But that's my point regarding the paper mentioned above: I have the impression that this viewpoint doesn't match with the way old beards used the language:

> In other words in each of these quotations, the author uses the graphic structure of a character to represent a key notion in the discursive reasoning to support or confirm a reality or a fact. The meaning of the character can be systematically related to the meaning of the graphic components

> [...]

> According to these texts, characters are analysed into pure semantic components: “west” and “rice” for “grain”; “cereal,” “entering” “rice” for “broomcorn millet”; “eight” and “ten” for “tree,” and the choice of the components is essentially explained in terms of the Yin/Yang and Five Elements theories.

Of course, I am well aware of the modern viewpoint, which disregards such systematic semantic interpretations, but I am wondering whether I understand the paper correctly: « were really characters components always understood semantically in the past? »

And I am cautiously wondering about it, precisely because all Chinese speakers I've talked to shrugs the idea off (« am I missing something? »).

> You're making a very unfair comparaison

Alright, but we were talking about ancient scholars, it seems fair to compare them to modern scholars. Let me try to rephrase myself. My point was multiple:

(1) ancient scholars were trained to do science, just not the way we do

(2) not being "trained to do science" in the modern sense isn't sufficient to disregard their viewpoint;

(2') the converse is also true: being "trained to do science" in the modern sense isn't sufficient to validate a viewpoint. Example: many trained scientists or engineers will hold democracy in high-esteem despite the previous "research" & history.

That's to say, I believe it's well worth studying their viewpoints, despite the fact that they "weren't trained to do science," as commonly understood today.