From: Kenneth Whistler (kenw@sybase.com)
Date: Fri Jan 18 2008 - 20:00:54 CST
> [after: · Yoruba]
> · German dialect writing
> · The lenght, attachment to the base character or slanting on use with
> italic fonts depends on local preference
Thanks. The editorial committee will take that and James'
suggestion on board to come up with something.
>
> KW> For the record, I agree with Michael and James about the
> KW> ogonek, as well.
>
> Here, I still not stay convinced. Look at:
> http://www.sprachatlas.phil.uni-erlangen.de/materialien/
> Teuthonista_Handbuch.pdf
> (German), where you find hooks to denote openness of vowels (like the
> one contained in my draft proposal) which are definitely distinct from ogonek.
> As far as I know now, these hooks are the ones related with the one proposed
> in my draft.
I agree, they are probably related. But all the more reason,
for plain text representation of the content of the transcriptions,
IMO, to associate these with U+031C COMBINING LEFT HALF RING,
which is semantically correct for this.
But there is still no contrast with ogonek demonstrated by
this, by the way, since Teuthonista uses tildes, not ogoneks,
to represent nasalization.
The Teuthonista hook deliberately doesn't seem to attach to
anything, either, which is more like the combining left half
ring. Formally, it is being used as a paradigmatic pair
with the dot below.
> I deliberately decided not to refer to Teuthonista in my proposal
> draft, but this maybe was wrong.
> Teuthonista is used by some universities in Southern Germany. The
> users I have contacted are happy with their legacy 8-bit encodings.
> Making a proposal to encode Teuthonista would probably rise
> interesting discussions, as it would add Tibetan-like stacking to the
> Latin script (unless you encode all possible vertical combinations as
> single characters, as the legacy 8-bit encodings do).
Yes. The stacking of the Grundzeichen (aeiou) could be handled
in Unicode plain text because of the set of combining Latin
small letters in the standard now. But the generalization of
that stacking to the Reduktionsvokale (alpha, schwa, upsilon,
dotless-smallcap-i) would pose a problem.
Teuthonista also has other characteristics not amenable to
plain text, including the indication of grades of nasalization
by using a lightface tilde for nasalization, a boldface tilde
for strong nasalization, and a lightface tilde enclosed in
parentheses for light nasalization.
The left-to-right composition of diacritics above implied
by the light nasalization convention is also seen in the
use of the doubling of the hook we have been talking about
to represent a greater degree of openness of vowels.
So one dot below for closed, two dots below for very closed,
one righthook below for open, two righthooks below for
very open. And then even two hooks between parentheses
for "somewhat more open".
The side-bye-side convention couldn't be accomodated for
dialect transcription, whether we decided the hook in question
was U+0328, U+031C, or required a new character.
If the users of Teuthonista ever decide that their system
would need to be convertible to Unicode -- as opposed to
staying in their legacy 8-bit encoding(s) -- I suspect it
would best be handled with a set of light-weight markup
conventions on top of the existing encoded characters,
the way Egyptian hieroglyphic texts will be represented
in Unicode, rather than trying to smoosh all these strange
conventions directly into plain text.
--Ken
This archive was generated by hypermail 2.1.5 : Fri Jan 18 2008 - 20:02:04 CST