Okay, I'm going to share a little bit of my experiences doing production vs. recognition.

Firstly, after finishing RTK I went straight into mining with a small vocabulary already. It took a long time to get 30 sentences to put into my SRS. Reviews for this deck are production of the word written in Kanji from seeing it in the context of the sentence where it is written in kana.

This type of production takes quite a while to review. I'm going to say roughly twice as long as pure recognition.

Then I moved on to doing the KO deck with full audio. Recognition only. So far I have found that a large portion of words I can remember how to write without having to do production. Words which only use a single Kanji are generally easy to remember. Words with two are medium - hard. Words with 3 - 4 get quite hard depending on what the word is.

What i've found is that after you know the meanings the more familiar you get with the readings, it becomes even easier to remember how to write words without drilling production.

For instance, we all probably learned 高 as 高い the first time we came across it. But you soon encounter it numerous times in other words as こう and it becomes far easier to remember in the compound as こう the more words you know that use it that way.

最高
高校生
高速道路
高級

Until finally, even if you have never encountered the word, you can begin to make assumptions based on your knowledge.

For example, I knew the word for low-pressure so I guess I just replace low with high こう and there we have it. High pressure. 高圧.

Now the point behind my post is this...

How much more efficient do you think it is based on an educated guess and some of your experiences to learn to read first (after completing RTK!) and then once you have a full working knowledge of reading, move on to drilling production of words which you can't easily remember how to write.

I estimate you'd cut your production reviews stage in half. At least. You'd also save yourself a whole lot of difficult reviews at a time when TBH you're probably not truly ready for them.

I feel that after learning how to read to JLPT1 level you will have such a good background knowledge of readings that will aide as a second mental hook in doing production writing.

RTK + knowing readings will be enough to write A LOT from memory but not everything.
But why drill how to write EVERYTHING from the beginning when it's probably not necessary and less efficient?. I know this is why RTK2 exists but lets not go there. However, what I'm saying is exactly in line with the philosophy of breaking things down in to layers and doing one layer at a time. I can see why now... It just makes things so much easier.

I definitely agree with the proposition that strong reading/recognition ability makes the transition to production much easier. Its part of the reason that I only plan to start doing production (writing) cards in ~6 months time, if my production hasn't improved (as much as I would like) by itself at that point.

My experience has been:

For some words- the ones I get a *lot* of exposure to- I can produce them without practicing them at all. I find this is only a very small subset of words, though. For most words I find that when I want to be able to write it, its basically just a matter of actively analysing what kanji it is made up of, with the intent of reproducing it. Its kind of like flicking a switch from "recognition" to "production"; I decide I want to write the word, I actively analyse it, then I can pretty much just start writing it from that point on. I liken it to flicking a switch because it really doesn't feel like learning something new.

yeah it definitely helps if you analyze which kanji make it up. Especially if it's logical. Not so much if it's random keywords. Those are really the ones that may require a bit of drilling.

And what do you guys think about "listening"? I have a deck where the cards are recognition and listening, and listening takes a lot of time to review. Do you think it's a waste of time to do?

I find that even random keywords/ seemingly random kanji don't bother me all that much. Its more the fact that I when I learn to read a word I don't closely look at the details of the kanji that make it up. I rely more on the shape of the word itself (and context too, I guess). This is particularly the case for characters with a common element that share the same reading. Like　生　性 etc. I won't necessarily take notice of which one it uses when I'm just reading, I'll just remember the meaning and the fact that it uses a kanji with reading of せい.

Asakk Wrote:And what do you guys think about "listening"? I have a deck where the cards are recognition and listening, and listening takes a lot of time to review. Do you think it's a waste of time to do?

On the contrary, my KO deck has full audio. Half of it is listening, the other half reading recognition.

It rocks.

I find it pretty fast to review listening cards. Reviews get easier with each subsequent review and especially with listening reviews I find myself going "hey that was an easy one!" at something which a week or two ago I wouldn't have gotten a word. WELL worth it.

Also, yeah the shape of the word as a whole is really how we read in any language. Works just as well with Kanji. I definitely find that words that have become very familiar can just be read at a glance based on the shape of the word as a whole. It's quite cool. Context too is very important... Just before I was reading a news article about an earthquake emergency warning system (or something of the nature) and I figured out from context that 緊急 meant emergency.

Hehe just now when I was bringing the word to mind I was able to write it from memory!! Having seen it only once and guessed its meaning!!! Why? Cos both those Kanji are familiar territory to me.

I also think having listening cards is great. When I started using them, I realized that my listening level wasn't so great. Now that I've been using them for about 2 months, it has improved dramatically.

For listening cards, do you just write it out? Or do you just listen, and pass it if you understand? And would listening cards help stuff that you haven't heard before? Basically I'm thinking of adding all listening cards to my deck from KO2001 to prepare for JLPT3 listening section (since my listening skills are atrocious).

Nii87 Wrote:For listening cards, do you just write it out? Or do you just listen, and pass it if you understand? And would listening cards help stuff that you haven't heard before? Basically I'm thinking of adding all listening cards to my deck from KO2001 to prepare for JLPT3 listening section (since my listening skills are atrocious).

Writing it out would take forever and a day. Grade based on understanding. If you understand the whole thing then great. Sometimes if theres a tricky word or two in the card that I don't get just yet but I get the rest of the sentence I might grade it hard and let it come back up in the next few days and keep grading it hard. I've found that the word I was having trouble with slowly but surely just becomes familiar enough to consider it mature when the card is about at that stage.

For the KO deck with audio just listen and comprehend. Then when you click show answer, try and read along at the same speed as the narrator reads the cards. I'm finding my reading speed to be improving a lot here too.

Also, when you read out loud or repeat what was said you're activating more parts of the brain and starting to form connections necessary for output. It's laying a foundation of sorts.

I personally think both production and audio cards are a waste, if you are beyond the early beginner stage.

Audio simply doesn't train anything and takes time. Hearing what is said in one sentence (which you have been exposed to several times by the SRS) is not even close to the actual problem area of hearing fluent fast native dialogue in a movie etc. It's only good for two things: Going from being able to comprehend almost 0% of japanese audio to being able to hear Japanese sounds. I'm talking the first 100-200 cards for a complete beginner. Second, it may let you get a slightly better grip on the pitch. Again though, I find it will help you a lot more to simply do normal recognition, then train listening when actually listening to real conversations.

Going kana to kanji is a waste because that knowledge is attained automatically if you're exposed to it enough. Like mezup said in the first post, if you know a kanji well, pretty much any compound using it will be a breeze to write after a while. I personally use a few production cards though, I make them for words I've known for over a year and still can't write, usually because it's the only word using a certain kanji, or maybe an ateji. By doing this, I get a bit better at using that kanji. By my experience on this, I'd say that anyone who has over 5% of their cards as production cards are actively wasting their time.

Tobberoth Wrote:I personally think both production and audio cards are a waste, if you are beyond the early beginner stage.

Audio simply doesn't train anything and takes time. Hearing what is said in one sentence (which you have been exposed to several times by the SRS) is not even close to the actual problem area of hearing fluent fast native dialogue in a movie etc. It's only good for two things: Going from being able to comprehend almost 0% of japanese audio to being able to hear Japanese sounds. I'm talking the first 100-200 cards for a complete beginner. Second, it may let you get a slightly better grip on the pitch. Again though, I find it will help you a lot more to simply do normal recognition, then train listening when actually listening to real conversations.

Going kana to kanji is a waste because that knowledge is attained automatically if you're exposed to it enough. Like mezup said in the first post, if you know a kanji well, pretty much any compound using it will be a breeze to write after a while. I personally use a few production cards though, I make them for words I've known for over a year and still can't write, usually because it's the only word using a certain kanji, or maybe an ateji. By doing this, I get a bit better at using that kanji. By my experience on this, I'd say that anyone who has over 5% of their cards as production cards are actively wasting their time.

I agree and disagree at the same time. Yes, it trains you to listen to the sounds, No it's not real native speed natural dialogue. However, it trains you on listening to common sentence patterns and certain things that you aren't likely to hear in every day convo's that you would miss if it was the first time hearing them despite being able to read them well. At least that's what I think.

It's not a total waste of time at all. All aspects must be trained. Audio cards are one way of training listening. Although it would probably be much better to use sub2srs for this purpose.

I used to think I had it all figured out with recognition/production, but as I started to type my confident recommendations, I realized my thoughts have changed. I'll have to wait till I work out precisely how they've changed. I can say though that it's better to ease into production (if here we mean production = dictation: audio/kana/image ---> kanji), and I think that dictation cards' quirks, the particular returns and disadvantages, are not conducive to easy, general comparison with recognition in terms of efforts and returns.

I also feel that if a card's audio is more than 4-5 seconds or it contains more than a cpl new words/concepts, probably shouldn't do production with it unless you've already studied it with recognition. (Splitting lengthy cards is tricky and I don't recommend it, but that's just me being neurotic.)

I agree that it's probably best do only do production cards with stuff mined from subs2srs (or otherwise extracted from native materials). And of course, I totally disagree with Tobbs regarding the value of audio for listening practice in Anki. I admit I seem to be alone in this respect, but I think every card should incorporate as many senses as possible, you think of the language like it's existing in multiple dimensions to be encoded and reviewed at the same time. You can do stuff as just reading or just listening, speaking, etc. outside the multimedia workshop and memory preservation chamber that is the SRS. Unless of course you're using the SRS to correct specific weaknesses or target specific language goals; I just mean for general all purpose cards.

Tobberoth Wrote:I personally think both production and audio cards are a waste, if you are beyond the early beginner stage.

Audio simply doesn't train anything and takes time. Hearing what is said in one sentence (which you have been exposed to several times by the SRS) is not even close to the actual problem area of hearing fluent fast native dialogue in a movie etc. It's only good for two things: Going from being able to comprehend almost 0% of japanese audio to being able to hear Japanese sounds. I'm talking the first 100-200 cards for a complete beginner. Second, it may let you get a slightly better grip on the pitch. Again though, I find it will help you a lot more to simply do normal recognition, then train listening when actually listening to real conversations.

Going kana to kanji is a waste because that knowledge is attained automatically if you're exposed to it enough. Like mezup said in the first post, if you know a kanji well, pretty much any compound using it will be a breeze to write after a while. I personally use a few production cards though, I make them for words I've known for over a year and still can't write, usually because it's the only word using a certain kanji, or maybe an ateji. By doing this, I get a bit better at using that kanji. By my experience on this, I'd say that anyone who has over 5% of their cards as production cards are actively wasting their time.

do you include sampling actual japanese ressources in audio card (and in your criticism of audio card)? or are you just criticizing the TTS technology and eerie misaki voice ?

Tobberoth Wrote:I personally think both production and audio cards are a waste, if you are beyond the early beginner stage.

Audio simply doesn't train anything and takes time. Hearing what is said in one sentence (which you have been exposed to several times by the SRS) is not even close to the actual problem area of hearing fluent fast native dialogue in a movie etc. It's only good for two things: Going from being able to comprehend almost 0% of japanese audio to being able to hear Japanese sounds. I'm talking the first 100-200 cards for a complete beginner. Second, it may let you get a slightly better grip on the pitch. Again though, I find it will help you a lot more to simply do normal recognition, then train listening when actually listening to real conversations.

Going kana to kanji is a waste because that knowledge is attained automatically if you're exposed to it enough. Like mezup said in the first post, if you know a kanji well, pretty much any compound using it will be a breeze to write after a while. I personally use a few production cards though, I make them for words I've known for over a year and still can't write, usually because it's the only word using a certain kanji, or maybe an ateji. By doing this, I get a bit better at using that kanji. By my experience on this, I'd say that anyone who has over 5% of their cards as production cards are actively wasting their time.

Going kana to kanji is not a waste of time.
When you're entering 30 - 100 sentences from natural sources everyday, and have 300+ reviews everyday, it's easy to go over the same card, and get caught on the reading of one kanji time after time.
When I keep getting the reading of a certain Kanji wrong, I flip it and put that particular kanji in kana, forcing myself to write it out everytime, 90% of the time this does the trick and I will remember the reading of the Kanji from having to write it out.

ghinzdra Wrote:do you include sampling actual japanese ressources in audio card (and in your criticism of audio card)? or are you just criticizing the TTS technology and eerie misaki voice ?

Yes. I consider any form of audio a waste of time. The benefits are way too small compared to the effort of ripping audio, the size of the decks, the syncing issues and most of all, the time it takes to review.

thecite Wrote:Going kana to kanji is not a waste of time.
When you're entering 30 - 100 sentences from natural sources everyday, and have 300+ reviews everyday, it's easy to go over the same card, and get caught on the reading of one kanji time after time.
When I keep getting the reading of a certain Kanji wrong, I flip it and put that particular kanji in kana, forcing myself to write it out everytime, 90% of the time this does the trick and I will remember the reading of the Kanji from having to write it out.

If you add 30-100 sentences a day, all the more reason to NOT add production cards. You don't have the time. If you keep getting the reading of a kanji wrong, add more sentences using that word, and add more compounds using that kanji and that reading. That will give you more sentences to work with, more grammar exposure and it will still save you time since writing a sentence out can easily take 5 or even 10 times as long as it takes to simply read it.

I think it's a waste of time to do cards *without* audio--the benefits are so large, and the effort in creating and reviewing them is so minimal, especially since we're continuing to improve our tools and expand our methods.

But to each their own. ;p

And here I was , pondering about exposing my brand new 3 level media strategy.
I guess on this one we really disagree Tobberoth
just a question :what does SRSing mean for you? how do you feel about it ?how much time do you spend on japanese each day , how much time are alloted to SRSing , how much time to live media ?

Tobberoth Wrote:

ghinzdra Wrote:do you include sampling actual japanese ressources in audio card (and in your criticism of audio card)? or are you just criticizing the TTS technology and eerie misaki voice ?

Yes. I consider any form of audio a waste of time. The benefits are way too small compared to the effort of ripping audio, the size of the decks, the syncing issues and most of all, the time it takes to review.

Are you referring to audio/kana-->Kanji cards here, or all audio entirely? I can understand your argument that you don't see audio/kana-->Kanji as time-effective. But I personally find that having audio on the answer side of my card is really helpful for pronunciation. I go from Kanji --> kana/meaning + audio. Answering is just as quick as normal recognition, but if I choose to I can shadow along with the audio on the answer side to improve pronunciation. Or I can choose to quickly move on to the next card if I don't want to spend time shadowing that day.

Blackmacros: I think that really isn't a problem. That's what I do on my deck anyway =)

blackmacros Wrote:I go from Kanji --> kana/meaning + audio.

I was having trouble with some of the readings, so I recently started doing this (I added the audio into my existing KO2001 deck so I didn't have to start again).
I think the audio provides a little extra context. It also reinforces the right pronunciation and pacing of the sentence.

It does take longer to review though.
I can't say for sure that it's worth it, but so far it seems to be helping.

This is sort of why I stopped doing production as well. Too much time vs. too little immediate benefits. I am also finding that the more exposure I get with recognition, the more often production comes naturally. I'm glad to hear that others are finding the same.

ghinzdra Wrote:And here I was , pondering about exposing my brand new 3 level media strategy.
I guess on this one we really disagree Tobberoth
just a question :what does SRSing mean for you? how do you feel about it ?how much time do you spend on japanese each day , how much time are alloted to SRSing , how much time to live media ?

SRS is an augmented memory. You learn something, usually by means of dictionary, and you want to remember it. Before I had an SRS, I often read Japanese stuff and used Rikaichan or just a dictionary, to understand it. When I ran across it a week later, I had to redo it. All SRS really does for me, is make sure I never forget what I learn. It's like I'm adding the dictionary I always needed to my brain, so I don't need it when I'm exposing myself to Japanese sources. While I find there are benefits to being exposed to grammar and sentences from real contexts, I really think that the massive and most important factor is simply to increase vocabulary. SRS let's you easily learn over 20 new words a day. By traditional studies, you'd be happy if you learned 1 or 2. Grammar, you can learn by exposure, by reading books and rules on grammar, conjugation tables. This works, and it's actually quite effective. In 6 months, I learned basic Japanese conversation to fluency using conjugation tables. SRS requires less effort and is a more fun way to learn it, but comparing the effectiveness over time, I don't think it's THAT big of a difference, not compared to the HUGE difference in expansion of vocabulary.

I spend a few hours on Japanese a day, 1 or 2, sometimes more. I allot the time I need to SRS, It has precedence over normal exposure for me. I review all my expired cards every day and I usually add 20 new sentences, sometimes from structured material, sometimes from mining. Normal Japanese exposure however, I don't even allot time to. I do it when I feel like it. Watch an anime, see an episode of london hearts, read a chapter in a book or manga. I'm not doing AJATT, I'm doing Japanese When I Feel Like It. I noticed early that exposure, while vital to language learning, is ineffective. Two hours of Japanese exposure gives me almost just as much effect as 10 hours, so I've removed all of the forced exposure, it just makes Japanese seem like a chore. I find that exposure only reinforces learning anyway. You learn tons of words and grammar, then you see them used and you understand how and why those things are used. If you don't know what you're being exposed to, you learn almost nothing, especially not at my level. My approach is: Use SRS to learn vocab and the meaning of grammar structures. When the stuff has been in your SRS for a while so it starts to feel comfortable, be exposed to it in various media. I do not believe one can turn this approach around. You have to learn the things before you are exposed to them for the exposure to have any effect. I don't think you learn new words from the exposure, you just internalize the words you already know.

Sorry for that really long post But it was a lot of questions.

Tobberoth Wrote:Sorry for that really long post But it was a lot of questions.

Thank you! You've just basically summed up the reason I study from lists, and now that I've completely finished with JLPT and few other books I'm stuck wondering around desperately looking for new lists too study(There's still so much vocab in books I can't understand)

One question: I know this is insane considering I'm 10,000 vocab deep into the language. But I find my biggest problem(espically for adverbs, I can't memorize adverbs! or verbs with no kanji/ strange kun youmi readings) is getting it into my head in the first place. Aka I can't get it into my short term memory! I look at it look away and bam it's out of my mind.

The only thing I've found that work is Stupid Stories　linking it the sounds to words I already know.


What is your technique in getting words(the sounds more then the meanings) into your short term memory?

I don't use any technique to get words into my short term memory. When it comes to SRSing sentences, I just go with rote memorization, and reliance on the readings of the kanji. If a word is especially hard to remember, I add more sentences containing that word, so I get extra exposure.

Tobberoth Wrote:I don't use any technique to get words into my short term memory. When it comes to SRSing sentences, I just go with rote memorization, and reliance on the readings of the kanji. If a word is especially hard to remember, I add more sentences containing that word, so I get extra exposure.

There's a bit of an overemphasis on memorizing kanji words on these forum almost to the point that adverbs and non-kanji vocab don't exist!

Yeah, due to kanji readings, I don't have any trouble with Kanji words, but non-kanji words, do you just repeat the word to yourself a couple of times/ write it out? or do you just have a really good short-term memory?