This would be good evidence for the hypothesis that the concept of "grapheme" is sloppy and western-centric.
As I pointed out, the "ideograms" or "pictograms" are also composed of radicals. 5,000 years ago, this was not the case. Over the centuries, Chinese has undergone repeated standardization and at this point all characters are composed of radicals, even those that most resemble an image.
Thinking of
@Limbo Pete , I am reminded of the word "donkey".
Here is an audio recording I just made of myself saying the word in Mandarin. Here is its Chinese representation:
驢
The left-hand side is again the character for "horse",
馬
while the right-hand side is a phonetic component 盧.
Many people believe the character for horse was invented to look like a horse's hoof. 5,000 years ago, it did not consist of the neat strokes e.g.,
but was later standardized to this point. Now it is composed of individual stokes as I pasted above. Some of these strokes can change the meaning of other characters, as I showed earlier with the example of
老師長得很帥! The underlined characters differ by only one stroke and have very different meanings.
Here is another example, the character for tiger. The modern Chinese expression is 老虎. Focus on the second character. Here is a proposal for how this character came to be. On the far left, we see a 1000s-year-old oracle bone inscription of something that looks like the side profile of a tiger. After thousands of years, the tiger don't look like a tiger no more. The new character has no phonetic components at all and is composed of "radicals" such as 几and more fundamentally "strokes". The final character even includes 七, the word for "seven". So where is the alleged "grapheme" here?
"A grapheme can also be construed as a graphical sign that independently represents a portion of linguistic material."
This definition seems vague. Suppose I wanted to construct the set of graphemes of the English language. Call the set S. Is 'z' in S? It's a graphical sign, but does it "independently represent a portion of linguistic material"?