I'd rewind the film and switch to English subtitles when I didn't understand a phrase. It takes about 6 hours per film at first, but becomes faster and faster as you learn more vocabulary/grammar and get used to the process.
It's best to use real French films, because it's important that the actor's mouth movements match their voice, especially when they're speaking fast.
I live in Paris now, and speak French fairly well with no American accent.
I also read bilingual books and memorized songs on the bus (pasted the lyrics into iTunes so they'd show up on my iPod when listening to the song).
The rule was never sleep without giving your brain something to work on. I spent an hour in the morning and an hour at night practicing the precise pronounciation of syllables, words and phrases, using tapes as a reference. There was little improvement during sessions, but noticeable improvement between them. Thus the rule about sleep.
As my pronunciation improved, so did my comprehension. I can't imagine it working the other way. Comprehension is passive, and no guarantee of accurate or even coherent production.
Following such a method you can learn a language in 6-12 months.
EDIT: I just read up on Krashen, and well, no, his theories are wrong. I wasted years on that crap. I doubt fluency is even possible with such a method. Just do what sounddust said and you'll be set.
EDIT2: The average person acquires fluency despite the Krashen method, rather than because of it. In my case, due to very poor audio recall, the Krashen method was a complete failure, so I was forced to experiment with other ways.
It turns out perception of foreign sounds is very weak in adults. We recognize a subset of possible sounds and toss out the rest, or munge it into something heavily accented. The repetition & production method compensates for this by making such sounds "real". Once they are real, the actual language follows more quickly.
EDIT3 (sorry): More on Krashen, from http://www.sk.com.br/sk-krash.html:
The only instance in which the teaching of grammar can result in language acquisition (and proficiency) is when the students are interested in the subject and _the target language_ is used as a medium of instruction.
This is so fucking wrong. The study of grammar is analytical. Language is the vehicle of analytic thought--if your proficiency in a language is weak, your level of analytic thought in that language is also weak. So if you want to study French grammar and you've only just started learning the language, by all means, study it in English.
Wow. So this is why the English-speaking world can't seem to learn a foreign language. Thank you xccx, you opened my eyes.
"In June 2004, at the ripe old age of 21, all post-pubescent and supposedly past my mental/linguistic prime, I started learning Japanese. By September 2005, I had learned enough to read technical material, conduct business correspondence and job interviews in Japanese. By the next month, I landed a job as a software engineer at a large Japanese company in Tokyo"
Basically, his method involved constantly immersing himself in Japanese media.
I lived in Tokyo for 2 years, studying Japanese full-time, and completed several courses in Finland before that. Speaking came pretty easy and I am quite fluent in normal conversation, but learning kanji has proven to be a real time-sink and a source of frustration, although at the same time I love them for their beauty. Also reading Japanese feels different than reading languages in the Latin alphabet, as the characters have an extra layer of meaning (not a huge difference, but helps to distinguish homonyms + create new kinds of puns / emphasis).
There are roughly 2000 kanji that you should know to be a high-school level reader. The latest test I took (http://www.speedanki.com) shows I know about 800. Reading a newspaper is not possible currently, as I would have to constantly look up kanji, and often the important words are the rarer ones.
But I will learn them. At this point it's an obsession, I'm not even sure why I need to know them, except to prove to myself that I can.
It may be painful, but it should be possible; yes, you'll have to look up the kanji, over and over -- writing them down as you look them up -- and it'll take you half an hour to read half of one article. But if you keep at it, and keep looking them up, every day, you will improve -- you will learn them! You just have to tolerate a lot of tedium on your way.
I picked up enough Portuguese to fumble along in Brazil by taking 40 hours of lessons from Berlitz, 2 hours per day twice a week, just me and the teacher. That was kind of the minimum to make continuous progress. Group classes are way less effective, and courses at the local JC are totally worthless. The only thing that counts is how much time you actually spend speaking.
It's expensive as hell, but I strongly recommend 100 hours of individual instruction before going off on your own in learning. It gives you a good feel for the pronunciation of the language. If you learn by reading but with the wrong pronunciation, it may take a long time to recover. The proof to me was when I was in rural northeastern Brazil. The person that I was haltingly talking to said that he could tell that I was from Rio by my accent. I wasnt, but my teacher was.
The two takeaways here are: learn pronunciation first, and learn the script of the new language immediately (or you'll just mentally transcribe the sounds using your first language)
Correct, and correct. All native speakers of any language are habituated to produce the sounds and perceive the phonemes of their own language, and NOT to produce the sounds or perceive the phonemes of any other language. Acquiring an understandable accent generally takes good training at the beginning.
What does "script" mean in this context? What is the "script" of a language?
P.S. I wrote a little Adobe Air application for someone who likes to learn a language while listening to songs or watching youtube. You can search lyrics/save/translate and bookmark your videos. It's a just organizes the whole process. Have fun ! http://www.singandstudy.com
The emphasis would have to be on fun in that formula for making a fortune.
I think various websites publish the skill set(s) the country and/or provinces are looking for. It's trivial for a US-ian to get work here, and it's a 3-6 month process for Europeans to do so (at least in software-related businesses). It also helps if you're young, single, ...
It seems that most Swiss speak at least two languages well, but these two are usually Swiss German and either French or Italian. Those working touristy jobs usually know enough English to take orders and direct tourists to bathrooms, but that's about it.
Edit: The irony of this article is, of course, that this guy probably now has a better understanding of French than most Canadians outside of Quebec.
But more power to anyone who takes the time and effort to learn another language well.
The United States government agencies that track this matter don't generally seem to think so. But if you have other evidence on this point, I'd be glad to hear it.
I agree in general that a grammatical feature such as minimal marking of nouns for case and verbs for person, number, or gender, making word order the main basis for grammar distinctions, makes Chinese eerily familiar for native speakers of English. But even though Chinese grammar is "easy" for native speakers of English, to the degree that some Americans say "Chinese has no grammar" (definitely a false statement), nonetheless lack of a lot of cognate vocabulary or similar phonological system usually means that the native speaker of English will thrive as a learner of any Indo-European language over almost any non-Indo-European language.
The comparisons I've seen most often assume native speakers of English learning various languages to an equal tested level of functional proficiency after government-sponsored training. Length of training to reach the required level is generally longer for the non-Indo-European languages than for the Indo-European languages. The "easy" languages for English speakers are the typically studied languages like French, German, and Spanish, perhaps because of prior exposure as well as close similarity. It would be interesting to see what studies of language learning starting from various native languages to acquire various target languages are showing these days.
http://news.bbc.co.uk/1/hi/programmes/from_our_own_correspon...
If that's not possible, sounddust's method seems the most likely for a hacker with limited access to Japanese people.
