I have just started RTK book 1 with very little knowledge of Japanese.
I'm good at remembering vocabulary (pronunciation). I can, with relative ease, learn and retain 300+ new words in 2 hours.

I memorised major parts of the body, along with location words to get familiar with Japanese. With this, I have a 98% success rate in anki on new words.

However, this kanji stuff is really hurting me!
I've seen many people here say that the first 300 or so kanji are easy. I'm really surprised by this.

I have been making many terrible mistakes such as putting the primitive in completely the wrong place.

And ive only learn the first 50!! My anki score is 72%. The first 50 or so took me 1 hour to learn to this standard. I literally have the entire day free, as I'm on medical-leave from my college.

My main finding so far is that Heisig's stories are not for me.

I have been using mnemonic systems for a few years, this is probably why vocabulary is easy for me to memorise, yet kanji is so new that im shockingly bad at it.

It seems the most difficult part is identifying the primitives, and putting them in the right location. I have a far higher success rate when I use my own stories/images, so I'm starting to go through them again adding my own.

Is there hope for me in this regard? As I'm beginning to feel a little hopeless.

No need to feel hopeless. For kanji, a normal retention rate can be anywhere between 50% and 90%. I've always averaged 25 learned kanji per hour. And as Heisig mentions himself, his stories are just cues to illustrate the method, you're supposed to devise your own if they don't work (or use this site's).

Both words and kanji derive meaning, but there's a more visual-spatial aspect to kanji. The mnemonic approach is different in that each kanji is a single 'picture'.

The stories are an issue, I presume.
Free association works best. Another option is to use this site's story database, as EratiK has mentioned.

In this article, Kushner describes her own approach to kanji.

You don't need to copy anyone else's cues.. You can formulate your own... Those were just suggestions, not the topic itself.. good luck!

Thanks for the suggestions.

I just learnt another 25 kanji and my anki "again" count for this was 51!!!!

Yes, it is the stories that are a big problem. I'm not remembering the stories at all it appears.

I have fanciful stories/images for vocabulary words which never disappear. I don't know why this isn't happening with kanji? Maybe because they don't actually look like anything? grrrr

For example, my image for こぜに (small change) is a child (こ) standing in a zen (ぜ) temple waiting in between two (に) vending machines to get his small change. It's *small* change because it is a child.

Now this seems convoluted, yet I've never once forgotten this. All of my other vocab stories, over 500 now, are as weird as this and are almost never forgotten; 98% retention.

Can someone explain why I'm bad at making stories with kanji, but very good with vocab?

Well, I'm looking to do 100 per day as I have buckets of free time.
I suppose I'll update this thread...

Are you doing both keyword>kanji, kanji>keyword? This might be why your count is double. In that case you should suspend all the kanji>keyword cards.

I don't think vocabulary images are substantially different from kanji images. I guess you will have more ease with more complex kanji. The kanji at the beginning have very few strokes, and most are the traditional radicals. I never had stories for moon, ricefield, or hand, in my opinion you have to more or less rote memorize those, which you will see is easy over time because of how frequent they are.

Also it's important to be relaxed. Give it time, let the SRS and your muscle memory do their magic.

Seems like your problem is mostly with the position of the primitives. As a student of superior study ethics, I have never done this and can't recommend it of course, but I think you should know that some people pass the kanji even if you have a minor mix up in the positions.

Even though I have 100% guaranteed never done this, I can understand where they are coming from. I mean, you are probably going to get it down eventually anyways from exposure, so really, whats the big deal? Hypothetically..

I'm not doing kanji > keyword. I'm only doing keyword > kanji.
I just keep getting the same kanji wrong over and over.

I just quit anki in frustration over this. So now I'm going back through lesson 2 and 3 in order to very careful create memorable stories.

For example, I'm having major issues with the story for "ten thousand". In my opinion, Heisig's story is awful, I barely understand it :-P

I'm also worried about creating my own stories as it seems his book builds upon his own. Eg, there was at one time he called a part of the kanji a "needle". If I changed this part of the kanji to something else then I will be screwed in the future; the author basically says this.

This is another reason I'm very hesitant in creating my own stories that are nowhere near his suggestions...

Similarly, it seems that stories cannot give you the exact position of the strokes. I put many strokes in the wrong place. I'm not sure how a story can encode the positions of each stroke, or the precise placement of each primitive.

Perhaps I rushed... Anyway, I'm going back :-P

Just to make sure, you have taken a look at the stories that this site provides, correct? That is the main function of this place. Most of the top rated stories are excellent.

If you don't like his stories, it doesn't matter that you won't be able to build on them. However, before naming an element something else, you should check the appendix at the end of the book that lists all of the keywords.

You probably don't want to use a keyword that he uses for something else, or at least you want to know that you're setting yourself up to need to change the keyword. You also should keep a list of what you changed the keywords to, so that you can keep the same change consistently and not forget and start using his keyword again.

I always tried to make my stories use the elements in the order they are written, and every once in awhile that wasn't enough and I needed to put explicit positioning into the story. (or still do, really, as I'm turning my cards into Japanese-keywords and the Japanese-keyword isn't always amenable to the original English-keyword story.)

[ファブリス]

Doctorate Wrote:I'm also worried about creating my own stories as it seems his book builds upon his own. Eg, there was at one time he called a part of the kanji a "needle". If I changed this part of the kanji to something else then I will be screwed in the future; the author basically says this.

Those are primitives. Pay attention in the book when Heisig puts a label to a combination of radicals.

You can change a primitive meaning if you want, or add a new meaning of your own. Just be consistent with it that's what's important. A good example that comes later is "person". Many people came up with their own. I chose "Mr T." It's possible to use more than one, eg. "Mr T" and "Superman". You can easily remember both are the "person" primitive. Usually I'd say it's best to use no more than two labels for one primitive, ideally only one.

Like Fabrice said, don't go confusing keywords, primitives, stories.

Keywords are global names given to kanji (often reflecting either the global or a specific meaning of the kanji); stories are images to remember the keywords; and primitive is word used by Heisig to refer both to radicals AND frequent simple combinations that aren't really acknowledged traditionally. Radicals usually have a clear meaning (like hand, vehicule...), but not the combinations like the right side of 渇 ( called "siesta" by Heisig) or upper one of 壁 ("ketchup"). These you can freely name.

As for needle/10/cross, the name is because it's later found in 針 needle. But also like Fabrice said, with a little bit of practice, you'll be able to associate more than just one meaning with primitives.

10 000 is a perfect example of the few you should almost rote memorize. Devising a story too complex for 3 strokes is a waste of time. (in my old jouyou kanji book, 31 have 3 strokes, that's less than 2%; if you can't think of a story they can easily be rote memorized).

Thanks. I will take all of your good advice.

I just did a review of 60 kanji and I am now at a 92% retention rate in anki.
I went back and made my own stories, as well as using some excellent stories from this site, so it seems I'm not doing too badly anymore.

I'll try to do another 70 today!

Not that I know what your goals are but it sounds to me like you're taking the numbers more seriously than necessary. I'd say as long as you're using a good SRS like anki, give it time, keep going, and it'll all eventually sink in. Even if a story is weak, theoretically you'd be making up for it by balancing it out with more reviews.

The stories aren't all meant to stay memorized forever, they're clutches until the characters themselves are memorized. Even Heisig himself has said that in the end, the kanji have just ended up being Japanese to him.

Yes, that sounds right.
I just get frustrated very easily if I'm repeatedly getting the same characters wrong, or totally blanking out on a set of characters over and over.

A few points, related to various things said in the thread:
1. Don't worry about forgetting the positioning or primitives. It happens early on, but by the time you're done with RtK and start using the Kanji, you'll know the positioning. It just happens naturally, as you learn more and more Kanji, because there's a simple order to where everything is that you'll pick up without even trying.

So, whenever you know the primitives but not the positioning, just take a quick glance at the answer before you write down the character to make sure you don't write it down wrong (so that you don't reinforce a mistake), and then set the card as good anyway.

2. Yes, Heisig's stories are fairly bad, the stories on this site are much better. Use them, or make your own for the first few hundred Kanji.

3. You don't really need to know all 2042 Kanji at the end of RtK. If a character frustrates you, suspend it. Don't overdo this early on though.

4. The stories are key. It is very important to know the story, for quite a while, until it eventually fades away. That's why it's important to have vivid, interesting stories for each keyword. And, don't forget, the keyword must be your anchor for the story. As soon as you see a keyword, you must know what its story is. Then, figuring out the primitives from the story becomes fairly easy.

5. When you do the new cards is important: always do the reviews of the day first; then pick out the stories for the new Kanji you're gonna add, and study them carefully; then take a break, and "learn" the same Kanji with Anki. In other words, don't just mix together the old reviews and the brand new Kanji. That makes it harder for no reason.

6. Consider adding clues to the more difficult Kanji. For me, the clues were usually there just to specify the sense the keyword is used in, in the story. Sometimes, they were also hints of what the story is about. As long as the hint does not tell you the actual primitive, the clues DO NOT take anything away from the method. All they do is make the keywords better, by giving them a more concrete meaning than just what the English word itself gives. This takes nothing away from how well you know the Kanji. It's a no lose modification of the method, the only thing you're losing is the time spent adding the clue to the deck (which is why it should only be done when you actually need it).

Doctorate Wrote:My main finding so far is that Heisig's stories are not for me.

I have been using mnemonic systems for a few years, this is probably why vocabulary is easy for me to memorise, yet kanji is so new that im shockingly bad at it.

Remember that Heisig is not asking you to use mnemonics, rather your IMAGINATION and thus imaginitive memory. So part of your problem could be there. In contraucting your stories, you should be doing something akin to daydreaming.

If you are having trouble remembering the location of each primitive, it's fine to be making your own stories if they help better. Heisig includes stories at first to help you see what he means by them, to see how they're different from ordinary mnemonics.

Also, try to imagine the primitives in their actual location. Eg for this character: 各 I actually imagine someone walking along carrying their mouth so it hangs below their knees, so I remember that the mouth goes below the walking legs.

Hope that helps!^^

Stansfield123 Wrote:4. The stories are key. It is very important to know the story, for quite a while, until it eventually fades away. That's why it's important to have vivid, interesting stories for each keyword. And, don't forget, the keyword must be your anchor for the story. As soon as you see a keyword, you must know what its story is. Then, figuring out the primitives from the story becomes fairly easy.

I'd just like to add to this. This is so important that you go with the first thing that pops into your head when you see the keyword. Even if its meaning is not quite in line with the keyword's meaning, just amend your story so that it includes the proper nuance too.
With so many kanji I made this mistake, and it ended up taking me longer to get them into my head because I tried not to use what first popped into my head as my anchor, because it didn't make sense, or it didn't match.