I have a lot of time on my hands so that’s why I was able to do what I did. Like I said though, I don’t think it’s an effective way to learn kanji. Matt’s idea of doing RRTK is that you do larger intervals (lazy kanji) and you don’t spend much time on them. I didn’t follow this and I did WK’s intervals and learned 30-35 kanji a day. Also, my retention of the first 600-700 kanji is close to 100%, but the last 300 is very low. Anyways, I was just posting that update to say my thoughts on doing it.
2 Likes
Keep going!
1 Like
I think this is a good way. I’ve actually been wondering if this may indeed be better than WK’s way of learning kanji individually first then use vocab as reinforcement. It’s hard to tell which way work better for most people without doing experiments with students and get the data.
2 Likes
For myself, learning the vocabulary on their own would have been a MASSIVE time investment… and I would have quit.
I’m curious how you are measuring your retention. Are you inputting the hiragana, or just going on general pronunciation? [Maybe you’ve answered this and I’ve missed it, but maybe not?]
Based on the way your personality comes across online and what you’ve written, I can understand this. The most important thing for you is that you’ve been able to choose your own learning path (with some guidelines) and have stuck with it. WK might’ve kept you annoyed by its pace and pre-selection of vocabulary. For someone like me, it has helped to have had the kanji/vocab process just handed and presented to me in order to push me forward, and the site’s game-like addictive quality has kept me going and thus my vocabulary and kanji knowledge are much greater than they otherwise would be had I not continued with it. What I find interesting is that my kanji level is above 700 and my vocabulary level is above 2,300 (according to WK stats) and there is a built in approach through the site to make sure everything at some points gets closer to that 90~100% retention rate (through the mechanism… not necessarily applied knowledge).
As others have suggested, though, I am working on consuming media + some amount of reading (and some grammar) + natural language in a somewhat natural immersion setting but also am trying to take steps to be an active learner.
Hopefully you’ll be able to continue beyond where are now and your lower retention items will rise higher. Also, I hope you’ll be able to find a way to apply this knowledge with native speakers soon!
2 Likes
The thing about RTK is that you only learn the meaning of the kanji. So I mean this doesn’t really answer your question too much, but if I get the meaning wrong then I fail the card. I used to be lenient on this, which is probably why I’ve had to restart the deck, but it’s now if I get it wrong I fail it. I’m also learning the readings through vocab if you were wondering that. Which I’m actually surprised, I know quite a bit of readings (more than I thought I did) by only using sentence decks. The thing is, I don’t question myself when recalling the reading of the kanji that has multiple readings. I can sort of figure out based on the context.
Yeah and I think this is what makes WK so great, that’s why I really like it. There’s a part of me that wants to continue doing it, but I can’t. Doing sentence decks (which I’m no longer doing) made more sense to me than just learning the individual word. I’m now doing audio decks which make way more sense than the sentence decks. Audio decks you have the individual word on the front (that way can’t memorize the word from the sentence) and you are trying to recall both meaning and reading of the word. On the back is a picture (if sentence mined), an example sentence, and a glossary for that word.
Yeah I’m trying for 1-2 hours of immersion a day. I had been getting close to 3 hours the past few weeks but I noticed I wasn’t listening well enough, so I cut it down and upped my studying time.
2 Likes
Is anyone a patreon of MIA? I’m a little confused as to what membership includes, and I’d love to know how it has been helpful!
I’ve found that I do not have any clue what my WK vocab words REALLY mean until I see them in context. It’s nice that I can usually work out the reading, but rote memorization of isolated words is a bit…let’s just say I’m in the camp that thinks there should be a “lite” accelerated WK version with half the vocab. Then combining with MIA makes more sense.
On the other hand, I’ll take WK over RTK any day. WK + plus a good sentence deck + a good text book + native readings is a good combination, provided you don’t have a full time job or something silly like that. I’m using the MIA N4 Tango deck and the Nukemarine N3 Tango deck right now, and they both reinforce the WK vocab quite well, along with new words and grammar practice.
2 Likes
Doesn’t WaniKani provide example sentences when you learn a new word?
I found when I did the tango book for N5 that I wasn’t really learning the word that well. I found myself memorizing the sentence structure (how it looked, what came first, etc.) and I would basically know the word based on that (rather than getting the meaning + reading from the word itself). Due to that, I would come across the same word in other sentences, and I had no clue what it meant. I ended up quitting sentence decks because of this.
Now, of course, sentence decks are pretty helpful if used correctly. But, another reason I stopped sentence decks is because they take up a lot of time. Doing single word vocab is a lot faster since there’s really no reading involved. I saw a comparison in time spent studying between a single word on the front vs a sentence on the front, and obviously the person doing the single word was significantly faster in their reviews (typically 10 seconds faster per card). If you think about that, the time you save doing that could be spent toward reading actual Japanese content. Which to me it would make way more sense to actually spend time in Japanese content rather than made-up sentences, especially since it’s all unique and not the same sentence every time.
If you want, you can read the sentence card section of this website here: https://www.animecards.site/
Talks about the effectiveness of sentence decks and talks about an alternative.
Anyways, good luck! Just thought I would share that.
Yeah, I don’t find the WK sentences that useful, since the sentence itself is not SRS’d. I agree about relying on sentence context to remember a word, which is why I think the MIA method alone is questionable. WK + sentences though, gives a good synergy between the two. The nice thing about sentences is the ability to identify collocations, which I wish WK used instead of just isolated words. Honestly, I think all vocab should come with SOME context so as not to try to 取る an お風呂 or other “eihongo” nonsense. 笑
Having said that, I am thinking about suspending all of my Anki decks to focus on just reading. I use Satori Reader for short graded readings, and just bought a copy of コンビニ人間 to try out. I may try that for a month or so and see if I feel like I’ve improved. I’m kind of mentally stuck in the idea that I need to be “doing” something like exercises or grinding through SRS to feel a sense of progress, even if my understanding of the language might move more my just kicking back with a book. Well, it’s not quite kicking back if I’m struggling to read it, but still…
1 Like
I found the best method for me after finishing wanikani was reading books and highlighting every N+1 sentence I could find. After I would go through it again manual inputting every sentence into a anki flash card. The process of typing out every sentence seems to have helped out with a lot of my grammar issues as well as slowly coming to understand which words pair well with each other. I think you get out as much as you put into it. I found that doing this from books I read it made it easier to associate meaning and context to each individual sentence. If you simply download a premade deck you can’t associate your own memories with each flashcard so it’s harder to remember. I highly recommend sentence mining yourself and physically typing out each flash card and not just copy pasting sentences from something like subs to srs.
2 Likes
Yeah, well that’s sort of the idea with the link I provided. However, instead of a sentence on the front, the sentence is on the back. The main idea is to save time that way more time can be spent immersing instead of studying. On the back is the sentence that was mined, along with audio of that sentence you pulled, and audio of the word itself.
Thanks for the suggestion. I’ll definitely give this a try farther along in my studies. I still struggle quite a bit to hear words I know while listening. I’ve also picked up reading, but I tend to skim over things I know too since I’m still overwhelmed by everything I still don’t know. I still haven’t started sentence mining because of this.
1 Like
Thanks for posting. This has been an informative thread. I had not heard about MIA. Immersion is important, but exclusively doing immersion, or even mostly doing immersion is…well…it may not be the best use of a person’s time. Asserting that because children learn a certain way, adults should as well, is just someone’s opinion and not actually a fact. Adults have a unique advantage over children in that they can learn a new language in terms of one they already know and it seems like an optimal learning technique would leverage that advantage instead of just throwing it away.
Actually, if I had to bet on which technique will prove itself to be the best approach towards learning a new language purely in terms of efficiency, I’d put my money on the same technique used to train AI - study thousands of complete sentences in two languages at once. This is actually the approach that Glossika takes. However, I found it to be mind numbingly boring. WaniKani is fun, and as one part of a multi-pronged approach can easily lead to fluency in a relatively short period of time.
Anyway, so many people have already said it in this thread, but its worth repeating - the best technique for a given individual is the one they can stick with over a long period of time. For me that is vocabulary building, alongside gentle immersion ala services like Fluentu. For others that may not be the case.
Fear of a better option is real, and a much greater threat to learning objectives than a less than optimal learning strategy.
2 Likes
In fact the whole “kanji with the same radical often have the same reading” thing is actually a double edged sword. When doing WK vocab reviews, I often find that I intuitively know the readings even though I can’t remember the kanji or don’t notice which of many similar looking kanji it is. Then when I get quizzed on the meaning, I’m lost because I didn’t actually recognize the kanji, just the compound inside the kanji which gives the reading.
Interesting. How is your progress now with the MIA path?
1 Like
Well MIA itself doesn’t even exist anymore lol.
Somewhere in September 2020 the owners had a pretty bad fall out and they went their separate ways.
Lucas created Migaku and Matt created Refold.
Anyways as for me I’m right now at 497 hours of total logged immersion. That includes 85 hours of reading raw text (i.e. novels). However I started MIA in november 2019 but didn’t actually start logging my immersion until May 2020 so actual immersion time is probably closer to 800 hours.
I hit WK level 60 around November. In total WK took me 348 hours to complete, in 400 days. I kept doing reviews for about a month after that but it started getting really boring. I had never taken a break from WK either so in December I put my account in vacation mode.
Right now I can read light novels with easy on the Kindle. Even if I don’t use the dictionary I can follow along quite easily.
ps. I ONLY read on the Kindle because it’s by far the most efficient reading medium I’ve come across so far.
I even bought a new Kindle Oasis (even though I had an old Kindle Paperwhite for Christmas).
Watching anime for the genres I’m familiar with, with jp subs enabled is almost effortless. Dramas and harder anime still require me to look words semi-frequently if I want to know exactly what’s going on.
Turning off subs and just listening though, even with regular anime, is bit challenging. It’s clear my reading ability has far surpassed my listening ability. But according to Matt and others it’s pretty easy to catch up.
So starting last week I’m started to just watch easy anime series like One Piece with no subs, no LLN (Language Learning with Netflix), nothing. Just raw with no supports.
ps- When I started Shirokuma Cafe was a series I found really efficient for immersing. But I only watched half of the series back then, and last night I picked up watching from where I left off with no jp subs. I had pretty much near 100% comprehension. 
For some drama and harder anime I will switch on LLN and do sentence mining.
I typically mine 10 sentences per day from anime/drama or books.
Speaking: Haven’t really concentrated on speaking much. I do take a casual Japanese Language Class and I speak there. My ability clearly surpasses anyone else in the class due to my vocabulary and due to the fact that I can read. Over the holidays my school was on break so I did do some casual conversation on italki. Found out I can express myself fairly easily.
So yeah anyways these days my immersion goal is around 4-6 episodes of anime or JDrama per day + 2-3 hours of reading. Of course many days I don’t hit these goals lol 
My SRS usage is about 40-60mins of anki. WK is on vacation mode and not sure if I’ll ever reactivate.
I listen to podcasts in Japanese most of the time I’m driving, shopping, doing light-work outs (sorry but when I workout seriously I still need Metal, Rock, Rap etc blasting through my earphones lol 
)
Grammar: I started grammar at the language school and I casually went through tae kim’s, watched some of nihongo no mori’s videos and did some bunpro. But right now I don’t do any grammar… Haven’t touched bunpro in months even though I still have an active sub lol.
Feel at this point I can essentially pick up grammar as I go simply through immersion.
I might eventually come back and run through bunpro though.
ps. Nihongo no Mori also have a new site even though right now they only have N1 lessons. Might go through that once they have N3 or N2 completed.
6 Likes
Interesting, thank you so much for your feedback!! I’ve been recently scrambling for real experiences from people who have used WK and MIA (Refold) methods seriously. And from that, I can say your feedback was the best. I appreciate the detail, I’ll try and shape my study habits in a similar way.
And one question. Would you say sentence mining and studying sentence cards helped you the most? If not, what would you say was the most beneficial to you?
1 Like
No, absolutely not.
Sentence mining is just a tool. Immersion is the only thing that makes you acquire language as far as I’m concerned…
However if you are asking if, among the various forms for SRS, for example grinding pre-made decks, vocab cards etc., whether sentence mining your own cards and then SRS’ing them is the most effective form of SRS, then yes think sentence mining is the best form of SRS for most cases.
But in the end SRS isn’t immersion. All of Migaku’s add-ons for anki have one overall goal: minimizing the time you spend on SRS so you can get more immersion time.
2 Likes
Interesting! I don’t want to be contrary here, i know that immersion is great, i use it a lot as well, and it seems like this approach works well for you. But i just want to add my perspective here, and for me SRS has been responsible for nearly all my progress.
I do WK, Bunpro for grammar, and vocabulary from my immersion in Anki. (I should add more sentence cards)
I went from usually not even being able to write down what people said (recognizing word boundaries) to understanding ~70% maybe in Anime purely with SRS and a very low amount of immersion, maybe half an hour or an hour a day at most, and not regularly.
To me, the biggest obstacle in immersion, in understanding what is being said, is almost always vocabulary or grammar. If i don’t know what an abstract concept like ぼっとう (没頭, ‘immersing oneself’) means, how am i supposed to pick that up from absorbing a random sentence (without translation)?
Now, i know what one could respond to this. I know that immersion is fundamental. I know that learning grammar formally, even with Bunpro which always uses example sentences, is not the same as absorbing it and gaining intuitive understanding. I know that you usually don’t get the nuance of a word from a dictionary, but by seeing how it is used in different contexts in immersion.
But I think especially for the beginning, to get a basic vocabulary and understanding of grammar, heavy SRS is hugely efficient.
And even now that my vocab is around 6000 or so (still not including an overwhelming majority of very common words), I find that heavy SRS is highly beneficial. Of course heavy SRS is not fun, and you can easily go into review debt. But once you’re at 6-10k words and most common grammar, you can go a bit more easy on it 
Then, after you have a general idea of vocab (vague meaning) and grammar, you can much more easily pick up the nuances in immersion.
But I like a formal approach in language, and maybe SRS is especially suited to me, while for others immersion is relatively much more effective.
3 Likes
Isn’t everyone just pretending his own experience is universal when they claim one method is better than another?
Like have there actually been any large scale studies testing how much time one needs using mainly immersion or mainly SRS? Pretty sure not.
Mostly likely scenario is that some middle ground is the most efficient way.
Imo you start out with “Lots of active study and very little immersion” and transition to “A lot of immersion and very little active study” over time. And I’m pretty sure anyone claiming to know the ideal curve is just being unscientific.
7 Likes
Good point(s)! This last one really makes a lot of sense to me, “much study and little immersion first, then transition to little study and much immersion”. (and to/for me, as long as i’m below 10k of vocab, it’s lots of SRS study)
2 Likes