First of all, I'd like to thank everyone here for creating such an helpful site! In the last few months, I have been lurking around this site and have been amazed at how much work has been put in. The help and inspiration I found on this forum may have been what kept me learning and enjoying Japanese!

My predicament is this: I have just moved into a more intermediate level with my Japanese and am now reading full-length novels (currently ゼロの使い魔.) However, I am experiencing one major problem. The only thing that really holds me back is vocabulary! For example, the word 鞭　and 照尺 aren't exactly common words but not knowing them can be the difference between comprehension and being totally baffled. Take this sentence: ルイズは鞭を持って、才人を近づいた。If you don't know that 鞭　means 'whip,' its hard to understand that the sentence is in fact quite sinister. Left with this problem, I have spent the last few weeks considering a solution.

I am currently using Anki 2, and have just added unknown words to a deck and made listening and reading cards with the help of Google Talk. However, my retention rate is not very high at all and I am forgetting most of the words the next day. I think this is because the words are too dissimilar. If I learned　心中、心配 and 心理 together, I could easily see that 心 reads しん in those compounds. Somehow, knowing the reading and the meaning from RTK1 really helps me when reading Japanese. 'Sounding out' the word in my head (eg: しん＋ちゅう= center) sometimes is enough to jog my memory. In addition, seeing the kanji　心 over and over helps me associate it with 'heart' (RTK1 was along time ago ) I was thinking that it would be more effective to study a few kanji at a time (ex, compounds with 心.)

When I used Anki 1.8, it was easy to group by kanji because of the plugin "Mighty Morphin' Morphology (formerly Japanese Morphology.)" However, an Anki 2 version was never made.

What does everyone think about grouping by kanji? Can it be effective or will it hurt me in the long run? Also, is there another way I can group by kanji in Anki 2?

I group by kanji on Core6k. I don't do anything fancy. All words not learned yet are suspended. I basically go down a kanji frequency list and unsuspend all cards that use the target kanji and any I already know. Repeat with another kanji if needed until I have enough words for the day. Sometimes I'll see kanji that are used together - 軍 and 隊 for example - and deviate from my frequency list. The exact order isn't important. But seeing several new words at once that all use the same kanji (often, but not always the same reading or two) helps me a lot I believe.

I'm not a very good programmer, so I don't have any automated way to do this, sorry.

I'm glad to hear that this method actually worked for someone: I haven't been able to try it myself. I think this is why Core 2000 was so easy; most of the kanji appear in more than one word.

Also, the frequency list is a great idea! I'll try that with my deck, just search the kanji in Anki and unsuspend the cards that come up. Hopefully learning vocab wont be as mind-draining now~!

HououinKyouma Wrote:What does everyone think about grouping by kanji?

Grouping vocab by kanji (and ordering the kanji by frequency) is what makes "Kanji in Context" such a good textbook.

I've already done RTK1 and Core 2000. Do you still think I should get the Kanji in Context textbook?

HououinKyouma Wrote:I've already done RTK1 and Core 2000. Do you still think I should get the Kanji in Context textbook?

Despite its title, "Kanji in Context" is more about vocab than kanji itself with aim of developing good reading skills. My estimate of the vocabulary from there is that KiC has around 6000 (maybe even 9000) words, so it's well beyond Core 2000. Whether you should get it or not, I can't tell. Will you be comfortable with just Core 2000? How are you going to reach higher levels? KiC is not the only answer.

Right now, I will probably just continue on to Core 6000. Most basic conversation is easy enough for me to understand but there are a TON of specific words that come up when reading. For example, the bench of a horse drawn carriage has a specific name! It feels so overwhelming to see words like that on every page of a novel. Most people say that you can discern a good amount of words from context, but it often feels like that's not the case, especially when those words are nouns. Kanji, being ideographic, help alot in guessing the meaning of some words, but others are just crazy! I know I cant study EVERY random word, but I want to be able to read adult novels without getting thrown off every time they talk about iron prison bars (監獄の鉄格子 apparently) and whatnot

This may or may not help you, but lately I've taken to putting my example sentences on the front of my cards. If I can't recognize a vocab reading immediately, I read the sentence. If I still can't remember the meaning of the word after a few moments, I fail the card. This gives me the benefits of a straight vocab deck along with the benefits of a sentence deck. I've noticed it's a lot easier to both recall and learn vocab using this method. Plus, it's helping me remember some useful collocations.

Thought I'd put that out there, as I'm doing the same thing as you, only with a different book. YMMV.

As for reading novels, I have to say that getting 3/4 of the way through きみにしか聞こえない has done wonderful things for my Japanese comprehension skills.

Yeah, I totally agree! I finished the first volume of Zero no Tsukaima and I've noticed my Japanese skills shoot through the roof! Hmmm... I think I might try putting sentences along with my vocab cards in the future, but I haven't added any sentences to the 800 or so cards I created from the first book. I'll try the sentences with the second book and compare, maybe then I can actually find a solid vocab building strategy. Lately, it feels like I've been changing my study method every other week! Hopefully I can find one that sticks

HououinKyouma Wrote:Yeah, I totally agree! I finished the first volume of Zero no Tsukaima and I've noticed my Japanese skills shoot through the roof! Hmmm... I think I might try putting sentences along with my vocab cards in the future, but I haven't added any sentences to the 800 or so cards I created from the first book. I'll try the sentences with the second book and compare, maybe then I can actually find a solid vocab building strategy. Lately, it feels like I've been changing my study method every other week! Hopefully I can find one that sticks

How are you culling words? Rikaisama? I'd highly recommend configuring the Rikaisama template to add the sentence automatically from the content you're reading. My template is:

$d$t$r$t$n$t$g$t$s

This isn't always valuable (some sentences are too short or too highly contextual), but 9 times out of 10, you'll be glad you have it.

I also have some early cards that don't have sentences. I review my failed cards every day using an Anki 2 Learning/Cram deck, which lets me review each failed card in detail and work on memorizing it. If I have a card without a sentence, I use Google or Goo to try and find a sentence. If I don't find one, and if I didn't get the word from a TV show and I don't recall running into it lately, I suspend the card until I see it in the wild again.

Wow, I completely overlooked the sentence import tag feature in rikaisama! Thank you for pointing that out! *goes off and changes template* Should have looked more closely when I set that up. Guess I was overwhelmed at the novelty of auto-importing words into anki

I just thought I would post that MorphMan 3 is up on Anki 2s shared plugins. It offers the functionality I talked about with grouping by kanji. Thanks a million to the author!

I just thought I'd point out another series of books that often seems to get overlooked. The Kanken study books. Now if you want to learn only the most common words this is probably not the greatest idea but since you want to learn words that use the same character in groups it's actually very useful.

The book I'm using at the moment (Kanken 4), which I assume works the same way as previous levels, has one page with 8 characters introduced followed by 3 pages of excercises. For each character there is up to 8 example words + any kun readings the kanji has. As an example, for the kanji 歓 the following 8 words are provided: 歓喜　歓迎　歓呼　歓声　歓送　歓待　歓談　交歓 and while all of these words aren't necessarily super useful it certainly cements your understanding of 歓 as a character, including the reading of course since it's pronounced かん in all the the above words.

Now a lot of people are against learning words in isolation but luckily that problem is (partly) taken care of in the book as well in the following exercises. There are some different kinds of exercises for each "stage" (group of 8 characters), but you're guaranteed to have at least 48 example sentences, and for potentially 60-70 words it's not too bad.

The strong point of this book is that you learn each character really well (which improves your chances greatly to guess meanings of new vocab containing that character). However there are the weak points as well. It's not too cheap since you're likely going to have to buy several books since each book only contains so many characters. 1000円 for a book is pretty cheap in my opinion but if you have to buy 3 AND pay for shipping overseas it can get a bit more expensive. Another point is of course like I said before, you learn a fair amount of words you may not find instant use out of (which luckily however has the nice side effect if giving you increased understanding).

I realize this isn't for everyone but it sounded very much like what you were describing and it has helped me a ton so I'll leave this post here for anyone who is interested.

HououinKyouma Wrote:Yeah, I totally agree! I finished the first volume of Zero no Tsukaima and I've noticed my Japanese skills shoot through the roof! Hmmm... I think I might try putting sentences along with my vocab cards in the future, but I haven't added any sentences to the 800 or so cards I created from the first book. I'll try the sentences with the second book and compare, maybe then I can actually find a solid vocab building strategy. Lately, it feels like I've been changing my study method every other week! Hopefully I can find one that sticks

That seems pretty impressive to me that you were able to read through the first Zero no Tsukaima light novel with RTK1 and Core 2000. And you say there were just 800 unknown words over the 2000? I may make that my first light novel to read. How difficult is the grammar in it?

PotbellyPig Wrote:That seems pretty impressive to me that you were able to read through the first Zero no Tsukaima light novel with RTK1 and Core 2000. And you say there were just 800 unknown words over the 2000? I may make that my first light novel to read. How difficult is the grammar in it?

For the most part, the grammar in ゼロの使い魔 is very basic. Lots of simple sentences of the kind that HououinKyouma quoted. It's fantasy, so you'll get a number of words that aren't very useful, but it's good, simple reading. You might need a little beyond RTK+Core2K to feel comfortable with it, though. Give it a try and see.

Yeah, gaislastlaugh said it well. There's some words that seem pretty useless like 風竜, but overall, it's very dialogue driven and there's not that many intricate descriptions. I noticed that when I tried to read the Steins;Gate visual novel, there were a lot of detail-heavy descriptions that really bogged me down and worst of all, bored me to death. They would spend paragraphs and paragraphs detailing how the heat made sweat drip down his face... I decided to stop reading SG because I already knew the story and it was too boring to trudge through it again. On the other had, the author's style in ゼロの使い魔 flows very well and keeps you interested, throwing in lots of good comedy along the way. Even though it was hard at some points, I kept reading because I wanted to know what happened to the characters. There were even times I forgot I was even reading in a foreign language I was so engrossed in the book! That was a cool feeling (even if it lasted half a paragraph cause I had to look up another word...) I would defiantly try reading it even if you haven't completed Core 2000.

I think the other thing about actually reading a novel is that you can use context clues. For example take this sentence: 才人は鞭で叩かれるたびに、心底情けない声をあげた。 You don't really have to know the exact meaning of 心底情けない to understand the sentence. If you didn't know that word, for instance, you would still understand that Saito is crying out because he's getting whipped. The best thing is that the context is the whole book; its not just one sentence. If you don't understand a sentence, move on to the next and see if you understand it. You'd be surprised how much we do it in our own language. If an English author but the word 'elephantine' in book, most people would only be able to guess at its meaning using context.

I think I'm going to try this title as my first light novel. I'm actually closing in on finishing adding to Core 6000. When I finish that I need to do some grammar studying/review and then I'll try to read it. Would you say Tae Kim covers the grammar for it or does it dip into N2 grammatical structures? I have the Kanzen Master 2 Kyuu and was going to get into it after Tae Kim.

Although Tae Kim's grammar guide was the only actual grammar 'textbook' or whatever you'd call it I studied, there were some grammatical forms I did have to look up. For example, I don't remember Tae Kim talking about the ず　form of verbs or みたい when it follows a noun or adjective. However, I found that the stuff I didn't learn from Tae Kim was very minor; Google was very useful for adressing small questions like that. Tae Kim probably covers over 95% of the grammar in ゼロの使い魔 but the rest isn't all too complex. The only thing that might trip you up is the specific vocabulary. Words like 呪文, 火炎and the like aren't likely to appear in the core decks...

I compiled a vocab deck for Book 1 only recording words I didn't know. I have completed core 2000 and have some other experience in reading Japanese (only very basic tho.) There were only 1123 words I didn't know in the book, so if you have only finished Core 2000 the number of unknown words would probably be quite similar for you (maybe closer to 1500). I can't speak for the Core 6000 vocab, but I assume your number of unknown words will be quite lower. Overall, I think 1000-1500 words isn't a bad number for a book this size.

I think it would be a smart idea to definitely learn vocabulary by some form of Kanji grouping. It enables you to remember which readings are used when, and at the same time the meanings actually get to make sense to you. They don't seem so arbitrary after a certain point. Of course, some Kanji like 尺 will be harder for you to find a variety of available words, but you can still attempt to. I wouldn't be troubled by not knowing a word or so. Even as natives we stumble upon words in reading that we've never seen before. I still learn at least 2~5 words a week in English because of my history course.

What I like to do, which I don't condone for everyone to try because it's certainly not for some people, is to go to an online in-Japanese dictionary and look up random words and type out their definitions for hours.

The reason I do this is because since I have such limited changes to speak Japanese, although my chances of quadrupled since coming to college, I get to see thousands of words and tons of examples. The words then don't become completely in isolation because you see them explained and how they're supposed to be used with more Japanese words.

Of course, at a certain level this would only cause the student to go in circles, and an English-Japanese dictionary would have to be used, but it's definitely something to think about later on.

I would also try looking up less common readings too. I try to read about 100 pages of Japanese literature a week, and that is how I continue to add tons of new information to my mind. I especially like the spellings of words in early 20th century works. I particularly like Matsumoto Seichou's book "黒い樹海”。Although I think it was published in the 70s, the guy was old by then, so it still had an interesting feel to it.

Say you look up a Kanji like 蛇
You should focus on words with its ON reading and KUN reading, or any irregularity that may exist.

If you freak out about a certain character, by all means go crazy on research.

One of my first books I got in Japanese was a Kanji dictionary that gave tons of words for all the elementary Kanji. I thought that was really helpful catapulting my vocab. Likewise, using something like jisho.org to compile great lists is good too. Even if you just skimmed like 1000 entries, you're still going to leave knowing something you didn't beforehand.

That's a very interesting approach. I think I'm going to have to give that a try. It could give me something to do when I'm bored in class . Reading a bunch of entries with the same kanji would probably cement those readings and build up a basic meaning of that character in compounds. It would probably be extremely helpful when inferring new compound meanings and readings; a skill I could really use as I don't really want to be tethered to rikai... Do use a monolingual dictionary for this? If so, did you have to import it from Japan? I have been looking for a hard copy of a dictionary but don't really know how to get one here in the US...

http://dictionary.goo.ne.jp

This is amazing of course. Yahoo has one too, and there are plenty of other in-Japanese online dictionaries.

There's 楽引き辞典　For the DS, which I love going through for hours at a time. There's just so much to learn from these sources.　

HououinKyouma -
When you come across a new word, what do you put on the flash card? A sentence mined from jisho.org to put in context? The sentence you found it in? Japanese -> English or English -> Japanese?
Could you upload the vocab deck for the book as a "Shared Deck"? It would be interesting to learn 95% of the vocab in a book and read through it similar to a native speaker would. I could even search for each word in 6k and tag each occurrence (although doing 1000+ would be time consuming) so that I could set them higher priority to follow core2k.

http://en.wikipedia.org/wiki/Chunking_%28psychology%29

@Invidious
I wouldn't mind uploading the deck, but it's Anki 2. I can export it to a text file if you would like to use it with Anki 1 (tell me if there's a better way.) I should say that the words in that deck are the one's I personally didn't know when reading, but my prior vocab background is only core 2000 and some other random reading (nothing to the extent of a novel or anything like that.)

For the deck I made with the first book, I did not add any sentences. Although some people suggest that context helps memorization, I tend to like the just word/defintion cards. I already have some association from reading them the first time and its probable that important one will appear as I keep reading the series. My card format involves two templates, one with kanji on the front, the other is listening first, then the term on the back. I used to do production, but synonyms got the best of me, haha. Besides, my short term goal is reading comprehension, so I want that to fall into place more than anything. I've been at it for more than 30 days and I have to say, its worlds easier now and I am enjoying the second book with much less lookups.

HououinKyouma Wrote:@Invidious
I wouldn't mind uploading the deck, but it's Anki 2. I can export it to a text file if you would like to use it with Anki 1 (tell me if there's a better way.) I should say that the words in that deck are the one's I personally didn't know when reading, but my prior vocab background is only core 2000 and some other random reading (nothing to the extent of a novel or anything like that.)

For the deck I made with the first book, I did not add any sentences. Although some people suggest that context helps memorization, I tend to like the just word/defintion cards. I already have some association from reading them the first time and its probable that important one will appear as I keep reading the series. My card format involves two templates, one with kanji on the front, the other is listening first, then the term on the back. I used to do production, but synonyms got the best of me, haha. Besides, my short term goal is reading comprehension, so I want that to fall into place more than anything. I've been at it for more than 30 days and I have to say, its worlds easier now and I am enjoying the second book with much less lookups.

That would be awesome if you could upload the text file exported from the facts from that deck. I could cross reference it with Core 6000 and see how many words are outside that deck. I'll can also manipulate it in a spreadsheet and import it into one of my decks. Many thanks in advance if you could do this for us.

I would be more than happy to, if only I knew how to upload it! I am not to familiar with the process. I have the text file but I need instructions from there.

Btw, I the deck contains words from the second book as well (maybe a fourth of it or so.) I don't know if its a problem or not. Anyway, its 1500 words in all but there are repeats. Anki 2 doesn't have an easy way to fix those.