Shimin dōji jizukushi anken 四民童子字尽安見, which I suppose means something like "Simple guide to all [written] characters, for the children of all four occupations," is a 1716 kanji textbook by Matsui "Tosui" Shōzaemon (松井庄左衛門/兎睡).

After the illustrations at the beginning, what it amounts to is a long list of kanji used for words in different categories. Some of the categories are conceptual, like "same Japanese reading, different meaning" (同訓別格), like 犬 and 狗: both are pronounced inu ("dog"), Tosui explains, but the first is used for big dogs and the second for little ones. Most of the categories are typological, like "animals," "equestrian equipment," and "utensils for the 'three beverages' [tea, booze, and tobacco smoke]."

Waseda claims that there's Ainu in there, too, but I couldn't find any. I did find a category called Ikoku sōmoku 夷国草木, which could mean "grasses and trees of the country of the Ainu," but it seems to be a more general reference to foreign lands in general, containing mostly Sino-Japanese vocabulary for continental plants (e.g. enbuju 閻浮樹 = jambul).

Anyway, the 64th and final chapter is called "Rigiji shū" 理義字集. Rigiji has a very specific meaning today: characters formed by doubling or trebling other characters, like 晶 ("bright"), which is 日 ("sun") × 3. And indeed Tosui's first few columns of rigi characters fit that description. Here are a few examples that (as far as I know) are no longer in regular dictionaries:

Three moons (月) for sayuka, i.e. sayaka, "clear." I suppose this is a de- and then reconstructed variant of a similar character sometimes used to write this word: 𣇵, 日 ("sun") and 明 ("bright", itself decomposable into either "sun and moon" or "window and moon").
Three men (男) for tabakaru ("consider," "discuss," or "deceive"), placed alongside the much more orthodox character 姦: three women (女), meaning "wicked" or "noisy." As you might guess, this one is usually first in line when it's time to write about how the Chinese-character system reinforces sexist attitudes in society.
Three individuals (个) for tagau, "differ."
Three goods (吉) or three completions (了) for satoru, "realize, awaken, become enlightened."
Three dots for bussho i-ji 仏書伊字, a.k.a. i-ji santen: the vowel i in Siddhaṃ (known in Japanese Buddhism as bonji 梵字), which according to the Digital Dictionary of Buddhism entry by Charles Muller linked above was "used as a trope for 'neither the same nor different'" and for "the relationship between the dharma-body 法身, prajñā 般若 and 解脫 vimokṣa, all three being necessary to complete nirvāṇa." Which is all very interesting, but please don't let it obscure your view of the fact that Tosui is now applying analytical techniques based on Sinitic script to an East Asian glyph evolved from a letter in Siddhaṃ.
The triforce, glossed only as kaki. Oysters? Persimmons? I don't know.

On the next page, things degenerate rapidly:

A branching abstraction for ki no mata, "fork of a tree."
A barbed abstraction for tsuribari, "fishhook."
Strokes going left and right for hyorohyo to (?), "staggering" (with a helpful note explaining the character: "turn left, turn right")
A broken 行 ("go") for tatazumu, "stop, stand still."
A broken 門 ("gate") for (w)ehen (ゑへん), the sound someone makes when coughing. (Can this be right? Maybe it's a write-o for (w)ihen 違反, "break the rules/one's word"? That seems pretty far-fetched too though.)

These characters are clearly not made through reduplication, which suggests to me that Tosui understood rigiji to have a broader meaning: characters (字) to whose meaning (義) there is reason (理), as opposed to characters with phonetic components and so on. That, or he has successfully pulled off a three-century troll. (I'll just leave this here.)

Popularity factor: 10


I can't read any of the script, but how about the rather silly six 有 under one roof, the open triangle, or the seal script looking thing at top left? That whole page looks like an entertaining read.

The use of triangles for "kaki" looks strangely familiar to me, but I'm probably just thinking of the beloved (by me) storefront "masu". 〼

L.N. Hammer:

Is that really the difference between 犬 and 狗? Or at least the modern understanding?

I've been lately (in between bouts with classical verbs) working on trying to get a handle on the nuances of alternate kanji for the same word. My most recent enlightenment was the connotations of 波 versus 浪.



I can't help but think that the big dog/small dog distinction comes from the 犬 = native dog, 狗 = continental dog (and 戌 = time to start preparing to have hair of the dog that bit you) distinction.


Wow, that photo of the book pages is lovely. I usually do fairly well with cursive hiragana but I can't read but about 10% of those glosses. I wish I knew of some study resources to improve my cursive reading.

Perhaps the degenerated kanji are like the fake words they bury in modern English dictionaries, so they know when someone just copied their entire list.

Leonardo Boiko:

> with a helpful note explaining the character: "turn left, turn right"

Couldn’t he be describing the eiji happō strokes employed in drawing it (taku and migi-harai)?

Which brings me to what surprises me more about these examples—they appear to not follow the construction of kaisho characters at all! The lines in the Triforce are too simple, the bonji dots don’t look like 点 dots, and those tapering curves of the fork make me think of seal/bronze/oracle characters.

“An abstraction for fishhook”—looks like the left part of 以。

Charles: I found this book in my dept library, it’s quite interesting—but would take a whole lot of dedicated study to really profit from it. You can also find countless calligraphy books on cursive kana, renmeitai etc. Perhaps just taking shodō classes would be better, though I’d recommend talking with the teacher about your goals to avoid being stuck drawing 永 for 10 years.

Leonardo Boiko:

I love the alternative character for よつ。 Dig those clocks that use IIII as roman for “4” instead of IV?

What’s いづる? Confusing but important distinction from usual 出 there— 出 has a single vertical stroke.

That he has things like what seems to be the left part of 以、 the frame of 直、 the top of 災、 the root hair component 彡 makes me feel like he was engaging in some early component analysis. Nice to see they were already amused by 凹 and 凸。


Really out of my depth here, but could the top-left seal-script robot be うるく~= intercalation (as in modern うるうどし leap year)?

And could the triforce be かき = fence, as in litle enclosures?


Okay, let's do this.

Seal-script looking thing is 鼎, /kana(h)e/ (yeah, it's actually an officially recognized kanji!). It means "tripod" and it is a picture of a tripod. (I can totally see where you got うるく from, though, anonymous, the kana do look similar.)

Open triangle: this page says that it's a type of cup made out of willow, also pronounced /ko/. (Kojien also gives /ko/ as a reading for 壺, with the meaning "cup (for alcohol)"

Six 有: This has the furigana /hasidate/, so it may be related to the tradition of assigning crazy compound kanji to "Ama no hashidate" (as in the view); search for 天橋立 at the link above to see an example.

Leo: "Couldn’t he be describing the eiji happō strokes employed in drawing it (taku and migi-harai)?" Maybe -- it just says "migi e modoru, hidari e modoru"

"they appear to not follow the construction of kaisho characters at all!" -- Yeah, it looks like this is also sort of his grab bag for glyphs or icons, which he had no category for other than "kanji".

Triforce = fence: Maybe!! I didn't think of that one.

Kana in general: The big problem isn't so much that they are just cursive as that they are hentaigana... like for "kanae" above, two out of the three kana are not the ones that survived to the modern day. I'll write a post about learning these next week!


Oh -- and "いづる" is just the old form of "でる". I think that the kanji 出 was originally two separate pieces, like he shows, but I don't have a dictionary to check.

Leonardo Boiko:

Judging from the examples in chineseetymology.org they already started to write it as a unit in Seal, though it was two parts in Bronze and earlier—but it was “foot” (modern 止) and something, not two mountains 山。 I’d like to see the evolution of the reisho/kaisho forms but have no idea where to look for that.

btw that anonymous was me, I simply forgot to fill the fields due to enthusiasm.

Comment season is closed.