I find that every so often i get the left/right side of my kanji swapped when I am reviewing. I'm surprised that it dosen't happen more often. But I just get a 'feeling' for which one should be where.

It could simply be because it seems like some radicals usually end up on the left, and others usually end up on the right.

Could there be more to the pattern though? What if there are tiers of radicals. Tier A always beat radicals in tier B, and end up in a more dominant position. A and B both beat C, and so on.

I imagine that there would be exceptions, but if they are few, you could simply add position to the stories for those kanji, and then use the hierarchy to figure everything else out.

Has anyone looked into this?

There are classes such as 偏（へん）, 旁（つくり）, and 冠（かんむり）when it comes to radicals etc. These can specify the positions of certain components e.g., radicals that are classified as 旁 should always be on the right. This is why 氵 is always on the left side.

I don't know any other rules. But as a native Japanese speaker, I had used tons of kanji everyday and kind of think you can form a set of rules that can roughly explain the position of each component in a given kanji in a logical way like the one you described. This is certainly an interesting topic, and there are researchers studying kanji and their etymology. So I'm guessing those rules are already studied very well.

But I'm not sure if they would be practical for non-native speakers learning Japanese. If you mastered the rules, I think you'd be able to explain why a kanji is written that way with certain logic or able to tell if it's an exception to the general rules. I suppose you could say your understanding of kanji would become much better and sophisticated than that of native speakers in a way. But obviously that kind of deep knowledge is totally useless for reading kanji. It's a lot faster to accept each kanji as it is without any explanation of why it's written that way.

If you try to write a kanji by applying the rules, I think you would get an unnatural kanji that kind of looks legit. Your kanji may always sort of make sense, and sometimes you could even get a 100% legit kanji. But in most cases your kanji would be of the "native speakers wouldn't write that way" kind.

Of course, it's not entirely useless to learn the rules. For example, rules might help you memorize very complicated rare kanji even native speakers can fail to write properly. If anything, they could work as good mnemonics for any kanji. It's a very interesting topic in itself too. But by the time you master the complicated rules and tons of exceptions, you would have mastered all the kanji through reading/writing/SRS. So I kind of think they're pretty much useless for reading and writing kanji like a native speaker.

So, I think it's a very interesting topic, and probably academia has studied it well. But your average native speaker doesn't know those rules, and if I were to relearn Japanese, I'd take the "Don't ask why because that's the way it is" approach to read and write kanji like a native speaker.

As magamo says, there's ones that appear on the left always, on the top always and on the right always.

There's a couple of techniques or methods you can use to encode positioning data into your mnemonic;

Firstly and this I'd recommend the highest: a good mnemonic system is a system that is systematic. Seems logical enough but Heisig's is a little too flexible when it comes to word order. It works and the flexibility allows for creative stories which is definitely a merit but if you'd like to try and design out the flaw of forgetting the order in which to write things then I suggest that when making a mnemonic, every single time start with the keyword (very important) and then use the name of each primitive in the order they are written in.

Secondly, and this is what I did when working through Heisig if one was giving me trouble: I used to sometimes include a word (same one every time, remember keep it systematic) that indicated the position of a radical. Mainly I used the word "strong" to mean that that primitive was written on the very left... so for example　校 which really means "school" when you think about it you might make a mnemonic "School has strong trees in succession"* so you KNOW that "tree" is the きへん radical because it's built into the mnemonic.

木　＝ tree, 交 = alternately/in succession, 校　＝　school. (following a true meaning rather than heisigs names). (this is just for an example so no flames)

If you want you could combine both techniques and it would work well. I recommend the first one over the second because if you always go Keyword, first primitive, second, third, ect. Then you're keeping it systematic aswell as tying in a bit of kinisthetic learning by doing it in the right order and each mnemonic goes in the same order. It might be a little trickier at times to make stories for but I've seen this technique used to awesome effect and the stories we're just as creative as i've seen here so it's doable.

[Katsuo]

Here are some more kanji from RTK1 where one component is in an atypical place. Sometimes this can be predicted because the left component is usually related to the overall meaning of a kanji. E.g. 吐 "spit" is more to do with mouth than soil, so we might expect mouth to appear on the left.

94 頂 place on the head: spike usually on right
129 順 obey: only time river on left
151 吐 spit: soil usually on left (see also 1092)
226 寛 tolerant: only time flower is under another primitive
278 導 guidance: road is usually at bottom
424 曇 cloudy weather: weather usually on top (see also 1068)
443 敵 enemy: antique usually on right
459 敏 cleverness: every usually on right
651 拓 clear the land: only time stone on right
965 休 rest: tree usually on left
969 信 faith: only time say appears on right
1068 漏 leak: weather usually on top (see also 424)
1092 社 company: soil is mostly on left (see also 151)
1540 静 quiet: blue usually on right
1718 須 ought: shape usually on right
1782 助 help: bookshelves usually on right
1846 都 metropolis: only time someone on left

igomadness Wrote:Could there be more to the pattern though? What if there are tiers of radicals. Tier A always beat radicals in tier B, and end up in a more dominant position. A and B both beat C, and so on.

"Taskmaster" and "head" are strong right-side components, but I'm not familiar with any comprehensive ranking system.

Katsuo Wrote:Here are some more kanji from RTK1 where one component is in an atypical place. Sometimes this can be predicted because the left component is usually related to the overall meaning of a kanji. E.g. 吐 "spit" is more to do with mouth than soil, so we might expect mouth to appear on the left.

94 頂 place on the head: spike usually on right
129 順 obey: only time river on left
151 吐 spit: soil usually on left (see also 1092)
226 寛 tolerant: only time flower is under another primitive
278 導 guidance: road is usually at bottom
424 曇 cloudy weather: weather usually on top (see also 1068)
443 敵 enemy: antique usually on right
459 敏 cleverness: every usually on right
651 拓 clear the land: only time stone on right
965 休 rest: tree usually on left
969 信 faith: only time say appears on right
1068 漏 leak: weather usually on top (see also 424)
1092 社 company: soil is mostly on left (see also 151)
1540 静 quiet: blue usually on right
1718 須 ought: shape usually on right
1782 助 help: bookshelves usually on right
1846 都 metropolis: only time someone on left

igomadness Wrote:Could there be more to the pattern though? What if there are tiers of radicals. Tier A always beat radicals in tier B, and end up in a more dominant position. A and B both beat C, and so on.

"Taskmaster" and "head" are strong right-side components, but I'm not familiar with any comprehensive ranking system.

This is one of the most interesting things I've learned through my KanKen study and I understand why all of these appear where they do. I'm however still unaware of the ranking system as to which is the strongest but whilst the kanji is made of all it's various components only one is THE 部首 and that gets the strongest position.

I'll go through each one.

94 頂 place on the head: spike usually on right

The main radical here is おおがい. Radicals on the right are named おおxxxx and since radical is based on 貝　it gets the name おおがい and is stronger than the other component.

129 順 obey: only time river on left

See above.

151 吐 spit: soil usually on left (see also 1092)

The main radical is 口 and not 土, if it we're 土 then you'd have 土 appearing on the left. Take a minute to think about which radical more strongly relates to the meaning, the dirt or the mouth? Of course it's the mouth and that's why 口 is the main radical here and thus beats out 土 for position.

226 寛 tolerant: only time flower is under another primitive

the main radical is ウ冠 which is the ウ "crown". Every 冠 (kanmuri) radical always appears at the top and is the main radical for that kanji usually. (sometimes there are weird exceptions in the world of kanji but that's a pretty solid rule).

278 導 guidance: road is usually at bottom

The main radical here is 寸 which always appears under things or to the right of them. The meaning of the kanji is guidance and it's very elegantly depicted as "stick to the road"... the roads just in the kanji because it's part of the explanation but the glue is the main reason behind the kanji, because you're being told to "stick to" something.

424 曇 cloudy weather: weather usually on top (see also 1068)

Think about the meaning of 雲 (kumo = cloud) and 曇り (cloudy). A cloud is a cloud but clouds make for a "cloudy day" and so in this kanji　日 is the main radical because it changes the meaning in a significant way. Make sense? Also 日 position as a main radical is usually above shit.

443 敵 enemy: antique usually on right

the main radical here is のぶん (the part on the right) and のぶん always appears on the right and in terms of ranking it out ranks whatever else it's put with (as far as i've seen). So if you see a kanji with のぶん　(task master) then the main radical is のぶん

459 敏 cleverness: every usually on right

See above.

651 拓 clear the land: only time stone on right

手編 is a hen radical which means of course it's going to be on the left. Think about how many kanji have 手編 as a radical... it's lots and lots and lots so of course it's going to outrank 石 which is in relatively few kanji. Mostly though if you think about the meaning of "clearing the land"... rocks can't clear land, only the hand can so it's the "hand moving the rocks" which makes the hand the main focus of this kanji and hence it's the main radical so it appears on the left.

965 休 rest: tree usually on left

人偏 is probably one of the strongest radicals there is, it appears as the main radical in so many kanji. When you see a kanji that uses this, it's the main radical. Think about the meaning... is it the person resting up against the tree? Or the tree resting up against the person? The person is the one doing the resting and this it's the main focus for the kanji and takes the position as the main radical, when 人 is the main radical it appears on the left as にんべん

969 信 faith: only time say appears on right

Again see above. 言偏 (the radical where "say" appears on the left) isn't the main radical here because here the faith refers to the person （人） that the people express their faith in. 言 is in this kanji because when people express their faith it's most often through words... think of the phrase "I have faith in you" or "I put my faith in you"... who is the main focus of that statement? Well it's the person isn't it? That's why 人 is the main radical (人偏)

1068 漏 leak: weather usually on top (see also 424)

This is an interesting one because it's very difficult to decipher. Whatever 屍 has to do with the meaning of this I'm guessing. This kanji is made up of さんずい　(the 3 water drops) 屍 (しかばね the corpse radical) and of course 雨 (rain). Nothing usually ever appears above 屍 and it would seem especially odd putting 雨 above it and if you did that where or how would you incorporate the water without ending up with a messy, ugly kanji? This configuration is logical for writing purposes and also the meaning has to do with leaking and often it's water that is being leaked, not necessarily rain water.

1092 社 company: soil is mostly on left (see also 151)

示偏　（しめすへん) is the main radical and that's why 土 is on the right.

1540 静 quiet: blue usually on right

This one is quite confusing. End of story.

1718 須 ought: shape usually on right

Again, see おおがい as it's the main radical here.

1782 助 help: bookshelves usually on right

力 is the main radical and as a main radical is always appears on the right. "bookshelf" is never ever a main radical. 祖 notice where it is used in this kanji? that's right, the aforementioned 示偏 is the main radical here and as such appears on the left. But yeah, 力　is the main radical in a couple of kanji but not too many. 励　(はがます、to encourage) is a good example of a kanji where ちから is the main radical and if you think about it's meaning you are being "strongly" urged to do something so the idea of strength influences the meaning in the greatest way.

1846 都 metropolis: only time someone on left

大里 (おおざと big village) is the main radical here and it always appears on the right remember the おおxxxx naming convention? 流石！阜偏　（こざとへん　the small village) is what it's called when teh radical appears on the left (like in 際) and again here if you think about which component is more influential to the meaning, it's obviously the "village" aspect of it as this refers to suburbs more than the people in them.

Wow, some really great info here.

I'm still working through rtk1, but I think from now on I will take some more time building position into my stories.

I notice that a lot of the kanji in the 'exceptions' list are ones that have given me trouble. It should be useful to at least flag the 'oddballs' with an asterisk or something so that people can take extra care with those stories.

It seems like there is enough here to warrant a bit of number crunching, play around and see what I can find. At the very least, it should be easy to find kanji with radicals in 'statistically unlikely' positions. I'm sure that there have been a lot of great linguists who have looked at this stuff already, But maybe those linguists did not have access to modern databases or have a pet bored software engineer at their fingertips.

Unfortunately before I can do that I need to create some data to consume, which means looking at a lot of kanji and creating lists of which radical goes where. Before I can do anything like that I suppose I need recognize the radicals, (im only in the 1200s in rtk currently) Oh well, something to come back to in a few months.

I'm pretty sure visual memory would play a bit of a role in positioning. I doubt that it is due to a 'perfect story' that makes one remember how to write 校 correctly, but rather you seeing 校 written as 校 and not 交木 or 木父亠 or 木亠父 or ........................ (Probably the amount of e-ink required to write the character). Just like how it's impossible to make a story for the character 木. If one were to write 木 with the last stroke running a little short, then it would be odd.

As you go on you'll find that your brain will create patterns and rules to the kanji you're learning since that is what we humans are good at, finding patterns. Once you start learning actual Japanese, which will take longer if you're too concerned about making ridiculous stories while paying attention to word positioning, which is to me, basically rote learning a mnemonic story since no one should learn a character by memorising a sentence/story in the word order. You can, and I do, and people do, for some characters here and there but it will be impossible to do a heap load. You ought to have a visual idea of something that makes you write the character correctly, not a stringed line of words that helps you string a character together.

Anyway, I don't think learning radical dominance will help you at all. It's interesting to know, but you only really need to learn that stuff if you're going to take 漢字検定 or something. If you're actually interested, by all means go for it learning anything is learning after all, but it won't help you become 'better' at Japanese at all, just like how studying etymology won't help you become better at Japanese.

I'd love to undertake a full study of radicals and if you do wind up putting something seriously awesome together it'd be a huge benefit to the Japanese learning community.
If one does do a proper study of radicals it can actually yield awesome benefits.

Just like there is an order in which you write the strokes of every kanji that follows a systematic formula, there is definitely an established pecking order with kanji.

The above examples of "troublesome" kanji (i'll admit they gave me some shit when I was doing heisig too) are actually not problematic at all when you what the main radical is and why it is the main radical.

However, HERE is a problematic kanji： 加

力 as a main radical is almost always to the right (and if it's not it's because there is something that always goes on top so it instead appears under (but is still the main radical).

You can see what I'm talking about here; 力 加 努 劫 助 労 劾 効 勁 勃 勅 勇 脅 勘 動 勤 勦 勢 勧

Which leaves us with 加 which seems to buck the trend, but why? Well, the answer is that pecking order is to blame. Take a look at the following kanji, it's a list in which 口 is the main radical for all.

口 叭 叶 叱 只 兄 叩 司 右 吃 叫 号 吐 吊 吸 呂 吻 呈 呆 吠 吟 君 告 吹 呟 咆 咀 咎 咄 呻 呪 味 呼 哄 咽 咳 咲 品 唄 哮 員 哺 哲 啜 唸 唾 唖 啓 唱 唯 喩 喀 喝 喘 喰 喇 喚 喋 喉 喫 喧 営 嘩 嗚 嗣 嗟 嗜 嗅 嘆 嘛 嗽 嘔 鳴 噌 嘱 噎 噂 嘲 嘘 器 噴 噬 嘯 嘴 嚇 噛 嚼 囀 囁 囃 囂 轡

Notice a pattern? As a main radical it's always on the left and if not it's underneath because there is a radical that just has to be on top. Also notice despite having 口 in it 加 is not on this list and rightly so because 口 is not the main radical,　力 is. You're probably clicking righttt about now...

The fact of the matter is due to pecking order, if 口 we're to appear on the left and 力 on the right then A: it would make it undeterminable what the main radical was B: 口 would most likely be deemed as the main radical if one had to choose but that's incorrect. So to solve the problem 口 goes on the right which indicates it's not in a strong position and therefore 力 is the main radical.

See how much sense kanji actually makes when you study how it really works? It's ***** amazing! The more I learn the more it wows me.

liosama Wrote:I'm pretty sure visual memory would play a bit of a role in positioning. I doubt that it is due to a 'perfect story' that makes one remember how to write 校 correctly, but rather you seeing 校 written as 校 and not 交木 or 木父亠 or 木亠父 or ........................ (Probably the amount of e-ink required to write the character). Just like how it's impossible to make a story for the character 木. If one were to write 木 with the last stroke running a little short, then it would be odd.

The more systematic your system is the less problems you're likely to run into with it. Visual memory is the absolute key, yes but that's not to say that Heisigs method couldn't be "tidied up" a little to economize it a bit further which is all I was suggesting.

Actually, the main reason I was replying is because whilst you don't need one, it is actually possible to write a story for 木. The way Japanese people describe parts of kanji that don't class as anything at times is just to simply use words like　横 and 建. So one could describe 田 as 建、横、建、建、横、横 or something to that effect. Of course it's just 口と十 really but Japanese people often use this way of describing things in Kanji which don't fall under a nicely defined category for them. It doesn't really play a role in the simple kanji because they're simple but it does in more complex kanji and in particular ones they're not familiar with.

I think if you learn how to identify the main radical of a Kanji (which is easyyy) then it gives you a one up in remembering how to write it easily. So in that respect it won't make your Japanese "better" but it will expand your knowledge and help you assimilate new kanji EASIER than flying blind all the time.

Of course Anki will take care of the rest.

I can barely remember what I was talking about now as I only half-applied them, but here's my thoughts on semantic/phonetic radicals in Japanese and 'new' ways to use them: http://forum.koohii.com/showthread.php?p...2#pid81762 (HBPK refers to my 'how the brain processes kanji' thread: http://forum.koohii.com/showthread.php?tid=4612 ;p)

igomadness Wrote:I'm sure that there have been a lot of great linguists who have looked at this stuff already

Obviously those guys who created simplified kanji/Chinese characters knew position rules, etymologies and whatnot. And most likely they created new rules or guide lines based on them so simplified versions follow similar rules as well. I'm guessing the rules are not that esoteric. I think it should be pretty common knowledge among experts.

By the way, if you google for 漢字　部首　法則 on google.co.jp, you can get tons of information like "How do I know which part is the main radical?" It seems many native speakers are saying the rules are so complicated it's better rote learn everything. But maybe there are some useful simple rules non-native speakers can take advantage of. I run into this book when I did cursory googling. Anyone know what this is like?

magamo Wrote:

igomadness Wrote:I'm sure that there have been a lot of great linguists who have looked at this stuff already

Obviously those guys who created simplified kanji/Chinese characters knew position rules, etymologies and whatnot. And most likely they created new rules or guide lines based on them so simplified versions follow similar rules as well. I'm guessing the rules are not that esoteric. I think it should be pretty common knowledge among experts.

By the way, if you google for 漢字　部首　法則 on google.co.jp, you can get tons of information like "How do I know which part is the main radical?" It seems many native speakers are saying the rules are so complicated it's better rote learn everything. But maybe there are some useful simple rules non-native speakers can take advantage of. I run into this book when I did cursory googling. Anyone know what this is like?

Yes I actually tried it out for a bit. It sucks. It's just like Heisig but less systematic. And

E.g teaches water radical as シ　and the 'school roof' radical as ツ　and makes stories around them. I only went through like 50 or so characters.

The best book (I know of so far) that goes through 'radical theory' if you may is Kodansha's Lets Learn Kanji - 1997


example^

lion: the "school roof radical" IS ツ (in Japanese). It's called つかんむり ("tsu crown").

Just found this site: http://www.nisk.jp/shodokisochishiki/sos...nmuri.html

Jarvik7 Wrote:lion: the "school roof radical" IS ツ (in Japanese). It's called つかんむり ("tsu crown").

Just found this site: http://www.nisk.jp/shodokisochishiki/sos...nmuri.html

Yeah I know but I'm saying it's not like that helps you learn anything new, better.