Anyone know of an available list of kanji that are easily confused due to visual similarity? A list of paired or mulitple kanji having similar form.

Well, actually I don't confuse a whole lot of kanji with the Heisig method, because every story has a different set of elements and a new visual image (of course, I have a problem mixing up certain key words instead, but you don't mean that).

What's left are kanji that have the same elements, but in different order. There aren't a whole lot of those in the first 2042. I can recall 含 and 吟, and 細 and 累. There may be a few more, but not that many I think.

As for the rest, I think it's rather personal which ones with different elements you tend to mix up, so for that ultimately, you might have to make your own list.

astridtops, thank you for your attention but your reply was a bit off target. I merely asked if you know an available list of visually similar kanji. Your reply may make some people think that kanji are easily distinguished. They are not.

Please see for yourself by going into the real world of Japanese writing and start reading real Japanese discourse. If you never confuse any kanji then you, indeed, sir or madam, have exceptional, I mean exceptional, perception. (And by the way, Japanese kanji is not limited to RTK kanji. RTK is just the first step!!)

While I use RTK, I want to make necessary visual discriminations, which only can be done by reference to kanji with similar visual elements. A ready-made list of such candidates would be an enormous help, I believe.

When you say visually similar kanji, do you mean primitives that looks very similar? Like the primitives altar and cloak?

Visually similar primitives isn't a problem with Heisig's method (I'm guessing you are not very far into RTK), because you review from keyword to kanji and it's a bit difficult to confuse an (mental) image of a cloak with an altar. They are written very visually similar (the difference is a drop), but such cases is solved by introducing one of them very early so by the time you reach the other, you know very well by then how to write (and recognize) the first one (or as with column-person, you meet the latter in 50 kanji in a row :S).

Personally, I had much trouble with king and jewel as primitives (jewel/ball was written as king when on the left), but solved that by ignoring the rule and simply decide: if it's written like king, it is king.

I don't have easy access to such a list but...

Three characters that are confusing based mainly on visual similarity, but also because the first and third share an ON reading are 緑, 縁, and 録


儀, 義 and 議 are confusing because they have similar, abstract meanings, and the same pronunciation. Some compounds can be alternately spelled with more than one of them, to further confuse the issue, but that is not to say that they are interchangable.

像 and 象 are confusing for the all the same reasons as above, with the added confusion that they have multiple readings.

In some cases, interchanging one of the above in a compound will result in a different compound with the same pronunciation but a different meaning.

Pauline wrote: "When you say visually similar kanji, do you mean primitives that looks very similar? Like the primitives altar and cloak?"

No, I don't mean primitives. I mean kanji (although some kanji are also Heisig's primitives).

"Visually similar primitives isn't a problem with Heisig's method (I'm guessing you are not very far into RTK)"

Actually, I did the program years ago but halted my J studies due to pressing life transitions. (They can be very pressing.) Now I am trying to recapture my study method and recover my RTK materials because I want to re-learn it.

Let me make this simple. Orthographic systems much simpler than kanji, like the Roman alphabet or Japanese kana intrinsically, have perceptual pitfalls for learners. Graphic forms that are similar will be transposed and confused. I am not talking about RTK at all; I am talking about Japanese kanji. Jimmy Seal has posted some examples.

If for one minute you think that because you can zip through RTK, you will soon be reading Japanese newspapers, then you are in for a big, big letdown. This RTK deal is just a drop in the bucket. Brace yourself for the long haul ahead because there is no shortcut to Japanese literacy.

You must go beyond RTK to understand this point. Surely within the deep blue sea of Japanese writing you are going to confuse kanji due to graphic similarity at one time or other. I am really surprised that my inquiry about a list of confused kanji has elicited some denials of the phenomenon. If some people so wish to, that's fine. But it draws attention away from my request.

So readers be on the lookout for any such list of confused kanji; or we can work together to make such a list.

I've seen books on this subject, but I don't know of any free online listing of easily confused kanji.

KANJI Wrote:Please see for yourself by going into the real world of Japanese writing and start reading real Japanese discourse. If you never confuse any kanji then you, indeed, sir or madam, have exceptional, I mean exceptional, perception. (And by the way, Japanese kanji is not limited to RTK kanji. RTK is just the first step!!)

KANJI Wrote:If for one minute you think that because you can zip through RTK, you will soon be reading Japanese newspapers, then you are in for a big, big letdown. This RTK deal is just a drop in the bucket. Brace yourself for the long haul ahead because there is no shortcut to Japanese literacy.

So readers be on the lookout for any such list of confused kanji; or we can work together to make such a list.

Kanji-san,

Your messages in this thread certainly do a good job of positioning the Problem: is there any particular reason you have chosen not to cut on the lengthy explanations by actually providing a few Examples? That would be a great start for working together and is surely within the grasp of someone who's gone into the real world of Japanese writing and touched real Japanese discourse.

As the existing resources go, http://www.zhongwen.com is probably the best reference point as it allows to group and review kanji by primitive elements - big and tiny.

CharleyGarrett Wrote:I've seen books on this subject, but I don't know of any free online listing of easily confused kanji.

Apparently there are quite a few books produced in Japan for native speakers on this subject 「間違いやすい漢字」 so this is clearly troubling to many Japanese! (Japansese confuse and forget kanji too, I've seen this happen numerous times.) There is a book for English speakers called "Kanji Connections: How to Differentiate 400 Easily Confused Characters".

Personally, I have enough to be getting on with learning the ones that aren't easily confused, so I'm not going to worry about this now. There's only so much one can take in at once; recognition of this fact is a pillar of Heisig's method and there's a danger of "dividing the attention" too much.

KANJI Wrote:Anyone know of an available list of kanji that are easily confused due to visual similarity? A list of paired or mulitple kanji having similar form.

I read a little. You're absolutely right about confusing characters. In my case the confusion started when I began reading at "normal" speeds, rather than staring at words for a long time. So when/if we get a list, what do you think the best way to study it is?

[ファブリス]

Whatever the method you use you're bound to run into this problem.

Visually similar kanji could be just a transitory problem.

For us non-natives I guess it will take several years of frequent reading, to develop a visual memory. That kind of memory that lets you recognise words simply by their length, and overall shape.

Now fastforward a few years of Japanese reading, you begin to recognise those shapes. At that point you're no longer trying to pick out each single kanji, you recognise whole compounds, for example I'm sure you've seen 部屋 quite a lot.

Then recognising individual kanji at quick speed becomes a non-issue.

Even for those kanji that appear alone, again, with frequent reading, I imagine that you get the drift of the sentence ALREADY as you are skimming through the sentence, and when you reach the character you guess which one it is by context, without carefully looking if it's got the differentiation stroke her and there.

The exceptions would be if you had another word like 部屋 where one of the kanji would look very similar, but woudl be different, AND the compound would have a meaning not so far from "room" so that the context of the sentence would not help. Nothing is perfect. In that case I'm sure the Japanese would have to deal with it as well, I dont see how you';re going to solve this other than eating plenty of carrots ? (good for the eyes I hear).

I guess it would be like this in english:
"For meditation of meditation to occur, the art of mediation needs to be improved through meditation. Could it be that meditation's mediation is a kind of meditation unto itself? I suggest that we mediate instead of meditate or perhaps meditate to mediate the problems of these times. Let us meditate the immediate thought: mediation or meditation, such is the question."

Can you read it all quick without hesitation? ^_^

[ファブリス]

Quote:Apparently there are quite a few books produced in Japan for native speakers on this subject 「間違いやすい漢字」 so this is clearly troubling to many Japanese! (Japansese confuse and forget kanji too, I've seen this happen numerous times.) There is a book for English speakers called "Kanji Connections: How to Differentiate 400 Easily Confused Characters".

That's very interesting Pangolin. But I'm not surprised... actually I believe we have an edge on this problem thanks to RTK.

KANJI: Ah, now I get the question. Apologize for making asssumptions. As I'm still in the beginning of learning to write/read, I tend to focus only of the obstacles I have in front of myself, and forgets that others isn't at the same low level (books and newspapers seems very far away, especially at a decent speed).

ファブリス Wrote:Even for those kanji that appear alone, again, with frequent reading, I imagine that you get the drift of the sentence ALREADY as you are skimming through the sentence, and when you reach the character you guess which one it is by context, without carefully looking if it's got the differentiation stroke her and there.

The exceptions would be if you had another word like 部屋 where one of the kanji would look very similar, but woudl be different, AND the compound would have a meaning not so far from "room" so that the context of the sentence would not help. Nothing is perfect. In that case I'm sure the Japanese would have to deal with it as well, I dont see how you';re going to solve this other than eating plenty of carrots ? (good for the eyes I hear).

I've experienced and read about that for other languages, but the same should apply to Japanese as the brain works that way. When I read, I skip a lot of words and phrases I know well, because the brain gives us the likely missing information based on shape recognition and the context.

Take ファブリス exempel with meditation and mediation; I see a lot of 'meditation' in the text and completely misses the first 'mediation', but when it shows up again I feel something is off. Would I have noticed it if 'meditation' had been replaced with 'mediation' and it was still contextually right? Or would my brain filter it away as a misspelling?

So the problem shouldn't be so much visually similar kanji as that they are part of words used in similar context, but also that one kanji/word occurs more than the other.

Thank you all for your postings on this. I feel it got us over a big hump. Please let me clarify.

RTK can be studied, absorbed, conquered in various ways. When I did much of the program a long time ago, I did so while trying to learn related kanji readings (the pronunciations) and kanji compounds (words). I had a stack of reference books along side RTK! I mastered half or more of RTK before I had to quit kanji study. (I am fairly fluent in the language however).

This time around in my study, I am going to do RTK "straight" but with a few modifications. Like for one I want to visually distinguish target kanji from similar kanji as an aid for memory. Some of this is naturally done in the RTK incremental approach.

Sorry, I don't have examples at hand to illustrate "kanji confusion." But some examples were offered by others in this forum. It should stand to reason that after mastering 2-4,000 kanji, you have a huge data base that needs to be maintained. Being acutely sensitive to graphic differences will be helpful I am sure.

Along the way of study of RTK I should also think that paying attention to graphic similarities of kanji will enhance not hinder your study. RTK kanji is not the whole cookie jar as others have pointed out in Forum discussions.

Again thank you all!

KANJI, I do understand better what you are trying to say now. It's true that when I said I usually didn't mix up kanji, I was thinking of the reading pace of kanji study, and not of the eventual goal of reading at full speed. I can 'read' the examples of JimmySeal without mixing them up, but it still takes the conscious process of first analyzing the radicals, and then putting the meaning to it, which takes at least 10 seconds per kanji at this stage for me.

On a related note, at the study level, I do see a difference between mixing up kanji with and without Heisig. My husband and I have both taken 3 years of Japanese classes, and are doing about equally well grammar wise, but I do Heisig, and he just follows Japanese for Busy People. When we are reading together, he cannot distinguish very well between different kanji, even if he studied them before, while to me they are very distinct and easy to remember in reading (of course, I'm still talking about study reading speed here, not normal reading speed). It does make me wonder whether doing Heisig early actually helps to minimize the mixing up problem of this thread.

Is there anyone in the advanced stages of reading who has done Heisig and has non-Heisig study mates who are equally advanced to compare themselves with? Are they better at not mixing kanji?

I think because RTK has a very analytical approach to all the elements of kanji (and not just traditional radicals) it teaches you to see kanji in a different way, perhaps, to other methods.

I originally learned to write kanji from "Let's Learn Kanji" (Mitamura & Mitamura) and this has a very analytical appoach too (plus an excellent section on the 8 types of stroke form or 一画), but I think Heisig goes further by getting the student to take kanji apart and put them together again having given each component a "personality" so that they almost become "friends" that you recognise immediately. I think this also gives you the ability to clearly isolate primitives you haven't learned yet, even if they are very like ones you already know. It's like seeing a person who looks very like someone you know, however close the resemblance, you just know it's not them.

Perhaps this works with other good methods, too, but I think the training in mnemonics that you get studying RTK definitely gives you an edge.

Anyone needing a topic for a PhD dissertation you got here, babe! RTK vs. traditional learning methods, ha, ha. Let me comment, hoping to consolidate previous postings to this thread. In my view, RTK is a stop gap measure in trying to control the multiple, multiple tasks required in acquiring Japanese literacy. With RTK, you nailed down the stroke order, stroke number, basic meaning, and more importantly, perceptual control of thousands of kanji--making them distinguishable. With that as a remarkable base, one can go on to the phonological aspects of kanji and their combinations in making words. Through much of this stage--cyphering individual kanji (rather than instantaneous conception)--is probably inevitable and probably not harmful.

But consider the difference between reading in Japanese and in English (my information here is offered tentatvely). In Euro languages, i.e. using a Roman alphabet, readers for the most part attain meaning through the phonology of the written symbols--letters. Japanese readers, on the other hand, do not; largely they bypass sound encoding. In other words, Japanese kanji is encoded without recourse to sound; apprehension of meaning is immediate. This is what I had heard from a American researcher in Japan.

I couldn't grasp the siginificance until I was well into Japanese and experimented with it. In English, I have always read through sound encoding of words. So I couldn't really understand reading without using phonology. In Japanese, I found that the times I tested myself, I did not go through phonology; mental imaging was immediate. Perhaps, this was just a mind game, a self deception. I await confimation from the related scientific literature.

However, what if reading as described here in the two languages is accurate? I wonder if that might factor into how we consider or should consider RTK. My interpretation is that RTK is a framework for the training of Japanese literacy. It is not literacy in itself; it is no more than our distinguishing between the letters of the English alphabet. Cyphering we do with kanji to conjure meaning is also a stop gap measure. After becoming a practiced reader, much will be instantaneously apprehened (however our overall reading speed may be). In other words, perfect mastery of RTK will matter little, for RTK is but an early stage in the process of acquiring Japanese literacy.

Afterall, many native Japanese will tell you that individual kanji do not have "meaning." Oh, surely, if you dig hard enough you can find meaning in anything. That is the defining characteristic that separates us from apes. But Japanese (for the most part) don't learn kanji through an "RTK" method. Many words rely on combinations of kanji; their combinations being the focus of study, not the individual units.

Although RTK is a very exciting learning tool, we ought to keep things in perspective. I sincerely hope that this post has illuminated rather than clouded our thread of discussion. And I also hope that my original postings in the Forum reap enough collective interest in hammering out the variable meanings or uses of Primitive Elements, such that we collaborate to chart out these variable uses by anchoring them to their physical positioning in kanji; that untypical uses of the Primitives be systematically renamed. I agree that some mental gymnastics can give order to these untypical uses; but I argue that that reguires more mental operations that may be the first to "rust" in periods of disuse. The mental gymnists do not need to participate, anyway. So I am looking for like-minded people who see the value of ironing out variable uses of the Primitives.

Cheers,
KANJI

I tried to look for some visually similar kanji pairs. I can't say whether they're commonly confused, but here's the best I found:

官/宮 (bureaucrat/Shinto shrine)
師/帥 (expert/commander)

Can anyone do better?

Not necessarily better, but I do think for instance 伐 and 代 look kind of similar at a glance. I mixed them up for a while, but that was mostly because the flow of the fiesta primitive is so nice I often put in the extra crossing stroke without thinking, even if it's the arrow or the quiver primitive instead.

How about 度 and 席 ? I get these mixed up for some reason.

This is fun. Here's another one:

替/賛

The neat thing about this game is that it shows how good Heisig's method is. The confusion I had when I was drilling brute force goes away entirely.

拐/招 whee

Stackz has a often confused kanji wordlist (to use with their software) with 402 kanji. It's in the middle of the page, but you can check it out here.

Some examples:

天犬太
下上
元売
力方九丸万
目日
白自
百首
小少
員貝買見

http://www.stackz.com/Stackz/stz.htm

Hope this helps.
J

It does help. Thank you.

Just ran into these on the study page:

侯/候