- Keywords
Having a set of more than 2000 characters without a way of identifying each of them is kind of difficult. It’s ok for a Latin alphabet to not name/identify those 26 characters, or not naming kana characters but I can’t see how our Western brains you can do that with more than 2000 kanji. So, here is what Heisig helps you with – he gives you a label for each character and when I say “sign of tiger” the other practitioner of Heisig’s method knows which character I’m talking about (it is 寅, btw).
- Meanings
Heisig’s intention was to give you keywords that are related to meanings of the characters. For the most part he does a good job but sometimes he fails (and yet at times he is plain wrong, like in village (町), which is town actually or town (村), which is village
. I don’t think this is particularly a problem – after all we know that a single word can have multiple meanings, so why this should be a problem with kanji? We are attaching a label to a character knowing that there may be other labels/meanings for it.
- Compounds
As we have keywords/labels for the kanji we can tackle learning compounds in a somewhat ordered fashion. Things like 新幹線 can be broken down into “new, tree trunk, line”, or 地下鉄 to “ground, below, iron”, and now you can have mnemonics for compounds as well. Again, having mnemonics equals to better and easier remembering.
- Reading/recognition
In my experience, having a mnemonic, either for a character or for a compound, is important at the time you start “real” reading. But as you progress, you tend to bypass mnemonics because your visual memory starts recognising characters just as they are. With more and more exposure, the mnemonics become less important but if it happens that you forget character/compound you can fall back on them – convenient parachute I would say.
- Writing
Writing kanji, however, is so different from reading that I can’t imagine being able to get rid of mnemonics here. When I think あさ, I keep recalling my story, identify the primitives and only then draw the character, primitive by primitive: “mist (needles, sunflower (sun, needle)), moon” and the character 朝 is jot in (of course, when you can resolve keyword to stokes without relying on composing parts then you do that).
- Writing practise
When you know the primitives you don’t spend much time on studying the strokes that a character is drawn with. If we follow with the 朝 example, just knowing that the character is a combination of “mist, moon” is almost sufficient to draw the character. If your story includes hints for placement of the primitives then I would venture to say that you don’t even need to see the character in order to draw it – the story has all the information needed for you to produce a legible kanji. All of that makes writing practise a relatively lightweight task and lots of time can be saved.
- Order
According to Heisig, in order to use his method efficiently, you need to follow it from the start to the end in the order that he provided. I don’t think this is all that necessary. You can pick up any kanji you want, back trace it to its primitives and come up with your mnemonic. If, on the way, some primitive is another kanji you do the same. SRS all those and you should be able to retain the character. With time, you will start seeing primitives that you already know and initially slow process of backtracking can be fast or not needed at all. The downside of this approach is that you spend some time on identifying the primitives. On the upside though, you save time because you don’t learn kanji that you are not likely to use in foreseeable future. Also, doing Heisig this way, allows for actual studying of Japanese, as oppose to studying kanji only.
EDIT: fixed some typos.
Having a set of more than 2000 characters without a way of identifying each of them is kind of difficult. It’s ok for a Latin alphabet to not name/identify those 26 characters, or not naming kana characters but I can’t see how our Western brains you can do that with more than 2000 kanji. So, here is what Heisig helps you with – he gives you a label for each character and when I say “sign of tiger” the other practitioner of Heisig’s method knows which character I’m talking about (it is 寅, btw).
- Meanings
Heisig’s intention was to give you keywords that are related to meanings of the characters. For the most part he does a good job but sometimes he fails (and yet at times he is plain wrong, like in village (町), which is town actually or town (村), which is village
. I don’t think this is particularly a problem – after all we know that a single word can have multiple meanings, so why this should be a problem with kanji? We are attaching a label to a character knowing that there may be other labels/meanings for it.- Compounds
As we have keywords/labels for the kanji we can tackle learning compounds in a somewhat ordered fashion. Things like 新幹線 can be broken down into “new, tree trunk, line”, or 地下鉄 to “ground, below, iron”, and now you can have mnemonics for compounds as well. Again, having mnemonics equals to better and easier remembering.
- Reading/recognition
In my experience, having a mnemonic, either for a character or for a compound, is important at the time you start “real” reading. But as you progress, you tend to bypass mnemonics because your visual memory starts recognising characters just as they are. With more and more exposure, the mnemonics become less important but if it happens that you forget character/compound you can fall back on them – convenient parachute I would say.
- Writing
Writing kanji, however, is so different from reading that I can’t imagine being able to get rid of mnemonics here. When I think あさ, I keep recalling my story, identify the primitives and only then draw the character, primitive by primitive: “mist (needles, sunflower (sun, needle)), moon” and the character 朝 is jot in (of course, when you can resolve keyword to stokes without relying on composing parts then you do that).
- Writing practise
When you know the primitives you don’t spend much time on studying the strokes that a character is drawn with. If we follow with the 朝 example, just knowing that the character is a combination of “mist, moon” is almost sufficient to draw the character. If your story includes hints for placement of the primitives then I would venture to say that you don’t even need to see the character in order to draw it – the story has all the information needed for you to produce a legible kanji. All of that makes writing practise a relatively lightweight task and lots of time can be saved.
- Order
According to Heisig, in order to use his method efficiently, you need to follow it from the start to the end in the order that he provided. I don’t think this is all that necessary. You can pick up any kanji you want, back trace it to its primitives and come up with your mnemonic. If, on the way, some primitive is another kanji you do the same. SRS all those and you should be able to retain the character. With time, you will start seeing primitives that you already know and initially slow process of backtracking can be fast or not needed at all. The downside of this approach is that you spend some time on identifying the primitives. On the upside though, you save time because you don’t learn kanji that you are not likely to use in foreseeable future. Also, doing Heisig this way, allows for actual studying of Japanese, as oppose to studying kanji only.
EDIT: fixed some typos.
Edited: 2012-02-02, 10:22 pm
