”.
I make this point because “Navajo” seems to map to Spanish phonetics and gave me quite start when my brain was like “oh Spanish” and read “Navajo during” as “navaho dooreeng”.
Yeah but not that alone. Basque has no direct relationship to Latin, but learning Basque declensions once you’ve understood the concept of “declension” from studying Latin is a lot easier than if you’ve never encountered it before; things like tenses with and without an auxiliary are common to Basque and to other European languages; the order of words isn’t completely “over the place” compared to French or Spanish (both of which have some Basque influence). Does Russian have declensions?
A few minutes later…
Almost.
It was Shibumi. As I recall, one of the books that he had was in French on one page and the Basque translation on the other. That was how he learned Basque but he didn’t really know how it was pronounced so his Basque accent was terrible.
True, but barring the possible loanwords for modern devices and the like, the vocab itself has no etymological relation one could “crib” to approximate a meaning. Even if our linguist hero can correctly identify the articles, ascertain that two different words really are the same word with different declensions, that a given word is the verb of the sentence, etc… he’s really no closer to cracking what they signify without a Rosetta Stone of some sort.
Before this thread I had no idea this went back to WW-I and involved a number of native languages. Hitler had known about this and tried to compensate for it: From wiki:
**Adolf Hitler knew about the successful use of code talkers during World War I. He sent a team of some thirty anthropologists to learn Native American languages before the outbreak of World War II.[3] However, it proved too difficult for them to learn the many languages and dialects that existed. **
Code talker trivia: the Comanches substitute code word for Hitler was “crazy white man”.
Our linguist hero in the post you originally responded to was handed a dictionary. I imagine that would be a big help, specially if it came with the usual section on grammar.
I assumed it was a Basque-Basque dictionnary. Working from the grammar section of one of those, he would probably have got the verbs “have” and “be” down pat, as well as the articles, conjunctions and probably the pronouns too but not much further. Let’s go with that, and Myrriam-Webster “aardvark” article, obscuring words the guy couldn’t know or pickup from the dictionnary’s grammar section :
Aardvark : a something something something something (Orycteporus Afer) of something Sahara Africa that has a something something, something something, something something, something something, and something something and verb adjective on something and something.
Well that’s one word elucidated ! 
I guess the point is moot if the guy has a French - Basque book to go with it, but that’s cheating
Or a copy of The Three Musketeers. I guess we’ll never find out whether he got the Batua dictionary (“unified,” the official, invented dialect), a Batua-otherdialect dictionary or a Batua-French dictionary, though, given that bengangmo isn’t sure of the reference. If it was the Batua dictionary it’s been heavily accused of over-Latinizing, so that’s actually a bunch of words he’d be able to figure out pretty fast.
A good question. I think it must be close to impossible. I’m a trained linguist and have worked out grammar systems from made-up languages as part of my coursework, but we were always given written sentences and partial dictionaries first.
I’ve also been immersed in a language I didn’t understand (Japanese) for many years, and for my first 4 or 5 years I watched only Japanese TV. You can pick up quite a bit by watching the action and memorizing the phrases. For example, when someone on the TV show walks into a shop that appears to have no staff present, you know what you’d say in English - something like “Hello! Anyone here?” and then you hear the actor say “Gomen kudasai” you can make a pretty good guess that these phrases are cognates. Not a perfect guess (see Grice on indeterminacy) but a pretty good one at least.
However, back to the problem at hand, here we are talking about listening to a radio transmission - the big hurdle here is that there are no visual context clues. You can’t see what people are doing so you have no frame of reference to begin. Having said that though, just listening to the sounds of speech again and again and again will eventually give you a better accent (when you finally get around to buying a textbook and learning the language). When I call shops or services here in Japan on the phone they don’t usually guess that I’m a foreigner.
Edit - I used the term “cognate” wrongly. Sorry, my English gets worse and worse every year. I meant something like “equivalents”.