Edit: this isn’t a “criticism” I’ve been learning very well on the current system, it’s more of a discussion on how does it actually work
Because I was under the impression that the system worked by keeping the times between kanji reviews longer and longer in order to improve memorizations. But I’ve been noticing that in this “time between reviews” I’m pretty much reviewing the same stuff over and over again in vocab.
Doesn’t that mess up a big the tictoc clock of the SRS they were trying to go for?
I mean, it’s not that it isn’t working to memorize kanji, but is this how it was meant to work?
It’s got to the point that I feel like I’m memorizing more from the repetition through vocab than with the proper kanji.
What do you mean by this? Yes, vocab are there to reinforce your learning of kanji. Ultimately kanji aren’t words, and to read Japanese text you want to be able to read the vocabulary words that kanji are in.
Once you finish the Lesson for an item it becomes Apprentice 1. There is a 4 hour wait until that item appears in your Reviews. If you get that Review correct, the item moves up to Apprentice 2, and your wait is now 8 hours.
But assuming it’s a level beyond guru and and the wait time is 4 days and you already have vocab. in that 4 days “wait” time, you’re gonna see the kanji on days 1, 2, 3 in the form of vocab. Which means by the time you review, you’ll have reviewed the kanji after just 1 day from the last review, when it’s supposed to be 4.
Which is why I was thinking that maybe the sheer amount of Srs vocab containing certain kanji is doing more in helping me retain the meaning than the SRS of the kanji specifically.
Or, in other words, quantity is being more useful than the spaced repetition which does not seem to happen all that well, when it comes to being properly spaced.
Yeah, you’ve pretty much got it! Yes you are reviewing vocab before your next review of the kanji it contains, but that’s the point. Essentially, if you’re not learning to read vocabulary words, you’re not learning to read kanji. They mostly come hand in hand. This is also referred to as learning kanji in context—you’re learning it in the context of the vocabulary that use that kanji.
As long as you’re honest, and WaniKani forces you to be honest, then an SRS system will sort itself out in terms of your knowledge. If you forget the reading of a kanji, but remember it’s associated vocab, you don’t need to stress. It’s more the reverse you might be worried about.
When you know a little more kanji, you may want to start reading, and that’s where you’ll really start encountering vocab, and that WILL be doing what you’ve described above. So, yes, it’s just how the system works, I guess. If you purely want to study kanji meanings without associated readings, that’s what Remembering the Kanji does. But I feel like it’s easier to just use the method that WaniKani employs.
In theory, yeah, the “duplication” of kanji + vocab and some other bits of overlap “pollute” the SRS.
Basically, sometimes you get “false positives”. Like maybe, I only got a vocab right because I had just reviewed the Kanji, or because I had just gotten a similar word wrong. (e.g. getting 左翼, “left wing” wrong, then getting
右翼 correct, even if I really didn’t remember 翼)
Or you might just happen to run across the word in outside-of-WaniKani-studies the day before it came up for burn review. Does that mean you should stop studying outside of WaniKani to preserve the purity of the SRS? Of course not.
If you burn a word and think you don’t really know it, you can always just revive it and study it again. It’s going to happen and it’s ultimately up to you to “keep yourself honest” about what kanji you really know.
There are times that I get kanji wrong, no matter how well I do on the vocab. Recognizing the kanji in isolation is different from recognising it in the context of a vocab card.
I get what you’re saying, but the readings for the kanji and the reading in vocab can be different. It depends on the vocab in question.
Each item has its own SRS clock ticking, so they’re treated as independent of each other. So, to answer the original question: no WK does not take into account reviews of other items.
I wouldn’t think too hard on it. The whole point of SRS is to get you exposed enough times to the readings and meanings that by the time you burn an item, it stays with you. If you do more repetitions on your own, such as with the Self-Study Quiz, that can only increase your retention in the long run.
I think it only makes sense to be wary about “polluting the SRS” as you put it, when it comes to burn reviews. That means you should avoid doing self-studying of those items before they’re up for burn, since those last months of waiting are meant to test long-term retention.
But, as for the lower SRS-stages, it hardly has any meaningful impact if you study more. That can only be a positive thing.
Think of it more in terms of efficient time-management: you don’t have to study each item more than the reviews given to you, but you absolutely can!
For example, moving forward you’ll acquire more and more leeches, and you’ll probably wanna study those items specifically. But, to not disturb the SRS too much, do so when they’re in Apprentice and Guru (and not Master or Enlightened).
I hope this somewhat alleviates your original concerns.
The purpose of the longer and longer review times is to avoid wasting time reviewing things you aren’t going to forget soon. Less reviewing doesn’t make you remember things better. So if you run into something a few extra times along the way, that’s just making you remember things a little better, if perhaps not as efficiently.
That is the part that gets me wondering whether there wouldn’t be a way to make this all more efficient.
This spaced repetition thing was meant for you to memorize it with the least reviews possible, but with the kanji+vocab combo it ends up making you review the same stuff way more than that, which makes me wonder if it even makes sense to apply these spaced repetition standard “times” the way they do here. Perhaps some other kind of proficiency testing
When koichi becomes a billionaire maybe they could restructure it in a system where you can level up the kanji if you get the reading right in a jukugo word (like if you get 友人 right you level up both 友 and 人) in a level system that obviously would look very different from the current one.
Same thing as you could level down a kanji if you get the reading wrong in a word, but only if you type a reading that is not an alternate reading for the same kanji.
This would all mean less isolated kanji reviews and a way to check for kanji proficiency through vocab reviews
If you purely want to study kanji meanings without associated readings, that’s what Remembering the Kanji does. But I feel like it’s easier to just use the method that WaniKani employs.