Learning a completely unknown language

How do people go about learning a language completely unknown to outsiders. I don't know how much it happens these days, but it must have happened quite a lot back in the days of colonisation where missionaries would go and try to convert natives and such?

So, what methods do you think they used and how effective would it be?

They probably used hand signs and stuff to communicate at first.

That must be a pretty ineffective method of learning a language. It must have taken ages.

I thought that learning a completely unknown language was necessary relatively recently with some (formerely remote) Brasilian tribes.

Papua New Guinea! LOLZOR! This place is like full of unknown languages! I heard there are like 20 percent of all earth's languages there (don't bet on my word thou, coz I might be wrong about the figures!) But even thou it still happenz there arent as many missionaries any more, coz of political correctness, they just leave em natives be in da jungle, as they outta!

It seems you would have to be well versed in general grammar (beyond your own language and those of your language family). Then you would have to learn basic vocabulary (rock, tree, food, etc.) and how they fit into sentences...

I have a friend who is in Africa right now helping record some random language that's spoken by less than 100 people... I'd email her and ask how they go about it, but chances are she wouldn't respond for a month or two...

It was quite common in the not-so-old days for researchers to go out to learn an unknown language. (The Summer Institute of Linguistics, which is missionary based, may still do it.) Not only missionaries, but linguists from the Structural Linguistics era (Leonard Bloomfield, et al) would often go to learn such languages, especially American Indian languages. This was called Field Linguistics.

They would go to live in the area where the target language was spoken. Besides learning like any other "immigrant" might learn--listening, observing, mimicking, etc.--they would often use what they called an "informant": a native speaker of the target language who also knew a trade language that the researcher knew. The researcher would systematically build up a corpus of the language with questions that would reveal the pronunciation, morphology, vocabulary, et al.

There's a classic by Eugene Pike called Learning a Foreign Language that dealt with just that issue. But there are newer ones too: See Linguistic Field Methods by Vaux, Cooper, and Tucker at http://www.amazon.com/Linguistic-Field-Methods-Bert-Vaux/dp/1597527645/ref=sr_1_1?ie=UTF8&s=books&qid=1215582182&sr=1-1 and click "Search Inside."

By the way, there was a great linguist named Kenneth L. Pike who had a monolingual approach to learning such languages. He used to do demonstrations, one of which is described here, by a professor I once studied with: http://www.sil.org/klp/klp-mono.htm

This is the most interesting topic I've seen lately.

I've seen quite a few foreigners trying to learn my native language (Kiristave) which is an Portuguese-Persian-Indian dialect spoken in Goa (West India) and in the beginning when learning the words, they find it very easy (especially those who have prior knowledge of either spanish or portuguese) but when forming sentences, they go crazy as the persian and indian rules of agglutination come in and make it 50 times more complicated and this is only for the written form of language!!!

The spoken dialect has 3 different forms for every sentence depending on who you speak to and how you want to sound (although previously these forms actually used to define the different social class systems that prevailed in India like the rich, upper class used a form of speaking...the middle class used another form and the lower a third form...but now one has to learn to mix all the three forms while speaking and one form can sound so different from the other....just an eg:

A rich woman came to my wedding
Riqo mulhera veya tiz mi casamenta - the "superir" tone (portuguese predominance)
Daultahonomiye kezishta izaab - the "meyo" tone (persian predominance)
Paisavalli mulera tami baraat avaiza - the "abaisa" tone (hindi predominance)

I wonder how some of the people I've met speak the language so perfectly. And many today consider that speaking in one tone is enough (and that's either the superir or the meyo tones) but it's better to learn the superir and the abaisa tone as more number of people speak that....

<< This was called Field Linguistics >>

-- ... which implies you must be on the spot and have visual contact.

But suppose you only have a phone contact with some alien speaking a fully unknown language: is there a way to reach a level of basic intelligibility in such conditions?

And even if you succeed to, how can you communicate the difference between 'right' and 'left', 'before' and 'behind', or what "clockwise" means?...