But like I said, there's only a couple of kanji where I have to pay conscious attention to the radicals in order to tell them apart from each other.

RTK entails:
-breaking down a character into all of its component parts
-Associating a keyword with each component part
-Associating a keyword with the kanji as a whole
-Creating a story that associates the keywords of the component parts with the keyword of the entire kanji

You can't consider it to be a "variation of RTK" if I just briefly note to myself that a certain character has a tree radical and another character has a horse radical unless you're stretching the definition so far as to be completely meaningless.

So you're saying that when you look at a character, you don't break the character down? You don't examine the parts of the character? You're starting to contradict the basic rules of human cognition.

Please break down precisely what goes through your head as you look at a character and determine what it means. I would be fascinated to find out.

i also very rarely examine the parts of a character.

You learn the kanji as a whole shape, in the context of a word. Occasionally you do need to briefly register what radical it uses, but you certainly don't need to break the entire character down, it's usually just a matter of noting where it's different from a similar kanji.

I totally don't see how that "contradicts the basic rules of human cognition". Which rules are those, exactly??

Thank you all for the helpful and insightful replies to my question/concern.

Since it seems somewhat pertinent to the discussion, I think I should give a few more details about my Japanese study related background. I studied Japanese through adult classes at a Japanese language school for the past 2.5 years and was in what the school calls an "intermediate" level. I also took and passed the N5 level of the JLPT this past December. So I'm not a complete beginner and I have a basic understanding of the language.

After 2.5 years of study, I found I had little to no retention of the kanji I was given in the classes. The school taught it in the traditional way - brute force rote memorization. They provided no organized, systematic method of learning the kanji. This method did not work for me and the kanji were just random scribbles even after hours of study. This frustrated me to the point of thinking Kanji were impossible and I had no chance of ever learning them. I actually thought about quitting Japanese entirely. Knowing the traditional method wasn't working for me, I set out to find a "better" way and that led me to RTK.

As I said in the original post, RTK is working well for me. I can now recognize and write 1000 kanji (and counting). I feel much more confident and comfortable with kanji now. It almost seems too good to be true which is why I posted my question. I was thinking RTK is great but what's the catch?

Anyway, thanks to all again for the helpful and reassuring replies. I will continue on with RTK knowing it is not a mistake or somehow detrimental to my study of Japanese.

I tend not to think about the meanings of a particular kanji; instead, I tend to think in terms of the readings and meanings of words.

If it's a word I know pretty well, then I tend to recognize it holistically, like in English I would recognize "onomatopoeia" or "miscellaneous" without having to sound them out. If I don't know it pretty well then my first stop tends to be making a guess at the pronunciation of a word -- I know enough of the on and kun readings that I can usually get it on my first or second try. If I don't know the pronunciation then I'll often think of what characters use the same right side radical so I can take an educated stab at the reading.

If an unfamiliar kanji has the tree radical then I usually guess it's a kind of tree, and if it has the bird radical then I usually guess it's a kind of bird, but for the most part -- either I know the kanji by sight or I don't, and if I don't then I don't bother trying to guess the meaning, and I don't worry about the component parts except to help me out with looking it up.

highfade Wrote:Knowing the traditional method wasn't working for me, I set out to find a "better" way and that led me to RTK...As I said in the original post, RTK is working well for me...I was thinking RTK is great but what's the catch?

I've had a very similar experience. Got to about 500 kanji the rote way, forgot them very quickly, then did RTK and it worked almost like magic. Kept waiting for something to *stop* working, but it didn't. As an added bonus, I then stopped reviewing (or doing *anything* in Japanese) for about six months, then came back and found my retention was still about 20%. Started re-reviewing, which in this case meant reading a story only once (not referring to the book at all), and my reviews have been progressing at 90%+ with almost no effort.

RTK has been working very well as a starting point for me, and it sounds like it's been doing just as well for you. That's really all that matters - if you're making progress and it interests you, stick with it. Once you're done (or get bored), find something else that challenges you and keeps you learning.

LittleFishChan Wrote:So you're saying that when you look at a character, you don't break the character down? You don't examine the parts of the character? You're starting to contradict the basic rules of human cognition.

Please break down precisely what goes through your head as you look at a character and determine what it means. I would be fascinated to find out.

wouldn't we all be fascinated? too bad we don't have EKG machines next to us.

for sure i don't break down english words into prefixes into letters and then into their requisite parts of "curly line," "dot," etc. at the very least, for sure i don't do it consciously.

highfade Wrote:As I said in the original post, RTK is working well for me. I can now recognize and write 1000 kanji (and counting). I feel much more confident and comfortable with kanji now. It almost seems too good to be true which is why I posted my question. I was thinking RTK is great but what's the catch?

the catch is that you still have a lot to learn and that RTK is just a drop in the bucket.

the advantage is that you know how to learn a kanji systematically, break it down, find it in a dictionary, re-write it, etc.

even if you know that, you still have to learn how to read the kanji, on yomi or kun yomi or whatever, know the meaning of the word of the kanji, know the word in the context of an entire sentence, and then the context of the sentence with respect to the paragraph, and then the paragraph in respect to the entire document. and that's not even including the writing part, that's just reading.

so while there are definitely advantages to the system, in the whole of things it's still pretty small. keep at it though.

i'm puzzled by people saying that the rote way doesn't work though. i mean, if you stare at enough kanji, you have to see the similarities, right?

kainzero Wrote:the catch is that you still have a lot to learn and that RTK is just a drop in the bucket.

the advantage is that you know how to learn a kanji systematically, break it down, find it in a dictionary, re-write it, etc.

even if you know that, you still have to learn how to read the kanji, on yomi or kun yomi...

so while there are definitely advantages to the system, in the whole of things it's still pretty small. keep at it though.

i'm puzzled by people saying that the rote way doesn't work though. i mean, if you stare at enough kanji, you have to see the similarities, right?

Just to be clear, yes I am completely and totally aware that RTK is just a small part of the overall process and there is still a long, long road ahead after RTK. My original post was to make sure RTK in and of itself is a worthy part of the learning process.

As for the rote method, please let me re-phrase my opinion. I wouldn't say it doesn't work at all but rather it's extremely inefficient and a harder way of learning. It seems the rote method doesn't engage the mind enough so things sink in very slowly and are easy to confuse and/or forget. Of course, I'm speaking just for myself and the rote method didn't work well enough for me.

Thanks for sharing your thoughts and opinions.

kainzero Wrote:i'm puzzled by people saying that the rote way doesn't work though. i mean, if you stare at enough kanji, you have to see the similarities, right?

Yes, if you stare at each new character that you're learning, and compare it to each character that you already have learned, you can see the similarities and differences. This works great when you only know 20 characters, it's a lot more troublesome when you start to know hundreds of characters and have no handle on how to describe the components.

You can study the radicals, you can use RTK, or you can come up with your own naming system, but as soon as you start thinking of any character as 'two trees in the middle of a bunch of squiggly stuff that outlines it' you're done for. If you fail to notice the differences between 又, the right half of 枚 and the bottom of 夏 you won't have any problem with those three characters - but in other characters they may be a significant difference (perhaps not the -only- difference, but the other differences may also be visually similar elements, eg, 米、木、糸、禾 as left hand elements... )

Given differences in fonts and writing styles, without some guidance it's hard to know which 'tiny' differences are important and which are insignificant until you compare two characters where they matter. Having noticed the same similar tiny difference you might try to keep an eye out for it, but there are, after all, many of these tiny differences.

It would be a lot easier to track them if you gave them a mnemonic name, once you identify all the important differences. Maybe names based on the characters that have the same strokes as the elements...

Or you could study RTK and/or radicals and not compare thousands of characters to each other in a search for the important differences that other people have already identified.

(The above, btw, is not hypothetical at all; I came to RTK late, after realizing that I could never hold more than 500-600 characters in my head unless I started approaching the kanji systematically. With a study of radicals and obsolete kanji that just happen to be reused as components, RTK methods wouldn't be necessary, although RTK-order lists would still be helpful.)

SomeCallMeChris Wrote:(The above, btw, is not hypothetical at all; I came to RTK late, after realizing that I could never hold more than 500-600 characters in my head unless I started approaching the kanji systematically. With a study of radicals and obsolete kanji that just happen to be reused as components, RTK methods wouldn't be necessary, although RTK-order lists would still be helpful.)

Sounds very familiar to me.

highfade Wrote:Just to be clear, yes I am completely and totally aware that RTK is just a small part of the overall process and there is still a long, long road ahead after RTK. My original post was to make sure RTK in and of itself is a worthy part of the learning process.

well, my experience with the rote method from textbooks is that they try to teach you everything at once. they teach you the stroke order, then different compounds and words and have you memorize them. but not all the words either, maybe just 2 or 3, and those vary in commonality. technically you do learn more practicality with the rote method. that's why i think RTK is really just rearranging the process so that you get the stroke order and recognition out of the way first before moving on to the other things.

SomeCallMeChris Wrote:You can study the radicals, you can use RTK, or you can come up with your own naming system, but as soon as you start thinking of any character as 'two trees in the middle of a bunch of squiggly stuff that outlines it' you're done for.

haha, that reminds me of when my friend and i were trying to find a place called teppanyaki ranma in roppongi and it was written in kanji. i described it to him in terms of AKB member names. "ran" as in "yamauchi suzuran" and "ma" like "shinoda mariko." (it was closed when we got there. twice. bastards.)

btw, i'm not necessarily trying to argue against RTK, i'm just wondering if it can be improved on. i did it and i can never take it back. i'm just curious to see if there are things that work and didn't work with it and if other things can be improved. like i 100% believe that the keyword is useless in all applications.

kainzero Wrote:like i 100% believe that the keyword is useless in all applications.

No, the keyword serves a very important purpose of uniquely naming the character so you can keep it mentally separate from the other characters with similar meanings. Otherwise you end up trying to think about 'the character that means warm more than hot, and the other character that means warm more than hot, not to be confused with either of the two characters that mean hot more than warm, but all of them mean warm and/or hot.'

Of course, if you already know a substantial amount of Japanese ahead of time, you can and probably should use Japanese keywords rather than English ones, as this is how you'll eventually end up thinking of the characters and keeping them mentally separate.

SomeCallMeChris Wrote:

kainzero Wrote:like i 100% believe that the keyword is useless in all applications.

No, the keyword serves a very important purpose of uniquely naming the character so you can keep it mentally separate from the other characters with similar meanings. Otherwise you end up trying to think about 'the character that means warm more than hot, and the other character that means warm more than hot, not to be confused with either of the two characters that mean hot more than warm, but all of them mean warm and/or hot.'

this seems somewhat like arguing that you're going to have trouble distinguishing between words with similar definitions. Each kanji has a certain feeling, which comes from the contexts you see them in, not a label stuck on them outside of all usage.

As usual, different stuff works for different people i guess... there's nothing inherantly wrong with using RTK, but there's nothing inherantly wrong in not using it either. Use whichever thing motivates you best, since whatever way that is will get you further.

IceCream Wrote:this seems somewhat like arguing that you're going to have trouble distinguishing between words with similar definitions. Each kanji has a certain feeling, which comes from the contexts you see them in, not a label stuck on them outside of all usage.

Well, that's truish (though perhaps it relies on a better visual memory than most people have), but there are essentially two ways to get up to speed with the kanji -
- learn the first grade kanji, and read first grade books
- learn the second grade kanji and read second grade books
ad nauseum (these may be 'graded readers' for foreign language learners steps rather than Jouyou school grades, either way.)

or

Somehow learn roughly 2000 characters (and be ready to learn more on the fly because there's no guarantee you have all that you need) and read light novels / novels / newspapers / whatever adult material you're interested in.

If you want to do the second, you need a handle on the kanji other than the context that you cannot yet read them in until you first have a handle on 2000+ kanji.

Or read and learn them as they come.

when i think 熱い vs 暑い i just think of different contexts. and i won't be mistaking 情暑 for 情熱 any time soon because one of them is a word, and one of them isn't. at that point there is no real "japanese" keyword but an actual vocabulary word.

(when i looked up the keywords, 熱 is "heat" and 暑 is sultry, and "sultry" in itself has more than one meaning. then there's 暑中, which i haven't seen before but i always hear 真夏 so... yeah.)

that's why i think the keyword isn't really too important, as a placeholder i'm convinced it's not useful and i'm hypothesizing that it's probably better to just recognize stroke order and the radicals and know of their existence before beginning vocabulary.

again, this is just me guessing, and i will never have the resources to compare having the "sultry" placeholder to learn 暑い vs. just knowing the shapes of 日 and 者 and proverbially gluing them together.

i'm finding that as i improve my reading, the definition of a word doesn't become what i see in the dictionary, but shaped through the many contexts in which i see a word and the "feeling" i get from it. it's like what IceCream was referring to, but my experience is more vocab than kanji... but i can see how it can easily carry across.

SomeCallMeChris Wrote:Well, that's truish (though perhaps it relies on a better visual memory than most people have), but there are essentially two ways to get up to speed with the kanji -
- learn the first grade kanji, and read first grade books
- learn the second grade kanji and read second grade books
ad nauseum (these may be 'graded readers' for foreign language learners steps rather than Jouyou school grades, either way.)

or

Somehow learn roughly 2000 characters (and be ready to learn more on the fly because there's no guarantee you have all that you need) and read light novels / novels / newspapers / whatever adult material you're interested in.

in my experience most of the students who complain about kanji usually don't know the corresponding vocabulary of that kanji anyway! whether it's 国会議員 or こっかいぎいん、 they still won't understand it. they can even download a visual novel where the words are read to them and includes displayed text that has furigana and they won't understand it, kanji or not, because they just don't have the ability.

in the case of a learner who is illiterate but a fluent speaker, i would think it's just as easy as a vocab anki deck since they just need to bridge the kanji with its reading and the appropriate vocab in their head.

SomeCallMeChris Wrote:

IceCream Wrote:this seems somewhat like arguing that you're going to have trouble distinguishing between words with similar definitions. Each kanji has a certain feeling, which comes from the contexts you see them in, not a label stuck on them outside of all usage.

Well, that's truish (though perhaps it relies on a better visual memory than most people have), but there are essentially two ways to get up to speed with the kanji -
- learn the first grade kanji, and read first grade books
- learn the second grade kanji and read second grade books
ad nauseum (these may be 'graded readers' for foreign language learners steps rather than Jouyou school grades, either way.)

or

Somehow learn roughly 2000 characters (and be ready to learn more on the fly because there's no guarantee you have all that you need) and read light novels / novels / newspapers / whatever adult material you're interested in.

If you want to do the second, you need a handle on the kanji other than the context that you cannot yet read them in until you first have a handle on 2000+ kanji.

well, the primary context of kanji is words, rather than phrases or sentences, so it's enough to learn them in the context of words, i think.

So, taking the 2nd option, you simply go through the kanji one by one (in an order, but can be any order) learning each in a number of high frequency compounds. Then, when you do go on to read novels, etc, you not only know the kanji, but also the readings, and the meaning of the words you encounter. If you're used to using anki, you can do that in a comparable time to the time it takes to do RTK.

It doesn't help you write them though, of course...

EDIT: to show what i mean, think about learning a number of words from this list at the same time:
http://jisho.org/words?jap=*%E7%86%B1*&e...&common=on
and then:
http://jisho.org/words?jap=*%E6%9A%91*&e...&common=on

it's enough to demarcate the type of heat each refers to in your head and start forming that feeling associated with the kanji that means you can predict reasonably accurately what a new word's about...

LittleFishChan Wrote:When a Japanese person sees a kanji, two things come to mind: a) what is means and b) how it is read. That is why my Japanese friends say that when they see something in Chinese they cannot read it, but they can get the gist of what it says.

People claim this, but it's not true. It might work for some very specific, small snippets of Chinese, but in general, a Japanese speaker can't even get the gist of Chinese writing.

yudantaiteki Wrote:

LittleFishChan Wrote:When a Japanese person sees a kanji, two things come to mind: a) what is means and b) how it is read. That is why my Japanese friends say that when they see something in Chinese they cannot read it, but they can get the gist of what it says.

People claim this, but it's not true. It might work for some very specific, small snippets of Chinese, but in general, a Japanese speaker can't even get the gist of Chinese writing.

I always thought it would have worked both ways... At least with traditional characters.

IceCream Wrote:So, taking the 2nd option, you simply go through the kanji one by one (in an order, but can be any order) learning each in a number of high frequency compounds...

Perhaps that would work for someone with an excellent visual memory who can simply memorize complex shapes, but pure recognition drilling absolutely did not work for me. Then again, I was using paper flashcards or 'dumb' flashcard programs up until a year ago when I discovered anki, so, with SRS who knows, but I don't think SRS would prevent the kinds of mistakes I used to make.

Of course, with dumb flashcards or SRS or whatever, I simply -cannot- learn words without context, so for me at least, 'learn this list of words to give the kanji context' is no good at all. The words might give the kanji a context, but I need a context for the word or it will not stick. My fail rate on 'plain' vocabulary cards is obscenely high (I don't create them any more, of course, and am editing my existing ones when I fail them.)

Also, I find learning multiple things simultaneously is a bad idea - this became obvious from the blind-stupid approach that I first approached the kanji with. Learn the kanji, it's meaning(s), it's reading(s) and the common vocabulary words it appears in.... and drill that by recognition of the kanji character. Obviously stupid, and yet that's what literally hundreds of flashcard applications encourage and what I tried. Anyway, extrapolating from this - 'learning multiple things at once is harder than learning one thing at a time', I always now learn vocabulary phonetically first, in a context with words I already know, and then learn the kanji for it (which is simple because the vast majority I have an understanding of from RTK if not other vocabulary yet, and the remainder I can quickly build a mnemonic for.) Anyhow, learning an arbitrary list of words is something I consider a painful waste of time, perhaps because of the effort that I put into each word that I go through the effort to explicitly learn (as opposed to those I simply pick up in context.)

For the characters, I really think some kind of component breakdown is necessary (and in fact, before RTK I was haphazardly doing that, half-consciously coming up with nicknames for components but then not being consistent with them because I wasn't writing down any notes on my components and it was chance if I remembered the nickname I had given a component. It wasn't a -process- just idle thoughts as I stared at flashcards), and I really believe some kind of pigeonholing is necessary. Keywords serve that role nicely; English keywords are okay; kun-reading vocabulary is ideal, but is incomplete because of multiple-kanji for the same pronunciation and words that don't -have- a kun-reading. So 暑 is あつい to me and 熱 is ねつ and 陽　is たいよう　の　よう, and characters that I don't have (compelling) vocabulary for still have their heisig keywords in my mind.

These posts keep getting so long, but I feel like it's really a very simple thing, or a very simple few things at any rate. I'm probably expressing myself poorly. Anyway, I'm not trying to say that RTK is the only way, just that it has some powerful points - providing a mental 'handle' for the character, learning 'one thing' instead of everything, component-analysis to simplify the task of remembering and reduce confusion, and of course, learning to draw the character which aside from enabling writing makes it much easier to recognize a character regardless of the font it is printed in (although one does have to learn to recognize variants for some components, but if you think of them already as components and not arbitrary strokes, that's easy to comprehend.)

blackbrich Wrote:

yudantaiteki Wrote:

LittleFishChan Wrote:When a Japanese person sees a kanji, two things come to mind: a) what is means and b) how it is read. That is why my Japanese friends say that when they see something in Chinese they cannot read it, but they can get the gist of what it says.

People claim this, but it's not true. It might work for some very specific, small snippets of Chinese, but in general, a Japanese speaker can't even get the gist of Chinese writing.

I always thought it would have worked both ways... At least with traditional characters.

No, it doesn't really work either way, unless you have a really, really liberal definition of "gist". You might be able to pick out a few nouns and verbs, but you wouldn't know which ones were negative or positive, what the subjects and objects are, and such. Of course a Chinese person might understand the characters on a Japanese sign, or something very short and with clear context like that. But if a Chinese person picks up a Japanese book or newspaper (or vice versa), their comprehension will be essentially zero.

For reading (though not necessarily for writing) there's another alternative to rote drilling and RTK-like methods: to just learn kanji in the context of sentences, in the context of vocabulary that you already know. Read a lot with furigana, and you can start with the meaning (of the word as a whole, not of the individual kanji) and pronunciation, and have some passive exposure to the kanji until you get used to them sufficiently that you can ditch the furigana. Worked for me.

By the way, for a long time I neglected my kanji writing, but now I'm studying for the kanji kentei. (Admittedly it's just level 5). I find that I've read so much that I only have to do a couple of repetitions of a kanji before I get it pretty solidly in my head, so -- well, if it's rote drilling, it's an extremely painless kind of rote drilling.

Chinese and Japanese people will tell you they can get the gist, but they can't. If you actually give them some Japanese or Chinese and ask them to tell you what it means, they won't be able to (in most cases). What they mean is that they can spot words they know, but that's not the same thing.

yudantaiteki Wrote:Chinese and Japanese people will tell you they can get the gist, but they can't. If you actually give them some Japanese or Chinese and ask them to tell you what it means, they won't be able to (in most cases). What they mean is that they can spot words they know, but that's not the same thing.

Yep. Exactly correct, IME.

なるほど. Makes sense.