Remember when you do your reviews, and you are asked for a kanji, and you accidentally enter the kunyomi? Wanikani shakes a bit and tells you “Wanikani is looking for the onyomi of this kanji” You then go “ah, right sorry.” and enter the right reading. It doesn’t tell you it’s straight up wrong, it tells you “look you are technically correct, but we’re looking for another reading”.
Now, when you are doing vocabulary reviews, you sometimes get vocabulary consisting of just one kanji. But when you accidentally enter the onyomi of that kanji, it straight up tells you, that you are wrong. Now I understand, that wanikani is looking for the kunyomi and that technically the standalone kanji is a “word” and therefore only the kunyomi is correct, but how about we get a message like we get with kanji review?
So instead of being straight up wrong, wanikani could tell us: “Wanikani is looking for the kunyomi”, just like with kanji reviews. Do you think that would be possible?
This has been covered a few times before, but essentially it boils down to this:
There is ambiguity with the kanji and therefore if you put the wrong reading in, that’s understandable. There is no ambiguity with the vocab readings, it’s wrong or it’s right. Also, it’s not as simple as “wanikani is looking for the N reading” because it might be an exception reading, like げ for 下 in 下品 since kanji can have more than just the two readings.
But that’s not what the message says. The message actually says that the reading you provided was not the one they were looking for. That’s why it simply shakes rather than marking it wrong because what you provided is a valid reading for the kanji just not the one they wanted you to answer with. It can actually be either on or kun they were looking for. Not all pink kanji reviews are looking for on’yomi. 川 being only one such example.
Because it is wrong.
Because for that specific vocabulary word other readings are not correct even though a kanji, as a building block or concept, can have multiple readings.
But what if the vocab word is asking for the on’yomi such as for 本? This is why it’s not a good idea to think pink kanji = on’yomi and purple vocab = kun’yomi. There are plenty of exceptions if even that is a common pattern.
There’s actually a reason it doesn’t. During vocabulary reviews you’re not rehearsing kanji, you’re rehearsing words. Unlike kanji, which have multiple readings, words tend to only have one correct reading.
An example that corresponds closely to something you’d see in English would be the word 水, meaning water. The kanji has the on-reading すい, and the kun-reading みず. If someone were to ask you: “what is the reading of 水?”, either would thus be correct. WaniKani wants to test if you know the reading it has given you during the lesson, so it will shake if you enter another one, but it will not be marked wrong because both are correct.
Now if you were to say you wanted to drink some water on the other hand, you would use the vocabulary word 水, using the kun-reading みず. If you use the on-reading and said you wanted to drink some すい, this would be incorrect. This is also why WaniKani will mark the answer as incorrect for the vocabulary reviews, すい does not mean water as a word.
This kind is kind of similar to the distinction between water and aqua in English. You can say you’re drinking water, or use aqua in a compound word to create something related to water (e.g. an aquaduct), but saying you were drinking aqua would be wrong.
I think this makes sense, and it’s a very good analogy. Still, I’d like to ask something (as a forum-only user): does WK provide context or include the option for hints in such situations, and are vocabulary tests and kanji tests separate in the SRS? Of course, I’m sure vocabulary lessons are clearly separated from kanji lessons, but when reading an actual Japanese text or writing a Tweet in Japanese, kanji readings will have to be chosen based on context, and not based on the colour of the kanji. I could be wrong, but I don’t think many people compartmentalise their knowledge into ‘vocabulary readings’ and ‘kanji readings’. Also, in case this interests anyone, there are some kanji pairs for which both the on’yomi and the kun’yomi are acceptable readings and mean the same thing, though I don’t know if they’re on WK.
From my own experience, when reading I rarely think in terms of kanji. In a normal piece of text every word will simply be a vocabulary word, so you tend to just use the vocabulary meanings and readings you learned for all the words. The only times I actually have to think about kanji readings are for words I do not currently know yet, in which case I’ll try to guess the reading based on what I know (e.g. if it’s a compound word consisting solely of kanji, I’ll try on’yomi). Usually being able to guess the reading of the word simplifies looking it up in a dictionary.
They are, there are separate items for the vocabulary words and for the kanji, which progress through the SRS independently. During reviews the background will either be purple for vocabulary or pink for kanji, so it’s pretty clear which one it’s asking for.
Hm… I guess you have a point. Still, what I meant was that you’ll probably choose a particular reading based on what the kanji appears with e.g. in a compound you know, alone, next to お or ご etc. But yes, I suppose you could say that you just see the words you know as units on their own, and don’t break them down into single kanji. That works too. As a Chinese speaker, I have a tendency to use the meanings of individual kanji to help me, but I can see how you might just work from the vocabulary you already have.
Just my view, but that seems kinda artificial. I would personally prefer to have an option for an example sentence to help me choose the right reading. However, since you said that the two test types are separate, and I don’t know many single, standalone kanji that allow for multiple readings (the only exception I can think of is 自ら, which can be read みずから or おのずから)… well, I guess users can’t really blame the system for not being clear enough. I just feel that example sentences might help with reinforcing those memories.
Example sentences for vocabulary words show up during lessons, but not during reviews. I guess it’s probably to prevent people from guessing the meaning of words from context, and instead forcing them to rely purely on their kanji knowledge.
I do tend to use the kanji to determine the meaning and reading of a word, but in that case I tend to rely on the entire kanji sequence to figure out the reading than relying on the individual kanji. For example, if I run into something like 今月, in my head I’ll parse it as now and month, yielding this month. To determine the reading I just look at the group and know that it’s just the on-readings, and you get こんげつ, pretty quickly. Looking at it as a group also prevents me from coming up with readings like こんがつ, こんつき, or いまげつ. So figuring out the readings for me is also easier when looking at them in a group. It’s also easy for cases like 大人, where the reading isn’t directly related to the individual kanji readings at all.
Oh, ok. I thought maybe reading reviews were separate from meaning reviews.
Hm… well, I guess it’s true that you have to memorise readings as a block to some extent. Japanese on’yomi for a single kanji aren’t as consistent as readings in Mandarin, even if on’yomi across kanji containing a particular common are far more consistent in Japanese than in Mandarin, so learning blocks (or ‘groups’, as you said) is important. Never really thought about it though. I just do it.
Out of curiosity, do you find it helpful to memorise single keywords or short phrases as the meaning of a kanji? That’s how WK works under the kanji section, right? Part of why I don’t use WK is because I feel like learning keywords would be really limiting, at least compared to the multitude of nuances that can be assigned to each kanji in Mandarin (which you tend to find in Japanese anyway). That doesn’t mean WK is bad, since I tend to try to extrapolate and link other meanings to the ‘core’ meaning of each character. The problem is just that some characters can’t be summarised with one word.
PS: just saw comments on this other topic
I guess these are moments where examples sentences might be helpful for reading-only reviews? Since it seems the reading WK decides to focus on can be changed by the WK team.
I think overall it’s helping me learn more kanji than I would otherwise be able to. I do agree that sometimes the nuance might be lost if you only use a single translation as the meaning of a kanji. However it does tend to speed up the learning process a lot. Getting one meaning down for the kanji tends to help you at least understand what it means, usually during the vocabulary lessons where the kanji then shows up in compounds you can figure out the nuances of the kanji a bit more. It also helps that if you understand a rough meaning of the kanji, you can also guess the meanings of words that show up when you encounter it in a compound in the wild.
I think one of the best examples might be the 気 kanji. Wanikani lists the main meaning as energy and has a secondary meaning as spirit. You then encounter the kanji in words like 電気, 天気, 人気, 元気, 本気 etc. So I know the kanji represents something more similar to spiritual energy, although the actual concept might be hard to fully explain in English. But learning the basic definitions at least gave me enough of a foothold to understand those words without having to read a multi-page explanation on the exact nuances of the kanji, which would have probably just confused me.
The “main” kanji reading they choose can change over time, although for as far as I know this only happens rarely.
During a review session which has a kanji or vocabulary word, the actual prompts will be split out over the review session, but the answer will be marked correct or not based on whether you can do both. So for example, say the vocabulary word 人, comes up. During the session it might first prompt you to give the reading, in which case it would require you to enter ひと, and a little later in the session it might ask you for the meaning, in which case you would have to enter “person”. The entire vocabulary word will only advance to the next stage if you do both correctly during the review session. In theory you could give a context sentence for only the reading part and not give one for the meaning part, although this would seem slightly inconsistent to me. Although I assume someone probably already made a userscript to do just that.
Hm. Interesting system. I personally would prefer to do meanings and readings together, because I feel that they should be treated as a single unit as much as possible, and that really is how I study when I try to learn new words. Then again, I usually study using a dictionary, so… it’s no surprise that I always have both in front of me. You’re probably right about the user script though. It’s pretty amazing how much people can customise WK, provided they have the knowhow for it.
Mhm. Fair enough. I tend to link everything to a ‘core’ meaning through lateral thinking, so I follow a similar procedure, I suppose. Just that my ‘core idea’ might be a bit more abstract. For example, I remember the core idea of かける as ‘contact’. However, you’ll probably never see that in a translation.
I hope you don’t feel that I’m messing up your mental ‘model’ for 気 by saying this, but the basic meaning of the kanji for me as a Chinese speaker (and 大辞林 lists this as the first definition as well, so I think that means Japanese people would agree) is ‘air’. That’s the most literal meaning of the kanji. ‘Energy’ and ‘spirit’ are just extensions of that. How? Well, I’m not sure, but I tend to see it as something that flows in and out of people and as something that envelopes everything (which is what the air actually does). Something like the expression ‘breath of life’, I guess? And so, when you think about the fact that 気 is also used to represent a person’s life force, it makes sense for it to be used to describe mental or physical states (e.g. 元気, 本気) and ‘flows’ (e.g. 電気, which would be electrical flows, and 人気, which would be the flow of… attention (?) from people). In Mandarin, in a context related to martial arts or geomancy, or even just as part of a joke, it’s not uncommon to mention 杀气 (the second character is the Simplified Chinese version of 気) – literally ‘killing qi’. ‘Killing energy’, I guess? You could explain it as the killing intent that emanates from a person who intends to cause harm. Think of the pressure you can feel when someone gives you a ‘death glare’. That’s what it is. It’s not so much energy as it is ‘aura’, a word which itself comes from a word meaning ‘breeze, breath’ according to Oxford. I really think that’s the closest word in English: literally, it’s the air, but figuratively, it’s a whole lot of things that are able to flow like the air.
No I actually completely agree with your definition of 気. The way I tend to visualize it internally is probably closer to aether, which I can use as both the connection to air as it related to terms like 天気, as well as a “life energy”-like concept which I can use as the connection to words like 元気. It’s just kind of hard to fully explain the concept as I find there isn’t a short clear-cut definition of the term in English.
Good pick! I didn’t think of that. I think that’s a very apt analogy, even if it’s not some sort of ‘fifth element’ as aether was (if memory serves) in Ancient Greece.
True. To be honest, my understanding of it is probably just a result of growing up watching Chinese dramas on TV, especially dramas involving martial arts and traditional medicine. My way of understanding it is visualising some white, smoke-like substance that flows, and which can be controlled to a certain extent. Probably the result of all those scenes in martial arts dramas where people manipulate their chi/qi/ki (depending on one’s preferred transcription) to do incredible feats.
For me it’s a bit of both but that’s part of the process. The words I’ve internalized and know really well are read as entire chunks, but ones that are less familiar I tend to still read each Kanji.
Only one reading is taught during the lesson, usually the primary or most common reading. There are still quite a lot of vocab that use other readings and those will be taught with the vocab word.
And I would have to agree with @BIsTheAnswer. Having example sentences would trivialize remembering it.
It’s not quite like that. While WK does give you a primary meaning, the nuances are found in the vocab that they provide. Granted, it’s not the full range of nuance, but it’s enough to get you reading so that you can begin to expand your internal definition.
But learning the primary meaning, or keyword, is critical because it gives you a foundation to build on. After using WK for almost 2 years now, I’ve found it much easier to focus on that foundation first as I learn a Kanji. Because that makes it much easier to later on refine the meaning or add more nuance to it as I learn the vocab or encounter the Kanji while reading.
I think I was just trying to come up with something that would imitate how I would learn the correct reading for a kanji ‘in the wild’. When I was given vocabulary lists in school (even if I admittedly hated memorising them), Chinese words generally came with readings, definitions and examples. It might also have something to do with the fact that the reason some Japanese readings have stuck in my head is precisely because I have a particular example etched in my mind (e.g. a few days ago on the forums, I gave the example of a line from an anime OP that makes it impossible for me to forget how to read 教えて). Then again, I guess such memorable sentences or phrases have to be personal, so there isn’t much of a point standardising them. I can also understand the desire to actually have to make an effort to recall readings and meanings. Memories become more strongly anchored when more effort is made for recall, provided recall is successful. Still, ultimately, perhaps why I feel it might be ‘artificial’ is that when kanji are read or used ‘in the wild’, there will always be contextual clues. I might choose a particular reading based on the meaning I wish to express, for example.
In all honesty, as I said above, I do a similar thing as a Chinese speaker by using literal meanings as a starting point and linking up with other meanings using lateral thinking. I guess the only difference is that WK’s keyword may not be the most literal meaning, but instead something that works for the vast majority of vocabulary words it will cover.
In any case, I think a lot of my comments come from my background as a fluent Chinese speaker. (Well, OK, I’m definitely getting less fluent by the day, since I almost never need to use Chinese.) I already have a structure in my mind for kanji, and I don’t want to change it since it works well. Plus, my awareness of other nuances probably makes me feel like ‘keywords’ are inadequate, even though I’m pretty sure they’re a very good starting point for beginners. (That’s also why I can’t understand why some native/fluent Chinese speakers choose to use WK to learn Japanese kanji. I’ve definitely seen at least one other WK user stated that he/she already spoke Chinese, which really surprised me because I feel that Chinese and Japanese maintain almost all the same nuances across kanji.) If I didn’t speak Chinese, however, I might have considered something like WK for kanji, even if the intensity of the reviews beyond the earliest levels might put me off.
For the meanings WK might not mean much to them, but you learn readings here, too! on’yomi that may come from Chinese, but sound different in Japanese on account of not having tones, not to mention being adopted from different eras with different pronunciations. And then there are all the kun’yomi. Wk is actually really good at teaching readings.
I’m sure it is, but I figured Chinese speakers would be used to memorising new readings anyway. Having multiple readings for a single kanji isn’t something unique to Japanese. Also, while it’s true that Japanese pitch accents don’t usually match tones in Mandarin for on’yomi, there are quite a lot of patterns that seem to work across on’yomi e.g. the pitch tending to drop shortly before a ん. Certain pronunciation changes relative to Mandarin are fairly consistent as well, meaning that on’yomi can be guessed if a speaker is able to see conversion patterns. I rarely get surprised when I check how a kanji compound’s on’yomi is supposed to be pronounced.
I guess my point is that I don’t really see why native Chinese speakers would need a kanji study system when they’ve already acquired all the necessary study skills when learning Mandarin or some other dialect of Chinese. Then again, maybe I’m an outlier since I grew up hearing other Chinese dialects, so ‘conversion’ comes naturally to me since I often tried to figure out what my grandparents were saying. On’yomi endings like つ and く also come fairly naturally to me because Chinese dialects other than Mandarin use lots of ending consonants, which Mandarin has mostly got rid off. Perhaps you’re right: they’re here for the readings, and perhaps also to see if Japanese contains nuances for kanji that differ vastly from those in Chinese (which is occasionally the case).
I don’t suppose WK includes recordings for readings or a text-to-speech engine? I think that would be valuable, but given how expensive the NHK Pronouncing Dictionary is, I strongly doubt anything like that’s been done.
In any case, the tools each of us uses are a matter of personal choice. If they find WK useful, then I’m glad that they found this resource. I personally tend to avoid flashcards, and have a preference for context-based learning and dictionary searching, so WK probably isn’t something for me.
All vocab lessons have audio recordings from a female and a male voice actor. But have you thought about at least going through the first level? Maybe some of your wonderings about the workings of the program can be answered that way.