Hacker News Reader

There are different valid approaches, but there are certainly techniques that are ineffective; I think the popular "just do something" advice is harmful as it suggests all approaches are equally valid. I think it's possible to develop an intuition about whether an activity is more or less beneficial for your learning.

Here's the best advice I could write down: https://www.reddit.com/r/languagelearning/comments/66myn1/a_...

There is a huge gap between "something is better than nothing" and "all things are equally good."

'Language learning advice is conflicting,' 'Stick with a language, Whatever your method is, don't give up.' 'If your method doesn't work for you, change it.'

I agree with the author. I think a lot of language learning advice is of poor quality, or it's hard to establish complex behaviours simply from reading an article. It's hard to write stuff down that you could expect to improve the odds of the student. Learning a language on your own is a kind of a funny hack, it's like learning to program without access to a computer, it can be done but it needs specific techniques. I make software for learning languages, writing proper instructions for using it is something that I consider difficult and I don't attempt it. :)

There is definitely a difference in the quality of content you consume though, which is a bummer - new language learners don’t know what’s good. I just ran across a YouTube channel [1] which if I had known about it when I was starting Japanese, would have probably prevented me from getting confused about a great deal of things.

But there really is no substitute for persistence, regardless of technique.

[1] https://youtube.com/playlist?list=PLg9uYxuZf8x_A-vcqqyOFZu06... (though her pronunciation is very unnatural, her explanations are spot on)

Okay, but can't we objectively experimentally compare them? Take a class of 20 undergrads, have them use method X. Have the next class use Y. Repeat with as many methods as needed, with as many classes/semesters as needed, control for as many confounding factors as you can. Publish papers, reproduce. Is the noise floor too high, the effectiveness too close, or am I missing something? (Yes, probably that last one; if so, what?)

I think that the field is fraught with a number of caveats, just off the top of my head:

* "I learned a language" is a very poorly defined statement. It's really difficult to quantify. The EU tried with their language levels [1], but it's still a mess. So you get a person that says "I speak 15 languages! Just follow my method forehead!" while an academic says "this person speaks two languages barely, and makes noises in the rest."

"I know a language" means different things for different people, depending on one's motivation to learn a language. Do you want to book a hotel room? Do you want to read a newspaper? Do you want to read Crime and Punishment in the original? Note that, for example, reading a newspaper is significantly easier in Tagalog, which uses a mixture of English and Filipino, than in Chinese, which requires you to know around 3 thousand Chinese characters.

* Learning a language is highly influenced by your native language. If you take 20 undergrads, you might conclude that "to learn language X as a native speaker of Y, the best method seems to be Z." Then, other papers will likely dispute your definition of "learning" (see previous point).

* Learning a language is highly dependent on your motivation. If you take a class of 20 people, their motivation to learn a language X is going to be all over the place. Learning a language is always really difficult, even when a language is deemed easy (like Spanish). Most people will drop off unless they are really driven to learn the language.

Long story short, I don't think it's as easy to apply the scientific method. There's a reason why we don't have a matrix of "if you're a German speaker wanting to learn Russian, here's the optimal syllabus for you."

On the flip side, I do think we can say something about language learning. For example, apps like Duolingo are not that useful. I don't think that's a very controversial statement in the language learning community.

Similarly, just moving to a country won't do anything special for you (other than utterly isolate you). For example, you won't pick up any Japanese from your one-room 40-squared-feet apartment in downtown Tokyo alone.

[1] https://en.wikipedia.org/wiki/Common_European_Framework_of_R...

> aha, you've been doing it wrong, you could have learned Chinese in 3 months!

Honestly there kind of is, you just have to put yourself in a "sink or swim" situation as the article says. You'll be surprised how far you can get in 3 months if you are in such a situation.

Also, don't believe any crap when someone tells you "X language is hard". This is perhaps the biggest myth I have ever seen. It's only hard if you think it is, and if you try to learn everything by rules and flashcards, which native speakers don't generally do. You can also try to make anything look hard that actually isn't. I find people in the US especially love to talk about how hard Chinese (and also Japanese and a few others) is but it isn't particularly different in difficulty than English or any other major world language. 5-year-olds can handle it, they just weren't told it's hard, they just needed to speak it if they wanted to argue and get their stolen pencil back.

Any grammar is surmountable, but the languages that are truly hard are ones which have sounds or consonant clusters that your native language(s) don't. For example, learning Georgian as a native English/Italian speaker.. I could hardly even say basic words like "forest" (t'ke), "water" (tsqali), or "frog" (baqaqi) for the first few months. Look up the ყ letter, or listen and judge for yourself ბაყაყი წყალში ყიყინებს (the frog cries in the water) Hear natives speak it: https://forvo.com/word/%E1%83%91%E1%83%90%E1%83%A7%E1%83%90%...

The one by shaburgiorgi is the most faithfully enunciated, while the others are what you would more likely hear casually/in conversation.

> I could hardly even say

If I may take the liberty to nitpick, that's a thing only an adult would say. A 5-year-old would just try.

That's a pretty unforgiving interpretation of what I wrote. Sounds came out of my mouth and people didn't understand what I was trying to say.

> Sounds came out of my mouth and people didn't understand what I was trying to say.

Right, and as a 5 year old who wants your pencil back, you find another way to say it. If you don't know how to say "backpack" you say "bag". If you don't know how to say "foyer" you say "that place near the door". If you don't know how to say "pencil" you say "pen" or "writing stick".

I realize it was an unforgiving interpretation. That was directed at the topic and not you personally. That's kind of my point. If you have a sink-or-swim attitude or situation you will learn it.

> A 5-year-old would just try.

And a 5 year old would be cut a lot of slack for screwing it up.

One of the issues with learning anything as an adult is that other adults are viciously unaccommodating. Foreign language learning turns that up to 11.

Personally, I think this is the primary issue with adult learning. You have to be a very strong personality to have an intrinsic motivation that will cause you to put in long hours of work because you aren't going to get any external motivation for a very long time.

Interesting take. Are they?

I know adults in the US are often sadly viciously unaccomodating of broken English, but I feel that isn't the case in most of the (especially non-English-speaking) world, where any attempts to speak the local language are usually very much welcomed.

There are some places where the locals tend to speak good enough English that if they realize you speak English better than their language they'll viciously insist on speaking English to get business done, but that is sometimes mitigated if you place yourself in a more rural area or less cosmopolitan city where English isn't common.

It depends how racist the country/situation is. Many countries will give English-speakers a break because many English speakers don't even try. So just being able to say something - anything - in the language counts as a plus. Anything after that is a bonus.

Even in English-speaking countries, you can screw up the grammar badly and still be comprehensible. US and British people may decide you're an idiot - and treat you accordingly - but they will mostly understand you.

So my personal ordering of importance is vocab, grammar, and finally accent/voice coaching. You will definitely need the latter for most languages, because even supposedly non-exotic ones like French require a novel set of mouth movements. Most lesson systems underemphasise this.

My working assumption as an adult is that it will take three months to learn basic conversation with full immersion and minimal distractions, and five years of constant practice to reach reasonable fluency in speaking, listening, and writing.

This is actually less time than it takes kids to learn a language, so the idea that kids are especially malleable or open is clearly wrong. Of course kids are learning language in general - and more - at the same time, so timings aren't absolutely comparable. But similarly adults are having to do a lot of distracting adult things at the same time, so it more or less evens out.

This doesn't match my experience at all, you just have to have a good sense of humour.

One of the first things I learn in a language is how to say some variant of "I'm a little puppy", spring chicken, I speak like a 2 year old, etc.

You're never going to sound like the Queen, no-one reasonable expects that.

Even listening is hard.

I spent quite some time with the Swedish retroflex fricative, and always heard something else than my teacher described. I finally settled on learning it by IPA, i.e. learning where exactly the tongue needs to be.

Even harder are Russian consonants for me: I can try as much as I want, but I will probably never be able to distinguish Russian hard and soft consonants, even with a Russian speaker saying them slowly and in contrast (and with lots of "you're kidding me, right? Those sounds are nothing alike").

I can completely relate to everything you wrote haha. I never mastered ш (which I guess is retroflex fricative?) vs щ and I've completely given up on hard vs. soft sounds in Russian because this

>and with lots of "you're kidding me, right? Those sounds are nothing alike"

Is too true. With practice and listening, my pronunciation is naturally improving over time, but I really don't know how to improve it in a faster way. Do you think learning and practicing IPA would do the trick, like you did with Swedish?

Learning about phonetics absolutely helps. It provides an accurate way of precisely describing how sounds are pronounced.

Do keep in mind, though, that the IPA classifications that you find on Wikipedia are sometimes rough estimates. With particularly difficult sounds, it may be a good idea to look at multiple phonetic descriptions.

Begin with the sounds of your own native language, which you already know how to pronounce, and try to describe them phonetically. A tip for learning to feel where your tongue is, is to pronounce the sounds without tone, i.e. whisper them, or even without air, i.e. simply mouth them.

Depends on your native language there are harder and easier languages for sure. It is much easier for a Chinese to read and understand Japanese comparing to someone who don't know the symbols and need to learn it from scratch. It is easier to learn other languages from the same group, like learning a different latin language, germanic language or semitic language if you know one of them already. There are other factors also like similarities in grammatical structure, in variations of the language around situations which require more politeness or formality like in Japanese, written vs. spoken language like in Arabic and many other factors.

There's a variety of factors that make language easier or harder to learn. Broadly speaking, you can divide them into writing, grammar, vocabulary, and phonology.

So there's languages that are hard to read and write. People love to talk about how poor English is when it comes to correspondence between written letters and sounds, but honestly, it's not close to the head of the pact. There are some Southeast Asian languages today where the gap between written word and speech is very much akin to writing Latin but speaking French. Japanese kanji is a moderately different level of burden, as every kanji has multiple different pronunciations, and it's at the point where--even for native speakers--it's not particularly expected that you can read someone's name and know how to pronounce it. Of course, for those of who use Latin scripts, any language that also uses Latin script is going to be easier to learn than one that doesn't.

Phonology can be a challenging aspect, because several languages don't have the same phonotactics as English does (or insert whatever native language here). There are multiple Slavic or Chinese sounds that my ears would process as "sh" and yet would be considered different consonants. Some phonemes are legitimately more difficult to pronounce--pronouncing click consonants literally scars your vocal tract. Some languages demand that you pay attention to tones, creakiness, breathiness, or aspiration, none of which is phonologically relevant to English or most Indo-European language.

Grammar and vocabulary are, as you imply, pretty mild concerns at the end of the day. They're not difficult enough to impede language acquisition, but there is actually some evidence to suggest that a language that becomes a major second language in the process goes through stronger grammatical simplification than their closest linguistic relatives--consider how much simpler English, Swahili, Malay, or Persian are compared to "normal" languages in their respective families.

So yes, I'm aware of analysis like this.

I think my point is, even "polluting" your brain with analysis like this of a language will make it actually hard. For example

> Japanese kanji is a moderately different level of burden

If you didn't read this, you wouldn't actually go into it thinking of it as a burden, and it would turn out to not actually be a burden. The fact that "even" native speakers from time to time have to ask how to pronounce something is just a fact of life, and not a difficulty of the language, but observers have come to think of it as a burden for some reason. But now that you've read that it is a burden, it will be a burden to you.

I have pretty firm conviction that "hard" is very much influenced within some range by how hard you a priori think it is, and that logic is not the way to approach this. Rather, being blissfully unaware of any logical or linguistic experts' arguments around how hard others think something is will actually help you learn the language much faster and more efficiently.

> going to be easier to learn than one that doesn't.

> There are multiple Slavic or Chinese sounds

> Some phonemes are legitimately more difficult to pronounce

> literally scars your vocal tract

> demand that you pay attention to

Without any intent to personally attack you, because a LOT of people have logical arguments such as the above, it's quite literally even hearing such logical descriptions of languages which may make something extremely hard. We aren't fully logical creatures, the languages we invented aren't designed to be logically analyzed.

I live in Japan, and every single Japanese person that I know, and a few Chinese coworkers too, have told me how kanji/hanzi is so ridiculously difficult to learn. I mean, native Chinese and Japanese speakers require at least 12 years of education to learn how to read and write their languages at a basic level. They are the most difficult writing systems in the world by far, saying otherwise is totally disingenious.

Difficult is subjective.

> at least 12 years of education to learn how to read and write their languages at a basic level

The idea of a "basic level" is fundamentally flawed.

Can a 10-year-old child in Japan function to a similar level in their society as a 10-year-old child in England? If asked to water the plants can they do so? Can they catch the correct school bus? Can they read a children's book for their age level? Yes? If so language has served its purpose, and there isn't really a fundamental difference in difficulty level.

Yeah, native speakers will tell you their language is hard, but they too are influenced by the media that all says it's hard.

You might be right that it takes 12 years for the average kid to learn a particular set of Kanji but that's a very arbitrary standard applied to a language that does not demand you know all of those Kanji to be functional in society. There is context, there is the simple act of asking, there is so much more to it. Likewise, English speakers don't demand that they know the entire Oxford English Dictionary to be considered literate. I mean, yeah, if you asked the 10-year-old to "irrigate the agricultural enclosure" instead of "water the plants" you might stump them, sure. If we had those kind of expectations, English would be considered "hard" -- look at that thick book of the OED! It's thicker than the Kangxi dictionary and Xinhua dictionary combined, and therefore it must be harder! (My point is -- logic isn't the way to analyze this.)

> Difficult is subjective.

You're not responding to the evidence presented in this thread that there is evidence that some aspects of language learning is objectively harder. Specifically:

* There's pretty strong evidence (comparison between literacy rates of children with different native languages) that some languages have writing systems that are harder to learn than others. (I feel obliged to point out that writing is the least important part of language acquisition--after all, writing dates to only ~6,000 years ago, whereas spoken language dates at least 100,000 years ago, possibly earlier. Of course, for most major modern languages, a language learner is equally expected to read/write in it as they are to listen/speak in it)

* There is some evidence, albeit more controversial, that there is a critical age range in which it is far easier to acquire pronunciation of phonemes than learning subsequently.

* Again, there is evidence that some phonemes are physically easier to pronounce than others, discovered in part by the order in which toddlers acquire various speech sounds. If you've got a toddler learning to speak, listen to the sounds he/she can pronounce and also the ones that aren't properly pronounced. I wouldn't attempt to chart out a full objective scale on which phoneme is harder than the other one, but I would be moderately confident in asserting that an Arabic ayin is objectively more difficult to pronounce than an `m' sound.

* I also mentioned evidence that some grammatical features may be more complicated: the developments of lingua francas, or additionally creoles, does suggest that languages used heavily as second language acquisition tend to drop several grammatical features.

(Now I should also addend that just because something is objectively more difficult doesn't necessarily mean that is specifically "hard." It is clearly objectively more difficult to lift a box 2 meters than to do so 1 meter... but that doesn't mean that it's hard to do so.)

> comparison between literacy rates of children

I would even go so far as to say that I feel like comparing things like literacy rates (which have very arbitrary definitions) and phoneme acquisition are a very "western" way to analyze a language, and by doing so, Japanese and Chinese seem hard. If they are not analyzed with that kind of logical mindset and rather just "Do people function in society", "can the workings of a complex machine be articulated well" and "Are people able to communicate needs and emotions" I think a very different answer would be seen.

Are you seriously trying to argue that literacy is not a requirement for functioning in society?

That's simply a fact that doesn't need to be argued. Lots of English speakers can't read or write well, because English orthography is bad and hard, or because of dyslexia or blindness. They function in society. They can't do everything, but people who aren't plumbers, programmers, or are terrible at math can't do everything either. They work around it.

edit: I think the functional illiteracy number in the US is something like 1 out of 6. R. Kelly famously can't read.

No, I never said that. I said that literacy has a very arbitrary definition.

Rather, the act of drawing a hard line between literate and non-literate is a very Western way of analyzing language learning and doesn't scale well past a certain set of languages. By drawing such a hard line you end up with ridiculous conclusions such as that a child needs 12 years to be literate in Japanese. Sure, if you go by your arbitrary definition.

You can say that the average kid in an English speaking country needs 18 years before they know what "erudite" and "ostentation" mean if you draw the line there.

Dude, why are you so stubborn? Have you ever tried to learn those languages? Do you really think that learning 20~30 simple characters is on the same level of difficulty as learning 2000~4000 characters, some of them very complex? But sure, Japanese and Chinese people think that kanji is hard because the media told them, not because they had to spend several hours a day of study for 12 years just to reach basic fluency.

> You might be right that it takes 12 years for the average kid to learn a particular set of Kanji but that's a very arbitrary standard applied to a language that does not demand you know all of those Kanji to be functional in society.

12 years is what it takes to learn the joyo kanji list (https://en.wikipedia.org/wiki/List_of_j%C5%8Dy%C5%8D_kanji) (I don't know the Chinese equivalent), a set of 2136 which is considered the *minimum* to function in modern society, what you will find in newspapers and government publication. Can you do with less than that? Maybe, but it won't be easy.

> Dude, why are you so stubborn?

Personal attacks not appreciated here

> Have you ever tried to learn those languages?

Yes? I grew up with both Chinese and English? What gave you the impression I was saying stuff out of nowhere?

> Do you really think that learning 20~30 simple characters is on the same level of difficulty as learning 2000~4000 characters, some of them very complex?

Yes. I'm happy to elaborate as to why I think so but if you're only going to stand on the opposite side and not interested in even trying to understanding my perspective it's a waste of time on my part.

> a set of 2136 which is considered the minimum to function in modern society

For an adult, fine, but (a) "number of characters remembered" not a good way to measure the ability to function, any more than "number of SAT/GRE words memorized" to an English speaker, and (b) to function at an 8-year-old level in society you most definitely don't need 2136 characters.

An 8-year-old in an English speaking country would trip up trying to read The Economist as well.

Ok, then maybe you are a genius, because everybody else has told me how painful it was to learn all the characters. I'm going through the process myself, and while sometimes I enjoy it, there is no point in denying that it is way more difficult than learning an alphabet-based language.

Experience is king. Language is used at a pace where it has to be intuitive. No amount of learning rules is going to give you that. As you use a language you build up intuition on it and you learn the patterns (combos) that go well together. It's actually quite similar to how you learn so many other things from mathematics to video games.

One thing that does set languages apart though is that they involve learning to listen and to speak. With other activities we don't usually have to learn to listen to a new audio pattern to start distinguishing it from noise. Even less frequently do we have to learn to make new sounds with our mouth. In fact, I bet most people couldn't even describe what they do in their mouth to produce the sounds of different characters in their languages that they can speak. (Think of the difference of French R and English R and how you produce them.) Learning those can be quite a new experience, especially if you're still learning up distinguish the sound from noise.

> Honestly there kind of is, you just have to put yourself in a "sink or swim" situation as the article says. You'll be surprised how far you can get in 3 months if you are in such a situation.

This is the problem in language learning. What you said could be true in certain contexts but is wildly misconstrued. As is, what you said is "bro science" without proper methods, context, and caveats. Finally, while immersion in a language is amazingly useful, you can't do much in 3 months.

Sounds like a recipe for burnout and total isolation in a foreign country.

There's not one method, because there's not one type of brain, not one type of learner.

Also, not one type of language.

That’s a myth.