I've tried searching the forums and the internet but all the lists i find seem to conflict on the number required and their kanji use. I got my JLP4 and JLPT3 lists from the JLPT Study Page but from looking over the JLPT2 list over there it doesn't seem too reliable.

Anyone, please? I don't want to enter an entire ~5000 or so list into trinity to find out it's wrong.

Thanks in advance.

All the lists I've seen are pretty poor, especially those with english definitions. They tend to just be the thbz list run against EDICT with a script, ensuring incorrect definitions for 20%+ of the words (either EDICT's poor definition, or the script matched the wrong entry for homonyms).

I just discovered playsay's (http://www.playsay.com) lists yesterday though. While they are still based on thbz, they have been gone through by hand to clean it up and add correct definitions and parts of speech. They list kanji for many of the words, but not others. I don't think they are trying to match kanji appropriate for the level, but maybe just what words are commonly kanji-fied. There is also a for-pay list that claims to be generated from the official test specs and professionally translated etc etc. I can't comment on that one since I'm not paying $40 or whatever it costs for a vocabulary list (and useless mp3s).

I'll probably be dropping it into Anki in a few days.

-edit-
If you request the free trial mp3s you can login and download the pro translated vocab lists for free - but only for the level you requested. You can modify the download url and get the rest though.

Thanks jarvik, the JLPT2 list looks useful, probably start transcribing it into Trinity after I do the JLPT3.

Also 200th post

I've learned JLPT 2 and a quarter of JLPT 1 from the playsay "pro" lists and they are maybe better than just plain Edict in a few spots, but for every spot they're better they're worse somewhere else. They were obviously done by a native Japanese speaker too arrogant about his English ability to see a native English speaker is also 100% essential to make a real quality list.

Example is one word meaning "spitting image of" and written like that in Edict became "splitting image of"... so a lot of little confusing shit like that. Then you have the usual Katakana crap like ダウン with a completely worthless shit translation of "down". As if I couldn't GUESS IT SOUNDS LIKE 'DOWN'. The problem is what does it mean? The direction down? ... So I go in Edict and see #1 entry is actually down (feathers), which I might not have thought of at all. The common separate uses of the word need to be delineated cause no language is going to match up on that exactly.

Everything out there is crap. Playsay is just the least crappy....

Unless you've found a better list (You mention EDICT but every EDICT based list I've found had countless matches to the wrong homonym.)

Yeah, it's not really surprising that the available vocab lists are so spotty, since they're put together by people who don't yet know the vocabulary (if they knew it, they wouldn't be putting together a study list). You'd think UNICOM or another company would realize there's demand for a simple professional vocab list, but I guess they're making more money with their thicker study books.
I'm still grateful to the people who have made their lists available.

snallygaster Wrote:Yeah, it's not really surprising that the available vocab lists are so spotty, since they're put together by people who don't yet know the vocabulary (if they knew it, they wouldn't be putting together a study list). You'd think UNICOM or another company would realize there's demand for a simple professional vocab list, but I guess they're making more money with their thicker study books.
I'm still grateful to the people who have made their lists available.

Considering the quality of the lists, I really wonder what the format of the data in the the original test specification is. I was looking at one of the lists and it had careless mistakes like having the kanji listed as 大家, the reading as おおや, but the meaning as the definition of たいか. If the original list just has the kanji listed, I sort of feel like it might be better to just post that unless they're really sure whether おおや or たいか is correct, for example.

Also, I think trying to study for the test based on edict's definitions would be rather difficult. I've been studying for the 1級 and there are a zillion words that would "synonyms" if you went by edict's definitions and on the section where you have to choose the best word this would really screw you over.

mystes Wrote:

snallygaster Wrote:Yeah, it's not really surprising that the available vocab lists are so spotty, since they're put together by people who don't yet know the vocabulary (if they knew it, they wouldn't be putting together a study list). You'd think UNICOM or another company would realize there's demand for a simple professional vocab list, but I guess they're making more money with their thicker study books.
I'm still grateful to the people who have made their lists available.

Considering the quality of the lists, I really wonder what the format of the data in the the original test specification is. I was looking at one of the lists and it had careless mistakes like having the kanji listed as 大家, the reading as おおや, but the meaning as the definition of たいか. If the original list just has the kanji listed, I sort of feel like it might be better to just post that unless they're really sure whether おおや or たいか is correct, for example.

Also, I think trying to study for the test based on edict's definitions would be rather difficult. I've been studying for the 1級 and there are a zillion words that would "synonyms" if you went by edict's definitions and on the section where you have to choose the best word this would really screw you over.

Yeah, when studying for 1級 you pretty much have to start using 国語辞書. Or you need proper examples etc for every word, just so you know the difference between the "synonyms".

[Katsuo]

mystes Wrote:I really wonder what the format of the data in the the original test specification is.

Here's a selection exactly as presented in the test specification book (vocabulary for levels 1 & 2*):

オーバーする
おおはば　（大幅）
オープン
おおまか
おおみず（大水）
* おおや（大家）
おおやけ（公）
* おおよそ（大凡）
* おか（丘）
* おかあさん・さま（母）
* おかえり（なさい）［感］
* おかげ
* おかけください［感］
* おかげさまで［感］

Katsuo Wrote:Here's a selection exactly as presented in the test specification book (vocabulary for levels 1 & 2*):

Thanks. That explains how that type of mistake would occur; they must have copied the reading from the specification, entered the kanji into a dictionary, and failed to verify that the dictionary's reading matched the one they wanted.

If you guys want a list, perhaps you should all, those with the book, type up a small section and share it. You could even make example sentence lists and such.

Unfortunately it's not a book many people have and bookstores don't normally stock.

Well, the thbz lists are supposedly the official specs entered "as-is", so they could be used (I have seen the specs book in a bookstore once but it's ususual to see it there. You can order it from Amazon, I believe).
I think that getting together a decent vocab list is a worthwhile endeavour but it might not be realistic to expect to get it done before this year's test. If you've been dealing with the existing lists you know what a headache it is trying to decide on definitions you feel confident about. It might be something to get started on on December 8th.

snallygaster Wrote:Well, the thbz lists are supposedly the official specs entered "as-is", so they could be used (I have seen the specs book in a bookstore once but it's ususual to see it there. You can order it from Amazon, I believe).
I think that getting together a decent vocab list is a worthwhile endeavour but it might not be realistic to expect to get it done before this year's test. If you've been dealing with the existing lists you know what a headache it is trying to decide on definitions you feel confident about. It might be something to get started on on December 8th.

I had forgotten where I got it, but I guess I actually used the thbz list. I'm not sure how much value there is to try to assemble a new list. Edict style definitions are useless, and since I pulled a lot of my sentences from Kenkyusha's dictionary I can't really post them online. Since you don't know what sense the words will be used in, it may be best to just work with a raw list and investigate them yourself anyway.

If you're actually going through the list anyway, it wouldn't be such a big effort to check the words with edict. I worked a bit with the http://www.jlptstudy.com list and I reported mistakes I found to the admin who made the corrections. If everybody did that there would be a good list in no time.

Well the thing is, it's not always as simple as "just check the words with edict" -- some of the entries are straightforward, but a lot of them require some informed judgement from someone with a high level of Japanese ability (ie higher than the level the list is for, preferably a native speaker), and also some familiarity with the test itself.
As for the jlptstudy.com list, I have a huge amount of appreciation for the admin of that site and the list itself, but it's been undergoing constant correction for *years* and it's still full of problems. This isn't at all a criticism of the site or the admin, but a reflection of how tricky the endeavour is (mainly because the only people with incentive to work on the list are people who are still learning it).