Ok I haven't actually GOT the book yet, so I am curious as to exactly why Heisig- and I assume more RtK followers- believe studying Kanji to Keyword is a bad idea?

Surely the point is to associate the keyword strongly with the kanji in both directions? Thus when you see the kanji in writing you can recall the keyword.

I don't understand...

In another post someone wrote...

"I just don't really see the point. Why waste time learning an english keyword from a kanji when you can spend that time learning japanese compounds instead? It gives you meaning to the kanji while expanding your japanese vocabulary at the same time."

But surely thats what RtK is all about!? Whats the point of this whole thing if the Kanji-Keyword link isn't two-way? Why bother learning keywords for kanji? I mean, surely the only ability that then develops is being able to write the character you remembered for a certain word, and nothing else.

I really don't understand....

stevesayskanpai Wrote:Ok I haven't actually GOT the book yet, so I am curious as to exactly why Heisig- and I assume more RtK followers- believe studying Kanji to Keyword is a bad idea?

Because it takes too much time.
Time that would be better spent studying new kanji.

Quote:Surely the point is to associate the keyword strongly with the kanji in both directions?

The point is getting to know the kanji, and being able to write them without mistakes.
The keywords are there for testability only.

"Why waste time learning an english keyword from a kanji when you can spend that time learning japanese compounds instead?"

Because:
1. There would be too many bits of information to learn
2. Unless you truly know the meaning of the compounds, there will be some vagueness
3. The keyword is only for testing purposes and to give you ONE meaning for the kanji

In Detail:
1. Without knowing any Kanji to start with, what happens when you see a compound? Right away, you are tempted to learn how to read it. Then you are trying to discern how much of the meaning of the compound comes from the first character and how much comes from the second character, if at all. Rather than focusing on the parts of an individual Kanji (as in RTK), you are already looking at combinations of characters.
2. Let's say you understand the meaning of the compound. Do you know the meaning of the individual Kanji? Maybe, maybe not...
3. If you use Japanese compounds as your hook, how would you test yourself for the Kanji? You could use hiragana, but then there are many compounds that sound the same. It gets messy pretty fast.

Finally, a good reason is, it has already been done for you by Heisig! Unless someone else comes up with a complete system (for example, based on compounds), you will have to spend a lot of time developing the system. It could very well take you more time to make the system than to complete RTK.

Studying ONLY kanji to keyword is a bad idea. Nothing really wrong if you do it both ways. Kanji to Keyword will be the easier of the two reviews.

However, if you do it, I recommend making it more a Kanji to Concept review. If you see 自, mark it right if you think "I, myself, self, me" even though the keyword is Oneself. The keyword is there to help you recall the kanji. Don't use it to force the reverse.

Anki is a great choice to do this.

[Katsuo]

stevesayskanpai Wrote:Ok I haven't actually GOT the book yet, so I am curious as to exactly why Heisig- and I assume more RtK followers- believe studying Kanji to Keyword is a bad idea?

Heisig says it's unnecessary rather than bad.

RTK1: Introduction Wrote:Third, the kanji are best reviewed by beginning with the key word, progressing to the respective story, and then writing the character itself. Once one has been able to perform these steps, reversing the order follows as a matter of course. More will be said about this later in the book.

RTK1: Chapter 11 Wrote:Let me repeat: study only from key word to kanji; the reverse will take care of itself.

You can download a huge chunk of the book for free from Heisig's pages here.

I was the one who wrote the quote above and I stand by it. Studying keyword -> kanji is already two-way. I've never studied kanji->keyword and I can name the keyword whenever I see a kanji. It gets stuck in the same process. Maybe not as well as the other way around, but it isn't as important.

So why waste time on it??

[ファブリス]

At the end of the day, Heisig recommends doing keyword > kanji because that's how you practice also remembering how to write the buggers.

Until you complete RtK 1, doing both ways reviews would be overwhelming in most SRS's. So unless you want to skip the writing, you do keyword > kanji reviews, regardless of considerations about the english keyword.

Katsuo Wrote:Heisig says it's unnecessary rather than bad.

Thus it's a waste of time, therefore bad.

I think the logic is that keyword to kanji recall is a "production" task, and production tasks are more difficult than recognition. So if you can do production, you will be able to do recognition easily, but if you train only recognition you might have troubles with production.

The research shows, though, that training recognition is important for fluency, so I do periodic kanji-keyword speed reviews, with the criteria for success not being "did I recall the keyword correctly", but rather "did I recall the keyword quickly".

joxn_costello Wrote:I think the logic is that keyword to kanji recall is a "production" task, and production tasks are more difficult than recognition. So if you can do production, you will be able to do recognition easily, but if you train only recognition you might have troubles with production.

The research shows, though, that training recognition is important for fluency, so I do periodic kanji-keyword speed reviews, with the criteria for success not being "did I recall the keyword correctly", but rather "did I recall the keyword quickly".

The only problem with that is that it roots the english keyword really hard in your brain which isn't really the point. You aren't supposed to know really quickly that 録 means record, you're supposed to know that it has to do with records, as in the word 登録.

There really isn't any point in being "fluent" in english keywords.

Tobberoth Wrote:

joxn_costello Wrote:I think the logic is that keyword to kanji recall is a "production" task, and production tasks are more difficult than recognition. So if you can do production, you will be able to do recognition easily, but if you train only recognition you might have troubles with production.

The research shows, though, that training recognition is important for fluency, so I do periodic kanji-keyword speed reviews, with the criteria for success not being "did I recall the keyword correctly", but rather "did I recall the keyword quickly".

The only problem with that is that it roots the english keyword really hard in your brain which isn't really the point. You aren't supposed to know really quickly that 録 means record, you're supposed to know that it has to do with records, as in the word 登録.

There really isn't any point in being "fluent" in english keywords.

The Kanji ultimately become a "concept" that no one language will have a monopoly on it. Now, the various Chinese languages and Japanese will be the main ones you use the Kanji with, but no one here is at native fluency in either language (I'm assuming). Having the concept of the Kanji rooted in your brain in English is not a bad thing anymore than a Chinese person will have it rooted in Chinese when going off to learn Japanese. You're going to root Japanese if not another language to that same Kanji soon enough.

Plus, the benefit of rapid recognition occurs with reading, as Costello points out above.

My recommendation (and in the end of the day, that's all any of us are giving) is that you must do Keyword to Kanji. If you add on Kanji to keyword, try to keep it more open to interpretation (you don't have to give the precise word, just know the kanji's meaning).

As Fabrice points out, you may overwhelm yourself early on if you try to do both. So maybe hold off on it and see if you'd like to try it later.

I have to agree with pretty much everything nukemarine said.

Start with keyword -> kanji. Not doing keyword->kanji will make the process very hard.

Tobb said he got two way recognition from doing keyword reviews. I didn't for all kanji (typically kanji which I didn't have any vocabulary for).

No matter what level your Japanese is at, start with keyword -> kanji.
If you're a beginner (some people are doing RtK before even knowing any Japanese!), keep up with the keyword reviews.
If you're at an intermediate or higher level in which you frequently encounter kanji and already know a decent amount of kanji vocabulary, consider slowly doing your reviews from kanji (once you have finished doing keyword reviews!)

The fact is that when you are doing Japanese, you will need to almost always go from kanji to keyword. And as Nukemarine already said, it should really become like a "concept" instead of a keyword. Going kanji to keyword will solidify this once you are regularly encountering kanji in vocabulary.

I'm In Japan at the moment and at a language school here so I have kanji overload at the moment. To be honest I don't really think that there is much practical use in going kanji -> keyword because as far as I can tell the actual amount of useful information you can glean from just knowing the Heisig keyword for a kanji is very limited (sometimes it is not even one of the translations of the kanji); that is not to say it isn't interesting, but it won't help you get more than a very basic idea of what a phrase you read means. So the question is really what do you hope to gain from going kanji -> keyword, because it wouldn't seem to really give you anything useful?

However, what I do find is that when I am reading the textbook I can understand it much better and get the reading of the kanji and if I'm doing my homework I can use the kanji I know and thus get the Japanese readings. Both things which seem to have been said adinfinitum are the benefits of RTK.