Hello everyone.

I have been learning Japanese for a while already and I just started reading RTK now. The book is pretty good so far and when I don't like a story, this site helps a lot. When I reached lesson 5 though, I found this: "Also, when you review, review only from the key word to the kanji, not the other way around."

I use anki daily for a while now, and memorized tons of vocabulary that way, but I always made my cards both ways. As such, as soon as I started this book, I always added keyword to Kanji and Kanji to keyword for every Kanji. However, Heisig tells you to only review from the keyword to Kanji and not the other way around. Doesn't this hurt recognition? I always reviewed both ways but if one way is enough, I could delete half the cards in my vocabulary decks too to save some time on reviews..

I'm going to stay non-controversial here and just state that the general consensus seems to be that as long as you are doing both, it won't really hurt you but that real reading recognition comes later as you learn words/kanji-readings and that what you can write, you will eventually be able to read as well.

Lightning Wrote:Hello everyone.

I have been learning Japanese for a while already and I just started reading RTK now. The book is pretty good so far and when I don't like a story, this site helps a lot. When I reached lesson 5 though, I found this: "Also, when you review, review only from the key word to the kanji, not the other way around."

I use anki daily for a while now, and memorized tons of vocabulary that way, but I always made my cards both ways. As such, as soon as I started this book, I always added keyword to Kanji and Kanji to keyword for every Kanji. However, Heisig tells you to only review from the keyword to Kanji and not the other way around. Doesn't this hurt recognition? I always reviewed both ways but if one way is enough, I could delete half the cards in my vocabulary decks too to save some time on reviews..

Kanji -> keyword is automatic for like... 90% of the kanji you learn. It will happen a few times that you won't remember the exact keyword but you don't have to worry since Heisigs keyword has little to do with actual Japanese.

For example, as long as you think of "looking" or "seeing" when you see 見 or 観, you're good to go. The specific keywords matter little, if nothing at all, and knowing the specific difference matters even less since Heisigs keywords generally don't show those differences. The mix up between Town and Village for example is a classic.

Since you just started, might I also recommend jumping to Lesson 11 and reading what Heisig has to say there as well about what you remember and such.

Why he waits till then to give those pointers I don't know.

Thanks for the replies.

I guess I'll remove the Kanji to keyword cards then to ease my reviews. I just read lesson 11 as suggested and I don't understand why he didn't say all of that earlier too. Should I do the same with vocabulary?

Lightning Wrote:Thanks for the replies.

I guess I'll remove the Kanji to keyword cards then to ease my reviews. I just read lesson 11 as suggested and I don't understand why he didn't say all of that earlier too. Should I do the same with vocabulary?

It's a bit more complicated with vocabulary I find. I personally do over 90% of my cards just vocab -> meaning (recognition) and never the other way around, but this is because I value recognition a lot more. If you read a word enough, you will generally be able to produce it eventually, but it's not failsafe and unlike kanji -> keyword, production is a needed skill when writing.

I would say, do just recognition with vocabulary at first and see how bad your production becomes. If you can't bear with it, start adding production cards for the most common and important words you feel you need to be able to produce. But I recommend keeping it down, because it's a lot more work to review production cards.

I mostly do it the way Heisig suggests -- keyword to kanji. I find this requires me to be more "active", and more precise -- if I fail to write a stroke, I mark it wrong and review it again. Working this way, I find that recognition (a more passive skill) follows pretty automatically. But I don't think it hurts to go both ways if you have time.