Hacker News new | comments | ask | show | jobs | submit login
Learn to read a sentence of Chinese in 3 minutes (memrise.com)
198 points by bmj1 on May 4, 2012 | hide | past | web | favorite | 73 comments

it's good - Chinese has an extremely steep initial learning curve, and this would be a good start. I also think the methods are sound (ie. mixing the word in, substitution, the way the actual tool tests you).

<pedantry> It does feel very limited. I think you'd outgrow this in a couple of weeks TBH.

For example, this article gives 子 as one of the first characters, and says it means "child". It does, but it also doesn't. On its own it is purely conceptual, and only has meaning combined with other characters. If you said "my 子 is 4" then that makes no sense. It can mean bullet (子弹), atom (原子), son (儿子), or a generic "thing" measure word (eg. 骗子, conman).

An early learner doesn't need to know the word for 'bullet' or 'atom', but it's not good either to tell them that "子" means child and then they think they can say "child" in Chinese. What's the parallel in English? It's like teaching you "pre-" and saying that means "early", and then you run around saying "He arrived pre-". Why not just teach the word for child (孩子) and then substitute that? Then you learn to recognise meaningful words and not conceptual characters, which is actually the key skill in reading Chinese.

You have to start somewhere, but I wish that more Chinese learning tools/books used modern learning methods like these, but with better linguistic accuracy. </pedantry>

Extending this farther, I'd say everything you read or listen to in a foreign language should be in some kind of context, the richer the better. Reading characters outside of their context in words is pretty useless, but reading words outside of sentences isn't that much better, and sentences can be pretty awful by themselves.

For example, in English: "Do you have any paper towels?" What does the sentence mean?

Context #1: At a supermarket, talking to a clerk. The sentence means, "Please direct me to the cleaning supplies."

Context #2: At a friend's house, after I spill soda on the floor. The sentence means, "Please give me some cleaning supplies."

Context #3: A friend asked me to help pack for a camping trip, and we're going over the list of what's packed. The sentence means, "I suggest packing cleaning supplies."

Context #4: A friend spilled spilled soda on someone else's rug, but is unconcerned. The sentence means, "You should clean that up."

Exercise: Come up with a context where the sentence has its literal meaning, "do you have any paper towels?"

Pragmatics (as above) are a pretty extreme example, but even fairly plain descriptive sentences can be indecipherable without context. E.g., in Japanese 「本を貸してくれた。」 you know that somebody lent someone else a book and that the speaker somehow benefited from this action, or perhaps feels grateful. It probably very clear in context, but you have to practice reading it with the context.

And this is why flash cards suck (not that they're useless...)

exactly. Actually, the Memrise thing would have been better if they had actually put it in context, ie.:

"My 孩子 is very naughty"


As a native Chinese speaker I'm not sure about how a foreigner feels when he's picking up these chinese chars, but I do feel much better to remember new english words after dividing them into their original form, prefixies and sufficies.Now I can guess for the meaning of an english word even if I have never met it before.

My point is, this method can also be applied to studying Chinese. Modern Chinese words are seldom built with a single char(as opposed to the ancient Chineses). A character in Chineses does have some meanings, but not too detailed. I do not have enough linguistic knowledge but it feels like we need "modifiers" plus these "meta meanings" together to build a word.

From the perspective of a programmer let me show you some examples:

子 -- Meta, means "sub", "child", "unit" 节点 -- Noun, means "node" 子节点 -- Child node 系统 -- Noun, means "system" (I believe this word comes from English because the pronounciation is similar to "system" :-D ) 子系统 -- Subsystem 弹 -- Meta, means "Ammo" or "something moving in high speed" 子弹 -- A small, single unit of ammo(just bullet, not bombs, a bomb is too big to be described by "子") 原 -- Meta, means "meta"( :-D ), or "original" 原子 -- A "meta unit" or "original unit" should be an atom shouldn't it?

Hope this will make you better understand how we construct words. :-)

And for grammar, I think the best way to learn grammar is to observe how native speakers speak. Every language have some fixed ways so it will be fast for one to adapt to it.

> extremely steep initial learning curve

Well, not really. Spoken Chinese its relatively easy to start with, because its grammar is much simpler than even English grammar and its phonetics is smaller. Writing using a pinyin IME is not too hard. Reading require a familiarity with characters building blocks. What is very hard is hand writing, reading handwritten script, special fixed forms like the chengyu.

I've always wondered what would happen if you read a big book (like Lord of the Rings) that was translated progressively into another language. So it starts in, say, English and finishes in Spanish. At the start the sentence structure could even be in Spanish form (word ordering-wise) and words are progressively switched to Spanish as the book goes on.

Probably wouldn't work at all but it'd be an interesting experiment.

Check out this Chrome extension that does exactly this, except for when you're surfing sites: https://chrome.google.com/webstore/detail/bedbecnakfcpmkpddj...

You can specify the level of immersion, and it uses Google Translate to swap out a certain percentage of words with the target language you're trying to learn. You can then hover over them to see the translation.

Oh my word. I've just installed the extension. That's absolutely marvellous. Nice implementation. And actually, it's similar to my idea except that instead of the large work being a book it's the entire internet!

Edit: On closer inspection - it doesn't really work that well because it's too much of a scattergun approach. You really want the words to be swapped out progressively. (Probably starting with the most common - articles etc)

I had that idea several years ago. A few years ago I hacked about and produced a story that progressively converted from English to Spanish, starting off with single words, then small phrases, then longer phrases, then full sentences. It used a mish-mash of Google translate and the same work but properly translated.

It worked remarkably well, but I lost it all in a disk crash and never did it again. I wonder how much of it stuck because I did the meta-work, and how well the actual resulting "book" would've worked.

Related: http://news.ycombinator.com/item?id=3921773

It would. Not necessarily with a book like that because fantasy tends to have way, way more difficult worlds than ordinary writings. I am not a native english speaker, but I am working my way through A song of Ice and Fire as an audio book.

I can keep up with ordinary podcasts just fine, I can keep up with television shows for the most part (extremly heavy accents can be a problem) but I really, really have to listen to the words carefully.

I do believe you would be better of with a book that wasn't written as well. Atlas Shrugged would be pretty awesome for this precisely because it has so many repetitions.

Hell, I'm a natural English speaker (and quite literate), and I still have to pay attention when reading Tolkien.

Ha, yeah, I was just trying to think of a long book. I tried reading Harry Potter in Portuguese once upon a time - quickly learnt that fantasy novels are a hard way to learn a language.

The only thing you could choose that is worse would be Shakespeare. I did read one of his stories, but only because I got edition where I could look up almost all the words in a little handy dictionary in the back.

A nice variation would be to practice the words for each chapter until you learn them then read it. You'd highlight the words you've learned. Words start to get repetitive from chapter to chapter so you'll only add the new words. I created a spreadsheet of words for Harry Potter and Around the World in 80 Days in Spanish. I'd like to do something like this in my iOS app.

I'll have to look for some of the classics that are in the public domain. Of course, the translations need to be in the public domain too.

There is a video of Tim Ferris, where he explained that he used only a few sentences, translated to another language to get grasp of it and guide his learning. I imagine if you would know the book by heart, it would not be that useless. I suppose if the book is interactive if you can "peek" to see which words are which then it might just work!

Tim Ferris blogged a guide to those sentences here: http://www.fourhourworkweek.com/blog/2007/11/07/how-to-learn...

I grew up in a Mandarin speaking household failing to learn Chinese when my parents tried forcing me to. I forced myself to learn in college for personal reasons. Oftentimes, this was hard, really hard.

Having said all that, this page was quite a joy to read. I know all those characters, and the page has one big flaw for learning these characters--the pronunciation. Nevertheless, it was a fantastic way to express the meaning of these characters.

If anything, I'd say that it captures the essence of Chinese characters--building up new meaning blocks by compounding basic pictographs together. This is the premise of the Chinese 'radical' system--radicals are the subcomponents of characters like the "child" and "female" characters that comprise "good".

Developer aside: the majority of written Chinese is comprised of about 1000 or so radicals. This may seem like a lot, but having learned enough to recognize them has helped me even in development--I feel like I learn hotkeys, plugins and VIM commands a lot faster than my coworkers because I forced myself to learn how to memorize.

The lack of pronunciation information doesn't seem to me like a flaw; it's just not what they were trying to teach. After all, Chinese uses semantic logographs, not phonetic representations, and lots of different spoken languages use the same Chinese writing system (thus a Shanghainese speaker can exchange letters with a Mandarin speaker despite difficulty actually conversing). Unlike learning to read, write, and speak English, or Russian, or Hindi (e.g.), learning to read Chinese and learning to speak some specific dialect of Chinese are much more distinct tasks, and this is tackling only one of them.

But there are phonetic similarities with characters that share the same right-radical. I can honestly say that many intermediate-level characters were easier for me to learn because I knew 2 things: they followed the pattern of symbolic-radical on the left and pronunciation-radical on the right. Learning the pronunciation can help learning to read characters as well.

You will be confused when there are much more characters to learn. It's not a general rule to learn characters like this. Even Chinese characters might be invented like this 2000+ years ago, the shape of them haved changed for several times. The meaning of most characters cannot be judged by its share.

Reading posts on Chinese forums( ruby-china.org , zhihu.com , douban.com , weibo.com ) I think would be benefit, the phrases and sentences used frequently in daily life are always better for people who are still learning, comparing to novels or something like that which contains many language skills. Ancient Chinese is not a good choice, it's a bit hard for a big part of us to read, Though many articles in acient Chinese was teached in school, it is rarely used during conversations. (Like Haskell, it's great but not widely used.) And I speaks Chinese well, I'm worrying about how to learn English well.

Actually, there are a few books that teach characters through mnemonics like this, going through the first 1-2,000 characters. (The mnemonics don't need to be correct - they simply need to give the right general meaning for that character.) Quite a number of learners have already used these systems to learn characters.

Learners don't rely purely on this to read - absolutely not. They learn reading and grammar separately, usually starting with everyday conversations and words. This is merely the initial base to develop a knowledge of characters quickly, and I definitely think it's helpful (though I studied Chinese in school and haven't needed to use this mnemonic approach).


Learning the first 1000 Chinese characters are the easiest. Moving from 1000 to 3000 characters is really hard. The first 1000 we reinforce seeing street signs when walking along the road, words in the menus, the newspaper headlines, subtitles on TV shows. But beyond that, we don't see the remaining 2000 in contexts often enough, and need to use flashcards and other more artificial methods to learn them.

I wouldn't worry too much about English.

Pretty soon it is going to be so dam ezy 2 get way with it lol.

This article unintentionally highlights one of the difficulties in learning Chinese characters, that understanding them and reading them (in Chinese) are two completely different things. While understanding the meaning (Chinese characters are fun to learn!) isn't too difficult, getting to how it would truly be read is more tricky (the Pinyin for Mandarin being hàn zì hǎo xué).

This is unintentionally an illustration of how LOUSY general (all-languages-included) materials for language-learning are. I am a native speaker of English who acquired Chinese as a second language at adult age during the late 1970s and early 1980s, gaining professional work as second-language teacher (Chinese for English speakers and English for Chinese speakers) and translator and interpreter. I don't like Rosetta Stone materials (as I have seen earlier editions of those) for the same reason that I don't like this Memrise lesson: the lesson is based on frameworks of learning Spanish or French for English speakers, and the lesson doesn't work nearly as well for Chinese, a non-Indo-European language.

The astute criticism already given in another comment has full force--the lesson here doesn't do a thing to teach a reader how to pronounce Chinese. Moreover, the lesson totally muffs up Chinese grammar, because "汉字好学" is not a "condensed form" of an expression that would include a copula verb in Chinese such as "汉字 [form of verb 'to be'] 好学" but rather the sole grammatical way to convey the idea in Chinese. Chinese grammar prefers stative verbs to combinations of copulas and adjectives. The word and character etymologies are also treated abominably poorly in the Memrise story. I never advocate filling one's mind with junk just to have memory hooks for learning new information.

This approach doesn't lay a good foundation for successful learning of Chinese by a native speaker of English. The tried and true textbooks by the late John DeFrancis from Yale University Press and their accompanying audio recordings reflect an older period of standard northern Mandarin, but are much better resources for learning Chinese than Memrise. Especially, DeFrancis's Beginning Chinese Reader is still the royal road for learning to read Chinese, the subject of the article kindly submitted here. DeFrancis made a very careful analysis of reading difficulties second-language learners of Chinese encounter. That is published in condensed form in the front part of Beginning Chinese Reader, and in full form in the classic article "Why Johnny Can't Read Chinese" in the Journal of the Chinese Language Teachers Association.

As Confucius said, 學而時習之 不亦說乎, so there is no substitute for practice in language learning. Language learning is overlearning, and learning languages well takes time.


Two kind replies below raise questions about what I've written above. I was asked about James Heisig. I have perused his books about Japanese (another language I have studied, not as much as Chinese). Doing some looking-up just now to answer the question, I would say that the James Heisig interview


gives, in Heisig's own words, cautions about using his texts as a comprehensive approach for learning literacy in Japanese. For memory aids for Chinese characters, I much prefer Grammata Serica Recensa by Bernhard Karlgren, a less popular but much more accurate reference book.

On the issue of "royal roads" to language learning, I am aware of the work of the Foreign Service language program and of its frequent failings. United States diplomats sometimes attain amazing success in learning languages--I met one once who was the best non-native speaker of Mandarin I have ever met, and who apologized for his Mandarin while saying that Lao is his stronger foreign language--but many United States diplomats are hobbled in their work by poor command of the relevant languages, which plays little role in the selection process for United States diplomats. I'm very respectful of differences among learners and agree with Israel Gelfand that "Students have no shortcomings, they have only peculiarities. The job of a teacher is to turn these peculiarities into advantages." That said, there is an irreducible body of fact in any new subject that each learner has to learn somehow, one hopes with the guidance of a good teacher. The late John DeFrancis was a very good teacher indeed of Chinese, and by validated test, most of the best readers of Chinese as a second language in my generation came into their reading ability with help from his textbook series. The approach taken by Beginning Chinese Reader is certainly better than that taken by the James Heisig popular books on Japanese, if I may say that to tie together the two kind comments.

Well, it's a 3-minute introduction, and it's specifically focused on character mnemonics. Of course it leaves stuff out.

But my real complaint here is the notion of a "royal road", an ideal path that works best for most students. This goes against 50 years of experience at the US Foreign Service Institute. Let me quote them:

Lesson 3. There is no “one right way” to teach (or learn) languages, nor is there a single “right” syllabus. Students at FSI and in other government language training programs have learned and still do learn languages successfully from syllabi based on audio-lingual practice of grammatical patterns, linguistic functions, social situations, task-based learning, community language learning, the silent way, and combinations of these and other approaches.

Let me back this up with some personal experiences. I currently speak French at a level between CEFRL B1 and B2. In other words, I can carry on a social conversation or follow a science documentary on TV, but I can't watch movies or function as an educated adult.

And I didn't even look at French grammar book until recently. Instead, (1) I overlearned hours of French audio and text with help from English translations, and (2) I spoke with native French speakers. I learned a huge fraction of French grammar from context and occasionally reading a footnote in Assimil New French with Ease.

So when I first read a grammar book that claimed, "French tends to avoid the passive voice," I thought, "Well, yeah, because I can always use an impersonal subject with 'on', or a reflexive thingy like 'il se dit'. After all, it's nicer that way."

So if I were going to learn Chinese, I'd begin by working very hard on tones and pronunciation. But after that, I would try to pick up the basic grammar from context. If Chinese strongly favors stative verbs to copulas, I want to figure that out by reading Chinese, not by reading grammar books.

This isn't to say that the "Beginning Chinese Reader" isn't excellent. I'm sure it works wonderfully for a great many people. But there's a half-dozen good ways to learn any language, and Memrise or LingQ or full-time conversational immersion might work better for somebody else.

I was born in Hong Kong and copied Chinese characters until I was in grade 4 when I moved to Australia. My parents recognised the importance of practice in learning Chinese so had me join a weekly 3 hour class to learn Chinese characters until the second year of university.

I am now 22 and I can understand 95% of the individual words in a Chinese newspaper. I don't claim to be able to read the newspaper, however, because the 5% of the words I can't read are the ones that actually matter. Also, for the phrase "學而時習之 不亦說乎", I know what each individual word means but when it is put together like that, it makes no sense to me. (I know ancient phrases are hard to understand in any language, but it's easier for me to point to this phrase in the parent post than searching for a random chinese multiple word-term that I actually don't know on the internet.) I'll have to sit here for 5 minutes to figure it out. My reading speed is also horribly slow. Imagine if you can only read English words at one word per second. Yes, like that.

Here I am, a Chinese born in China (technically an English colony at the time...), more fluent at my second language English than my native tongue. Sad as it may be, when I speak English I have a slight Chinese accent and when I speak Chinese I have a slight English accent so I don't sound native in either.

I hope that can attest to you how much practice learning Chinese takes.

Very well said. I was immediately reminded of those awful "Remembering the Kanji" books when I read this article. You don't "reason" meanings of characters. You drill the meanings into your brain through practice, practice, practice. There are no shortcuts, and no substitutes for rote memorization and drills. Any book or article about language learning with the words "in just X minutes" or "easy" or "in only X days" is selling snake oil. It's just friggin hard.

How do you respond to the folks that say it actually worked for them?

I've only experience with Remembering the Kana, but it worked, and worked very well, for me -- not because I immediately knew them by intrinsic sight, but because of the silly stories with each one would quickly bring them back into memory.

I would first ask them what they mean by "it worked for me". How long did it take? How well can you read?

If you're talking about learning Kana, then my argument would be that some flashcards or computer based drills are going to be a more efficient method of learning, if perhaps less pleasant, than reading a story for each character. And why go through this circuitous thought process of seeing the character, recalling the appropriate story, and then divining the meaning? We don't do that for our letters, and we don't do that for simple arithmetic.

I have also done RtK (book one). It took about 4-6 months over the course of two years (I lost my way a bit in the middle). Anki says I have done 34,000 reviews over 174 hours.

I agree that eventually the stories should be forgotten, but initially, they are significantly better than my normal memory.

In fact I am starting to forget the stories of the kanji that I am exposed to regularly. However, some of the general use characters appear pretty infrequently, and for those the stories are still invaluable. I like Greg's training wheels analogy below.

Another advantage is that the stories provide a check for characters I also know visually. So if I'm writing a character from my visual memory I might be reasonably confident that I'm writing it correctly, but if I can also remember the story then I am sure.

Of course doing RtK doesn't teach you any Japanese, but since I did it I find Japanese vocab acquisition much easier. Maybe it's like learning English when you already know Greek and Latin.

Memory is made of associations. I remember very few flashcards; they're just not at all connected to real life. I remember a lot of silly picture stories, though, like 古 looking like an old grave. Do I have to go through that process every time I look at it to remember that it means old? No, I do not. But the memories stick in ways that simple flashcards do not.

Maybe other people don't or can't learn that way, but these silly picture stories are enough to let me memorize characters I've never seen before so that I can look them up later.

> How long did it take?

About a year (~2000 kanji). That's casual study though, and I restarted once after an idle period.

> How well can you read?

I can recognize and write just about any of the general use characters. But kanji is not vocabulary. Instead, going through RTK has made studying vocabulary tremendously easier. It's given each character a kind of "identity" in my head. I can also guess the meaning of many compounds by the keywords given in the book, but this is just a bonus, real reading comprehension should be attained with vocab study.

I'm not sure if you've actually read the book, but it's no snake oil. It makes no claims to the amount of time or effort you'll need to finish it. It's simply a method to learning kanji that tries to make the most out of your memorization capabilities. It does this by presenting them in a more convenient order (progressively building up the characters), and throwing etymology (to some extent) out of the window in favor of your own personal mnemonic stories.

> I'm not sure if you've actually read the book, but it's no snake oil.

I would not be surprised if the parent hasn't read the book or tried it. Most of the complaints with his method seem to come from people who have never tried it.

It took me three months at about 1-2 hours per day. I still review occasionally and can get about 75% recall. I think a lot of people completely miss the point with Heisig's books.

The point is to just get you familiar with the kanji so they don't look like a foreign language anymore. After that, you learn how to read them in sentences and words.

Also another point is that you don't have to "practice, practice, practice." Some characters I can remember after only writing it a few times because I have the story associated with the character. It sure beats the Asian method of writing out each character hundreds or thousands of times.

The Heisig method is just meant to be the first step.

To finish Heisig with > 95% recall: 2 months. I had a general idea of what I was reading in Japan as soon as I arrived. Passed the JLPT 1 in 1.25 years. (I had a year of university study before Heisig, but the # kanji I knew was likely only in the 100-200 range.)

I never thought I would be able to upvote someone on HN for a reference to Bernhard Karlgren. I spent many fond years with GSR and Wang Li's dictionary in the stacks of UW's Asian Library. Good times, though programming is perhaps better as a job.

> The word and character etymologies are also treated abominably poorly in the Memrise story. I never advocate filling one's mind with junk just to have memory hooks for learning new information.

Etymology is a complex and often convoluted topic, particularly for a language learner who may not be able to understand etymological information in the target language. Do you think there's no value at all in simplified memory aids, such as James Heisig's approach?

I don't care about etymology per se, but if you want to see it that way, Chinese character etymology is a mnemonic built right into the language. 90% of characters them are half pronunciation, half meaning. It is (even etymologically) okay to explain some of the base characters like 子 and 口 and 馬 with a story or picture, but that's where it stops.

Some people argue that the pronunciation part of most characters is seldom 100% accurate, but neither is English spelling, and at least it has some connection to the pronunciation. And like Chinese characters, to really understand spelling you need to know some things about the etymology.

I think the approach in the original link is great for learning Japanese, not Chinese. I use a handful of "story" mnemonics to disambiguate really similar Hanzi, though, when the "built in" mnemonics fail.

Dammit, I stumbled across what looked like a master's or PhD thesis comparing Heisig's approach to others, quantitatively. I didn't read it even though I knew I wanted to. Mental note to find it again.

But if you want to be proficient, you have to be able to read words as a chunk with associated meaning and pronunciation, without the intermediate mnemonic device. So at best it's something you'd want to forget, and at worst it's something that you wouldn't want to learn in the first place.

I'd be totally fascinated to hear about this if you ever find it! I'm greg at memrise dot com.

There's lots of evidence to show that mnemonics boost recollection by a factor or three or so across a wide range of domains, abilities and time ranges. See e.g. http://www.unforgettablelanguages.com/studies.html

Re the intermediate mnemonic device, here's the way I picture things. The mnemonic provides training wheels for your brain, helping you get the answer right a few times. Then, after enough correct responses, mediated by this (hippocampal) mnemonic representation, you rely less and less on the training wheels, and your cortex has had a chance to form a longer-lasting and more direct semantic link.

Disclosure: I'm one of the co-founders of Memrise, so it's not too surprising that I think there's merit in this approach :) Drop me a line or reply here, and I can try and follow up in more detail. Maybe I should write a blog post...

Training wheels is actually a good metaphor, since whether or not one should use training wheels for learning to ride a bike is also fairly hotly debated.

>since whether or not one should use training wheels for learning to ride a bike is also fairly hotly debated.

Really? By who? Because sure as hell training wheels worked wonders for millions of kids worldwide...

> For memory aids for Chinese characters, I much prefer Grammata Serica Recensa by Bernhard Karlgren, a less popular but much more accurate reference book.

I'd never heard about Grammata Serica Recensa until just now (and only looked it up on Wikipedia), but unless I'm misunderstanding its contents/purpose, I'm pretty sure it's an apples-to-oranges comparison. GSR looks like a reference book, which is _not_ what Remembering the Kanji is. RtK is a guide for teaching oneself the writing and meaning (in English) ONLY of the 2200 (in RtK 1) most useful characters (according to Heisig) in written Japanese.

I have a unique perspective on this issue, as I'm someone who has studied Japanese for a few years in college and is now using RtK. The courses I took focused heavily on spoken Japanese, using the Japanese: The Spoken Language series, which is also from Yale University Press and uses a very similar approach to the Chinese series by DeFrancis you mentioned (and also has the deficiency of being out-of-date). Although it gave me a much stronger handle on grammar and pronunciation than other courses would have, their teaching of the written language left much to be desired, to put things lightly.

While the courses did teach the stylistic aspects of written Japanese well, they didn't teach kanji fast enough, and that turned out to be an enormous problem in Japan. So I started using RtK (in concert with the OSS SRS software Anki) about a month ago and have gotten through 600 kanji so far, and I have to say, it is nothing short of magical.

I don't know anything about linguistics or language pedagogy, but I do know that RtK has produced the results I'm looking for. With an investment of approximately 45 minutes/day, I'm learning 20 kanji/day, and remembering them long term. Now, when I see a word I already knew the meaning of, but didn't know the kanji for until now, I find it much easier to remember - and, of course, I can write it by hand perfectly, something I have seen many native Japanese struggle with with my own two eyes far too often. I struggled with it too before starting to use RtK.

> in Heisig's own words, cautions about using his texts as a comprehensive approach for learning literacy in Japanese.

You're absolutely right. Learning the kanji is only the first step in achieving literacy. After that, you need to learn the readings of the kanji, the meanings and readings of the words they form when used in combination, and of course the stylistic aspects that are peculiar to written (vis-à-vis spoken) Japanese, and none of those things is a cake walk by any stretch of the imagination.

However, as Confucius said, 學而不思則罔 思而不學則殆, so if you approach kanji with the rote memorization method used by the vast majority of foreign learners (often because they, like MikeMacMan below, hold the misguided belief that it is a sine qua non of language learning), you will miss out on a lot and unnecessarily waste time and effort.

I'm not disagreeing, but I think one important thing here is to set up the perception of learning, even if the lesson isn't technically right. Few lessons are harder to unlearn than 'I can't learn this.'

(though these inaccuracies should be noted by the end of the lesson.)

Almost reminds me of the Lingua Latina method I used to learn Latin.

You basically read a text that is totally in Latin, with occasional images to guide you.

It starts off very simply, and gradually introduces new vocabulary and grammar. You intuitively learn as you go along.

The books are out of print, but you can still get them on Amazon. http://www.amazon.com/Lingua-Latina-Pars-Familia-Romana/dp/1...

Yes, was going to mention it too. It's incredible and works so well... It's basically a khan academy for Latin and a shame not more people know about it. Definitely the way to go if you want to learn Latin the way a roman child learned it :-)

字学 is the kind of confusion you run into only when you learn simplified Chinese. If I have to interpret 学, it looks more to me like a person sitting in a house on fire; whereas the traditional form 學 is a man holding 爻 in his hands in order to learn his fate, hence "learn". Similarly, 覺 is one who sees (見) his fate. Chinese characters are created with all sorts of clues for you to pick up. It is unfortunate that some of these clues got taken away with the simplification of the language. 學 is not only more distinguishable than 学, but takes the same amount of effort to type.

Well put; I share the same sentiment. I feel that one of the strongest arguments for traditional characters is that it retains the original meanings behind the characters themselves. Simplifying the characters just throws information away. It's a lossy compression!

The amount of rage in these HN comments makes me think MemRise must really be onto something.

I'm from China, I think the method for familiar with Chinese characters is creative and useful. But if you want to learn the real Chinese, the most effective method is using it. I often watch the video on Youtube and make tweets on Twitter. And I'm benefit from it for my English. So you can communicate with Chinese people on Youku(like Youtube) and Weibo(like Twitter) directly, though those websites' designs are not so friendly. We're pleasure to communicate with you.

A while back I tried signing up for some Chinese BBSes to practice my written mandarin (I believe I tried mop.com and 163) and I found it utterly impossible to figure out what to do. I was expecting something similar to reddit, HN or traditional phpBB sites and it was nothing like that at all. Do you know some sites that might be easier to figure out?

I would suggest http://www.douban.com/ Much cleaner user interface (just forget about mop.com or 163.com, those are ancient sites...) Better content, people discuss about books, movies, music and much more(you can try "groups", there're so many with a wide range of topics) And www.guokr.cn -- Geek topics goes here. :-D

Yeah, I also suggest douban,Very clean design website.

Sorry, as far as I know, we don't have web service like reddit or HN. Like @v-yadli suggests,douban is good site, most of its user is student and young artist(文艺青年).If you want to write long articles, you can try www.diandian.com, a light weight blogger like tumblr.

Chinese forums live in the bad old days of the early 2000s. If you're afraid of the plunge, you can come post in the Chinese-language-only section of Chinese-Forums.com; we're a friendly bunch.

With all the respect in the world for the great nation of China, I keep wondering why people would learn Mandarin. I mean, isn't it time we all go for a common alphabet and language? I'm French, I quite fancy my language, the culture around it and its sophistication, but it's about time we all focus on English. Let's make it as ubiquitous as possible, please. It's just so simple, flexible and expressive...

The way the Hanzi are introduced in the newspaper article could be the way to go, i.e. embedded within English text and grammar. E.g, your post could be rewritten:

"With all the respect in the 世界 for the great nation of 中国, I keep wondering why 人们 would learn 汉语. I mean, isn't it 时间 we all go for a common alphabet and 语言? I'm 法国人, I quite fancy my language, the culture around it and its sophistication, but it's about time we all focus on 英语. Let's 让 it as ubiquitous as possible, 请问. It's just so 简单, flexible and expressive."

Chinese isn't going away considering how many speakers it has and how influental China is.

Even if it would, it would be sad because losing languages means merging different cultures into one big blob of metaculture, losing a lot in process.

And not that English is so good as the language. It is simple, yes, but the phonetics are tricky and inconsistent and the dictionary is long (without providing as much utility) and fragmented. Also, being very simple (and limited) is fine for pigdin conversations - least common denominator - but not necessarily a good thing to force on the entire humanity. What if there is actual value lost in leaving aside languages with more complex grammar?

If we're after a common language, why not Esperanto? It's simpler than English and certainly simpler than Mandarin.

Esperanto has a huge bias towards latin-based languages. It's FAR from being universal.

I have been using their service for Spanish for almost a year... Really useful for expanding your wordlist.

Memrise is pretty cool, I've been using it to learn the lojban gismu and it seems quite effective so far.

Started losing me at: 学 字 - creative teaching & boring teaching, I don't understand how those two stack and what they mean in unison. The last paragraph lost me completely.

I was disappointed with that section as well. Thought it could have been written clearer to differentiate the two.

学 - study

字 - character

So the author actually meant boring teaching of 'characters' with the section on 字 but I don't think it came across well.

In my point of view, as a native Traditional Chinese user (I'm from Taiwan)


学 is actually the simplified version of the character 學. And the character 子 in both of these characters doesn't really have much connections

The character 子 (literally means children) in 學 stands for "student", but in 字 it is just a rebus (phonetic loan) character.

学 means "learn"

字 means "characters"

>字 means "characters" Not exactly! Single 字 has meanings in Chinese, but single character generally has no meaning in English.

That was an interesting idea. To bad it I nearly couldn't finish it because of all the bigotry.

Lighten up. You shouldn't be getting offended by something as minor as what's in that article.

If they had been making the same claims against women would you also dismiss it?

I am Polish, but I am convinced that we as a human kind should just pick one language and stick with it. All these languages are there to prevent us from freely exchanging the ideas. With the age of the Internet all of us should just start using English.

This didn't work for me at all. I think telling a story like that is good, but you gotta tell us what the characters are/mean at some point. You can't just start using them an expect us to get it. At this point I think the character for children is one, the character for characters is children with a roof and the character for learning is an excited child with a roof. I don't know if this is correct or not because the sentence "thus X means Y" doesn't appear before X started getting used... when you have 3 uncertainties in my head, then I stop reading because its hard to know what you're talking about because there's too much you haven't answered, and I don't expect you'll get more clear as I go further.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact