Hi

I just recently started doing the Heisig method of remembering Kanji through stories, although without the actual Heisig book.

I was thinking of setting up my learning methods to start learning Kanji by the frequency of use. There is this site where the author data mined a couple news sites and then tallied up the results. Jim Breens dictionary files also provide frequency values for most of the kanji.

I'm thinking this would be a more efficient method of learning to be able to start reading as soon as possible. Judging by that sites statistics, you can learn about 1000 kanji and be able to read approximately 93% of the writing you encounter which I think is an excellent measure.

What do you guys think?

Bad idea, for these reasons:

* You're going to want to learn all the jouyou kanji anyway. Knowing the 1000 most common kanji will let you read a lot, but 93% sounds WAY high. It depends on the subject of what you're reading etc, you won't simply learn 1000 kanji and be satisfied with that.
* Learning 1000 kanji can be done in 1-2 months. 2000 kanji in 3-4 months. So is it worth it to mess up the order just to get 1-2 months of extra reading (where you will still run into kanji you don't know from time to time)?
* Heisigs order is based on logic. You learn the parts of the kanji first, you don't learn a kanji and learn it's radical next.
* Heisig teaches radicals, if you don't have the book you will probably not be able to split kanji up properly, making it much harder to work on the stories.

Overall, following Heisig is the safe and best way of doing it. The pros outweighs the cons easily.

Well obviously I won't stop at 1000 Kanji. Also I don't have the Heisig book, and won't have it for a few weeks even if I order it online due to the region I'm in. So for stories I basically just have my own and the ones available on this site, as well as a bunch of Anki flash card decks.

If you're not going to do Heisig, you can not do Heisig any way you want -- but it will be hard for us to give you suggestions grounded in experience, because all our experience is with doing Heisig. Except for those of us who did it the usual way first and failed so hard we're now even more rabid Heisig partisans; and alyks, who invented an interesting method based on Heisig for learning readings along with meanings.

deathtrap Wrote:Well obviously I won't stop at 1000 Kanji.

If you're planning to do all 2000 anyway, why not do them in the "proper" order? I don't think it's impossible or anything to learn by frequency of use, but is it worth it?

I seriously also think that this is a bad idea. I have been living in Japan for around 3 months now and when walking with my friend who has finished Heisig we can even spot at least 2 or 3 non-standard Kanji out in the field. xD
I am at around 850 now, and I also feel that most basic Kanji really ARE covered in the first 1000. Even with Heisig's order of Kanji it's usefull from the very beginning!
I really think just by wanting to learn most frequent used ones first you should change the order, since it will make studying more difficult and I think in the end you will take more time doing that, than if you would have used Heisig's order... That's what I think. Really, with Heisig's order you can do a lot of Kanji much more faster for reasons that you can link kanji with the same story subjects, and kanji with the same primitive elements come after each other what makes story making a lot easier and faster. I covered around 50 kanji all with references to the same anime series. ^^ And I am sure there are lots more reasons why everyone doesn't encourage a diffrent order.
But it's still all up to you! xD
-Mesqueeb

There are a lot of characters high up on that list which are not all that frequent outside of the news. Considering newspapers are typically one of the harder things to reach reading fluency in, I would consider a different frequency order if you go that route. Try a textbook which also teaches kanji. The ordering tends to be more in order of usefulness rather than frequency of appearance newspaper. If you can't read yet, then most likely the first thing you'll be able to read (or want to read) won't be the news.

But I highly suggest Heisig order. From experience, I can say it takes much longer to learn 1,000 kanji the frequency way than 2,000 kanji Heisig's way. If the overall goal is to learn 2,000 kanji, Heisig is the best way to go.

You can always do a trial. Try 200 kanji the Heisig way. Then try learning 200 the frequency way. See which one you prefer.

I had a similar notion back when i was starting, but mine was a tad more sensible, my aim was to kinda mix heisig a little with the order that appears from primary school up to high school (i.e following JLPT from 4 up to 1).

In any case when i learn juko/readings and such. I plan to follow, as close as i can the JLPT order. This is so i can start reading elementary texts with kanji as early as possible. From there, i'll be able to grab a firm grounding in Japanese from basic grammar, up to advanced grammar.

The site you linked did the frequency of use on a newspaper, where the language (not just the kanji) is at a very high level. I think you know where i'm headed with my point.

deathtrap Wrote:I just recently started doing the Heisig method of remembering Kanji through stories, although without the actual Heisig book.

This is how Heisig works:
* He assigns a relatively *unique* keyword for every Kanji character.
* He only introduces one radical at a time and then he lists and explains all the Kanji that can be constructed using the new radical and the learned radicals so far.
* He helps with stories at the first two parts of his book.
* He provides the stroke order especially at the first two parts of his book.

So, you can see that it is not only about stories.

deathtrap Wrote:I was thinking of setting up my learning methods to start learning Kanji by the frequency of use. There is this site where the author data mined a couple news sites and then tallied up the results. Jim Breens dictionary files also provide frequency values for most of the kanji.

To address your second point, it is impossible to mix the frequency method with the Heisig method. I wish I can do that myself, but it does not work. You have to dig your way through the Kanjis one by one till you reach the last one of the 2000. So, using the Heisig way, it is more like one package and you won't be able to recognize all the frequent Kanjis till you finish the whole of the first volume.

Last but not least, if you want to study Heisig's book, you cannot use your knowledge of Kanji characters that you learned using other methods. Meaning that let's say that someone is very comfortable with 500 frequent Kanjis and he wants to use Heisig, he will have to start with Kanji #1 in the book and he *cannot* skip any!

So now, you have two options. Either order Heisig book and wait till you get it and only then start studying Kanji, or you can use another method.

joxn_costello Wrote:and alyks, who invented an interesting method based on Heisig for learning readings along with meanings.

Haven`t been here for awhile. Where can I find more information on alyks`s method?

@deathtrap: You can download a sample pdf with 250 kanjis on the internet. That should be enough to get you started with Heisig.

watashimo Wrote:

joxn_costello Wrote:and alyks, who invented an interesting method based on Heisig for learning readings along with meanings.

Haven`t been here for awhile. Where can I find more information on alyks`s method?

@deathtrap: You can download a sample pdf with 250 kanjis on the internet. That should be enough to get you started with Heisig.

Search on the forums for "the movie method" and you should find it. The basic idea:

Take a movie. Pick a fitting sound for it (Let's say The Matrix for ま) and put every kanji with the reading まin a specific scene in that movie. The mnemonic for the kanji is built upon this, so when you see the kanji, you remember that scene from the matrix and thus you remember that the reading is ま. At least, that's how I think it works, I haven't tried it myself. But Alyks learned all the kanji AND one on'yomi per kanji in just 50 days, so there's no doubt that it works.

Not sure if you've got this or not:
http://www.nanzan-u.ac.jp/SHUBUNKEN/publ...sample.pdf

Recently I came to accept such idea. It seems like a very pragmatic solution. You should check the "RTK light" thread to see a better discussion on the subject.

I would still not do it. RTK kanjis already seem to few.

Frequency of use depends on what you're reading. For example, 本 is the most common in newspapers, but I could go through a whole video game without seeing that kanji.

It's better if you learn kanji as you encounter them, chances are you're going to see it again. Say you encounter a new kanji, write it down 20 times to get stroke order and then put into your SRS.

Or you can just go through Heisig and make this significantly easier.

As mentioned on the RTK light thread, there are already systems such as De Roo's that are very similar to Heisig's but are more open as to the order of learning. As Fabrice has stated on that thread, Heisig would probably not be a huge problem in other orders either so long as you started out with a list of all the primitives beforehand. Perhaps you could learn the kanji in the order of Kanji Odyssey since they are based on frequency.

Squintox, the frequency lists based on newspapers are not very good because they tend to over-emphasize economics and business terms, but that doesn't mean that a frequency list based on generalized vocabulary isn't possible or more useful. I don't think anybody knows how K.O. gathered their data, but it seems pretty accurate as far as widespread frequency in practice.

Squintox Wrote:For example, 本 is the most common in newspapers, but I could go through a whole video game without seeing that kanji.

本当に？

iSoron Wrote:

Squintox Wrote:For example, 本 is the most common in newspapers, but I could go through a whole video game without seeing that kanji.

本当に？

The game I'm talking about is 99.9% hiragana >_>

Dragg Wrote:Squintox, the frequency lists based on newspapers are not very good because they tend to over-emphasize economics and business terms, but that doesn't mean that a frequency list based on generalized vocabulary isn't possible or more useful. I don't think anybody knows how K.O. gathered their data, but it seems pretty accurate as far as widespread frequency in practice.

I don't know about that, in 400 sentences, I still haven't encountered 寺, even though it's one of the first you learn in K.O.

Of course, I'm not going to descredit the whole book because of one kanji, but knowing the 500 most common kanji (which apparently means you can read 50% of text) doesn't necessarily mean you can read 50% of a manga, infact it may be as low as 20% or as high as 90%.

@squintox

First of all, the order of the K.O. books are not literally the exact order of frequency. What I meant to say is that book 1 does seem to truly represent the 555 most common kanji in widespread generalized usage.

However, there is a huge logical reason that they choose to introduce that "temple" kanji right away. It is placed right next to the kanji for "time" which uses the exact same "temple" kanji inside of it. And btw, the "temple" kanji is extremely common as well. 400 sentences is not a great indicator of anything.

Btw, I agree with your last paragraph that you might not be able to read 50 percent of a given manga with the knowledge of those kanji. However, I would bet that you would be able to read 50 percent on average of ALL manga which is what makes it useful for anybody who plans on using their Japanese language skills for more than just reading a particular genre of comics.

Dragg Wrote:@squintox

First of all, the order of the K.O. books are not literally the exact order of frequency. What I meant to say is that book 1 does seem to truly represent the 555 most common kanji in widespread generalized usage.

However, there is a huge logical reason that they choose to introduce that "temple" kanji right away. It is placed right next to the kanji for "time" which uses the exact same "temple" kanji inside of it. And btw, the "temple" kanji is extremely common as well.

Yeah, I think we both can agree on that. But what I'm saying is you're not guaranteed to see お寺 before 望遠鏡, I know I didn't, even though they are "fairly accurate" I think it's better to learn kanji as you encounter them, because they are truley 100% accurate, at least to you.

But you are guaranteed to see 寺 more OFTEN in most generalized media and that's why you learn it first.

Dragg Wrote:But you are guaranteed to see 寺 more OFTEN in most generalized media and that's why you learn it first.

Yet after learning 1,000 words and 600 kanji, I had not encountered 寺 until an hour ago, that woudl be a lot of energy spent that could've been put into another kanji I encounter more often. Kanji frequency is different for every form of media and every person.

@nestor, I agree, but at the same time I don't think that learning the kanji from a K.O./Heisig combined method would be hugely damaging for everybody. It might be okay for some people who are in such a rush to get into reading, but I also agree with Mentat that they need to learn ALL the kanji at some point. (I just read your last post and it looks like we are in complete agreement.)

@squintox
No offense, but the reason you don't see a lot of 寺 is probably for the same reason you haven't seen a lot of 本: you haven't had a lot of exposure yet. You yourself admitted that you are playing games that are 99.9 percent hiragana. I have seen the word for temple hundreds of times within the last few years, whereas I have only seen the word for telescope a small handful of times even though I expose myself to a wide variety of media.

I'm uncertain how learning vocabulary based on what you're encountering constitutes a short-term focus. I am fairly certain that vocabulary will still exist in the long-term.

I saw 寺 often in Japan, but rarely in video games.

Anyway, I think the topic was on the order best suited for learning kanji, then drifted into vocabulary learning order which really isn't the same thing.

Personally, I would say learn 寺 early because it's used in 時, 待, and 持 which are used quite often in most forms of media, but it doesn't really matter if you know 寺 is pronounced てら until later or whenever you feel like learning that. Learning the character isn't the same as learning the word. I'm not entirely sure which we're talking about now.

If you play an RPG game set in Japan, there is a really high probability that you will see 寺, but perhaps not so much in the RPG games that follow a more European mythological tone.

It's true that kanji and vocabulary are not exactly the same thing, but if we are talking about kanji frequency based on vocabularly usage as opposed to usage as primitives/radicals in other kanji, then the two become inextricably linked which is why this topic is looking kind of muddled.