Hi there !

I think i m hitting the wall lots of people are talking about on this forum.

I started to learn 10 character a day for 2 weeks.
Then, 20 a day. until i reached 1300, where i had to switch back to 10 if i wanted to keep my sanity.

Before 1300, i managed to get a retention rate of 80 - 90 %, but now, it s failing below 80% each time i review things. The story seems more and more stupid and confusingly complex, the key words all similar and related to some other...

Come on: Drowning is different than Drown, Seethe different than boiling, Dart actualy use a character meaning arrow wich is already used as a keyword for something that most probably means something completely unrelated, daybreak is another thing than nigthbreak and means the opposite as one could guess it mean. measuring cup, measure, measuring box... monme...WHAT THE HELL IS A MONME and why should i care ? clean vs cleanse, lie down vs lying down , speakth and sayth, and it seems to goes on and on and on.

I m extremely dissapointed by the rtk at this stage, and for now 3 days, each time i study again, i feel frustration at the bad choice of keywords. As a result, of course, i dont memorize things as correctly as before.

So, to all of you who went there and did that: When does it start being sensible again and stop being this fuzzy mess of similar ill chosen keywords ?

It's frustrating. If you really can't get over seethe vs boiling, then replace that keyword with a Japanese word that has that kanji in it (煮る "to boil" and 沸く "to get hot/excited"). But if you don't know the Japanese word in the first place, it'll probably just make it more confusing, because you have to memorize both sides of a new flashcard.

Anyway, someone else can help you better.

Let me add, the ideal retention rate is around 80%. It's high enough that stuff is sticking and you're making progress, but low enough to indicate you're keeping the materiel challenging, which is great.

I've heard other people complain about the 1300s. It's a rough patch. But power through, and you'll be reading Japanese before you know it. If you really want to see the power of RTK, try reading an easy manga (I'd recommend 夏目友人長) and see how much you understand through sheer kanji recognition. That should get you pretty excited about the magic you've been doing.

がんばって！

You're at 1300 now? I say keep reviewing without adding anything new, until you have a free week to spend on this. In that week, add 100 a day, each day, and that will be that.

If you have a free week right away, then you should do it right away.

nightbreak Wrote:monme...WHAT THE HELL IS A MONME and why should i care ?

I think it's a measuring unit. And you shouldn't care. Suspend it.

nightbreak Wrote:clean vs cleanse

Suspend one of them.

nightbreak Wrote:lie down vs lying down

Suspend one of them. But I don't think those are both keywords, I'm pretty sure at least one of them is just a primitive. Primitive names can be changed. There's a wiki page around here somewhere with alternate names that are much better, that are used in the top stories.

nightbreak Wrote:speakth and sayth,

Do you see the pattern here yet?

You don't even need to study all 2042 Kanji, and you certainly don't need to study them all until you learn them. If you end up studying 90% of them, and eventually learn 70% (the rest you either end up suspending as you go along, or they just fail to stick), you're still in good shape and RtK was still a resounding success.

nightbreak Wrote:The story seems more and more stupid and confusingly complex, the key words all similar and related to some other...

Come on: Drowning is different than Drown, Seethe different than boiling, Dart actualy use a character meaning arrow wich is already used as a keyword for something that most probably means something completely unrelated, daybreak is another thing than nigthbreak and means the opposite as one could guess it mean. measuring cup, measure, measuring box... monme...WHAT THE HELL IS A MONME and why should i care ? clean vs cleanse, lie down vs lying down , speakth and sayth, and it seems to goes on and on and on.

I m extremely dissapointed by the rtk at this stage, and for now 3 days, each time i study again, i feel frustration at the bad choice of keywords. As a result, of course, i dont memorize things as correctly as before.

So, to all of you who went there and did that: When does it start being sensible again and stop being this fuzzy mess of similar ill chosen keywords ?

Read my post here. It explains how I dealt with keyword meanings that I found were similar:

http://forum.koohii.com/showthread.php?tid=11732

Basically, I made a special study of confusing keywords by reviewing them in groups (by reviewing I mean looking at the keywords and writing the kanji).

As someone who finished RTK1 (and is now working on RTK3) and is now reading real Japanese, I can tell you it's worth the effort to remember the keywords because the keywords give you a way of referring to specific kanji and telling them apart. E.g., aid/help/abet are three different kanji so having three separate keywords is extremely useful (despite the similar core meanings).

Even if you study all 2042 kanji, you're going to encounter new kanji in native material. In other words you must not study all 2042 before studying "real" japanese. The problem with RtK is that some of the most frequent kanji are near the end. I did RtK lite plus some other kanji used in Genki but not in RtK lite, that sum up to 1300 kanji, and I understand how you feel. Thanks God I did RtK lite, there is no way I was going to endure RtK until the end, too much time needed. But I think 1300 kanji are enough to just make you learn the other you'll see in native material (both those in RtK and those which are not in it).

You should try studying from kanji to keyword. That's been the most helpful for me concerning similar keywords. When you see "没" or "溺", you'll know that both have to do with the word "drown". If you forget to add "-ing" to the end of the correct one, it won't feel like a loss because you almost got the word correct. Whereas, if you write the wrong character while reviewing, it might feel like more of a defeat.

In time, by doing this, you'll start to distinguish between the two. And anyway, is remembering how one keyword is conjugated so important? Remember, the end goal is to recognize them in context.

I feel your pain with similar keywords, but this is mostly an unfortunate by-product of many kanji having similar meanings. Like there are something like 6 kanji that all mean to jump or dance. So he had to get a bit creative to give every one its own distinct keyword. If you use this site to review, and not anki , you'll see on the public stories page there are a lot of suggestions for alternative keywords that don't clash later in the book or RTK3.

Since there is no real-world situation in which being able to write a kanji in response to a Heisig keyword is a useful thing to be able to do, you might as well put a hint on the front of the card to disambiguate similar keywords (e.g. the radical).

anotherjohn Wrote:Since there is no real-world situation in which being able to write a kanji in response to a Heisig keyword is a useful thing to be able to do, you might as well put a hint on the front of the card to disambiguate similar keywords (e.g. the radical).

God, this. I put the story in tiny hard-to-read gray-on-black text; IF I need to glance at it, it takes me 2-3 words to immediately know what they keyword is referring to, and to fully recall the story. I also annotate the keyword itself, adding various English meanings from Jisho, as well as Japanese references that I might know (くん for 君, おれ for 俺, なに for 何, etc.), anything to break ambiguity.

Stansfield123 Wrote:Suspend one of them. But I don't think those are both keywords, I'm pretty sure at least one of them is just a primitive. Primitive names can be changed.

I'm almost done with a RTK1 kanji dependency graph that'll make it a snap to (among other things) tell whether a given RTK kanji is used as a primitive; if not (if it's a leaf node in the graph), you know you can suspend it without worrying about whether a future kanji uses it. And if a kanji isn't jouyou and is only used as a primitive in other kanji, you can modify it as much as you wish (e.g., 勺). It hopefully won't take more than a few days to finish. It's gonna be awesome!

Hello, I believe I hit a wall at around 500 kanji, and got through it. Below is a method that might get you through it. I'm glad I went through the wall, because I think I got something out of it which will help me finish the book (currently at around 1400 and counting).

Note: the below will feel like backtracking, in that you may have to relearn several kanji that you thought you knew. On the other hand, you already kind of know these kanji, so that's encouraging. If you made it up to 1000, that's a big accomplishment, even if you think you're forgetting 20% of them, so you've already primed the pump.

1. First, stop adding new cards. No point in adding things if you're shaky, it will just stress you out.

2. Then consider the cards that you find easy to remember, you hopefully have some stories that really work for you. See what it is that makes them that way. For me, it was people, and action/movement. There are some good posts on coming up with good stories.

3. The tedious part: starting from the very first kanji, go over each one of them, and make their stories better. This was the hard part for me ... rather than just picking up someone else's ready-made story, I had to invest the time to really make it work for me. Some characters are really tough (eg. "each" ... "yearning" ... basically, anything non-concrete! ... so I try to tie these back to concrete things). This took me a good couple weeks of investment, but at the end of it I had a better sense of what was needed for me to feel I had a good handle on it.

Regarding similar keywords, I'd say change them to make it work for you. I look at jisho.org for kanji, and view some of the words for a kanji, and try to come up with my own keyword (eg, "transparent" wasn't working for me for stories, so looking at jisho.org it seemed that "look through the gaps" was a better meaning, and I could make a story for that). Bend the keyword to make a story, sure, don't go too far off base though.

4. My Anki (assuming you're using that) settings weren't that good ... I was only looking at a new card once, and then it was going into review. I changed it so that my "new card" settings are something like "10 30 60 120", so I'll see it a few times on the first day.

5. One thing that's important for me, and might be for you, is to make a clear distinction between "learning" time and "reviewing" time. By this I mean that you might just be adding new cards without giving much thought into them before adding them, so when it's first shown to you you're just learning it. For me, didn't feel as good as spending time learning it, and then putting it into Anki. What I currently do is suspend every card I've not looked at yet, and only unsuspend them when I've put in the time to make a good story (and adjust the keyword etc). This means that when it first shows up, it's really a review, and not a new thing. It helps for me to do this the night before the card will first be shown, because it lets the story bake a bit in my brain overnight.

Best of luck. You're already quite far along, and if you manage to batten down these loose concepts you'll likely be the better for it. It's a slog, you can do it. Slow and steady wins.

ps - "monme" really is silly.

71% today. Most of the one i saw the past couple of weeks did not stick at all. That s it, i'm stopping learning new one. I ll have to find better story for the one i fail, and will wait to get a more decent retention rate before trying to learn new one.

Some comments about the tools i use. Actually, i use both anki and this site. On the beginning i was only using anki and set the deck to show 10 new card a day. Then, i went a bit faster with this site, and decided to just try to see how things would be with 20 a day, and use anki as "the second blade" of the learning razor. My retention rate for the original stack is still really good (97%) but i am only at 740 cards there, plus there is the biais of card i may have seen recently on this web tool.

I'm not waiting to finish rtk to study japanese. Actually, i waited for a year since i started to study japanese before deciding to spend time with kanji. I'm already reading stuff daily (my main exercice book is a bible i got that comes with furigana. It s much more easier to read something i'm already familiar with the text. I also like nhk easy web, but it still is difficult to understand).

Since i started rtk, reading has been much more intersting. Before, i never associated character with words (It took me 6 monthes or more to recognise 食 and that s the first one i recognise i think), now, i sometime "hit" on a word i have been looking at for monthes before ever realising what the kanji where about. Remembering them and associating words will be much more easy, so i understand how interesting being familiar with the 2200 set is, however, my brain just cannot stand the pace i was trying to follow.

I do not plan to spend any time kanji to keyword. I want to recognise character when i read stuff with furigana, and memorise real japanese words this way, without any proper method right now. When i ll have finish rtk, i may start looking at secific words, probably using jlpt level to gain vocabulary.

Either use this site or Anki-- don't use both. That's part of the problem. Reviewing more doesn't make it better, reviewing at the right time does, and letting either this site or Anki just "do it's thing" is the best way to do it. (This problem pops up a lot. People try too hard when they don't have to.)

If a keyword doesn't work for you, then by all means, CHANGE IT! There is no RTK referee who will give you a yellow or red card for making stuff up as you go along. RTK itself is just one guy's study trick that worked really well for him, so changing it to fit your needs is not only fine, it's a lot better than using things that don't work for you. (Also, isn't Jim Heisig a professor of religious studies? So of course his keywords will suit his inclination!) Make it work for you in a similar way if you feel like you have to.

I'm a big believer in dumping whatever doesn't work, or bending it into something that will work for me. RTK is especially flexible that way.

Get the keywords in your brain as best you can, and don't lose sleep over it. It's just something to make the rest of your Japanese study easier. It's not an actual end in and of itself. It's only a trick to make the real thing easier.

nightbreak Wrote:I do not plan to spend any time kanji to keyword. I want to recognise character when i read stuff with furigana, and memorise real japanese words this way, without any proper method right now. When i ll have finish rtk, i may start looking at secific words, probably using jlpt level to gain vocabulary.

But isn't keyword to kanji practicing production? If you are interested in recognizing kanji when you read them, it makes sense that you would want to practice recognition - Hence kanji to keyword. As a bonus, most of your present problems will be minimised and things will go a lot smoother and faster.

yogert909 Wrote:But isn't keyword to kanji practicing production? If you are interested in recognizing kanji when you read them, it makes sense that you would want to practice recognition - Hence kanji to keyword. As a bonus, most of your present problems will be minimised and things will go a lot smoother and faster.

kanji to sound or kanji to real word is the goal, not kanji to arbitrary keyword.


The only goal i set for rtk is to be familiar with kanji, and write them.

Once i m familiar with a given character, the next step is to learn how to read it, but that will happen with exposure to real sentence, real word, real sound.

My only present problem are the bad choice of keyword who are either completly esoteric or too similar to some already used keyword, and the fact my brain does not store data as fast as i wish it would. If i was to try to remember kanji to a keyword on top of all i am trying to remember it would be more effort, and no gain (as i dont plan to read text by guessing what it may be talking about giving a list of keyword).

Quote:The only goal i set for rtk is to be familiar with kanji, and write them.

Going kanji-to-keyword now doesn't preclude doing it the other way around later, once you're familiar with kanji*. You'd be able to largely skip the story business (in both directions) in favor of visual memory, and you could use Japanese keywords for the eventual keyword-to-kanji part.

*And I think this way would be substantially faster than the one-pass approach if done from the beginning. Switching at 1300, not so much.

Nightbreak, there are French keywords, and maybe they will be easier to use? Quite often it's not clear if Heisig's ones are verbs or nouns or adjectives or what, and I found that confusing too, so I edited them a bit to make their meaning clearer. I have a kanji look up option on my dictionary app which I found helpful to understand what heisig meant by a keyword.