The Movie Method of learning onyomi.

I have developed a different method of learning kanji that I call the Movie Method. What's unique about it, is that it's possible to learn the onyomi at the same time as the kanji with minimal extra effort.

Basically, you learn the kanji in groups of onyomi, and put them in locations that correspond to onyomi readings. For example, all the kanji pronounced カン will be put in the Ocean's movies. This makes it necessary to learn kanji without the benefit of Heisig's order, but it's really not that hard, actually.

What's great is that I've personally finished learning the RTK1 kanji successfully with this method in fifty days (rate of 50/day), and I'm now able to pronounce (EDITSmile most compounds I come across. Another side bonus is that I don't have to study signal primitives at all, because I've pretty much learned all of them without trying.

I'm documenting my methods on my blog above (also as my website on my profile), and plan to add more information on how to do it. Ask me any questions and I'll answer/put in my queue to write an article about.
Edited: 2008-08-09, 11:19 pm
I wouldn't be so quick to say you can read any compound you come across >_> Japanese is a hard language for a reason. Not only do many kanji have more than one on-reading, but there's also the possibility the compound you're looking at uses kun-readings, or an irregular reading. Can you guess how the word 様子 is read? This one stumped me until I looked it up... Maybe you already know, but the on-readings can change depending on the reading of the other kanji being used in a compound(ex. 薬局 is やっきょく not やくきょく).
Oh yeah, it's easy to learn the multiple readings. I'd guess for 様子, it would be pronounced either ようし or ようす. I mean, I already know both possibilities, it'll be pretty easy to remember which it is when I learn the word.
August Sale (14th - 25th): 30% OFF Premium PLUS - 25% OFF Premium
Very interesting method. In your example of KAN, there are actually over 20 kanji with that ON reading. Do you squeeze them all into that trunk scene or do you create new scenes from the same movie?
I put each kanji in it's own location in the movie. KAN has a lot more than twenty to learn, somewhere like forty or fifty, which is why I choose all three Ocean's movies. For the larger groups like ショウ and カイ, etc, you have to pick movies you really know well.

The trunk scene represented one kanji. For 漢 (Sino-), I place in the scene where they're meeting with Yen. (The little Chinese guy, the grease man). For the kanji 堪 (endure), I put where George Clooney and the big guy are in a room and he hits him. For 閑 (leisure), I put in the jail cell they're sitting in after they get caught in the second one.

By doing this each kanji is independent of one another, and chaining is unnecessary.
Edited: 2008-08-10, 2:30 am
Some kanji have multiple ON readings, I'm assuming you are just using one principal ON reading for each kanji? What do you use for reference for the readings?

Also, any problems with the more abstract keywords using this method? With Heisig's method usually the hardest thing is coming up with stories for the abstract keywords, at least for me.
Edited: 2008-08-10, 2:48 am
No, I learn multiple onyomi readings. I made a new post on my blog in response to Yukamina describing how I learn them.

I used a combination of the Kanji learner's dictionary and the index for RTK2. But I wish I could have found a better list, since I found a lot of inaccuracies (first edition) and it was limited to the readings for compounds he gives.

Abstract keywords: It was easier to find abstract keywords, because you have an entire movie to derive action/intentions/ideas from. Not only that, but you can imagine characters performing actions for things, even if the scene is way off. Like I put "vote" in Once Upon a Time in Mexico by imagining the president in the scene where he picks a song.The hard ones were things like "Imperial we" (朕) to put in the movie Signs. Or "pear" in the movie Red Planet. But I managed, since when you really can't do it, you can be arbitrary and let the SRS teach you the meaning.
Edited: 2008-08-10, 2:57 am
I must be really slow because I don't seem to be getting it. (Or maybe I am, and just don't understand how it works.)

Do you make a movie clip with the kanji on screen, or make a still image with it, or write down the details? Or does it all just stay in your head, and making up the scene 1 time is enough?
It sounds interesting, but I'm having some trouble understanding as well. You see a kanji in a newspaper, and then you immediately remember a certain movie, a certain scene in a movie, and the ON reading you assigned to that particular scene? And you have to do that twice (since ON usually comes in compounds) for every word you encounter?

Sounds like you have a photographic memory, in which case you don't actually need any method. Cool

Anyway, it works for you, so kudos for 2000 kanji in only 50 days.

But maybe you can describe for us step for step what you do when, you see for example, the word 確認. (Picked at random, but was curious how abstract keywords and pronunciations come together in a movie).
Edited: 2008-08-10, 9:23 am
Each movie has an onyomi assigned to it, not individual scenes.

Basically, I find ways to change scenes from movies by incorporating elements into the scene that plays out. I only need to do it once, and I don't need to write it down. It's a lot like how memory palaces work.

As to how much I need to recall a scene to remember the onyomi, think of it as like when you need to recall a mnemonic to remember the meaning of a kanji. If I'm familiar with it after seeing it in context enough, then I don't need to recall anything, I just know (for example, 国 is very familiar to me, meaning and reading). If there's a signal primitive, like in 認 (ニン), then that's pretty easy to recall without thinking. It's about the same as how well normal RTKers can recall the meaning of a kanji. Eventually I won't have to recall any scenes or movies, and my mnemonics will fade away.

I didn't really learn words, that would have been too much for me to do then, but I would have learned 確 when I was learning the KAKU kanji, and 認 when I was learning the NIN kanji.

For all the kanji pronounced KAKU, let's say you were using Mr. and Mrs. Smith. So for 確, 'certain', we're going to find a scene within that movie to place it that has something to do with certainty. So I can imagine the scene where Brad Pitt is in his car on his way home after Angelina Jolie tried to kill him. But in the movie, see, he still loves her, so he's not very certain about trying to kill her. To remember the elements I put a floating turkey with a crown on in the car with him, and he's throwing rocks at it. I think of this once, and I usually remember till the next day. SRS takes care of the rest.

For NIN, I use The Mask. So when I want to put 認 (recognize) in there, I use the scene where he sees a big lump of garbage in the water and thinks it's somebody drowning. To remember the elements, I imagine him going down and trying to help the guy, only to find a sharpie (changed the 'word' primitive). Then the garbage somehow starts moving and stabs him through the heart with a katana.

These all take a fraction of a second to think of, and are pretty easy to remember. If it's simple, like the two examples, I pretty much never have any trouble remembering the elements. If it's complex and has a lot of elements to remember like 爆 which has five elements that are hard to create an essential image for, then I might forget an element or two by the next day. But with SRS, it's all pretty easy.

I do not have a photographic memory.
Edited: 2008-08-10, 2:35 pm
Have you guys checked out his link at all? I'm not endorsing the method or anything, but these questions leave me baffled when the answers are literally a click away. He explains it really clearly.
Yup, I read this whole thread, his whole blog, and even stopped and tried to imagine it before I posted.

I suspect the problem is that I'm don't have a good imagination. If you ask me what my mom looks like, I'll be like .. uh... blonde hair? I can't see things in my head well at all, even if I'm extremely used to seeing them.

After this last post by Alyks, it seems that's an important part of this method and I just don't have it.
Dear Alyks,

you already mentioned that for large groups you need to know the movie well. This is, I assume, because otherwise you could not recall that many scenes.
So I wonder if you choose a movie do you think about how many scenes you remember or would you just start and see how much Kanji you can place in it.
And in general how much times have you seen those movies?

I can imagine that it is an awful lot of work and I wonder how you organize it at all, because I did not find any exact explanation how you do it. I mean how you choose which Kanji to learn next and the like. You already mentioned that it is not important but I guess you must have some system...

Warmth greetings Sarina
I decide a movie based on how well I know it, and how much the reading will remind me of it. カン reminded me of the word "con", and that word reminded me of Ocean's movies. カイ reminded me of Keanu Reeves and so I picked The Matrix. Both of these are relatively big, so it worked out. There really aren't enough big groups to worry about, since most of them are ten - twenty.

I generally avoid movies I haven't seen for along time, unless there are less than ten kanji to put in. I don't try to think of how many scenes I remember, but rather how familiar I am with it. I've gotten more use out of short movies I liked a lot, than something big and long like The lord of the rings that I've only seen once. It's not the length of the movie or how many different scenes there are, but how well you know the movie. Of course one thing I liked to do was turn on the movie I was using and memorize with it playing on mute. (Good times...)
I can always fit all the kanji I need to remember in a movie, it's just a matter of good selection.

I don't have to organize it all, it's all remembered pretty well. This is really the essence of the system, lack of any organization and complete independence of kanji from one another. I'm remembering individual kanji, but I only remember as a group for convenience. If one were really dedicated, they could even do this by frequency, but that would be very difficult.

But other than that, I simply use the RTKII index as a guide to which kanji to memorize and show how many in a group. Before I sit down to memorize, I would pick out groups until I had enough to learn fifty, think of movies for each group and start memorizing one group at a time.

Edit: I made a new post on the blog. This one is relevant for people having completed RTK already.
Edited: 2008-08-10, 3:16 pm
I think that's pretty damn cool. My only question is, do you bother giving the movie treatment to kanji with unique onyomi? Or onyomi sets that only share two or three kanji? Like じょく/辱 for example. That would be nearly 300 movies you have to be familiar with. I've been playing around with the Kanji Town concept lately but I'm only applying it to groups of 5 or more, not sure what I'll do with the rest yet. Which, speaking of

alyks Wrote:Kanji-town is essentially the same as Kanjichain.
This isn't true, unless I'm using it wrong. Kanjitown is just location based memory similar to what you're doing. Maybe it seems like chaining because that guy liked to make story adventures out of it? But the way I'm using it, I create or take an existing place (locations in RPG games work great) and simply relocate the images from RTK1 to that place. Eg. サ is a beach, so I relocate all the kanji with that reading to a beach. 査 used to be a tree growing out of a shelf in a lab, now it's a tree growing out of a shelf washed up on a beach (with some scientists around it). The kanji mostly exist independently.

The reason I bring this up is I think it's pretty easy to apply a method similar to what you've used after RTK1, and without the drawbacks you mention (link dependent, time consuming, no location)
Everyone has their own approach, so whatever works best for you is just fine. I have a different take on language acquisition, so I'm not using memory palaces or any of that.

I'm a big fan of Dr. Krahsen. He's been studying this stuff for a Very Long Time, and has done a lot of interesting research on second language acquisition.

Here's a link to a writeup of his basic approaches. It's sort of a summary.

I'm a big fan of comprehensible input. It doesn't require me to build memory palaces or anything like that. I just keep on doing i+1 and letting my brain do its thing.

RTK is about as far as I'll go in creating any sort of mnemonics, and that's mostly because I'm simply using it to carve out placeholders until I add the correct Japanese. It works for me.

I don't like the idea of putting too much "stuff" in the way of my target language. About as far as I'll go is remembering that the びん in 郵便局 is the same as the べん in 便利. I never needed mnemonics to remember English words, so why should I use them in Japanese? I just let my brain do its thing and give it lots of good comprehensible input.

For remembering other stuff... well... there's Anki. Otherwise, I'm a firm believer in GTD, so other stuff that I don't want to bother remembering gets dumped into a reliable storage system, and I try to unclutter my brain as much as possible. Sure, it's great that someone can remember a 10,000 item shopping list, but all I can think is "Why would I want to do that when I have a perfectly good piece of paper or a PDA?"

EDIT: Okay, I think Signal Primitives are kind of hot... but other than that, yeah... that's about as far as I'll go. Big Grin
Edited: 2008-08-10, 10:51 pm
Rich_f, I agree with everything you said. I don't learn vocabulary with mnemonics very much. But I don't see how that's relevant in this thread. Since it's about how I learned the onyomi at the same time as I was memorizing kanji.

shakkun Wrote:I think that's pretty damn cool. My only question is, do you bother giving the movie treatment to kanji with unique onyomi? Or onyomi sets that only share two or three kanji? Like じょく/辱 for example. That would be nearly 300 movies you have to be familiar with. I've been playing around with the Kanji Town concept lately but I'm only applying it to groups of 5 or more, not sure what I'll do with the rest yet. Which, speaking of
Yes. It's not too hard as long as you maintain a balance of smaller groups and big groups.
My way of learning onyomi: don't bother trying silly methods and techniques, just learn vocabulary.

Jarvik7 Wrote:My way of learning onyomi: don't bother trying silly methods and techniques, just learn vocabulary.

Now that's just silly!
rich_f Wrote:I'm a big fan of Dr. Krahsen. He's been studying this stuff for a Very Long Time, and has done a lot of interesting research on second language acquisition.

Here's a link to a writeup of his basic approaches. It's sort of a summary.
That's a very good article. It really clarifies a lot of things I've been hearing on the web, especially about using Anki as a study tool. (It suggest that the real learning from Anki isn't the review, but rather the gathering of sentences... And I've heard that before, too.)

For anyone who is needing a little more motivation to start using their Japanese, I recommend reading this. It will definitely push you to start reading immediately and learning that way, even though it seems really slow.
Yeah, and I was talking about on-yomi, too. The last time I checked, the on-yomi were a significant part of Japanese language, so I would argue that Krashen's theories apply to those equally. So then it becomes a learning vs. acquisition question as to how to approach on-yomi. (And kun-yomi, because none of this happens in a grammatical vacuum.)

You've chosen a rather complex learning approach to it, and I'm saying I'd prefer something less time-consuming and less brain-intensive. In other words, an acquisition approach. Your approach will probably be faster, but mine will be be easier, with nothing to be forced to be remembered. I choose sources for my comprehensible input system that show various on/kun readings in a targeted fashion, and I don't fight it. It works for me.
Well yeah, you guys are already done with RTK. I would say that you're methods are fine and wouldn't tell you to use this or any mnemonics. But when I got started and thought of this before learning any kanji, I thought "why not just learn the onyomi with the kanji?". It wasn't a lot harder than RTK. I don't review the onyomi at all. All I recall is the kanji. After completing it, I really don't have to recall anything to remember the onyomi. They're just... there.

My thought process is this: it was a simple trick to add on the onyomi to RTK. But it's not like they're very hard to learn anyway. So if you want to, it doesn't mean a lot of extra work to try my method out. If you're halfway through RTK and don't want to, then more power to you, keep going with what you're doing.
I think it's a good method, and I've given it some thought and consideration.

I don't really think knowing the readings of the kanji by themselves is exceptionally useful, and if you learn vocabularly related to the characters you will know the possible readings already. I know 後 has the readings あと and ご because I often say 後で in Japanese (after) and also 食後 that I hear in restaurants a lot.

I learned two readings simply as a matter of course.

If you want to read words you don't know out loud then it might be useful I guess.

I think the biggest gripe I have with this method is just coming up with the 250+ movies you need for all the onyomi sounds. I'm not even sure I've seen that many movies to be honest. I still think it's a very interesting method but I won't be trying it out because I'm just too lazy to do all the prework required.
It is very useful to diminish the stress of learning vocabulary!