I have found that 更 {694} has different meanings on my kanji dictionaries. Rikaichan and other english dictionaries only list meanings simmilar to "grow late", while my spanish dictionaries only list meanings simmilar to "renewal". This disparity also happens with my translated version of "Remembering the Kanji", and I wonder which keyword would be more useful to memorize.

Thanks in advance.

Always memorize the Heisig keyword, it's tempting to try to replace it with another when we don't like Heisig's but the problem is that if we don't know the full list yet, we can't make an informed decision about it. And it's going to be confusing every time you find references to the normal keyword.

In many cases, the Heisig keyword isn't the main meaning of the kanji sometimes it's only related through a specific compound word but that's OK. Later you can refine your knowledge of the kanji meanings.

In this case, I find words both relating to the "grow late" meaning as well as "renewal" in EDICT, so I don't see any reason to change.

The problem is that both meanings are "Heisig", in that one appears on the english edition of Remembering the Kanji, and the other on the spanish edition. Heisig participated on the spanish edition, too, so this is not that the translator just made up the word.

Also, because the full list of keywords appear at the end of the book, I can make changes with care (specially in cases when Heisig is way off). Looking at the list, it might be that Heisig didn't want to create confusions with "again" {1815} and chose *grow later* instead.

I ask this question because an inadequate meaning will make harder to learn the kanjis that use the primitive, I'm still a bit bitter at the "computer/house of the fathers" fiasco and don't want to repeat the mistake.

But the whole Heisig method is based on assigning arbitrary meanings to the primitives. I think that using "computer" was brilliant. It's obvious that it's made up and does not reflect the etymology but it was easy to learn the kanji using it.

Codexus Wrote:But the whole Heisig method is based on assigning arbitrary meanings to the primitives. I think that using "computer" was brilliant. It's obvious that it's made up and does not reflect the etymology but it was easy to learn the kanji using it.

I won't fight on this point, but I just disagree on all things said on this post.

Now, granting that both meanings are Heisig and chosen arbitrarily, which one do you think will be easier to use for building stories on kanjis featuring that primitive? How many times do 更 appear on the books as a component? I know it appears at least one more time on "stiff" (#695).

Thanks.

I've replaced a lot of Heisig keywords when I thought the meaning was too far from any popular Japanese words. You just need to look a bit through the index and see if there's a conflicting keyword somewhere. As for choosing the keyword, I just looked for popular compounds in Wakan. I searched for 更 and it seems to be used in both ways: 変更 へんこう更改　更新　or 夜更け　よふけ　今更　いまさら。In the end though there is no perfect keyword. 変更　へんこう　means change as in change the date of our appointment. It's not really that close to either renewal or grow late. Just having a keyword helps you remember the word, even if it's not closely related. So just choose one and move on.

Unfortunately English words are never going to truly encapsulate the meaning of the kanji. After some careful work, you may find that you can choose another keyword that approximates the meaning more closely and doesn't clash with other keywords, however I don't believe the results you get from doing this will be worth the effort. There are so many factors that you have to take into account: which meaning of the kanji should be chosen? does my candidate English keyword have other unrelated meanings? does my candidate keyword clash semantically or phonetically with any others already in the list? etc.

I would choose one of two options. 1. Choose a Japanese keyword and get rid of the English one. 2. Accept the Heisig keyword and don't waste time and effort trying to choose something else that won't be a huge improvement.

The meanings assigned to the primitives aren't arbitrary by a long shot. A great deal of them are closely tied to their actual etymologies.

I agree with wrightak here, with emphasis on number 2.

I also think it would behoove you to stick to the English or Spanish edition of the book and not try to use both at the same time.

I should point that I'm using the Spanish edition (with minor changes), and I just noticed the discrepancy while using the greasemonkey script to "rewrite" the english keywords into spanish.

I just wanted to know if 'grow late' was an easy primitive to create stories, or it made as much sense as computer for carp 鯉　こい (that is, none at all). But I guess I'll stick with 'renewal' for the moment (the sense is closer to 'another time', though), and use 'grow late' if I get any difficulties.

Thanks!

DrJones, just out of curiosity, why do you not like calling the 里 (ri) primitive computer? The reason I ask is that 鯉 (carp) was a character I found really easy to remember using the story Dr. Heisig gave in his book. I thought giving 里 the primitive meaning of computer was especially clever because the character looks like a computer and it's also a very nonabstract object. (I tend to remember stories with less abstract primitives more easily, I find.)

I'm not after a debate, but I am quite interested in your point of view since it seems so completely different to mine.

Deon

You can change the keywords if you think you have a better one, but note that it's not enough to just search the index for an exact match. You need to take care not to conflict with different keywords that have similar meanings. Unless you fancy carefully searching through 2042 keywords looking for similar meaning words you're better off just going with Heisig's keyword.

In your example, there are the following common use words: 更かす = to sit up late; 更ける = to get late. So this keyword seems to be suitable.

Some of Heisig's keywords are not the best choices, but statistically they are few and fairly insignificant. A kanji almost never stands for just one word. Jukugo words are far more numerous than individual kanji words. So knowing an "exact" meaning for a single kanji is not very important. It is not a good idea to try to translate Japanese exactly into your native language anyway. The keywords are just memory hooks.

As for "ri" -> "computer" I think it was a great idea from Heisig, and I think as you progress through the book maybe past about half way point you will appreciate the value of following Heisig's advice. Doubting the method and/or not following the advice seems to be the #1 route to nowhere with RTK. Note that Heisig warns not to mix RTK with other kanji learning methods because he knows of no cases where this has been successful.

It seems to me you are trying to mix RTK with other stuff which is not such a good idea, DrJones. But, it seems you are already making good progress to 800 which is a milestone; many people give up before 800. So keep at it and my advice is to follow the book's notes, and the tips in the stories on this site. I found many of the shared stories here very helpful. You will find that using good memorable keywords and sometimes personifying primitives will be much more helpful than looking at etymology.

According to my book, carps symbolize vitality because they can swim up a river (just like salmon do). Salmon start their life on the river, and according to wikipedia:

"they are born in fresh water, migrate to the ocean, then return to fresh water to reproduce. Folklore has it that the fish return to the exact spot where they were born to spawn."

The kanji 鯉 represents a "fish returning to the house of their fathers", which is one of the most direct kanji-to-story translation in the entire book. But, instead of using that powerful reminder, the book features some stupid story featuring 鯉のぼり and computers. Utter nonsense!! (if you take a look at the study section on "carp" you will find many people telling that there's no easy story for this kanji).

It's not the only case of easy stories turned hard because use of computer (the spanish version is worse, it uses a calculator), either. 童　わらべ　"juvenile" (#436) shows a person that "stands on the house of his fathers, because it's too young to buy one". All stories using that primitive improve by using this meaning, which is common sense, because chinese created the kanjis in a way not very different to the one we use to learn them: by creating stories featuring those elements. That's why using arbitrary meanings to relate them only messes up the stories.

If you look at the study page on ri, you'll find at least another person concerned because Heisig didn't use this meaning.

This is not the only choice for primitive that I dislike. I was thinking on opening a thread about them, to ask someone that has already finished the book which ones I should modify (for example, "crotch", that so much pain brings to many people, represents a hand, that's why it appears on the lesson about hand-related kanjis).

I didn't yet finish RTK1 but almost, I'm half way through chapter 53.

Personally I didn't have problems making up stories with "computer" or "crotch". Actually for Carp I used Heisig's story and never forgot it. On a trip to Japan I saw the carp hanging in the wind outside houses during boys festival, and it was easy to picture that with a computer hanging there too as Heisig depicts it.

The thing about "crotch" is that in Chinese etymology there are loads of different shapes that all relate to hands. With RTK method it doesn't make sense to have lots of different primitives with the same keyword. That's why he used hand, finger, reaching hands, two hands, crotch, elbow, fist, etc. so that its easy to make up stories that help to remember which shape to write. I found the "hands" chapter difficult when studying it but looking back now its a piece of cake to remember most of it. Don't worry if things seem hard at first because they will seem easy when you look back a week or two later.

For "house of fathers", it seems ok to me to use that if you prefer it. There are some kanji with similar keywords like "house", "home", "home town", "home country" etc. but not too much trouble to differentiate.

However I don't see much point. There are not so many direct uses of this primitive in the book and computer is quite logical for several of them.

"quantity" -> just use "ri" since it's a quantity. Also computers can be used to calculate quantities.
"rin" -> small quantity: computer smashed to tiny pieces when thrown off a cliff
"bury" -> bury your beloved computer under some soil.
"logic" -> computer = obvious.
"black" -> computer on fire turns black.
"juvenile" -> a juvenile computer hacker worked fine for me.
"back" -> hidden computer in the back of top hat & scarf, used for Blackjack counting.
"plains" -> ri+beforehand = before computers people worked in the field. Or use "ri" which was a distance measure used to measure fields before.

DrJones, it's funny how I seem to think the exact opposite of what you think. The "crotch" primitive is one of my favorites as well since it allows all those easy to remember stories, from the "dirty uncle" to the "soiled underwear" primitive and many more.

Off course, what matters is that it works for you. The Heisig method is often criticized for the arbitrary meanings assigned to the primitives and if you can make it work for you using etymologically correct meanings that's a powerful argument to convince people who avoid using Heisig for that reason!

[ファブリス]

DrJones Wrote:(...) example, "crotch", (...) represents a hand

You sure it's a hand? Because the "finger" 扌 is already the "compressed" version for the hand, when it appears on the left. This list of chinese radicals says bushu 29 ("crotch" 又 ) is "again".

If the etymology can help you make a story that in turn helps you to remember that's great. In a way it's still a kind of mnemonic technique, remembering by association. For the Heisig method though it's not really important, especially if you notice that often the primitives have a relationship in meaning to the chinese radical.

According to Henshall's "A guide to remembering the Japanese characters", 又 can be "hand", "take" or "help".

That book's focus is on the etymology of the kanji. Interesting but as a way to remember the kanji that's way too complicated.

ファブリス Wrote:

DrJones Wrote:(...) example, "crotch", (...) represents a hand

You sure it's a hand? Because the "finger" 扌 is already the "compressed" version for the hand, when it appears on the left. This list of chinese radicals says bushu 29 ("crotch" 又 ) is "again".

If the etymology can help you make a story that in turn helps you to remember that's great. In a way it's still a kind of mnemonic technique, remembering by association. For the Heisig method though it's not really important, especially if you notice that often the primitives have a relationship in meaning to the chinese radical.

Just to be sure, I just checked again in an awesome kanji dictionary titled "Kanji: La escritura japonesa" (ISBN: 84-7517-110-9 1991) that includes old forms of the Kanji dating back to the original pictographics. Here are some examples:

反: Shows a hand pressing against a wall.
友: Shows two hands doing the high five.
受: Shows a hand giving a ring to another.

According to the book 又 is just a hand and means "and, again". Curiously, Edictj says that it also has the meaning "or", so we have a case in which a kanji is used in opposite concepts. The keyword I assigned to it is "por otra parte", which is a spanish conjunction that "parks" the previous sentence to talk about something related. I chose it because it can be used on both meanings; is not very different to the one used by Heisig; and because it has an easy mnemonic: I put aside my previous sentence with my hand. I even have done that gesture in some occasions, heh.

又 is one of many kanji pictographs representing a hand. There are a great deal of them. Heisig was very deliberate in giving each one a distinct image. That's what makes his technique work, when you recall the story, you're not going to confuse 又 with something else that also represents a hand.

There are many problems with following an etymological method of memorising the writing of kanji. One of these is that there are many etymological theories for each kanji.

Quote:反: Shows a hand pressing against a wall.
友: Shows two hands doing the high five.
受: Shows a hand giving a ring to another.

The above may be the generally accepted etymology for each of these kanji but for every one of them there will be many proposals put forward by experts giving different reasons for the origin of the shape of these characters.

Even if you take one of the most basic characters, 人, many people thought that the second stroke represented the forward striding leg of the person. However, the most recent theory to gain widespread support is that the second stroke is in fact the dangling arms of a labourer with a heavy load on his back.

My point is that you may think that a story about computers for carps is ridiculous but for all you know, the story about the fish coming back to the home of their fathers may be utterly false as well. If you want to study etymology then it's very fascinating but I don't think it's the best course of action if you want to achieve what RTK1 sets out to do: memorise the writing of the characters.

DrJones Wrote:According to the book 又 is just a hand and means "and, again". Curiously, Edictj says that it also has the meaning "or", so we have a case in which a kanji is used in opposite concepts.

The translations of "and, again" and "or" both reference the word また. また can have all of these meanings depending on the context of the sentence.

"友: Shows two hands doing the high five."

I doubt that the Chinese were doing high five's when this character was invented... but anyhow, this is a perfect example of why etymology is not as good as RTK for memorising the writing. The story is great... memorable... but gives little clue which components to write. You have two hands being used but they look different. If you don't give them distinct keywords you'll be relying a lot more on your visual memory to remember how to write each part, rather than learning the kanji by parts with Heisig's method.

In my experience having tried to learn the kanji from etymology based books, they were interesting and useful for learning some of the more simple kanji but nowhere near as effective as RTK for studying kanji en mass.


I'd consider the amount of time spent learning each kanji multiplied by 2042, and just how much benefit you get for your time spent. I average 2 to 3 minutes per kanji, because very often I find shared stories that work for me. That's one of the advantages of sticking to the tried and tested keywords. When I have to make up my own story it adds significantly more time.

Now consider if you add 1 minute to look up each kanji in the dictionary or etymology (just 1 minute seems optimistic to me), you add 34 hours extra study time, which for me would have added more than 1 month to completing RTK. That would suck IMO.

vosmiura Wrote:"友: Shows two hands doing the high five."

I doubt that the Chinese were doing high five's when this character was invented... but anyhow, this is a perfect example of why etymology is not as good as RTK for memorising the writing. The story is great... memorable... but gives little clue which components to write. You have two hands being used but they look different. If you don't give them distinct keywords you'll be relying a lot more on your visual memory to remember how to write each part, rather than learning the kanji by parts with Heisig's method.

In my experience having tried to learn the kanji from etymology based books, they were interesting and useful for learning some of the more simple kanji but nowhere near as effective as RTK for studying kanji en mass.

I'd consider the amount of time spent learning each kanji multiplied by 2042, and just how much benefit you get for your time spent. I average 2 to 3 minutes per kanji, because very often I find shared stories that work for me. That's one of the advantages of sticking to the tried and tested keywords. When I have to make up my own story it adds significantly more time.

Now consider if you add 1 minute to look up each kanji in the dictionary or etymology (just 1 minute seems optimistic to me), you add 34 hours extra study time, which for me would have added more than 1 month to completing RTK. That would suck IMO.

I think I have misguided people a bit with too much assumptions from my part, so I'll try to write some clarifications.

1st. The kanji descriptions are mine, what the book shows is a row of primitive pictographs of these kanjis (so no discussions about the etimology possibly being wrong) and descriptions took me less effort than scanning the pages. Maybe I've exaggerated a bit with the "High Five" description, heh.

2nd. I think the Heisig's method is good, of course (I'm using it afterall), also, I've read his advice, too, so I'm aware of problems that can appear and the problems with making changes or using other approaches. More than that, I agree with him on that part.

3rd. I think Heisig made some mistakes accross the book, which I can fix with careful consideration. That he ignored etymology is one of them, which happens almost exclusively due to poor choice of meanings for primitives, as Heisig uses known etymology everytime he can.

4th. You don't have to do a look-up in an etimology dictionary. You just "read" (as what I have is a bunch of drawings) from start to finish, and then you use that story for Heisig's. I think it takes less than 20 seconds per kanji following this method; also, you don't have to learn 2000 kanji this way (partly because I don't have pictographics for all of them). 214 radicals suffice.

5th. Even though there are many hand primitives, there are many hand gestures available, too (you can give 又 the meaning "palm", for example), and even though I don't have problems specifically with 又 being "crotch", many people do, according to the study page. Seeing that RTK I (the proven method) has a very high drop ratio, I would try anything that could reduce this number reliabily.

I hope that didn't sound as if I didn't agreed with your points.

[ファブリス]

DrJones Wrote:That he ignored etymology is one of them, which happens almost exclusively due to poor choice of meanings for primitives, as Heisig uses known etymology everytime he can.

I suspect James Heisig was experienced with mnemonics having learned other languages before, maybe that's why he wasn't so worried about departing from the historic background of some components of the characters. If you're not dry up on cash after this Christmas you could check Father De Roo 's "2001 Kanji" book. Have never had a chance to see it myself yet, though I'd love to get a copy some day. Of course this won't help you finish RtK faster, but not incredibly slower either, the only risk is too loose heart when it takes too long (if you don't finish RtK within a year you might loose heart).

Thanks, I think I understand better now DrJones. If it helps you then that's good.

Btw, which study page are you referring to?

I don't see the point in trying to cram in a bit of etymological knowledge by changing the primitives' names to more "correct" ones. You're only going to learn a bit more etymology and it risks making learning the kanji harder. You can always learn the etymology later if you want. It sounds a bit stupid for sure to call the "ri" a computer, but it makes for better stories. You're going to be making up so many stories that anything that makes things easier will be appreciated. And the name of the primitive is very important.

It's good that you're trying to improve the method, but I think that this is not a good idea Dr. Jones. Changing "computer" to "house of our fathers" is probably ok, but changing crotch to hand or palm is a really bad idea. There are so many of the hand/fingers primitive that you definitely don't want anything that can be remotely confusing. When I begun doing the Heisig book, I did many things to "improve" upon the method and that turned out to be a bad idea. I didn't like the idea of making weird stories with weird primitives, but that's precisely why the method works. I'm not saying that you can't possibly adapt some primitives, but I think you're better off using proven primitives and just learning afterwards the correct meaning. A bit like you're using the crotch primitive all the time for your stories, but then learn that it really means "or again". (I didn't like that primitive btw and I ended up calling it a dowsing rod).

vosmiura Wrote:Thanks, I think I understand better now DrJones. If it helps you then that's good.

Btw, which study page are you referring to?

I'm referring to this one: http://kanji.koohii.com/study/
You enter a Keyword on the search bar, and then you can see other people's stories (and their frustation with some of the primitives). You can post your stories, too! But I think you already know that page.