Re: An Aburdly Brief Introduction to Unicode (was Re: Perception ...)

Date: Sat Feb 24 2001 - 09:10:49 EST

On 02/23/2001 05:06:03 PM Kenneth Whistler wrote:

>Peter expostulated:

OK. I have no problem with Ken's response, since he made clear what he was
talking about. This was not the case in the earlier messages to which I was

I have one minor quibble and a comment, for neither of which would I go to
the wall:

>Actually, {a with ring above} is an abstract character in either of
>the two senses I have talked about above.

The introduction of a distinctive notational mechanism, {...}, suggests
that there is an attempt to refer to something *distinct*. Notational
distinction or not, (abstract character)1 and (abstract[able] character)2
are different types of objects. The former is an element in an information
system construct, that construct being a character set / encoding standard.
The latter is an abstraction of abstract real world objects, graphemes. In
this latter sense, {a with ring above} might be thought of (using an
extensional definition) as the set of all such graphemes from all past,
present and future writing systems that contain instances of it. (If you'd
prefer an object-oriented programming metaphor, the latter is a class which
is instantiated in various writing systems.) I'd like to make the
suggestion (meaning that I'm trying on a new idea for the first time and
haven't yet decided whether to buy it or not) an (abstract character)1 = an
element in an IT character set corresponds to or represents but is
different from an (abstract[able] character)2.

All this is very abstract, and very likely won't make any difference to
anything. What will, though, is understanding what each other means, and
for that, I think we need to know when we mean (abstract character)1 and
when we mean (abstract[able] character)2.

Given the use of "abstract character" in the standard to refer to the
former, I'd suggest that we limit the use of that term in our discussions
to that meaning. As for the latter meaning, Ken has coined "abstractable
character". What about "abstract grapheme"? (Neither of these really roll
off the tongue.)

- Peter

Peter Constable

Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485
E-mail: <>

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:19 EDT