vosmiura wrote:

Also as someone pointed out before, the frequency list is almost the inverse of the importance list.  If you learn 3000 words maybe you can (hypothetically) understand 90% of words, but you miss the 10% that are the most important details.

I'm living proof of this. With that number I can follow basic conversations. Very, very, very, very, very basic conversations.

Nukemarine wrote:

So should I:

You should think less and do more.

Srsly, some people here think way too much.

Evil_Dragon wrote:

Nukemarine wrote:

So should I:

You should think less and do more.

Srsly, some people here think way too much.

I am doing. I'm just posting what I did and what I'm doing now and the current results on literacy and fluency. Good call on the thinking too much charge, though. I'm guilty of that.

But if you fail to plan, you're planning to fail. People come in here with different plans, but if they know the outcome ahead of time, there's less chance for disappointment or disenchantment.

If you learn just 1110 kanji in KO2001, you're going to run into a lot of unknown Kanji (on AVERAGE about 25% of the time. If you 2042 of RTK, you'll run into unknown's less (about 10% of the time) at the cost of 100 extra study hours. If you did the 3007 of RTK 1 and 3, that's down to 3% for cost of another 100 extra hours.

Similar ideas go with Vocabulary and Grammar.

So when you finally move on to the sentence method or variants, don't be surprised that the less you bring to the table the more you have to work. It's all a trade off based on how much of each area you want to work.

Codexus wrote:

Worrying too much about my listening skills at this point seems really useless to me, I just don't have the basic vocabulary knowledge required. Guessing and filling in the blanks is not understanding.

Listening will be the most used skill, it should be the strongest. I think one should worry about it early, even with the little skill assets one has to work with. Honestly, I thought AJATT's advice to listen to Japanese would work for my listening skill as I worked on the basics. It just didn't do it.

Now I'm of the mindset one should build a collection of close to 100% comprehended material (so you're not guessing and filling in the blanks here). It should be varied to cover large range of accents and speaking patterns ie one movie or about 60 to 120 minutes from a tv series. It can be built over time, and listened to in small, random chunks of 3 to 5 minutes on a consistent basis on your free time.

If you learn just 1110 kanji in KO2001, you're going to run into a lot of unknown Kanji (on AVERAGE about 25% of the time. If you 2042 of RTK, you'll run into unknown's less (about 10% of the time) at the cost of 100 extra study hours. If you did the 3007 of RTK 1 and 3, that's down to 3% for cost of another 100 extra hours.

I don't know where you got those % from, but I don't think it's efficient to do RTK3 unless you're already at an advanced level.  It would be very unbalanced to learn all kanji up to Kanken Semi-1 Level while the rest of your Japanese is still lower intermediate.

Sure, you're going to encounter some kanji outside of 2042 from time to time - so either: a) skip them: there are other bigger fish to fry, or b) spend a minute or two to learn them as you come across them.

Anyway at lower intermediate level you don't really need 2042 (as a priority).  I did the 2042, but in terms of actual usage I'm still around 1300 and that number is increasing very slowly.

http://www.tidraso.co.uk/kanji_frequency.html

The above is just newspaper frequencies, but it shows 93% coverage with 1000 kanji.  It also shows that the 2000th kanji appears about 11 times less frequently as the 1000th, and the 3000th kanji appears about 17 times less frequently than the 2000th.

Similar ideas go with Vocabulary and Grammar.

Yes, and so I find it quite lop sided to think that you DO need 2500 kanji, but you DON'T need more than 3000 words.  The former covers kanji past the advanced level, while the latter is covering vocab below intermediate level.

The way you're counting 100 hours for 1000 kanji or 1000 vocab, then I think the best is for every 100 hours of kanji to do 400~500 hours active study of vocab & grammar. 

But anyway, I think your new method sounds good.  I'm just trying to make sense of why your past effort was less effective - and it sounds to me like lots of kanji + lots and lots of listening + little vocab is not so good.

You said you can read well - is that because you know a lot more vocab but you just don't count it?  Do you guess the meanings of words from the kanji?  Or do you have to look up a lot of words as you're reading?

Last edited by vosmiura (2009 August 06, 3:32 pm)

vosmiura wrote:

it sounds to me like lots of kanji + lots and lots of listening + little vocab is not so good.

I'd agree with this, based on my experience.

Since starting KO2001 (up to 650 kanji) I've noticed a big jump in my understanding.
Before that I had a lot of trouble understanding what people were saying just because of the number of words I didn't understand, even though the speed and grammar weren't such an issue.


I haven't tried Nukemarine's method, but I did start using the "時をかける少女" subs2srs deck, and watching the movie again last week I understood much more clearly - before I could follow what was going on but must have been filling in a lot of blanks.

I've also spent time looping Tiger & Dragon, and find that I gradually pick up more and more, but there's still large sections that are too hard.
I think getting Japanese subs for that would help a lot.

bandwidthjunkie wrote:

@Nukemarine <-- It sounds like your nuts and bolts Japanese is pretty good. Do you think maybe that you have got to the stage that you would benefit more from actively conversing in Japanese than concentrating on using an SRS and passive input? I appreciate that being in the military it might be hard to find Japanese people to talk to, but you might be able to do a language exchange with someone via MSN messenger, or something like that.

Since there's about 200 Japanese military on my base in Djibouti, it's not hard to find Japanese people to talk to. Being the most experienced in Japanese on the base makes it very easy to strike up conversations. However, it's been my experience this does not help my Japanese though.  Talking to Japanese will be like reading a manga or watching dramas. It'll be for leisure and pleasure, but not for learning even though I may learn something.

Last edited by Nukemarine (2009 August 06, 8:02 pm)

I understood 90% of newspapers after doing my modified AJATT for just one month! So listen to me: Just watch reruns of Bleach! You can trust me, I'm a person on the internet making claims of rapid progress as proof that you should follow my tips.

nest0r wrote:

I understood 90% of newspapers after doing my modified AJATT for just one month! So listen to me: Just watch reruns of Bleach! You can trust me, I'm a person on the internet making claims of rapid progress as proof that you should follow my tips.

So that's sarcasm which implies the guy above you is lying? O_o

Why?

My take on AJATT is this:

"It probably isn't going to work for most second language learners."

Why do I think this?  Because Khatzumoto is not a second language learner.  Japanese was at least his fourth language, so he is probably pretty good at picking up new languages by the time he started studying it.  This probably gave him a great deal of comfort picking out the comprehensible parts of input in an otherwise unknown language.  Us previously monolingual types are going to find his method very tough going IMHO.

That said he makes some very good suggestions, and I love reading his stuff.  It's great just as long as you just don't automatically assume it will be easily applicable to your studies.  Just tailor everything to fit and throw out anything that isn't working for you. 

my 2 cents,

C.J.

btw: the khatzumoto != second language learner is derived from where he mentions his native languages in "The African Way of Learning…Just Do It".

Last edited by cjon256 (2009 August 12, 12:58 am)

I don't know how much of a factor knowing other languages is, but I think that not everyone has the time to do AJATT.  If you have an occupation that requires lots of concentration, a family to spend time with, other hobbies, friends, etc. AJFTFHAD "All Japanese for two freaking Hours a day (tm)" is hard enough to do sometimes.

Last edited by vosmiura (2009 August 12, 1:42 am)

mafried wrote:

cjon256 wrote:

Because Khatzumoto is not a second language learner.  Japanese was at least his fourth language...

Where'd you get that from?

http://www.alljapaneseallthetime.com/bl … just-do-it

Khatzumoto wrote:

Dude, I am a native speaker of Dholuo ...

I am also a native speaker of Swahili...

He had also been learning Chinese before starting Japanese, although I don't think he was anywhere near fluent.

James wrote:

Meh, sounds like you're just making excuses. AJATT will work eventually; might take different amounts of time depending on the person, but it is a valid and efficient method of learning and it will lead to fluency.

Oh.  Well, I guess I was misinformed.  I didn't realize it was "valid and efficient."

End of discussion as far as I'm concerned then...

C.J.

cjon256 wrote:

Again, just my opinion.  Thanks for the lack of ad hominem.

C.J.

I'm not sure if you're being sarcastic or not...? Regardless, I didn't mean for it to be.

I don't think *any* method is going to be successful for all, or most, language learners. We're all individuals after all. But I think it is an exaggeration to claim that AJATT isn't going to work for "most" learners because, anecdotal or not, it doesn't take too much digging to uncover a number of genuine success stories.