dizmox Wrote:

blackbrich Wrote:Quite the gamblers. Your gonna need srs, mnemonics, and your gonna need to eat and breathe kanji. Like when your taking a break from using SRS ur gonna need to be doing more kanji on the side. 2000 is a lot, 4000 that's just...

I hope you set some specifics of this bet. If you have to remember the meanings and a reading also. That's gonna hell.

He doesn't necessarily need mnemonics... they certainly didn't suit me.

I did a first run through of RTK in about 40 days, with around 2 hours a day of study, no mnemonics, 95% retention or so, but I guess I had a visual familiarity with about 700 before I started. OP will be fine if he has a decent memory and doesn't quit, provided the goal is just to memorise RTK1.

I am an intermediate mnemonics user & have a decent rote memory, but by no means the best. I felt that those two coupled together would give me a good shot at pulling this off.

Stansfield123 Wrote:You should negotiate down the error rate to 90%. And then work your ass off, but don't spend ten hours a day on this. 5-6 hours max, and carefully follow the method. All the extra time will achieve is make you sick of reviewing Kanji, it won't help you remember better. Even if you spend 5-6 hours, you'll have to tweak the Anki settings, to make sure you review each card much more frequently than you would with the default settings. (Anki is set up for learning over a longer period of time, not just one month).

I think 160-170 hours, over 4 weeks, might be enough to get a 90% retention rate, right at the end (though it won't be permanent). But 95% is much, much more difficult. It doesn't look that way on paper, but it is.

I spent about that time over the course of almost six weeks, and I think I was just a little bit under 90% at the end.

P.S. There are also several other ways to tilt the test in your favor. For one, make sure there's no time limit. Second, you could (and this is only for the test, not the way you review them), do the test in reverse order (recognize the Kanji).

A third idea would be to do the test in the right order, but get the primitive names listed in the question. Then I think it would be a pretty easy test. Useless, as far as actually measuring your level of knowledge, but easy.

I asked, my dad said that would give me a very reasonable shot of actually succeeding, which would defeat the point of the bet. LOL

uisukii Wrote:

Crawdaddycon Wrote:

blackbrich Wrote:Ok that's more manageable. For the love of god ask him not to test the kunyomi. If he does your chances of winning are quite low.

You will probably still need to eat and breathe kanji though. To learn meaning and reading.

At least I am getting a few hopeful words here and there. I appreciate that.

So how is your old man going to test you at the end? I mean, I started and finished the initial review for RtK1 in 14 days, but I didn't "know" any of them. I mean, I could write them all out, and with (from memory) about 80% retention, or so, using English keywords in Anki. Wouldn't be able to tell you their common readings, though. Well, maybe a few, but that was due to unrelated factors.

We didn't have these specifics listed out initially, but because I changed the initial conditions of the bet he said I needed to write out the kanji with the keyword he gives me. He said the test would consist of 200 random kanji. I talked him down to 100 equally spaced throughout RtK Volume 1. Because of the 95% required retention rate, I'm allowed to make "minor" errors to the kanji when drawing them out. I don't know what minor means.

Crawdaddycon Wrote:We didn't have these specifics listed out initially, but he said because I changed the initial conditions of the bet he said I needed to write out the kanji with the keyword he gives me. He said the test would consist of 200 random kanji. I talked him down to 100 equally spaced throughout RtK Volume 1.

This is good to hear. Better than RTK1 + RTK2 or 4000 Kanji.

Crawdaddycon Wrote:Because of the 95% required retention rate, I'm allowed to make "minor" errors to the kanji when drawing them out. I don't know what minor means.

I have a feeling your dad will be lenient on your Kanji test. Goodluck and let us know the result.

Haych Wrote:I've got some experience with this. I did RTK3 (1000+) in 10 days. It wasn't a painful experience. I was just motivated because I was thinking in terms of progress-- not time spent, and progress was going fast. If you think in terms of hours, and just say "I am going to study 12 hours a day for the next 4 weeks", that will turn into 12 hours of drudgery real fast. Instead, think of getting the most out of every hour. You need to timebox. Try to get the numbers up on a 5 minute window. Monitor how fast it takes for you to do 100 reviews and try to bring that time down. Set up a baseline of 60 new cards a day and do it AFTER your old reviews (because new card adding time varies more than review time). Do more if you feel up to it. If you're using anki, add new cards in batches of 20, and monitor the time for that too. With all this, I bet you could probably reach your goal in about 5 hours/day.

As for the stories, think about the KEYWORD first and find a story that uses it in an ACTIVE way. It is absolutely critical that just thinking of the keyword will evoke the story in your mind. The use of the primitives within the story is important, but less critical. Focus on highly visual stories rather than joke-y ones (unless you think the joke is especially memorable). Save time by using stories from the site as often as you can find good ones.

It's going to be harder for you because you don't know the primitives already. That might become a hindrance for you. RTK1 expects you to learn them by rote. If its a problem, I'd advise looking up a list of RTK primitives and doing a bit of practice with them on your downtime.

Also, you said that RTK2 is part of the deal, but no one really studies RTK2 by itself. The book is essentially a reference text. This site has recently included a common on-yomi reading with each kanji. I think learning that, and doing it as you go, is probably your safest bet. As for how to memorize them... I'd say just do it by rote unless you keep forgetting. In that case, include it in the story in some way. Honestly though, I think you might want to bargain with that portion of the bet. Learning on-yomi out of the context of vocab has limited benefits, since there are many of them per character, and with an unknown word, you don't know which should be used. Its still sort of a good thing to do, though. It will make the vocab-learning process a bit easier.

Very good advice, Thank you. The test he is giving me will be over only the kanji from RTK Volume 1.

bimspramirez Wrote:

Crawdaddycon Wrote:We didn't have these specifics listed out initially, but he said because I changed the initial conditions of the bet he said I needed to write out the kanji with the keyword he gives me. He said the test would consist of 200 random kanji. I talked him down to 100 equally spaced throughout RtK Volume 1.

This is good to hear. Better than RTK1 + RTK2 or 4000 Kanji.

Crawdaddycon Wrote:Because of the 95% required retention rate, I'm allowed to make "minor" errors to the kanji when drawing them out. I don't know what minor means.

I have a feeling your dad will be lenient on your Kanji test. Goodluck and let us know the result.

I hope so, but knowing him I still have a very high chance of losing the bet even if I do complete the course, but as you said, he will likely go a bit easy on me if I forget a drop on a rather difficult kanji.

Animosophy Wrote:So that's 2000 kanji in 28 days.

I maintained a pace of 100 new words/day for 10 consecutive days while I did RTK (before life interruptions and whatnot). You'll need to go for triple that, which means no distractions. I hope you don't have any responsibilities.

I just about maintained 95% retention on any given day by using Anki's custom reviews on top of regular reviews the day before (review cards first > 100 new cards > Custom Review (cramming) cards from the last 2-7 days, depending on time/balls). I did only one >1000-card review day, and although challenging, it was very rewarding. You'll need to do at least two of these right before you're tested in order to cover all 2000 kanji.

Writing the kanji while you review (you'll need to do this for proof) will slow you down to a pace of ~150-200 cards/hour. I averaged about 180 review cards an hour (the green number in Anki, not total number of cards seen).

100 new cards studied at a natural pace (for me), with music, with koohii's mnemonics, took me 3.5 hours average.

If I could do it again, or if I were you (which I won't and which I'm not) my study days would look more or less the same:

Wake up
Start the playlist
Cook a day's worth of food, keep it beside desk, bring water
Kanji reviews (~150-200 cards/hour)
100 new cards (3.5 hours)
Breaks whenever mind wanders.

Optional except for the last few days: mass cramming (~200 cards/hour)

Good luck :p

Edit: I'm not really concerned about whether you continue to review after the bet (which is necessary if you want your efforts to mean anything, ESPECIALLY if you lose the bet by a small margin), I just think a laptop and bike is 10x worth the effort, haha.

Oh and you may need a playlist. I was airbanding in front of Anki at least a third of the time. It doesn't matter how passionate you are or what incentive you have; when you're spending 7+ hours a day on kanji, you need energising music to maintain a good mood. I could share the one that kept my going until the end if you cared at all, but your tastes may vary.

Great reply. If you would share that playlist with me that would be awesome. I'll start using your strategy tomorrow as it seems really well thought out.

I don't have the money to lose this bet lol.

Crawdaddycon Wrote:I don't have the money to lose this bet lol.

That would be ... a problem. I suggest you need to be studying -almost- constantly. Whenever your unfocused - just woke up, mind is wandering during studies, etc., set a timer (kitchen timer, desktop timer, cel-phone timer, whatever) for a 20 minute break and then return to studying for at -least- 20 minutes. (You can play with times, but... to be honest... you don't have the free time to have the luxury of working it out. 30 days from now you can search around the web for 'timeboxing' strategies for future studies.) Physical exercise in some of those 20-minute mental breaks will probably help.


If you have school, work, sports teams, etc, then you're basically screwed, because you don't really have time for that.

I do agree that learning keyword->kanji is the -only- way to learn to write the kanji and is the better way -long-term- to recognize the kanji, however I definitely -don't- think it's a good idea to review -only- keyword->kanji and then be tested kanji->keyword. I've never tested kanji->keyword and I found my kanji->keyword recognition was quite low in general reading, even though the character would feel familiar. After looking up a character just once or twice the keyword becomes quite recognizable for me (until forgotten in favor of actually Japanese vocabulary but that's a longer-term story). I think reviewing kanji->keyword just a couple times near the end is enough, but to go into a test cold with -no- practice quizzing in the same direction as the test would, IMO, be suicide.

(perhaps my brain doesn't work like other human brains though. I only have my own learning experience to draw on here.)

.

Arupan Wrote:And I don't see why he has to spend more than 3-4 hours a day. It's just the writing people...

Following a standard SRS schedule, the number of reviews quickly becomes -very- large when you're learning hundreds of items per day. -Not- following the standard SRS schedule or a similarly rigorous schedule of a different pattern means not reaching the 95% success rate.
I don't know of anybody that has completed RTK1 in 30 days in 3-4 hours per day with a 95% success rate.

SomeCallMeChris Wrote:

Crawdaddycon Wrote:I don't have the money to lose this bet lol.

That would be ... a problem. I suggest you need to be studying -almost- constantly. Whenever your unfocused - just woke up, mind is wandering during studies, etc., set a timer (kitchen timer, desktop timer, cel-phone timer, whatever) for a 20 minute break and then return to studying for at -least- 20 minutes. (You can play with times, but... to be honest... you don't have the free time to have the luxury of working it out. 30 days from now you can search around the web for 'timeboxing' strategies for future studies.) Physical exercise in some of those 20-minute mental breaks will probably help.


If you have school, work, sports teams, etc, then you're basically screwed, because you don't really have time for that.

I do agree that learning keyword->kanji is the -only- way to learn to write the kanji and is the better way -long-term- to recognize the kanji, however I definitely -don't- think it's a good idea to review -only- keyword->kanji and then be tested kanji->keyword. I've never tested kanji->keyword and I found my kanji->keyword recognition was quite low in general reading, even though the character would feel familiar. After looking up a character just once or twice the keyword becomes quite recognizable for me (until forgotten in favor of actually Japanese vocabulary but that's a longer-term story). I think reviewing kanji->keyword just a couple times near the end is enough, but to go into a test cold with -no- practice quizzing in the same direction as the test would, IMO, be suicide.

(perhaps my brain doesn't work like other human brains though. I only have my own learning experience to draw on here.)

I have a treadmill I run on daily but other than that I have no obligations for the next month.

I also get what you're saying. When I learned spanish the translation from spanish to english would be difficult if my method was to learn english to spanish. When I practice I'll make sure I confront it from both keyword-->kanji and kanji-->keyword. Thanks for the advice.

I'm sure someone knows where to find the RtK Volume 1 Anki deck. I can't find it in english... German, ja, but not English.

.

SomeCallMeChris Wrote:

Arupan Wrote:And I don't see why he has to spend more than 3-4 hours a day. It's just the writing people...

Following a standard SRS schedule, the number of reviews quickly becomes -very- large when you're learning hundreds of items per day. -Not- following the standard SRS schedule or a similarly rigorous schedule of a different pattern means not reaching the 95% success rate.
I don't know of anybody that has completed RTK1 in 30 days in 3-4 hours per day with a 95% success rate.

Could you please explain to me what a "SRS schedule" is?

Arupan Wrote:Why does he has to use any kind of software? He just has to learn X new kanji and review those he learnt the previous day. He'll finish in 2-3 weeks anyway, so he'll be left with 1+ week for general review.

He doesn't -have- to use software, but unless he already understands SRS theory and how to arrange physical cards and is willing to take the effort to move physical cards from box to box, he won't get the benefits of SRS scheduling. Ultimately the only reason to use software over moving physical cards around is because software is faster and more accurate, but SRS is in fact a scheduling plan not a computer program.

If, OTOH, rather than software vs. physical, you're suggesting he spend 3 weeks inventing mnemonics and 1 week reviewing them at the end, I think that won't work well as studies do suggest that material learned once and then not reviewed until 3 weeks later is unlikely to be remembered, so he'll lose those early characters that aren't also primitives.

As for an Anki deck, try this, https://ankiweb.net/shared/info/2654705267
Of course, you'll want in this case to suspend all cards after 2000.


Edit:
SRS is a complicate topic, but most people mean something derived from the Leitner system when they say SRS,
http://en.wikipedia.org/wiki/Leitner_system
Essentially, the idea is to review newly learned material the next day, two days later, four days later, eight days later, etc. resetting any individual fact that isn't remembered on review. Software can adjust timing more subtly than simple doubling (or simply multiplying by any other factor.) Plus physical cards means maintaining a written list or well-organized box.

.

uisukii Wrote:

dizmox Wrote:How to write the kanji and their general meaning..

Which volume of Remember the Kanji are you referring to?

The first one

Arupan Wrote:Oh, you just had to write -have- there. Look at the the edit time.

??? My edit time or yours? I don't know what you're talking about.

Quote:I see. People can't determine which kanji they suck at by themselves for 1 week, so they have to use a software. That sums it up.

Well, yes, basically. If you have an enough personal organization with physical cards or mental organization with hundreds of facts, then perhaps the software is not a great benefit. If you don't have that much organization, then it is a great benefit. But when I said you don't -have- to use the software I meant it.

There are other ways to learn, they just may not be as efficient and may involve more effort in personally doing record-keeping instead of letting the software track records for you. For this sort of cramming of facts though, I think the software is of benefit to the vast majority of people. If we were discussing 'learning Japanese' in general, maybe not, but when you discuss 'learning 2000 facts in 30 days' ... SRS is good at that and it's easier to SRS with software than with personal record keeping.

dizmox Wrote:

uisukii Wrote:

dizmox Wrote:How to write the kanji and their general meaning..

Which volume of Remember the Kanji are you referring to?

The first one

I see. I've got a few questions in mind but they would only disrupt the thread. Might look to see if the search function could answer a few of them.

Vempele Wrote:Do you have a source for your claim that production is better than recognition for learning to recognize?

Yes, Heisig's book. He makes it very clear how it should be used (by producing the Kanji). As for whether that will lead to recognition of the Kanji, I tested it myself. So has everyone on this site who did RtK and is now recommending it to others.

We all did the method by producing the Kanji, and at the end we were all able to recognize the Kanji.

Note that recognizing the Kanji is not the same thing as being able to read Japanese (or any other language written with Kanji). That, of course, requires practice.

Vempele Wrote:It also presents the kanji in a convenient order, tells you what they're made up of and gives them names.

The Kanji are made up of strokes, written in a specific order. They're also organized by radicals.

But they're not "made up" of Heisig's primitives. Heisig's named primitives are his own invention: they ARE a mnemonic device, not actual components of Kanji. If you told a Chinese or Japanese person about them, you'd have the same reaction as mentioning any other mnemonic device you came up with yourself, to learn the Kanji..

If someone learns the Kanji using Heisig's primitives but not any stories, they are using Heisig's mnemonics, they're just not using them to their full potential.

Splatted Wrote:It's no worse than listening to someone who gives a bogus reason for an arbitrary claim. As for my reasoning, I thought it was obvious: doing something you haven't practiced is harder than doing something you have practiced.

Is it? Would climbing Mount Everest be easier with practice than eating a muffin without practice?

Not everything requires practice. Some things you can do without practice. For instance, if you know the names of the primitives and have a good story associating them with a keyword, you can recognize a Kanji and name its keyword. No practice is required.

As for the most efficient way to learn something, according to James Heisig (who writes specifically about learning the Kanji) and almost every expert on learning I've ever heard from (speaking about learning in general or about learning various specific things), it's by producing that something, not by recognizing it. That is also true in my personal experience.

Crawdaddycon Wrote:I asked, my dad said that would give me a very reasonable shot of actually succeeding, which would defeat the point of the bet. LOL

Yeah, from the sound of it, you've been trying to con a hustler . Just get a job and fill your own damn wishlist, instead.

Ahahah... sorry, found that more amusing than I probably should have.

Stansfield123 Wrote:We all did the method by producing the Kanji, and at the end we were all able to recognize the Kanji.

Of course. But were you able to do so faster, with less total effort or with a higher retention rate than than someone who only did recognition?

I did RTK1 in 5 weeks, recognition only, about two hours a day with inferior tools (laggy website, uncustomizable SRS schedule, ready-made stories were scarce after 1500 or so). From what I gather (and from what Heisig says - "20 or 25 characters per day would not be excessive for someone who has only a couple of hours to give to study"), it takes most people a lot more effort than that to learn the kanji by learning to write them.

Just an out of the topic maybe advice.. But I can swear it's effective... Don't think that you have a bet with your dad.. Think of it as a challenge to yourself,.. Just saying though...

Vempele Wrote:Of course. But were you able to do so faster, with less total effort or with a higher retention rate than than someone who only did recognition?

All I can tell you is that I'm confident that I would have had a harder time doing it that way. All the evidence I have (and mentioned in my previous posts) points to that.