Chinese literacy for foreigners: are flashcards the miracle cure?

A former MBA student of mine was asking about my experience learning Chinese characters recently. He grew up overseas. His mother speaks Mandarin with him, but he doesn’t have any formal schooling in the language. I started to write him a long email reply, and realized as I wrote it up that it might be a good Sinoglot conversation. Our collective advice has to be better than mine alone.

So I’m going to start with my story. From 2001-2007, I developed basic conversational ability without knowing much about characters, mostly learning from my mother-in-law and pinyin-based self-study. I learned pinyin pretty quick like everyone else, of course, and in the early years I tried to write a few characters 500 times each, or whatever someone told me was necessary to remember them. But then I just kind of ignored characters — until 2007ish.

I think now that I could have learned characters much faster if I had followed the approach I use now. If I was doing it all over again, I’d also start with characters sooner than I did. Six years was too long to be illiterate and starting to work on literacy at, say, 2 years would have dramatically speeded up my vocabulary acquisition.

Vocabulary acquisition — and, to some extent, acquisition of characters — is still something I work on today. What I’m going to describe below is how I do it and why. It’s a lot of gritty detail and I don’t recommend it unless you really care about this stuff. But if you do and you’ve got thoughts, I’ll be happy to get even nittier and grittier.

Here’s the advice I wish I could have given myself in about 2003. It’s the approach I later developed and still use now.

  1. Use flashcards. Specifically, use electronic flashcards with spaced repetition. Anki is good. The software is free and there’s a free app for Android (called Ankidroid) [well, ok, there was no Android in 2003. whatever]
  2. Make your own flashcards. Never use someone else’s cards.
  3. Have two separate decks of flashcards: one for words, one for individual characters

My rules for word flashcards 

  1. Use the cloze method, where the front of the card is a sentence that’s missing a word and the back of the card is the word it’s missing. Corollary: Don’t learn words in isolation and especially not by translation.
  2. Use sentences that you find anywhere, as long as the sentences are by native Chinese speakers: comic books are a good place to start, since they are all dialogue and often use pretty simple everyday spoken language. Novels are good too. I use sentences that people email or text me when I don’t understand a word. You can also use sentences that people say, if you can get them written down.
  3. Be choosy. Only put a word on a flashcared if you could see yourself wanting to use in the near future. For example, don’t bother learning 面首 because you saw it in a 王小波 novel and found it amusing
  4. But don’t be exclusive. I don’t worry about putting the same word on more than one flashcard, because it helps get a more rounded sense of the word’s meaning. For example I swear I have half a dozen cards with 进行. That’s okay with me because it’s a word that translates into different English words in different contexts. So every time I find it in a context where I wouldn’t think to use it myself, I just create a new card.

Here’s an example from my flashcards. Someone once sent me a text that had this sentence in it:

有个问题我得请教你
Yǒu ge wèntí wǒ děi qǐngjiào nǐ

At the time, because I knew the person and the context, I understood that this was a polite way of asking for my help. Something like, “I have a question I’d like your help with”. But I didn’t know the word 请教. So I put this sentence in my flashcards.

FRONT: 有个问题我得________你
BACK: 请教

The great thing about this method for me has been

  1. I learn the context in which the word was used — how it’s used in real life. Often I remember much more than just the sentence. I remember the tone in which it was used, the setting — everything. And all that context gets tied to this usage of this word, just like it does for native speakers. At least that’s how it seems to me. It seems much easier to remember to use the word, then, in an appropriate context in the future, because of all those connections in the old neural network.
  2. I do a lot of reading practice. You might think reading the same sentences sentences over and over wouldn’t be much use. It sounds crazy, but it has actually made my reading much faster.
  3. As I sort of said in (1), I very quickly find that I can use the word myself, often in just the right situation.

In the beginning, if you don’t know many characters, you could even use pinyin

FRONT: Yǒu ge wèntí wǒ děi ________ nǐ
BACK: qǐngjiào

But you can quickly switch to partial characters, maybe like this

FRONT: 有个wèntí我děi________你
BACK: 请教 (qǐngjiào)

And I recommend using all characters as soon as you are able, because it will really improve your reading.

My rules for single character flashcards

  1. Don’t bother learning a character until you know a word that uses it. In other words, don’t learn characters just because they’re on the “top 3000” of some frequency list. Instead, just learn the characters that appear in words in your word deck
  2. Don’t consider that you “know” a character until you can look at it and tell both its pronunciation AND a word that it’s used in. For 多音字 (characters with more than one pronunciation), you don’t “know” the character until you know all its pronunciations and a word to go with those pronunciations. For example when you see the card for 行, you should say to yourself xíng as in 行人 and háng as in 银行. However, per rule 1, don’t bother learning multiple pronunciations of a character if you don’t know the words that go with the alternative pronunciations. For example, for a long time my 没 card had just one pronunciation, méi as in 没有. It wasn’t until I got my brain around 没收 that I added mò as a pronunciation.

I still add characters to my character deck occasionally, but not often. After 3000, the returns diminish rapidly. If I come across a new character in a word I feel like I should know, then I’ll put the word into my word deck, and the character into the character deck. But a lot of times the characters I don’t recognize are in place names or people’s names. With those, either I see them a lot and memorize the pronunciation without needing the formality of a flashcard, or I only see them once and guess at the pronunciation and never come across them again, in which case memorizing the character seems kind of pointless.

Roughly, that’s it. My path to literacy has been paved with Anki, and I’m grateful for it. At the same time I’ve wasted a lot of effort trying to find the right way to use Anki, because it can certainly become a time-waster too. I hope this story plus the comments will save some time for future learners.

PS: I rarely handwrite anything, for the record. I’ve learned how to write characters, stroke order and all that. But I haven’t written most of the characters I can recognize. And I never practice handwriting, so from memory, I have trouble writing even the simplest characters. All of my communications are 输入法 assisted, i.e. using pinyin input.

PPS: Really, really gritty detail, for those in Ankidroid. My word deck actually contains three fields. Field 1 is the cloze sentence. Field 2 is the answer (in hanzi). In Field 3 I will sometimes put the pronunciation or even (gasp) an English translation. However, the way Ankidroid presents the cards, it’s just Field 1 as front and Field 2 as the back. I only see Field 3 if I edit the card. This way, most of the time, I’m doing my flashcards in a pure native hanzi & mandarin environment: no English or pinyin. But if I worry that I’ve forgotten a pronunciation or meaning, I can edit the card and look at field 3 as a crutch.

35 responses to “Chinese literacy for foreigners: are flashcards the miracle cure?”

  1. Aaron says:

    I started studying Japanese in middle school in the US, and achieved literacy (~2,000 kanji) largely thanks to self-study with flashcards. Of course back then we only had paper flashcards, which we studied while walking up-hill both ways in the snow.

    I agree wholeheartedly that you *must* make your own flashcards and not use pre-made ones. Once I internalized the kanji structure and had a few hundred characters under my belt, I would often be able to learn a new character (at least its meaning, if not its myriad pronunciations, compounds, etc.) simply by making a flashcard for it.

    • Steve (Syz) says:

      Ah, paper flashcards. In general I’m anti-progress and thus a fan of old school methods. But in the case of flashcards, the modern convenience of storing them on my phone and having the computer worry about when to show them to me has made me suppress my inner Luddite.

      But yes, sometimes simply the act of making the card is enough to practically have a word memorized.

  2. Sam says:

    Nice detailed article! I especially like your Anki format – I might have to try making some new decks with the fields you use.
    On only learning a single character when you know a word that uses it, I guess it depends on whether you already have a large (speaking and/or recognition) vocabulary or not, but it seems to me that learning single characters, even obscure ones, first, can help you guess at the meanings of words during reading, and I even find that knowing a bunch of the surnames that round out the top 3000 odd most frequent characters helps a lot when reading as for me I’ve always stumbled on surnames and place names when they are not immediately clear from context.
    For me personally, I think I learn more words from learning single characters first, than vice versa. Then again, I’m also interested in the character-component and etymology side of things, and I want to be able to write every character I know, so my goals are a bit different.

    • Steve (Syz) says:

      I’m more of the opinion that characters get their meaning from words than vice versa, but maybe Sima will show up to this part of the conversation. I’m pretty sure he’s got a different opinion.

  3. Katie says:

    I think the bit about writing characters is a bit more nuanced. I too never write in Chinese by hand and can remember embarrassingly few of the characters I can read. (And it is embarrassing because I have a ridiculously complicated Chinese name that I can never remember how to write, and so can’t ever fill out a form in Chinese.) But I think writing it has its place. First of all, not relevant to your friend, I assume, but in case others are looking for tips here: if you ever want to take Chinese classes at a university, so far as I know, you’ll need to be able to write characters by hand. Second: there are a lot of characters that look a lot alike. The less you know about things like radicals, the more true that is. If you write them out by hand, you’ll be better able to make sense of what those differences are. So the bit about learning proper stroke order and such is, I think, more important in learning to read than it seems. Also, in real life, on rare occasions, I do actually have to write something down, and if I can at least copy a character correctly, it does prove useful.

    Also: I agree that 1-2 years in seems to be about the right time to start learning characters. I was first exposed to Chinese through a college class, and despite the fact that I spent more time studying for that class than most of the classes in my major, when all was said and done, we probably learned 10-20 words a week. Take out a week or two at the beginning for pronunciation drills and that leaves you with a ~200 word vocabulary for a semester’s worth of work. I think, especially if you don’t have the luxury of waiting two years to be able to ask where a bathroom is, that time would be much better spent up front learning oral language. By the time you have enough vocab in that you could, in theory, make sense out of actual Chinese text, learning characters will start to seem much more useful.

    Also: If people do decide to learn to write characters, I highly, highly recommend avoiding the “write 500 times” method. Although I never had time to finish it, I started with the “Remembering Simplified Hanzi” book and found it incredibly useful. No, it’s not etymologically correct, and yes, some of their key words are a little weird, and yes, some of it’s pretty silly, but I could learn and retain characters at a much, much faster rate that way, and I found that I was much better at recognizing characters I had learned to write in this way. (As with your advice, I think the book is much more useful when you actually know words with those characters in them.)

    Also: any recommendations for novels that are easy reading? I read my 4-year-old’s books sometimes, but something a bit more interesting might be nice :)

    • Katie says:

      I forgot to mention, another project I started and never finished in my attempts at reading was trying to find a good way of making a list of all the words I knew that had a given character in them. It seemed like it had good potential for making associations between words, but I never got far enough to know for sure.

    • Steve (Syz) says:

      Writing Characters
      Well, yes, you gotta be able to write your name 😉

      I’m sometimes in Chinese-language meetings where I need to take notes. I have to admit that sometimes I wish the characters would just flow out of my pen. They don’t, so I use a mix of pinyin and English. It works well enough to jog my memory when I get back home. I’m guessing the same would suffice for a uni class.

      I also agree it’s important to learn about stroke order and such.

      But on the issue of handwriting helping distinguish similar characters, I think my jury’s still out. Like all learners, I often mistake one character for another, especially in isolation. My solution in the flashcards is to make a note on the back of the card, so then I can see the similar characters side-by-side. On the back of my 衣 card, then, there is a note that says, tersely, “not 农” On the back of 秦 it says “not 奏”, etc. I’ve found that this approach forces me to focus on the component of the character that’s causing problems, and I still think it saves time over learning how to handwrite the whole darned thing.

      When to start learning chars
      Couldn’t agree more that Chinese 101 is NOT the right time. Your experience I think perfectly captures why so much foreign Chinese education is fruitless.

      Easy novels?
      This should be a separate post. Maybe I’ll start one sometime. The first novel that I finished was 活着 by 余华. It is short, engrossing, and has a very simple style. Not lighthearted, however. I also then read 余华’s 兄弟. It’s much longer but also has a very simple style. It is, however, somewhat NSFS (not safe for subway) in the sense that the person reading over your shoulder might be unable to suppress expressions of disgust at certain scenes…

      • Katie says:

        Depends on the uni class, right? If you have to take written exams and can’t write, you’re in trouble. I assume most Chinese classes for foreigners at Chinese universities still require writing knowledge on par with your speaking/listening. I know the one at our local university does.

  4. Sima says:

    I started very early with flashcards for individual characters, largely because I found talking so daunting, so I guess most of my experience is irrelevant here. Just a few points to try to contradict Steve, as a matter of principle.

    I made cards by hand. I bought large sheets of coloured card and cut them into squares, as required. I didn’t use the coloured card for any kind of categorisation.

    I wrote a single character on the face of each card using a big fat marker pen. At the beginning this made for very ugly characters, so I practised writing the character on squared paper, with an ink-gel pen – twenty times for each character, or more if I was struggling with the form – before attempting to write with the marker on the card.

    On the reverse of the card, I wrote the pinyin for the character, plus the pinyin for any words I knew which used that character, and English translations. I added further words when appropriate.

    I would not attempt to write a new character until I’d looked it up in a paper dictionary, using the radical index.

    By the time I’d made a card, I was already well on the way to having memorised the character.

    I typically added a few characters per day, being quite strict about not getting ahead of myself. I never added more than twenty, and often it was just three or four. At times I had to be quite brutal about which characters I added to my decks. I only added characters that I thought I ought to know. Even so, some of those seem amazingly complicated, given my overall language level at the time. For some reason, I thought that knowing 橘 was important, because ‘even children know the names of fruit’. Still, I did manage to leave out a fair number characters too. And I do think it’s important to do this. One has to learn to accept that forgetting is ok. If something’s important, you’ll bump into it again and, when you do, it’s all the more likely you’ll attach significance to it, and to its context. I think of this as a kind of rational acquisitiveness.

    I would test myself by taking a card from the top of the deck and either:

    1) reading the pinyin and writing the character (on squared paper)

    2) reading the character aloud and checking the pronunciation

    I would also usually say a word or phrase which used the character.

    I was very strict on mistakes. If I made a mistake in writing a character, it would go on a separate pile and I would write each of these mistaken characters twenty times before retesting. Mistakes in pronunciation would be similarly treated.

    I had a very strong preference for option 1) above (writing) and, to some extent, indulged it. I believed (and still do) that learning to produce all the strokes in the correct direction and order, and each character in a balanced fashion, was important. I bought a decent fountain pen, and I enjoyed having to refill it, and even clean/flush it out from time to time.

    Fifty cards went were bundled in a numbered paper slip. Once I had about 150 cards, I made a tray in which I could store them. I ended up with two trays, each containing fifteen packs of 50 cards; 750 cards per tray, 1500 cards total. I still have them, and they get dragged out whenever I’m asked to give a talk on learning Chinese. They’ve become treasured possessions.

    Let’s be clear about a couple things:

    A) Few people (now, I think) would recommend learning to write at the very beginning of one’s Chinese studies.

    B) Learning to recognise (and write) Chinese characters does not constitute reading any more than learning to decipher letter combinations constitutes reading in English.

    C) 1500 characters is nowhere near enough for a non-native to attempt to read meaningfully.

    BUT:

    D) I never looked on flashcard testing as study. This is what I did when I didn’t want, or didn’t have time, to study. And it was an absolute pleasure. Better than crosswords; better than sudoku; a hell of a lot better than Angry Birds.

    E) 50 cards fitted easily into a spare five or ten minutes (and I found several such times through the day)

    F) I kept the cards, pen and ink, paper, in a given place (in the summer months, the balcony) and I would go to that place to use them. I never tried to carry them anywhere, unless I was going away for an extended period.

    G) Those 1500 characters, learnt over the course of about 18 months – 2 years, set me in great stead.

    H) I think that the process itself, and the physical nature of the cards, was actually important for me. Making the cards and using the paper dictionary were part of the learning process. Counting, stacking, shuffling and turning the cards were rituals. Even dropping them on the floor, smudging the ink, or accidentally creasing a card, were part of the process. Seeing the cards accumulate in trays, giving a visual and physical representation of my progress, was significant to me.

    I) It’s all highly personal. You develop your own rules. You learn what you want to learn. You make things yourself. You develop your own little habits and routines. You instill meaning in the process. And for me, the physical dimension was important.

    For someone who already speaks the language (and has much of the cultural background), I wonder whether one might approach things slightly differently:

    Give much greater priority to simple characters (characters with fewer strokes).

    Give greater priority to characters which, though rare in their own right, occur frequently as either radicals or phonetic components.

    I suspect this would slow the process at the start (against simply seeking to memorise the most frequently occurring characters first), but may build a more solid foundation and result in accelerated learning later on (remembering that the native-speaking learner has a head start in terms of anticipation of words he’s likely to encounter in any given situation).

    It’s frequently stated that learning Chinese is hard. And that this is in large part due to the difficulty in learning to read and write Chinese characters. This is both kind of true and highly misleading.

    Learning how to recognise, and to passably write, Chinese characters is EASY.

    The process requires time – rather longer than learning abc – though people are prone to overstate this.

    It requires a certain amount of commitment and discipline. But so do most things worthy of study.

    But it really doesn’t require any special skill or talent*.

    ( *Disclaimer: Learning to read ink-on-paper scripts can be difficult for those with major visual impairment. Learning difficulties associated with reading may occur in different forms with different scripts. Skill and talent will usually be evident to all, once you’ve mastered it.)

  5. Aaron Posehn says:

    Nice article. You make a good point about making your own flashcards in general, and I specifically liked your idea about seeing them in context from sentences you’ve previously come across.

    As well, I’ve always found that beginners often find it useful to become familiar with the “smaller pieces” of characters, and not only just the radicals, but the other “random” parts of characters too (often the part with pronunciation clues). For example, if you see the characters 疫, 役, and 没, although they’re totally different in meaning, and the last one differs in pronunciation, you can still see the similarity on the right-hand side, and use that knowledge to “recognize” characters you might come across in the future with a similar structure. At least the new characters won’t totally look like little scribbles, and you can start to guess at their meanings and pronouncations more quickly.

  6. Steve (Syz) says:

    That was poignant, Sima.

    I’m reminded of my very first character-learning foray, which I’d managed to block from conscious memory. It was early 2007, and my approach was the antithesis of what you’ve described above: brute force, brute functionality, no love. Specifically, I had read enough descriptions of written Chinese to know that 3000 characters were enough for basic literacy. I considered my speaking ability already reasonably up to par (laughable, in retrospect) and so considered that memorizing 3000 characters was going to be just like learning a complicated alphabet that I would be applying to my pre-existing vocabulary. How hard could it be, really? I’d put my mind to it, I’d put the grind to it, and at 300 characters a week, (only 50 a day, including a sabbath somewhere!) I’d be literate in 10 weeks. Everything was on the computer. Not even fancy flashcards, just a simple excel spreadsheet with the character in one column and pronunciation and “English meaning” in another.

    When after the first week I was forced to admit that it would be a 20-week endeavor rather than 10, I should have had an inkling that all was not well in efficiency-ville. Yet I was confident. After all, I’d actually done 157 that week, not just 150, and with some unexpected business encroaching on my study time. The second week I even managed, barely, to make it into the low 300s, presumably enough to keep me on the 20-week pace.

    Honestly I can’t remember when the whole thing blew up. 400, 500, 600 characters? At some point I hit the wall. Hard. Bounced off it, maybe. I think there was a week when my test showed that the number of characters I “knew” had actually decreased from the previous week. I suddenly found myself confronting characters I thought I knew — I swore I knew! — that claimed to have completely different pronunciations and meanings than what I had memorized. Of course you’ve probably already guessed why. Essentially, I was memorizing the general shape of characters, and almost nothing about the detail. That works fine when you’re recognizing shapes as diverse as 大 and 那, for example, but it doesn’t work for much more. I was blindsided by the 农衣,秦奏 kinds of examples I gave in response to Katie. There were no two ways about it. I was going to have to spend a LOT of time looking at the details and thinking much more carefully about what to pay attention to.

    In a way, this is a response to Aaron Posehn too. I agree that beginners “find it useful” to learn the smaller pieces of characters. In fact, I’d argue that it’s mandatory. There’s no other way, unless your character-memorizing goals are limited to a few hundred.

    Anyway, as you can imagine, eventually I abandoned the speed-memorization approach entirely. The system that I’ve come up with, in my own way, is also based on love, which sounds like hyperbole but is something I actually thought as I settled into the process. Unlike some, like you, Sima, my love doesn’t extend to the act of writing the character itself. Maybe it will someday, I don’t rule that out. But it does extend to the components of the characters that distinguish them from their close relatives, and to the words that give those characters pronunciations and life. As I retooled myself for long-term character acquisition, I quit fretting about how long it was taking me to internalize each character and tried simply to enjoy the process. Every time I’d confuse one character for another, I’d take it as an opportunity to slow down and focus on the differences, like a scientist examining a new specimen of beetle when he already knows 2500 of them.

    I’ll wrap up the story before it stretches its metaphors any further. I’ll be anxious to see your cards the next time I’m in your neck of the woods.

  7. John Pasden says:

    Reading the various replies, you might be tempted to conclude that different personalities learn best in different ways… :)

    I’d just like to add one little nitpick about one of the statements in the original article: “Don’t consider that you “know” a character until you can look at it and tell both its pronunciation AND a word that it’s used in. For 多音字 (characters with more than one pronunciation), you don’t “know” the character until you know all its pronunciations and a word to go with those pronunciations.”

    This is kind of a slippery slope. For example, your dictionary may tell you that 吃 has the readings “chī” and “qī,” and 喝 has the readings “hē” and “hè.” And yet most of us learners are perfectly happy knowing only one basic reading for each of those characters.

    So I’m guessing what you actually meant was that “For 多音字 (characters with more than one pronunciation), you don’t “know” the character until you know all its common pronunciations and a word to go with those pronunciations.”

    The problem then becomes that as a learner, you really don’t know which readings for a character are the common ones. To which I say it’s best to take the same approach as you did with words. When you encounter a new reading for a character you already knew, just add it to your old card with a new word for context. Otherwise you run the risk of burdening your memory (or your SRS reviews) with an obscure pronunciation you may never, ever encounter.

    • Steve (Syz) says:

      Hey John, I agree with you and it might not have been clear, but that’s what I was trying to say with

      However, per rule 1, don’t bother learning multiple pronunciations of a character if you don’t know the words that go with the alternative pronunciations. For example, for a long time my 没 card had just one pronunciation, méi as in 没有. It wasn’t until I got my brain around 没收 that I added mò as a pronunciation.

      Basically, my “don’t learn a character till after you know a useful word to attach to it” approach should prevent you from making that kind of mistake: learning an obscure pronunciation. Believe me, I’ve made that mistake and pretty much every other one in the book.

      • John Pasden says:

        WTF? Was that there originally?? How did I miss that?

        Apologies… (And yes, I think we’ve both made that mistake….)

        • Steve (Syz) says:

          lol. hey, i’d be the last to throw stones at someone for having skimmed over a sentence.

          Yeah the whole issue of obscure pronunciations is interesting. Early on I memorized a pronunciation of some character — damned if i can remember which one right now — listed in a seemingly reliable source, only to have some well-educated native speaker tell me she’d never heard of this pronunciation.

  8. Ben Ross says:

    Great points on learning with flashcards. And I totally agree about characters. I learned how to write. It was fun. But other than my address, I rarely if ever handwrite anything in English even, let alone Chinese.

    One thing I will add from my experience however, is that I found QQ to be immensely helpful for improving my Chinese literacy. There is an infinite number of Chinese people online at any given moment who will gladly chat with a stranger, especially a laowai. And having multiple QQ conversations going on at once is a phenomenal brain exercise for Chinese literacy. It also greatly speeds up your typing ability.

    • Steve (Syz) says:

      Somehow I missed your QQ note before, Ben. I’ve heard this from a couple other folks and I’m guessing it’s gotta be good. Personally I’ve found over the years that one of the hardest things to do is just get myself to read more. It’s a vicious circle: your reading is slow, so you don’t enjoy it, so you don’t do it, so it doesn’t get any faster… Any kind of chat is going to force you to read, and to process stuff quickly, as you say.

  9. aelephant says:

    Sentences sound good, but I found that they didn’t work well for me. I found that my cards were often too ambiguous (there might be 3, 4, or even more correct answers to one cloze deletion). They are also much more time consuming in the way that I was doing them (read the card aloud with correct pronunciation, fill in the blank with correct pronunciation & word choice). I drove myself crazy trying to have the tone correct for every word in the sentence. I find that my reviews are much faster & more enjoyable if I make them as simple as possible — just 1 piece of information required on each card. So if I’m asking for pronunciation, I’m not also asking for meaning. I’m sure this method will also evolve in the future. For now, it has dramatically reduced my stress & frustration levels vs. a more “comprehensive” method.

    • Steve (Syz) says:

      Well, maybe Pasden’s “different personalities learn best in different ways” was right on the money 😀

      Three points:
      1. It’s quite possible my approach works only at a certain stage of my learning. I’m sure it wouldn’t have worked when I knew fewer characters and words. But then I would have used more pinyin…
      2. You’re right of course about ambiguity. The tack I take is that my answer is Wrong if I choose the wrong word, period. It doesn’t matter if the word I thought of would have worked in that blank — I insist on remembering the word that the native speaker used in the original. Sometimes this is a little frustrating, and I do delete cards that are too ambiguous and don’t provide enough context, but for the most part it seems to work.
      3. I agree with you about testing only one thing. I saw this advice somewhere for flashcards before. It’s served me well. I’d say if you’re worrying about pronunciation of other words in the cloze sentence, then maybe that’s a good argument for putting those in pinyin so you can just focus on the word in the blank.

      Ultimately, yeah, it’s a matter of constant tinkering to find something that works for you. I guess my reason for writing this post was more to have a forum for talking about the tinkering than it was to say that any one approach was right.

      • aelephant says:

        Thanks for your advice. I think I may add *some* sentences back, but include pinyin for “borderline” characters (I know the meaning & pinyin, but sometimes get the tones mixed up). Sometimes it is actually *easier* to remember the word in context than it is to be presented with 2 characters floating in space.

        Another key part I was missing was deleting cards on a regular basis. Sometimes I have to remind myself I’m learning Chinese because I *like* Chinese — if certain parts are frustrating me now, I can always skip them & come back to them later.

  10. Kat says:

    I’ve been using Memrise flashcards. They’ve really helped me remember tones and character components.

    • Sima says:

      I have a great deal of time for Memrise, largely because I’ve read a little about Ed Cooke, one of the founders.
      I was a little concerned by some of the mnemonic offerings when they first rolled out Memrise, because it looked likely to confuse more than help, but it’s interesting you say it’s helped you with both tones and components. How have you found it helpful for tones?

      • Kat says:

        Repetition, mostly. In general, it helped develop my ear for how each tone rises or falls, and it’s been especially helpful with two-syllable words. The repetition also helps tighten mental connections among character, meaning, and sound. I also like how there are at least three different voices for each word; it’s helped me recognize words and tones despite subtle differences in pronunciation (e.g. “ying” / “ing” / “yeung”).

        Mnemonics are user-generated, but the Mandarin courses have several longtime users who produce really useful mnemonics. And if you can’t find one you like, you can always create your own or ignore them altogether.

        • Kat says:

          Oh, one drawback to Memrise is that it teaches words in isolation. I’ve only been using it as a supplement to the formal language classes I take; if I was using Memrise alone, I might be really confused by some of the definitions. 吗, for instance, is only defined as “question mark”; only a speaker/reader would know how it’s actually used. So, I’d only recommend Memrise as a supplement, not as a starting point.

          • Sima says:

            Thanks for that, Kat.
            I think that’s a good general rule – if all your study is based on the same activity, you’re going to miss out quite badly.

  11. TWteacher says:

    Chinese-Characters.org
    the single greatest, free character studying tool online. As well as putting in text to mdbg and reading word by word with its fantastic interface. Reading is the best way to improve your characters in any language. Read more and you will get better. There are many complex characters that I know how to write not because I’ve practiced them or used flashcards, but because I’ve read them enough times. If you pay attention to the parts, reading can only improve your character memory.

  12. Peejayrocks says:

    Brilliant! Brilliant! Brilliant! Ideas I need to incorporate to help my students. We haven’t done flash cards yet–largely because I hadn’t been able to think of effective ways to make them–ahem, effective. Thanks for your ideas. Also, couple thoughts: 1. I ONLY teach pinyin the first two terms. Otherwise, we’d get so pigged down that students lose interest. Last two terms are when I introduce characters. 2. Some research has shown that physically writing out characters assists in faster RECOGNITION acquisition than just learning them via computer, etc. I’ve found this to be anecdotally true with both students and myself. Just some FYI. Take care all. Great blog!!!

    • Sima says:

      Peejayrocks,
      Could you point me to the research about writing and recognition you mentioned? I’d be very interested to see it.
      Who do you teach and to what level?

  13. Tanya says:

    I really love the idea of doing flashcards with sentences. If I could go back I would probably do that. I did something similar when learning a new set of vocab attached to a new job a few years back, but for the most part I was still fixated on learning characters in isolation. When practising spoken it was all about the whole sentences, fixing the flow of the intonation in one smooth piece, but I never thought to use that same approach when studying written as well. These days I almost never handwrite – it’s all texting and typing – and it’s a bit sad how bad my recall is without an input device at hand!

  14. Bruce Humes says:

    Enjoyed this conversation immensely. I should note that I am a native speaker of English, have something like a university-level reading knowledge of modern Chinese and have served as director of several Chinese magazines and web sites for businesspeople and IT professionals. A few things occurred to me as I read everyone’s entry.

    1. Flashcards
    — Absolutely essential for me; and yes, you gotta make ’em yourself. It never occurred to me to leave blanks and use the other side of the card to provide the “answer.” On one side I wrote anything from a two-character word to a short phrase or sentence in character form with pinyin too, and then put the translation on the other side.

    2. Testing yourself
    — I usually looked at the English side first, and then guessed at the correct chinese character + pinyin. I “punished” myself by putting all the cards I didn’t get right into a pile, and then went through them again and again until I could write the Chinese correctly.

    3. What about reading?
    — What struck me most about most of the notes above is how there is hardly any mention of reading natural text in order to better master Chinese reading and writing. I used all of these methods over a few days, but never focused too much on just one: a) Made flashcards and tested myself using them; b) Read “natural” Chinese (i.e., not a textbook) regularly, using a dictionary wherever necessary, but usually after reading a large chunk of text. Never allowed dictionary checking to interfere with reading a meaningful section of the text and guessing what it means as a whole; c) Once I had done 1-2, “rewarded” myself by reading short items (news worked for me) without checking any of the words in a dictionary or making flashcards for them.

    I think the secret to learning how to read Chinese is to allow yourself to read more than you can write and not feel “guilty” about that . . . writing/copying new vocabulary at the outset . . . achieving a happy balance between reading for vocabulary acquisition (i.e., look those new words up in a dictionary!) vs. just reading for fun!

    Bruce Humes
    Ankara, Turkey (just started learning Turkish . . .)

    • Steve (Syz) says:

      Hi Bruce,
      Did you find that “translation” flashcards sometimes gave you a wrong impression (or a poorly nuanced impression) of the meaning of terms?

      In general I’ve been against it, preferring the word-in-sentence-no-English approach. But I’m sure sometimes that unnecessarily complicates the task…

    • TWteacher says:

      Yes! I wrote almost exclusively about reading. The best way to learn characters is to read, the best way to learn conversation is to watch tv.

  15. Bruce Humes says:

    Interesting question. I also try not to “translate” when learning. Occasionally when I could, I defined the word in Chinese.

    But I guess for me the important thing was the act of writing out the cards and becoming familiar enough with them so that, when I was reading Chinese and came upon that character, I would feel familiar with that hanzi. It was like finding a friend among a crowd of people I didn’t know!

    I didn’t find it a problem to essentially forget the translation I had learned, and adopt the newer meaning suggested by the context in the Chinese where I came across the character.

    • Sima says:

      Some great points, Bruce.

      I think reading real texts is a very important part of the process and, obviously, recognising characters is not reading in any meaningful sense. I think it’s still difficult for learners to find much material that they can read for pleasure. I found that news really didn’t work for me. Ploughing through long papers on linguistics would occasionally sustain my interest, and the occasional novel helped, but I never really found a good source of appropriately sized – manageable – texts. QQ conversations helped, and I guess learners now might get sufficiently absorbed with something like Weibo.

      Your “friend in the crowd” analogy is a good one.

      One of the great benefits of learning Chinese in China, one which I think may often be overlooked, is that one is surrounded by written language, in the form of signs; shop names, building names, street names, advertisements, and the like, formed a great crowd, from which friendly faces started to appear with increasing regularity.

      Good luck with Turkish!

  16. Bruce Humes says:

    I also note that a fair number of foreigners are using their smartphones to input Chinese by writing. That’s also a good way to practice both writing and recognition.

Leave a Reply to aelephant