Memorising Chinese Characters - Mnemonic Approach

[Sixwinged]

https://www.youtube.com/watch?v=weZOrFei90E

Hey guys! I have recently uploaded a video on memorising Chinese characters by using mnemonics. What do you think? Which other systems for Chinese have you tried which are more effective? 

P.S. Feedback is always welcome

[Shelley]

I am always very sceptical about these types of memorising methods. I would prefer to put all that effort you put into your mnemonics into learning the meaning and character by my preferred method of practising writing the character until I know it.

 

I take it you haven't learnt chinese but are just trying to apply your memorising technique to chinese. I say this because you pronunciation is not too good and I think your understanding of characters is sketchy, and you have other videos that apply this method to other subjects.

 

Yeye is 2 characters 爷爷 , it is not one character made up of 2 identical parts, it is 2 separate characters, ye 爷 can be used by itself with other characters to make other words.

 

My personal opinion is there is no shortcuts when it comes to learning chinese, it will take time, hard work and persistence, but it is well worth the effort when you start to see your progress.

 

There may be other people who like your idea but its not for me.

[Xiao Kui]

Wow, what a complicated way to learn Chinese characters. Now do that for 2500 of them, and do that for the thousands of words made up of 2+ characters and you can completely waste your life before you have enough to read a newspaper or novel.  I think memory palaces have value for certain learners and certain information but I think simply writing the characters 5  times each, reviewing with flashcards, and applying them in context is the best way to remember them.  This worked for me, and I know that Chinese students have to write characters many times over in school as a learning method.  You tried to simplify learning the character, but in fact you actually complicated it.

 

You did seem to amuse yourself with your "bizarre and entertaining“ yeti scenario so at least you're having fun. i hate to criticize this video because it was very polished and it looks like you put a lot of work into it, but there seems to be more style than substance here.

[Gharial]

The problem with these memory palace approaches is that they try to "reduce" things to a parlour trick almost (invariably quite badly explained regarding just the procedural details let alone the overall purpose!), all the while overlooking the relatively painless associations that natural (less encumbered) language will provide (that is, a language isn't just a load of unrelated facts like a list phone numbers or a Generation Game conveyor belt of items https://www.youtube.com/watch?v=N0r4dDgde0M , even though a foreign one is admittedly foreign at first and may have genuinely complex things lurking).

 

That is, any beginner of Chinese will soon be able to tell you basic vocab like 爸爸 and 父亲 (from a character learning point of view, note the common component (radical)), and they really aren't that hard to learn (especially when the phonetic elements haven't been simplified - that 爸 vs 爺/爷 , stuff like http://zhongwen.com/ will speedily help in such instances: "Oh, I see, that bottom part's been simplified to a seal or stamp, but it doesn't seem very phonetic, fine, no problem").

 

Mnemonics may have a contribution to make regarding tones (e.g. an ad-hoc silly image such as "grandfather's expectantly raised eyebrows") and perhaps syllables generally (but note for example the horrors in a previous discussion concerning the book Knee Howdy's "A proxy mate's hounds" (i.e. 'approximate sounds', sort of similar to ice cream = "I scream as my sensitive teeth spasm at the coldness of the Haagen Dazs", thought the person with the pathological inability to remember 'ice cream' without a lot of pain)), but you skated right over actual pronunciation issues because you were so busy stocking the mental larder up with tinned yeti spam, Gritty Kitty cat litter, Thor's staple gun, etc etc etc. A spring clean may be in order LOL.

 

NB: Shelley's explained well your sketchy grasp or wooly phrasing (from 2:03 to 2:16 in your clip) of the character involved. You seem to be confusing the concept of words (often polysyllabic) with characters (necessarily monosyllabic, or morphosyllabograms, to give them a technical term). In the bisyllabic word yeye the same character is obviously repeated (i.e. occurs twice), for a total of two identical characters (compare mama, baba etc) separated by an actually-noticeable space (if you look closely enough and understand every character as proportionally occupying roughly the same amount of squareage as any other). Pretty sloppy, and undermines your whole venture if not approach. FWIW though there are or have been some "polysyllabic characters" knocking around, such as those given here http://languagelog.ldc.upenn.edu/nll/?p=3330 , but they seem to be frowned upon, and one could imagine the difficulties that such compounded characters might cause some learners (many people think the changes [that is, simplifications] like 爺 > 爷 were bad enough!). Mind you, apparently regional characters like beng2 甭 (bu4 不 + yong4 用) and fiao4 覅 (wu4 勿 + yao4 要), if we run analogously with the 'tuan'-like shortening or merging that Prof Mair mentions, don't seem to cause too much consternation.

[wibr]

@Sixwinged Welcome to chinese-forums.com, mnemonics are a controversial topic here :-). In general I have the feeling that people who have learned Chinese for a couple of years often form some strong opinions about the "right way" to learn Chinese, so we get lots of opinions and not enough science. I think your video gives a reasonable introduction, although I am not sure if your memory palace scales to the number of character you would have to learn eventually. The most common mnemonic system for Chinese is Heisig, which doesn't use a memory palace, only stories and images. Heisig also ignores pronunciations, which is one of its big issues. Another one is Matthews, which covers pronunciations but much less characters.

 

The way I see it, if you start out on a long project like learning Chinese you have to develop a certain toolbox and mnenomics can be one tool in this box. I like to think of it as glue, because that's kind of how they work. Some people don't like glue for whatever reason, so they never consider trying it. Some people use a little glue here or there to make things stick faster and easier, for tones or certain troublesome characters. But you can also build a whole framework of characters and maybe even words just with glue and then fill in the rest of the language later. If the glue gets old and dry, either it holds anyway because by then it's surrounded with lots of other stuff or you can just reapply it until it you don't have to worry about it anymore.

[Gharial]

Nice response Wibr, thanks for welcoming Sixwinged to the forums LOL. Not sure where the science or learning is in his clip though, but I digress.

 

I think it would help if the character he'd chosen were more complex or "disconnected" than apparently chosen at random (though I'll admit that relationship terms can be tricky).

 

I mean, in the learning sequence I proposed above (which I've no reason to believe is uncommon), students might well learn the radical if not the character 父, then the character 爸 (possibly after seeing that particular phonetic as used in 吧), then perhaps sometime later the character 爷 (but for which the traditional form 爺 will be of more use in understanding what the phonetic was rather than now is[n't] in the simplified form). I just don't see the great difficulty in that set or sequence, and the natural language (and "traditional" or everyday "plodding" thinking) surely provides the necessary scaffolding without the need for the baroque overlays. Maybe that stuff can come in once one knows hundreds~approaching thousands of characters, but I think it is definite overkill if used from the very beginning of one's Chinese studies. In short, I think it tests one's patience if not "imagination" too much, especially if potentially used unnecessarily. Give more "rote" methods or standard approaches a chance first - or are memory palaces now the new standard?

 

Matthews by the way was given a bit of a mauling over on Language Log by Prof Mair, and rightly so IMHO: http://languagelog.ldc.upenn.edu/nll/?p=1811 . The currently final comment, that Mair added several years after the main discussion had run its course, is probably one of the best pieces of advice one will get regarding how to learn to actually read Chinese.

[dtcamero]

ya i think Wibr put that very well. the video's ideas are much better explained by heisig, so attacking the video is kind of a strawman.

I used heisig's mnemonics to learn japanese to a comfortable fluency in 3.5 years.

i'm doing chinese the same way now it's worked very well here too.

i can show you a whole message board of people who agree mnemonics are great over at koohii.com.

if you don't like them that's fine but in my experience a great many people have used mnemonics to great success in self-studying languages that use chinese characters.

like srs, it's worth a try. you'll find out if it works for you pretty quickly, so why not right?

i think in self-study we should really always be trying new methods and exploring for some other thing that might improve our progress.

[dtcamero]

also...

http://languagelog.ldc.upenn.edu/nll/?p=1811

wow...

 

a lot of vitriol directed at a self-study technique.

 

i think it's funny how in the heisig controversy it almost always seems to be a bunch of people struggling trying to learn chinese characters on one side, talking about how heisig and mnemonics make no sense... they'll turn your brain into mush... try doing that for x,000 characters it's impossible...

 

and a few people explaining how they finished heisig for x,000 characters and have now moved onto more advanced areas of study or speak their target fluently.

 

maybe it's an inelegant solution that simply works (at least for some). practical people are succeeding using it, doesn't matter that it seems rediculous or has x many problems supposedly if you accomplish the goal does it?

[Zeppa]

Whatever the value of the mnemonic here, the video is not very convincing.

It uses 'word' to mean character, confuses the two characters forming one word, and uses 'character' to refer to a radical or element (the term 'element' would work better for part of a character, because the element might not be a radical).

It gives a vague idea of how a family tree might be based on the image of a house, but that doesn't tell us how to link the many members of the family tree, with their very detailed Chinese equivalents, with the Chinese characters involved.

 

It seems fair enough to me to direct vitriol at it! Of course there's the argument pro and con mnemonics too, but this video doesn't do much to fight its corner.

[roddy]

"a bunch of people struggling trying to learn chinese characters"

Hey, we're trying REALLY hard here.

[Gharial]

I don't think anyone's saying that mnemonics can't at least in principle be useful, "just" that it is important to get basic details right before trying to pass oneself off as an expert with stuff of supposed value to peddle.

 

But no, we (the worthless unwelcome critics proffering mere opinion rather than any clear facts ) are the ones struggling here, righttt.

[Guest realmayo]

a lot of vitriol directed at a self-study technique.

 

I agree, I actually used the book that was savaged by languagelog and I found it pain-free to 2000+ characters very quickly. Before I tried it, I had really struggled remembering characters. The book changed everything.

 

It's not for everyone, it requires imagination and trust, and you need to be happy to spend time learning material that you're inevitably going to forget: you don't forget the characters, but you do end up forgetting the mnemonic-stories. I think some people don't see the point in learning things you're going to forget, or things which will be superseded by new knowledge (i.e. true etymologies). But for me it was completely worth it, because all of a sudden I could read Chinese characters. Which, for me, was the important thing.

 

Victor Mair, like so many people, misses the point: the stories are just scaffolding, to help you build the house much faster. Once the house is built the scaffolding comes down and the building is as strong and sturdy as any other house. The stories aren't the foundations or load-bearing walls, they just help you move around much faster on site.

 

I think:

 

+ mnemonics (designed by someone who knows Chinese)

+ SRS

+ writing a character 10 times a day for a few days (I've no idea why this is out of fashion)

 

= quick way to learn characters.

 

 

However that is a bit off topic because while the OP's video might be perfect for learning the family tree, it's not very good for learning more characters than that. As mentioned above, a mnemonic system which makes use of the coincidences and genuine etymological relations within the Chinese writing system will always be more efficient than one, like the OP's, which does not.

[dtcamero]

ya i think people get hung up on the etymological irrelevance of mnemonic stories...if they are going to spend a lot of time memorizing something then that thing should be an etymologically accurate analysis of how the characters relate to one another.

 

@realmayo   as you say, the way Mair misses the point is that 5 years later no one remembers any of the stories, you have gone on to using the characters themselves. everyday use is the srs and you don't need the crutch of the mnemonic any longer.

the act of creating the mnemonic is a form of study, and after repping a character to maturity you generally don't need the story any longer. it is a useful link during a period of infamiliarity with the characters.

also, just as chinese characters build off of themselves, the stories build off of one another in a very efficient way... so a character with 25 strokes can be a very simple memorizable story.

 

i appreciate that people criticizing heisig don't want to use it, but since they have no experience with the system they also end up repeating arguments that are patently untrue.

lets go over mair's criticisms...

1. etymologically inaccurate - irrelevant to efficacy

2. outlandish explainations - in heisig you create your own stories, so not relevant

3. impossible to scale up - realmayo, myself, and koohii.com grads have each memorized thousands of characters with no mush-brain

4. must be frustrating - not yet

5. you'll go insane - debatable

6. opportunity cost re normal study... this one is probably the best, with a caviat. the argument for heisig i guess would be that chinese characters are a unique challenge that can be systematically attacked using a special method, before or separate from normal study. heisig def takes time but on the back end you know the meaning (and possibly) reading of so many characters that I would argue it is actually very time efficient in the long run.

 

i know a guy who actually met heisig, an accomplished academic of japanese philosophy, and asked him a little about the method. he seemed very embarrased actually, since it is such an inartful, etymylogically ridiculous approach... totally irrelevant to his academic work.

i don't care i can speak japanese now, probably would have burnt out otherwise.

[OneEye]

A few things:

 

1. You didn't "learn Japanese to a comfortable fluency" using character mnemonics. Fluency in a language is only tangentially related to orthography. You had to learn the whole language—learning a few thousand characters is a relatively small part of that project.

 

2. What do you do when you forget a character you haven't had to think about in a long time? This may seem like a trivial problem, and it usually is, but here's a scenario my company's co-founder Ash Henson likes to talk about. He was taking his PhD qualifying exams—handwritten, all in Chinese—and forgot how to write 頓 because he hadn't had to write it in well over a year. He was able to use his knowledge of character etymology, which Heisig completely ignores, to recall how to write it. His thought process: "What components can represent the sound 'dun'? 屯、享、尊、etc. The meaning has something to do with the head—what are possible semantic components that could be in the character? 首、頁, etc.

 

屯 and 頁 both rang bells, then he wrote 頓 and [頁+屯], recognized the former as correct and moved on. Had he not understood how characters work (etymology), he wouldn't have been able to do that. 

 

3. I don't agree that etymological accuracy is irrelevant to efficiency. I'm not against mnemonics. In fact, I used Heisig's books fairly extensively once upon a time (my now-defunct old blog is, embarrassingly, linked to in the comments in the Language Log post above) and I found them useful. I just think that mnemonics are much more powerful if you tie them in with how the character is actually structured—etymology. And why not do that? Mnemonics are flexible, so if you know that component A is a sound component and component B is a meaning component, why not base your mnemonic on that knowledge rather than purposefully ignoring how characters actually work, the way Heisig does?

[Gharial]

I can maybe see the relevance of Heisig's methods to learning Japanese, but quotes like the following (from Unger's Ideogram, in a note following remarks regarding the treatment of Kun or Japanese readings of kanji in volume 2 of Heisig's original course) make me wonder about its relevance to Chinese: "Interestingly, for On or Sino-Japanese readings of kanji, Heisig relies heavily on the fundamental fact that most kanji contain phonetic elements that provide a cue to the syllables in the reading. He does this in full knowledge that this signific-phonetic analysis is often at odds with the mnemonic linking stories he develops in volume 1".

[wibr]

@OneEye I think the Outlier approach will enable us to create mnemonics that are rooted deeper in the language and therefore provide stronger connections and better understanding. That being said, the superficial Heisig approach in combination with the keyword system also makes it very easy to use. Here are the components, those are the keywords, storytime! It's the same formula that can be used for all characters. Adding information about the role of components, pronunciation etc. will require additional effort. I hope that there will emerge some kind of systematic way to form mnemonics using the Outlier approach, like standard indicators for different kinds of components, tones etc. that can be included into stories or images.

[Shelley]

I have to agree with OneEye, why not learn something that will be much more useful in the long term than some quick fix temporary "scaffolding".

 

I personally really dislike Heisig, some of the mnemonics are so far fetched and tenuous that I find it hard to attach them to the character and I find the liberal scattering of Christian imagery is really off putting for me.

 

Any mnemonic system is going to have its problems, some worse that others, so why not use the system that already exists, use the etymology of the characters.

[Guest realmayo]

2. What do you do when you forget a character you haven't had to think about in a long time? This may seem like a trivial problem, and it usually is, but here's a scenario my company's co-founder Ash Henson likes to talk about. He was taking his PhD qualifying exams—handwritten, all in Chinese—and forgot how to write 頓 because he hadn't had to write it in well over a year. He was able to use his knowledge of character etymology, which Heisig completely ignores, to recall how to write it. His thought process: "What components can represent the sound 'dun'? 屯、享、尊、etc. The meaning has something to do with the head—what are possible semantic components that could be in the character? 首、頁, etc.

 

I think this is indicative of another slight misunderstanding about using mnemonics to help memorise Chinese characters. Someone who has learned those characters and who has been reading and using characters for a couple of years -- he will have realised that 屯 contributes to the sound of 頓。Use of mnemonics doesn't mean you're forever blinded to existing patterns.

 

It's also worth pointing out that the 頓 story above, if relied upon, would produce an incorrect pronunciation; a mnemonic would give you the correct one.

 

Why do 烦 and 顺 and 颇 and 颁 and 颤 and many more all have 頁? If the 'correct' reason is easy to remember, fine. But if it's complex, you've got to remember far more information that you would have to with a simple mnemonic. If you forget the etymological reasons why all these are 'head' then you will be scratching your 頁 thinking: 'I'm sure the fan in mafan is 烦 but I can't think what it's got to do with a head. So perhaps it's not that fan, must be another character'.

 

I do however agree that Heisig's method for Chinese, while perhaps great for Japanese, should take account of the pronunciations. But that is easily remedied: these days when I read 頓 I don't see its constituent parts, the same way that I only see television, not tele+vision, in English. But I looked up my original mnemonic for the character (I remember the character, I've long forgotten the mnemonic): it was:

 

The food was stored 屯 in empty heads 页 of slaughtered enemies, giving the guest a reason to PAUSE before the MEAL. Coco wanted to escape from the meal, but she was surrounded by sand DUNEs .

 

Although I started the mnemonics using the Matthews book, I did most of them on my own, using Wenlin for the etymology or the parts-breakdown.

 

 

And actually, the 'correct' etymologies are, for most learners, mnemonics too. The salient issue is whether strict adherence to those etymologies places less of a memory burden on a learner than using selected short-cuts instead. My sense is that the correct etymologies are too varied and inconsistent if adhered to strictly. But that they are good for cementing knowledge of characters, once someone has memorised a few thousand and is using them regularly.

 

 

Edit: OneEye, I completely agree with this, so long as room is left to ignore the 'correct' etymology in cases where that's deemed expedient:

 

Mnemonics are flexible, so if you know that component A is a sound component and component B is a meaning component, why not base your mnemonic on that knowledge rather than purposefully ignoring how characters actually work, the way Heisig does?

[Guest realmayo]

I have to agree with OneEye, why not learn something that will be much more useful in the long term than some quick fix temporary "scaffolding".

 

The reason is that you learn how to read Chinese faster. But I appreciate not everyone is in a rush.

[Shelley]

No, I am not in a rush but I do want to progress, but a more measured and thoughtful system is the way I want to learn.