This issue gets discussed fairly frequently.  The keywords aren't "meanings", and it's considered counterproductive to drill from kanji to keyword.

Yeah this is a solution looking for a problem. I understand why someone would feel the need to do this, but trust everyone here saying it's not necessary. Those links will form in your head even if you don't explicitly work to make them. Keyword -> kanji is enough.

In Lesson 31 Heisig writes in his cryptical fashion: "There is yet a fourth stage to be reached, as you have probably realized by now, but one you ought not trust until you have completed the full list of the kanji given here. In this stage, the primitive elements are suggested according to form without any immediate association to meaning. Quite early on, you will recall, we insisted that visual memory is to be discarded in favor of imaginative memory. It may now be clear just why that is so. But it should also be getting clear that visual memory deserves a suitable role of some sort or other, once it has a solid foundation. This is a process not to be rushed, however appealing its rewards in terms of writing fluency."

What he's trying to say (I think) is that at some point you will make the transition from connecting meaning to character to seeing kanji as a signifier of an abstract meaning or a reading, and sometimes both. But it really depends on the context.

Let's say you know 子 and 捻 (screw). The compound used is 捻子 [ねじ] - screw. What's the meaning of the 子 in this? I would say it only denotes the じ sound. As Inny Jan says, the keyword meaning is just a label of some sort.

Ok I think I'm starting to get what's being said here and I appreciate all the detailed responses.  I find Heisig's explanations of his method pretty convoluted and difficult to follow, so maybe someone can make sure I'm understanding things correctly.

So a keyword is a meaning (in english) associated with an individual kanji such that words in japanese which have meanings similar to that keyword often contain its associated kanji, is that correct?

I guess what I still don't quite get is say all I want to do is read japanese and not write it, then what advantage do I gain by knowing the keyword -> kanji direction?  In this case I would have all the kanjis in front of me and thus know all their associated keywords and then would be able to figure out the vocabulary based on the way these keywords combine, correct?

ファブリス wrote:

The keyword is first and foremost a remembering device.

Conveniently, it does in fact often (not always) suggest closely one of the meanings of the kanji.

This.
Keywords don't always mean the same thing as the Kanji does, and it isn't always the most frequent meaning. The only purpose of them is so when you remember a word, you know how to write the Kanji.

construct wrote:

Ok, but I still haven't heard a reason as to why learning in the keyword -> kanji direction provides any advantage when it comes to reading.

It may not, but kanji->keyword basically doesn't provide any advantage for reading either.  RTK 1 isn't about reading.

In this case I would have all the kanjis in front of me and thus know all their associated keywords and then would be able to figure out the vocabulary based on the way these keywords combine, correct?

No.  This is a common misunderstanding about kanji; having some idea of a kanji meaning can help you learn words, but (as a beginner) you cannot reliably guess the meaning of a kanji compound just from knowing English meanings or keywords. 

If I had to guess, I would say that only about 20% of Japanese words have a completely and totally transparent "meaning 1 + meaning 2 = word meaning" construction.  Another 15 or 20% are totally opaque; they either use obscure/archaic meanings or have nothing to do with the meanings at all.  The majority of compounds make sense once you know the meaning of the word, but if you tried to guess the meaning based on the kanji you would have a good chance of guessing wrong or being misled.  And even then, just knowing the meanings usually won't let you see the finer shade of meaning the word has.

Last edited by yudantaiteki (2012 June 08, 10:18 pm)

JimmySeal wrote:

Let's say you know 子 and 捻 (screw). The compound used is 捻子 [ねじ] - screw. What's the meaning of the 子 in this? I would say it only denotes the じ sound. As Inny Jan says, the keyword meaning is just a label of some sort.

In 捻子, 子 is a suffix that means "small object."  I'm almost inclined to say that 子 has no sound at all in this compound, because 捻 carries the sound of ねじ all on its own.

Yes, I think this is right -- ねじ is the renyokei/stem form of ねず.  捻子 is the Chinese word for "screw" -- I'm not sure if 子 actually means "small object"; I thought that in words like this it was just a meaningless suffix tacked on to a morpheme to avoid a single-syllable word.  But maybe it is only used for small things.  Anyway, 捻子 read ねじ can be considered an 当て字 -- I think it's just a coincidence that the on-yomi of that compound is close to the native Japanese word.

construct wrote:

Ok, but I still haven't heard a reason as to why learning in the keyword -> kanji direction provides any advantage when it comes to reading.

Because you want to be sure to be able to recall the characters from the keyword (which often is a distinct concept/idea similar to the kanji meaning). This ensures that you know how the characters are built and distinct from one another, which is important because there are many very similar looking characters.

Then, in due time, simply through reading and studying vocab you'll get the kanji -> keyword. You may have to check the keyword now and then but the connection will be made.

Until you're done with RTK, kanji > keyword is counterproductive. You want the more difficult keyword > kanji that creates the memory connections between the idea/concept and the components.

I was actually wondering this same thing regarding kanji to keyword.  Yes maybe it's not necessary to know it this way around and YES Heisig says it and everybody says it numerous times that we should learn keyword to kanji only and the other way around is basically insignificant. 

However, I'd like share my personal experience as I"m absolutely delighted learning vocabulary words after knowing Heisigs keywords.
I'm nearing the last couple hundred kanji of RTK1 and so i've started using iKnow to learn vocabulary this week and the readings and I MUST SAY, that having done Heisig beforehand makes it so much more rewarding and I personally find that it is easier to remembering the writing of the vocabulary word when I can recall the keyword from the kanji.  When I play the brainspeed game I pick up the written form of the vocabulary word faster for example:  見る- look, see. 、使う - use、秘密 - Secret
Being able to recall the keyword allows me to remember the writing usually after seeing it just once.  Like 'Secret' (himistsu) is the character for Secret and Secrecy together.  I love it!  I'm sure it may not always be the case but I think it really speeds up the process a bit.

That's the benefit I personally find in recalling kanji to keyword.  I suppose since flashcards aren't available, the vocab shuffle, staring at a kanji poster will have to do.