right now, im at 500 kanji from studying for a little over a month. i'm now spending the majority of my time listening/reading/speaking instead of kanji studying. so how do i go about learning how to pronounce the kanji? the method i'm doing right now is whenever i come across a kanji that i already know in my studies, i go to my flashcards and write the pronunciation next to it.

my question is, should i be doing this? i'm sort of in a rush since i'll be in japan in 3 months, and i want to focus more on my speaking/listening. fyi, i have the second volume as well

This is just my method by I go with a systematic approach to the readings.
I learn by using compounds and the sole verb/noun for other readings. The way its approahced is by starting from square one and building up.

Example, I learn the readings for 年. Toshi and Nen. I use compounds that have a kanji I already know
生年月日
五年間
年金

Now since 年 has been added to my resevoir, the next kanji I learn uses one's I already learned plus that one(if it comes up)
前 Zen and Mae
以前
前年
分け前

And it builds up. Obviously these are very basic kanji so its just for example's sake.
The book "kanji and kana' and many others build up this way. With the examples only using previously gone over characters.

It works for me and I add about 60 a day and actually do learn at that pace. Needless to say I have about 400 reviews a day on just vocab alone but because its set up with this system its fast. Just like RTK where you only use primitives already learned. I used the same structure.

Aside from this I of course go over grammar, read, listen, etc like normal.

Focusing on your listening/speaking is definitely a priority for you now seeing as you're going to Japan soon.

I hate to break it to you but it takes a long time to learn to read, even to get to the point where you can read basic stuff to an OK level. (Unless you blitz KO2001 in 1 - 2 months)

I suggest if you want to get up with the play on reading stuff you have a couple of options. First off learning to write is not necessary for learning to read so if you want to learn to read maybe you could put writing on the back burner in order to focus on being able to functionally read. Second of all frequency will get you pretty far in terms of reading basic/important things so learning to read kanji in order of their frequency will help you out in that regard.

Nothing beats actually reading Japanese and looking things up in the dictionary and then entering into your SRS to learn kanji readings. In fact you can learn to read other ways by learning every on/kun reading of tonnes of Kanji and then just trying to read but you still wouldn't understand absolutely anything. Sounds like you don't have time/skill enough for sentence mining so vocab in isolation would work for a short term goal. Kanji Word - J-E definition cards.

3 Months is a decent time frame to prepare yourself IMO you've got a good chunk of time that if you work hard in you should be able to come out with a decent enough ability to read functional every day things that you can get by.

Perhaps if you set yourself a goal to be able to read all the words in a JLPT3/4 list you'd come out of it being able to walk instead of crawling. I guess trying to balance that study in with your speaking/listening practice is the real trick.

Good luck!

If you've already started reading, then you'll soon enough get a sense for when a word uses onyomi (chinese derived reading) or kunyomi (japanese reading). One way to learn readings is to read a lot from texts with furigana. Note down words those readings you don't know and add them to your srs, testing kanji->hiragana+meaning, and mentally keeping note whether its kunyomi/onyomi/name reading etc. Even if you don't add cards to your srs, if you do enough reading, a lot of the common words will be learned quite quickly, especially if you're reading plenty of furigana-less text too. You want to be srsing rarer words.
Another step I highly reccommend during the kanji learning (RTK) phase, is attaching one common onyomi (chinese reading) to each kanji as you learn it. Some will argue learning readings in isolation from actual words is a waste of time, and this is probably true for kunyomi readings. As for onyomi however, most kanji have one base reading which is used in 95% of compound words. When words use alternate onyomi readings, they are usually either very common words (oft encountered = easily learned) or very rare words (less important). So if you can rapidly learn the one most common onyomi reading for each kanji as you learn it you'll get a lot of mileage from it in the long run. Pre-learning onyomi in this way will make it easier when you start reading, just as pre-learning how to write them does. "pre-learning" the onyomi can be done extremely simply by using a consistant memonic for each onyomi reading. So imagine you encounter a kanji such as 缶, with onyomi かん. Think of an appropriate memonic for the sound かん (I use ghenghis(spelling??) KHAN) and work it into kanji stories (eg. Ghenghis KHAN likes crushing beer CANs on his forehead). Then when you go through RTK, test yourself on onyomi and writing.

Also, make sure at least some (if not all) of your reading material is furigana-less.

I recommend not learning readings at all. Learn words instead. How a kanji is pronounced in a word you don't know yet doesn't affect you at all, and when you learn the word, you will learn that reading automatically. If you actively want to learn the readings for a kanji, I would recommend looking up a few common words using it and simply learning those words instead.

What Tobberoth said.

先生 「せん・せい」
友達 「とも・だち」
温泉 「おん・せん」
寿司 「す・し」
芸者 「げい・しゃ」
弁当 「べん・とう」
忍者 「にん・じゃ」
歌舞伎 「か・ぶ・き」

8 simple words which you probably already know. 17 readings.

Why learn them separately?

My suggestion, was to learn 1 onyomi reading with each kanji WHILE doing RTK, simply because you may as well because it's so easy. And it DOES make it easier when you start reading. And I'll add as an afterthought, that it could improve guessing ability while listening. Note this is a recommendation for people at the RTK stage. Obviously if you've already completed RTK and are reading plenty, there is less (no) need to systematicly do this.
Another suggestion, though I haven't tried this, you could check one common word for each kanji and work it into your memonic. This would be more beneficial for kanji which appear more frequently in kunyomi form (such as verbs and adjectives). Though the obvious problem is that it's difficult to create sound memonics for kunyomi because they're usually multi-syllabic.

[Katsuo]

harhol Wrote:Why learn them separately?

Those who are adept at using mnemonics may find it to be a more efficient method.

Using a mnemonic to attach an ON reading to a kanji usually gives me a quicker, stronger and more dependable connection than relying on having seen the same kanji used as part of another word.

Also, using only the "It's the same kanji as in word XX" method would mean that you don't take advantage of "signal primitives" as outlined in RTK2. There are quite a few kanji where the ON readings can be learned in groups, for example, 剣険検倹験 all include a similar element and share the reading ケン.

On the other hand, if a word uses rare kanji (and those kanji do not have signal primitives I know), then I generally find it more efficient to learn the word without reference to the kanji readings.

nadiatims Wrote:My suggestion, was to learn 1 onyomi reading with each kanji WHILE doing RTK, simply because you may as well because it's so easy. And it DOES make it easier when you start reading. And I'll add as an afterthought, that it could improve guessing ability while listening. Note this is a recommendation for people at the RTK stage. Obviously if you've already completed RTK and are reading plenty, there is less (no) need to systematicly do this.
Another suggestion, though I haven't tried this, you could check one common word for each kanji and work it into your memonic. This would be more beneficial for kanji which appear more frequently in kunyomi form (such as verbs and adjectives). Though the obvious problem is that it's difficult to create sound memonics for kunyomi because they're usually multi-syllabic.

Totally second this idea for anyone at the RTK stage. Wish i'd done it. I can see how it would be virtually no extra work to remember 1 onyomi.

Either that OR when learning the kanji learn a very common vocab word with it as a way to reference it.

What I mean by that is for a kanji like 生 to learn a reading with it you have a word on the card "先生”　and when you think of this kanji you can think of it like 先生の生。 This approach works really well in giving you a JAPANESE mental hook to remember kanji instead of a "keyword". It works well for me when I want to write or remember which kanji goes in a word like 価格 for instance I forgot which kanji the かく was and then when I checked the answer my immediate thought was ahhhh　性格の格です！ This is a great way to clarify kanji when talking to Japanese people.

Katsuo Wrote:

harhol Wrote:Why learn them separately?

Those who are adept at using mnemonics may find it to be a more efficient method.

Using a mnemonic to attach an ON reading to a kanji usually gives me a quicker, stronger and more dependable connection than relying on having seen the same kanji used as part of another word.

Also, using only the "It's the same kanji as in word XX" method would mean that you don't take advantage of "signal primitives" as outlined in RTK2. There are quite a few kanji where the ON readings can be learned in groups, for example, 剣険検倹験 all include a similar element and share the reading ケン.

On the other hand, if a word uses rare kanji (and those kanji do not have signal primitives I know), then I generally find it more efficient to learn the word without reference to the kanji readings.

Actually, you get the signal primitives anyway. I have no idea how, but you usually automatically guess the reading from the signal primitive even if you don't know the kanji from a word, without having learned that it was a signal primitive. I guess it's just that logical of a system (as long as they actually follow it).

The truth is that there is no pattern when it comes to kanji readings. Most onyomi originate from different periods in Chinese history. So really there is no logic to kanji readings. It is what it is. So honestly, any system is as good as any other. This is why I prefer to just learn readings as I encounter them. No need for radicals or anything else. Kanji do not have primary and secondary readings. Just different readings.

Then of course, there's always systematic methods like memory palace, kanji chain and "movie method".

For "common" Onyomi (according to a list that Katsuo puts out), there's only a couple hundred kanji from RTK that have two common Onyomi. The rest have just one. With that bit of insight, it's useful to learn Onyomi in a systematic fashion.

For Kunyomi and words with uncommon Onyomi, you just have to suck it up and learn those as they come along. I don't think it's worthwhile to learn those in a systematic fashion.

I think another good point to learning pronunciations from words like I recommend is the knowledge that you can skip a while "arbitrary step". I know a lot of people like to take everything in arbitrary steps (Rtk -> RtK2 -> KO2001 -> Smart.fm) but a lot of people want to just learn right away and are only doing RtK because they can't find another good way of doing it. For such people, I think it might feel good to know that you're learning actual words which have uses immediately.

Hint: don't ask for the "best method." This place is full of personal experiments, not divine revelation. Not a bad thing at all--this is the bleeding edge of self-directed language acquisition--but that the concept of "best" is quite hazy and tempts people to express their own experience as settled scientific law.

Now, my experience is that studying readings during RtK 1 or soon after is a bad idea, and I wish I had quit it sooner. There are two main reasons for this.

First, my pronunciation really sucked back then. Learning a couple dozen characters have the sound カン is only helpful if you're in the right place pronouncing カン. Unfortunately, final ン is quite tricky for native English speakers. You need to experiment with it a while in order to become accustomed to how the exact pronunciation changes with context. With kanji-chains you can associate those few dozen カン字 with a sound something like "con" or "corn" or "Cannes," but unless that sound is decently close to 「カン」, you haven't gained anything. Even if it is, you still have to learn the other contexts: カン alone differs slightly but significantly from 関 in 関係（カンケイ）.

Second, you're learning something abstract, and this necessitates some fairly heavy-duty mnemonics (like chains) and can really sap your motivation. I tried it, it was slow and hard even with the funny stories, and I was much happier after I switched over to Core 2000.

I now have the opinion that you shouldn't learn hard stuff at the expense of easy unless you have good evidence the difference in utility outweighs the difference in cost. Isolated readings are fairly low utility: they'll certainly help you guess new words, but you still have to learn whether that guess is right. You can do the exact same thing if you know other words that contain the characters of the new word. They're also comparatively expensive: there are all kinds of useful sentences and words you can learn without any mnemonic at all.

If you want to learn readings, my advice is to first, have a good grasp of pronunciation (do a few hundred sentences with audio first), and second focus on useful, easy to learn 音 groups--don't go for completionism unless and until you need to know all the isolated 音読み. (e.g. studying for the 漢検)

A trap beginners like myself (yeah, I'm still considering myself a beginner) fall into is using a systematic approach for everything before getting into "real" Japanese. So we do RTK as a systematic way to learn kanji, then move on to Tae Kim or UBJG as a systematic way to learn grammar, then onto Core 2k or 2k1KO as a systematic way to learn vocabulary.

Thing is though, I'm feeling the sooner you get into "real" Japanese, the better. Yeah, I'll pimp out doing subs2srs as a qualifier but only because it helped me. Perhaps after getting real, used in actual situations in Japan type Japanese under your belt then it's a good idea to go back and fill in gaps with systematic approaches.

Example: You learned 4231 vocabulary words via subs2srs or sentence mining Death Note, My Girl, Jin, 20th Century Boys and Prison Break (Japanese dub). Still, you note when looking at Wikipedia and 2ch there's words you kind of know due to kanji from RTK, but pronunciation isn't there because they didn't come up from your mining. Maybe it's worth the 10 to 20 hours of studying to learn common Onyomi for the Jouyou Kanji via Movie Method.

Perhaps after 5000 sentences, you think you're missing nuances on some lyrics of Arashi and pages from Neon Genesis. Maybe it's time to systematically go through the 700 sentences of Kanzen Master.

Later at 8000 sentences, you might want a break from the Nodame Cantabile mining and decide to round out Onyomi for upto Kanken level 1.5 and do sentences from Kanzen 1.

These systematic approaches can be left on hold, and will be there when you hit a natural pause. Hell, there may even be a better list to draw from by the time you get around to doing them. It's just that I saw such an enormous boost from doing subs2srs (which is still just sentence mining) in my natural use of Japanese I regret putting it off for so long having done a lot of systematic items.

On the other hand, systematic approaches give the bang for the buck and covers a lot of useful material. One would be a fool to blindly ignore those as well. Hell, who knows, you may get to such a level that you're able to use systematic lists made in the target language itself. I base that off of one of Khatzumoto's post about reading a grammar book about Japanese (in Japanese) and realizing he knew so little. Plus, Tobberoth has had good things to say about Kanzen Master which is in Japanese.

Long post short, there's advantage to learning Kanji pronunciation systematically AFTER you've learned a lot of them pseudo-randomly (vocabulary lists, sentence mining and regular exposure).

At ~1000 I started to move to monolingual sentences.
And around ~4000 sentences I stopped with the SRS completely.

The SRS is great to get you off the ground, but as soon as your get your speed, go for real texts and real audio.

This is of course entirely subjective, but I'd say forget about learning any pronunciation for _kanji_ altogether. Concentrate on learning the pronunciation for words instead. IMHO, you are going to gain relatively little by memorizing onyomi readings beforehand. There are enough exceptions that you'll have to check the reading of every new vocabulary word anyway. Besides, as you work through vocabulary, you are going to catch the common onyomi readings automatically, so there's little point spending extra effort for that beforehand.

So, if I now started everything from scratch again, I'd do it like this:

1) Finnish RtK 1
2) Read through Tae-Kim (just once or twice, so that you have the basic idea. No need to SRS grammar stuff at this point)
3) Start reading native material with a dictionary and add every new word you run into to your SRS deck. Check grammar that you don't recognize from Tae Kim or something more extensive, like the "Dictionary of Basic/Intermediate/Advanced Japanese Grammar".

@shang

I almost totally agree with you except for 2 things. That's what I'd do if I was starting again.

One is listening. Listening hours are really important and you don't have it in your schedule.

Other is that I think that general knowledge of on yomi is important. You don't need to review them like you review the writing, but you should have a method to remember them.

ZuMash Wrote:my question is, should i be doing this? i'm sort of in a rush since i'll be in japan in 3 months, and i want to focus more on my speaking/listening.

Could you give us some idea why you are going to Japan and for how long? Since you will be leaving in 3 months, I wondered if you're going for the Japanese academic year?

Kanji: If you'll be studying, then kanji study might make sense. Why not combine Nadiatim's and Katsuo's ideas and learn on-yomi only for kanji in the pure and semi pure groups in your RTK2 book. Since you're learning primitives anyway, it's not much extra effort to note that a particular primitive always gives kanji a certain reading (or every kanji but [exception]). You then need only associate the reading with the primitive rather than with each kanji. (Note that a phonetic primitive doesn't meaning other readings are impossible, just that it will have that reading). Learning RTKLite might also be a good idea (if that wasn't mentioned). Where you already know a suitable Japanese word, use that as your keyword. To learn readings after RTK, I suggest learning common vocab for kanji in Kanji in Context order. (2001KO?)

Spoken:If you're not a student, then I would leave kanji for later and focus on learning situational spoken Japanese. You may already be doing this, but I'll describe it just in case. [in separate post, I'm out of time]

If you have 3 months, you can get thru at least KO2001 book one for sure if you buckle down. It's systamatic, teachers you plenty of relevant voc and readings, and will set you up to be a pretty good reader by the time you get to Japan.

I wonder why most people do it in a first do this and than that way. Why not just mix it up?
What's keeping from already going through the first parts of tae kim, some systematic vocab phrases to get the pronunciations down. Or starting to get used to the 'real' japanese with sub2srs decks?

If you continue to do about 15 kanji average with rtk, you'll be up to 2000 in three months, or you can work you're way up to the rtk lite level within the next 1 to 1,5 month.
With one lesson a day of Tae Kim you can be finished with it within two months or so, 45 ko2001 or smartfm phrases for vocab for 15 kanji a day from those you've already learned, will take something over two months. And in the last month you can almost fully concentrate on how you do with native level material and let everything you learned sink in.
And of course if you set up a low leech threshold, sifting out everything you find hard to learn, you can fully concentrate that last month to fill in those blanks.

In this way instead of working yourself up from basic material to harder material over and over again. You can start with several bases at once, the basics are the most important anyway, and they can strengthen eachother on the go. Even if you don't finish it all in time, you're sure to have a lot of the most basic stuff down. And you'll be more skilled to understand Japanese than by just managing to go through all of rtk 1 with some readings. the more advanced kanji's, grammar and vocab is probably better suited to learn while you are in japan anyway.

Just set some goals, see what you need to do to get to those goals, what materials you have available, and how much time you have to spent to get there. See how you manage and just change your ways accordingly.

At least that's what I'd do with a deadline like this.

sugarlevi Wrote:And of course if you set up a low leech threshold, sifting out everything you find hard to learn, you can fully concentrate that last month to fill in those blanks.

In this way instead of working yourself up from basic material to harder material over and over again. You can start with several bases at once, the basics are the most important anyway, and they can strengthen eachother on the go.

I like this idea, but there's a problem with it. RtK grows your vocabulary of primitive elements over time. Tae Kim jumps in to the full set from the start. Thus, RtK will be playing catch-up for a long time before it makes kanji easy.

As just a random example, here's example # 254 from The Guide (thanks to Nukemarine's deck).

重い荷物が持てます。

In RtK:
0660 持
1013 荷
1050 物
1675 重

Example 254 / 815 is about 31% of the way through The Guide. 31% of the way through RtK 1 is around frame 640--you'd know 艹,可,寺,牛 but would still be waiting on the primitives 亻,勿,重 used in that sentence.

Starting Tae Kim before finishing RtK will be similar to starting Tae Kim without RtK because TK introduces primitives much faster (but much less memorably) than RtK.

Side note: it's somewhat unusual (though not unheard-of) to be productive at all after only three months of study, even with many hours of instruction a day. Even then, you can't expect to have more than an extremely basic output ability at that point. http://www.eslteachersboard.com/cgi-bin/...;read=1677 . Sorry. Language acquisition doesn't happen in Pimsleur Time.

wildweathel Wrote:

sugarlevi Wrote:And of course if you set up a low leech threshold, sifting out everything you find hard to learn, you can fully concentrate that last month to fill in those blanks.

In this way instead of working yourself up from basic material to harder material over and over again. You can start with several bases at once, the basics are the most important anyway, and they can strengthen eachother on the go.

I like this idea, but there's a problem with it. RtK grows your vocabulary of primitive elements over time. Tae Kim jumps in to the full set from the start. Thus, RtK will be playing catch-up for a long time before it makes kanji easy.

As just a random example, here's example # 254 from The Guide (thanks to Nukemarine's deck).

重い荷物が持てます。

In RtK:
0660 持
1013 荷
1050 物
1675 重

Example 254 / 815 is about 31% of the way through The Guide. 31% of the way through RtK 1 is around frame 640--you'd know 艹,可,寺,牛 but would still be waiting on the primitives 亻,勿,重 used in that sentence.

Starting Tae Kim before finishing RtK will be similar to starting Tae Kim without RtK because TK introduces primitives much faster (but much less memorably) than RtK.

Side note: it's somewhat unusual (though not unheard-of) to be productive at all after only three months of study, even with many hours of instruction a day. Even then, you can't expect to have more than an extremely basic output ability at that point. http://www.eslteachersboard.com/cgi-bin/...;read=1677 . Sorry. Language acquisition doesn't happen in Pimsleur Time.

This is true.

Although I'm guessing your post is from a standpoint of being able to write everything in Tae-Kim? Otherwise what do primitives (or even knowing kanji) matter to learn to read them?

I highly suggest RTK-lite. 1000 kanji covers 80% frequency (so true) and can be a quick way to get you up to speed on all the kanji used in Tae-Kim (don't quote me there) and KO2001.

You don't actually need to have "learned" a kanji in RTK to be able to read it and understand it in a word. At first I thought kanji was so complex that you actually did have to have learned it before so that you knew what it meant and could come to terms with it in the real world. I find myself learning new kanji often that i've never studied through RTK and I can read and understand them just fine.

Just to reiterate my point, if you want to invest more time in learning to read you can do so by dropping RTK and moving straight into Tae-Kim and then onto KO2001. Pick it up when you want to learn to write or want a solid foundation of kanji skill.

The choice is yours.

mezbup Wrote:Otherwise what do primitives (or even knowing kanji) matter to learn to read them?

Only about as much as letters matter to reading alphabetic scripts. Or phonemes matter to spoken languages. Of course you need to recognize primitives!

You may not have to study them explicitly. You may not even use the same set as other people, whether you make 夋 one primitive or three matters about as much difference as if "ing" is one letter or three. However, explicitly or implicitly you'll either end up learning them or stuck at low proficiency.

If you've finished RtK it can be hard to remember how much kanji used to suck. So, it's helpful to look at what other people put themselves through: http://www.road2japan.com/kansai-gaidai-...ace-ified/ . Yuck. That's what the "kanji are ugly pictures" approach gets you.

As for "perspective," well, I'm just someone who thinks "whole word" sucks for English and would take a lot of convincing to become a fan of "whole kanji."

PS: Yes, I'm aware of the apparent hypocrisy of supporting sentences for learning written Japanese as a foreign language. Forget "whole word," isn't that "whole phrase?" It isn't, and I'll explain on request.

wildweathel Wrote:

mezbup Wrote:Otherwise what do primitives (or even knowing kanji) matter to learn to read them?

Only about as much as letters matter to reading alphabetic scripts. Or phonemes matter to spoken languages. Of course you need to recognize primitives!

You may not have to study them explicitly. You may not even use the same set as other people, whether you make 夋 one primitive or three matters about as much difference as if "ing" is one letter or three. However, explicitly or implicitly you'll either end up learning them or stuck at low proficiency.

If you've finished RtK it can be hard to remember how much kanji used to suck. So, it's helpful to look at what other people put themselves through: http://www.road2japan.com/kansai-gaidai-...ace-ified/ . Yuck. That's what the "kanji are ugly pictures" approach gets you.

As for "perspective," well, I'm just someone who thinks "whole word" sucks for English and would take a lot of convincing to become a fan of "whole kanji."

PS: Yes, I'm aware of the apparent hypocrisy of supporting sentences for learning written Japanese as a foreign language. Forget "whole word," isn't that "whole phrase?" It isn't, and I'll explain on request.

You make an Ok point. Yes, it definitely helps. I still think it's not entirely necessary although I definitely agree it's much much easier when you're familiar with the various primitives. Hence why I'd advocate an RTKlite > Tae Kim > KO approach. If you can pump that out in three months it's good bang for buck IMO!

LOL@jace. God that's a horrible method of learning to write.

edit: Interesting page you linked to there about reading. It's crazy to try the whole word method for a language with an alphabet! I totally agree with what they're saying there that the child will never actually learn to read. Although I guess we learn to read via phonetics and as adults the words by now are fairly ingrained as to have become pictures almost.

I think kanji's largely different in that it doesn't provide phonetic clues and is highly visual. Like I said, I don't have much problem assimilating how to read new kanji but we also must take into account I'm quite familiar with the subset of primitives used to construct the characters and this helps me in telling them apart.

At any rate, I love both systems of orthography.