Does anyone know where any of these JLPT kanji lists originated?


#1

I’m trying to get more organized about the kanji that I learn outside of WK.

I started with the wkstats site list of JLPT N1 kanji that aren’t included on WK. Of course, I know that some are unlikely to show up even if they are listed as joyo kanji, but I’m curious about how something like this got on there.

This is a jinmeiyo kanji, meaning it’s not taught as part of joyo and it is in the list of kanji for use in names.

wkstats and Jisho list it as JLPT N1. Jisho also notes that it’s 1149 of 2500 most used kanji in newspapers. So maybe that’s why it got on the list, even though it’s not joyo?

But as far as I can tell, the reason it’s ranked that highly is because it is used for names that include ひろ and ひろし. The JLPT will never ask for a name reading, even on N1, in the kanji section.

So… like… any idea where these lists even came from in the first place? Is wkstats just referring to Jisho? Or are they both referring to some other source?


#2

I think they got the data from the kanjidic file, which is similar to jmdict.

The project summary page mentions

The level of the Japanese Language Proficiency Test (JLPT) in which the kanji occurs (1-4). Note that the JLPT test levels changed in 2010, with a new 5-level system (N1 to N5) being introduced. No official kanji lists are available for the new levels. The new levels are regarded as being similar to the old levels except that the old level 2 is now divided between N2 and N3.

which sounds like there was an official list at some point.


#3

So, there’s no real documentation or justification for any given kanji’s proposed JLPT level in this database, just the hunch of whoever put it together.

Because it really does seem like with a bunch of these that they are just name kanji that show up in newspapers a lot.

Kinda silly to call them N1.


#4

The lists were originally derived from the pre-2010 tests, which were publicly available (after being administered, of course). I don’t remember the provenance of the later lists, but they were at least partly derived from the earlier ones. To be honest, though, I think they were derived too soon after the 2010 change to be reasonably accurate reflections of an average test.


#5

I’m just wondering how much vetting the kanji went through.

Like, I can see 浩 showing up on N1, as a name, in a reading passage, with furigana.

I wonder if the people who put these tags on there just sorted all the kanji that appeared ever and said “that’s the list of kanji for the test” without considering if there was ever any reason to learn some of them.


#6

I think they made the old level 1 the new level N1 and did some magic to distribute the level 2 kanji to levels N2 and N3.

It’s hard to find the originals lists, I guess this can do (taken there from a 日本語能力試験の出題基準):
http://web.ydu.edu.tw/~uchiyama/data/index.html

They probably just added more kanji, it’s better to be save than sorry. Nobody is involved with creating the tests anyway, the lists are pulled out of thin air (and nobody can prove them wrong) …

I guess the lists are now “all jouyou + stuff that oddly appeared on the old list”.


#7

The problem is that when you see a kanji listed as N1, there’s an implication that it will appear in important vocabulary or as a pure kanji reading question, so it’s easy to waste a lot of time on these name kanji if you’re not double checking that part of it.

If you just toss it in a flashcard set with all the words that go with it, it’s probably going to be fruitless. They all have associated words, even if they are jinmeiyo. They are just super obscure words or readings.


#8

They decided to abandon the lists probably because the people started treating it like a checklist. I think it’s actually more liberating because many of the obscure kanji that they included in the list “just in case” are now more unlikely to show up. If it’s only in names and obscure words don’t learn it, if they actually ask it in the test they are unfair; how does that measure your ability in Japanese?

I’m not even sure if the kanji in the list were questions or just included in the texts with furigana as names or something, that is, not really part of the test?


#9

Right, that was my hunch.

I’m disregarding them, but I still have to wade through them first and after the 25th one it started to get annoying.


#10

I’ve done some mapping of the lists against the official practice tests, and they’re not very accurate. I wouldn’t put any faith in them unless you just have no idea what to study, in which case they’re better than nothing.


#11

Yeah, all in all, I think I’ve chosen about 10 of the kanji that appear in the wkstats “261 of 2211 JLPT kanji not yet covered by WaniKani.”

Some of them in that list legitimately could appear on the test, but yeah, this was mostly about “who the hell thought these would be on N1”


#12

@Leebo Just out of curiosity, how do you go about to learn the kanji not included in WK? Do you make physical flashcards or do you use an electronic flash card system? Do you list kanji and related vocabulary?


#13

I put them in a list on the Kanji LS app I have for iPad. In addition to just normal flashcard functionality (without SRS), it has a nifty “test by missing kanji” function. This lets you do stuff like see __炭 (しんたん wood and charcoal, fuel) and I can write the kanji for 薪, which is the only one that makes sense given that reading and meaning. I don’t choose which words are tested so it could be anything, unlike KaniWani, where I already know all the words, or a custom deck, where I chose them myself. The challenge is good I think.


#14

Thanks that sounds interesting! Will keep that in mind for once I get through WK.