I will finish RTK1 by the end of January. I've already looked ahead at 2&3, read a lot of the topics concerning them, and researched various other learning methods people use for these.

My question for all of you at any stage of working on RTK2/3 is: with your experience now, do you think it's better to simply continue with RTK2? Or skip to RTK3, add the kanji, then return to add the readings for all 3000?

I'm leaning towards jumping to 3. Heisig admits afterall that the 3000 mark is a more realistic set of kanji necessary for fluency, but that he didn't include them because they weren't part of the government mandated 2000.

Also, I assume this would make the methods for learning the yomi (KanjiTown, chaining, etc.) more concise since I won't have to add 1000 more stray readings later on.

What do you think?

Many of the RTK 3 kanjis are so rare that you shouldn't be putting anything off until you've learned them and their readings. I mean, Heisig recommends studying RTK 1 first and only then starting other studies, and while that may work with RTK 1, I wouldn't recommend it with RTK 3. Having gone through RTK 1 you have all you need to study whatever kanjis you like on your own.

Cherry-picking kanjis from RTK 3 as you encounter them in real texts isn't bad either. Also, there are useful kanjis outside RTK 3's scope and useful non-jouyou readings, so you shouldn't really expect to ever completely "finish" learning kanjis any more than you will ever finish learning vocabulary. The frequency you have to learn new things just gradually falls.

I'd venture that learning new kanji has pretty much (by the end of RTK1) become something that you're going to be doing pretty much every day for the rest of your life. Learning or reviewing, I guess I should say. So, one might start doing readings of kanji after RTK1 (ie, RTK2), but I've just continued with RTK3 as well, and in fact I'm giving more time to RTK3 than to RTK2. But then I'm trying to read Japanese in context too, so there is some time to those skills, even when it is not strictly speaking RTK2...just my opinion.

I used the frequency index in Wakan which counts the 2501 most common kanji (I don't know what index he used for this). For all 3007 Heisig index kanji, I found:

RTK3:
~140 kanji occur from 745 to 2042 on the frequency scale.
~406 kanji occur from 745 to 2501 on the frequency scale (601 are unranked)
All but one are "uncommon" or "name-only" kanji.

RTK1:
~140 kanji occur from 2042 to 2501 on the frequency scale.
~35 kanji are unranked

This is directly related to what laxxy started in his RTK3: Easy Kanji thread.

I guess a better question might be, do you have trouble learning non-RTK1 kanji and fitting them into your RTK2 reading framework? Would it have been beneficial to at least have added a few of the more common RTK3's?

I'm still wondering whether it would simply be better for me to jus learn 'em all while I'm at it rather than add them sporadically as I encounter them in daily life. For example, 丼 is a kanji I see every single day. It's a simple kanji even the youngest J children know. But it's 2884 in the Heisig index, and 2088 in frequency.

Maybe what I'll do, just for the sake of goals, is learn all 2501 most common kanji regardless of which book they're in. That should put me at 2448 kanji. That might be kind of difficult though considering the progressive method of adding primitives...

Let me just chime in again to say that part of the secret sause of Heisig method is to learn them in a specific order that makes sense *structurally*, without getting out of place due to simplicity, meaning, frequency, reading, etc. If that philosophical approach is valid for the first bit of kanji, then it will still be a valuable guiding principle to follow for later kanji. I don't really think it's all that efficient to allow our indroduction of new kanji to be governed by "as we come to it" or a frequency index. IMO.

I've started RTK3 after RTK1 (not because I need it for my Japanese at this point, but because I like learning kanj). The kanji themselves aren't always hard to learn, since you've learned most of the primitives by now, and even the few new ones will look familiar since they already occurred in RTK1. So, many of them are easy to learn (actually, many are easier to learn than the last quarter of RTK1). The difficulty of RTK3 so far is mainly that I got a lot of plants, trees, flowers, body parts, etc, to confuse. Those are still the hardest for me.

In way of support, you won't find many stories from other users. I've tried to put one story on each kanji so that they won't be empty, and Charles Garrett has added his stories to the first 200 or so as well, but that still won't give you much choice. I do encourage people doing RTK3 to share stories as much as possible, even if they don't consider them especially good. We need more choice over there...

However, if you fully intend to keep up with reviewing your RTK1 as well, the combined total of kanji can become burdensome. I had to make a temporary halt around 2500 because the total reviewing burden is becoming increasingly heavy. My recall rate of the RTK3 kanji is not as good as the RTK1, and I have to review a lot to keep up.

Sometimes I wish I could separate the Leitner stacks for RTK1 and RTK3, that would make it a little easier; I could give the RTK3 a lower priority.

dingomick Wrote:My question for all of you at any stage of working on RTK2/3 is: with your experience now, do you think it's better to simply continue with RTK2? Or skip to RTK3, add the kanji, then return to add the readings for all 3000?

Personally, I'm bored just learning how to write kanji without being able to read let alone use them so for me I've been dipping into RTK2 since about the 1500s of RTK1. Definitely going to go for RTK3 at some stage but waiting for my motivation to return on that front.

As far as readings are concerned I'm studying for the 漢字検定 coz I find that a lot of the vocab is stuff I use but I get to learn the kanji/word relationship. Also it gives me a specific goal as opposed to simply learning vocab. The tests do require you to know ON and KUN readings which is a bit of a deviation from 'pure Heisig'.

For me RTK2 isn't the bible that RTK1 was, more of a reference, i.e. learn vocab for 漢検 then check RTK2 for signal primative value etc. Incidentally, I've got RTK2 second edition and it seems riddled with small mistakes (particularly the cross reference numbers, sometimes missing; other times incorrect).

I'd say be your own judge regarding what to get into next, go where your motivation takes you.

CharleyGarrett Wrote:If that philosophical approach is valid for the first bit of kanji, then it will still be a valuable guiding principle to follow for later kanji. I don't really think it's all that efficient to allow our indroduction of new kanji to be governed by "as we come to it" or a frequency index.

So when should we break this principle since we must eventually? RTK3 wasn't published until 1994. So for almost 20 years prior everyone who used his method HAD to break that principle and learn every new kanji on a "as we come to it or frequency index" basis. Even now, everyone else MUST also break this principle for kanji they learn beyond the 3007 in the Heisig index.

There are many non-Heisig kanji that could have been included in his framework. Just a cursory glance provides dozens of examples of simple kanji that are in the 2500 frequency list but not in Heisig's index. Some examples I choose because of their simplicity (and even possible early inclusion in the Heisig progression) and because they also even include compound examples in Wakan:

嗅 | Freq 2480 | 嗅ぐ trans-verb smell, sniff, scent

贅 | Freq 2386 | 贅沢 na-adj luxury, extravagance

瘤 | Freq 2373 | 瘤 noun bump, lump, protuberance, swelling

薔 | Freq 2356 | 薔薇 noun rose | 薔薇色 noun rose-color

峙 | Freq 2279 | 対峙 p-suru confronting, holding your own with

(I'd make those bigger, but I don't know how to change font size in BBC code, just html...)

I'm trying to figure out how I should proceed after RTK1. Whether it would be most beneficial to
1. continue through RTK2 then into RTK3,
2. do all of RTK3 then all of the readings,
3. pick the ~400 kanji of RTK3 that have a higher freq, then do all the readings.

I think I'll probably do option 2. Now that I've dedicated myself to these kanji I might as go all the way, and I think this is the most conducive path. First all the meanings and strokes, then all the readings.

(Astridtops, I made new accounts for my mass reviews (dingomick1026, dingomick2042) to ensure good randomization. I was going to recommend starting a new account just for RTK3, but then realized it would include 1-2042. New feature Fabrice, woelpad, greasemonkeys...?)

I delayed a lot of my goals of tackling spoken Japanese to finish RTK1, so I am tackling RTK II slowly while I am starting to do more Japanese for Everyone and Japanese the Mangaway. I will do Pimsleur III during my summer vacation. I've been approaching RTKII with the blue kanji box, setting up a Leitner system with post-it notes. It's a slow go because I try to only have 10 cards to review each day unless I fail a lot of cards on review. On a good day the most I add is 6 new cards so it's taking a while.

I've also been using a Lonely Planet phrasebook and converting the phrases to both their kanji and hiragana equivalents while I create flashcards on flashcardexchange.com. I find that with my time I need to limit my kanji review or I won't have time to learn how to communicate in Japanese. What I've found is when I make my flashcards for phrases the kanji don't seem so foreign. They are actually helpful in helping me deconstruct the meanings of sentences.

For me, it's important to set limits on the kanji review because I want to be able to live in Japan for a while and actually communicate, but if you've fallen in love with kanji study, I say go for it however you want once you finish RTK1.

CharleyGarrett Wrote:Let me just chime in again to say that part of the secret sause of Heisig method is to learn them in a specific order that makes sense *structurally*, without getting out of place due to simplicity, meaning, frequency, reading, etc. If that philosophical approach is valid for the first bit of kanji, then it will still be a valuable guiding principle to follow for later kanji. I don't really think it's all that efficient to allow our indroduction of new kanji to be governed by "as we come to it" or a frequency index. IMO.

The order in RTK 1 is useful so as not to introduce huge number of primitives at once and to give enough examples of each primitive. There aren't many new primitives in RTK 3, and once you're done with those, I don't think the order matters any more.

astridtops Wrote:[snip] However, if you fully intend to keep up with reviewing your RTK1 as well, the combined total of kanji can become burdensome. I had to make a temporary halt around 2500 because the total reviewing burden is becoming increasingly heavy. My recall rate of the RTK3 kanji is not as good as the RTK1, and I have to review a lot to keep up.

Sometimes I wish I could separate the Leitner stacks for RTK1 and RTK3, that would make it a little easier; I could give the RTK3 a lower priority.

I'm hoping to finish RTK1 in the next 4-6 weeks and am looking to continue on with RTKIII. So the above quote from astridtops has great relevance to me as well.

I think astridtops raises a good point about having separate Leitner stacks for RTK1 and RTK3. It is sad to hear that people have had to interrupt their chosen study plan (continuing on to RTK 3) because the review burden is too much under the present system. This would go a long way to easing the burden and allowing people to continue using this marvelous site to study all 3007 Heisig kanji.

Another possibility would be that if Fuaburisu would enable people to add kanji out of sequence someone could set up another account name (as suggested by dingomick) for RTK 3 study only.

How about it Fuaburisu? I think either one would be a worthy addition to this wonderful website.

Just for completeness sake, you could also have more "boxes", so that the well known kanji would come around even less frequently. And then another is to mark a kanji "done", so that it would NEVER come up again. For example, the numbers 1 to 10. I've reviewed them for so many years now, that I don't think it would be possible for me to ever forget them. That would also allow some one to say "I don't care whether I know this kanji or not, I'm still done with it." Something about leeches.

dingomick Wrote:I was going to recommend starting a new account just for RTK3, but then realized it would include 1-2042.

You can pretty much do this already, because the blue stack always empties from the most recently added first, and is grouped by the day cards were added.

Create a new account today, add 2042 cards but leave them in the blue stack. Then tomorrow, add your first 10 or whatever from RTKIII. The blue link will immediately say Test cards added today (10). Clicking on this will review those ten, then take you back to the main review page afterwards. You can leave the first 2042 in the blue stack forever, without them interfering with your RTKIII cards.

The trick is to leave one day after you add the first 2042.

That is correct colonel. I salute you! I will do that in the future.

[ファブリス]

Quote:I think astridtops raises a good point about having separate Leitner stacks for RTK1 and RTK3.

I'll look at that this weekend.

Quote:Sometimes I wish I could separate the Leitner stacks for RTK1 and RTK3, that would make it a little easier; I could give the RTK3 a lower priority.

I would have thought you would have wanted to give rtk3 a priority.

I don't know just how many primitives that appear as kanji themselves are used in RtK3, but I'm guessing that by reviewing RtK3 alone you would be reviewing a fair bit of RtK1 kanji as well ?

Anyways, I think if I put a Leitner stack display that shows only RtK3 kanji it would give a better estimate of one's progress through RtK3.

ファブリス Wrote:Anyways, I think if I put a Leitner stack display that shows only RtK3 kanji it would give a better estimate of one's progress through RtK3.

That leaves the cherry-picking as a possible greasemonkey project. I've pondered it for a day and I'm pretty confident that I can trick the system into reviewing a user-selected list, provided the user starts with a blank slate, e.g. by creating a new account.

I would just go on to RTK3 while the method and stories are still fresh in your mind. RTK2 will take much more time to go through, and it will be easier like you said to group similar on-yomi together from the start after going through RTK3. Personally I don't feel it's urgent to know how to sound out the kanji when I can get the gist of Japanese text from the meaning of the kanji and the kana words that i know. For me, meaning is more urgent than sound, but of course this is IMO.

Also, props to astridtops, I appreciate all the stories you've contributed for us RTK3 learners. I've gotten about as far as you in the book, but on principle I don't share a story unless it's gotten my kanji to the fifth stack, whereupon I can safely say it's a memorable story for at least me. I just started the book last month, so a lot of the kanji haven't made it there yet, but the ones that do will definitely get shared. Good luck with the rest of the book!

ファブリス Wrote:I would have thought you would have wanted to give rtk3 a priority.

I don't know just how many primitives that appear as kanji themselves are used in RtK3, but I'm guessing that by reviewing RtK3 alone you would be reviewing a fair bit of RtK1 kanji as well ?

I don't consider RTK3 a high priority because I don't need them very often. It's something I really do for the pleasure of studying kanji, while with RTK1 there is a necessity to hammer them really into my head to be able to use them in my Japanese classes. I try to write all vocab during lessons in kanji instead of hiragana, and the more kanji I can quickly write, the better I can keep up. The RTK3 kanji do not feature often, so they'd be more of an optional element.

As for the learning of kanji, it's true that I don't have to learn the primitives, but I'd still have to learn the combination and memorize the story as attached to the key word, and if I'm not succesfull enough, the review stacks are overflowing (which is still the case at this point). I'm currently taking a 25 limit per day for stack one, which means I stop studying the last stack if that number of failed kanji is reached.

PS - Fabrice, If you can do that separation of Leitner stacks for RTK3, I'd be much obliged if you could do it in such a way that I can transfer my current RTK3 kanji and the stories attached to them to the new stacks. Well, the stories mostly. I wouldn't mind if they'd all start in stack one or so, but rewriting 500+ stories is not something I'd look forward to, as you probably can imagine

mantixen Wrote:I've gotten about as far as you in the book, but on principle I don't share a story unless it's gotten my kanji to the fifth stack, whereupon I can safely say it's a memorable story for at least me.

Well, that's one way to look at it, and I certainly respect that point of view. But on the other hand, for me it really helps if other people write some different angles of thought down. Say you write something down that you'd have to change later and thus isn't succesful for you - I could still use that line of thought for my own story and make it somehow work for me. That's why I've written 500 stories for RTK3 - not because they're any good (I'm convinced many of them are utterly lame) but at least anyone looking at them can say: "that line of thought is too lame, I can do better than that", and so it may offer a starting point.

So, in that regard, I'd rather see people share some ideas even if they're not perfected yet. Don't be shy

Astridtops, I'd just like you to know that you have been an inspiration for me. (And there have been many others for RTK1, of course.)

I think it is very generous of you to share stories for all RTK3 kanji. It is great that you put yourself out there with stories for every single kanji in the hopes of making RTK3 less daunting and to inspire people to add stories of their own. I know that when I get to RTK3 it won't be such a cold and lonely place. Thank you!

[ファブリス]

Quote:I try to write all vocab during lessons in kanji instead of hiragana, and the more kanji I can quickly write, the better I can keep up.

That's great astridtops! I also started a few weeks ago, finally, to write down all sorts of words I lookup and come accross while composing a japanese email or writing a letter. I don't really study them yet, but the ability to write them as is is there! Go go RTK!

Quote:Fabrice, If you can do that separation of Leitner stacks for RTK3, I'd be much obliged if you could do it in such a way that I can transfer my current RTK3 kanji and the stories attached to them to the new stacks. Well, the stories mostly. I wouldn't mind if they'd all start in stack one or so, but rewriting 500+ stories is not something I'd look forward to, as you probably can imagine

What I have in mind works just like a filter. The reviewing status is not affected. You simply get to see all of them, only the rtk1 ones or the rtk3 ones. And correspondingly you will review all together, only rtk1, or only rtk3. So only the cards that get picked for review are affected.

Hmm, my reply didn't post.

Anyway, unless it's simply a direct copy of someone else's story, I post a story for every kanji. It may be lame, it may make sense only to me, I may change it later. But there might be a nugget in there that makes all the difference in the world for someone.

The collaboration is the foundation of this website. It's like brainstorming. The more options of stories and images everyone has to choose from, the better everyone's learning will be.

dingomick Wrote:Anyway, unless it's simply a direct copy of someone else's story, I post a story for every kanji. It may be lame, it may make sense only to me, I may change it later. But there might be a nugget in there that makes all the difference in the world for someone.

The collaboration is the foundation of this website. It's like brainstorming. The more options of stories and images everyone has to choose from, the better everyone's learning will be.

I like your stories most of the time. I post my stories if I think that my story is better than what's there or summarizes it (and it happens way to often). It's interesting that with this many registered users only so few actually post the stories, and of those even fewer constantly post reasonable ones.

Problem with current state of the site is that there's no process for refinement of stories. Bad stories are left there intact. The more stories there are, the more difficult it is to look through them all. It incites responsible contributors NOT to post mere story refinements, as it makes even more stories and makes story lists less user-friendly ("Although No Liability Is Implied Herein, the Consumer Is Warned That This Process Will Ultimately Lead to the Heat Death of the Universe"). Even when some reasonable story is achieved after all, it soon becomes buried under many unreasonable deviations, and even if it's selected by most users there's no way to show that so everyone needs to look through every story again. If there were some ranking for stories and members, it might be fixed. By story ranking I mean story "copy" feature, when other's story is referenced (so that only single story for kanji may be ranked) - this is "must have" either way as stories are more frequently adopted than invented and manual copypaste is not a civilized way to carry it out. Member ranking can be achieved through statistics on 'friends' lists which I suggested in Feedback forum. Alternative is some kind of wiki system, but with current contribution rate it's unlikely to succeed.

To summarize for stories around 600-800, Fabrice made basic stories for everything (though some of them are too farfetched for my taste, I guess goal was to post at least something for each kanji to bootstrap storymaking), raulir contributes reasonable stories most of the time, smithsonian, darg_sama, matticus and Immacolata sometimes come up with something better than others and dingomick completes underdeveloped storylines. It becomes a process for story selection as there's not enough energy to read everything.

[ファブリス]

RoboTact Wrote:Fabrice made basic stories for everything (though some of them are too farfetched for my taste

hahaha, they are for me too, believe me ... thanks to everyone's contributions I can finally put some to rest and use simpler alternatives ^_^

Regarding stories ranking / filtering / ordering let's continue the discussion in Rating System for suggestions / hide ones which you don't want to see, as this is getting out of topic. Thanks.

RoboTact Wrote:To summarize for stories around 600-800, Fabrice made basic stories for everything (though some of them are too farfetched for my taste, I guess goal was to post at least something for each kanji to bootstrap storymaking), raulir contributes reasonable stories most of the time, smithsonian, darg_sama, matticus and Immacolata sometimes come up with something better than others and dingomick completes underdeveloped storylines. It becomes a process for story selection as there's not enough energy to read everything.

I think my taste is not too different from yours, but some people seem to actually like far-fetched stories a lot (take a look at "favourite stories by someone else" thread, for example). Even I sometimes notice that a far-fetched story that I thought was rather ridiculous at the time actually stuck pretty well. So this varies a lot from person to person.