I’ve been using Wanikani for about 15 months now, and I’ve begun amassing a large number of burned vocab. But I’ve been noticing more and more, that even in the review that I burn the vocab, I’m still very often having to “parse” out the vocab from the kanji that make it up.
This alone didn’t bother me too much for awhile and I figured it was par the course, but I’ve been realizing that when I listen to comprehensible input, outside of a lot of the very common vocab, a lot of burned vocab I have I don’t recognize until I go back and read the transcript of what I listened to.
I’ve been using Anki in parallel with Wanikani for about the last 6 months now for sake of seeing what works best for me, and I don’t see this occurring nearly as often with vocab pulled solely from Anki, and in many cases I can recognize even relatively new anki vocab in my listening input.
So my question ends up being, has anyone hit the same wall with Wanikani? If so, what did you do to overcome it? Am I relying too heavily on the mnemonics and end up stuck only ever seeing multi-kanji vocab words as their individual kanji and readings and not one cohesive word and meaning? Do I not take enough time to read example sentences during reviews? Am I not doing enough comprehensible input to help solidify the vocab?
Edit: For note on my current wanikani review habits, I do them daily, with 10-20 lessons per day. I also almost always play the audio of a given vocab word after a review and say it aloud as well.
Listening and reading are very different. Are you reading anything? It should get easier over time as you make more connections. For example I can probably understand around 60% of an anime episode but reading manga is pretty comfortable with maybe 1 or 2 jisho trips per chapter.
Funny things is that this part is even harder when learning English.
This might be the key. WK doesn’t stimulate memory in the direction from word pronunciation to recognizing it, so training that separately could be useful.
I like to re-play the same sentences several times to pick out every word. It works best with a video player that has hotkeys to navigate to subtitle lines, e.g potplayer.
Another approach is to create Anki decks out of show episodes. Card front: sentence audio scene screenshot, card back: full subtitle line. It’s a bit of a hassle to set up and SRS on the go, but burns the words into memory pretty well.
I’ve had the same problem. I’m still in beginner books, where many words are written in hiragana, and I often find that I have no idea what a hiragana word means if I don’t have the kanji to lean on. I often just don’t recognize it as a word I’ve already learned in WK. Listening is even harder, and WK does very little to improve that skill.
I’ve found that KameSame.com helps a bit - it gives you the English and makes you think about how to spell out the Japanese translation, which ties the words a bit more meaningfully to the sounds, rather than just training kanji recognition. And it leverages your WK account, so you’re practicing from the same vocab list. I’ve found that it helps me learn the WK vocab much more solidly.
I’d love to find a similar website to directly help with listening - one that speaks out loud a Japanese word from my WK list (without showing it written) and requires me to type the English translation. I understand that it’s better to hear words used in context, but it’s effectively impossible to find occasions to hear most of the WK vocab words in context, especially in beginner-level materials.
The best option I’ve come up with is copying the context sentences into a Japanese text reader - current generation AI text readers sound fairly natural in English, so I’m hoping the same is true for Japanese readers. A native speaker would be much better, though.
Really, WK and Anki should be supporting your reading / listening / speaking, not the other way around and after 15 months most of your time should be spent on real skills not apps, unless you’re revising for an exam or something.
Also, 6 months of Anki is coming up to 2000 words? Assuming you’ve done them in a sensible order like a most common 2k deck or something then 2k words is a good time to switch to your own sentence decks made up of things you’ve read or watched. Or possibly pick a book at the right level and use koohi.cafe to work your way through it.
I think this suggests that for that subset of words it isn’t worth spending the time to try to do pronunciation to meaning training on them at this point in your learning. Focus on the words you are finding in context first.
(The other problem with sound to meaning is all the homophones; you don’t want to be trying to do bare word sound to meaning without context for those either.)
Was hoping for some more specific feedback relating to my specific case rather than the general feedback here, but thanks for trying everyone. I think I’ve just come to the conclusion that Wanikani’s method of learning vocab just does not work for me on many fronts. In addition to what the original post mentions above, seemingly forever being unable to differentiate 理論 論理, 勝負 勝敗 as examples, while in similar cases where I learn just whole vocab in Anki I never get different vocab so tightly entwined, maybe since I don’t know the individual meanings of each Kanji and rely on mnemonics that relate to those individual meanings causing kanji combinations with similar meanings to be indistinguishable from one another.
I’m sure Wanikani works for many folks, but it seems like for me, learning the Kanji into learning the vocab actively hurts my ability to learn the actual vocab, as much as I enjoy learning what the individual kanji mean. This will probably be where my Wanikani journey ends. </3
I know you said your WK journey ends here and by the way that is totally valid as everyone is different.
But just in case I wanted to chime in that I don’t have your issue but I also don’t rely on the mnemonics. Most of the time I don’t even read them. When a vocab word comes around I read the meaning and the reading. If I struggle in reviews I’ll think harder about the kanji makeup (still no mnemonics). I wait until I struggle in reviews multiple times before I consider mnemonics. I do this specifically because I don’t want to end up defaulting to mnemonics when I’m reading real content.
So if you do decide to try WK again someday, maybe that’ll suit you better. (Keep in mind that learning is not linear so what works for you now might not be what suits you later, and vice versa)
This is essentially what I ended up trying to do, I’m gonna give it 3 or so months of ignoring the mnemonics to see if that’s the sore spot for me since that’s one of the major differences in the two vocab sources. If I notice some change in those 3 months I’ll probably extend it bc I want to see how I manage after getting vocab from new to burned and if I recognize them more easily. Seeing another have some success with it gives me some hope, thanks <3
I guess the other difference is I’m also not going to think too hard on the individual kanji when I get a mulit-kanji vocab. What I’m trying to break is the process I go through of remembering the reading of each Kanji, smushing them together, recalling what each Kanji means, and then connecting that to the vocab meaning. That’s a whole lot of steps, and these steps remain there for many vocab for me all the way up to burned. Which I think is the main reason I never recognize the vocab when I hear them, and why I still have to go through this when I read them, in comprehensive input. (Which I know this is what the process is meant to be to an extent, but I assume your brain is supposed to shortcut this at some point with enough repetitions, but for whatever reason my brain doesn’t jive with this well and rarely shortcuts this)