I came across this article that discusses the element-based approach to learning (teaching, actually) kanji that Heisig devised as well as possible modifications to it.

http://www.sabotenweb.com/bookmarks/about/scott.html

One idea that I found interesting is that the author advocates teaching ALL of the "primitives" BEFORE learning to combine them into kanji. He divides "primitives" into two groups: "open elements" (that are not kanji themselves, like "top hat" or "flower") and "pure kanji" (the simplest in form and meaning, like 口、木、of which he says there are about 50).

And he claims it is best to teach/learn them in this order: "open elements" then "pure kanji," and after that combine them into kanji according to a specified order like in RTK.

What do people on this site think? Would you have preferred to have been presented all of the primitives first before setting about the task of combining them?

I think that such an approach might leave us more freedom and certainty when we want to devise additional primitives. (Especially if we are not concerned about going on to RTK3 formally.)

Do you think there is any value to separating out "open elements" and "pure kanji" and teaching them in this order?

Finally what other things have helped you in your study of RTK? What do you wish Heisig had expanded upon or included?

Six of one half dozen of the other. A lot of Heisig's kanji use a full kanji plus a "closed" or "open" element. And then you never see that Kanji again. So I don't see learning 200 elements first to necessarily be beneficial.

If I were to write a program that would work like Heisig's vol1, I would explain the principles, and then let the user name the components. With each component the kanji that use it (and their meanings) would be displayed. Perhaps also popularity of the kanji would be displayed. Then one could try to name the componenets with names that might work with those kanji.

I had a terrible time with the Person on the left component. I could not think of a person I wanted to be in all of my stories, and so it remained a generic person which was hard to make stories for. I tried naming it a Toothbrush since it sort of looks like one, but that just didnt work. It is hard to rename a component once you have started using it.

I hated Pinnacle. I switched over to Kanjican's "Acropolis" (or some classic Greek building up on a hill) and had better luck.

Didn't and don't like TaskMaster. Somebody here uses Fidel Castro. I love it. I'll probably start using that as I am reviewing all the kanji I never learned very well (most of em LOL).

That DOS based RTK program that came out about 15 years ago had an option to let the user rename the components.

Another thought. Looking at the types of stories that work for people it is REALLY clear that the mind works through one's OWN memory threads. Some people make a vivid story that is silly and memorable (at least to the creator), while others make some obscure comments that are probably usuable only by the creator.

Well, first of all, I'd have preferred it to be a Wikibook (http://en.wikibooks.org/wiki/Main_Page) with links for each kanji to Wiktionary and best stories from the community collected there for each kanji ;-) This site mostly makes up for that lack, except that you don't get a printed version with everything (the community stories and other notes!) included like you can with Wikibooks (from a PDF form), and that it's harder to refer other people to a relatively obscure physical book rather than something on the net.

The book itself was mostly useful to me for its list of primitives and their assigned meanings. All the other useful parts to me were on this site, and the primitive meanings themselves were useful mostly because that's what the community here produces stories for. That's why I think that this site is the more useful part of the combination.

I'd also have preferred a list of the primitives in one place and hyperlinked (yes, you need an electronic format for that) to the kanjis where they are used. Less errors would also have been nice, as I ended up not trusting the stroke order and stroke count information in the book at all, checking them elsewhere instead. (http://kakijun.main.jp/ is nice for non-jouyou kanjis, btw. The site has some peculiarities, such as 虎 being written horizontal stroke first, but the site owner insists that this is the more popular way to write and that it is not incorrect)

Information about the true etymologies of the characters and true meanings of the primitives would have been a nice addition (so you know what you're ignoring), as would true names for them (草冠 etc., so you can easily verify what to call them when speaking to the Japanese) and Japanese versions of the keywords (with identical meaning; some here complained that switching back and forth between Japanese and English is an annoyance, which I can sympathize with).

The keywords could've also been more accurate, always giving a meaning of the kanji that is widely in use (and I do mean meaning of the kanji, not meaning of a random compound word it is used in) as well as a good description of the meaning with examples, like you get in Japanese kanji dictionaries (but in English). They should also have been much less obscure English words. I don't think the uniqueness of keywords is a particularly useful thing, you could use "keyword (in a ... sense)" which would be both more accurate and readily understandable. Also, gratuitous skew of meanings would've been nice to avoid, e.g. 草冠 being "flower" instead of "grass". I used it in the sense of "grass" many times in later adaptations of my stories (which I have not posted here as they diverge from Heisig) because it just fit better in those cases -- often because the real etymology happened to be also easy to remember (something that they often are not).

Many of these things may not be useful to many who go through the book for the first time. Due to my own peculiar way of going about it - note that I replaced keywords with Japanese examples - I dug up most of the things mentioned here myself when going through it. Looking at the forum posts, there are others who find parts of what I mentioned here useful. It all depends on what level you are at when reading the book and what you find interesting/useful/necessary to learn at the same time. In any case, such things are a useful reference later ("later" does come, as you need to review). And, umm, why exactly did you want to know? :-P

Chadokoro_K Wrote:I came across this article that discusses the element-based approach to learning (teaching, actually) kanji that Heisig devised as well as possible modifications to it.

http://www.sabotenweb.com/bookmarks/about/scott.html

One idea that I found interesting is that the author advocates teaching ALL of the "primitives" BEFORE learning to combine them into kanji. He divides "primitives" into two groups: "open elements" (that are not kanji themselves, like "top hat" or "flower") and "pure kanji" (the simplest in form and meaning, like 口、木、of which he says there are about 50).

And he claims it is best to teach/learn them in this order: "open elements" then "pure kanji," and after that combine them into kanji according to a specified order like in RTK.

What do people on this site think?

Interesting article. I'm impressed that he actually tried to teach this a little. I was a teacher for a while, and I can't understand why people think Heisig would be hard to teach. To me, it seems it would be easier to teach than just about any other aspect of the language, although very time consuming. Comments?

After he taught all the "elements", I was surprised that he didn't go for the holy grail of kanji learning - learning kanji by frequency number using the Heisig method. Sure, you lose the benefits of the domino effect, but why not just go for it and make everybody happy? I've heard of several chinese students learning Hanzi this way; they learn the primitives first, then every new hanzi they encounter they just Heisig it. Piece 'o keiki.