Can i rely on kanji koohii for mnemonics

I’m doing RTK one right now and come up with some stories but mostly use the ones provided by the users, while reading them out loud and visualising them but still i keep forgetting kanij. Do you know of any research , or your own experience, saying that mnemonics are better used if you come up with them yourselves

1 Like

Whatever works for you. Don’t overthink it, you’ll probably use your own mnemonics for some, theirs for others and probably no mnemonics at all for a good chunk of the rest.

It’s just a temporary crutch anyway.

4 Likes

I’m around 1200 kanij now and heard people say that you can learn the first 1000 kanji in 2 months and the other 1000 kanji for 2 weeks but haven’t reached that point yet. Is there any method for that

Whoever told you that gave you very important information: that you should dismiss their advice completely.

9 Likes

it was meant more like at around a 1000 kanji you get so proficcient that you get so much faster at learning them but that isn’t the case for me.

1k kanji in 2 months is really aggressive and not very many people can manage that without burnout. then another 1k kanji over 2 weeks is worse to boot.

3 Likes

I think it becomes a bit easier to learn kanji past a certain point because it builds on top of your existing knowledge of radicals but at the same time the new kanji also becomes less common and will be harder to practice.

So overall it’s easier in some ways and harder in others.

3 Likes

i mean i did 1200 in one month but what i paraphrased there was mostly a figure of speech

And will you be able to recall all of them a month from now?

Also, with RTK, are you doing just meaning or readings too?

1 Like

Frankly I can’t imagine what would possess someone to decide to spend even more time studying kanji one you’ve already learned all the super common ones. Assuming for the sake of the discussion that you can actually manage to meaningfully learn 1000 kanji in two months (a cool 15 kanji/day), the last thing I would advise after that is spending even more time studying 60+ less common kanji every day. It sounds miserable and not very efficient.

If you’re doing them in RTK standard order then they’re not in anything even close to frequency order, so you kind of have to do the whole thing rather than bailing out early.

4 Likes

I think the idea is not 1000 kanji learnt in two months, but rather the first 1000 kanji will need 2 months each (you can learn several in parallel, but it will take two months for each ho be learnt), whilf for the next 1000 learning is faster for each one (presumably because you can make linkx to already known ones).

Well, that will make sense.
Learning 2000 kanji in 10 weeks doesn’t makes any sense.

2 Likes

I wasn’t aware of that. Seems odd.

I don’t see how anyone could feel like they learned 1000 kanji in two months from scratch. Are we just talking about “learn kanji = associate one English keyword with each kanji”?

That’s more believable, but it also isn’t anywhere close to everything you need to know, so calling it “learning the kanji” seems misleading.

1 Like

imo, whatever works. RTK may help with keeping each Kanji distinct via a keyword each, and giving “some” meaning to Kanji via the keyword. But then, in an actual vocabulary, the conferred meaning of the Kanji may be different from that keyword.

Not to mention that a keyword may got interpreted differently from the intended meaning. Otherwise, a rare English keyword that doesn’t have a stable meaning in my brain.

But then some of those may be re-learned later on with vocabulary learning, which doesn’t need to be in Kanji order.

It may help though, that RTK is more meticulous with Kanji details than WaniKani, and keywords are more distinct.

1 Like

The idea behind RTK is that you go through all the kanji quite fast up front, associating each with a single keyword, and practicing keyword-to-kanji recall. The author states in the introduction to the book that (as a beginner student of Japanese) they originally devised the method and learned 1900 kanji meanings with a month of full time study, with no particular further need to review after that. The theory is that this then puts you in a similar position to a Chinese student of Japanese, who doesn’t know the Japanese words but does have the ability to recognise the kanji and know their core meanings, and you can then learn vocabulary separately in a way closer to what you might for any other foreign language. Since the author thinks it won’t take too long to go through the whole set, he therefore prefers to order them purely for efficiency of memorisation, with all the characters for a particular component clustered together.

Personally I have a feeling that Heisig’s visual memory is perhaps rather stronger than average – he suggests in the introduction that with a good visual picture fixed in your mind based on the keyword and the components, “many characters, perhaps the majority of them, can be so remembered on a first encounter” with no need for further drill or review. Certainly my own mind does not work like that, but then I am very non-visual as a thinker, close to aphantasic.

When I was doing it I mostly leaned on the kanji koohii site’s user contributed mnemonics, picking through the top rated ones to find one that seemed most memorable to me, but sometimes I didn’t like any of them and wrote my own or tweaked an existing one. That worked well enough for me at least to get through the whole set using the kanji koohii SRS system with keyword-to-write-the-kanji recall.

My main issue with RTK was that I found it didn’t tie very well into the rest of my Japanese study. The problem I was trying to solve at the time was having difficulty recalling how to write words when handwriting Japanese, but without any link from the Japanese word I wanted to write to the RTK keywords of its characters, being able to recall kanji from keyword wasn’t super useful. Subsequently my requirement to handwrite Japanese went away and so I became less interested in trying to find solutions to this problem (though I did get about 1000 characters into a “RTK from Japanese keywords” deck I was working on; also not convinced this is the best way either).

1 Like