That would be endorsing the strategy which I personally think is abominable for learners of Chinese...

Serge wrote:

That would be endorsing the strategy which I personally think is abominable for learners of Chinese...

I've read your comments on RTH and hope you elaborate a bit more on them.

But as for what you've said already, I don't agree that the phonetics are all that helpful to learning hanzi.  I know that hanzi are more phonetic than kanji and I've already heard one other critic saying that phonetic markers are the most efficient way to learn, but honestly in my own time studying hanzi, I've found them to be more confusing than helpful.  I see a character containing 分 and I've had trouble remembering whether it's /fen/, /pen/, /ban/, /pan/, or something else.  It might be easier if they had completely different pronunciations!  And that's completely aside from the issue of tones, which raises another question: should he separate tones from the pronunciations to avoid overloading the learner?  Sounds controversial to me.

One of the main principles of the system is that he adds primitives as slowly as he can, to build up as many characters as possible with them.  I can't see any way to group the characters by reading without breaking up that progression, and in the end, whatever benefits can be gleaned from grouping characters phonetically can't make up for what would have to be given up in order to do that.

I know that hanzi are a bit easier than kanji in that maybe 90% have only one reading, but the quantity alone makes learning them a gargantuan task and, I think, merits the use of this book.

Last edited by JimmySeal (2007 August 31, 8:05 am)

I very much appreciate the experience pool on this forum and a generally very healthy discussion so will try to respond with my most detailed comments on why I believe this approach to be abominable for Chinese students. I will start now and - time permitting - I will try to substantiate my comments with some Chinese-specific examples in the next couple of days. This might turn out a little lengthy but then again, I'm no longer a frequent poster.

So as not to replicate the neighbouring Mega-Discussion on Phonetic Components and How-I-Think-Heisig-Is-Not-Efficient-In-the-Long-Eun-Even-Though-I-Never-Finished-It (which I stopped following a long time ago fearing for my sanity), let me qualify the word 'abominable'. Let me take it back, in fact. Any method is a valid method, including learning one character a day by doing nothing else. And I'm sure RTH will deliver what it says: teach its students X number of English keywords and the associated shapes. Whereas for Japanese it can be argued that this method makes sense (although I vehemently opposed learning English keywords when I first came across the book a long time ago), I think that it is not quite as useful for Chinese studies. One cannot blame Heisig for his attempts to capitalise on the success of his method but imposing the book on the Chinese students by the enthusiastic - yet generally poorly informed on the actual subject - Japanese learner community may be a step too far.

While Heisig's method has meant a lot to me, personally, in my studies, I believe there is a general trend on this forum to forget that it is just a Method (i.e., a collection of guidelines on how to approach a task) and to take every word in his books too religiously. Let me also state what I believe the true essence of the Method to be:

- recognising that characters break into components
- associating some concepts with each component
- associating some concepts with each character
- finding a mnemonic to link the above concepts together
- bulk-learning the characters by breaking them into manageable groups...
  - ...displaying similarity in components
  - ...yet regardless of their meaning
  - ...and regardless of their frequency

Anything else, such as (I will refer to this as 'the 5 points' later on):

- keywords associated with the actual characters
- concepts assigned to the individual bushou or components
- the list of characters to be learnt
- the order in which the material comes
- the absence or presence of readings, etc.

...is highly irrelevant and needs to be driven by each student's individual abilities, starting level and end goals. This is what Heisig himself repeatedly states in his text. Yet he needed to fill the book and hence provides one possible filling to the above 5 points. And there will always be people who will follow it religiously. Which in a way will work for Japanese because of the way Japanese works.

So, back to RTH, I do believe that the Method as stated above is a very valid one for the study of Chinese. But one does not need a book to follow this method: the info is generally available via websites like www.zhongwen.com (which by the way is how the AJATT guy claims to have done it), the rest is one's personal effort. And the actual filling of the above 5 points in the RTH book and - in particular - the absence of tones and readings is, in my opinion, far from being the most efficient way of approaching the Chinese language.

I am currently in the middle of my project in bulk-learning the Chinese script (striving towards the magic 6,000 number by year-end) and with around 3,000 of them - with readings and tones - firmly under my belt so far I will try to base my comments on my personal findings. (And sadly, or luckily, I have no Teacher figure to hide behind.) I know that Jimmy-senpai is similarly advanced yet we rarely agree on things so this might turn out to be an interesting discussion.

Examples to follow.

Last edited by Serge (2007 September 03, 6:09 am)

Hey Serge, I agree with a lot of things that you said.

I think we need to keep in mind though, that people have varying skills at self-learning. I think the process you use is not very difficult, but at the same time I think many people would be lost. So in that sense a method like RtK helps. It probably explain why many learners having finished RtK felt kinda lost with RtK2. RtK1 is such a straightforward goal.

As for your 5 points I'm not sure to understand them all: how do you study without the keyword/meanings? Do you mean perhaps that without doing readings as a second-step there is no need for the temporary "hooks"? So you study several readings for this small percentage of characters that have 2 or 3 readings, or do you learn just one reading and add others later (this would then be akin to the kanji chains method) ?

I disagree a bit about zhongwen.com. It's a great site, no doubt, but the book offers guidance I think like I said above it has its purpose. And zhongwen's explanations are not always easy to work with. I can see that it's doable though.

All the best with you study of Chinese!

Are you living and working in China?

And now, as promised, some examples to illustrate why I believe that RTH is not necessarily a good idea.

I will try to be very brief yet explain myself as fully as I can.


POINT 1: There is no point in delaying learning the readings in Chinese.

New things learnt in Japanese can only be applied on a very limited scale. New things learnt in Chinese can be applied immediately and with much higher efficiency.

One can easily write a book on all the grammar contained in something as simple as 「初めまして」. And as far as the kanji goes, one will have to learn that it stands for  「はじ」in「はじめる」.  And then discover about 「しょ」and「はつ」. And 「うい」in「初産」. And just when one thinks one is done, there comes 「そめる」. And this is just the first kanji that crossed my mind...

Now, in Chinese, one might learn the shapes and readings for 你 [ni2] and 好 [hao3]. Both immediately useful and highly recycleable in all sorts of expressions that even beginners can construct intuitively. The progress becomes TANGIBLE and applicable to all spheres of language learning.

Back to the Japanese, this very second I can think of over seven frequently used ways to express 你 and don't even get me started on the concept of 好... My point is - the progress will be much slower and the process much more frustrating.


POINT 2: Very few people will bring up the courage to embark on the project of learning 2000+ English keywords and the related hanzi shapes without actually getting other info on the Chinese language.

It is not impossible but it would take enormous determination to carry out. Most people will learn stuff about tones, will listen to some dialogues, will pick up some grammar... And realise that.... -> go back to point 1.

At which point they are likely to become frustrated with the method and will want to modify it in some way so the info they are getting becomes immediately applicable.


So we have established that there is immediate award for learning the shape, the reading and the tone at the same time. Now, on to point three.

POINT 3. Studying reading and tones at the same time as the hanzi shapes comes at very little extra cost.

Let's have a look at some real life examples.

Example 1: Piece of Cake

〔皇〕huang2
〔凰〕huang2
〔隍〕huang2
〔惶〕huang2
〔徨〕huang2
〔煌〕huang2
〔煌〕huang2
〔篁〕huang2
〔蝗〕huang2
〔鰉〕huang2
〔遑〕huang2

That's 11 characters in one go. No need for any further comments.

Groups like that (but may be smaller in size) are actually not that infrequent.

Example 2: Mixed Tones

〔漿〕jiang1 (paste)
〔鱂〕jiang1 (a fish sort)
〔將〕jiang1 (future tense, a general)

〔獎〕jiang3 (prize)
〔槳〕jiang3 (oar)
〔蔣〕jiang3 (surname)

〔醬〕jiang4 (soy sauce)

Once again, very little to comment. How hard is it to learn the tones? Easy if you think of it as an extra element in the character - you just need to add another bit to your mnemonics.

(In my case, for the above, I think of shapes and emotions: the first tone is a straight line or something automatic (like a general's way of thinking or the way one spreads the paste...), the third tone is something surreptitious (like awarding a prize, apparently - it helps if one has already learnt 賞 and 給) or circular (like oar movements)), the fourth tone is an order or a downward movement - like pouring the sauce).

Now, groups of this type are quite common too.


Example 3. Mixed Readings.

〔甘〕gan1
〔坩〕gan1
〔柑〕gan1
〔苷〕gan1
〔疳〕gan1

〔紺〕gan4

〔酣〕han1
〔蚶〕han1

〔鉗〕qian2
〔箝〕qian2

〔嵌〕qian4

Clearly, this is more complex and resembles RTK2 yet there is enough similarity to link all of this together. And, as stated above, once learnt, it becomes immediately useable (contrary to Japanese).

My feeling is that groups of this type are the most frequent ones in Chinese. They usually share the same final ([-an] in this case) which makes learning them much easier.

Example 4. Complete Nightmare.

〔路〕 lu4
〔露〕 lu4
〔璐〕 lu4
〔鷺〕 lu4

〔各〕 ge4
〔落〕 luo4, la4
〔格〕 ge2, ge1
〔閣〕 ge2
〔擱〕 ge1
〔咯〕 ka3, lo5
〔胳〕 ge1
〔骼〕 ge2
〔貉〕 hao2, he2
〔賂〕 lu4
〔駱〕 luo4
〔珞〕 luo4

Yes, one occasionally sees groups like this. Yet by now one would have acquired some mnemonic to deal with tones and enough concepts associated with the readings so one has lots of material to fall back upon. And then, just imagine that there are two extra strokes to learn with each of the above...  And note how the first [lu4] subgroup is actually a small group of its own with the 〔路〕component present in all cases...

Etc., etc...

POINT 4: With relatively small extra effort, it's possible to acquire all aspects of a hanzi and start applying it in real life from day one.

See above...


As a final thought, I think people who are most likely to come to the RTH book are students of Chinese frustrated with their progress in learning characters. And they might pick up some ideas but are unlikely to be happy about missing the readings bit...

我完了。 (^_^)/*

Last edited by Serge (2007 September 03, 6:04 pm)

Ok, here are a few of my points:

1.  Increased mnemonic complexity.

If I'm not mistaken, you're suggesting including the reading and tone in with the Heisig-like stories for the characters.
I think almost everyone here has felt the struggle to remember stories, and this is even more true for characters with more elements.  Anyone who, like me, tried to remember 隔 or 襲 with Heisig's mnemonics knows what I mean.
Trying to include two extra elements in every mnemonic story would create a huge mental burden in a task that's already taxing enough.

2.  Foreignness of Chinese syllables

I am curious how you are suggesting that learners group the characters by reading.  For a beginner, Chinese syllables will be quite unfamiliar.  Would they associate each syllable with an English word?  This would add about 200 extra elements into the pot (about 200 syllables) and require tying all those elements to syllables in one's memory.

3.  No immediate returns for added work

Serge wrote:

And, as stated above, once learnt, it becomes immediately useable (contrary to Japanese).

Indeed a Chinese learner generally needs less than a Japanese learner to immediately put a character into use, but this generally requires learning several compounds and just basically learning how each character is used.  This is not so bad for useful characters like 你 and 好, but not so for a lot of others.  And it's highly unlikely that the learner will become able to use all the characters they are learning, in the order they're learning them if they're grouping them by reading.  So as Ｉ see it, they are not getting such a large reward for tying the extra information of reading and tone into the characters at that early stage.

4.  No gradual learning of elements

In Heisig's method, primitives and other elements are presented in an order such that the learner can form (more or less) the maximum number of characters while taking in the minimum amount of new non-character (primitive/element) information.  And complex primitives and characters are almost always built up from simpler ones.  I think this is a major strength of the system.
To group characters by reading and still use a cumulative approach like this would be tremendously difficult, if not impossible.  Just to look at one of your groups:

serge wrote:

皇 凰 隍 惶 徨 煌 篁 蝗 鰉 遑

The learner would need to know 白, 王, "wind", "pinnacle", "state of mind", "going", 竹, 火, 虫, 魚 (and presumably, its various parts), and "road-way" in order to decompose these characters into their component parts.  Another set would require a considerably different set of components.
One could choose to throw all of these at the learner at once, but that again would require them to take on a glut of information that wouldn't soon be terribly applicable.
I suspect this was not a large problem for you as you were already quite familiar with characters when you started forming these theories.


So that's how I see it at the moment.  I look forward to seeing what you have to say.

As a student of Japanese with some experience in Mandarin, this discussion is interesting, but somewhat too academic.  I find the fact that Serge has developed a RTH-equivalent system interesting, but since keywords just give us a rough idea of a character's meaning anyways, I'm indifferent to what kind of keyword is used as long as I don't spend a ton of time learning the keyword itself.  After a while, the keyword falls away anyways...

What would I have to lose if I do the following?

1. Use RTH to learn English keyword -> Hanzi association
2. Learn the readings when I learn word combinations (Chinese equivalents of Japanese 熟語), naturally...

It is my opinion that learning Modern Mandarin (ignoring the gazillions of historic Hanzi), is much simpler than learning Japanese, if one can get a good grasp of pronunciation/pitches in Mandarin.

I figure RTH would be at least half the battle for Mandarin, where RTK is kind of a starting point for Japanese.   

The only ugliness is having to learn Simplified and Traditional...  ...I suppose, in parallel, is better than sequentially?!?  Anyone have insights?

I know this is a rather old thread, but isn't RTH scheduled to come out by end of this year? 

It would be fantastic if Fabrice can create a section for RTH practice/learning!!!

Last edited by kfmfe04 (2008 October 20, 9:38 am)

The way I see it, whatever method you decide to use to learn the Hanzi readings, it will be simpler and more consistent than learning the Kanji readings.  So there is no need to dwell on the methodology.

If Fabrice could add the readings as a feature to future RTH software, then it would be up to the student whether he/she would want to learn the readings at the same time.  Problem solved...

...the bigger headache may be learning both Traditional and Simplified concurrently.  I'm also curious as to how Heisig will deal with the many characters that are the same in Japanese and in Chinese, but the meanings/usage is quite different.  Different keyword or same keyword?  Maybe it won't matter either way, as the keyword is just a step towards learning the combined characters anyways (the REAL vocabulary).

kazelee wrote:

Also a question about Remembering the Hanzi... Are the traditional characters all the same one's covered in RTK?

According to:

http://www.nanzan-u.ac.jp/SHUBUNKEN/pub … zi%201.htm

PROGRESS REPORT (17 October 2008):

The books have been printed and are schedule to be at the
binders on 24 October. They should be ready for shipping
within a few days after that.

---------------------------------------------------------------------

So we should soon know the answer to your question, IF we can find a way to order some copies.  Anyone know where to go to order Book 1?!? 

I have a feeling many users of RtK will wolf down Book 1 before Book 2 ever hits the printing press.

Last edited by kfmfe04 (2008 October 20, 10:04 am)

kfmfe04 wrote:

...the bigger headache may be learning both Traditional and Simplified concurrently

I note that the introduction to RTH (in the downloadable extract) suggests that both at once is a bad idea, and recommends starting with T if you ever want to be able to write it, and S if you think you need to write S but will only ever need to read T, not write it.

Is there a Chinese equivalent to KANJIDIC that you guys know of ?

kazelee wrote:

Are you suggesting (with point one) that learning readings with the Hanzi is the approach you recommend? There are many who would agree with that. There are some who have devised methods of somewhat easily doing so in Japanese. So, I can agree mostly with what you say.

I'm just not sure as to why you call Heisig's method abominable to Chinese learners. For some adding a Chinese reading on a study card might be the way to go. For others it might be a it of an overload. Also, I'm not clear as to why one should omit the English keyword. Associating the Kanji with an English keyword, I would think, would give it much more weight than associating it with a Chinese word one barely comprehends, wouldn't it?

Or perhaps you mean it abominable when compared to other methods you've come across? Please elaborate.

Also a question about Remembering the Hanzi... Are the traditional characters all the same one's covered in RTK?

I have elaborated on all of the above in my post of 2007 September 03, 11:07 am.

Chinese traditional characters are in some cases different from Kanji, as some Kanji, too, are in fact simplified forms.

There was actually a METHOD behind the simplification madness and phonetic components played a significant part in choosing the simplified forms. That's why it's easier to learn traditional and simplified at the same time, together with the relevant phonetic components.