Greetings.

As suggested in other thread, I'll try to ask direct questions and not deviate much into other topics. I'll thus ask a question first, and then elaborate on what I mean and how I think (skip it if you want).

So, I have a question to people who are japanese or live in japan or know japanese friends... How do the japanese, native people, percieve kanji? I mean, what do they associate the kanji with? How do they "relate" to them, how do they remember them? What are kanji to them?

As you may have seen in my threads, I'm kinda clueless.

I thought (many times) I have finally grasped the idea, but was (many times) proven wrong. From the different methods suggested to learn kanji, those praised the best (and being considered the fastest) associate kanji with a keyword - a single definition, usually one of the dictionary definition. However, as I learnt, japanese does not work that way - kanji are often never used to mean that definition, or never used alone at all. I want to understand how it's "supposed" to work - I mean, as with every system, it must have some sense of logic to it, right?

So, can someone explain, how do japanese people themselves percieve kanji?

I mean, with words, we european-minded people just remember "certain combination of sounds has this meaning and is spelled this way". We usually don't even remember spelling, just exceptions - like, for the word "like" or "light" or "can" there's no need to remember spelling, but for the word "one" there is. Writing is actually "transparent" after you know grammar, alphabet and pronounciation rules - you just have to remember every exception to the general rule "attached" to the word. So, in mathematical terms, for us object from set of "meanings" is translated into set of "words", and written form of a word is a function of word and different rules that convert it from sound form into written form (and back from written form into sound).

But how does it work in minds of japanese people?

*) Do they remember kanji as a big alphabet of many "letters" (like 木, 人) and many "combinations of letters"　(like 人+木=休), and spell words using the alphabet - like, "I'm okay" is spelled "big-robust-husband", or "giant" is spelled "huge-big"?

*) Do they make difference for kanji that are used solo and those that are always used in a compound? Like, 木 vs 召 - do they remember both kanji the same way, applying same rules, or do they remember them differently, because first is used to mean what it means, while second is not?

*) How do they remember those kanji that have a meaning that is never used (旦 召) - do they associate such kanji with a most used compound, and just remember "once = 一旦” and never bother what 旦 means by its own? Do they learn the previous meaning, dictionary meaning, or only the current one?

For example, so that you understand how I approach this in my language, in Russian (as I belive it is in English too), there are words that have severely mutated over time.

For example, "прелестный" (prelest'nij) = adorable, charming, lovely - a word which is actually derived from лесть (lest') = lie, with a prefix пре-（pre-) conveying, in this case, a sense of "extremity" (kinda like over- or super-), and originally it meant "an extremely dangerous and cunning diabolic lie", and used mostly as a state of being (as, "to be in the state of prelest'"), used to describe a person who is charmed by demonic forces to think of himself as a holy, righteous person, it was used mostly in religious literature. Most of the non-religious people nowadays in russia don't even know it's original meaning, while most deeply-religious people tend to avoid using this word and even can get insulted if called that (as a compliment)!

Same for стерва (sterva) = vixen (usually used by women to mean "strong independant attractive woman" and by men to mean "heartless manipulator, bitch"), originally meant "dead body, rotten flesh" (russian word for "vulture/carrion bird" is стервятник - stervyatnik - derived from this word).

Same thing for many other words - like влагалище (vlagalische) = vagina, used to mean pouch, purse, small bag, nowadays means (in scientific sense) woman's genitalia (same evolution, if I'm correct, as with the original word, which originally meant "sheath/scabbard"). So, when someone encounters this word in, say, older bible texts, they are utterly confused or amused.

So, at least for russians, this is the way it works - most people know the common nowadays use of the word, but some people dig deeper and know the original definition or where-this-word-came-from. It would make no sense, when teaching Russian, to teach people "стерва" as "dead body / rotten" flesh, later commenting that nowadays used to mean "vixen". At least, if I'd be told that, I'd be like "well, why tell me it means rotten flesh if I can't use it when I talk about rotten flesh, what's the point? Just tell me how I can use the word in my everyday life!".

Does it work the same way for japanese? Like, typically people remember a kanji by its current use and meaning but more literate / educated people know it's original meaning? Or does it work in a completely unrelated way?

I don't really know how to look for this kind of information, and for some reason, no method that helps you learn kanji fast actually tells you that - they seem to be focused on how to get those pictures into your head ASAP, and let you sort them out later... But to me, it seems to make no sense... well because I'm not japanese! I think, understanding this will help me greatly in getting ahead with learning the kanji... Learning to "navigate" the system, to "operate" it. Because right now, my mind just rebels at the fact that I'm learning kanji associating them with God knows what and getting more and more confused on the way.

So, maybe someone out there can shed the light on this?

Others can provide better responses probably. But..

I think Japanese remember kanji like we remember/read most words. A large part of it is context and familiarity. Most Japanese probably aren't looking at the whole kanji, they are simply seeing small pieces (if that) and are reading the sentence quickly. The okurigana on kanji really takes a lot of the processing difficulty out of the equation.

Sometimes they might read it and think its one thing and get someways into the sentence and then realize something didn't make sense and go back and look again.

For individual kanji, they probably just decompose the parts up till they know what it is, but its really fast. When you've been reading and seeing the same kanji off and on for your entire life, its very quick. For less common stuff, they might use context (just like any person would in other languages), and if that fails they use their phone/dictionary/etc.

Looking at reading studies/eye tracking studies in English, you'll see most people aren't even reading the whole word. The same is true in Japanese.

2 vix86
Thanks. I asked about a bit different aspect though, that's why I tried to elaborate it in a rather lengthy OP.

I'm not asking about how they recognize words, but rather how do they perceive individual kanji, if at all.

Like, 旦 is only met in several compounds, and never means anything even close to it's dictionary "meaning", so do they even bother remembering its dictionary meaning? Do they associate 旦 with daybreak or morning? Or just recognize it as part of the compounds and never as an individual (like we never recognize "-ess" at the end of "press, bless, caress, excess..." as some individual block, and just remember the whole word)?

Sorry. The russian explanation threw me.

From talking to some Japanese, it really depends. If its a common radical/part, then some people might be able to say what it means, but I have asked many Japanese questions about parts of kanji before and they go "I don't know." and we both look it up in my electronic dictionary and they go "Oooh." with me.

For more common and used radicals/parts, they probably have a natural idea of the meaning of the parts but it may not be the common dictionary meaning.

I don't know if that will of any help, but you might find interest in the following books:

1. Japanese Psycholinguistics: A Classified and Annotated Research Bibliography (Library and Information Sources in Linguistics) by MIYAMOTO/KESS

2. The Handbook of East Asian Psycholinguistics: Volume 2, Japanese by VARIOUS

Istrebitel Wrote:So, I have a question to people who are japanese or live in japan or know japanese friends... How do the japanese, native people, percieve kanji? I mean, what do they associate the kanji with? How do they "relate" to them, how do they remember them? What are kanji to them?

This is a hard question to answer without linguistic research, but remember that by the time Japanese people begin studying their writing system, they already have a huge spoken vocabulary. So they have a lot of Japanese words to hook the kanji too.

Quote:*) Do they make difference for kanji that are used solo and those that are always used in a compound? Like, 木 vs 召 - do they remember both kanji the same way, applying same rules, or do they remember them differently, because first is used to mean what it means, while second is not?

*) How do they remember those kanji that have a meaning that is never used (旦 召) - do they associate such kanji with a most used compound, and just remember "once = 一旦” and never bother what 旦 means by its own? Do they learn the previous meaning, dictionary meaning, or only the current one?

Meanings are associated with kanji for convenience, based on the words they're used to represent in Japanese. Japanese people learn kanji based on the words they appear in, not some abstract keyword. Even an illiterate person will have some concept of morphemes -- for instance, a child probably has an instinctive notion that there is some morpheme in Japanese "shoku" that is associated with food, because the child is familiar with words like "shokuji" and "shokudou". Later the child will be taught that this "shoku" is represented with the symbol 食, which is also used for the word "taberu" that they already know means "eat". But usually when you hear Japanese people talk about kanji they will say things like 食事の「食」 (the "shoku" in "shokuji"), or maybe just "taberu" in this case.

But because language is messy, the writing system can't be exact. It's almost never the case that you can assign a single, precise "meaning" to a kanji that covers every use of it.

Apologies in advance if you would prefer me not posting on your threads

But I think many of your questions are sort of answered in the other thread with this:

yudantaiteki Wrote:

Istrebitel Wrote:So it's hard to grasp the fact that they do.... aaaaand they don't at the same time! Like, 召 is seduce, but seduce is 誘惑.... 旦 is daybreak, but daybreak is 夜明け / 黎明.

That's because the language comes first, then the writing system to represent the language (as Chris said above). When we say that a kanji "means" something, we're really saying that the kanji is used to represent words in Japanese that are associated with that meaning. It's less exact when English is being used because that creates one more level of separation between the Japanese words and the symbols used to represent them.

"yoake" means daybreak whether that word is spoken, written in romaji, written in hiragana, or written in the combination of kanji and hiragana that's standard.

Foreign learners of Japanese try to attach lots of "meaning" to single characters, but really its backwards. The spoken language comes first, and then characters are used to represent whatever they say. When anyone talks of any "meaning" of a Japanese character, the meaning isn't intrinsic to the character, its just whatever character tends to most often represent such an idea.

If I ask a Japanese person what the single character is for "the sun", they will probably tell me "There isn't a character, its 太陽 which has two characters". Its like when people get tattoo's of Chinese / Japanese characters, they often make no sense because they are taking a single "meaning" which doesn't really exist in the language itself. Its working backwards and often the natives themselves would have no idea what its supposed to "mean".

Kanji characters aren't remembered by meaning. If they wanted to refer to the 旦 character for example, they wouldn't have a single meaning to describe it, they would use a word with how its used. For example they might say 旦那の旦. Then I'd know which "dan" they are referring to.

All characters in that sense are basically the same. Its an alphabet, albeit a large and overly complicated one. Only for a very small number of characters can you say things like "Its 木 because it looks like a tree!".

To answer your questions a bit more directly:

Quote:*) Do they remember kanji as a big alphabet of many "letters" (like 木, 人) and many "combinations of letters"　(like 人+木=休), and spell words using the alphabet - like, "I'm okay" is spelled "big-robust-husband", or "giant" is spelled "huge-big"?

Kind of, its not the meaning though but how its used in a word. The best example is when spelling our their name over the phone or something. Lets say a persons name is 品田 for Shinada (I stole this example from here: http://japanese.lingualift.com/blog/spel...-out-loud/) , over the phone she could spell it out as "The shina in Shinagawa (place name), the ta in tanbo (rice field)", or in Japanese 品田… 品川の品、田んぼの田.

She isn't saying "The first character means item, the second means field" because that would make no sense. She explains the characters by using examples from other words the person listening would know. She isn't using any intrinsic meanings of the Kanji themselves, because they don't have a meaning. Any meaning is merely taken from whatever words they tend to be used in (which itself can be very many for some characters, which is why picking ANY keyword is meaningless)

Quote:*) Do they make difference for kanji that are used solo and those that are always used in a compound? Like, 木 vs 召 - do they remember both kanji the same way, applying same rules, or do they remember them differently, because first is used to mean what it means, while second is not?

No, both remembered the same way. Besides 木 isn't always used on its own, it could be used in a world like 木造, where its pronounced "moku" and not "ki". Most characters are the same.

Quote:*) How do they remember those kanji that have a meaning that is never used (旦 召) - do they associate such kanji with a most used compound, and just remember "once = 一旦” and never bother what 旦 means by its own? Do they learn the previous meaning, dictionary meaning, or only the current one?

They don't learn *any* "meanings" for characters. Meanings is just something foreigners are obsessed with but have no practical use in Japanese (maybe you are starting to see where my anti-Heisigness comes from?). They would associate those Kanji with the words they are used in, so if they wanted to describe 召し they would say "召しの召", or 召喚 's "shou" or something.

Most characters have multiple readings and whether they can be used singularly or not is completely unimportant. Given single characters can be used in so many totally unrelated words too, there is simply no "meaning" except for that which can be loosely derived backwards.

I hope that helps somewhat

Well, for one, words can take on broader meanings or change meanings over time. With 一旦, while the primary definition in my modern Japanese dictionary is "once", it does also have the definition of "one morning", and this is the primary definition in my classical Japanese dictionary. So, the 旦 in 一旦 isn't completely divorced from the meaning of morning.

And in some cases the kanji mean nothing. For example in 旦那, it's actually originally a sanskrit word. The kanji are purely phonetic, only representing sounds here, so the word cannot really be read with kanji meanings.

As yudantaiteki has explained previously and Nightsky below:

NightSky Wrote:Foreign learners of Japanese try to attach lots of "meaning" to single characters, but really its backwards. The spoken language comes first, and then characters are used to represent whatever they say. When anyone talks of any "meaning" of a Japanese character, the meaning isn't intrinsic to the character, its just whatever character tends to most often represent such an idea.

Going out on a bit of a limb, but I think perhaps some of your questioning on how Japanese approach kanji may be due to the fact that native speakers of English and other languages with a phonetic alphabet and nothing really like Chinese hanzi/kanji, what have you, are are sort of approaching it upside down.

While overly simplistic, generally speaking the Japanese learn their kanji similar to how you become literate in your own language: with a mass of words floating around in your head, little free radicals, waiting to be given a more systematic and definable shapes such as the written language.

This is I think a "base fluency" stems from, for native speakers. They have all these familiar sounds for various concepts in which are slowly and systematically reinforced with various characters (for the words which generally use kanji). They are familiar with the basic layout of the terrain and are getting used to the map; which in turn helps them explore places they were previously unaware of.

Also, as patriconia highlighted, some kanji are phonetic markers:

patriconia Wrote:And in some cases the kanji mean nothing. For example in 旦那, it's actually originally a sanskrit word. The kanji are purely phonetic, only representing sounds here, so the word cannot really be read with kanji meanings.

NightSky Wrote:They don't learn *any* "meanings" for characters.

This may be going a bit too far -- the books they learn from in elementary school and the dictionaries do provide meanings for the kanji. It's just not their primary means of remembering the characters. Even though my kanji dictionary gives よあけ as a meaning of 旦, they're not going to learn that shape as よあけ, and then that it's used in 元旦 or 一旦.

Incidentally, it's possible to learn kanji without meanings at all. The main books I learned to read Japanese (Japanese The Written Language, Basic/Intermediate Kanji Books, and Kanji in Context) did not provide any meanings for the kanji, just the words they were used to write. (BKB/IKB may have had meanings in the back, I don't remember.)

yudantaiteki Wrote:

NightSky Wrote:They don't learn *any* "meanings" for characters.

This may be going a bit too far -- the books they learn from in elementary school and the dictionaries do provide meanings for the kanji. It's just not their primary means of remembering the characters. Even though my kanji dictionary gives よあけ as a meaning of 旦, they're not going to learn that shape as よあけ, and then that it's used in 元旦 or 一旦.

Really? I won't doubt what you'd say, but I didn't think Japanese kids ever learnt any Kanji's with "meanings" separate from the words they are written in. Are those Kanji dictionaries used so much as part of the standard curriculum though? Given the first 1000 or so characters are pre-decided which they learn year by year, I thought they were just used in words they would already know.

Quote:Incidentally, it's possible to learn kanji without meanings at all. The main books I learned to read Japanese (Japanese The Written Language, Basic/Intermediate Kanji Books, and Kanji in Context) did not provide any meanings for the kanji, just the words they were used to write. (BKB/IKB may have had meanings in the back, I don't remember.)

Same here, I've never learned meanings / keywords and have no regrets with that decision

NightSky Wrote:Really? I won't doubt what you'd say, but I didn't think Japanese kids ever learnt any Kanji's with "meanings" separate from the words they are written in. Are those Kanji dictionaries used so much as part of the standard curriculum though? Given the first 1000 or so characters are pre-decided which they learn year by year, I thought they were just used in words they would already know.

That's mainly what they do, but the textbooks and kid dictionaries give meanings for reference. As far as I know it's not something they need to memorize and they're never tested on whether they know the "meanings" apart from the words, but the information is provided.

I think that most Japanese people are familiar with the concept of a kanji having abstract meanings, it's just not the main way they learn or remember them.

http://users.bestweb.net/%7Esiom/martian.../5-5-2.swf

buonaparte Wrote:http://users.bestweb.net/%7Esiom/martian.../5-5-2.swf

Interesting.

I'll jump in on this one.

When you see the words/images 2 or 3, etc.. how to you perceive it ?
You could think of those as kanji.
Just lines drawing some random shape that we have some meaning/words for.

After a while kanji become like that, you just immediately know what it represents.

Thanks everybody, especially NightSky (your long explanation was most informative)!

I have spent some free time going through Kanjidamage and http://www.guidetojapanese.org/learn/grammar and it kinda clicked for me. Now I read your explanations and they do make sense!

When I was going over the kanji in kanjidamage, I was like: okay, in order to remember that 計 is ケイ and keyword is measure I have to remember the mnemonic "It is o'KEI (ok) to measure SayTen (satan)!"... And so on. But it kinda felt awkward. Will I have to remember 2K stories / mnemonics just to remember all the kanji? I could probably, but this felt just awkward, kinda... wrong? What would I do then, most of these kanji's "keywords" are derived from compound words, so in order to actually "spell" this keyword I'd need to remember which word to "attach".

And then, when I met words I already know from my long story of watching anime / super-sentai / kamen rider, I was like "oh so that's it!". Like, with that "eienni" that keeps popping a crap lot in romance scenes in anime, how can I forget that 永　is エイ? Or knowing that "denwa" is telephone, how can i forget that 話 is ワ? And so on. Daijobu, Shinkansen... Remembering these kanji, including pronounciation, felt just natural when I knew how I would use them.

I guess I'll keep studying actual japanese, and learn kanji in the order they pop up in the texts I'll encounter. I don't think I'll aim to "know all 2K kanji in 3/6/9 months" but rather to just be able to hear, read, write and speak as much as possible.

However, I must say, without Heisig's method (or kanjidamage) I wouldn't get past the "kanji are intimidating" stage. Even if the method of "associate kanji with a single keyword" may not be the best way to learn kanji, the method presented in typical "classic" books on japanese like "Genki" is just no good as it is, because nothing gets explained to you! You just get kanji after kanji with no explanation what they consist of, what's the logic, etc. "There's your 時, there's your stroke order, remember all strokes in this order and ask no questions."

I suppose for me, Heisig's method / Kanjidamage / other such methods were most useful because they showed me how kanji actually work (even if it wasn't their intention, meaning, they're methods of learning all 2K kanji quick, not books on kanji theory). Like, for example, how 熟　and 塾 have same ON because they have same "picture" which is "modified" by fire or earth below, or how 間　and 燗, or 寺　and 時　have the same ON because the composite kanji "inherits" the ON of one of its components, or what are typical "building" blocks out of which bigger kanji are made (and how kanji "mutate" when included into other kanji). Just by going with either order - Heisig's or any other, it naturally occurs to you after you process about a hundred or two. And then learning any kanji becomes... how should I say, easier than before.

And also, without kanjidamage (or Heisig, but it's harder to use it this way) it would still be very hard to learn complex kanji when you encounter them before you encounter every part of them. With kanjidamage I can just let it tell me what they consist of, and learn every part first, at least without any meaning/sound, as a pure radical, so this is useful - don't have to remember over 20 strokes of pure nonsense.

PS: I must also mention, http://www.guidetojapanese.org/learn/grammar is AMAZING. The author is just a natural teacher, he explains so well, so much and in such a compact manner.

I'm glad you've found a method that works for you!

Exactly; Because that's what really matters.

Apparently people above learned the Kanji without any attached meaning for each character. I can say that that idea would never have worked for me. My brain *needs* each kanji to have a distinct meaning. It's how I remember things.

There isn't any reason why *not* to assign characters meanings (no matter how convoluted they may be) if it helps you remember the characters. Just keep in the back of your mind that these meanings aren't necessarily intrinsic to the characters and you'll be fine.

Avoiding English meanings can help you associate the kanji with Japanese meanings instead, which I find tends to be better because the Japanese meanings are the actual words the kanji is used to write (in most cases). I would rather associate 密 with 秘密, 密度, and 密着 rather than whatever English word you come up with.

yudantaiteki Wrote:Avoiding English meanings can help you associate the kanji with Japanese meanings instead, which I find tends to be better because the Japanese meanings are the actual words the kanji is used to write (in most cases). I would rather associate 密 with 秘密, 密度, and 密着 rather than whatever English word you come up with.

That seems to be a common approach to studying kanji amongst learners who collected sufficient amount of Japanese vocabulary. For the beginners however, the question that arises is: "How to associate a meaning with a kanji when you don't know enough Japanese?"

There are couple of answers/approaches that you could apply:
1. Use meanings with words from your first language (ala Heisig).
2. Delay kanji study until you know sufficient amount of Japanese vocabulary (meanings of words + pronunciation but no kanji readings yet).
3. Learn vocab together with the kanji (ala KO, KiC)

Which of those is the most efficient is not quite clear to me...

Inny Jan Wrote:That seems to be a common approach to studying kanji amongst learners who collected sufficient amount of Japanese vocabulary. For the beginners however, the question that arises is: "How to associate a meaning with a kanji when you don't know enough Japanese?"

Why do you have to learn or associate any meaning with Kanji you don't know any words for? Why do you write that as if it is a necessity?

If someone asks me "Hey, I don't know any Japanese so how am I going to remember all those characters?"

The pretty obvious reply is "Well learn some Japanese then!"

NightSky Wrote:Why do you have to learn or associate any meaning with Kanji you don't know any words for? Why do you write that as if it is a necessity?

You seem to be criticising Heisig's advice, which is ok by me BTW, but there people out there who start their studies of Japanese with zero knowledge of the language and Heisig's advice is what they follow. (Others, like AJATT, support doing RTK at the very beginning as well.)

I'm just pondering about relative value of those three strategies for kanji study when you start from ground up.

I'm sure I've seen instances where individual kanji have been pointed out ("This person's name has the kanji for x and y"), or used to make a pun based on changing a kanji with a different meaning but the same pronunciation.

That's common for names; today I was telling someone how to spell the personal name of Kitamura Kigin and I said 吟味の吟と季節の季; the person verified the first kanji by asking 口に今？ A common myth is that Japanese people are constantly doing that (plus drawing kanji in the air or on paper) even in casual conversation. The truth is that it's hardly ever done except for names or occasionally some technical or obscure word that the speaker thinks the listener might recognize from the kanji rather than just explaining the meaning.

I don't think that has much to do with English keywords, though.

NightSky Wrote:

Inny Jan Wrote:That seems to be a common approach to studying kanji amongst learners who collected sufficient amount of Japanese vocabulary. For the beginners however, the question that arises is: "How to associate a meaning with a kanji when you don't know enough Japanese?"

Why do you have to learn or associate any meaning with Kanji you don't know any words for? Why do you write that as if it is a necessity?

If someone asks me "Hey, I don't know any Japanese so how am I going to remember all those characters?"

The pretty obvious reply is "Well learn some Japanese then!"

If it works for you. What I found that RTK helped with is that removing that "learn some Japanese" is that I don't have to juggle sounds that are removed from my native language and tie them into something completely alien (In terms of writing I mean, though I have been told my English handwriting comes off looking as bad as some kanji).

Oddly enough, having done RTK, it's helping me with core2/10k as I can very easily assign the kanji that I know to the words now.