Continue or change focus

Hi everyone looking for a bit of advice here. I have made it to level 38 now since starting last July. I am now struggling a bit and my accuracies are dropping badly.

Grammar wise I have done both genkis, and am about 20% through Tobira. I would like to get as close to N2 level as possible for March 2022 (at least solid N3 with good vocab)

Is it worth pushing forward to level 60 until the end of the year or focussing on just reviewing the past 38 levels in Anki and spending more time with grammar/listening?


I’d say you’re at a good level to start focusing on other stuff outside of WK. You can only get so far with WK anyway. So immersion is the way forward, especially if you haven’t done much of it before.

But, stopping your momentum on WK might not be ideal. You could just decrease the number of lessons to still steadily progress through WK, though with longer leveling times than before.

If you feel your accuracy is truly the main issue for you, you can use the Self-study Quiz, the Item Inspector, and Shin Wanikani Leech trainier to address your problem items. That should also free up time and energy for other things outside of WaniKani (when the leeches stop falling back down and progress again= less apprentice/guru items).

Good luck!



Ppl were able to get good at Japanese before wk appeared, so it’s not the only way to learn. And if you feel like you are not making progress here, maybe it would make sense to look for it elsewhere. I’m not even planning to go further than lvl34 here or so.

I think it depends on what your priorities are and how you learn best: are you more interested in vocabulary or kanji or grammar? Which of the aspects of Japanese you study on WK are you comfortable with tackling on your own? Do you have a plan for doing it yourself?

I don’t think falling accuracy is necessarily a sign that the WK system isn’t working for you: you might also need to examine other factors, like whether or not your workload (both inside and outside WK) has changed and if the way you’re studying the kanji and mnemonics has changed, particularly since I’m sure that the kanji at the higher levels are more complex and contain more strokes. Even if the default mnemonics from WK aren’t working for you, the SRS will at the very least ensure you review kanji regularly, which is important for learning regardless of whether you’re doing it inside or outside an SRS. I’m not a fan of flashcards and SRSes myself, and I don’t use WK because I’m a Chinese speaker and don’t feel the need, but if the SRS seems to be helping you to stay on track, it might be worth sticking around. To put it another way, even though it’s true that you don’t have to use the WK SRS to keep learning Japanese, and it doesn’t seem to be going well for you right now, there’s no guarantee that you’ll get better at retaining new words and kanji outside of it either: I’m not insinuating anything negative here or being condescending, but I think that when we struggle with learning something, we need to reexamine how we’re learning it. Facing difficulties is normal, but the real question is whether there’s something we haven’t tried yet that could help.

I don’t see why you can’t do both, though you might want to slow down a little on WK so you have more time for other stuff. That should also give you more space to work on the items that you’re struggling with on WK. I think that exposure to real usage is very important and helpful in learning languages, especially when it comes to learning what exactly something means and how it’s used. However, you could also see learning vocabulary on WK as a form of preparatory study so that you’ll have an easier time recognising it in the real world. The fact that I know various words in Mandarin and that I know how to guess on’yomi based on readings in Mandarin often allows me to identify these same words when I encounter them in Japanese, and, when I know the kanji used but not the compound, I often have an easier time guessing what the compound means. I think it’s quite rare for sources outside of WK to do a kanji-by-kanji breakdown for new words, and so, unless you’re comfortable with doing that yourself – by looking up single kanji on Jisho, for example – I think this is one aspect of vocabulary study that you might lose after dropping WK. I think kanji-by-kanji breakdowns are essential to understanding how compounds work, however, so this might be quite a great loss unless you acquire the skill yourself. (Of course, there are compounds that are never dissociated, and various people will tell you that you can learn compounds as single units. Sure, that’s true, and I’ve done that in Mandarin to an extent, but when the day comes when you want to interpret a novel compound that’s not even in the dictionary yet, maybe because it’s slang, you’ll be much happier if you know what each individual kanji means.) I don’t really think about it much, but as much as I love learning things in context, I believe I’m vastly under-appreciating just how much of an advantage my Chinese knowledge affords me in understanding Japanese, and I think WK offers a similar advantage by giving you similar ‘foreknowledge’ of what words mean before you see them in the wild.

Just another thought about the importance of immersion and exposure though:

I’m at the beginning of chapter 13 of 15 in Tobira, and I’m taking a long break. I certainly did pick up some vocabulary from the textbook, but a lot of it is utterly useless for me because it’s so tradition-specific that I think you wouldn’t need it even if you lived in Japan, unless you were talking to friends about traditional arts and culture. As for the grammar… I’m not trying to push this on you if anime isn’t your thing, but I sincerely and very seriously can say that I learnt almost all the grammar in Tobira just by watching anime and constantly looking words up in the dictionary and online. For the past 12 chapters, three grammar points per chapter – at the very most – were new, and most of the time, I would only learn one new thing if I was lucky. That’s how much you can learn through exposure alone. I even ran through the grammar point list on (though yes, I know it’s not an error-free site), and I think I know or can extrapolate what all of the N2 grammar points mean, along with about 50% of N1 grammar. Ultimately though, I think what you need to examine is whether you’re willing to put in all the work on your own (pausing videos and looking for transcriptions or copying words from articles; looking them up in dictionaries, including and monolingual dictionaries because sometimes Jisho just doesn’t provide enough examples; and trying to commit them to memory in any way you can before moving on) or if you find it more effective/efficient to absorb things on WK with regular revision and then supplement that learning with additional exposure.

Just one caveat with regard to everything I’ve just said: again, I already speak Chinese, so I can’t guarantee that what I do is just as feasible for everyone else. There’s probably a whole lot of words I’m hardly looking up or understanding 50% of beforehand just because I know the kanji in Chinese already. I don’t know what it will be like for you, which is why I think that kanji learning systems, while not strictly necessary for learning Japanese, are probably very helpful for learners who have no experience with kanji in other languages.


Nowhere near where you are, but taking a pause on new lessons might not be a bad idea. Get that accuracy back up, feel a little less stressed. Then get back to a steady pace you’re comfortable with. By lvl 51 you’ve hit all the N2 kanji, so I don’t think you need to make lvl 60 your goal for end of year. Putting some more focus into grammar and listening will help you get your N2 more than leveling in WK.


If I got to 38, personally I’d want to at least round off to 40, but then you have some fast levels I think (around 40? Or 45?) so might as well go to level 50.

But then you’re level 50, so might as well just get 60.



Well, I always thought I’d slow down once I got further also. But, I never did. But then again, I had done a ton of immersion learning before starting WK, so I had something to build on. I think that makes a big difference: whether you can connect the items WK teach to actual language use.

If you have less experience of Japanese in the wild, the 30-40 is a good time to start working on that: listening, watching, reading, and gaming - anything that takes your fancy. And it probably best to not go full speed on WK if you switch priorities to things outside of WK, to not overdo it and burn out. It’s supposed to be fun. And with more immersion done, that will also make WK more fun when you do progress through the last 20 levels. :slight_smile:


I can’t wait to get to a point where I can use comprehensible input. At the moment I’m basically Jon Snow, I know nothing. So there is almost no comp input I can do.

I’ll get there!


Yes, you’ll need to have the very basics of grammar before reading is an option. But then I’d just go for it. Or watch anime. Maybe some simple game.

I don’t believe there is a “too early” for these things, but it’s a matter of how patient you are. Stopping to look stuff up can be frustrating if you let it. But it could also be just fine if you have the right attitude. That’s how I managed before - patience, loads of it. ^^’


Why would you want to learn things on Anki a 2nd time? Or do you mean completely moving to Anki and dropping out of WK?


I think that’s pretty normal for the late 30’s.

Aye agreed.

The difficulties that start cropping up in the later levels are because that’s about where you start learning a lot of kanji that are similar to ones you already learned in previous levels, which drops your accuracy for both the new and old ones.

You’ve also built up a big enough kanji vocabulary that almost every single lesson item can’t help but be similar to something else you learned.

That’s something you’ll have to sit down and decide for yourself.

You’ve already got close to 98% of the N3 Kanji learned already* so it may be a good time to slow your leveling pace and work on other things.

  • Caveat: The JLPT hasn’t published official lists since 2010, so those estimates are going to be based on that. Funnily enough, both N3 and N2 lists hit 100% at level 51. So you may want to push though until then.

Can’t agree with you there. Tried playing a video game with no kanji knowledge. Looking up stuff took up all the time, so there was basically no Immersion left. That led to bad retention and me looking up the same words over and over. Decided to fix that problem using wk. Going to give it another go in a couple of months. Wish me luck. :slight_smile:

1 Like

Well, that’s what I did also, because I wasn’t actively trying to memorize words. I just did on-the-fly translations. Even so, a lot of those kanji that I’ve come across have been very familiar to me here on WK, so you’re definitely learning things, though not as explicit knowledge perhaps. But, you also get an overall sense of Japanese from immersion. It also makes it easier to swallow all those kun’yomi that WK teach us as part of vocab.

Still, playing games, more than reading manga with furigana, requires patience that’s true.


Exactly. This is key. Part of the difficulty, of course, is that kanji are unfamiliar for most Japanese learners, so they’re harder to retain without extra practice beyond immersion and lookups. Something needs to be done so the symbols stick. However, just looking words up without making an effort to remember or understand them beyond a translation isn’t going to provide anything beyond a cursory understanding of the sentence or phrase in which the words appear. That’s why I personally use dictionaries that provide examples and nuanced explanations, so that I know how to use the words I come across and can internalise them. I also use monolingual dictionaries as much as possible in order to learn finer shades of meaning, because translations in English rarely match Japanese words perfectly.

I agree that doing so with no kanji knowledge, or at least no system for remembering new kanji, was starting too early, because that means there’s no way for the learner to attach meaning to the symbols he or she is encountering. However, I disagree about the idea that ‘looking up stuff taking up all the time’ necessarily means ‘no immersion’ and ‘bad retention’. I watched The Rising of the Shield Hero maybe… five times? I looked stuff up a lot each and every time, and the final time I watched it, I probably caught (i.e. could decipher) 80-90% of the words spoken in terms of proportion without subtitles. However, because I was going for 100% comprehension in that final viewing and aiming to understand everything, I still stopped the video every… 30 seconds, maybe? to look something up. Sometimes it was every 10-20 seconds. Do I still remember everything? I can’t say for certain. I’m about to rewatch the series yet again, so I’ll find out. It’s been eight months. However, I’m fairly sure I remember most of it, because the looking up clearly worked all the previous times.

I’m not saying that your experience didn’t happen or doesn’t matter, but what I’m trying to say is that looking things up ridiculously often, even to the point that you’re spending more time in the dictionary than consuming the content, doesn’t mean you’ll experience poor retention. You just have to do it in a way that boosts retention, provided you have enough basic grammatical, kanji and vocabulary knowledge for the Japanese to be meaningful for you. I could keep looking things up because some part of me knew what the sentence meant overall and how to break it into parts, so I could place the missing pieces into their slots. I was immersed in the content because I knew what it meant and I was holding everything I associated with it, including emotions, in my head, ready to attach it to what I learnt from looking words up. Without some sort of structured, meaningful storage space, the information isn’t going to stick.

I still remember random words from Quintessntial Quintuplets(五等分の花嫁)Season 1 even though I last watched it… probably more than three months ago? And I only wrote the words down the first time, which was more than one year ago. Why? Because they mean something to me. I remember ぱくぱく as the sound of noisy/energetic eating because I drew the くs with teeth on a whiteboard in my dining room next to the window and imagined ぱくぱく as the sound of clashing teeth and air popping in and out of a mouth. I remember 無言 because one of the girls in the story was speechless when she was asked a question and stood there making an embarrassed, frustrated sound. I can see the scene in my head, and I remember what colour her hair is and what accessories she has in it, along with the male protagonist’s shocked tone. Retention is all about association and meaning: if new information isn’t meaningful to you and isn’t attached to something old/familiar that is meaningful to you, then it’s not going to stick around.


Yup, particularly since few games come with furigana, or so it seems. However, immersing yourself in anything that requires look-ups and deciphering is going to require quite a bit more patience than a textbook or other system with prepared, organised explanations, so even though I think the threshold for starting immersion learning is relatively low, finding suitable immersion material for one’s level and working through the frustration involved can be challenging. Not everyone will be prepared to do it from the get-go.


I find Wanikani’s usefulness veining as I’ve reached levels above 10 as kanji gets more complex. I’m currently doing Heisig’s “Remembering the kanji” and it’s basically witchcraft. It will truly help you remember the kanji as you learn how to write them. So if differentiating between the kanji is your issue then I recommend checking it out.

I’ve also experienced what I learn through Wanikani never really sticks to memory, but what I’ve learned from “Remembering the kanji” sticks like glue. The real focus is to write the kanji and remember its meanings and not its readings, so it affords you way more options when it comes to mnemonics which helps a lot.

I’d like to know what you intend to do when you finally need to learn kanji readings. I’ve looked at a few pages from Heisig’s books, and while I think his mnemonics are probably more intuitive for some people because they focus on meaningful components (which he calls ‘primitives’) and link them to kanji meanings, I’d personally be a little wary of misattributing credit. I think the main reason you’re remembering the kanji better is because you’re writing them. As a native speaker of Chinese, I know that even after I’ve forgotten all the origin stories linked to a kanji, the one thing I’ll never forget it how to write it. The hand movements and wrist flicks are ingrained into my memory by hours and hours of writing. If I forget how to write a kanji, that’s also the point when I’m not able to visualise it anymore. As such, what I’d really like to ask is… are Heisig’s mnemonics really that much better for you? And if so, how or why? I definitely like some of them because they’re closer to actual etymology, but because I’m an etymology nerd, I got turned off when his explanations didn’t match the sources I usually refer to. However, WK does meaning mnemonics too, so… what’s the difference? Because if they’re not actually that much better, then all I can say is… it’s because of writing. A friend on the forums took my advice the other day to write out problematic readings on paper in kana. The mistakes started disappearing in that friend’s reviews almost instantly. This works for just kana, which have no inherent meaning and can even be replaced by rōmaji systematically if one so chooses. Wouldn’t the effect be so much greater with kanji, with all their intricacies?

I personally think that the biggest failing of both Heisig’s and WK’s systems is that neither system links reading to meaning in a clearly meaningful way. In WK’s case, it’s because the mnemonics are designed to be random and whacky and to not fit into real life, possibly to avoid memory interference. It links mnemonics to a system of familiar characters instead, like Mrs Chou and Koichi, which you have to get to know for the system to work. Heisig just doesn’t even try, which he of course has the freedom to do, but I’ve heard that people who use RtK often end up forgetting many of the kanji they learnt by the time they start studying readings. I just hope you’re not using RtK in isolation, and that if you are, you won’t end up being one of these people.

I mean, if any system works well for anyone, including you, then power to you all. I’m just personally extremely suspicious of any system that claims to be able to teach the symbols of a living language using only one facet of them (meaning) while leaving out another major facet (reading) entirely even though both can be learnt at the same time.


Heisig actually stops making the mnemonics for you after a while and only gives you the primitives. And the ones he does give are of way higher quality than that of WaniKani. Wanikanis mnemonics are tied to the readings, but with Heisig, they only focus on meanings, which makes it easier to remember them.

And when it comes to not relying on stories but only the writing, yes I agree. But the mnemonics are a great starting point. It helps me to remember the kanji for a while in the beginning until it’s ingrained into memory.

I learn their readings by playing games without furigana. Whenever I see words that consist of the kanji I’ve learned from Heisig I look them up and add them to my SRS. It works very well actually, and it’s easier to remember readings that way because I don’t have to focus on their meanings.

And in addition, I’m learning kanji at a way quicker pace with “Remembering the kanji” since it becomes so much easier to differentiate between them, so I can just pile on more kanji with ease. I’ve reached kanji nr. 434 in the book, and my SRS success rate is at 94%. Just a few weeks ago I was at kanji nr. 200.

1 Like

OK, from the few I’ve seen from each method, I have to agree, though I’m probably biased because Heisig’s are primitive-based and therefore closer to etymology, which I like. I guess higher-quality stuff does help, in any case.

This actually surprises me because I thought WK separated meaning and reading mnemonics, so I figured there would be no reason for that to happen. However, I’ve just tried looking at a few of the level 60 kanji, and I have to say… yeah, OK, I see that they’re linked. I guess that creates restrictions…

I’m not an SRS fan, but I know such systems work, so I’m not complaining. I’m glad that you’ve found an enjoyable way to tackle readings, and that you’re doing readings while also learning meanings. I think that learning readings in a context that helps you understand what those kanji mean is also helpful. Learning kanji in isolation, regardless of whether you’re focusing on meaning or reading, is much more difficult. I still think it’s really important to learn the reading of a kanji soon after learning its meaning though.

Could I ask if your SRS success rate includes readings, just for the sake of understanding your approach better? It’s true that a lot of WK users complain about confusing multiple kanji, which is an issue I know rarely exists if you learn to write kanji (though mistakes are certainly possible and happen even among native speakers when they first learn kanji). I can see how RtK saves time relative to WK in this fashion.

Now, I just figured I’d explain why I don’t like the idea of learning kanji meanings alone, even though I like Heisig’s approach of reducing kanji to meaning keywords and primitives, because it’s very similar to what I do myself: when I learnt Chinese growing up, doing reading comprehension passages in which I couldn’t read words was always really frustrating. I would generally be able to guess kanji meanings from context and components alone, and even give the kanji a likely reading as a temporary reading so I’d be able to get by, but just having a rough idea of what something means doesn’t get you very far when it comes to long-term retention, even if it’ll let you muddle your way through a test. I believe that learning to pronounce things is essentially, if only so that you can read entire sentences aloud and let them flow through your head so you can grasp meaning. It’s like how studies on extremely fast speed readers has debunked the claim that they can read without hearing sounds in their heads: they still do, but they just do so much faster, because reading aloud is that essential to human comprehension. That’s why I think it’s a bad idea. However, seeing as nothing forces one to use RtK in isolation, and you’ve clearly found a way to pick up readings on the side… well, I’m very happy for you!

Just one final question though: do you still use reading mnemonics when the time comes then? Or do you have another approach to remembering them?

PS: on my own mnemonics thread (here), I typically try to link readings and meanings because I feel that’s much faster. However, my goal with my mnemonics is to make meanings/readings obvious based on the mnemonics, which I feel WK doesn’t do – WK’s are memorable if the story appeals to you, but not obvious. My point is though, that I don’t think it’s necessarily bad to try to integrate reading and meaning mnemonics. However, yes, sometimes, it’s impossible to link the most obvious meaning mnemonic to the most obvious reading mnemonic, so I just link both to the kanji and leave them be. There’s no point restricting yourself like that, especially if you’re able to create two super memorable, unlinked mnemonics instead of two dull/confusing linked ones.


I’m only lvl 25, but i’m in similar situation. My accuracy is dropping a bit and I am behind on grammar (way more than you). My plan is to slow down on WaniKani and try some reading in addition to spending more time on grammar. I also intend to tackle my mountain of leeches which i’ve been putting off. I won’t completely halt WK, but I doubt having Kanji knowledge lightyears ahead of other areas of study is the most efficient means of learning Japanese.

Thanks for posting this topic! I was considering doing the same, but now I can just read through everyone’s advice for you :wink:

1 Like

I only have to spend a couple of minutes per kanji because I don’t have to restrict myself by learning readings at the same time. And because WaniKani’s mnemonics have to link both meanings and readings I feel that they become way too whacky and in a bad way. They often make no sense and feel like they were written in just a couple of seconds.

And Heisig gives you primitives in a very thought out way. The order the kanji are given to you in are frankly nothing else than pure genius. It just makes sense. Almost every kanji you learn builds upon the previous ones, which also is a major help in remembering them. The order in which WaniKani hands out kanji feels non-sensical in comparison.

When I link a kanji with a word I already know in English, it saves me a lot of time trying to remember the kanji. And I feel that it gives me a stronger link, because I don’t have to tie it to the Japanese word as-well. Essentially I can focus on remembering the kanji and what it looks like. As long as i remember what it looks like, and its meaning, I can easily look it up when I find it while I’m reading stuff.

Luckily I know two languages already; Norwegian and English. So when it comes to tying a reading with the words consisting of the kanji; I have tons of sounds I can make stories out of from both English and Norwegian after what suits the reading best.
Reading mnemonics that I make on my own makes it way easier to remember. And I’ve already mastered the meaning, so that is out of the picture entirely.