I am a big believer in chunking the stuff down to 1-to-1 associations for learning. So RTK makes a lot of sense with the english-to-kanji association.

However, the kanji also typically requires learning at least 2 sounds for it, the kunyomi and omyomi.

This seems to require a whole 'nother step for learning the readings. So it seems that RTK learners compensate by focusing on sentence mining and context-based learning? Perhaps some will doggedly go through the kanji list again and learn the most common kunyomi/omyomi for each?

Wouldn't it be more efficient to focus on learning lots of hiragana words first, then learn the kanji from the hiragana, thus getting both the meaning and readings for free?

For example:

First learn hiragana words like: あたらしい -> new

Then learn the kanji and reading at once: 新しい -> あたらしい

This is almost similar to how Japanese pick up kanji, by learning the hiragana/sounds first and learning the kanji from school & furigana. However, there's nothing stopping us from using Heisig's simple-to-complex order & radical system with hiragana keywords.

What are your thoughts?

I reckon it's probably best to learn the kanji from within compound words like the above example, rather than on it's own. Or perhaps that would make it too easy?

You have fundamentally misunderstood the purpose of RTK. You should re-read the intro chapters again.

Short form: It's just a mnemonic to help learn the kanji.

Yes, I am talking about english-keyword to kanji. Part mnemonic, part meaning (sometimes)..

scuda Wrote:First learn hiragana words like: あたらしい -> new

Then learn the kanji and reading at once: 新しい -> あたらしい


I reckon it's probably best to learn the kanji from within compound words like the above example, rather than on it's own. Or perhaps that would make it too easy?

As for the first part of what you said, I often will learn a word and WHICH kanji to use at the same time, but I don't have to sit there memorizing a new kanji in ADDITION to a sound at once. You are still doing 2 separate pieces of work, the reading and the kanji itself. RTK just makes the kanji portion efficient and easier.

Saying that it would be "too easy" to learn a kanji from within a compound rather than the way some of are doing is flat out insulting to me. I picked up this method because learning kanji individually from within words was too inefficient, and I sucked at it. This made it MUCH easier. I don't get at what you're trying to say about us.

I meant "too easy" in the sense of being able to remember "新しい" == "あたらしい" in spite of the kanji, i.e. if the "しい" part could provide too much of a clue. But that entirely depends on the size of the deck and the items in it.

For example, in smart.fm or similar systems, sometimes there's clues that make it too obvious. It probably doesn't matter in the long run though. Exposure is exposure.

I'm looking at the other posts out there about japanese keywords for RTK, so that might be what I want.

You do raise a good point about having to learn the kanji anyways, and that the additional reading from the hiragana keyword might not get picked up..

I see what you are saying in wanting to emulate the process of how Japanese children learn kanji. The problem is that Japanese children already know how to speak Japanese, and are simply learning the hiragana, then the kanji that matches it. For English speakers, having to learn the word sound, hiragana and kanji for each new word is a lot of work. I think that's the whole reason Heisig wrote his book.

Ultimately, if your method detailed above works for you, go for it. But there's a reason there's a whole website devoted to the RTK method. Heisig's method might seem roundabout, but it is actually quite efficient for English speakers.

I don't think learning the english keywords is harmful or requires any extra effort or anything. Think about it. When you learn a word with hiragana, you have to learn the meaning of that word in English, right? And if you skip the kanji for the word to begin with, then you are having to memorize a sequence of seemingly random kana for every word.

If you learn the English keywords, then what happens is whenever you encounter some words that you might not know, you can sometimes (quite often in fact) sort of figure out a rough meaning of the word. You still don't know the reading, and you don't even know if your guess at the meaning of the word is correct, but it can really help you understand a sentence if you don't have a dictionary or something handy.

scuda Wrote:I'm looking at the other posts out there about japanese keywords for RTK, so that might be what I want.

You do raise a good point about having to learn the kanji anyways, and that the additional reading from the hiragana keyword might not get picked up..

I've got no problem using Japanese keywords for rtk, and if you already know some Japanese I highly recommend using them instead. The problem is someone that knows too little japanese, there is more to learn per card, so instead of dividing the learning of how to write the kanji, the student has to memorize the particular japanese keyword first, to know WHAT that word is referring to THEN tie that to the japanese kanji that is also being learnt.

I changed many of my keywords to Japanese keywords, but if I didn't know a Japanese keyword, and the kanji was super easy to memorize with the given keyword, I didn't mess with it, it's fast and efficient

Another pro for Japanese keywords is making them less ambiguous, such as the town/village characters, similar in english meaning so easy to mix up based on english keywords. I've never mixed up まち　to　町 but in rtk HAVE responded to both town and village as 町。

Sorry for being more insulted by the 新しい comment, I took it the wrong way, perhaps you can see why, but thanks for clearing that up

Quote:Wouldn't it be more efficient to focus on learning lots of hiragana words first, then learn the kanji from the hiragana, thus getting both the meaning and readings for free?

No. Picking up a reading or 2 by learning words is going to be a lot easier than "picking up" a 22-stroke kanji that's in a word you know.

I wanted to clarify something, there seems to be a bit of misunderstanding.

The idea is to learn all of the necessary hiragana words FIRST. Then learn the kanji using japanese keywords that give you the meaning and 1 of the readings. Using RTK order for the kanji most likely (simple to complex)

Basically, it would be a 2 step process, not an attempt to learn both kanji & japanese words at the same time.

The RTK way seems more like a 3-step process in comparison?

Here's an example:

RTK:
1) english word => kanji
2) hiragana word => english word
3) kanji => readings/japanese word

Alternative method:
1) hiragana word => english word
2) kanji => hiragana word

Sorry if what I'm saying kind of flies in the face of what RTK is meant for, I'm just curious if this alternative path could actually cut out an extra step of learning in comparison..
Thanks for your responses!

What it comes down to is for a lot of people, me included, RTK is efficient. If I could remember the reading, meaning and Kanji all at the same time and it was still efficient then I would do it. Unfortunately, when I was starting out, and perhaps now, this was impossible. Hence, RTK made it possible for me to speed through learning Kanji, which in turn made Kanji filled sentence and reading accessible. The reality is 99% of us want to get fluent fast, so choose the fastest method. I'm not keen to wait the extra time to get good! (Incidentally, I did the lazy Kanji method from Kendo. It's faster, but not as deep in terms of understanding as the pure RTK method.)

Cranks,

That's my goal too. To get fluent fast. I'm talking about removing an extra step that I see in the standard RTK process.

Well you might be the first to try it that way.
Go ahead and do it and let us know how it works out for you.

I imagine that if you don't know the hiragana words to begin with, it might be difficult to choose good ones for each kanji.

What Heisig does is giving you one keyword in English for one Kanji. Since you already know English, you are Canadian, you don't have to memorize the word. The keyword is there as connection between keyword -> story -> kanji.

So the way is more like: (k) keyword -> (l) story -> (m) kanji -> (w)rite the kanji

You (k)now the keyword, you are (l)earning the story -> to (m)emorize the kanji -> to being able to (w)rite it from memory. You don't have to memorize dozens upon dozens of stroke orders either. With Heisig method this comes naturally for he is breaking down kanji in smaller parts. And because you have to write them so often, you will be so familiar, even with kanji with many strokes, that you don't even have to think about that.

Once you jump into Japanese learning, there is still the thing with having to learn ひらがな words, but with a difference. You don't have to stare at kanji trying to see a difference between similar looking ones. 靴, 鞄 for instance. You can connect ひらがな to 漢字 and connect that to the English meaning. To take your example 新しい　(あたらしい) new. Since you know how to write 新 and that it is part of 新しい you can write both without any effort on your side.

If you are going to do Heisig the way -> memorize 2042 Hiragana words, the English word for the Hiragana -> memorize the Story -> memorize the kanji you could go rote memorization. Then you can learn all things at once -> English word -> Hiragana -> Kanji -> write it over and over and over again until it sticks, learn the On reading -> Kun reading -> not to forget about Grammar, pronunciation, have fun.

I hope it is clear to you now that you would just making the whole process more complex. You could however use japanese keywords for those kanji you knew the Hiragana and English word already. That is what some are doing while working through the book. See the story section for examples.

But the thing is, even if you know 2042 words in Hiragana from which you then learn the kanji, there is still one thing you have to consider. If you see the word mould, which is frame 680 型, which Hiragana word would you learn for that? The one part that makes up this kanji かた? Or just カビ which has no kanji but means mould? Or would you go for かたわく which would give you the whole word? 型枠 mould? Even for the English word it was not easy to find the fitting definition for that word. Is it mould as in "to form," or more likely "mould as in bread that moulds."

With Heisig you don't have to worry about that, work through the book, and learn the other things after that without the need to learn several things at once.

Edit: If you want to have it that way, you could try to fetch a copy of Kanji isn't that hard! This book too works with mnemonics in a way, gives you stroke order, a text to memorize, a reading passage both english and japanese and some reading on / kun and a picture pointing out what is what. But it is so dirt dry to read that in fact you are still better off with RTK. But, who knows, it might be what you really are looking for.

Example:
学　「子」は子供です。「ヮ」は帽子です。「ツ」は飾りです。「学」の形を見てください。子供が、帽子をかぶります。その帽子に飾りがあります。帽子の飾りは、学校のしるしです。子供は、その帽子をかぶって、学校へ行きます。「学」の意味は「勉強する」です。

子is the kanji for "child." ヮ is a cap. ツ is a decoration. Look at the outline of the kanji 学. A child is wearing a cap. The cap has a decoration. The decoration on the cap is a school logo. The child is going to school, wearing the cap. The kanji 学 means "to study."

Learning a huge number of words first just from sound or hiragana may be how Japanese children pick up the language, but that doesn't necessarily make it the best system for foreigners learning the language. It does after all take about 12-18 years (depending on reading ability) before they can read something aimed at an adult audience. I think most Japanese learners would like to progress a bit more quickly than that.

When I first started learning Japanese I just learned Kanji as they came up in the textbook, and then forgot them pretty soon after. I progressed a lot quicker when I started trying to learn kanji with flash cards, and then learnt new words that could be made with those kanji at the same time. I never tried using RTK to learn just the meaning without any of the readings.

In summary of my experiences-
Learning lots of words/hiragana first then the kanji used to make them = v. slow progress

Learning new kanji (with english meaning) and then at the same time a couple of words it can be used to make = much quicker progress

Having not tried the RTK method I can't really compare how quickly it would work compared to what was successful for me. I think the key to both methods though is the early emphasis on learning as many kanji as possible early on. By doing this I think it makes it easier to remember words and also to read more interesting materials.

Knowing the kanji that make up words help you remember them.
Especially homophones.

edit: disclaimer: imo. ymmv. etc.

scuda Wrote:RTK:
1) english word => kanji
2) hiragana word => english word
3) kanji => readings/japanese word

Alternative method:
1) hiragana word => english word
2) kanji => hiragana word

I think there's a misunderstanding here. I know I don't learn any 'hiragana words', I learn words. They often use kanji. Going by your reasoning, I'd describe the process like:

RTK:
1) english word => kanji
2) japanese word, with or without kanji => english word (or monolingual definition)

And that's it.

Considering vocab - especially compounds - is easier to learn with kanji, and that you will also want to learn when it is appropriate to use kanji for a given word or not, you might be doing yourself a disservice in terms of overall efficiency by doing hiragana vocab first.

tokyostyle,

That's true, since so many words would be identical if I was learning 2,000-3,000 words from hiragana only, I would need some way of providing context. Kanji could be one way, or the questions could make the context clear.

astenda,

In your example of doing something like "japanese word, with or without kanji => english word", what happens when it's a kanji-based word? Then you won't know the reading of it. Unless you include it as part of the question or answer.

But then it won't be "i+1", there'll be 2 bits of information to learn associate each time. So even if it's 2 "steps" it's still 3 bits.


But yeah.. the hiragana vocab approach would probably be at a disadvantage unless there was a good way to provide context.

I think you guys have convinced me. I didn't realize how difficult going from hiragana to kanji would be. Can't think of any way to provide context besides the actual kanji themselves.

However, there doesn't seem to be a good incremental/"i+1" way of associating the reading and meaning to a japanese word that contains the kanji.

For example, with "i+1" we would have separate flash cards for each of the two situations:
1) 新しい <=> NEW
2) 新しい <=> あたらしい

That seems a bit repetitive.

Sorry, I was being a bit blunt. I don't think you'll get much closer to i+1 than doing kanji word -> reading+meaning, though. As for your suggestion with separate flash cards, all you really need to do is to employ multiple card models; this way you can have several cards for the same 'fact'. No need to enter it twice. IIRC, Anki uses some sort of fancy scheduling to keep these 'variant' cards apart. It might be a good option.

Keep in mind though that some kanji or combinations thereof can have different readings with different meanings/nuances associated to each of them, thus in effect being different words that look exactly the same. Might want to have a backup method for that.