Jarvik7 Wrote:Personally I think textbooky sentences aren't a problem if you use them as a tool to learn and not as a model for your output. Mining sentences to put into an SRS turns them into a model though. It's the same story for every other flawed input source (girlfriends, anime, manga, etc).

I'm not following... there shouldn't be any output model or there should be an output model that has nothing to do with sentences? What's your output model?

I don't SRS sentences at all. My output is modeled after thousands of conversations with hundreds of people.

Jarvik7 Wrote:I don't SRS sentences at all. My output is modeled after thousands of conversations with hundreds of people.

Maybe I'm mistaken but didn't you take traditional classes for quite a while (hope I'm not confusing you with someone else)? If so then you weren't you essentially doing the same thing as an SRS for a while, just without an SRS? That meaning constantly reviewing text book type sentences over and over again?

If you are advocating some traditional text book learning (which I agreee with), and you LEARNED from seeing those examples then they became your output model. Just because you didn't happen to enter them into a computer program like Anki would't change that fact.

Also remember a lot of us don't have access to Japanese speakers. We're trying to establish a good foundation for when we finally do get to actually start interacting in a Japanese environment.

In a way I actually agree with thermal in theory. If I could grab the most frequently used vocabulary from natural sources as easy as I can take sentences from KO2001 I would in a heart beat, as would anyone else. The problem is I need efficiency. I don't live in Japan, I don't have Japanese friends all around me, I work 40hrs per week, and I'm taking 20hrs of college credits and will continue to do so until the end of this year.

I've tried mining sentences from a complete natural source and it just sucked due to my limited vocabulary. I want to do just that eventually of course, but I didn't have the time to sit there and break down sentences where 90% of the words were unknown or to go around looking for an i+1 sentence that had that same word. When I would do that I would get a sentence that was just as out of context (in terms of a large general context, such as a complete story) as anything from KO2001.

Then there is the problem of trying to determine usefulness. Some people will argue "Well if it is something you are interested in reading then it IS really useful". I don't buy that. If I don't understand what I'm reading UNTIL I learn it then it is impossible for me to know exactly how useful something really is in terms of my ability to use and encounter it elsewhere.

When I do KO I constantly encounter the words in all of my other natural sources of media which tells me one thing.....the word order it teaches is extrememly efficient. And that is what I need, efficiency and lots of it. If I went around reciting KO sentences as conversation practice or really dwelling on them as I read them then yes I could see the concerns people would have, but I don't do that. I read through the sentence as quickly as possible, make sure I get the general jist of sentence, and check myself on the specific vocabulary word I have highlighted. The rest of the day is then spent either listening to Japanese audio, watching Japanese media, or reading Japanese books.

activeaero Wrote:Maybe I'm mistaken but didn't you take traditional classes for quite a while (hope I'm not confusing you with someone else)? If so then you weren't you essentially doing the same thing as an SRS for a while, just without an SRS? That meaning constantly reviewing text book type sentences over and over again?

I did take classes, but thanks to self study I skipped the second semester of 1st year and all of second and third year and went directly into 4th year. I took 4th and 5th year Japanese in Japan, but the classes didn't use a traditional textbook, instead focusing on actual texts, movies, conversations, etc. I then took Japanese linguistics and Classical Japanese back in Canada, as well as a whole host of JP literature, religion, and art courses. I also was an officer of the 日本語会話クラブ and started/ran the 日本人飲み会サークル.

Considering my somewhat unique education, I don't think I did much repetition of textbooky stuff, but did spend a LOT of time interacting in Japanese in a formal environment. I did go through all the textbooks of the classes I didn't have to take in my own time so I am familiar with them and think they are very useful for beginners though.

I understand the point you're trying to make thermal, but it's a point we already understand. No one is advising to only ever study from KO or textbooks. The textbooks are a way to bootstrap yourself for when you do encounter that word in something native. I go through about 300 textbooky sentences a day, but I watch a few hours of Japanese television everyday as well. I play and understand Japanese video games. You can't tell me that it's not working. I'm understanding more as I study more. I'm understanding specific words better as they come up more often. KO is the entry point just like Heisig is the entry point.

You can make a point about how this won't necessarily improve speaking, but speaking is a completely different category. KO, KiC, and 例文で学ぶ are all about literacy, and that's what the JLPT is mostly about. The only way to get better at speaking is to speak. Everyone knows that. You can't fault conversation practice for not teaching you how to read anymore than you can fault reading and kanji practice for not teaching you how to speak.

I think everyone studying KO, etc. already understand this. It's just that you're choosing to make a huge deal out of it.

KO on smart.fm + Japanese TV = Da Bomb!
How can it get any more fun?

It can be easy to make a breakthrough in your ability, and forget about the benefit of what you did earlier.

For example I remember one time, I lacked the ability to understand most speech on doramas. It was too fast to process. I went on a trip to Japan for 2 weeks for business meetings & presentations where I listened to spoken Japanese most of the day, with the same presentations being repeated for different customers. When I came back to US I realized I started to understand lots more speech in doramas - like 10 times more.

Is it because like Antonio Banderas I learned Japanese just by listening to it, and everything I learned before was a waste of time? Far from it. I had been on this kind of trip many times before. Most of what I learned was from KO, and it was just at this time that the pieces came together with a bit of concentration, and my comprehension level leaped.

This seemed like the right place to ask my question, so I'm reviving it.

It seems clear from most of these posts that KO2001 is sufficient for the vocab section of JLPT2.

How significant is the vocabulary jump between 1 and 2? I assume KO2001 is not going to be enough for JLPT1. I plan to sit JLPT 1 in December and the section I am most unsure about how to proceed with is vocabulary.

i thought the vocab in jlpt 1 included the entire jouyou kanji...

Hashiriya Wrote:i thought the vocab in jlpt 1 included the entire jouyou kanji...

I had a pamphlet laying around but can't find it now, but it said something along the lines of

JLPT2: 6000 words vocabulary
JLPT1: 10,000 words vocabulary

JLPT2 covers 1000 kanji
JLPT1 covers 2000 kanji

or something along those lines.

So I know the "technical" jump that occurs. But I'm more interested in what the jump is in reality. KO2001 is only ~3600 words, but as a number of people have noted its enough for JLPT2 because- even though it technically covers 6000 words- if the words that come up are all from the ~4000 that you know it doesn't matter that you're technically 2000 words short of the requirements.

blackmacros Wrote:So I know the "technical" jump that occurs. But I'm more interested in what the jump is in reality. KO2001 is only ~3600 words, but as a number of people have noted its enough for JLPT2 because- even though it technically covers 6000 words- if the words that come up are all from the ~4000 that you know it doesn't matter that you're technically 2000 words short of the requirements.

I am struggling with this myself right now. Having completed KO2001 some time ago and the iknow Core 6000 just recently(only adding words I didn't know), I am now working my way through all the past JLPT1s.

Right now its not pretty, I'm getting around 30-40% right. While I can read about 90% of the test its invariably the 10% I don't know that is actually being tested.

In the JLPT2 to JLPT1 jump from 6000 to 10,000 I think the problem is that you go from what are basically everday words to many somewhat rare words, and its those new 4000 that are being tested. I have about 4200 SRS sentences and estimate that I know about 5000 words and could probably own the JLPT2, but I have yet to find a systematic way to learn the JLPT1 words. If they ever get around to making the Core 10,000 then that would probably go a long way.

I've tried sentence mining books but usually come up with a bunch of words that are not on the JLPT1 (on the online vocab lists.) They are either too technical, old, literary or whatever. I suppose the last resort is to get the JLPT1 vocab list and just start tracking down example sentences for all the ones I don't know, but that seems really boring.

that s what I m doing and indeed it s real pain in the ass...
especially since the biggest list available is the result of the merge of several list so there a lot of to delete .... not to mention all the stuff you already know .
I think creating such a thing would be the next step of collaboration but unfortunately it a very touchy process as everyone must agree on the keywords , then on the sentences (and I can spend a hell of a time before choosing the good sentence....).And you must double check on top of that.... and there s not any logic and you dont have audio except misaki but you t can t even really batch it as you do it one sentence at a time....

and the KO2001 experiment and the other projects have shown it : either you complete those kind of things quickly or it become dead in the water after a couple of weeks.

So it s a bit tricky.

Would you be able to link that list? Right now my thoughts are to use a combination of vocab lists and mining past tests (the rtk share google group has a link to a download of about 20 past tests).

I use an offline epwing dictionary and for most words -as long as they aren't obscure (and therefore don't have example sentences)- I can usually add an example sentence within 10 seconds. Using a vocab list wouldn't be too bad for me.

Mining a lot of past tests is probably the best way to go about it though. If you mine enough tests, it will generate a frequency list of sorts and you'll end up learning the words that are most likely to be used on the test.

actually I suspect that the list has been based on JPLT tests.... because each and every time that I didnt know a word in a test , just by looking into my Anki database I found it in my JPLT deck ....

wait a moment for the list ... I downloaded it 2 or 3 months ago and after putting into my deck I started removing everything useless for me or redundant . So I can t even export it as it s now designed for me . If I m not mistaken I think the first file was about 7000 words big. It must be somewhere on my secund harddrive as I reinstalled my OS last week.

Was it from this thread?

http://forum.koohii.com/showthread.php?tid=937

I got a file from there around 7500 entries long.

I think so .
make a test . Take a word you dont know in one ofthe test.... and then look for it in the spreesheat. Until now I think I never found a word that was in a test and not in the list.

I quickly searched for 2 or 3 words and it found them all. Excellent.

On a side note I used the 2007 1kyuu, and I didn't do too bad on the first couple of questions. I didn't know a few words, but was (mostly) able to make an educated guess at the correct readings.

as I said i m pretty sure it is based on the test . Heavily.
Because there are some words that I barely found elsewhere to the extent of having difficulty to find sentence exemple in ALC dictionary like 清濁 while there are pretty common words that I mined from anime, news ,etc.... and didn t found in the list like 阻止　or 視線

blackmacros Wrote:

Hashiriya Wrote:i thought the vocab in jlpt 1 included the entire jouyou kanji...

I had a pamphlet laying around but can't find it now, but it said something along the lines of

JLPT2: 6000 words vocabulary
JLPT1: 10,000 words vocabulary

JLPT2 covers 1000 kanji
JLPT1 covers 2000 kanji

or something along those lines.

So I know the "technical" jump that occurs. But I'm more interested in what the jump is in reality. KO2001 is only ~3600 words, but as a number of people have noted its enough for JLPT2 because- even though it technically covers 6000 words- if the words that come up are all from the ~4000 that you know it doesn't matter that you're technically 2000 words short of the requirements.

I don't really know if that's true, that depends on the JLPT2 test. Remember, it's 6000 official words, the test makers are allowed to take 20% of the words on the test to be OUTSIDE of that range. So not even 6000 words is enough.

Tobberoth Wrote:I don't really know if that's true, that depends on the JLPT2 test. Remember, it's 6000 official words, the test makers are allowed to take 20% of the words on the test to be OUTSIDE of that range. So not even 6000 words is enough.

I think it's more like about 5000 are official, and in addition there are others outside that range that are unspecified, so in total a range of about 6000.