See title. I feel like 99% of the errors I make are due to the addition of the addition or subtraction of う and it’s really annoying me and making the experience frustrating. It’s either that or small mistakes with similar letters like る vs. ろ. I feel like, just like if you have a small spelling error, you should get that shake where it lets you know you’re almost there. Idk why it’s so much more unforgiving on this point than in english misspellings. Maybe it’s bad for learning? but also like, why do I need to go all the way back down to apprentice for something i’ve mastered just because of this small mistake? Also sometimes it’s accidental??
A mistake is wrong no matter how small. How are you going to learn how to do it right if you are lenient for the mistakes yo do? Marking yourself as wrong when you are wrong is part of the learning process.
I won’t deny that being frustrating, but I’m on WK’s side on this one because the vowel length difference is very important to Japanese and it’s something learners might want to gloss over as close enough. You want to be treating the distinction as just as important as the difference between any two other sounds.
That said, you mention accidental – your choice, but I wouldn’t have personally survived WK without an add-on to allow for marking typos, close enough English synonyms, etc as correct when WK said they were wrong. But you really should not use that to mark genuine mistakes in long vs short as correct.
Japanese people struggle with the difference between Ls and Rs. Doesn’t mean it isn’t important to learn.
I’d also point out that the disparity between how forgiving the app is for English spelling and Japanese spelling is that the app doesn’t claim to teach you English. Whereas it would be a problem for it’s claims of teaching you Japanese words if you had the word sounds and spelling wrong.
The only time I mark a mistake in the Japanese as correct is when I’m 100% certain I actually know it and it was merely a typo.
With the English translations/glosses, it’s often fine if you just get the idea close enough, as the specific senses will cement themselves into your head as you come across the words in the wild, in context.
But if you’re, say, trying to type the word for “hallway,” ろか won’t get you it. 廊下 is only read ろうか. る and ろ may look similar, but there’s a big difference between 見る (plain form verb) and 見ろ (informal—even rude—command). Generally, learning spellings that are merely “close enough” will cause you more problems than learning glosses that are merely “close enough.”
which addon are you using?
I haven’t used WK in a long time, past that now, but I believe I was using Double Check.
Please do heed the warning and use with caution though . Typos and frustrating WK English glosses are the way
What helped for me was memorizing one specific kanji for every common sound and then use that in my own mnemonics. Often beats the english-based mnemonics WK uses for these subtleties.
So for instance 上 is じょう. I memorize this and then reuse it for other kanji with the same pronunciation. So 場 (also じょう) is like seeing a place from above. 情 is feelings, and I like when I feel up (not down) etc…
But for 助 or 序 (じょ) I would use something else, like 女. This way in your brain you start having distinct groups of kanji that reinforce each other. You know that certain kanji are pronounced the same and, as long as you remember the exact reading of one of them, you’re good. Meanwhile you also separate strictly the long and short versions of the syllables. Long is 上, short is 女, completely different.
Basically the idea is that you want to consider long and short vowels as distinct as m and n are distinct. Try to see them as completely different things, not “variants”.
The last paragraph of Long vs Short Vowels | KANJIDAMAGE might be useful.
Interesting, that confirms my gut feeling that long vowel onyomi are more common than short…
Think about it this way. People have built-in error correction for languages. A small error is the one that we can easily correct from context. It could occur due to a slip of a tongue, or from noisy environment, or from reading something you cannot quite clearly see.
So for example in English, if somebody consistently swallows a sound in “eror” instead of “error”, it’s not going to hurt understanding too much, since there are no other close words.
But when it comes to long vowels in Japanese, often you cannot error-correct them because words/kanji would exist with both the long and the short vowel. So the notion of what a “small” error is changes with the language.
For example, you could look up しょうじょ、しょうじょう、しょじょう、しょじょ in a dictionary. Do you really want to add potential misunderstandings due to short/long vowels to the already long list of words with each pronunciation?
As for WK’s attitude of “no takebacks for typos”, I think many agree that it’s counterproductive and use addons to fix this.
There’s also the fact that Japanese is spoken with less “flow” than English. That’s because there is no stress accent. Knowing this is probably more important than worrying about “getting pitch accent (which is what it has instead) right.” So in English people say “ba NAYA nuh” but in Japanese it’s pretty much three exactly even beats of 3/3 on a metronome, “ba-na-na.” This kind of word really doesn’t exist in English anywhere. We can say “REE -negade” or “adMIII-ni STRAA-tion” but we never say ad-mi-ni-stra-shon and it would sound absolutely bizarre and robotic to try. In any case, that’s how more or less every word is spoken in Japanese, and it means this difference is much more audible than it usually is in English. Sentences are like a machine gun rat-tat-tat-tat-tat of syllables, and long vowels are two whole tats. I highly recommend practicing it when you’re sounding out words in your head - a word like shoubu should pretty much be sounding out the “shou” exactly twice as long as the “bu.” It doesn’t hurt to really over-exaggerate it either, because you’ll inevitably slip back to pronouncing it more like you would in English if you’re ever actually speaking. S s s h o o o u u u u u u - - - bu.
It’s easy to overlook this, and think you’re hearing clearly when you’re not, because ん as a single syllable and double vowels kind of sound similar to stress accent. Shougun is sho-u-gu-n with n held for a whole beat just like sho and gu, and the two halves end up being the same duration despite the double vowel. Saying “banana” the English way sounds a lot like saying it with a double vowel in the middle in Japanese (ba-na-a-na).
Agreed. I would probably reduce the list of statistically short ON to フ and ク… I’m pretty sure ソウ has come up more often than ソ on WK so far
I don’t think you understand what mastery is. It’s inconceivable to say that you mastered something when it was incorrect. Perhaps, instead of getting upset because it’s frustrating how “close” it felt you were, use that as motivation to learn better how to distinguish the differences and why they are important.
I also have doubts that you’ve even got a good understanding of the fundamentals of the language, but I won’t say anything further about that. Especially since many people have already offered useful suggestions regarding your current pitfalls.
Anyway, in regards to using scripts to circumvent this…
The great thing about language learning is that it’s a solo game.
I’m reminded of this legendary quote:
You cheated not only the game, but yourself.
You didn’t grow.
You didn’t improve.
You took a shortcut and gained nothing.You experienced a hollow victory.
Nothing was risked and nothing was gained.It’s sad that you don’t know the difference.
I’m fully in favor of using tools to correct actual typos, but like most people here have said. The difference between o and ou is important and really definitely should be flagged wrong. A quick one that comes to mind is the difference between 最小 and 最初.
Also in terms of “cheating”… I really think that videogamey aspect that Wanikani injects with the level system is the real underlying issue here. Getting things wrong too many times does stop you from advancing in this program, making you feel stuck. Even though SRS especially should allow for failure in a non-stressful environment so you can do better next time. But you do have time pressure with a continuing sub you’re paying for. So as much as that cool dopamine hit gives you great feelings from hitting that big Level Up, it can also have the adverse effect of making you feel much more stuck than you need to be feeling. So people try to find ways around this when the issue really is just…
There shouldn’t be a leveling system gating continued learning the way there is in my opinion. It’s unnecessary stress. Alternatively, you just need to be okay with taking longer. But eh.
It is indeed a solo game. If you’re not doing it for bragging rights to show how elite and cool you are, but because you genuinely want to be doing this, you will correct yourself and actually learn over time. And learning a complicated language with a new typography set should take time. That’s completely okay.
I think the issue here is that it feels like starting from the beginning. Some sites, like Bunpro, will move you back only by one tier.
I would like to point out (mainly for OP) that even going down from master level to the very beginning is about 7 weeks, and going down by 1 tier is 4 weeks. Bulk of the time is in last tier. This is just how SRS works. Bunpro compensates for this by using more SRS stages and shorter jumps between them. WaniKani uses larger jumps, but pushes back more. It has to be done this way or SRS wouldn’t work.
I really appreciate this and wish I could figure out a good way to do it for myself because it makes so much more sense than trying to remember them one by one.
I find I can’t track connections I make between different kanji by just doing the SRS but struggle with the best way to lay out handwritten notes that would meet that need.
Do you just have one of those memories that can recall your associations and groupings or do you write it out somewhere? If the latter, I would love to see anything you might be willing to share in case it might help me or others do the same!
I think for the most part it’s just bruteforce. I do write kanji down a lot however, because I practice kanji drawing with this deck: Anki deck for Kanji drawing practice with all WaniKani kanji + jouyou + more
But I definitely abandoned WaniKani’s mnemonics for pronunciation very early on because I’m not a native English speaker so a lot of the time they would actually teach me a wrong reading (since in my accent a lot of these mnemonics simply don’t work).
Basically I just try to tell myself little stories involving those kanjis as I mentioned in my comment. As I progressed more I would focus on phonetic components that help guess and remember the reading of a good chunk of the kanji:
This is now fixed in version 2.24 of Advanced Context Sentence 2 (the problem was a bugged version of the Custom Icons library script). I also went ahead and updated Keisei and Niai to the newest version of the Custom Icons library script so that all scripts use the same version.
The more kanji you know the more these components are useful.
Thanks so much!!