Re: Phonetics
| From: | Mark J. Reed <markjreed@...> |
|---|
| Date: | Wednesday, March 28, 2007, 22:37 |
|---|
On 3/28/07, John Vertical <johnvertical@...> wrote:
> >But that semantics is defined in the standard -
>
> Hold on - what semantics exactly?
The equivalence of (a + combining macron) and (a with macron). In
general the equivalence of combining sequences with precomposited
single characters.
> This is the part where I disagree. If your abstract caracters are not based
> on their encoding, nor on the actual appearence, then what are they based
> on? I mean, I don't see a reason to postulate a metaphysical intermediate
> level here.
It may be metaphysical, but that intermediate level is nonetheless
explicitly defined in the Unicode Standard. The supposition is that
"a with a macron" exists, as a conceptual entity, between and
independently of any particular graphic realization (glyph or
combination of glyphs) and any particular sequence of code points
(single character or multiple characters in different orders). Of
course, the abstraction continues on beyond those points in both
directions - above glyphs are words and below code points are actual
bit representations in UTF-8 or UTF-16BE or UTF-16LE or UTF-32 or SCSU
or . . .
> >No, it's not. Unicode has nothing to do with fonts!
>
> It definitely depends on the font used what exact sort of a glyph a giv'n
> piece of code comes out as, so either you misunderstood my point or I have
> misunderstood yours...
My point was simply that Unicode characters have an existence
independent of any particular glyphic representation of them. The
ones in the tables are just samples, after all.
> And I got the impression that Philip was suggesting that cedilla and comma
> belo' are "the same diacritic" in some manner independant of their encoding,
> appearence or history...
Well, I can easily imagine that "squiggle attached to the bottom
middle" might be an underlying concept that gets realized as either a
cedilla or a comma depending on context or geographical location or
whatever. That was presumably the position of the Unicode Consortium
prior to 1999... but apparently no longer. :)
--
Mark J. Reed <markjreed@...>
Reply