This is the study method for Kanji I decided on after a lot of looking around at what was out there. I'm curious to hear what people think.

I studied Japanese with a textbook (all romanji) for about 10 months, getting a good vocabulary base and grammar knowledge. Once I learned hiragana and katakana I stopped using romanji and made my new flashcards in kana only, as well as converted a few old flashcards each day into kana.

Once I decided I wanted to learn Kanji, I downloaded the top 2000 most encountered kanji based on a search of hundreds of Japanese websites. They are listed in order here: http://po-ru.com/projects/kanji-project/list/. I decided I only wanted to learn kanji I was most likely to see often.

I also made the decision not to learn to write kanji. Not only is learning and doing it time consuming, but I’m not sure it’s worth it. The vast majority of Japanese write these characters exclusively using their phone or computer. To read (and write with a computer), I only need to be able to recognize the character.

So, I created 20 cards on Anki for the first 20 in my list. Then I broke them into their radicals and started a separate deck of radicals with their English meaning. This would allow me to make a story for each kanji using their radical parts. As there are only 214 radicals, I knew eventually I would have them all memorized, but only learn them as I need them.
My cards look like this:
Front:
Kanji: 年
Story: SLASH and DRY ONE rice field each year
Back:
on reading:ネン
kun reading:とし
English: year
dictionary #: 45
popularity #: 9

The dictionary # is from the book “Japanese Kanji and Kana” by Hadamitzky. The English reading and the story are the same color as the background, so that I can only see them if I highlight the card. Often, if I knew the English definition really well, I wouldn’t even add it.

Each radical in the story is capitalized and the English definition is underlined. Once I made a kanji card, I would search a Japanese website (like http://www.goo.ne.jp/) for that character and look for popular words that used that kanji character and 1-2 additional characters to make a word. For example 大学　– daigaku. In this way I could apply the on reading, which by itself isn’t very useful, but is very commonly found in these types of kanji combos. This helps you to remember the On reading.

I realize it’s difficult to learn the Japanese and English definitions at the same time, but it didn’t make sense to me to do the RTK method. My reason: after completing RTK you can write the characters, but the keywords aren't all that accurate as definitions, you can’t really understand any texts, and you haven’t learned any Japanese readings. You will have to do a few more steps after RTK to get anywhere. I like my method because you IMMEDIATELY learn the most useful kanji and how to read and understand them. It will go a bit slower than learning the RTK keywords, but it is immediately useful and it feels much more fulfilling to me to be progressing in Japanese speech and reading/ writing as I do these. I think in the long run, you get the same result faster.
I've only got 90 cards in this deck so far, but I'll make it public if anyone wants to try it out/ build on it. "Perry's Top 2K Kanji" is the deck name.

Well, good luck with that. In a previous life I would have been righteous and taken issue with the quote “I think in the long run, you get the same result faster” (from my experience, you will get less results, much more slowly). But you seem determined and empowered to follow your own path. Best of luck to you.

Any chance you'd release the full list of 3200 kanji and the frequency statistics?

This is really interesting!

[ファブリス]

Quantity (of information on your flashcards) does not make quality. Heisig had the nerve to say NO: one keyword, one kanji.

I would add example words instead of individual readings, unless you have grouped and ordered the kanji by ON or KUN readings (as in RTK2).

What you are doing is of course legitimate, but speaking from experience (I was doing something similar last summer), I got annoyed to progress ever so slowly. Add that to the fact that I was mixing up kanjis that looked similar - it may happen to you once you get past a couple hundreds of kanjis. Also, I was annoyed at not being able to write simple kanjis. Writing is in fact useful and (my personal opinion) fun and enjoyable. You don't always have a computer handy when you need to write.

RTK does NOT teach you japanese, but it teaches you a systematic way to remember 2000 confusing glyphs. For about 6 months' work, it is definitely worth your while. It is however only a first step on your way. After you finish RTK, you need to start revisiting the whole kanji set and add readings - for that you can use RTK2 or Halpern - I personally prefer Halpern, which I find more authoritative and complete than RTK2.

I read the explanatory part of the RTK book a couple times and I spent a lot of time looking around at how other people had liked it. I do think it's a great idea how he simplifies it down to one english keyword, I definitely respect what he's done. I ultimately came to the decision to try it this way because of the reasons I listed. Probably the biggest reason is because I began to doubt the usefulness of being able to correctly write the kanji. Sure, nothing beats RTK in teaching you how to Write and Recognize the kanji.But how often are you going to use that? Are you going to write long handwritten letters to japanese newspapers? Probably not, you're going to read papers or websites or books, etc. Because my vocabulary was already around 800 words, I want to have only japanese readings on the card or both japanese and english.

I figured I'd try it my way for at least the first 100 of the "top two thousand kanji", which includes (will include) over 100 additional cards that combine those kanji to make words I found on popular japanese websites. After the first 100 I can re-evaluate the system and how it's working for me.

@ファブリス "I would add example words instead of individual readings" - The card just has the individual On and Kun reading, but I immediately add another card that uses the On reading of the previous card and one or more kanji to make a common word.

@Mafried "you will get less results, much more slowly" I don't know, I've only got 100 cards, but I'll let you know. "Any chance you'd release the full list" Sure, when I get there!

You progress through individual kanji slower, yes. But you can really use and read the kanji you just learned right away. I use Lang-8.com (where you write a journal in Japanese and native japanese people correct it), already I'm starting to use these common kanji in place of a lot of words I used to spell with hiragana.

whoa now, let's not call anyone a liar here. Perhaps it's an exaggeration. Regardless, I've read the more recent generations are losing their ability to write kanji because after school they mostly depend on their phones/ computers to write kanji. #1 most important thing to me in my studies is to be able to converse with ease. #2 is to be able to read texts. Coming in last place is being able to write, which I can get along quite well in Japan writing by just electronic means. I challenge you to find a person in any Japanese city with out a phone on them with which they can write with.

I would say about half the people I know in Japan do not have a phone with which they can write. And that would be several hundred.

onafarm Wrote:I would say about half the people I know in Japan do not have a phone with which they can write. And that would be several hundred.

even i don't have a phone that writes(yet)

Among the japanese people I know, there is indeed someone that claims "not being able" to write any kanji. It is indeed a generational thing: she's young and never went to university. However, that self-proclaimed "illiterate" can recognize oodles of kanji instantly and read japanese novels and newspapers without a dictionary. No way a foreigner can reach that point without extensive exposure to the language, and the printed word is the easiest way to get that exposure.

The relevant question is, do you want to start getting exposed to the language before mastering most of the glyphs it is written in (RTK), or do you want to start getting exposed to only the glyphs that you know, that you collected bit by bit at much cost, and be condemned to do this for years on end before being able to tackle the simplest of native materials?

For me the choice was clear. But RTK is a long slog and somewhat of a punishment - especially when you get to around 1500 kanjis ;-), so it is best to try your approach first.

Regardless of its utility, learning how to write is the most tangential, fringe, unexpected side benefit of Heisig's method. It's Remembering the Kanji, not Writing the Kanji. It just so happens that the most efficient way to remember the kanji involves involves learning to write them. When you get to the kanji with visually similar primitives (there's a lot of them!), you'll understand why. Many young adult Japanese have trouble differentiating some kanji--and it is a generational thing--because they learned them the way you are (and it took them years to do it). Note the wording: they don't have trouble remembering a kanji, they have trouble differentiating how it is written compared with other kanjis. And they practiced their kanjis for years. RTK learners never forget how to write a kanji they know, and often complete the book in just a few months.

I suggest you look at RTK-2. RTK-1+2 will take you where you want to go much, much faster, and without all the long-term memory problems.

mafried Wrote:RTK learners never forget how to write a kanji they know

Lol, come on now, be serious. I agree with most of what you said in your post, but RTK learners definitely can and WILL forget how to write kanji.

I'm usually in the Yokohama/Tokyo area and the trains are packed with people typing on their phones and each of my friends/ acquaintances has internet/ e-mail on their けいたい。

Anyway, I don't want to get off topic and onto a petty argument. I just am presenting my idea and my reasoning for people to consider. If I eventually decide it is no good then I'll switch methods. Either way I will have learned to read, pronounce, and write some quantity of the most commonly encountered kanji in the WWW. Right now though, I can say it's been pretty fun, because I can look at an article and read lots of the characters I've learned over just the past 2 weeks.

I think the real question people should ask themselves is "honestly, how useful is it for me to know how to write kanji with a pen and paper in modern day japan?" People who know zero kanji get along ok, so is it that ridiculous for me to suggest you would get along just fine by only knowing how to read in Japanese and write with an electronic device in japanese?

"Kanji they know" It's part of the system. If I know a kanji then I know how to write it. Maybe others' experience is different, but I've never once encountered a kanji in the wild that I had learned but didn't know how to write, and if I'm writing and know which kanji I want to use, I know how to write it. Yes, I do forget some kanji. But in those cases I really have no clue. RTK makes remembering meaning+writing a package deal. I've never been in a situation where I knew which kanji I wanted to write, but dumbfounded as to how. But I've seen native speakers like this, and that was the point I was referring to.

Zarxrax Wrote:

mafried Wrote:RTK learners never forget how to write a kanji they know

Lol, come on now, be serious. I agree with most of what you said in your post, but RTK learners definitely can and WILL forget how to write kanji.

Well, one could argue it's stored forever in your muscle memory... Though it's not always available, that's for sure...

mafried Wrote:"Kanji they know" It's part of the system. If I know a kanji then I know how to write it. Maybe others' experience is different, but I've never once encountered a kanji in the wild that I had learned but didn't know how to write, and if I'm writing and know which kanji I want to use, I know how to write it. Yes, I do forget some kanji. But in those cases I really have no clue. RTK makes remembering meaning+writing a package deal. I've never been in a situation where I knew which kanji I wanted to write, but dumbfounded as to how. But I've seen native speakers like this, and that was the point I was referring to.

You've never know the kanji, had a basic mental picture of what it looked like, but then either messed up the writing or drew a blank when you tried to commit it to paper?

If not, then you are the new God of Kanji. Congratulations.

My guess, based on the experience of studying Japanese in a large undergraduate program without knowledge of Rtk back then: Your approach will become inefficient once you reach around 700-1000 characters... depending on how good your visual memory is.
The main benefit of of Rtk is better recognition, not writing.

First, very few people have the photographic memory capacity to accurately recognize 2000+ kanji without confusing many of them quite frequently, but virtually everyone of us has the biological capacity to store the writing of thousands of character in muscle memory, which then (counterintuitively) feeds into recognition. You don't need to do Rtk to acquire this muscle memory. You can also write each character thousands and thousands and thousands of times. But without the muscle memory aid, your memory capacity is likely to be tested hard.

Second, working through Rtk gives you an excellent foundation to understand the role of the radicals in character formation, which in turn makes it much more likely that you will be able to either recognize a rarely-used character accurately or at least have an idea of its meaning and/or reading.

Here's some evidence from a comparison between L1 and L2 Kanji tests:
http://books.google.co.uk/books?id=yDLVa...es&f=false

I'm not saying your approach won't work, but you would have to have an extremely unusual memory capacity for it to be more efficient than Rtk.



Neuroreport:
26 April 1999 - Volume 10 - Issue 6 - p 1335-1339
Involvement of motor cortices in retrieval of kanji studied by functional MRI
http://journals.lww.com/neuroreport/Abst...of.33.aspx

"FUNCTIONAL magnetic resonance imaging was successfully used to study the activation of the motor cortices during retrieval of Japanese ideogram, kanji. The subjects performed kanji completion tasks to generate a kanji in response to an element which is always written first. In most of the subjects, the contralateral premotor cortex, the presupplementary motor area (pre-SMA) and the bilateral intraparietal sulcus were activated during retrieval of kanji without actual writing nor intentional mental writing. Activation associated with actual writing was shown in the contralateral primary sensorimotor cortex and the SMA proper. These results suggested that retrieval of kanji would share the neural basis of motor representation with writing of kanji except for regions directly working for motor output."

Well it sounds like what he's doing is pretty similar to RTK--breaking the kanji up by radicals and then coming up with mnemonics to remember which radicals form the kanji.
I can't really see this as being a whole lot worse than RTK, but with RTK, there is the advantage that the kanji are ordered efficiently for learning. For instance, you learn 1 new primitive, and then you can learn 20 new kanji right off the bat.

If I understand correctly, it's not writing the characters that is most helpful in memorizing them, it's breaking them into parts, and then creating stories with those parts that will lead you to it's definition. I'm making my own stories using the radicals. That gives me the english definition. If I already know the reading in japanese that's great, but otherwise you have to memorize the readings right then. I admit that can be tough. but, hey, you decided to learn japanese... you have to learn the readings eventually anyway.

eubankp Wrote:As there are only 214 radicals

It seems you're working with the Bushu radicals.
Have you considered supplementing those with the radicals that Heisig added?
I just counted the number of radicals from the appendix of the book and found 230, so I guess the difference is not huge.

Also, how to you "call" the radicals you're learning? Are you using the standard Japanese names? Considering how memory works best, it's likely that the information will stick better if it's flashy and thus it might be a good idea to give them your own names (like Mr. Bean for 豆, or the St. Bernard dog for 大), like those found in the RevTK stories.

In my opinion, your approach seems very interesting. I mean, it seems like a shortcut to reading. I'd be happy to hear an update in a few months on how it's working out. Good luck in the meantime.

I was just using the "Kanji and Kana" dictionary definition of the radicals. eg: tongue, cow, body. I mean, the one for "woman" is the same as the kanji for "woman" so I worry about changing the names of the radicals. Besides that, I would prefer they had more interesting names.

I'll update here after a while, even if I'm hanging my head in defeat.