For example: 漢字 from both 日本語 (JP) and 中文 (ZH) are derivative from semantic concepts based on what they saw around them (when both languages were first created), kind of like this:

There’s more examples of it I can come up with such as the following:

Also, both languages allow this writing orientation (R-L vertical):

I mean, they even have words that are “swapable” (mainly how characters are positioned but retain the same definition) as shown:

The thing is, both languages are logographic which allows for this form of conceptualization regarding vocabulary based on “shape”:

That is in contrast with languages that use alphabets (as those are based on sound) while 日本語 (JP) and 中文 (ZH) are “pictorial” if that makes sense, meaning each character conveys a word or concept rather than a single letter that has no inherent definition.


Interesting. Never thought about it this way.
First thing to come to my mind: logograms. &, %, +, @. They are quite similar to Kanji. They each have different “readings”, like “&”, which can be “et”, but also “and” or “und” or “y”, depending on the kontext of the reader. So it is not phonetical, but defined pictorally.
Also, maybe Onomatopoesia? Buzz. Hiss. Sizzle. Pop. Slurp. Tick-tock. But I think that does not really meet your criteria.
Or concepts like a U-turn, T-shirt, V-neck, A-frame. They use the shape of a letter as a descriptive element.
And the last thing that comes to my mind would be zigzag (GER: Zickzack). It kind of depicts a broken, alternating line. It’s not really fully visual, but the reduplication and the harsh contrast of syllables try to immitate a stark alternation.
Maybe “synaesthesia” is a distant topic that might interest you, although it is a very broad field. Visual concepts may be distantly related to phonetic concepts and vice versa.