I’m going through my second time through Wanikani, and while doing one of my many lessons, I realized that barely any characters in the Japanese script look like other characters if you were to rotate or flip them, unless you severely simplify them.
In the English alphabet, there are many such cases.
N → Z
d → p → q → b
u → n
f → t
etc.
However, in Japanese this is quite difficult.
One could make a case for possibly く → へ, or maybe even 三 and 川 if you get really creative, but apart from those, it is hard to find many examples.
I find this quite baffling, as Japanese has more characters by multiple magnitudes, using similar radicals in many of them. You would think that statistically, it would have much more occurrences of being able to flip or rotate on character to get another.
I’m struggling to find many more instances myself, so let’s turn this into a little game! If you know any rotational symmetries between different kanji, hiragana, or katakana, post them down below!. The more complicated, the better!
But yeah, most of the radicals have only mirror symmetry and not rotational symmetry. They pretty much all have a right-side-up, and don’t exist in any other orientation. When you combine two of them together, that symmetry goes away. For example, both 木 and 東 have a vertical axis of symmetry… ish. However, 棟 has no symmetry at all.
(Though while we’re on mirror symmetry, 上 and 下 are almost mirror images of each other.)
Which is also kind of surprising. For how complicated some radicals are, you would think that every now and then an upside down 大 would be good enough to represent something. It makes me wonder whether there is a reason why the language did not naturally simplify to have this occur more often, like how modern Latin script evolved over time. Often, script slowly becomes simplified to accommodate more speedy writing, while at the same time not making characters too similar to make them easy to confuse.
Maybe there is a reason why rotational or symmetrical similarities would be confusing in Japanese, like maybe how signs or books were read or something. Or maybe the more rigid adherence to tradition in Japanese culture compared to Western-European culture could slow this process?
Interesting observation. I wonder if it has anything to do with writing with a brush and (not sure of the correct term but) for one thing directional tails wouldn’t naturally have a symmetric pair. Maybe.
I would expect that it’s just a probabilistic thing. Since characters are not designed to be rotatable, and being rotatable is not really a desirable quality (it may even arguably be a slight disadvantage since it can create ambiguity if the direction of the text isn’t obvious, see 9 and 6 for instance), I expect that the only reason a character becomes rotatable is random chance.
From there I think it’s fairly intuitive that the more strokes a character has, the less likely it is to be randomly rotatable. Since most kanji have many more strokes than the average Latin character, it’s not surprising that rotating them doesn’t randomly yield some other character.
And of course as soon as multiple radicals are combined together, the probability of the compound being rotatable becomes increasingly unlikely.
But if you consider that f and t are close enough if you flip them (a bit of a stretch IMO), I would argue that Japanese kana have more instances of this… if you’re willing to squint a little:
And btw N → Z would also require a fair bit of squinting to judge them to be a rotation of each other.
Being able to judge text orientation at a glance definitely seems like an advantage for a font.
It seems that accidentally symmetrical characters appear when the font is simplified to be composed of a small amount of straight lines, i.e. when it need to be chiseled in stone, or drawn by early low resolution computers/printers for that matter.
When a font is brush-written, all such symmetries immediately disappear and you see asymmetrical curved brush strokes.
Æ, œ, ß and the like are ligatures. Apparently ø also originates from a ligature of oe:
The letter arose as a version of the ligature ⟨oe⟩. In Danish manuscripts from the 12th and 13th century, the letter used to represent an /ø/ sound is most frequently written as an ⟨o⟩ with a line through, but also ⟨oe⟩. The line could both be horizontal or vertical.
As for å:
[…] a minuscule O was placed on top of an A to create a new letter which was used in place of the digraph Aa. It was first used in print in the Gustav Vasa Bible published in 1541 and replaced Aa in the 16th century.
only sort of on topic, sorry, but my favorite thing about Japanese/Kanji has got to be the sheer number of variants of 田 or 目 there are. like, basically any place that you could change the stroke order or leave something out, that exists. 田 目 日 口 月 耳 身 基 角 用 皿 組 買 母… i’m sure there are more!
I originally wrote this about just Kanji and then I got overconfident and added Hiragana and Katakana . Good to know I was wrong about that.
I think that’s a good consideration. I would have kind of expected these details to be streamlined over time though, for ease of writing.
I would have kind of assumed that with the sheer amount of symbols, there would be more matches Like two symmetrical radicals being used in opposite sequences.
Oooh, that’s actually a really good observation. The brush history came into my head as well, but the chisel history of western script is something I didn’t consider.
. Good to know I was wrong about that.
Like two symmetrical radicals being used in opposite sequences.
