Evgeny-Mnemonic techniques for writing kanjis – any tips?


I would like to create some mnemonics like Evgeny Uskov (who succeeded – several times!- in Kanken 1), but I don’t know how to start it properly.

References :
https://roshiajin.jp/kankenlevel1/ is worth reading!
https://community.wanikani.com/t/anyone-tried-kanken/17115/ is the main post for preparing Kanken

@efugeni Thanks for taking so much time sharing your thoughts.
And of course, if you add time to read this topic too, it would be great!

So, Evgeny breaks components down kanji into invidual parts, and then he applied his own “original” recipe:
for example kanji
彬(refined, gentle) becomes ki-ki-san
簪(ornate hairpin) take-kiba-kiba-hi

What’s the goal ?
With these mnemonics, you avoid information load, and you give a “keyword/expression” by kanji. It could be less memory intensive than the story version.

My goal ? Enjoying learning kanjis and know how to write them. (Writing is a challenge by itself!)

Questions :

How would you proceed to emulate such mnemonics ?

Would you take https://kanjialive.com/214-traditional-kanji-radicals/

and give each traditional kanji radicals a small “name” – they already have a name but maybe some names could be shortened ?

Would you take kunyomi or onyomi? or would you just go “wild”?

Mnemonics are man-made creations, so ヘアピン could be nice for𧘇…

Let’s take for example the kanji 園
What are its constituents:
囗 and 袁
囗 the first one is くにがまえ (kanji box or country radical) – so the name of this component could be “kuni”
the other one袁 is a long kimono
So how do I remember “long kimono” with this method ?
袁 is a variation of衣
袁 could be simplified to 土 口 ⾐
⾐ clothing ころも is not a perfect choice (since the top part has disappeared) but well… 𧘇 is a variant
So it would be 園 kuni-tsuchi-kuchi-koromo ?
let’s try again: tsuchi is a bit long, let’s call it “do”
So 園 would become kuni-do-kuchi-koromo - Would it work ?

I would be tempted to remember:
kuni-en – but I would have to remember that “en” is the long-kimono part ^^

Would I need to keep a table of names I chose along the way for each radical ? (so as to keep naming consistant across kanjis ?)
If i choose some shortcuts (“en” for 袁, there’ll be plenty “en” along the road so it’s a tough choice!), would i have to update “mnemonics” (change them) as i learn more kanjis ?

Any tips ?


I wouldn’t rely on mnemonic being proper and inambiguous, but I might adapt it later until I get the writing correct every time. Also, there are writing details beyond Kanji components, and that might be incorporated into a mnemonic as well.

I don’t think mnemonics would be all vocal, but probably also visual and muscular/sequential. Location information might also be in the mnemonic, but I haven’t yet put stroke orders or direction / stroke type in the mnemonic.

袁 = 土 口 (⾐ minus the lid)
⾐ = a lid with some exaggerated bottom writing (𧘇, which also appears to be tofu radical).


I tried running a few experiments with grade 2 kanjis:

R 亠 nabe
交 nabe-chichi
京 nabe-kuchi-shō
会 yane-ni-mu (yane is a variant of hito: ひとやね)
R乍 saku (not a radical)
作 nin-saku
体 nin-hon
儿 hitoashi
元 ni-hitoashi
兄 kuchi-hitoashi → ani
R 冖 wa (couronne) - wa(kanmuri)
売 samurai-wa-hitoashi
R 冂 maki (fem name)
内 maki - hito
公 hachi-mu
同 maki-ichi-kuchi
R冫 nisui
冬 nomata-nisui

The interesting part is being interested in radicals and looking for components.
It’s true, mnemonic is missing location (top-bottom, left-right, around)… but that reminds me of a game with kanjis components and you have to write the whole kanji with them. As you said, it also lacks tome/hane…
I have a feeling going from words to kanji may be a reversed way (one not used by Evgeny i understand correctly), but i like the idea too: if you have seen/written some times: 元々 i think you can get 元 visually / with a few writing repetitions (by hand).

So, for these kanjis, it may be a bit overkill, but i think the concept is worth exploring.
Naming-convention consistency would be the hard part. How to define a good mnemonic keyword without going too far…


first, thanks for mentioning and referencing

the only thing i can add to what i have already said/wrote on this platform and elsewhere with regards to the topic of kanji mnemonics

  1. obviously, different people have different memory traits, and respond better to different cues:
    hence, if something works for you - that’s fine, if something works (or is alleged to be working) for other people but for whatever reason does not work for you - following that approach would not make much sense for your personally

  2. in this sense, as a general rule one needs to be careful not to mistake the goal and the means, or over-emphasize the latter over the former:
    any mnemonic technique that makes achieving the ultimate goal of memorization of the target content more difficult / complicated (compared to not using it at all) should probably be avoided

  3. you can name specific components as you like (as long as it helps with memorization FOR YOU) but you can also use conventional radical namings
    used in Japanese (wikipedia page on 部首 has them)


Thanks for your answer!
I’ve read in one of your article, that you didn’t attempt Kanken first, so i was wondering: have you noticed an evolution in your own creation of mnemonics ? Did you develop your system for Kanken or was it already “in place” ?

I got the feeling that according to our own level of japanese-proficiency, our mnemonics will evolve too.

Do you change mnemonics over time ?

For exemple, if you forgot how to write a difficult kanji, how do you approach it a second time ?
Would you reinforce your already-existing mnemonic (taking the risk of “forgetting the goal, overloading slightly your memory for a single kanji”), or would create a new mnemonic, or well, just write it a few times ?

  1. well, obviously when you have just started you can’t avoid learning things by mechanical memorization:

yes, you can memorize “彬” as “ki-ki-san”, but to do it you first need to know that 木 is “ki”, etc.

  1. when we talk about forgetting the kanji-related material specifically in terms of individual kanji (i.e. not particular compounds comprised of them) i usually:

a) either rely on that ‘auditory’ approach if there happens to be one: yes, for 彬 i have “kikisan” but that does not mean that i have developed my own special incantation-style readings for all kanji

b) or try to deduce them by analogy - most often through thinking about other kanji with similar on-yomi (i.e. those that have the same common 音符 element):

thus, for example, on the last >>kanken test<< (2023/02/12), to which due to certain circumstances i did not have any spare time to prepare for (actually, this condition has already been going on for half a year), in the writing section i had to recollect the word “キホウ” (機鋒) which is figuratively used in polemic sense, about which i could figure out the meaning from the context, but was totally unable to recollect its constituent elements, namely:

  • which kanji should i use for the キ part (despite having seen this word in the past, during the test i mistakenly thought that it was 危)
  • which kanji should be used for the ホウ part (i.e. i certainly remembered it was “kissaki” (blade), but was not able to come up with its shape (though, i was more than sure that it must be the 金 - because, blades are made of metal) ----- so here we have the exact case you are asking about ----- “if you forgot how to write a difficult kanji, how do you approach it a second time ?”):

i was thinking along the line of which other kanji had the same “onpu” 音符 element (in this case it is ホウ) - like 方 or 包…

in the end, i was not able to recollect the correct kanji but that was the method i tried to use


This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.