Keep the kana-only vocab

But it is still a word with a meaning, that you can find in any normal dictionary, and that can be efficiently learnt using SRS like wanikani.

1 Like

Yeah, WK has always taught grammar points as vocab, like 毎に, or 以来. I’ll still object to teaching the kana form of it when the kanji exists, but we’ve beaten this horse to death in countless other threads.


They should really always teach the kanji versions… Makes no sense for a kanji learning platform to make you have to learn its actual kanji form outside of the site.

1 Like

I don’t think it makes sense for a kanji platform to teach grammar points as regular vocabulary either way. Without proper, concrete explanations and examples that’s likely misleading.


When I think of “a kanji platform shouldn’t be teaching grammar point” my image of a “grammar point” is something like ~なければならない or ~んがため or something. To me, things like 以来 or 殆ど are just words you can learn. Sure, they might appear in some resources grammar section, but they feel totally different to me. Any word is going to have details around its usage that go beyond the simple meaning, and for some words it will be more complex than others, but still.


Fair point. Some of those words are also flexible enough to be used outside of fixed grammar points (or what is being taught as such) as well.

1 Like

Everybody: “Make kana vocabs optional!”
Strawman: “Everybody wants the kana vocabs removed. I want more of them.”


WK: To learn how to read.
adds kana words
Some people: Why?
WK: reasons.

People join WK in order to learn how to read, for kanji. Vocab comes secondary, if they were interested in vocab they would get a common vocab deck.

Solution: Split the deck in two, one for kanji, one of kana. People could choose to do whatever. Would also benefit absolute beginners since they would be able to learn vocab before getting deep into kanji.

So they would essentially get kana srs → kanji SRS. That way they would be able to learn the most common kana vocab, dive into kanji and then reinforce with immersion.

I have no problem with new crap in my deck, even if I know it already. It’s just the principle and it ruins what WK was good at. I like recommending WK to people, and it’s hard to do if they ruin what they are good at by flooding their deck for kanji knowledge with kana.


Flashcard based SRS are fundamentally not efficient, just convenient and easy to use. There are a hundred more efficient ways to learn the word ナイフ than typing the meaning into a console 10 times over multiple months.


Especially in the format WK does it.

1 Like

I disagree. Looking at my anki, many words of mine take me about 5-8 seconds per review. I have personally set doubling intervals with minor adjustments, but a card appears about 8 times the first year I see it typically and I have 95% accuracy so most words I don’t forget. If we take an average review to be 8 seconds for me… Then 8 times over the year is 1 minute of review time over an entire year to have it in my memory for the entirety of that year.

That seems pretty efficient to me. Could you list some of these other hundred ways that are more efficient?


I struggle with 擬音語, 擬態語 and the other onomatopoeia type words, and I would be happy to see more of those included. There’s got to be some way to help remember these more easily!

I believe they meant SRS is not efficient for learning new vocabulary (based on the rest of their post). Learning new words out of context is usually not going to allow them to stick because you don’t understand what they really mean or how they’re used.

Of course SRS is good to efficiently get consistent exposure to a word you already know in order to not forget it. That point would be very hard to argue.


The whole point of SRS is to efficiently learn stuff. Could you list these ‘100’ methods that are more efficient to memorise vocabulary?

well, technically one is meant to learn first. Then use SRS to review.

Just semantics, but for some people it can be eye opening if they’ve been skipping the learning stage and having trouble with SRS.

edit: but yeah, not to detract from the point you were making that SRS is designed to be a highly efficient means of getting a word into long-term memory.

Though the example of ナイフ for an English speaker was humorous for both sides of that argument. English + katakana knowledge should suffice for an immediate eternal burn.


Most English speakers would probably guess what ナイフ means if they didn’t know already, but probably most English speakers who don’t speak Japanese wouldn’t know how to say ‘knife’ in Japanese, so there is still merit in teaching that word. In general there is value in teaching borrowed terms as one can get used on how Japanese tend to convert foreign words into Japanese. There are some clear patterns and it’s still beneficial to get exposure to them.

As per the ‘learn’ vs ‘review’ I agree, SRS is meant to review stuff not to learn new things. But wanikani is not a pure SRS system, it has ‘lessons’ where one can learn new things and see them used in actual sentences, so the following SRS cycle is meant to reinforce what one learnt in these lessons.

1 Like

Still disagree. I mean what one would call the initial learning takes place outside of srs since srs is review, but no matter what we take the definition of learning new words to be, I feel like srs remains possible to be used in an efficient setup. The idea that it’s fundamentally inefficient is about as wrong as it can get imo.


I originally thought they were speaking to the kana vocab part of it, but now I’m not sure if they’re talking about the SRS part lol.

1 Like

To begin with, WaniKani uses mnemonics to support memorization in Lessons, before putting into the SRS of the Reviews part.

Strictly speaking, SRS is the timing system. Flash cards – what to put as the question, what to answer, and what to show after answering – aren’t the SRS itself, but ride on the SRS. Questions that encourage bigger answer, trying to recall more, would help with memorization. What are shown after answering may help as well.

Actually, attempting to recall more, doesn’t need SRS at all. Only that a good question needs to be asked. That is, depending on how a good question is formulated.

I see SRS as a good method to carry short term memory into medium term, if the learning interval is short enough. (Up to 3 days in WaniKani, iirc.)

After that, SRS might be a good method to selectively increase exposure, not really the memorization itself.

1 Like

It’s only inefficient if you’re spending all your time doing SRS and no time outside of that using the knowledge.

Like the people that dedicate all their time grinding Anki and that’s all they do for the day is a waste.