As an alternative point of view, I used the Heisig method as well as this site briefly (for a few weeks last year) but gave it up because I didn't like the method.

At that time, I knew virtually zero Kanji, and my Japanese level was sub JLPT4. After giving up on Heisig, I tried a variety of different methods, but it was mostly just learning by context, studying the radicals and etymology (which is how Chinese students study Hanzi), and looking up any new character I encounter through an electronic dictionary. I have not used SRS in my studies (I simply have no time - in total I probably only spend a few hours each week learning Japanese.)

Six months later, I have no problems using a laptop with the Japanese version of Vista. I also use a variety of other electronic components (camera, Sharp Zaurus, projector, GPS) using the Japanese user interface. I managed to read the operating manuals for these devices (in some cases cover to cover) in Japanese. I have also managed to successfully read a few manga and children's books in Japanese. I can usually understand the headings of magazine articles (though not the full text). I converse with my Japanese teacher in email using Kanji. I can follow recipes in Japanese.

Recently I went on a 2 week trip to Japan (my first time!) where I spoke in Japanese almost exclusively. I checked into hotels, made bookings, ordered in restaurants, gave directions in taxis, bought items in Japanese. I had no problems reading train and bus timetables in Kanji, and I found I can easily read most advertisements in trains, public places etc. I can understand some of the commentary provided by monks in temples (as well as signs giving directions, suggested routes etc.).

My spoken Japanese is good but not perfect. People have no problems understanding me, but I get lost if the conversation becomes very complex (TV is a bit hard to follow, they speak so fast!!!). Many shopkeepers assume I live in Japan, and keep offering me discount point cards (which I politely refuse) - I notice they seldom offer them to people they perceive as "tourists." Some look really surprised when I show them my Australian passport and ask for a tax refund. One person in a kombini was very intrigued by my Australian drivers license when I opened my purse.

One man in Osaka asked me whether I was from Brazil, and refused to believe me when I said I was not. I couldn't figure out why he was so insistent, then I did some googling and discovered there has been a strong wave of Japanese migration to Brazil (particularly from the Kansai region) early in the last century, and therefore there are lots of vaguely Japanese looking people from Brazil who cannot speak Japanese well (he must have thought I was one of them!).

By the way, I am not Japanese, and have studied Japanese only in my spare time for less than two years. I do not know Chinese or Korean either. I don't have any close Japanese friends, so my only interaction in Japanese is with my teachers and classmates. I don't believe I am particularly gifted in languages, but I can attest there are far more effective methods of learning Kanji than Heisig (I think starting with a good working knowledge of radicals is a must. I also like writing out the characters multiple times - which is completely contrary to the advice Heisig gives - I think writing helps me remember the shape of the characters simply through motor memory).

So I can definitely attest it is possible to build a reasonable (but basic knowledge) in Japanese (listening, speaking, reading, writing) without using Heisig, and without using SRS. Learning over 2000 characters is not a prerequisite. I think I barely know more than a few hundred. However, that doesn't stop me from doing lots of reading, even complex technical documents like operating manuals.

More often than not, I can guess the meaning of any characters I don't know, simply through context and using the radicals of the characters as hints. For example, in one of the hotel rooms, the controls for the airconditioner was purely in Kanji, and even though I didn't know many of the characters, I had no problems figuring it out (the setting for "cool" for example had the "ice" radical, the button to operate the unit had the "movement" radical, etc.). By the way, radicals are not the same as Heisig primitives (I think this is one of the worst flaws in the Heisig method - it complete ignores the fact that the Chinese has already invented a really good system for decoding the reading and meaning of characters - all one needs to do is learn it). I am now reaching the point where I sometimes don't even bother looking up unknown words in a dictionary, I just guess and move on.

Last edited by Christine_Tham (2008 June 20, 7:59 pm)

Dragg wrote:

@christine

How are primitives so much different than radicals?  As far as as I have seen, they exactly the same almost all the time. I don't see how this is even close to a flaw, let alone the serious one that you make it out to be.

Read the following books (more or less in order of increasing difficulty):
- Let's Learn Kanji
- The Complete Guide to Everyday Kanji
- 例解学習漢字辞典
- The History of the Japanese Written Language

The short, but possibly obtuse, answer to your question is: Radicals are semantic markers. Primitives may not be.

The best book out of the above is 例解学習漢字辞典, but unfortunately it is written entirely in Japanese. But if you ever wonder why is it a certain character is written a certain way, this book will tell you. I have had plenty of "a ha!" moments reading it.

There are also quite a few Chinese textbooks that explain all this, but I would recommend avoiding them, because the Chinese use Hanzi very differently from the way Japanese use Kanji. That's why Chinese students tend to struggle so much learning Japanese.

There are also some excellent linguistic research papers, unfortunately these tend to be rather dry and heavy going. I do remember an article in a book called "Japanese Linguistics" that made it very clear for me.

Being pedantic here, 210 of them were admitted. Four were omitted.

In any case, while this does help the mass memorization of the kanji, it seems like it will be even more annoying later on. It also causes confusion with the rikai-chan kanji lookup.

Christine_Tham wrote:

So, any teaching method that entirely relies on associating each Kanji character with a single keyword (in a foreign language) is not based on the way native Japanese speakers actually read and use those characters.

Well, I plan on forgetting all these useless keywords. I already know a lot of Japanese, and am just using Heisig to assign arbitrary placeholders for all the kanji. For future use.

@ christine

You still haven't really proven anything... Just because we have learned differently than a Japanese native doesn't necessarily mean our method itself is any worse. 

I'm not surprised if the average Japanese person does not think too much about semantic origins, and instead focuses on readings.  But that's because the characters roots are often cloudy or sometimes a radical was meant to be a phonetic marker, in which case it would be essentially meaningless beyond hinting at pronunciation for the Chinese.

Heisig essentially adds a semantic layer through the use of stories, which although often somewhat artificial, still gets the job done.

Last edited by Dragg (2008 June 20, 10:27 pm)

playadom wrote:

Well, I plan on forgetting all these useless keywords. I already know a lot of Japanese, and am just using Heisig to assign arbitrary placeholders for all the kanji. For future use.

I think based on the research I have quoted earlier, we can safely assert the future use of these keywords is nil.

Therefore, the only use of these keywords is to act as an interim indexing (and possibly cataloguing) mechanism into the shape of the characters, kind of like a mnemonic "hash" into the character's shape (and associated story).

But, there are more efficient ways of indexing these characters into long term memory. Again, I would encourage reading some of the books I have mentioned earlier. None of this is new. People have been studying these characters for thousands of years. The Chinese and Japanese school systems know how to do this well. The reason why adult learners struggle is I suspect as adults they are too proud to scribe the characters on an exercise book and learn the nursery rhymes that helps memorization of these characters.

Try maintaining a Kanji writing practice book. Try chanting some of the verses taught to Japanese school children (some of them are quite cute!). You will be amazed how well you will start recalling, even without using SRS.

if you guys are telling me that those phonetic markers are actually useful to anyone other than the Chinese, I would love to hear more about it...   I have a kanji etymology book and they seem kind of useless...

For example, the kanji for "block, exact" which literally depicts a nail is used as a phonetic marker in the kanji for "hit, strike".  However, the on-yomi for the two characters are completely different.  Therefore, whatever phonetic purpose it served for the Chinese is completely lost for Japanese making it useless for all practical purposes.

yukamina wrote:

Phonetic markers are great(it's true, I look at them first when reading kanji words), but they aren't very reliable.

The statistical analysis done by de Francis et al suggests they are very reliable, much more reliable than semantic associations.

Read "The Complete Guide to Everyday Kanji." That outlines a method of teaching Kanji via phonetic markers. You will find the method to be well described, robust, and reliable.

And best of all, if you use SRS, you can probably master the method used in this book in roughly the same time as finishing RTK1 (my educated guess, unproven assertion). And you'll end up knowing the readings as well as the phonetic and semantic associations.

RTK1 is not necessarily the fastest or best way if your goal is to memorise over 2000 characters quickly.

playadom wrote:

So I can find some of these nursery rhymes in the books? Do you have any examples?

Just pick up any Japanese grade school textbook. Start with Grade 1, work upwards. Lots of examples.

Best of all, you can pick these up dirt cheap from a second hand Japanese bookshop.

From experience, you can probably work through, say Grade 2 or 3, in a matter of weeks.

If you can afford it, hire a Japanese teacher for a private lesson, and get them to read through the book and tape them doing it. Listen to the tape in your car. It will improve your vocabulary, as well as speaking accent.

Dragg wrote:

I have read a good book on Chinese etymology and I also have what I believe to be a good kanji etymology reference book as well (Henshall).  So don't make it sound like what you are saying is over my head.

Although this subject matter might be a little confusing to the unitiated, it doesnt necessarily warrant buying a handful of books just to get the gist of it.

If you recall, I specifically advise against reading Chinese etymology.

Henshall unfortunately is unnecessarily academic in his prose, and contain a number of glaring errors. I wouldn't recommend it (I used to, but I have been misled too often). The problem is Henshall based his book on Chinese etymology. Enough said.

All I am asking is, before you start questioning, and say nothing is proven etc, why don't you simply follow my advice and read some of these books? Is that too hard? You have clearly demonstrated you are not adverse to reading more than one book.

Borrow them from a library - why buy them?

I'm not going to buy a book just because someone says "buy it".  I'm a relatively poor college student so, unfortunately, I need to be convinced first.  As far as the Japanese phonetic book that you described, about how many kanji really have phonetic markers  that enable them to be learned  this way?  If you say something like 50, its not worth my time or money.  If its on the order of 400, then it sounds worth a look...

Last edited by Dragg (2008 June 20, 11:20 pm)

Heisig is pretty quick going, I did mine in two months.  I doubt the radical method is very different anyway, and definitely it will take longer.  You admitted yourself, Christine, that you only know a few hundred kanji, so this must be weighed.  You know a few hundred with readings, I know 2000 without readings.  It's kind of even, and everyday I learn more and more readings.

Personally, after Heisig I completely stopped trying to do kanji reps and remember keywords, but this didn't matter, the kanji are in my head now.  When I see them they are easily discernible, as are their radicals.  This is all I wanted from the method, now I learn words and recognise kanji easily.  I do sentences of course, which reinforces kanji daily and sorts out the commonly used from the esoteric.  As for writing them out a lot, why?  Just why?  I'd prefer to learn ten words containing that kanji with an SRS.