Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Nobody Knows How to Learn a Language

blog.usejournal.com

Post reply on HN
Print view View original

51–60 of 69 posts

Re: Nobody Knows How to Learn a Language

#51

Post by james_s_tayler »

mikekchar wrote: post #22

I'm going to skip my usual really, really, long explanation. Long story short: TFA is wrong. Many people know how to learn a language. Many people learn several languages because after learning how to learn one, they realised it's not that hard. Main problem people have: they have unrealistic expectations. Children don't have a magic ability to learn languages. They take 10-20 years to learn up to adult level profici…

Although I agree with plenty of things you say, and we probably have similar experiences learning language, the strength of your assertions to always use i+1 (95%) and to always go from native to target when doing flashcards are unwarranted.

That's A method. Or A way to do it. There are many paths up the mountain. Granted I appreciate it when people explicitly lay out "here's the path I walked up" as that's good information. But there exist other paths equally as fast and maybe some even faster, many slower of course but there really is a surprising multitude of techniques that wind up at similar results.

I remember sitting and engaging with content that I had fuck all idea what was being said but I was so damn delighted every time a word came up that I knew that it just lit the fire and kept it burning. I did plenty of i+1 flashcards which were sentences (both reading and listening) purely from target to native. I quickly established a foundation on which I could parse any sentence, I just needed the vocab. Then I did nothing but engaging with content I was interested in regardless of difficulty and hung out with friends getting in 10+ hours conversation (one on one plus group situations) per week. End of my second year of doing that I could hold my own in a lengthy one on one for hours on end and end of year 3 I could hold my own indefinitely in groups.

Was a blast. So long ago now.

Re: Nobody Knows How to Learn a Language

#52

Post by physicles »

The hardest part about learning a new language is maintaining motivation.

Learning a new language isn’t a hobby you can just add to your life like knitting or cooking. It’s more like an ambitious diet and exercise program, except the payoff is even less motivating. If you look good, everyone sees. If you can speak Japanese, nobody cares but other people you talk to whose English is worse than your Japanese. Like exercise, if you invest anything less than four hours a week, you won’t make meaningful gains, and it’s often months before you get anywhere. It’s a slog.

The trick (if you can call it that) is to find something you really want to do, badly, that you NEED your L2 to do. For Japanese learners, maybe maybe wanting to watch anime is enough. For English learners, maybe it’s watching American TV and movies. For Americans learning an L2, immersion is the pretty much the only viable approach.

Immersion is just a hack to motivate yourself by making your life almost unbearably difficult, with learning to speak your L2 the only way out.

Looking back on six years living in China and learning Chinese while working (borderline fluent in some topics but woefully inadequate in others), if I could do it over again I would:

1. Spend a few months getting the fundamentals down: learn pronunciation, drill it with a patient native speaker, go hard at 10+ hours a week book studying. (I did this)

2. Move somewhere to either work or study only in your L2, and where there’s barely any English speakers. Live there for a year or more. (I didn’t do this; I lived in a large city and had mostly English speaking friends and my work was in English)

There’s no way to fluency without spending hundreds of hours talking with people in your L2: ~500 hours for Spanish, ~1500 for Chinese or Arabic.

Re: Nobody Knows How to Learn a Language

#53

Post by mikekchar »

jasim wrote: post #42
mikekchar wrote: post #22

I'm going to skip my usual really, really, long explanation. Long story short: TFA is wrong. Many people know how to learn a language. Many people learn several languages because after learning how to learn one, they realised it's not that hard. Main problem people have: they have unrealistic expectations. Children don't have a magic ability to learn languages. They take 10-20 years to learn up to adult level profici…

If someone wants to learn a language just to enjoy its language & literature, and not necessarily be conversant in it, then would it be okay to never speak it? Or is conversation an essential practice?

If you are asking, can I learn a language enough to enjoy literature and studying the language without engaging in conversation, the answer is "yes". I know lots of people who do this (especially in Japanese since a lot of people only study it in order to enjoy manga and anime in Japanese).

Interestingly, there is a prevailing theory that even if you never have a conversation, you can eventually acquire the language (see the work done by Stephen Krashen). There are examples of severely abused children who were able to acquire language without being allowed to converse. I would say it's not the fastest way to learn, but it seems it may be possible.

The main thing is that you need to be able to make a mental map of the language and then check your mental map against reality. Conversation is a good way to do that because it allows a native speaker to quickly point out problems. There are other ways to do it.

A lot of language is culture, though, so it can sometimes be hard to truly understand some language without having the same experience. Usually this requires experiencing social events with real people, which also usually requires considerable conversation. But if enjoyment is the goal, it isn't necessary to understand absolutely everything.

Re: Nobody Knows How to Learn a Language

#54

Post by WheelsAtLarge »

james_s_tayler wrote: post #41
WheelsAtLarge wrote: post #2

The answer is no secret but it's hard to achieve. The fastest way to learn a new language fluently is to speak it, constantly. Yes, it's easier for children to acquire a new language but they learn it because they are constantly forced to use it. I've known a number of people that have lived in a foreign country yet they did not learn the language. Why? Because they could get away with using English most of the time.…

I pretty much did this with Japanese but there are a few caveats. Namely informal grammar study as fast as possible and learn to read quick as you can are also massive, massive boosters. Of all the language learning methods I've seen I've only seen one thing consistent across all of them that in the end leads to results and that is you have to legitimately feel that it's deeply important to your life story to actuall…

Excellent point, caring at the deepest level will make any method work. But a system that pushes you along will help. The goal is to be fluent in a language but it's easier if you get pushed towards your goal by circumstances rather than actively having to push yourself.

Example, you can read and practice a conversation by yourself and learn, something that gets harder as time passes, but if you have no choice but to communicate with someone then you will do it even if you don't feel like it.

Re: Nobody Knows How to Learn a Language

#55

Post by tomxor »

injb wrote: post #3
WheelsAtLarge wrote: post #2

The answer is no secret but it's hard to achieve. The fastest way to learn a new language fluently is to speak it, constantly. Yes, it's easier for children to acquire a new language but they learn it because they are constantly forced to use it. I've known a number of people that have lived in a foreign country yet they did not learn the language. Why? Because they could get away with using English most of the time.…

This. The author of the article says: "If there were someone who knew how to learn new languages, we would all know it. They would be uniquely and unquestionably skilled at producing new language speakers. They would be very visible: everybody would be flocking to their doors and imitating them." He seems to be confusing "learn" with "teach". All of the above would be true if he said "if there were someone who knew h…

> He seems to be confusing "learn" with "teach". All of the above would be true if he said "if there were someone who knew how to teach new languages...". Lots of people know how to learn, but no one wants to hear about it, because the answer is: work really hard.

Exactly, the author is talking about teaching from the perspective of a _shortcut_ (which most teaching does, since there would be no point to teaching if it didn't enhance the learning beyond making information available), but many skills or working knowledge are "irreducible" from a learning perspective, in the sense that they must be acquired from the ground up.

He implicitly states this goal much later in the article here:

> learn a new language like a baby! Start learning by babbling single syllables and eventually move on to whole words and sentences. Culminate in advanced classroom instruction. The problem is that, taken as a literal model, this will take you 18+ years. This is not a practical way to learn a new language.

We can't assume that something is always reducible (mathematically there are many things that a provably not!), that there is a trick to it that we can cheat nature - that is the old scientific way of thinking.

All language teaching methods i've been exposed to assume this to be the case as they are all based on the idea of bootstrapping the process with our existing native language... I suspect this never actually works, just that a few individuals who are intuitively clever enough quickly and unconsciously discard this relationship as a basis of the new language.

TL;DR when the author debates whether it is actually possible to "teach" a new language, what I see is the question of whether the process of fluently learning a new language is computationally reducible.

Re: Nobody Knows How to Learn a Language

#56

Post by mikekchar »

james_s_tayler wrote: post #51
mikekchar wrote: post #22

I'm going to skip my usual really, really, long explanation. Long story short: TFA is wrong. Many people know how to learn a language. Many people learn several languages because after learning how to learn one, they realised it's not that hard. Main problem people have: they have unrealistic expectations. Children don't have a magic ability to learn languages. They take 10-20 years to learn up to adult level profici…

Although I agree with plenty of things you say, and we probably have similar experiences learning language, the strength of your assertions to always use i+1 (95%) and to always go from native to target when doing flashcards are unwarranted. That's A method. Or A way to do it. There are many paths up the mountain. Granted I appreciate it when people explicitly lay out "here's the path I walked up" as that's good info…

Sorry. You are correct. I keep getting obsessed with efficiency :-) I also did what you did. However, I have discovered that you can learn faster than that if you maintain 95% comprehension rate (it's the amount that's required to infer language from context). I used to have a list of references on this topic, but sadly I lost it somehow (along with a book on teaching English I wrote -- astonishingly I did not have backups). Since I returned to IT, I haven't had the time to go back to this :-(

But anyway, I've tried the 95% comprehension technique on my students and the results were pretty fantastic. I highly recommend it.

Re: Nobody Knows How to Learn a Language

#57

Post by mikekchar »

v4ult wrote: post #48
mikekchar wrote: post #22

I'm going to skip my usual really, really, long explanation. Long story short: TFA is wrong. Many people know how to learn a language. Many people learn several languages because after learning how to learn one, they realised it's not that hard. Main problem people have: they have unrealistic expectations. Children don't have a magic ability to learn languages. They take 10-20 years to learn up to adult level profici…

I found your comment interesting. If you have a longer version or blog post written anywhere I'd like to read it.

If you look through my history long enough you'll see several versions :-) I'm afraid I don't have anything current that's public. Some day I'll try to find time to write up something more concrete.

Re: Nobody Knows How to Learn a Language

#58

Post by james_s_tayler »

mikekchar wrote: post #56
james_s_tayler wrote: post #51

Earlier quoted context omitted.

Although I agree with plenty of things you say, and we probably have similar experiences learning language, the strength of your assertions to always use i+1 (95%) and to always go from native to target when doing flashcards are unwarranted. That's A method. Or A way to do it. There are many paths up the mountain. Granted I appreciate it when people explicitly lay out "here's the path I walked up" as that's good info…

Sorry. You are correct. I keep getting obsessed with efficiency :-) I also did what you did. However, I have discovered that you can learn faster than that if you maintain 95% comprehension rate (it's the amount that's required to infer language from context). I used to have a list of references on this topic, but sadly I lost it somehow (along with a book on teaching English I wrote -- astonishingly I did not have b…

I don't disagree with 95% it's definitely a magic formula. What I did was do my focused study where I would sit down and do flashcards etc in the 95% format and then for hours a day after the focused study side of things was taken care of I'd dive into completely unstructured learning. That's been my go to ever since.

Also, I suppose I should clarify a little around target to native. I tried that a couple of times. Maybe even 3 times seriously. Intellectually I always came to the conclusion that if you can pull off doing pure output in native to target over a wide enough range of sentences then the whole structure of the language will crystalize really fast. I could never make the emotional side of it work. It always felt too grindy for me and I had to overextend on spending effort against that one task and couldn't sustain it. So I'd just slide back to my tried and true method because tough as they were I could always make the emotional side of that sustainable.

I think it's actually a critically important point. I love it the idea of it as a method and if the emotions of it are workable then it's probably an epic path.

Re: Nobody Knows How to Learn a Language

#59

Post by jhbadger »

jasim wrote: post #42
mikekchar wrote: post #22

I'm going to skip my usual really, really, long explanation. Long story short: TFA is wrong. Many people know how to learn a language. Many people learn several languages because after learning how to learn one, they realised it's not that hard. Main problem people have: they have unrealistic expectations. Children don't have a magic ability to learn languages. They take 10-20 years to learn up to adult level profici…

If someone wants to learn a language just to enjoy its language & literature, and not necessarily be conversant in it, then would it be okay to never speak it? Or is conversation an essential practice?

Of course. For example, while there are a small number of "living latinists" who actually want to speak the language, the vast majority of people studying Latin just want to read it, and the teaching material reflects this (focusing on interpreting text rather than composing it or speaking it) Same with other classical languages such as Ancient Greek, Sanskrit, Ancient Egyptian, and Babylonian.

Re: Nobody Knows How to Learn a Language

#60

Post by james_s_tayler »

WheelsAtLarge wrote: post #54
james_s_tayler wrote: post #41

Earlier quoted context omitted.

I pretty much did this with Japanese but there are a few caveats. Namely informal grammar study as fast as possible and learn to read quick as you can are also massive, massive boosters. Of all the language learning methods I've seen I've only seen one thing consistent across all of them that in the end leads to results and that is you have to legitimately feel that it's deeply important to your life story to actuall…

Excellent point, caring at the deepest level will make any method work. But a system that pushes you along will help. The goal is to be fluent in a language but it's easier if you get pushed towards your goal by circumstances rather than actively having to push yourself. Example, you can read and practice a conversation by yourself and learn, something that gets harder as time passes, but if you have no choice but to…

The caring is the push. The caring is the pull. The caring is the forward momentum.

So many people I know lived in Japan 5 years and can barely speak it. Some are even wholly dependant on their wife to manage all their affairs because they can't function in society.

I never set foot in Japan before I was fluent because I moved to the city where all the exchange students came to learn English and I did nothing but hang out and speak Japanese with them.

I know what you mean. Methods can make or break your success in a way. But the method isn't the reason for it. It's the conduit. I remember when I first started learning I went to the bookstore and got one of those shitty teach yourself Japanese phrase books. I got 1/3 the way through then I got stuck on this page and I just couldn't get it. I couldnt get how to count some stuff and I couldn't move on until I got it. I threw the book in the trash but not my dream. I went in search of another way to learn it. I came across many obstacles and the caring led me to always search for another way.

Today the last of the caring is ebbing out of me after 10 long, fun years. And with it is going my skills.

Post reply on HN