Berthold Frommann wrote:

> Well, regarding Han-characters, there are quite a lot of
> graphical elements which do have a meaning but are not
> part of any of the various lists of "radicals" (the most
> frequently used being the Kangxi-radical system).

In fact, in analyzing Chinese characters, I would prefer the term
"component" to "radical". To me, a "radical" is basically just a section
heading in a dictionary ("section heading" is perhaps the most literal
translation of Chinese "bushou").

A "component" is any one of the basic items in which logograms may be split
up, either bearing a semantic information, or a phonetic information, or
both, or neither.

The fact that most characters have exactly one semantic component and that
this component normally matches the character used as a "section heading" in
a dictionary is clearly not incidental, but does not authorize to unify the
two concepts, IMHO.

> (e.g. 寺 ("temple"), which appears in many characters (詩, 侍, 時, 特...)
> but is NOT a Kangxi-radical.)
>
> Therefore, I would call the radicals an incomplete subset of
> the graphemes of Chinese writing.

I would call them elements of a lexicographical indexing system which are
loosely inspired on some semantic components of Chinese writing.

> Mind by the way that some of the graphemes can be further
> subdivided into other graphemes, so graphically speaking,
> there are combinations of certain graphemes which have a
> distinct meaning. E.g.: 音 is not just a mixture of 立 and 日,
> but a unit on its own.

I guess that the fact that "音" ('sound', Kang Xi rad. 180) looks like a "立
" ('standing', K.X.r. 117) on top of a "日" ('sun', K.X.r. 72) is pure
coincidence (or perhaps a graphical "folk etymology").

It sounds like a case similar to letter "d" which is not composed of a "c"
plus a "l", despite its appearance.

But take for instance "見" ('see', K.X.r. 147): it is probably no
coincidence that it is formed by a "目" ('eye', K.X.r. 109) on top of a "儿"
('human, human legs', K.X.r. 10). Diachronically, it probably represented an
eye on top of two legs, suggesting something like "looking around".
Synchronically, however, it is just an indivisible semantic component of
characters having something to do with vision.

This second case looks more like the one of letter "w": diachronically it is
a digraph formed by two "u"'s but, synchronically, it is just a single
indivisible letter (BTW, notice that its capital form is "W", not "Vu" or
something like that).

> P.S.: I wouldn't use the term when describing writing systems
> other than Han, logographic cuneiform, hieroglyphics and the
> like. As Marco Cimarosti pointed out, it's extremely silly to
> analyse "B" into I and 3.

OTOH, analyzing "W" into "V" + "V" (or, say, "G" in "C" + "-") is not so
silly: just very diachronic (i.e., appropriate when dealing with the history
of writing, not when teaching children to write).

Moreover, even analyzing "B" as <vertical stem> + <arc> + <arc> is not
always silly: when teaching a class of calligraphy, or when designing a
typographic font, that might be the proper thing to do.

_ Marco