Nagareboshi Wrote:My favorite example is atsui. If i have to read a sentence without any further context, that contains atsui instead of 熱い, 暑い, 厚い, 篤い, i would not know the meaning of the word. The first two could have to do with objects and weather, while the latter two could refer to a person being warm (hearted), or cordial, or kind. With only the word in a random sentence, written in romaji, i would not know which is which.

You are very unlikely to encounter a "random sentence" with no context in the real world. Of course "Atsui desu ne" is ambiguous, but so is "I am hot" or "This is hot" in English.

Kanji are often of benefit to learners who don't have the native-level vocab and contextual knowledge of Japanese that are sometimes necessary to disambiguate the words (although "atsui" is a pretty simple one), but this isn't a problem for native speakers. It also becomes less of a problem the better one's Japanese gets.

Quote:So i oppose the idea of changing a writing system in favor of another, just to make it simpler for the rest of the world, to learn the language.

See that's my whole knock against it. I don't think it would be easier to learn the language. I think it would be a great deal harder. Japanese has mnemonic devices practically built into the language that help you learn. The very reason I've done so well with learning Japanese is because of kanji. For that to have not existed or be forced to go away feels criminal to me.

When I'm watching Japanese drama, for example, when I hear someone say something my brain converts a lot of what people say into kanji. When I think in Japanese, I think in kanji first because kanji hold the meaning of the Japanese language for me.

A lot of beginners talk about how they have trouble with reading or listening comprehension because their brain is trying to translate the foreign language into their native one for them to be able to understand it. A lot of people at an intermediate level start describing how they've finally been able to quiet the translator part of their brain and go right from foreign language -> intuitive understanding rather than foreign language -> english -> intuitive understanding. This is helpful because it shortens the amount of time it take for your brain to understand things, and that frees up a lot more of your brain to actually try to understand what's going on.

For me, that destination I labeled "intuitive understanding" is kanji. I have very real feelings and images tied to kanji, not even from RTK, but from where I see them used in native sources. I have a relationship with each and every single kanji, and they all have a kind of personality. For me, at my level, the kanji have ceased to be symbolic or pictographic, and may as well be full on photographs. When I look at a kanji like 本 or 母 or really any other, I do not see a symbol. I see a book. I actually see a mother.

It actually trips up some people around me. When my girlfriend came to visit me in Japan while I was on study abroad there I once gave her bad directions because I had forgotten that these symbols that I've spent time learning, that appear to have very obvious meaning for me, don't have an obvious meaning for her. There was a store I told her to turn by that had a big sign with 本 on it, and I just told her, "You'll be able to tell the place because it'll say book on it, I mean it's pretty self explanatory." Then later on she was like, "WTF? I don't know Japanese. I have no idea what you're talking about. It's all just a bunch of squiggles to me."

For someone to want to take away that relationship I have with kanji and the Japanese language, I find that extremely disconcerting. Like I have an emotional sort of reaction to it because it would be so damaging to how my brain operates when I think in Japanese.

For an idea of my level of Japanese. I've been studying and exposing myself to native sources for about 6 years. I've lived over there for a year. I've passed JLPT2, am studying for JLPT1. I listen to Japanese news podcasts every day from NHK, and I understand them to about 90% accuracy. I watch dramas without subtitles, and understand them just fine. I play tons of video games in Japanese without trouble.

It seems like anyone who's around or past the same level I am would have this same relationship that I have, but I don't know. Maybe I learn differently. I tend to read a lot more than other people, and I knew people who were fluent from a speech standpoint that were garbage at reading. For me, that's not what my relationship to language is. My relationship to language is based on text. That's how it is in English for me, and that's how it is in Japanese.

I can see how someone who approaches language from the direction of speech could think that Kanji are a hindrance. In that case the kanji are just there to obscure the true meaning, which is the syllables they represent. For me, and many other people, it's the other way around. That's kind of what you have to understand about my point of view, and why I'm kind of a jerk about this stuff.

To take away kanji would be to take away the very thing that allows me to think in Japanese, and I think there are a lot of other people like me. I understand that not everyone is like me, though, but I think you people that are anti-kanji need to understand the kind of benefits they do have for people who learn language like I do.

well, a lot of this discussion has focused on whether or not it's possible for Japanese to lose kanji... which, i'm not sure is a real issue to begin with.

The best reasons for losing kanji would be social issues to do with inclusion.

The best reasons for keeping kanji will probably be cultural, literary, and psychological ones.

So far, i haven't seen any concrete evidence of any deep social inclusion problems due to kanji, given the high educational standards in Japan. That's not to say that they don't exist, but it does seem like it's certainly not an issue like it was in Korea for example.

In that case, all the cultural, etc, reasons to keep it probably win.

Moving to something slightly related to the original topic, has anyone ever looked at Japanese sign language?

I did some BSL and in the alphabet I could kind of see how the signs resembled the letters but with the Japanese I find it quite hard.

http://www.h-hope.net/ud/column/column2/image/img1.gif

erlog Wrote:I understand that not everyone is like me, though, but I think you people that are anti-kanji need to understand the kind of benefits they do have for people who learn language like I do.

I think that the point of what yudan was saying is that it shouldn't be about people learning the language.

So yeah, I know what you mean about 'having a relationship' with kanji...but it's not about you or other people trying to learn it. It's about the people who's language it is.

pudding cat Wrote:Moving to something slightly related to the original topic, has anyone ever looked at Japanese sign language?

I did some BSL and in the alphabet I could kind of see how the signs resembled the letters but with the Japanese I find it quite hard.

http://www.h-hope.net/ud/column/column2/image/img1.gif

that chart has gotta be wrong!!!?? Some of those are impossible hand movements... try り。 Either that, or you're supposed to make the signs the opposite way to the pictures, which is highly counterintuitive...

erlog Wrote:

yudantaiteki Wrote:So i oppose the idea of changing a writing system in favor of another, just to make it simpler for the rest of the world, to learn the language.

I did not write this.

IceCream Wrote:

pudding cat Wrote:Moving to something slightly related to the original topic, has anyone ever looked at Japanese sign language?

I did some BSL and in the alphabet I could kind of see how the signs resembled the letters but with the Japanese I find it quite hard.

http://www.h-hope.net/ud/column/column2/image/img1.gif

that chart has gotta be wrong!!!?? Some of those are impossible hand movements... try り。 Either that, or you're supposed to make the signs the opposite way to the pictures, which is highly counterintuitive...

I think you're meant to mirror the action on the page. So for り you make a peace sign and then flip your hand over and downwards. Maybe. Ignore the 'twist your hand and mangle your arm' implications in that picture

yudantaiteki Wrote:There are also many audiobooks available for purchase (and have been since the 80s):
http://www.febe.jp/
The JP wikipedia article on audiobooks does not mention kanji at all and the audiobooks are just straight readings of the books like in English.

Also the news subtitles are sometimes used as support for the idea that Japanese people need kanji to understand, but the news is also broadcast on the radio (and via podcast) and there's no kanji there....

For the most part you can understand things with context, but often when you hear an ambiguous homophone the easiest thing to do is to think of the kanji, haven't you ever had a native say this to you or experienced it yourself?

You can get by without kanji, just as you can get by with just pinyin, but I love the kanji, they're my favourite aspect of the Japanese language, and I would vigorously oppose any attempt to eliminate them.

I personally don't visualize kanji when I'm listening or speaking. But my position on kanji is not really based on my personal ability. I find kanji annoying in that I've been studying Japanese for 12 years and I still have constant issues with kanji, but since I'm working with Heian to Edo-period Japanese, changing the script now would have no effect on my personal needs so it has nothing to do with my position.

yudantaiteki Wrote:

Nagareboshi Wrote:My favorite example is atsui. If i have to read a sentence without any further context, that contains atsui instead of 熱い, 暑い, 厚い, 篤い, i would not know the meaning of the word. The first two could have to do with objects and weather, while the latter two could refer to a person being warm (hearted), or cordial, or kind. With only the word in a random sentence, written in romaji, i would not know which is which.

You are very unlikely to encounter a "random sentence" with no context in the real world. Of course "Atsui desu ne" is ambiguous, but so is "I am hot" or "This is hot" in English.

Kanji are often of benefit to learners who don't have the native-level vocab and contextual knowledge of Japanese that are sometimes necessary to disambiguate the words (although "atsui" is a pretty simple one), but this isn't a problem for native speakers. It also becomes less of a problem the better one's Japanese gets.

Yes this is a very simple example. Maybe i choose this for an example, because it was the first word i learned, that had two different kanji in writing, used in a different context. Let me give you a more concrete example with atsui desu. Say you read the sentence basu no naka ha atsui desu. What would you think it means? This sentence could stand perfectly well on it's own, i guess, and it could mean: "Something is in the bus that is hot (object)," or "It is hot inside the bus." Would a Japanese person be able to tell without kanji? When the sentence would be バスの中は暑いです。 It becomes clear, that it is hot inside the bus, no more is needed.

erlog Wrote:

Nagareboshi Wrote:So i oppose the idea of changing a writing system in favor of another, just to make it simpler for the rest of the world, to learn the language.

See that's my whole knock against it. I don't think it would be easier to learn the language. I think it would be a great deal harder. Japanese has mnemonic devices practically built into the language that help you learn.

It was a mistake. The sentence should have read: "making it easier to learn to read the language," instead of "making it easier to learn the language." Because that is not what i think. I think about the kanji almost the same way as you do. Kanji have been an integral part for me, even when i had no clue how to write them, except with the IME, i saw them as an integral part of learning the language. And i couldn't imagine learning from a book that's containing romaji instead of kanji ...

So, if anyone would say that learning Kanji is stupid, because there is Romaji - well, i would tell them that this is ignorant. Same way as telling someone, don't learn the ABC, stick to {input non roman language here}. It would never be the same.

yudantaiteki Wrote:

erlog Wrote:

yudantaiteki Wrote:So i oppose the idea of changing a writing system in favor of another, just to make it simpler for the rest of the world, to learn the language.

I did not write this.

There was a quote pyramid, and I got confused. I took your name off it. Sorry about the confusion.

Asriel Wrote:

erlog Wrote:I understand that not everyone is like me, though, but I think you people that are anti-kanji need to understand the kind of benefits they do have for people who learn language like I do.

I think that the point of what yudan was saying is that it shouldn't be about people learning the language.

So yeah, I know what you mean about 'having a relationship' with kanji...but it's not about you or other people trying to learn it. It's about the people who's language it is.

You are assuming that the Japanese people don't have the same relationship with the kanji that I do when I think there's a lot of evidence that quite a lot of them do have a similar relationship. The popularity of kanji-based comedy, puns, calligraphy, and poetry points to kanji being an important cultural staple of Japan.

Nagareboshi Wrote:Yes this is a very simple example. Maybe i choose this for an example, because it was the first word i learned, that had two different kanji in writing, used in a different context. Let me give you a more concrete example with atsui desu. Say you read the sentence basu no naka ha atsui desu. What would you think it means? This sentence could stand perfectly well on it's own, i guess, and it could mean: "Something is in the bus that is hot (object)," or "It is hot inside the bus." Would a Japanese person be able to tell without kanji? When the sentence would be バスの中は暑いです。 It becomes clear, that it is hot inside the bus, no more is needed.

I'm not sure the sentence can have the first meaning, but even if it can, real conversations and texts don't have contextless sentences like that. It's very hard to look at an actual Japanese text and find examples where a native Japanese person would not know which of several homophones was meant, assuming the words are not obscure or technical jargon that the person doesn't know. Like I attempted to say earlier, if kanji were really necessary for Japanese people to understand Japanese texts, audiobooks would be impossible.

暑い and 熱い are pronounced identically in speech, so if a native Japanese speaker cannot understand the difference between them in romaji, they also cannot use them in speech without confusion. This obviously is not the case.

erlog Wrote:You are assuming that the Japanese people don't have the same relationship with the kanji that I do when I think there's a lot of evidence that quite a lot of them do have a similar relationship. The popularity of kanji-based comedy, puns, calligraphy, and poetry points to kanji being an important cultural staple of Japan.

I agree, I do think a lot of Japanese people think of it in this way. However, I don't think comedy and puns are necessarily a good reason to keep it around. Historical things would obviously remain the way they are. Sure there are cultural things keeping it around, but in my opinion, if that's the only thing holding it back, then I do think that it should be replaced. Studying kanji from elementary through high school so you can pick up on a few gags, and maybe read some old poetry?

I think the writing system should stay as it is. I don't have any real justification for it other than "it's worked so far, so why change it?" But...I also don't see any actual significant factors keeping it from being changed.

The cultural element has some validity as an argument for why kanji should be retained (I tend to avoid these more subjective views on whether they *should* abandon kanji or not). However, it is worth pointing out that just because something is cultural doesn't necessarily mean it should be retained. Even if we just look at Japan, it used to be a very important cultural element that men were able to read and write classical Chinese. Texts from the pre-Edo period especially are often written either in pure classical Chinese (or hentai-kanbun) or in a mix of Japanese and classical Chinese. Since virtually no one in Japan today can read or write it, this textual element has disappeared.

You also used to have a much greater freedom in using hentai-gana (this persisted in some form all the way up to WW2), so that writers could choose from a number of different kana for each syllable. This has also virtually disappeared. (You could argue this isn't the same as multiple kanji, but I tend to mostly disagree with the "shades of meaning" claim that's made for kanji since for the most part they're fixed by convention, and writers only occasionally depart from convention to make a pun or suggest some additional meaning.)

Not to mention that the modern usage of kanji/hiragana/katakana is relatively recent and only came into near-universal use after WW2. Even some of the Genji books I use that were published in the 1950's still have introductions written entirely in kanji and katakana.

With respect to puns, the changeover to modern kana usage has eliminated some of the older puns that were possible (for instance, 葵（あおい） with 会う日, since both were once written "afuhi"). You also find puns in (much) older writing that depends on the fact that dakuten were often not used in writing, so that there are puns between words that differ in unvoiced and voiced syllables that you cannot do anymore.

So there have already been a number of significant changes to the writing system that can be said to have discarded things that were once culturally important, and also arguably reduced the expressive potential of Japanese writing. (And this is only in the writing system -- no civilization's culture is static; things that were important are constantly being discarded and replaced by other things.)

I don't claim to be a linguistics major but as a person learning the language I find Kanji to be a useful tool (once you can recognize them and pronounce them) for quickly going through large bodies of text. It makes things a lot faster - especially when I compare it to those hard times in Japanese class in high school where everything was Kana... I didn't understand a word I was saying when I read aloud and I had to read quite slowly.

If they were going to put all of Japanese into Romaji, they'd have to introduce putting spaces into everything and plus, wouldn't it get confusing with all the homophones? I mean yeah, context goes very far in helping that but I imagine that it would come up a lot more often that people can't distinguish which of 4 homophones a word is.

I was just wondering if yudataiteki could elaborate on why he thinks romaji (or kana) should only be used.

erlog Wrote:It seems like anyone who's around or past the same level I am would have this same relationship that I have, but I don't know.

Nope, I think people's mental representations of language are surprisingly varied. I read Japanese the same way I do English -- unless I'm reading so fast that I'm nearly skimming then it all converts into internal mental voice. (Except when I don't know a reading, in which case I get nani-nani-suru or mmm-suru :-)) I have very little 'visual' memory (after all, where would you put visual recollections? My field of view is full of the stuff I'm actually looking at...)

That doesn't mean I ignore the hints-to-meaning kanji provide in written text, though.

Incidentally, I agree that going directly from what you hear to intuitive understanding is a key step in language learning, but I'm inclined to think that if you really go from what you hear to kanji to understanding then there's still an extra step lurking in there. I suspect you actually don't do that and it's a simultaneous-activation thing of the kind nestor loves to bang on about.

thecite Wrote:For the most part you can understand things with context, but often when you hear an ambiguous homophone the easiest thing to do is to think of the kanji, haven't you ever had a native say this to you or experienced it yourself?

I assume most natives wouldn't have the "kanji visualization" problem because they learnt the language before they learnt how to write, unlike us.

The support I've mentioned for the benefits of kanji does not rely on native vs. learner. It's objectively beneficial, thanks to things like the depth-of-processing effect and the way humans become literate.

Kanji are easy to learn with methods that rely on visual chunking and spaced retrieval which makes usual Japanese use of muscle memory even more efficient. This is not a native vs. learner thing, it's a human thing. Their complexity also makes them superior for elaborative encoding (see previous mention of depth-of-processing) and iconicity, cohering the constituent parts into wholes during reading.

By a wide margin, most literate peoples' new words are learned via text, not speech. Even the ones that are acquired early are beneficially impacted by orthography not only through speech segmentation, the development of moraic awareness from syllabic, but through adding the textual medium for contiguity and/or reinforcement in the multimodal learning process of lexical items, another human thing, and where kana is not used and (increasingly larger as pedagogy evolves, kanji learning can begin quite early, studies show) kanji is used, even better.

Humans don't live in a society of primary orality, so having to rely purely on sound and sounding out words in text is, well, primitive. Especially as language is a multimodal, emergent phenomenon, one needs to address the general functions of the brain and the most effective environmental variables for learning. Being able to invest meaning visuospatially rather than relying on speech is superior and opens up the nuances of literacy beyond the phonocentric domain. The focus ought to be on picking up words encountered in text not due to prior speech knowledge of words but due to the properties of the text and orthographic knowledge that aid parsing and learning new words on multiple sensory levels beyond sound. Additionally, as one develops their language skills, this becomes less of a matter of one's spoken lexical inventory lagging behind (despite being able to use kana in materials as an extra bridge) and more a matter of simply taking advantage of a feature rich orthography for learning vocabulary, especially when it's naturally in order to use the primary feature of writing, which is to communicate information.

It's also easier to write/spell kanji words than kana or letters alone because of the use of less characters in a linear sequence and also the aforementioned properties of the icons, and/or easier to type depending on which modality one is more familiar with, with the IME increasing usage, where letters typed operate on a lower kinetic level and immediate visual feedback is given for the smaller, visually richer clusters of charaters, which is also more suited to the way the eye works for quickly scanning text (and taking advantage of stronger visual-semantic associations), I believe.

I guess if you disagree with me, which would make no sense, we'll have to compare literacy development across these societies. Where are those accurate figures, again?

English has too large a syllabic inventory, by the way, it ought to be reduced. There will be more homophones, but this can be offset by logographic integration and contextual awareness. Of course, with logographs mixed in, there won't be a need for spaces, so that'll be nice.

nest0r Wrote:easier to type depending on which modality one is more familiar with, with the IME increasing usage, where letters typed operate on a lower kinetic level and immediate visual feedback is given for the smaller, visually richer clusters of charaters,

Ignoring the rest, I think this is flat wrong. My experience is that what makes things fast to type is when what you're typing has a straight one-to-one relationship with what you're trying to make appear on the screen. This means you can rely on muscle memory and effectively "type ahead" of the computer. In particular, errors are usually accidentally hitting the "wrong" key and the feedback loop is very very short because you can feel that you've done that and start hammering on the backspace key. One of the reasons that dictating using voice recognition software is a frustrating process is that the feedback loop for errors is long: "say something; wait half a second for the computer to process it; read what the computer has entered, detect that it's wrong, start correction process", so you are forced to break up your input into artificially short sections, and the input process itself intrudes into getting your thoughts onto the screen. The traditional Japanese "cycle through the options" IME has some similar properties in that it's guessing what you meant, and you have a feedback process for checking and correcting it that, while not as long as that for speech recognition, is still not as fast as for plain keyboard use, and still intrudes somewhat on the flow of thought from your brain into the computer via the keyboard. (I'm assuming you're not using a t-code style input method here ;-))

I disagree, I think having to type one-to-one compounds the problem of having to spell out phonetic sequences of letters and relies too much on pre-typing assemblages of letters. It also increases the number of ‘fail points’ for both typers and spelling correction. It's much better to have an ever-evolving IME system and a complementarily ‘endo-’ contextual orthography. ;p

Voice dictation software probably works better for Japanese than English, with the differential aspects of the Japanese phonological properties vs. orthographic for machine recognition. Easier to analyze the input and determine the proper output.

nest0r Wrote:It's also easier to write/spell kanji words than kana or letters alone

This is a testable hypothesis, incidentally. You could take people who are fluent first-language readers/writers of, let's say, Japanese, Chinese, Korean, English and some roman-alphabet non-irregular spellings language, all educated to the same level (eg "college degree"). Ask them to write a set of words, and see how many errors they make and whether they're correlated to language. Anything like that in your big pile of papers? (You could also try some "complete-the-sentence" style tests to try to check the 'brain-to-writing' rather than the 'heard-speech-to-writing' angle.)

nest0r Wrote:I disagree, I think having to type one-to-one compounds the problem of having to spell out phonetic sequences of letters and relies too much on pre-typing assemblages of letters. It also increases the number of ‘fail points’ for both typers and spelling correction. It's much better to have an ever-evolving IME system and a complementarily ‘endo-’ contextual orthography. ;p

Well, obviously it might be nicer if what you were typing out one to one was actually something that bore closer resemblance to how the word actually sounded, but I think English spelling reform has about as much chance of success as Japanese writing reform that throws out the kanji. But you have to type something, and anything where the output produced is the result of the computer making a guess rather than being totally deterministic is just a recipe for annoyance.

Quote:The support I've mentioned for the benefits of kanji does not rely on native vs. learner. It's objectively beneficial, thanks to things like the depth-of-processing effect and the way humans become literate.

I don't see any reason to believe this. Most cultures don't use kanji or logographs any more, and I don't see any signs of suffering.

I agree that there are cultural reasons to use the kanji, and I'm not arguing in favor of dropping them. That's for the Japanese to decide. On the other hand, both Japanese and Chinese can be read using Roman letters, as evidenced by the fact that it's done a lot. Do I think using letters would be more efficient for learning and reading? Well, yes, in the end.

It took me five months or so to learn the Monbusho kanji (including onyomi), as opposed to a few days to learn Cyrillic, for example, so saying kanji are efficient to learn makes me cringe.

bertoni Wrote:

Quote:The support I've mentioned for the benefits of kanji does not rely on native vs. learner. It's objectively beneficial, thanks to things like the depth-of-processing effect and the way humans become literate.

I don't see any reason to believe this. Most cultures don't use kanji or logographs any more, and I don't see any signs of suffering.

I agree that there are cultural reasons to use the kanji, and I'm not arguing in favor of dropping them. That's for the Japanese to decide. On the other hand, both Japanese and Chinese can be read using Roman letters, as evidenced by the fact that it's done a lot. Do I think using letters would be more efficient for learning and reading? Well, yes, in the end.

It took me five months or so to learn the Monbusho kanji (including onyomi), as opposed to a few days to learn Cyrillic, for example, so saying kanji are efficient to learn makes me cringe.

Despite attempts to paint it as the product of superior design, the development of the alphabet was incidental to the properties it is now lauded for, as I understand it from reading scholars on the development of writing systems. I think in that sense it was a great thing, but calls one to keep in mind that it's not objectively superior by any means. As anime characters would say, kanji is kanji, letters are letters. ;p

Just because you were able to learn a small set of letters doesn't mean you're able to acquire the vocabulary. You can learn a small set of kana and then progress from there into kanji, which is better in the long term for literacy than just having the one set.

@pm215 - I'm pretty sure I've read studies that say spelling is easier in Japanese, but then there are other difficulties in writing Japanese, obviously. I just happen to think that mitigations of these errors, such as balancing determinism and automaticity in typing, leans towards mixed logographic/phonographic as being ideal. ;p

I was mostly being trollish, since in my mind these conversations are sort of nonsense, starting from biased hypotheses from the perspective inside one system and then promoting it over imaginary cross-linguistic comparisons. Trollish in copying others and pretending my views can patently cut across culture and language in comparative fashion. Not because I don't believe the fundamentals are objectively true across the board. ;p

I also don't really believe any voice dictation/speech recognition will be objectively easier for any natural language. Well, I do think they'll evolve to support the ideal mix of determinism/automaticity/multimodal mixed phonograph/logograph I mentioned before.