I also want to point out that I believe that Khatzu's prioritization of input over output to be deeply flawed.  He claims that people shouldn't try to speak Japanese until they feel ready.  This is one of the biggest mistakes a new learner could possibly make.  I know because I made the same mistake.  If you consider how a baby learns a new language, you will notice that they try to imitate their parents as soon as possible.  They learn by making mistakes constantly, and being corrected when wrong, and receiving positive reenforcement when right.  When Khatzu speaks against classrooms, he is recommending that students get rid of one of their best chances to replicate this early learning process besides actually living in Japan.  A good college course features a native speaking instructor and encourages dialogue in Japanese early on and often.  As long as you remain conscious about how you speak and care about improving, you don't need to worry about these mistakes becoming ingrained.  People who speak in a thick accent don't do this because they tried to speak to early.  It is much more a problem of perception and lack of proper  accent practice.  This is not so hard to correct, and I have helped a few international students improve their English accents in a matter of weeks even after they had spoken improperly for years.  One of the best ways to impede fluency is to abstain from speaking.

Dragg wrote:

He claims that people shouldn't try to speak Japanese until they feel ready.

I don't know about Khatzumoto, but I know that Antimoon emphasizes that you shouldn't try to produce a language until you know for sure that what you're trying to say is correct.  I believe it's even encouraged to mimic and imitate material from native speakers.    The problem comes when you have somebody who isn't being exposed to the spoken language who's trying to construct their own sentences based on grammar rules.

Mcjon01 wrote:

Fluent is a tricky word, it can mean lots of things to lots of people, and it's great advertising.  All we know for sure if we take what he says at face value, is that at the end of the 18 months he was competent enough to read manuals dealing with a field he was already familiar with, conduct a job interview, and get a job with a Japanese company.  That could certainly count as fluent, since it's enough to get by on, but it doesn't necessarily imply near-native or anything like that.  Keep in mind though, it's been three years since then.  I doubt he's just stopped getting better at Japanese in the interim.

Yes. Perhaps where we differ is that I am putting emphasis on the -maybe not native level- part of the description. Now it seems to me that if you see yourself as maybe not native level then you must see yourself as pretty damn close, otherwise there would be no *maybe* about it. But sure *maybe* I'm being pedantic there.

Perhaps one of the regular commentators at AJATT could politely ask Khatzumoto if he wouldn't mind outlining in greater detail (as far as his memory allows) what he was and wasn't capable of after that 18 month period. That might have the dual benefit of relieving me of the anxiety that he might actually be James Rudi and provide a clear idea of what is possible over a period of 18 months employing the AJATT method. Of course others might accomplish more or less, and it'd be nice if they would also similarly report after their first 18 months.

Last edited by bananarama (2008 April 03, 1:54 am)

Dragg wrote:

I also want to point out that I believe that Khatzu's prioritization of input over output to be deeply flawed.  He claims that people shouldn't try to speak Japanese until they feel ready.  This is one of the biggest mistakes a new learner could possibly make.  I know because I made the same mistake.  If you consider how a baby learns a new language, you will notice that they try to imitate their parents as soon as possible.  They learn by making mistakes constantly, and being corrected when wrong, and receiving positive reenforcement when right.

Ah, but babies don't try and imitate their parents from the day they're born.  It's normally about a year before babies start to make proper attempts at speaking.  Do you think babies are just sitting around doing nothing for that year.  No, their brains are frantically trying to make sense of what's going on around them.  Only when they understand most of what people are saying will they feel comfortable to start speaking.

This is known as the silent period and applies to second language learning as much as it does to our first language.  Stephen Krashen seems to be one of the original proponants for the silent period.  He has some articles here http://www.sdkrashen.com/

James Asher also believes in the principle of understanding first, production (naturally) afterwards.  He started the Total Physical Response movement.  He also has some articles online at http://www.tpr-world.com/

To be honest, I believe if Khatzumoto showed up at that guy's house and re-inacted the closing scenes from "Tiger and Dragon" to prove fluency, Dragg will bring up some other complaint.

Dragg, you're right, AJATT doesn't work. Go about learning Japanese via your method.

That said, it would be interesting if he posted about 4 or 5 three minute audio blogs starring himself, perhaps just answering various questions people have written in.

As for what others are saying, yep, adapt what ever advice you want. About the only solid "golden rule" seems to be LEARN THE KANJI FIRST. The sentence method is a great suggestion, but that can be superceeded by great instruction and practice should that be your thing. Heck, even that sentence method is a work in progress. Khatzu's posted some interesting updates. The best being going Kana sentence to Kanji sentence reviews (it's encouraged me to stress doing Kanji to Keyword reviewing on top of Keyword to Kanji). It's due to this that somehow a Kana to Kanji sentence is added to Trinity (as I really want to use Trinity).

Just personally, I'm reading Japanese. I can actually sing a Karaoke song not from memory but for the kanji and furigana on the screen. I can read basic sentences that shocks the hell out of people. Yeah, I'm equivalent of a semester in college as far as vocabulary goes but I'm way advanced in other areas. What ever it is, it's working. Yes, this is due to RevTK and AJATT and Anki.

Dragg wrote:

If you consider how a baby learns a new language, you will notice that they try to imitate their parents as soon as possible.  They learn by making mistakes constantly, and being corrected when wrong, and receiving positive reenforcement when right.

I think we have to be careful drawing conclusions first about how babies learn and secondly how that relates to us as adults. My niece didn't speak until much later than most kids but when she finally started she spoke at a very advanced level and was able to read at the 3rd grade level before entering kindergarten. This to me, intuitively, is evidence that the listening/quiet period is very important for early childhood language learning. Also, parents usually speak in special way to their children and I read somewhere that this is a very important part of acquisition. I'm not sure if there is evidence that the correcting we do is as important as the examples we provide.

In any case, we are not babies and our brains are fully developed (I hope!). What matters is how do we learn best. I like reading and listening so I think using sentences in an SRS is best for me if I add the audio. It takes more time to make up the cards but I use an SRS that allows me to do this (Vtrain) fairly conveniently and I am happy with my progress, although slow, it merits the time I invest.

So far, I am just adding sentences and words from some simple dialogs I have digital audio for (DK 15 minute Japanese, Genki vol 1 and living language Ultimate Japanese level 1). I have all 3 Pimsleurs but I have to admit it was too boring for me to continue past about lesson 11 of level 1 and I found it hard without a transcription of the audio. I feel that pimsleur is a bit of a gimmick in that the decay time and repetition cycle according the actual Pimsleur article he wrote himself is not a fixed variable for all people. Therefore, an SRS is a better way of reviewing according to his principle in that it automatically tailors itself (sort of) based on your either remembering or forgetting. But I'm not here to bash anything that works for someone else. I just think that although there may be some universal principles that apply to most people, I think that individual learning style is most important.

I think that Khatz sounds a little geeky (and maybe a lot of RTK members are too) in that he can be so obsessed with the 24/7 thing and he's into computers and thousands of sentences and has his own web page etc. However, I do love his enthusiasm and ideas and I agree that he has a lot of good stuff on his website. For me though, my attention span is better if there is a good story to follow or an interesting real life person speaking to me face to face. I lean heavily toward the feelings side, big picture, right brain dominant. I tried supermemo with thousands of sentences in Spanish and German before I ever heard of AJATT or those Antimoon guys but I would become bored and go into a kind of autopilot trance mode when reviewing. Unfortunately, I can only take sentence reviews in limited doses. I probably get as much out of making the cards as I do reviewing them! The main thing for me is to keep it fun.

Meanwhile, I continue to trudge the road to happy RTK nirvana where some day I might know the Kanji and I can move on to more reading!

P.S. This is a very interesting thread to me and I am enjoying the posts very much. Thanks to all.

In my opinion, the most important thing to remember is the uniqueness of the individual.  YMMV.  His mileage may vary from yours/ours.  I think he pretty much did what he said he did.  I think it probably did work for him.  He believes that it would work for anybody.  He could be wrong about that, but I think it probably did work for him.  For me, though, the only right way to do it is the way that works for me.  That doesn't make him a liar or a cheat or anything.  It just makes him Khatz.

It strikes me that a person could also use the silent period as an excuse for intellectual laziness.  You could sit and listen to Japanese noise for ten hours straight, have a nice relaxing time while learning next to nothing, or you could spend a good hour of proper speaking practice to make the connections your mind really needs to learn.  Newborns have an excuse for not speaking because their brains and bodies are still developing.  However, this period is by no means silent.  They communicate constantly through crying, gestures, and various noises.  Babies have a right to make excuses; adults don't.  :p

Last edited by Dragg (2008 April 03, 3:32 pm)

Forgot to include this in my last post as I was in a hurry.

Dragg you seem to be misunderstanding what AJATT is trying to communicate. Khatz doesn't say you shouldn't speak Japanese, but that you shouldn't go about creating Japanese you don't know how to say. It might've been Khatz or someone else, but I remember hearing, "if you have to think "how do I say this in Japanese" and can't produce it quickly, you're probably going to be wrong". I think this is mostly true.

AJATT doesn't at all encourage maintaining complete silence while doing Japanese. Anyone is plain wrong if he or she thinks this. This may be my own interpretation, but I don't think it also doesn't say you can't make your own Japanese sentences, but rather that just ones you don't know how to say should be off bounds.

For example:
Obviously, if you're beginning and you learned ピザを食べた, I think you'd be safe to say すしを食べた. You can go on all day making sentences about things you ate until you are content. That same beginner though, shouldn't try to say "the pizza I ate was cheap" if they don't know how. Let's say they haven't learned about relative clauses, so they try: ピザを食べたは安かった. This is incorrect. Time would've been better spent actually seeing the correct form first and trying to mimic it.

vosmiura wrote:

Regarding the 7,500 sentence thing... with language learning there comes a point where you don't need to worry about consciously learning and cramming stuff into an SRS.

It would not surprise me if learning 7,500 sentences, each covering at least one new word, grammar or idiom, would get you pretty close to that zone.

Plus it's not just about what's in the sentences in your SRS.  The daily exposure part, and incorporating Japanese into your life as much as you can, is an even more important part than cramming sentences.

You won't get as as fluent by just doing 7,500 sentences, without watching lots of Japanese TV, reading Japanese, listening to Japanese music, etc.

Please don't get me wrong, I'm very aware that there is more to the AJATT method than just the sentence thing.  And I do agree that learning several thousand sentences will tend to give you the ability to know, infer, inteligently guess things that you haven't been directly exposed to in the sentences in your SRS. But that isn't going to be the case for the 2 points I raised above; unusual kanji readings and the readings of proper nouns. Learning those thousands of sentences will make you familliar with certain patterns in the way kanji are used that will allow you to accurately read words you have never seen and even write words you hear but have never read.

But unuusual readings and proper nouns don't fit these patterns so you can't simply use your general body of knowledge to deal with them. Further, becasue they don't fit the patterns, they are harder to remember. You can't learn them without being exposed to them and once exposed to them how do you remember them if you don't stick them in your SRS? Khatzumoto says he is still using the SRS method for Japanese, so he clearly hasn't, even now, reached some point where he can reliably learn / remember without it. It just doesn't make sense that the SRS be essential to remmeber stuff that's realtively easy to rememeber, but unnecessary for the harder stuff.

So, after 18 months and 7500 sentences is it possible that these readings will be in your body of knowledge? I don't think so.

Now if Khatzumoto isn't really claiming that he was close to native-level after those 18 months then I have no problem and I'm waffling at length over nothing. If he is claiming such a feat, well I'd like to hear some further explanation of how he did it. So I think it would be useful if he would clarify what he could and couldn't do after those 18 months.

In anycase, one of Khatzumoto's maxims that I agree with is that it's futile to argue in English about Japanese on the internet. Sadly I sometimes weaken as you can see.

I believe in the value of audio input from my own experience in learning english.

I worked two years in UK as a software engineer and before that I never spoke English to anybody except at secondary school (I never completed a university level degree or even graduate). Needless to say at my job I had to communicate in english on a daily basis. It was hard at times to follow pub talk. Only after a few months there and watching UK TV I started to understand what they were talking about : comedians, TV programs, local foods, etc.

I think I was speaking relatively fluent, as I had no difficulty in expressing my thoughts. Everyday conversation was really easy. More technical conversation such as enquiring about places to rent was difficult early on because of words I didn't know, not because of grammar.

Before that my exposure to English was mostly reading programs and games on the computer, especially text adventure games. Interestingly for the text adventure games, such as the classic "Guild of Thieves" on Amiga, I had to produce sentences. Those games went way beyond "go north" or "take key". You could type full sentences like "say to thief 'do you want to play cards?'" and the game would know you wanted to speak to one of the characters and would give you meaningful replies nearly every time. Being new to programming I was fascinated by it and tried all sorts of trick questions to see where the limits were in the game's understanding of what I typed. This could count as "production" even though I didn't speak aloud, let alone with a real person!

Audio-wise I listened to BBC almost daily. Not all day long but probably an hour a day or more. I think I liked the feeling of being somewhere else, I liked most talk programs. On the BBC they have many programs where they talk with people all around the world. And yes, that was background noise very often. I certainly made efforts to understand what was said, but often times I would just leave it on in the background as I was programming.

I'm convinced listening regularly is a big help to learning a language.

In fact there is a small percentage of vocabulary that accounts for a large percentage of the spoken language. Given that it's not surprising that even if you feel like you don't understand anything to a Japanese audio program, if you listen to it several times, you start picking up words that you didn't recognize before, words that are already in your basic vocabulary.

I've found this quite interesting. This idea that of all the Japanese you hear and you think is beyond your grasp, in fact a large part of it is simple words you know but you are not familiar with the grammar yet, and your brain is simply not trained yet to understand the new language.

It's really easy to demonstrate to yourself if you listen to Japanese audio, no matter what level it is. Say you know 100 words and the program contains the 100 words spoken at one time or other. Invariably the first time you listen you won't pick them all up, you will pick up maybe a quarter of those or even less. But if you listen again you get used to the sounds and rhythm and you will pick up more.

I think audio alone isn't efficient, and reading alone either, but the two together is very productive. From my experience with english I believe the producing part is secondary to both reading and listening.

I think this is valid for Japanese too. However with Japanese reading and writing is such a headache that the reading can easily lag behind listening, so listening is not as effective

rich_f wrote:

If you don't like the method, don't use it.

AND THOU SHALL NOT SPEAK OF IT!

rich_f wrote:

If you have issues with Khatzumoto, then take them up with him.

Hasn't he made it clear he's not going to get into arguments of that nature? Actually. it's a stand I respect.

rich_f wrote:

If you're worried about your friends being losers, then get new friends. A language site can't really help you with that one.

Oh I don't know, you could meet new friends on a language site.

Dragg:

I think it's pretty much pointless for you to continue posting in this topic. It has become quite clear that no one here will be able to change your mind. Furthermore, your arguments are starting to become nonsensical. So, because you start saying, "The best way to learn Japanese is to attend a German class!", I think you should just give up.

That being said, in your post replying to mine before, you said you should "be creative." Well, that's all good and fine, but you must, I repeat you MUST, be doing this with a native speaker. You NEED someone to be there to correct you. Without that, "being creative" is time poorly spent. It would be much wiser to spend that time learning how to say it correctly, by means of examples. Oh wait, look... that's exactly what Khatzumoto is trying to say!