Can vocab just reject readings put in the meaning field instead of counting them wrong?

I’m still early trying to clear out the basic stuff I’m familiar with. It’s already annoying going through 一、一人、and 一つ all individually for each number so please forgive me if I’m just zoning out and speeding through. The problem I’ve had repeatedly is that I’m focusing on the only remotely hard part, the reading, and every other review I put in the reading when it asks for the meaning for something and it gets marked wrong. I’ve looked at some discussion and see various complaints about requiring certain readings in kanji vs vocab and how it makes sense that one is more forgiving than the other about simply doing the wiggle instead of marking it wrong but in this case that’s really not so applicable.


I agree, it would be much less frustrating.


You have problem with to see difference between roman letters and kana. Wanikani uses a ime for readings. It autoconverts to kana. It also is written over the on the input box. So what’s your problem. ?


If you trust yourself to not abuse it, you could install the WK override script that allows you to disregard incorrect answers and re-answer them. I could also probably throw a userscript up that shows some blatant indicator on the review page that tells you if it is asking for reading/meaning… Because while there are indicators right now, they are (probably intentionally) not very noticeable when speeding through reviews.

Personally I don’t think this is much of an issue though because after doing reviews for a while (level 5+) you start to get used to recognizing the indicators for reading/meaning almost instinctively. I had this issue when I started as well, but nowadays I fly through reviews and almost never put readings in meanings and vice-versa.


Imo, what you’re suggesting is not a solution to the problem. Your problem is being sloppy. Razorcat’s solution works to remove the symptoms, but doesn’t remove the real problem.

I suggest learning to check what you wrote before submission. Not only will this help reading/meaning confusion, it will also help guard against typos or silly mistakes due to misreading kanji.

One minor thing I also dislike about rejecting readings instead of counting them wrong, is that there are a number of meanings of kanji that are equal to it’s reading (think hokkaido for example). This will introduce an incentive to try out the reading, when you don’t know the meaning, which I would consider a bad thing.


Because it takes basically no time to type “roku” and slam enter.

Yes, I could slow down and double check. This is not a situation where that should be necessary. This minefield does not need to exist. Like I said, I even see where there’s a point to having it be picky between kanji/vocab reading. There is no point in what I’m asking about though.


Is it sloppy to not want to proofread? I don’t think so. This is a learning tool and I’m here to learn, not to please the robot masters. I can see the educational value of some strictness but this is not one of them. Even in your example is that really a problem? Trying the reading and having it rejected is still going to be something to remember.

I’m not even asking for typo forgiveness. Outright wrong answers submitted can be marked wrong without attempting to anticipate and forgive fat fingers. This is me knowing the right answer and typing it in correctly but on a question that wasn’t being asked at the moment. And I wouldn’t be so irritated if it didn’t have and use a different response (the wiggle error) when that happens in other contexts.


Tempting. I saw mention of a few scripts but when I get to the stuff I don’t know yet I want to stick with the default behavior to start.

You’re probably right. I stumbled once or twice learning the color difference and I’ve been doing better about that. The meaning thing has happened more and intermittently. Probably will get better about it eventually. I saw mention that vocab doesn’t count against progression at all right? Radical and kanji progress is what matters for unlocks so missing a vocab or two isn’t a big deal? I kind of expected that just dealing with it would be the best option but I wanted to vent a little and have the feedback out in the open. Thanks.


Yes. It takes only a few seconds to double check the answer field before submitting. Just be more careful. Do you not verify that when you’re typing you didn’t accidentally hit a wrong key?

Except in the case of the wiggle, you didn’t actually put in a wrong answer. That’s why you aren’t marked as being wrong. That is signifying that you typed in a valid reading just not what was expected. That’s not the same thing as providing a wrong answer which is the case when you don’t provide a meaning when it asks for meaning.


Well, to be honest, the whole point of the SRS system is to cement the correct reading/meanings into our memories. If you’ve typed in the wrong reading for a vocab word, for example, you did get it wrong, and in a way that has an impact. If it wiggles, it’s not doing you a favour by letting you make that mistake.

Whereas when you accidentally put in the meaning/reading instead the reading/meaning, while you were wrong, that mistake wasn’t relevant to the memorisation of the kanji and so the SRS shouldn’t take it into account if it wants to reflect the current state of the reading in your memory.

I’ve gotten used to seeing “reading/meaning”, though, and noticing if what I type turns into kana or not. And if I do the occasional fat-finger (like accidentally using an adjacent vowel) and am 100% confident that I didn’t think it actually was what I typed (see the proper hiragana and go “that’s exactly what I said”, especially as I say it in my head as well), I use the “ignore answer” script, so that’s what I recommend OP do.


A few additional seconds would double the time it takes for me to review


I don’t think you’re being sloppy. I used to do that too, but after a few levels I got used to it and hardly think whether it’s asking for meaning or reading. Someone mentioned the override script, but there’s also the double check script, which gives you the option to retype your answer, I find that more useful than just submitting it as correct.


The wiggle is only for the individual kanji not vocab. Vocab is always marked incorrect for the wrong reading. The wiggle for a kanji’s reading exists because both on and kun readings are correct even if it wasn’t what the system expected.

Then you deserve to be marked wrong for not paying attention to what you type. I do it occasionally, too. I accept that I didn’t type the right answer and move on with my life. I then try to pay better attention. The override script is for people who can’t accept this.

Also, having done level 2 multiple times, it’s not like you have 1000 items in a single review so a few seconds of being extra careful hardly adds much time.

1 Like

Wow. Your horse certainly is tall. Nevermind that I’ve done hundreds of thousands of reviews, you have the real answer, and it is the only answer. Just wow. lol

I was flagged for this. Apparently calling someone out is not appropriate.

If you only take a second or two to answer in the first place it is going to make a huge difference


I feel like OP is trying to ask whether there could be the addition of a chance/warning for accidentally typing reading in meaning field.
While I see where OP is coming from, and i would honestly be lying if sometimes i didn’t think ‘Wouldn’t that be nice?’, I agree with Toomanywarm. The SRS is about cementing readings and meanings into your memory, and by not letting you put the reading under meaning and vice-versa, it makes it easier to call up a reading when a reading is needed and a meaning when a meaning is needed; instead of having to sift through the mnemonics to find one or the other.
After time I found that my subconsciousness would pick up on the black for reading and white for meaning when I’m doing reviews.
The last thing is it’s a computer, so it can’t tell the difference between a wrong meaning and the reading for a kanji; it’s only looking for what’s been listed in the synonyms.
Part of learning is to make mistakes, sometimes we may need to adjust things slightly before trying again.

1 Like

If you think my suggestion is ‘pleasing the robot masters’, you’ve obviously learnt nothing here. Do with my advice whatever you will.

1 Like

The strictness in this matter is completely pointless, wanikani aims at teaching you how to read kanji’s and what they mean. When you actually read something your brain doesn’t have to decide if a reading or a meaning are needed, as you probably need both simultaneously. And you have to retrieve both in a fraction of a second and you most definitely do not ‘proof read’ and spend ‘a few seconds’ per kanji, otherwise reading a page would take one hour.


This thread has gotten way more vitriolic than it needs to be. Just relax. It is okay to have differing opinions.

I think razorcat is pretty spot on. This was a big issue for me at the start but eventually WK trained me to not do this anymore. With more time I think you’ll find it’s not a problem for you anymore but I certainly empathize with the struggle.

From a practical standpoint, I think the main reason that this will never happen is because it is simply an issue of opportunity cost. There is a finite amount for developer time to make changes to WK. They could certainly do this for the 8000+ kanji and vocab, but that would be at the expense of doing other things. For an issue that seems to go away for most people after enough time, I think it just simply isn’t worth it, at least for the moment.


It definitely gets better with time, even in the next few levels or so you will see yourself recognizing what it is asking for without even noticing too much. I think if it bothers you a bunch the easiest option would be one of the scripts people have suggested. (I am not familiar with any of them because I don’t use scripts) They seem to work out great for a lot of people though.

The only thing I would offer other than that is an opinion from the other side of the viewpoint. It is very important for me specifically to pay attention to what I am doing, I tend to go on autopilot for a bunch of things and Wanikani doesn’t let me do that which is helpful when in the future I will be taking tests etc on Japanese. If I was say taking a test, and I filled in an incorrect bubble because it was asking for one thing over another then it wouldn’t shake at me when I put in a wrong answer. I like having personal accountability on Wanikani for that matter. If I don’t pay attention and put in something when it is asking for something else then I am happy it marks me wrong because I feel like it makes me a better person.

Definitely something a lot of people wouldn’t worry about but I do like that aspect about it, if Wanikani just made it where it shook then it wouldn’t really teach me to pay more attention for what something is asking, in my case it would just kind of cement that I didn’t need to pay attention because it would let me off the hook. But anyway, the good thing about it is there are scripts to help you out if you don’t like the base mechanics of Wanikani. If you go all the way through level 3 and don’t like it the way it works, then maybe another system would be better suited for your particular reasons for learning anyway. Don’t really know your reasons for learning kanji, so that one is difficult to answer.

1 Like

Another script you could use (that is not the override script) is the reorder script. You can choose the “1x1 mode” to always answer reading and meaning back to back. That way you’ll always have the reading first (or meaning - up to you) and you don’t have to think about whether WK wants the reading or the meaning now. It definitely helped me speed up my reviews.