I talked to a Self Learning student (who claims to be at N1) and he told me this. Is this even possible? Or its an ridiculous exaggeration?

(Paraphrasing)
"I took Heisig's advice to the heart and adopted the divide and conquer strategy. I finished RTK, and then completed one of those 1000 words or so vocab decks. I ditched readings and listening all together after that, just learning maximum vocabulary a day using my Kanji mnemonics (I also added New Kanji meanings since Heisig's keywords stopped working after a while). I did grammar on side too. In 7 months, from total noob, I had done over 10 ,000 words without readings, and about N3 - N2 grammar. Then probably for 4-5 months (busy because of family), I just reviewed the anki decks, read articles (like nhk easy) , easy Visual Novels , Stories (adding unknown words to Anki). After that, I only paid attention to readings while watching anime with subtitles, visual novels and making a mental note (no Anki), and they came automatically to me. "

It's possible. If you learn words, you are also learning the readings of the kanji. Also you know that you are actually learning readings that are used, rather than rare ones that somehow made it onto a list or into a dictionary.

Regardless of what strategy you take, you should never try to learn a kanji reading in isolation. Always have at least one word to go with that reading. Learning isolated readings doesn't help you know which readings go with which words.

Sure is.
I have no idea why you think it would be impossible. Learning words in Japanese is just like learning words in any other language.

I assumed he meant learning the words without learning how they're pronounced at all (beyond the first 1000). Which would be seriously weird.

You can learn the words while being unable to pronounce them. In the end, ideograms are just pictures to which you assign meaning, so you can memorize them as "pictures" rather than "words" and it doesn't change the end result. This has the disadvantage that a simple font change can render a word unrecognizable, but nothing prevents you from "relearning" a word with a different typeface. It's not efficient, but you can do it nevertheless.

Vempele Wrote:I assumed he meant learning the words without learning how they're pronounced at all (beyond the first 1000). Which would be seriously weird.

This is how I read it too and I agree. I wouldn't recommend this to anyone because you won't be able to pass the N2/N1. The speaking section doesn't have subtitles (lol) and there are sections in the reading part that actually check that you know the proper readings for words (if I recall correctly).

Honestly I wouldn't call this "learning vocabulary" I'd call it "Learning Kanji compound meanings without knowing the readings."

If you meant "doesn't know how to pronounce the words at all":
Possible, but impractical and inefficient. You can memorize just about anything...

If you meant "didn't learn the readings of each kanji separately from vocabulary study":
Easy to do, efficient, practical.
You pick the readings and get a feeling for how common they are (for example: 発 is pretty much always はつ, but it does have other readings... that are almost never used).
I had no trouble not studying kanji readings separately. And as long as it must take, I'd suggest no one do it, unless they're going to take kanken or something (I think you need that knowledge for that test?).

sholum Wrote:If you meant "doesn't know how to pronounce the words at all":
Possible, but impractical and inefficient. You can memorize just about anything...

If you meant "didn't learn the readings of each kanji separately from vocabulary study":
Easy to do, efficient, practical.
You pick the readings and get a feeling for how common they are (for example: 発 is pretty much always はつ, but it does have other readings... that are almost never used).
I had no trouble not studying kanji readings separately. And as long as it must take, I'd suggest no one do it, unless they're going to take kanken or something (I think you need that knowledge for that test?).

There is no need to study readings separately for kanken, since you're always going to be writing out words. So to take your 発 as an example, sure you'd be expected to know it can also be pronounced ほつ, but the only way they're going to test that knowledge is by asking you to write/read 発作 for instance, given in full context. So just learning the readings and not the words that come with them has pretty much no use, since even if you know ほつ is a potential reading, you wouldn't know when to apply it.

manman2a Wrote:just learning maximum vocabulary a day using my Kanji mnemonics

What does that mean? Something like looking at the word 一生懸命 and thinking, oh, "one" + "born" + "hang" + "life" [*], it means "with all one's might" - but without being able to recognize it when written in kana, without recognizing the word when pronounced, without mentally saying いっしょうけんめい when looking at those kanji?

It's possible, I guess. But it sounds a bit ridiculous. Every writing system (including kanji) was created to represent a spoken language.

[*] Just guessing, never done RTK so I don't know what the Heisig keywords are.

Oh wait, did I misunderstand? Kanji are absolutely *not* ideograms or "pictures" and should never be learned that way. If you mean learning words without learning how they are pronounced, definitely not. Terrible idea.

One problem is that it's essentially impossible for most people to take written symbols and go directly to some abstract meaning without any association with language. If you don't know the Japanese pronunciation, you will be associating it with sounds in your native language. Reading will be switching back and forth between Japanese and your native language, and you can't become a fast reader that way.

Taishi Wrote:There is no need to study readings separately for kanken, since you're always going to be writing out words. So to take your 発 as an example, sure you'd be expected to know it can also be pronounced ほつ, but the only way they're going to test that knowledge is by asking you to write/read 発作 for instance, given in full context. So just learning the readings and not the words that come with them has pretty much no use, since even if you know ほつ is a potential reading, you wouldn't know when to apply it.

I see. I wasn't sure how they tested kanji knowledge for kanken.
I guess that drives the point further; there's no reason to study kanji readings separately from vocabulary.

I guess if one's Japanese class requires you to recall readings... but even then it'd be easier to learn those readings by studying words that use them.

manman2a Wrote:In 7 months, from total noob, I had done over 10 ,000 words without readings, and about N3 - N2 grammar.

Has this friend also won the Grand Prix and the Golf World Cup?

I'll have to raise my game - sounds like he's catching up.

I went the direct opposite way and learned readings (dominant onyomi) before building a large vocabulary. My thinking at the time was if I'm going to make my own stories to learn the kanji (something I decided I would do), I might as well integrate the sound of the onyomi into the story and that way I'd kind of get that knowledge for free in a way. I was already very familiar with how to sound out the Japanese syllabary, and honestly this method did sort of work. I was reading way above my level thanks to it, though I'm not sure how helpful that was in the end. I can't go back and undo it and then just learn vocabulary instead.

Doing it this way lead to some interesting exchanges when I did reading practice with my tutor at the time. She was always impressed with my reading speed and how fluent I sounded when reading, but then she'd ask me to explain what I just read and on some occasions I had no idea whatsoever. I suppose it wouldn't be unlike an English speaker reading a whole load of words they could pronounce fine but of which they didn't know the meanings.

I can't even say whether it made it easier for me to learn vocabulary or not, again because I have nothing to compare it to, I just know for sure it gave me a premature boost to my reading ability and made things far above my level always feel somewhat in reach instead of completely alien.

Roketzu Wrote:I went the direct opposite way and learned readings (dominant onyomi) before building a large vocabulary.

I also did this for a while. Memorizing readings is not as hard as it sounds, as most characters (as much as 90%, if Wikipedia is anything to go by) have a phonetic component that's there to indicate the onyomi (青, 精, 清, 晴, 情, all read セイ). Too bad that not many learners seem to use this. (I know that this is covered by RTK2, but that book doesn't seem to be very popular.)

fzort Wrote:

Roketzu Wrote:I went the direct opposite way and learned readings (dominant onyomi) before building a large vocabulary.

I also did this for a while. Memorizing readings is not as hard as it sounds, as most characters (as much as 90%, if Wikipedia is anything to go by) have a phonetic component that's there to indicate the onyomi (青, 精, 清, 晴, 情, all read セイ). Too bad that not many learners seem to use this. (I know that this is covered by RTK2, but that book doesn't seem to be very popular.)

It's pretty funny that you said you never did RTK, but you basically did RTK2 without realising it.

ktcgx Wrote:It's pretty funny that you said you never did RTK, but you basically did RTK2 without realising it.

You mean me? I did say that and maybe I did just do a variation of RTK2 (honestly not sure what it is exactly), but it's because I was copying someone else who was maybe influenced by RTK2. I just changed the order I did them in and made all of my own stories. The person I copied didn't go as far as that and actually just stopped studying not long after starting.

fzort Wrote:

Roketzu Wrote:I went the direct opposite way and learned readings (dominant onyomi) before building a large vocabulary.

I also did this for a while. Memorizing readings is not as hard as it sounds, as most characters (as much as 90%, if Wikipedia is anything to go by) have a phonetic component that's there to indicate the onyomi (青, 精, 清, 晴, 情, all read セイ). Too bad that not many learners seem to use this. (I know that this is covered by RTK2, but that book doesn't seem to be very popular.)

I think most learners realize this as they learn the kanji -- I never used RTK2 or anything like it but I figured out that certain components signalled pronunciation. You have to be careful with this because sometimes those readings aren't very common -- for instance, 情 is more commonly read as じょう than せい.

That's very interesting. I will end with RTK1 soon and I would like to continue with vocabulary and grammar, because RTK2 is not that popular (I have it bought though). But, I think that part of RTK2 might be worth doing, like those common signalling primitives you are mentioning? If someone went this way and has useful advice how to tackle it, I would appreciate it. Thanks, and sorry for bit of off topic.

[Katsuo]

Taishi Wrote:There is no need to study readings separately for kanken, since you're always going to be writing out words. So to take your 発 as an example, sure you'd be expected to know it can also be pronounced ほつ, but the only way they're going to test that knowledge is by asking you to write/read 発作 for instance, given in full context. So just learning the readings and not the words that come with them has pretty much no use, since even if you know ほつ is a potential reading, you wouldn't know when to apply it.

When I took various levels of the KanKen, one method I sometimes used was to learn the words with rare readings and then assume any unknown words that came up had the common reading.

For example in the case of 発, check a list of, say, the 100 most common compounds and note that 95 are read hatsu and 5 hotsu then focus on the latter group. Of those five I knew four already, so just learn the other one. Then I could be fairly confident that any unknown word with 発 on the test would be read hatsu.

So is it better to learn how to pronounce kanji while I learn them from RTK? Because that's not what Heisig said at all. I had already learnt 100 kanji or so, but without pronunciation, and now I read it was a mistake...

I'm already learning a bit of grammar while I learn new kanji, and I'm not so good at multitasking. So you're saying it would be better to learn readings of kanji after all? Even if I care only about translating japanese directly into english?

[Zgarbas]

Possible? Yep; just look at Chinese learners.
Preferable to traditional methods or beneficial in any way? Not so much.

Zgarbas Wrote:Possible? Yep; just look at Chinese learners.

They still know readings for the kanji, just not the Japanese ones. (I'm not just nitpicking; having any semantic value to associate with each character makes it very different from trying to use an abstract meaning.)

yamineko Wrote:So is it better to learn how to pronounce kanji while I learn them from RTK? Because that's not what Heisig said at all. I had already learnt 100 kanji or so, but without pronunciation, and now I read it was a mistake...

I'm already learning a bit of grammar while I learn new kanji, and I'm not so good at multitasking. So you're saying it would be better to learn readings of kanji after all? Even if I care only about translating japanese directly into english?

No, what is being discussed here is learning to translate words without learning how to pronounce them. For Japanese learners, RTK is useful in that it allows you to learn to distinguish kanji in a familiar environment by using keywords (I don't recommend practicing how to write them as a beginner). While many of those keywords relate to a common meaning for that kanji, you can not successfully translate Japanese by only using RTK keywords. As you continue to study Japanese, you will forget the keywords, so don't waste time by being a perfectionist and going hardcore on RTK reviews.

When RTK was written (and even now in college Japanese classes), you're introduced to kanji and are immediately asked to learn all their readings at the same time; that is what Heisig is advising against. We are suggesting that you don't have to study readings separately from vocabulary, but that you do need to know how to pronounce the words you're learning (think of what it'd be like to look at 'learning', but not know how it's pronounced at all).

You will pick up the readings of each kanji as you study vocabulary. If you wish to study words based around the pronunciation of a kanji in them, that's fine, but the point is that you're studying a written word with its pronunciation; studying readings in isolation is fairly useless and doesn't give you any context to work with.

You can't translate directly from Japanese text into English (or if you can, it'd be very difficult to learn and do); to translate (well), you need to know what's being conveyed in the source language, which requires knowledge of vocabulary and grammar, along with cultural knowledge (idioms, imagery, etc).

yamineko Wrote:Even if I care only about translating japanese directly into english?

Do you actually mean you want to translate? Or do you just mean you want to read and understand something?

Read and understand.
Japan have so many awesome (and untranslated) fiction that I decided to learn it as opposed to wait, who knows how long, for translations.