"(ex. 薬局 is やっきょく　not やくきょく)."

there is a rule for this, it isnt abnormal

in example, kanji where the last hiragana ends with ku and first letter of next kanji is k, they will merge. I can read kanji like these correctly without ever seeing them before.
作曲 heres another example, sakkyoku, sakukyoku, theres a bunch more types and what-not, but I dont know them on the top of my head

I thought this was 'make a movie' method, ie, make those short kanji videos with lots of keyword associations. I wonder what happened to that project?

@raseru
I guess the general consense here is that even if it follows any logic it only helps you acquire the word faster, with less effort.
For acquiring the language, it doesnt matter if there are any rules, as we have to find them by themselves.

So I guess alyks still have not noticed the full potential of this method. He'll see it while doing the sentences.

raseru Wrote:"(ex. 薬局 is やっきょく　not やくきょく)."

there is a rule for this, it isnt abnormal

in example, kanji where the last hiragana ends with ku and first letter of next kanji is k, they will merge. I can read kanji like these correctly without ever seeing them before.
作曲 heres another example, sakkyoku, sakukyoku, theres a bunch more types and what-not, but I dont know them on the top of my head


That's not a rule, it's a tendency.
counter-example:  幕切れ (まくぎれ) undergoes rendaku instead of gemination.
Also, the tendency is usually on the (weak in Japanese) 'u' vowel, not on く in particular.
(everyone's favorite) example: 打っ掛け　（ぶつ＋かけ.る）
It also has nothing to do with the following mora starting with a k sound.
example: 一石二鳥（いっせきにちょう）
And of course there is stuff that has abnormal readings.
example: 取っ置き（とっとき）、佐々と（さっさと） etc

Personal name kanji also totally ignore all rules when it comes to rendaku and gemination. They are pronounced following free variation with no rules or tendencies.

In short, you can't predict the readings with any sort of reliability with "rules" (this is also true with rendaku) and as such one should put more focus on learning readings through vocabulary instead of the reverse. If you have good study habits (SRS) then vocabulary study is pretty simple and the readings come free. Like I said earlier, 一石二鳥.

-edited with examples. maybe too much linguistics terminology?-

Alyks,

Out of interest, what are some of the movies you picked for the most common On-Yomi?

For those doing Kanji-chaining or Kanji-palace, what were the locations you used for the On-yomi?

raseru Wrote:"(ex. 薬局 is やっきょく　not やくきょく)."

there is a rule for this, it isnt abnormal

in example, kanji where the last hiragana ends with ku and first letter of next kanji is k, they will merge. I can read kanji like these correctly without ever seeing them before.
作曲 heres another example, sakkyoku, sakukyoku, theres a bunch more types and what-not, but I dont know them on the top of my head

It's not irregular, but for someone who learned the readings without any vocabulary and then pieces words together with isolated readings...they might not know when the sounds are likely to change. So if they see 薬局　and piece together やく　and きょく, it's logical but still wrong. Even if you know every reading for every kanji, you still have to look up the pronunciation...

By most common onyomi I suppose you mean the ones with upwards of 40-60. The real trick to picking good movies for them is pick ones you've seen a million times that you know really well. Jeez, thinking back, I must have used at least ten different Nicholas Cage movies.

コウ - American psycho - 68
キ - Kill Bill - 36
キョウ - Scarface - 30 (this was somewhat arbitrary)
シ - The Exorcist - about 50
ショウ - Victor/Victoria - 69 (Broadway movie)
トウ - Boondock saints - 44
ジ - The Rock

This are probably all of the biggest ones. But they weren't too hard, because I've seen them all to the point of have the entire movies memorized. Finding locations within a movie is all about knowing the movie really well.

After starting the whole sentence deal, I'm beginning to see the real benefits of this method. The most immediate is how I'm able to read signal primitives. Reading kanji are very easy when you can spot signal primitives a mile away. As for irregular readings, well, that's what I have anki for. I wonder if Japanese would do any better with the irregular words they've never seen before? I'm a native English speaker but I can't spell onomatopoeia by sounding it out. The point isn't how many exceptions there are. I can still do a lot more with what I know than what I'd be able to with just RTK.

Another big thing is learning vocab. Words that use onyomi are practically memorized already. You guys think learning 飛行機 is easy when you know "Fly - Go - Machine". Imagine how easy it is when you also know the readings. ひ - こう - き. The side to fly is unique to ひ as far as I've seen, and the junk on the right of 機 signal き.

Thanks Alyks, perhaps you can round it out by presenting some samples from a couple of your "kanji movies".

For the negative comments, to me they sound similar to some of the anti-RTK stuff I read in TJP forums (just not hate filled, though). There will be flaws but would it be enough to dismiss it out of hand. Alyks isn't saying to learn pronunciation then stop there. That's no more than we'd say we're done learning kanji with RTK. He is using a stop gap before formal reading, just as we did with RTK.

With the sentence method, you're kinda of randomly adding readings (unless you are using a coordinated source for your sentences ala KO2001). RTK2 offered a method to get a large part of your reading consolidated with minimal effort. This movie method appears to be the same, much like the kanji chain thread from way back when (at that interesting kanji palace idea).

Keep us updated alyks. I remember how I saw the benefit of RTK when I began the sentence method.

Anyway, movies I've seen ALOT:

Groundhog Day
The Wonders
Joy Luck Club
Nottinghill
Matrix
Star Wars
Lord of the Rings

TV shows:

Babylon 5
Battlestar Galatica
Doctor Who (Tom Baker, New Versions)
Simpsons
Frasier

Comics:

Sandman
Miracleman
Supreme Power
Rising Stars
Watchmen
Hikaru no Go

I'm guessing that movies, tv shows, comics would work equally well. Now it's just a matter of trying it out, kick the wheels, look for faults and strengths.

I am interested in this concept.

A couple of questions:

1. Has anyone put together a list of onyomi that include kanji from both RTK1 and 3 (I finished RTK3 just today)? Alkys-sensei's seems to only have RTK1 - which is fine for most people, but it'd be very convenient for people who've done both volumes to have everything in one file.

2. I, like others in this thread, don't watch many movies - at all. However, I did used to play my share of video games, which may be more applicable to some here.

Shenmue (Dreamcast)
Legend of Zelda series (all nintendo systems)
Super Mario Bros (old NES)
Earthbound/Mother 2 (SNES)
Donkey Kong Country (SNES)
Crash 'N The Boys (NES)


The problem with the above, of course, is twofold:

1. Some of these games/shows don't have much of a story. In Donkey Kong Country, you have to get bananas, or something. In Crash, you run fast and hit people with sticks. Images that thin would work better for single kanji, which brings me to my next point:

2. Let's say that I was to use a TV show - like Star Trek: The Next Generation. I could probably do fairly well with that one (I don't remember many plot sequences, but I definitely know the characters well). The problem with that is that I used Data for my State of Mind primitive, which eventually ended up with things like Picard slipping into the picture now and again, etc. It worked just fine for the kanji themselves - but now if I choose Star Trek for a "movie", I worry that I risk muddling things up in my head.

Another example is that for the combo on the right side of 陸, I used pirhana plant, and goomba for the right side of 凌. Worked great for kanji, but now I probably can't do Mario Bros as a "movie" location.

Any thoughts?

also, I'd like to second the request for an example from one of your movies. This would probably take some time to type up, but would be great proof for the effectiveness of the method:

think you could give us, say, the 36 kanji from Kill Bill or so? Or even a 15-kanji-or-so onyomi group would be great.

dihutenosa Wrote:also, I'd like to second the request for an example from one of your movies.

Another request for a concrete example.

Since this thread has been started, I think I have just now begun to understand what you are trying to do. Even so, I would love to see some examples.

If someone has already done an RTK like study on the kanji in a group, it does seem like they could conquer a single group fairly quickly if they are creative.

alyks Wrote:コウ - American psycho - 68
キ - Kill Bill - 36
キョウ - Scarface - 30 (this was somewhat arbitrary)
シ - The Exorcist - about 50
ショウ - Victor/Victoria - 69 (Broadway movie)
トウ - Boondock saints - 44
ジ - The Rock

How do you differentiate between コ and　コウ,　ショ　and ショウ, etc? Since you are tying the readings to similar sounding English words, which don't have long and short sounds...

Quote:I can still do a lot more with what I know than what I'd be able to with just RTK.

Another big thing is learning vocab. Words that use onyomi are practically memorized already. You guys think learning 飛行機 is easy when you know "Fly - Go - Machine". Imagine how easy it is when you also know the readings. ひ - こう - き. The side to fly is unique to ひ as far as I've seen, and the junk on the right of 機 signal き.

Definitely, you have a huge advantage by having learned the yomi along with RTK1 :) But you still have to learn the readings of each word you learn, even if it's just to double check.
Your method just seems more specialized to you(alyks). It has a lot of arbitrary points to it, like which movie gets which reading. I myself don't watch that many movies, let alone rewatch them until I know then so well. I'd probably end up using anime, or made up scenes, if I needed this method.

alyks Wrote:Another big thing is learning vocab. Words that use onyomi are practically memorized already. You guys think learning 飛行機 is easy when you know "Fly - Go - Machine". Imagine how easy it is when you also know the readings. ひ - こう - き. The side to fly is unique to ひ as far as I've seen, and the junk on the right of 機 signal き.

I guess for me though the thing is that I never want to learn just ヒ　－　コウ　－　キ　seperately. I'd rather just learn 飛行機 （ひこうき） all at once.

From my experience, I find that it is better just to learn vocabulary. After you learn 飛行機　銀行　行動　the コウ reading will take care of itself.

And then when you learn 行儀　and 行, the ギョウ　reading will start to take hold.

I think that learning on yomi in isolation uses a lot of time that could probably be better spent. Also I think it is really boring.

I'm not singling out the movie method presented in this thread...I feel the same way about kanji chains, and kanji palaces, or whatever other systems are out there that focus on onyomi in isolation.

I think the only exception for me are "signal primitives" like in RTK2, pointing out 砲　胞　包　can all be read ホウ　seems a lot simpler and useful than making mnemonics to remember 報　砲 and 方　can all be read ホウ

Oh and I've also seen the ヒ　reading of 飛　in Mario galaxy....when a comet comes to a galaxy, it says 飛来中　＾＿＾

Quote:I'd probably end up using anime, or made up scenes, if I needed this method.

Wouldn't it be like doing a more advanced version of RTK at that point?

If this can help speed up the reading process, I'd like to learn more too..

IT'S GROUNDHOG DAY! What is this, Miami beach? God that's such a great movie. I use it for ヒ. The scene where the kid is falling from the tree? Yeah, that's 飛. He's holding on to a stick with rudders on both ends floating down into your person's arms. (Although I personally changed the primitive for イ because I needed to keep it different from 人, since 人 appears so often when you only use the bear minimum of primitives.)

The Matrix is カイ, because it reminds me of Keanu Reeves. I had to stick a bunch of weird stuff into that one. 灰 is right after he talks to the creator in the second one. He's bursts the door into flames flying out and ash falls to the ground. (The cliff primitive became more of a "falling down" primitive.)
開 is in that scene where he's about walk through the door to meet the oracle for the first time. He puts his two hands up to it and pushes it open.

I found tv shows a little harder, just because you tend to watch one episode and move on. I only was able to do it for House, because I own the series on dvd and watch it every now and then. As for comics, if I was able to use a gameboy game for オク (pokemon), I don't think you'd have too many problems with a comic book.

dihutenosa, for both RTK1 and 3 there's a file called "RTK1 vs RTK2 numbers.rtf" I put on the yahoo group. It will give you want you want. As for an example movie, I'd have to do a short one. Describing 36 different kanji mnemonics would take a lot of time. I might put コク (Pulp Fiction) up on my blog, because it's a staple movie with only 9 kanji.

How do I keep the readings like ショウ and ショ from getting mixed up? They both get different movies. Easy. Done. The movie you pick can even be arbitrary. The point is to have different movies for each reading group. I just instantly know that Blast from the Past is ショ. This is totally different from Victor/Victoria, which is always ショウ.

Johzep, ひらいちゅう? That's great. If you think the time used to learn onyomi could be better spent, fine. But remember that it didn't take any additional time to learn them, as I completed learning the kanji and onyomi at the same time in only fifty days without actually reviewing or studying the readings directly, just the kanji.

PS. I learned the signal primitives without trying. I honestly didn't know how well I knew about the signal primitives until after I finished the method. In this sense, it was simpler than RTK2. Not only because it was easier, but the exceptions are also secondhand.

dihutenosa Wrote:also, I'd like to second the request for an example from one of your movies. This would probably take some time to type up, but would be great proof for the effectiveness of the method:

think you could give us, say, the 36 kanji from Kill Bill or so? Or even a 15-kanji-or-so onyomi group would be great.

http://drmoviemethod.blogspot.com/2008/0...ction.html

As promised.

[Katsuo]

I think Alyks' method is interesting.

But I wonder if it wouldn't be easier to combine it with RTK1's order and primitives.

Before starting RTK1 you would need to make a list of which movies to assign to which ON readings. Then, as you go through RTK1, make a point of locating each story/image in the appropriate movie.

Alyks himself learned the kanji in an order determined by ON reading/movie. But is there a compelling reason for doing that?

The origin of the project was to be able to adapt Heisig's system to classroom study/order. To me, because of the way I "crafted" the system and created new iterations, it didn't seem like an option to learn in his order. All that aside, I never decided to try and adapt Heisig's system for these reasons:

1. For ease. Switched around movies would be very confusing with a big potential to make a mistake. Not only that, but I don't think learning signal primitives would be quite as streamlined. Part of my recall for them is just having strong associations between the essential images and the movie. Phonetic components tend to be very specific to it's onyomi group. Part of why it worked was just having all the kanji with a certain combination of primitives all grouped together. Using my method without following groups would be like trying to learn RTK2 out or order- you could do it, but it would be difficult and tedious.

2. I wanted to prove Heisig wrong. Simple as that. When I read his introduction for the first time after trying it out a little, I disagreed with the order being the reason it works.

3. This is probably the biggest reason. Using my method requires tangible images to work with. Because the essence of the system is using locations, which you need to visualize clearly in your mind, you must also visualize the components clearly in your scene. Quite a lot of kanji have intangible meanings and would be ineffective when trying to incorporate into a scene.

I completely disregarded the use of "stories" and logic and relied purely on my mind's ability to remember imagined images. This I feel is necessary to make clear connections to the movie. It also allows the meaning to be derived from the scene, and that is very much needed to lighten the load of things to visualize.

4. It wasn't hard to learn out of order. It was certainly a lot easier to memorize the components beforehand than it would be to learn all the movies before hand. I
Seriously, it was really very easy to learn out of order. Even complex characters were not hard.

Katsuo Wrote:I think Alyks' method is interesting.

But I wonder if it wouldn't be easier to combine it with RTK1's order and primitives.

Before starting RTK1 you would need to make a list of which movies to assign to which ON readings. Then, as you go through RTK1, make a point of locating each story/image in the appropriate movie.

Alyks himself learned the kanji in an order determined by ON reading/movie. But is there a compelling reason for doing that?

I'm going to have to disagree with this. While doing RTK you want to worry about keyword and stroke orders. After you have that under your belt, there's this movie method.

Alyks, you say you did about 50 a day (finishing in little over a month). Roughly how long per day would you say those 50 and your reviews (if you did them) took?

For those thinking of doing this with flashcards, I think "Trinity" once again can help out. Create a new account (unless you're not using Trinity already), and add a vocabulary word (or two) that uses the On yomi you're reviewing. The only thing you care about when reviewing is getting that one kanji's onyomi correct.

Nukemarine Wrote:Alyks, you say you did about 50 a day (finishing in little over a month). Roughly how long per day would you say those 50 and your reviews (if you did them) took?

I'm glad you asked, but I honestly expected somebody to ask sooner.

Everything was the same. Same speed as doing RTK, same speed to review. Since I was doing fifty a day, my daily reviews got up to 200 about halfway through. It would take me about 2 - 2 1/2 hours to review (incrementally throughout the day, I'm going off of anki's ETA). A straight run and I could get up to a hundred done in 45 minutes (not writing out the kanji, but still visualizing the kanji in my mind).

I said this in an earlier post, which sums up my speed of learning new kanji:

Quote:Slow and steady. Go through ten and review. Review while learning new ones, and review after. 95%+ retention rate with this. 10 kanji/half hour (average).

Fast and sloppy. Go through twenty at a time, and stop to review once. Go through twenty more. The point is to make an image as best you can, then move on without looking back. 70-80% retention rate and 20 kanji/half hour (optimal).

It's worth mentioning that I looked up each kanji in my dictionary before I learned them.

Yo, after trying it for some time. I have a comment to do:

It has a lesser use after you finished RTK. The reason is that words work better as the places than movies. And there are soo many readings.

But it has a great thing to contribute to RTK. Using places for your stories is much better than using characters, expressions or actions. Moving between locations produce better stories. Using a known story to do these "movings" works even better.

So learning the readings is not the best thing of your method, but the way you visualize your stories. About the Hesig order. It is very hard to you say anything against it, as you have not used it.

I personaly building framework for stories is much better than recurring to basic primitives all the time. And creating primitives on-the-fly it is even better.

And deciding everything in the begining is generaly a strategical mistake. You want to be able to change things as you go.

Yo, update on my usage of the method.

It is going very well. I can't measure it because I'm doing it while doing the sentences in a very informal way.

So, I recommend to everyone at least to know what this method is about.

Anybody have a link to a spread sheet or Anki file that lists: Kanji, Heisig Number, Stroke Count, Kun yomi and On yomi entries? I'd like to begin this method, and I guess Anki is the best tool to do it. Problem is Anki pre-loaded Heisig deck does not have onyomi.

I don't think it would be easy to find what you're looking for there, Nukemarine. The best I can find for you is this:
http://www.hellodamage.com/tdr/archive/7...yfreq.html

But the meanings aren't exactly Heisig.

[Katsuo]

Nukemarine Wrote:Anybody have a link to a spread sheet or Anki file that lists: Kanji, Heisig Number, Stroke Count, Kun yomi and On yomi entries?

I don't have any links, but I can generate a spreadsheet with that data and email it to you. Any other feature requests in addition to the above?

alyks Wrote:The best I can find for you is this: 2501 kanji by frequency

That data is a selection from KANJIDIC.