[demetrius]

> The Chinese "text" unification was done two millennials ago, but people "speak" differently of every individual "character", as dialects or mutually unintelligible or whatever.

It's like Latin in Middle Ages. Everyone speaks differently (in Old English, Old High German, Old French, etc.), but people write things in the same way (in Latin). And often don't even learn to write their native language.

> I'd argue Chinese characters don't actually represent a sound

If this were true, people wouldn't need to switch from Classical Chinese to Written Mandarin.

If Chinese writing didn't represent sound, it wouldn't matter if you wrote 學而時習之，不亦說乎？ or 學習知識以後，常常溫習它，不也很快樂嗎？

But people stopped writing in the first style, and started using the second one, to better represents Mandarin speech.

> if a person reads a "character" completely wrong, he/she may never realize until some awkward moment happens during a speech or a conversation

This happens in most languages, just to a smaller degree. E.g. for a long time I thought Septuagint was Septugiant, because I've only encountered this word in writing and never cared to read it letter-by-letter.

[est]

> people wouldn't need to switch from Classical Chinese to Written Mandarin.

Yeah and when did that happen and what's the phenomenon called in Chinese?

> it wouldn't matter if you wrote 學而時習之，不亦說乎？ or 學習知識以後，常常溫習它，不也很快樂嗎？

That's kind of the point. If you were a LLM, the 學 and 說 would get most attention, the rest of characters is trivial and you don't need precisly recall the "sound" representation. You can even swap the order like

習學識知後，常常習溫它，不很也樂快嗎？

ppl get this in "writing", but not "speaking"