I finished 3/4 of Heisig using this site, but choked and died on reviews. I stopped using Heisig and moved on. Anyways, many people learn Japanese just fine without it, right?

But, as I found out, Heisig did help me trememndously. When I encountered kau (to buy) for the first time, I learned it extremely easily. When I encountered a Kanji that I hadn't learned through Heisig, it was indeed a bunch of scribbles. (I'd give an example, but I can't since I don't remember any.)

I'm starting to doubt the need to thoroughly drill Heisig Kanji. How many people do you think benefit from the book without finding this site for drills?

A passive knowledge of the Kanji is extremely easy to acquire from Heisig, an active knowledge isn't. Regardless if you can associate the Kanji with an English word, you'll be able to connect an extremely vague concept (and you can break new Kanji into radicals.)

Do Heisig up to, say, a 1300, review them a bit until you have at least 70% down, and then drop the reviews altogether.

After a month or two of study through Assimil and other immersion materials, go back and finish Heisig. With the knowledge accumulated, you should be encountering Kanji frequently enough (through immersion materials) such that you will learn them fully and reviews won't be necessary.

Using immersion materials like Read Real japanese and Assimil rather than textbooks is important for this approach. Immersion materials somehow expose you to kanji in a way that complements this approach.

Here's an analogy:

Think of it like this, can you tell me the exact plot of Mario game you've played before? No? Will your knowledge of the Mario universe make a new Mario game seem less bizzare? (Compare the first time you say a chain chomp in a game vs. the second.) Yes, it will, and the Kanji are like Mario games. Having experienced them but not memorized a plot to go with them will help you learn them when you do get into Kanji study and immersion (with assimil and other programs.)

I've used this approach by sheer accident and it is working wonders for me. Any thoughts?

I've been doing Heisig simultaneously with Native materials like video games, immersion, Core 2000, and reading through Japanese the Manga Way. I'm at 1650 in Heisig, and I have a somewhat similar feeling towards Heisig.

Many people on here are very strict with their Kanji Reviews and take their stories very seriously etc. But honestly that's not important to me as far as I can tell.

What Heisig does for you, in my opinion, is make you able to see all the different parts of a Kanji very very easily. You can break down new Kanji on your own, because you are able to see all the Radicals, and you understand the stroke order so you can write these new Kanji very well. This is the main benefit from Heisig.

What Heisig doesn't do for you - give you a concrete meaning for each individual Kanji. In this regard I honestly find the English Keywords fairly unimportant. If you remember them, great, but I see no reason try so hard to make a really good story developed in your Mind's eye. The reason being these stories are really a crutch that your going to want to discard along with these English keywords. Some of the time the English keywords aren't even accurate.

So really I don't see much reason to take Heisig Reviews very seriously. I never fail Kanji anymore when doing reviews. I half ass my stories. I intend to keep up my reviews for Heisig simply because I think it helps with handwriting, and keeping all the primitives fresh in your mind. But I don't know for how long I will keep up Kanji Reviews.

In short: RTK is great and certainly worth working your way through. But I don't really feel that it should be taken with seriousness I see on these forums. It's simply to ease your way into seeing real Japanese material. And it's by far the easiest/fastest way to acclimate yourself to the　常用漢字 and how to write them. But it will not outright teach you the Kanji. And the stories/english keywords are a crutch. They are a placeholder, and the more you work to remember them the stronger that crutch becomes.

I agree with whats been said here. For me Heisig is just the beginning, so that I can mentally distinguish one kanji from another. It's by no means a be all and end all (but still noteworthy, getting thousands of characters into your head in any form at all is an accomplishment).

I tend to agree that this is a possible route and I meant to post a little about it. I think the main point is how RTK prepares your brain to have a space for the kanji. You can then use native material to learn the readings/vocabulary. If you are only interested in reading the kanji and not remembering how to write them I don't think that painstakingly reviewing RTK is the most efficient way.

Here's a possible study plan:

* Use a "lazy RTK" approach (I think someone mentionned such a method, where the story is part of the prompt instead of just the keyword; or there's this method). The aim is to create the "kanji space" quickly in your brain. Recall rate should be near 100% so RTK should go by quickly.

* Start reading native material and drop the RTK reviews once you feel comfortable enough. Read a lot.

(* If you want to also know how to write the kanji, once you've read enough you can go back to strict RTK with Japanese keywords.)

I'm currently doing something like that, although I did the strict RTK in the beginning (so of course I cannot assert that the "lazy RTK" approach is good). After I started reading I just dropped the reviews and I'm now starting again with Japanese keywords. I'm using wrightak's deck but modifying it rather heavily. I'll upload my deck when I'm finished.

Hey guys,
I kind of agree about what you feel,
yet it all depends on the fluency you aim to attain through the method.
For instance, one of my goals is to read a detective novel in japanese. And to achieve that you need to know at least 3000 kanjis. And in context, you can get a feel for the meaning, but not the actual definition of a compound, which means you would need to learn it anyway, which in my perception, is done quicker and easier when done systematically. So the writing drills are kind of working for me, but I'm kind of a graphomaniac, so I'm probably not objective about that matter.
I mean once you know the kanjis by heart, sure, to hell with the reviews.
You might want to read the thread about the japanese key words by the way.

I did RTK normally, but about a year later, found that since I had been neglecting my reviews a lot, it was becoming really annoying and I was forgetting a lot.
I also noticed that I often had trouble recognizing the kanji when I was reading. I suspected that there was a very good reason for this. RTK trains you to go from keyword to kanji... but in reading, you see the kanji first, and have to recognize it.

I decided to drop my normal reviews altogether, then downloaded the anki RTK deck, and went through the entire thing again, only in reverse. That is to say, I looked at the kanji and produced the keyword (or a word with similar meaning). Having already gone through RTK normally, doing it the reverse way was a breeze. There are a few that I have trouble with, but I'm seeing a lot of benefit from this. For one, the reviews are much much easier, so I don't get frustrated with them. Also, when I'm reading now, I see kanji and actually recognize them.

So in summary. I think if you aren't interested in learning to write the kanji, then doing RTK as kanji -> keyword can be very beneficial, and more manageable. Of course I would still suggest at least going through RTK the other way around, otherwise you might not learn the kanji properly.

Don't forget RTK Lite so that you can do 1000 kanji that are used 90% of the time (or roughly that). YMMV

EratiK Wrote:For instance, one of my goals is to read a detective novel in japanese. And to achieve that you need to know at least 3000 kanjis.

So, I've read several detective novels in Japanese, and I definitely don't know anywhere near 3000 kanji. I don't even really know the whole Jouyou set. What you will need to know are words, and grammar.

As it happens, I think that detective novels are comparatively straightforward for several reasons:
(1) a lot of the pleasure of reading them is in the plot rather than any literary style, so they're enjoyable even if subtleties of style go over your head, or you're reading for gist and skipping unknown words
(2) they're generally set in the present day so there's plenty of commonplace vocab that you probably know already
So if you enjoy them, you definitely shouldn't wait until you have some insane number of kanji memorised first...

I'll second pm215. I began reading novels when I knew ~500 kanji, and vocabulary posed more of a problem for me than kanji ever did -- although you'll probably need more than that if you're doing Heisig because Heisig doesn't go by anything like kanji frequency.

I know probably ~1400 now, and a couple of months ago I finished Kasha by Miyabe Miyuki, an excellent detective novel focusing on debt and loan sharking in Japan. It's long but straightforward in style, and despite the subject matter you don't need much knowledge of Japanese financial vocabulary to read it. (A reader's poll picked it as the best mystery novel of the past 20 years, by the way).

The question is, do you want to make sure that you never ever never have to look up a kanji, or do you want to read a detective novel? Because, if you want to read a detective novel, read a detective novel.

I think doing it the "proper" way is far more beneficial if you're new to the language, but for people with experience it can be annoying to "miss" a kanji in a review session you know like the back of your hand simply because the keyword is wrong. In any case I find it to be more useful for identifying components than anything else, and slogging through the whole thing in a few weeks seems more useful than doing it in chunks in my opinion.

@pm215 and Fillanzea: Could either of you recommend a couple of detective fiction authors to start with? I'm thinking the Japanese equivalent of (the late) Robert B Parker, if you know his work, would be perfect because of his spare style--good plots, short chapters, snappy dialogue. Hmm, I wonder if Parker's been translated into Japanese....

After falling too far behind a couple of times in regular RTK I discovered RTKLite and have been doing much better with it. I'm at about 1300 and it's made a huge difference already in my interaction with written Japanese in the wild.

I'm not actually much of a reader of detective fiction. I won't recommend Kasha only because it's about 600 pages long and it's difficult enough for me, as slowly as I read, to hold all the threads of plot in my head that long. Miyabe has a couple of short story collections: 我らが隣人の犯罪, とり残されて, 地下街の雨, 人質カノン, and some others -- but I won't list the historical ones.

Quote:So, I've read several detective novels in Japanese, and I definitely don't know anywhere near 3000 kanji. I don't even really know the whole Jouyou set. What you will need to know are words, and grammar.

Good to know. Don't know why I wanted to wait... I don't have any statistics, but you know, linguists say technical Japanese uses between 3000 and 4000 kanjis. I should have mentionned I'm a student in literature, so yes, my long term aim is to read heavy, barely understandable Japanese shit (like who haven't dream to read the 11 tomes of the Hagakure in Japanese? Okay, maybe a lot of you...).
But a lot of answers above are intersesting (like the Anki anti-RTK),
so thanks everyone for the input (I carefully wrote down the authors).

I haven't read much detective fiction in Japanese, only a few books.

点と線 by 松本 清張 : this is an alibi-breaking kind of mystery; lots of following the detective as he wades through train timetables and the like.
佐渡伝説殺人事件 by 内田 康夫: this is one of a series which I think I've seen described as combination detective fiction and travel writing. The protagonist gets involved in a case which requires him to go to some random part of Japan, which is described in loving detail. (In this book it's 佐渡, obviously.) I've also read another in this series but forget the title as it was a library book; that one was set in Tokyo's shitamachi and involved conflict over the remodelling of Ueno station (which dotes it a bit!)

Anyway, those are just the books I've encountered (although 点と線 at least was a recommendation from somebody, I think). You might try looking to see if there's some sort of Japanese mystery fiction prize and try reading the books that have been winners in past years, perhaps.

EratiK Wrote:Good to know. Don't know why I wanted to wait...

The person who wrote RTK3 claims that she was unable to read detective novels with just the RTK1 kanji. I can't say for certain without knowing exactly what she did, but my feeling is that she just gave up too quickly or had problems other than kanji that she didn't realize. I just don't see how it's possible to know 2000 kanji and still have kanji be an insuperable barrier to reading any modern text.

But a lot of people find that to be a big mental block and have a very tough time getting past the idea of "3000 (or whatever number) kanji are used in modern Japanese, so I can't read anything until I learn 3000 kanji." There's also a related idea that if 3000 are used and you know 1000, then you're only going to be able to understand 1/3 of a book, which totally ignores kanji frequency and everything else that goes into understanding something.

Quote:I don't have any statistics, but you know, linguists says technical Japanese uses between 3000 and 4000 kanjis. I should have mentionned I'm a student in literature, so yes, my long term aim is to read heavy, barely understandable Japanese shit (like who haven't dream to read the 11 tomes of the Hagakure in Japanese? Okay, maybe a lot of you...).

Literature actually doesn't require as much kanji as a lot of people think. Many editions of anything pre-WW2 will have a lot of furigana. Hagakure is available in an 岩波文庫 edition which will almost certainly have furigana over any non-Jouyou kanji and maybe even more than that. If you're doing anything Heian period there will probably be furigana over almost every kanji. (Although the more obscure the text the less likely that is.) I need to know more kanji to read Internet pages than anything I read literature related.

As has been pointed out many times before, the more kanji you learn, the less useful each new one is. I honestly don't know how many kanji I know, but I would be surprised if it's 4000 or even 2500 (depending on your definition of "know").

You're right: to hell with the mental blocks! (that's why I love this website).
I'll go buy my first kids story tomorrow (because if I'll soon know 1000 kanjis, my vocab is more like 50 words, so I need to take it slow).
I think I just need to change my conception of "learning all the kanjis once and for all and never look up another one ever again because it's such a pain in the ass to learn".
Which I just did.
Thanks.

Nukemarine Wrote:Don't forget RTK Lite so that you can do 1000 kanji that are used 90% of the time (or roughly that). YMMV

I second Nukemarine, if your only going to do a portion of RTK, at least get the biggest bang for your buck. You should be able to visually break down most kanji you don't know with Lite.

If you know 1000 kanji but only 50 words, grammar may be a big problem depending on what else you've studied, but give it a try at least.