As long as we're throwing out theories about romaji.........

Scholars and teachers give us very logical reasons why we should avoid it all all costs. But they developed these reasons to explain the historically poor literacy rates in students. Japanese takes 500 to 1000 hrs to get proficient with the spoken language, and 2000 more hours to to get proficient at the written language. Some people are willing to spend 500 to 1000 hrs, some are willing to spend 2500 to 3000 hrs. You can't trick people into adding an extra 2000 hrs of study by making them avoid romaji; they'll probably just quit.

There's nothing wrong with using only romaji until you become proficient in the spoken language, then switching over to Japanese. If I could do it all over again, that's what I'd do. It's certainly not for everyone, but many have learned this way. In fact, every time I hear of a westerner who has learned Japanese really fast, it turns out they used this method (I don't count Mr. 10,000 sentences).

Theory:
The reason why people fail to become literate is not because they use romaji, it's because they lack motivation. Romaji success stories have proven this.

Corollary:
The reason people fail to master kanji by studying them out of context is not for all the reasons the scholars and teachers mention, but because they were not using a method that works. Heisig sucess stories have proven this.

Top That!:lol:

Everybody is making good points, with which I do not disagree in any major way. There is a perspective, though, that may be being overlooked. Maybe. If you want to learn to speak a language, as opposed to learn to read and write ---those skills are necessary if you're going to be functional in the society--but if you're learning to speak and hear the language, then you're more focused on pronunciation. In any of that context, you use a notation system that as effortlessly as possible DOES convey to you the sounds. Romanization has it's issues, in that to an English speaker a character may suggest a sound that a Japanese speaker wouldn't use. OK, International Phonetic Alphabet (IPA) would be more accurate, but most english speakers don't know IPA. In a similar vein, they don't know カナ. So, my little bitty point is, you can teach japanese pronunciation, and use the roman alphabet, or IPA or カナ. But if you use the later two, you're teaching 2 things. IMO, our english alphabet is no worse at notating the pronunciation than カナ. You just have to be aware of the language you're pronouncing as you read it.

Learning カナ is fun! Just like learning 漢字！　And it is a stepping stone to functional literacy in the culture. Yes. And depending on it for too long is a crutch that isn't helpful. But you can be fairly fluent in spoken language without being comfortable with カナ. You're just illiterate.

laxxy Wrote:I don't know what's ISAA, but I understand what you are talking about oh boy they do swear. This system is like a club tie for them. Well, for each his own I guess. Luckily we don't have many romaji supporters of the same belligerence

ISAA = Institut Stran Azii i Afriki (this forum won't let me use Cyrillic or Kanji, for that matter...)

astridtops Wrote:It's very much an 'amateur' course at what we locally call a 'folk university', 25 x 75 minutes, so the level isn't terribly high, as people from all kinds of backgrounds are allowed to participate (though most of them have some form of a higher education).
Also, it's fun, since I'm doing the course with my husband and one of my best friends. So, it's also something of a social activity .

[kana]naruhodo[/kana]... That's understandable. And also confirms my point that most Dutch study groups are more of a 'gezelligheidscluppie' than a serious enterprise...

astridtops Wrote:Well, since I'm mainly learning Japanese for listening to anime and reading japanese manga and novels, my aims are not quite the same as the average student here.

There is quite a gap between manga and novels... Your interests seem really diverse!

wrightak Wrote:The most frustrating thing for me when I was an English teacher, was not the pronunciation difficulties that came with the katakanisation of English, but that I could SEE students thinking about English in terms of katakana.

Very well said. This is exactly the reason why many Japanese never learn proper English pronunciation.

CharleyGarrett Wrote:Romanization has it's issues, in that to an English speaker a character may suggest a sound that a Japanese speaker wouldn't use.

Good point. I keep forgetting that Japanese is a second language, after english, for lots of students. I keep thinking most Americans know Spanish pretty well, and romaji is a no-brainer. Unfortunately, this isn't the case. I've neglected other native english speakers too; sorry 'bout that.

But one has to learn romaji anyhow (to type) right? Is it easier for one to learn romaji or kana? If romaji is easier, it makes some sense to use it for learning grammar.

This is analagous to the reason english keywords are used in RTK1. One could learn a Japanese keyword, but that makes the process more difficult. That's why I choose to use the english keywords and add Japanese meaning by studying vocabulary.

Is it quicker, or better, to (1) use kana only to learn grammar or (2) use romaji to learn grammar first, then learn kana? I assume one would be at the same point in their studies after finishing either of these options. Both ways are probably fine. Romaji doesn't damage a person's reading; lack of motivation does.

The only benefit I notice myself from using kana over romaji is that it forced me to drill the kana into my head very thoroughly. Otherwise, it's still a lot slower to use kana instead of romaji for me, and it has no added benefit for anything else. Therefore my answer to Leo's question would be that romaji are faster for learning grammar, unless you need the kana practice, in which case you might well combine the two.

Reading and writing kana will always be more awkward than roman letters. I know my kana, but I still have to spell them instead of them coming together fluently in my head like roman letters, so studying grammar in kana takes up 5x the amount of time it would do in romaji. I don't think I can ever overcome the speed gap when reading Japanese. I'm resigned that it will remain a slow affair forever.

Anyway, at the point where I do not need kana practice (because I know them when quizzed, and extra practice does unfortunately do nothing for speed), I see no longer the point in reading and writing in kana only. That's why I am now switching to writing in kanji (+ necessity kana only) all the time. At least then I get some meaning benefit out of my slower speed, and some good practice in writing my kanji.

astridtops Wrote:I see no longer the point in reading and writing in kana only. That's why I am now switching to writing in kanji (+ necessity kana only) all the time.

Most information in a sentence comes from kanji - once you know kanji, reading kana-only phrases in a study environment becomes rather challenging and unnatural.

In fact - for real-life situations - there's been a study which showed that when pressed for time (like reading subtitles, etc.), native speakers scan the kanji in the sentence, whereas foreigners prefer to read the kana bits - and usually miss all the info.

Most people seem to agree that using romaji is fine in certain situations but it's best to use kana if you can. I think that romaji acceptable only if you are dealing with Japanese for a short time. If you do anything vaguely serious with Japanese (a month of study or more) then I think using romaji is pretty crazy.

It surprises me that anyone on this forum would say that kana is hard to learn. How long did it take people to learn the first 100 kanji in RTK 1? A week? Maybe two? There are about 50 hiragana and that's all you need to get going. Get Heisig's book and do it in a weekend or just get some flash cards and start drilling. Whatever method you use, it won't take long if you concentrate on it and do some practice every day.

Also, if you think that using romaji is fine then think about what that must imply. If you were teaching a Japanese person the sentence, 'My name is Hideki and I live in Tokyo', would you be happy for them to write it down as

マイ　ネーム　イズ　ヒデキ　アンド　アイ　リッブ　イン　トウキョウ

OK, so Japanese people already know the alphabet, but even if they didn't, I don't think writing down the above would be advised for any student. It represents English badly, it's inelegant and totally the wrong approach. I think that romaji is no different in its representation of Japanese.

The two main arguments for using romaji seem to be that it's quicker to read and kana is hard to learn. I don't think kana is hard to learn and if you think that you read Japanese quickly with romaji, wait until you get really familiar with kana and kanji and you'll triple your speed.

astridtops Wrote:Otherwise, it's still a lot slower to use kana instead of romaji for me,

Reading and writing kana will always be more awkward than roman letters. I know my kana, but I still have to spell them instead of them coming together fluently in my head like roman letters, so studying grammar in kana takes up 5x the amount of time it would do in romaji. I don't think I can ever overcome the speed gap when reading Japanese. I'm resigned that it will remain a slow affair forever.

Based on my experience and that of many people I know, I think you will find that the situation will improve. Give it time. The reason I raised this topic is because I was reading romaji in posts and it was taking me time to figure out what was being written. I read a lot more quickly with kana and kanji and I'm certain that the same will apply to you. Please stick with the kana and keep practising!

Raichu Wrote:The point I was making was that IMNSHO as a writing system, it's ludicrous, arguably the most difficult in the world. Romaji in spite of its imperfections would be a more practical alternative--or any writing system that is capable of representing their spoken sounds.

It is pretty crazy isn't it. The thing is that they already have a phonetic alphabet, the kana. So your point would surely be that they should use kana all the time and give up with the kanji. In a very similar way to how the Koreans now use their phonetic alphabet and don't use kanji at all any more. Kids in Japan certainly do this. Unfortunately it gets difficult to read. Sentences like　ははははながすき come to mind.

I understand your comments that phonetic alphabet systems are better than the kanji and kana combination of Japanese but I don't see why it would follow that romaji should be used. When you right 'ka', you use two characters and immediately indicate that there are two sounds there. In Japanese, this is just one sound not two.

Hi Wrightak,
I understand your opinion, and agree that learning and using kana instead of romaji whenever possible is a legitimate option for learning. I also agree that most teachers and students advocate this method. I am using this method myself, and it's working for me.

But there is another option. Quickly learn grammar first with romaji, then turn around and start learning reading/writing. Learners will accomplish the same thing, through different paths. The path that is best for some will not work as well for others; people learn differently. I don't think we can say that one method will always lead to a better result.

I would expect to see decent support for both of these methods. Maybe the pressure to get people to avoid romaji in the beginning has influenced some people to take the harder path, which can be a contributor to quitting. Unfortunate. So, I'll keep saying romaji doesn't damage a person's reading; lack of motivation does.

The thing I find best about studying my Japanese vocab in kana is that you generally only have half the characters to remember. Also it makes reading it out loud so much easier and it 'sounds' more japanese as well.

I studied Japanese in romaji for about a year before I finished learning kana. It probably was another three months before I was fairly comfortable with them, and maybe a year past that before I started trying to avoid romaji.

I don't feel like romaji has adversely affected my studies in the long run. When I read kana, I see sounds. I don't see English letters superimposed over them, and I think my Japanese reading speed is pretty high for a non-native speaker. I don't feel like I had to "un-learn" the romaji at any point either. Rather, I think that if I had to learn all the kana before I started learning vocabulary and grammar, I would have probably given up at the beginning. In fact, this is the very reason why I gave up learning Greek, Russian, Hebrew, and Tibetan before I got anywhere at all.

I think it's rather fascist to expect everyone on this board to avoid romaji. It should be a matter of preference, as with anything involving BBCode. I personally think Hepburn is a lot nicer looking than kunrei, but that's just my opinion.

To those who don't think romaji accurately represent Japanese sounds: you've got to stop thinking in English sounds when you see romaji. The D in the word "juventud" is pronounced like a /th/ and the V is pronounced like a /b/. The X in "xiao" is pronounced like an /s/, and the Z in "zai jian" is pronounced /dz/. Roman letters do not only represent English sounds, and once you break yourself from that mindset, you will probably stop seeing romaji as such an evil device.

wrightak Wrote:I have had another thought though. It generally appears that people who are less concerned about romaji are those whose native language isn't English. It appears that English speakers get particularly upset about romaji even though there are plenty of other languages that use the roman alphabet. A Swiss friend of mine once said that she preferred French to English because when she read a French word, she knew how to pronounce it immediately. When she saw an English word, she was often uncertain. I may be well of the mark, and this is just a random thought, but do you think that English speakers develop a different mental relationship with the alphabet than French people do because there are no easy rules for pronouncing words?

I think what really happens is that Japanese is rarely a second language for a native speaker of a language other than English. Certainly not for anyone here, as this site itself is in English. So by the time one gets to Japanese, any impulse to read 'ri', etc. exactly like in his/her native (or any other) language is probably long gone (and if his native script wasn't Latin, it might not have been that strong to begin with). So romaji is less of an issue.
I have also observed English speakers complaining about Chinese pinyin script for the same reason, saying that letters like c,q or x denote something completely different from what they stand for in English, and I don't have any issues like that. For me they are just letters, and the sounds they represent can be anything depending on the language, it does not bother me at all.
On the other hand, I do have all these problems when I look at transliterations in my native language (Russian). Even though I know how the Japanese words should be properly pronounced, the mental connection between letters and sounds is too strong and quite distracting.

leosmith Wrote:Quickly learn grammar first with romaji, then turn around and start learning reading/writing. Learners will accomplish the same thing, through different paths. The path that is best for some will not work as well for others; people learn differently. I don't think we can say that one method will always lead to a better result.

The option to 'quickly learn grammar' hardly applies to Japanese. Besides, books only seem to teach general rules with few examples - it takes a great deal of reading before one properly learns to recognise grammar forms and patterns in real life usage. And the reading cannot be done in romaji.

Serge Wrote:The option to 'quickly learn grammar' hardly applies to Japanese. Besides, books only seem to teach general rules with few examples - it takes a great deal of reading before one properly learns to recognise grammar forms and patterns in real life usage. And the reading cannot be done in romaji.

I think I understand your point. What I meant was to do the initial learning quickly, for example get through "Japanese the Spoken Language" in less than 6 months. For me, the grammar wouldn't sink in until a lot later. But I meant after the 6 months, when your down to flashcards and audio, switch over to kana/kanji. Familiarize yourself with the written language while the grammar is still sinking in.

I think you can learn grammar through speaking/listening. Probably not the most efficient way for most people, but possible.

JimmySeal Wrote:I think it's rather fascist to expect everyone on this board to avoid romaji. It should be a matter of preference, as with anything involving BBCode. I personally think Hepburn is a lot nicer looking than kunrei, but that's just my opinion.

I think that I've put forward some of the strongest arguments for using kana instead of romaji so I assume that the above was directed towards me. It's the first time anyone has ever used the word fascist in relation to something I've said, especially after being such a leftist university student for so long! First time for everything I suppose...

As I said in some of the first posts in this thread, I would never expect people to avoid romaji. Allowing people to contribute to this forum is the top priority and if the choice is between a contribution with romaji or no contribution at all, I'd go for the former every time.

After learning kana, have you not found that your reading speed has improved drastically over when you read romaji all the time? If so, would it not have been nicer to have had that benefit earlier?

I think your comments about 'juventud' don't relate as well because that word is written with those characters normally anyway. Romanisation of Chinese is an interesting topic though. I don't know much about it but I think a few other members of the forum are heavily involved in Chinese.

laxxy Wrote:On the other hand, I do have all these problems when I look at transliterations in my native language (Russian). Even though I know how the Japanese words should be properly pronounced, the mental connection between letters and sounds is too strong and quite distracting.

I found your post really interesting laxxy, especially the comment above. I think your points about Japanese being a second language for many English speakers is a good one too. It's not my second language and I'm used to reading roman characters in different languages and yet I still find reading Japanese in romaji difficult.

leosmith Wrote:I think I understand your point. What I meant was to do the initial learning quickly, for example get through "Japanese the Spoken Language" in less than 6 months. For me, the grammar wouldn't sink in until a lot later. But I meant after the 6 months, when your down to flashcards and audio, switch over to kana/kanji.

I'm finding it a bit hard to sympathise with your point Leo because I can't see how you can promote learning grammar quickly before learning the kana. Learning the kana can take you a weekend, what can you do with grammar in a weekend?

I think that in order to make good progress in Japanese, it's important to advance all four skills: reading, writing, listening and speaking. So if you're learning the kana at some stage, learn grammar too. Don't divide it all up. Advancing one of these four will help with the others. Personally, my listening and speaking and reading streaked way ahead of my writing. About a year ago, I put more effort into writing and it has benefited the other three a lot.

wrightak Wrote:After learning kana, have you not found that your reading speed has improved drastically over when you read romaji all the time?

Well, I couldn't read at all before I learned kana, so yes, my reading speed improved dramatically after I learned it. Be that as it may, I was still illiterate because I had 2000 kanji to learn, so what was the huge benefit I got from learning kana?

wrightak Wrote:If so, would it not have been nicer to have had that benefit earlier?

No. I already said that if I had thought I had to learn kana before I learned anything else, I probably would have given up right at the beginning.

wrightak Wrote:I think your comments about 'juventud' don't relate as well because that word is written with those characters normally anyway. Romanisation of Chinese is an interesting topic though. I don't know much about it but I think a few other members of the forum are heavily involved in Chinese.

They're both (Spanish and Chinese) perfectly valid arguments. If someone studying Japanese hears the first syllable of "really" when they see "ri," and the first syllable of 林檎 when they see り, then the problem is with their mindset, not with the character set, because they are two representations of the exact same thing. People studying Spanish have to learn that J is not pronounced like an English J, and romaji is the same idea.

wrightak Wrote:I'm finding it a bit hard to sympathise with your point Leo because I can't see how you can promote learning grammar quickly before learning the kana. Learning the kana can take you a weekend, what can you do with grammar in a weekend?

I know people who can speak fairly decent Japanese and can't read a single kana character. Learning kana offers no benefit to a beginner learning Japanese, and someone can learn kana in a week, but it takes far longer than a week to be comfortable with them. Being able to communicate in Japanese requires knowledge of vocabulary and grammar, not kana.
One could go so far as to argue that kanji are more useful to a beginner than kana, although I'm not going to attempt that claim. But someone who knows kanji and no kana can probably figure out what 止まれ means, but someone who knows kana and not kanji would only know that it's a word ending in "mare."

Jimmy, would you agree that someone learning Japanese using romaji is the same as a Japanese person learning English using katakana? Referring to one of my earlier posts.

wrightak Wrote:Jimmy, would you agree that someone learning Japanese using romaji is the same as a Japanese person learning English using katakana? Referring to one of my earlier posts.

No, I wouldn't. Both romaji systems are able to distinctly represent all Japanese sounds. "sa, ri, and ku" can only be pronounced one way, and the same goes for any other romaji. One could raise the small issue of the particles は and へ being romanized as "ha" and "he," but for such reasons, I think it's better to romanize them as "wa" and "e," though it's only a small exception and can be easily dealt with in either case.

Representing English with katakana, on the other hand, is highly imprecise. If a Japanese person has the words "razor" and "lazer" both stored as レーザー in their head, then they won't know whether to pronounce them with R or with L. In addition there's the issue of extraneous vowels. My name, James, has two vowel sounds in it, but when it's written out in katakana as ジェームズ, it takes on two extra /u/s. Lastly, there's the issue that katakana can only represent five vowels while the English language uses about 12 different vowels. There are ways of using kana to halfway accurately represent English pronunciation, but no Japanese ESL student would go to the trouble of learning that system.

There's a one-to-one correspondence between romaji and Japanese pronunciation. On the matter of the romaji <-> kana correspondence, the Hepburn system ignores ず/づ　じ/ぢ and sometimes お/を, but learning those exceptions is far less effort than Japanese students have to put into learning English spelling.
Nobody has yet laid out a way to accurately represent English pronunciation using kana, and even if they did, no one would bother to learn it, so it causes ambiguity and bad pronunciation, and therefore is a completely separate issue.

JimmySeal Wrote:But someone who knows kanji and no kana can probably figure out what 止まれ means, but someone who knows kana and not kanji would only know that it's a word ending in "mare."

Of course as I was reading this I read the last word as mare - female horse, and not as ma-re in the japanese sense. But then when I saw 止まれ I read it in my head correctly.

I guess what this showed to me was that a lot of it has to do with a mind-set. I was reading in English and came across a collection of letters which happens to be a word in English, so that was the first thing that came into my head. Had I been reading something in Japanese romaji then that probably wouldn't have happened. I think reading in kana helps me get in the right mind, especially when reading beginner grammar books and the like, which have a mix of English and Japanese.

JimmySeal Wrote:There's a one-to-one correspondence between romaji and Japanese pronunciation. On the matter of the romaji <-> kana correspondence, the Hepburn system ignores ず/づ　じ/ぢ and sometimes お/を, but learning those exceptions is far less effort than Japanese students have to put into learning English spelling.
Nobody has yet laid out a way to accurately represent English pronunciation using kana, and even if they did, no one would bother to learn it, so it causes ambiguity and bad pronunciation, and therefore is a completely separate issue.

The one to one correspondence that you mentioned is extremely important and is a very good point. For this reason, it would certainly seem that romaji can represent Japanese sounds very well and katakana can't come close with English. However, if you look at how katakana has developed, there are now characters such as ヴぇ or ヴぁ which are used to distinguish between the 'v' and 'b' sounds in English. I'm sure that there are methods for representing the difference between l and r and so it seems clear that the amount of sounds that katakana fails to distinguish between is reducing.

You mentioned that when using romaji, you just have to remember that 'ha' in 'hana' isn't pronounced in the same way as it is in English. Nothing in Japanese is pronounced in the same way to anything in English. Some things are very similar but they're all different. Your point was that we remember differences with Spanish, so just do the same with Japanese. So why can't a Japanese person do the same with katakana? They just remember that the ム in ジェームズ is pronounced weakly and with no 'u' sound. They could apply a system of rules, just as you do with romaji, and they could create a one to one correspondence by creating more characters as was done with ヴぁ and ジェ and others. You think that no one would bother to learn it but a Japanese person would probably find it much easier to read because they're really familiar with katakana just as we are with roman characters. I think they would bother and I've met students that do.

You didn't point out the big difference between kana and romaji though. English is a language where every word has a single stressed syllable. The other syllables in that word are much weaker. There is no stress in Japanese and it is a rhythmical language instead. Perhaps the biggest advantage of using kana over romaji is that each kana represents one 'beat' in the rhythm of Japanese. With romaji, it's usually two characters ('ka', 'te', etc), sometimes one ('a', 'i', 'n', etc) and sometimes three ('tsu', 'shi', 'kya', etc.) So when reading kana, the rhythm of how it should be pronounced is much more obvious.

So whether you're learning English using katakana, or learning Japanese using romaji, there are lots of rules that you have to learn in order to get the pronunciation right. As you said in your post, with romaji there are far fewer rules but there are still a lot of them and more than most people realise.

So why do people bother going to all this trouble learning these rules when they could save themselves a lot of effort and just use the system that the natives do?

wrightak Wrote:So whether you're learning English using katakana, or learning Japanese using romaji, there are lots of rules that you have to learn in order to get the pronunciation right. As you said in your post, with romaji there are far fewer rules but there are still a lot of them and more than most people realise.

So why do people bother going to all this trouble learning these rules when they could save themselves a lot of effort and just use the system that the natives do?

Because the way people learn and their strong points differ? I doubt there is one true way in all this debate. I will never dispute the need to learn kana if you're serious about Japanese, but I refuse to go as far as to say that romaji are always worse than kana. For some the superiority of kana seems an obvious conclusion, for others, it's not. People have told me in this thread that my kana reading with speed up over time, and that eventually I will be faster in kana. At this point, I'm not prepared to accept that as the unconditional truth. My readings speed with roman letters is extremely high, a skill I've developed for over more than 35 years now. Kana are simply 35 years behind in catching up.

For instance, while at Japanese class yesterday, I noticed that while I write all the practice sentences in kanji-kana, I'm now creating 'romaji furigana' above the lesser known kanji readings instead of kana ones. This works way better for me than scribbling small unreadable kana above the kanji - especially since the tokens between voiced and unvoiced varieties are so difficult to decypher if you write very small. This may be because my handwriting is not as good as print, or because of size issues. And sure, I may overcome that issue ione day. But I don't see it as a drawback at all. I write kana were kana are best, and I use romaji where romaji work best.

wrightak Wrote:So why do people bother going to all this trouble learning these rules when they could save themselves a lot of effort and just use the system that the natives do?

No effort for me. I was able to read romaji immediately, zero learning time (probably due to my knowledge of Spanish, etc). Kana took me about 15 hours (spread out over a month) to learn, and perhaps 6 months to get proficient at with my leisurely pace.

astridtops Wrote:People have told me in this thread that my kana reading with speed up over time, and that eventually I will be faster in kana. At this point, I'm not prepared to accept that as the unconditional truth.

If you get to a point where you want to improve your kana quickly, you might get one of the nice little kana workbooks. I noticed big improvements fast when I used mine. Of course, I didn't use it consistently, and sort of took my time (I don't have to worry about a class). I'm going to go back and drill some day because I sometimes get stuck when I write katakana. And as an aside, why do we learn to write "wo" in katakana?