[Bug?] JLPT N5 vs 100 Kanji for JLPT N5

Hi, I am going through the JLPT N5 deck rather than the kanjis only equivalent, as I prefer to learn kanjis as part of words rather than in isolation. I do have a couple of issues with the deck though. The first one is already being discussed here, and the second one is that even if I am almost halfway through the deck it seems that I have only studied 4 kanjis from the 100 Kanji for JLPT N5 deck. I am quite sure this is not correct, as I have opened that deck (and rejuvenated it) and there are so many kanjis that I have definitely studied in the other deck as part of words. In fact, I suspect I have studied most of them, so why are they not flagged as studied? This makes me doubt the global stats about words known and unique characters is correct, as it seems to consider the individual kanjis in that deck as “words” and not unique characters. Am I missing something?

Skritter doesn’t mark a word like 行 as learned if you’ve learned a word containing it, like 行く or 行を改める, because they are separate words.

In Japanese, for single kanji, learning one set of readings/definitions doesn’t necessarily mean you know another set of readings/definitions. So Skritter separates these out into separately learnable words. What I mean is this: obviously the writing is the same, but I think you’d agree that
行(い) “to go; move towards”
行(ぎょう) “line (i.e. text); row”
行(こう) “counter for banks; counter for groups or parties of people”
are fairly separate and independent meanings and pronunciations, and only one of them could potentially be reasonably derivable from a word that uses the kanji like 行く.

In other words, knowing 行く doesn’t necessarily imply you know all the meanings and readings of 行 by itself, or to generalize to Japanese in general, any of them, really. And even on the single kanji level, if you had learned the word 行(い) “to go”, it doesn’t mean that you would know or be able to derive the reading/meaning for 行(ぎょう) “line (i.e. text); row” without explicitly studying it in some other context. Thus, Skritter doesn’t make any assumptions about kanji contained in words and keeps them as separately learnable vocabulary.

However, the mobile client, Skritter: Write Japanese, does still count the kanji in longer words like 行く towards your unique characters count on your overall progress tab!

Thanks for the answer! I have to admit I am still a bit confused. Sure, a kanji can -and generally does- have multiple readings and meanings, and sometimes can even be a word in itself without any okurigana. But the deck is called “100 kanji for JLPT N5”, not “100 one-character words”, and various kanjis in the list are not indeed words by themselves, so I am not sure why they would be counted as “learnable words” if they are not actually a word.

I suppose what I really would like to see is a list of the known kanjis (the ones I can write, regardless of how many readings or meanings I know), so basically rather than just a total of “unique characters” a list of what these characters actually are, perhaps with the option of sorting them by JLPT, Joyo or frequency.

There are a few reasons why single character vocabs are counted as learnable things in Skritter’s system. First and foremost, sometimes a single character is a standalone word: 犬, 猫, 魚, 女, 男… Second, we focus on helping learners tackle the unique problem that the Chinese and Japanese writing systems present, which is largely based on the characters; why would a writable character in the language that could be read and understood by a native speaker not be counted as something learnable? Skritter supports many styles of learners with different goals and approaches, so if single characters don’t suit your style, I hope that we do support another approach that fits better with how you want to learn!

Here’s a few use cases on why a user would study single kanji, even if they generally are not standalone words: depending on which list/deck or textbook a learner is using, it may introduce certain individual kanji, and usually with a specific definition and reading that the learner needs to study. So Skritter tries to match this data in its decks. For Japanese specifically, Heisig’s Remembering the Kanji is a popular resource that solely deals with single character vocabs. Alternatively, if a learner wants to practice writing–not just acquiring or memorizing characters, then studying individual kanji rather than words makes for a faster, more focused review session. Or, some advanced learners simply want to review only individual characters as they already have a large vocabulary that would be tedious to SRS but want to refine their differentiating and writing skills (e.g. with a custom definition, a user can make prompts like “schoolの学__” to test on whether the second character is 佼 or 校).

Depending on how deep down the rabbit hole a learner wants to go, Skritter even has data for many character components (“radicals”) like 扌(not 手 but 扌), which can help learners achieve a greater understanding of character composition. Many of these are not even, strictly speaking, individual kanji or standalone words, but they do have readings and definitions in Chinese and Japanese and are counted as words in Skritter’s system. For instance, knowing that the component ⺝ often represents 肉, not 月, helps the decomposition for characters like 脳 make a lot more sense and can speed up memorization. Learning about ⺝ alone and independent of 月 is just as meaningful and worth as much as having learnt an entirely independent word in a learner’s target language.

Admittedly, I personally think the individual character approach makes a bit more sense on the Chinese side of the fence than the Japanese, but we see learners on both sides taking an individual character approach partially or fully for any number of reasons.

We currently don’t show a list of a user’s unique known characters, but it would be a cool feature to have, especially for test preparation. I will pass the idea along to the rest of the team!

1 Like