Thank you for your contributions everybody, they are really valuable. So I think we agree that the issue boils down mostly to the ability to develop an identity in a foreign language by giving up part of your native language identity. I think this also depends on how "usual" it is for you to use a language other than your own in your early years.
I have some observations on the subject. I grew up in a bilingual area of ex-Yugoslavia where it was customary to use at least two languages (Serbo-Croatian and Hungarian), and I was often exposed to Slovenian or Macedonian with which I had no direct contact (I mean the speakers), but could understand most of them. So learning English with this background really presented no problem at all. When I came to Hungary (a largely monolingual country with patches of multilingual areas), I realised that for many people it was really hard to master even one foreign language on an acceptable level. Some of the older generations had compulsory Russian at school (up to the late eighties), but they were not really motivated to learn it, so couldn't really speak it, and I guess this affected their attitude to learning foreign languages in general. And I did indeed feel that people living in a monolingual area found it really hard to give up their native language identity, because it had never been really "threatened" (I mean in a challenging way).
Speaking of the "age factor", I think that the capacity to learn a foreign language does decrease a little with age, but if you are "trained", you can "keep fit" up to a very high age. And I do feel you have to take risks. If you cannot converse to people speaking a given foreign language on a daily basis, but happen to have internet access, go and join a forum on one of your interest and try contributing as best you can, not bothering much about mistakes, and you may be feeling your ability to express yourself improve rapidly, because you learn a lot by reading the posts of others alone. Try reading newspapers or books in a foreign language, and by deciphering the meaning you will pick up a lot of vocabulary and idioms which will affect your fluency as well. And one more method that works for me is listening to all kinds of songs in a target language (posibly by getting the written lyrics): the melody will help the words and expressions creep into your mind like nothing else can. :-)
Sorry, Tom, but you didn't learn Polish effortlessly; you just don't remember how hard it was. And you've forgotten that it took you many years to master it. Nobody learns his native language effortlessly, but most people don't remember just how hard it was to learn their native languages.
Living in a country doesn't work for second languages because the motivations of the native language acquisition are no longer there. When you are learning your first language, your motivations are extraordinarily strong: you must learn to speak just to communicate at all with anyone. Just asking for food is impossible unless you learn to speak. And even after you learn a few words, your peers laugh at you unless you improve your proficiency, so you are motivated to get better or suffer ridicule. Overall, the motivations for acquiring a first language are much stronger than they will ever be for any other language—for most people in most cases. But that doesn't mean that you _cannot_ learn a second language just as well, nor does it mean that learning a second language is more difficult … it's not.
I agree with your point on cognitive dissonance. Some people lack linguistic aptitude, although usually this shows in their native language, too (they are not very strong readers, they express themselves poorly or simply and with difficulty, etc.). For them the difficulty of acquiring languages is considerable (as it was for their first language, but they had no choice in that case). These people will rationalize their inability to easily learn new languages by saying that the languages are too objectively difficult, or by claiming that some insurmountable obstacle prevents them from attaining native fluency. It should be kept in mind, though, that ultimately it's their attitude that stops them (if they applied themselves as they had for their native languages, they'd achieve the same levels of fluency).
In any case, anyone who doesn't believe that one can achieve native pronunciation need only listen to your own recordings of your speech in English on this site. You don't have any accent. Proof that it can be done. But most people would rather not expend the effort to achieve that, and since they don't want to admit that it's a question of laziness, they try to claim that it's impossible instead.
And sorry, that old saw about puberty having some mysterious effect on language acquisition is total bunk. Most linguists are so religiously attached to this idea that they aren't even willing to question it, but it has no proven basis in fact. All we know is that people seem to learn languages less well with age—but that does not prove cause and effect, and so it doesn't show that languages cannot be learned perfectly with increasing age, it simply shows that, for whatever reason, few people do it. But we do have counterexamples of people who achieve native fluency well after puberty, which demonstrates that it _can_ be done, and that there is _no_ magic change at puberty.
Most adults students who apply themselves and are reasonably intelligent will achieve greater fluency in a language in ten years than most ten-year-olds have achieved in their native languages over the same period. We expect ten-year-olds to make mistakes, and so we ignore those mistakes when we hear them in children—we still claim that they are "fluent." But if an adult makes the same mistakes, we claim that he can't learn a language because he's too old. Our standards of fluency are different for children versus adults.
And still another factor is that, in many cases, adults are in no position to evaluate fluency, anyway. American parents may claim that their child speaks fluent French because he picked up a few words during a 6-month stay in France. In fact, his French is abysmal, but his parents don't know that because they don't speak any French at all. I've seen this on innumerable occasions. Some parents are further hoodwinked by the fact that children tend to speak quickly, even when they are not fluent (because they aren't afraid of making mistakes). Adults tend to assume that anyone who speaks quickly is fluent, but that's just not true.
Well, I don't know about the others, but I could never achieve native pronunciation, even if I learned to speak and write better than the natives. I believe that anyone can learn a foreign language, but the fact is that some people are better at it than the others, because learning a foreign language takes talent. But if you expose yourself to a foreign language long enough, you will learn it eventually, even though you probably won't be as fluent as in your native language.
Predictions about language learning are self-fulfilling prophecies. People who say they'll never achieve native fluency are right. And people who say they _will_ achieve native fluency are also right.
Living in a country doesn't work for second languages because the motivations of the native language acquisition are no longer there.
I'd say the above proves my point that learning a second language is harder than learning your first language.
When you are learning your first language, your motivations are extraordinarily strong: you must learn to speak just to communicate at all with anyone.
I disagree. You cannot speak of motivation when no conscious decision is involved. I didn't make the decision to learn Polish. It just happened. (Though I do agree it took at least 5 years.)
To learn a second language, you have to make the decision and then follow through. Much harder.
There is also the problem of your first language interfering with the acquisition of a second language. For instance, people tend to interpret the second-language sounds they hear in terms of the sounds of their native language. People also tend to "transplant' phonological rules of their first language to the second language and they're not even aware of it (e.g. de-voicing of final consonants).
In short: For your first language, input does the trick. For your second language, you have to pay attention to things like grammar and phonetics. Input alone does NOT do the trick for many people -- or else all immigrants would speak perfect English.
"There are no lousy language learners who really want to learn and are willing to observe and absorb the language, but are unable."
There are people who lack intelligence and/or motivation who will profess to have very strong "language identity", so that they don't have to admit that they are just stupid and/or lazy.
I agree with almost everything you say, I just don't buy this "language identity" theory. In my view, the biggest obstacle to language learning success is lack of motivation.
"But if an adult makes the same mistakes, we claim that he can't learn a language because he's too old."
- EXCELLENT POINT
"And still another factor is that, in many cases, adults are in no position to evaluate fluency, anyway." -- EXCELLENT POINT
"...your peers laugh at you unless you improve your proficiency, so you are motivated to get better or suffer ridicule"
You know what, I have this pet theory (which was sort of meant as a joke when I said it in a discussion with Michal Ryszard Wojcik) which explains why children of immigrants learn the local language better than their parents. The classic explanation for this is the Critical Period Hypothesis (= the parents' brains are physically different).
My explanation was that parents can get away with speaking the local language imperfectly -- they will not be openly ridiculed by their co-workers, for instance. It's not culturally acceptable to make fun of someone who came to your country, just because he cannot speak your language perfectly.
On the other hand, children are cruel idiots who have no such qualms. They will ruthlessly make fun of anyone who's the slightest bit different, which means that immigrant children are under enormous pressure to "blend in" lingustically -- and do it fast.
I understand that this would correspond to your own views. The problem is, I'm not convinced this theory holds for first language acquisition. I certainly don't recall being ridiculed by my parents because of my mistakes.
You say "I just don't buy this "language identity" theory".
I think this is a crucial factor in the attitude of a good language learner. I have heard your American accent. It is simply American. There is no way you could achieve this accent if you did not subconsciously want to be an American. I do not mean in terms of naitonality or citizenship, but in terms of behaviour. You have projected yourself into an American identity, you have imitated not only the words, but the pronunciation, as wll as other forms of social behaviour of an American. You could not do this if you were not able to see yourself outside of your Polish identity. I know, I have done this for European languages and for Asian languages. I set aside my native identity and become on of them.
When I have Chinese learners say to me, I want to express the more complex levels of politeness of our culture in English, how do I say this etc.. I say forget it. If that social habit does not exist in English, don't try to create it. Imagine you are a native speaker. Act like them when you speak English. when you speak Chinese you can be Chinese.
On another point. Children are not ciriticized but encouraged by their parents. That is why early insistence on correct peformance is counter-productive with adult learners.
Young children learn languages effortlessly. They are curious, interested in learning and eager for challenges. Once they reach puberty this often changes but it need not.
Young children are not self-conscious about using the language, and they are observant. They do not have preconceived ideas about what the language should be like or sound like and they do not have grammar theories. They do not ask why the language is spoken a certain way. They would have no interest in the fine points discussed on this forum.
The best way to learn a language is to imitate children.
"Man is most nearly himself when he achieves the seriousness of a child at play"
500 BC (or whenever)
That is the attitude needed to be a good language learner, the attitude of a child at play.