Nobody Knows How to Learn a Language
blog.usejournal.com
blog.usejournal.com
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. They were never forced to speak it. They could get away with only speaking it a little. And if they got in trouble they could always ask someone for help.
The stress caused by having to come up with the proper word combination helps immensely. We tend to remember what caused us stress. Also, repetition will always help in acquiring new words.
So, basically, start a learning program but make sure you practice every day by speaking with someone that's fluent in your new language ASAP and let them correct you as needed. Also, NEVER speak your native language with that person. Only use the language you want to learn with him or her.
Do not start book studying, grammar and such, until you have a good grasp of the language. Reading a new language will reinforce accents since we tend to read with our language in mind.
The more you have to think in the language you are trying to learn the faster you will reach fluency.
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 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.
"Everyone wants to be a bodybuilder, but no one wants to lift no heavy ass weights."
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.
I live in a foreign country and this is my life, while english is not really the best here - I've had literally zero issues with either speaking english or having some help every now and then (like for 'official matters').
>>> The stress caused by having to come up with the proper word combination helps immensely. We tend to remember what caused us stress.
I also think this is very accurate, most of the words I know are really set in stone from these kinds of occasions. Some specific event happened related to the word, maybe I misspoke something and it was pointed out. And so on.
You don't have to be quite that strict, it's enough to have dedicated time slots where you won't allow yourself to escape learning by switching languages. The most efficient way is to find someone who wants to learn your language and do tandem learning, so they're motivated to slog through correcting your mistakes over and over. Since I'm not that good at small talk, I had some trouble finding activities where there's actually enough to talk about, and eventually settled on museums.
Another thing that helps is to watch videos in your target language. It doesn't do much to improve your vocabulary unless you pause to write new words down, but it helps with listening and by proxy also improves pronunciation. I've found https://viki.com to have a nice collection of subtitled shows in the languages I'm interested in. (Though by now I've progressed enough to keep subtitles off.)
For quickly learning and actually remembering vocabulary, it's probably unavoidable to use a spaced repetition system like https://apps.ankiweb.net/
Listening to the radio, reading one newspaper article a day whilst following an intensive language course helped the most.
It also helps to get drunk from time to time, because your inhibitions drop and you practice better uninhibited (like a child does).
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 actually learn it to fluency. At some level you actually just have to care about the outcome. It's an emotional switch that once flipped, actually the method is irrelevant. I've seen every which way work. The constant was that it mattered to those people. And no one knows how to turn that switch on/off. Therefore I agree with the sentiment no one knows how to do it. I guess I'd be more specific and say that no one knows a fool proof method to engineer the outcome because it relies on emotional states and belief systems. As for the intellectual understanding of how to execute against learning a language. Plenty of seasoned language learners know full well what works and what doesn't.
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.
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.
Surely that's not true: otherwise, children wouldn't have learned their native tongue(s) the way they famously do.
Not sure what kind of theory the author's been immersing themselves with. When I was studying for a language tutor, along my linguistics program, the context was put forth as the most important part of learning anything, let alone a language.
> I studied French in elementary school — supposedly the “ideal” age to learn a new language, according to some people — and somehow continued to get As and Bs while retaining almost no ability to speak the language at all.
Perhaps they haven't because language acquisition, like most any mental skill¹, requires an interest to maintain, let alone improve. My suspicion is this is also why people dislike maths so much: because it's a tenuous process that requires considerate effort to develop. Without a passion for it – often gained either through a natural aptitude or an inspirational teacher – it might be difficult, in those two fields, or in any other of similar requirements, to obtain and retain the skills necessary.
¹ I have a sneaking suspicion – and might well be wrong – that, in physical skills, muscle memory or related mechanisms may affect retention. There've been studies suggesting that writing things down helps memorization, which is a rather "muscle" activity. Perhaps the latter is related to the former?
Is this really as notable as the author suggests? Are there more people proficient in other subjects learned in the classroom? I don't think more than 1% of people from my classrooms are proficient in maths / chemistry / physics to a "fluent adult language" degree.
Similar: It's notable that I've got 2 legs - as a living creature, due to adaptation, etc. It's not notable in a town of people - they all have 2 legs.
For example :
"I know" : Ya znayu
"We know" : My Znayem
"He/She Know" : on znayet/ona znayet
"You know" : Ty Znayesh
"They know" : Oni znayut
etc
Knowing that 'znat' means 'To know' might not be enough to place a word in a sentence
Some suggest that it might be better to memorize sentences or phrases instead
Doing it in real-time when speaking is more of a challenge though.
1. search for e.g. french (https://quizlet.com/subject/french-basic-vocabulary/)
2. look for the list(s) with the most words
3. check quality (if you can :)
4. copy them to your language directory
5. repeat
if you speak German, the addition of "Langenscheidt" could help...
Example (contains about 6000 words): https://quizlet.com/ThomasBisig/folders/franzosisch-wortscha...
One thing I think I got from this: languages are spoken not written. Once you start writing a language it becomes something else. In some cases, you even end up with an official language. A standard.
Human language is the use of sounds to negotiate meaning. So yes, immersion really is the only way to learn because with immersion all you really care about is whether meaning can be transferred around, not whether it's "right" or not. All of that correctness stuff is a different thing entirely than language. Perhaps the author's goal might be "speak language X well enough that others don't realize it's their second language". It's a laudable goal. I'm just not convinced it qualifies as learning a language. (In the colloquial sense, sure.)
People are actually experts in creating and learning languages. They take what they know, extend it a little bit, and work it into some pattern of speech that conveys meaning. Any time you hear somebody who's new to your country speak in a way that half their old way, half the new way, you're hearing somebody create a language.
But if you want to start fooling people, to invisibly blend into a society and not give anything away with your speech, I think you're going to need stressors, either internal or external. Otherwise your brain is going to learn just enough to get by, just like it learns everything else.
You don't quite get the grammar and native speakers will notice that, but so long as they correct you, eventually you'll just pick up the rules.
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.
I don't think of English as a formal structure except for when I'm editing a written document.
Fluency is the ability to carry on an uninterrupted conversation with another fluent speaker.
Immersion (true immersion, where you don't try to cheat) is the best way to become fluent.
You've gotta be Daniel Day Lewis.
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 proficiency. They take 3-5 years to learn fluency up to a basic level. If you learn as fast as a child, it will take you 5 years to be able to speak like a 5 year old.
Technique is important, but it is late and I don't want to write a massive post about it (again). Here is the short version. Study every day. Don't miss a day. If you study 3 times a week, you will plateau at a very low level. If you do less than that, you won't get past baby level. You will eventually have to "study" at least 1 hour a day, but you can start with any small amount you want (even 5 minutes is enough) -- but don't skip days. If you miss a day (or several), you can't make up for it. Don't miss days.
Only study language you understand. If you can not understand a sentence at least 95%, you will have troubles (the 95% number did not come out of my ass... it is important). Always aim for 100% comprehension. Whether you study grammar explicitly or not is up to you -- do it if you enjoy it, don't if you don't. However, do not practice constructing sentences from "first principles" using grammar. This will mean that you acquire non-idiomatic language. Only use grammar to check that your grammar is correct, not to produce sentences.
In order to learn sentences, it is enough to expose yourself to them (as long as you understand them). I found it faster to memorise exemplars. YMMV. If using flash cards (with or without SRS), always practice from your native language to your target language. Never the other way around. It is important.
You must also practice forming sounds. You can "mirror" recorded audio, but sometimes it is too fast. To combat this, get recordings of songs and learn them. Try to perfect them. Record yourself and check to see how close you are.
Always try to practice with native language sources. It can be a live person, TV, radio, or written language. Reading written language is usually best for the bulk of your learning. This is because you can easily get a lot of it and you can easily go through it again and again. When reading, make sure you understand 95% (there's that number again). If you don't, find something simpler, or find out what the content means and read it again. If the latter, repeat the section every few days until it seems like your native language.
Take every opportunity to speak the language. However, beware. Most people go through a "closed period". This is a period where you can understand a lot more than you can say. Often you can't say anything. This is normal. It means you are around a 2-3 year old native level of speaking. Congratulations! Don't let it bother. Keep going and you will definitely break through it. Most failures happen because people quit (duh). Most people quit at this stage.
There is a lot more, but that's probably enough to get anyone started. Many languages share a lot of grammar and vocabulary. If your native an target language are related, you can often "learn" a language (to a child level of ability) pretty quickly (say a year or two). If you are choosing a very different language, it will take you longer (3-5 years). No matter which you choose, you won't get adult level of fluency and proficiency in less that 5-10 years. Anybody telling you differently is selling you something that doesn't exist. Languages are huge -- you have to be in to for the long haul.
It's very fun and rewarding to just dive into pop culture and literature and enjoy that one aspect. I've loved it every time I've done just that dimension. Though you inevitably wind up feeling like you want to speak with people.
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.
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.
But anyway, I've tried the 95% comprehension technique on my students and the results were pretty fantastic. I highly recommend it.
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.
I think the problem, at least in schools, is the dedicated grammar classes, which have no apparent meaning or use. It's best to tackle grammar using practical examples, with an understanding that grammar is not just an end in itself, but something useful to be exploited during speech.
The thing is... There are people like that. They may not know how to teach that to others, but if you look at simultaneous translators, they don't just know another language. They know 4 or more fluently. And maybe pick up another one over summer holiday just for fun. (At least the ones I know did that)
- For a child, every lesson in everything is a language lesson. Everyone you come across will correct you. Everyone will add to your vocabulary. Say you're cooking, or going to the zoo. You'll learn what pots and pans are, all the animals in the zoo, some verbs that are relevant.
- English is very relaxed about its pronunciation. If you're British, "scarf" and "giraffe" rhyme. In American English, they don't. Yet nobody meeting someone from the other side of the pond considers either to be wrong. When English speakers go and learn another language, there's a lot of these close sounds that are wrong. In fact just yesterday in a Mandarin class a guy asked the teacher what the difference between "车" (car/vehicle) and "吃"(to eat) is. He couldn't hear it until we tried a few times.
- English is everyone else's second language because pronunciation is not a big deal. You can often tell if someone is a foreign speaker from their accent, but it's still understandable and nobody complains. By contrast most people in Europe will just swap to English if they hear you say something slightly wrong.
- Institutions are bad at teaching languages because they're not geared towards teaching, they're actually just indices. When you learn something like Linear Algebra or Data Structures, you don't learn stuff in class as much as you learn that they exist. So you go to a DS class and you find out there's a thing called a hash map. You go home, fire up your toolchain, and play with it for a few hours until you get it. Everything you need is out there on the internet somewhere, in English. Same with just about everything else, you learn on your own what's been mentioned in class. If you sit in a language class, you can't do that. Nobody can hear if you are pronouncing things correctly. Also, there's a minimum amount of structure and vocab you need to be able to say anything useful, and you're constantly running into missing ideas, eg "how to I say 'used to' in this language?".
- I suspect the real reason people find it hard is economics. Contrary to this article, I don't think it's actually particularly valuable. This is why immersion doesn't happen. Kids have low opportunity cost, this is why they are able to spend their youth learning a bunch of things. If you move to a new country as an adult, chances are your livelihood depends on some specialist skill that is supported by English, and the rest of the time you're taking care of your family. So you're not getting immersion. You also don't get cultural immersion from the language class. I'm not talking about the fact they eat baguettes in France; your knowledge of France as a culture requires you to access umpteen levels more information than people normally do. How many people are going to know what the French think of secularism, or the Dreyfus affair, from just studying French? I don't think I'll be coming across essays about the cultural revolution in my Mandarin class, either. And then add to that cultural items that every local knows, but isn't considered historically/politically important. The most popular Danish song of all time?
They don't rhyme in British English either. Maybe in one of the regional dialects - I think that you could probably force it in a scouse (Liverpool) accent.
But yep if you go oop North then A's become short and giraffe will rhyme with gaffe (like bath and laugh will).
Furthermore, cantonese is a bad comparison to english because it's a tonal language; i.e. the inflection of the word changes the meaning. So on top of having to learn more or less vowel sounds, students have to struggle with a fundamentally different way to conceptualize language.
> English is everyone else's second language because pronunciation is not a big deal
That is not true. English has become the lingua franca of the day because England & America have dominated the cultural and commercial global market for centuries. The adoption of lingua francas has nothing to do with how easy it is to learn a language (after all, this is incredibly relative) and more with how international trade and history have shaped a region. For example, in the late bronze age the lingua franca was Akkadian, a semitic language that would've been pretty tough to learn for, say, indoeuropean speakers like the greeks.
Having moved to an english-speaking country because my husband is a native english speaker, I find that even with the language immersion is difficult. Unless you have a very particular hobby that allows you to connect to other people, it's difficult making friends because of the cultural differences. And even though at this point I'm so fluent in english I've written and published fiction stories in it, it's still a difference that can be felt - imagine being barely fluent or speaking barely a few words.
Sometime the voices in the BBC are different too now that they've loosened the standard.
The underlying linguistic issue here is that when we're young our brain trains itself to classify speech sounds into categories based on the speech we hear around us -- so sound differences that matter in our native language(s) go in different buckets, but differences that don't matter in that language are ignored. That makes it harder to learn a different language later where the bucketing is different because our brains naturally ignore sound changes that should be significant, and need retraining.
Yes, most times you don't really need to know a language. In the old times in europe, many village people never left their home village. As long as you can understand everyone in your village, everything is fine. Traders, priests etc had to learn multiple languages and often used latin as lingua franca.
Nowadays, in the western world, knowing english is often enough to communicate with most people you have to do in various areas of your life.
The exception confirms the rule: sometimes, languages are useful for jobs. In the CIA for example, they actually do encourage employees to learn languages. Because knowing the language of the country your work is about is actually helpful.
I used "came" and not "come" since it just feels more appropriate considering it's two different countries countries now.
I don't think I've ever heard about "split infinitive" before.
Perhaps I did - and it's just that the "school way" of learning foreign languages always confused me.
My first foreign language was German and at this point I basically can't have a conversation using it - at best I can kind of understand what someone is saying.
Though to be fair living in The Netherlands Dutch kind of highjacked it.
The way I learned English was simply pure interest. Watching Hollywood and British movies and TV series, reading computer/programming books and other "stuff on the Internet".
Teachers were equally confused by the fact I have had problems with recognizing/naming different tenses and grammar forms - and still getting it (mostly) right just by "gut feeling".
And actually I've been able to "think in English" for a long time. So long that these days I occasionally have trouble coming up with idiomatic way to say something in my native/mother tongue (Serbian).
> Case in point: did you catch the split infinitive in the previous sentence?