From: jcowan@reutershealth.com
Date: Fri Apr 30 2004 - 11:20:24 EDT
Philippe Verdy scripsit:
> Suppose that a modern Hebrew text is speaking about Phoenician words, the script
> distinction is not only a matter of style but carries semantic distinctions as
> well, as they are distinct languages. It's obvious that a modern Hebrew reader
> will not be able to decipher a Phoenician word, and even understand it if it is
> transliterated to the Hebrew script.
I own a book about Modern Irish which uses the Gaelic script for all Irish
words and phrases, so the situation is analogous. Yet we do not encode Gaelic
separately. I also own a 19th-century book about German which does the same
thing with Fraktur print and Suetterlin handwriting. I can read German in
Antiqua more or less, but I have considerable trouble with the Fraktur, and
the Suetterlin is utterly opaque to me. Yet a Unicode transcription would
be in the Latin script throughout.
> Even though there's a continuum here, having the choice between a historic
> script and the modern Hebrew script will be useful to allow writing texts with
> mixed scripts (notably for didactic purposes, and vulgarization books). Without
> the distinction in the code, it will be difficult to read a text using mixed
> scripts unified with the same Unicode code points.
To represent such documents one needs markup.
> Modern Hebrew with its pointed extension for historic religious texts is already
> complex enough without adding new historic script styles to that complexity.
But all the archaic versions are far simpler, being plain R2L letters without
complex-script properties like vowel signs, ligaturing, or even final forms.
If this is not the case, we need evidence of it.
> It may even be possible that some branches be disunified to cover the
> case of left-to-right scripts or early ancesters of Greek, or the case of early
> Brahmi and Arabic scripts.
Archaic Greek is well-represented by Greek. Karoshthi has its own encoding,
and rightly so. Nabataean, the ancestor of Arabic script from a glyph point
of view, is yet another 22CWSA.
Nobody proposes unifying the Greek-descended alphabets (though Pontic uses
the Greek alphabet with a different set of conventions from standard Greek,
notably the avoidance of final sigmas; see Nick Nicholas's site), or unifying
Phoenician with Arabic, Syriac, or any Indic script. It's the 22CWSAs that
are at issue here.
> One day the Hebrew script will need to be stabilized to work correctly with
> modern and Biblic usages.
That day came long ago. We need a few adjustments, a few conventions,
a few new characters, and support for non-Tiberian pointing.
> Then writers and scholars will have the choice between the best scripts to use
> to represent the printed texts. I quite sure that each branch will have their
> distinctive orthographic system, their own sets of properties, etc... even if
> there's a superficial one-to-one mapping from one to the other.
It's not superficial. It's deep. Phoenician shin is Hebrew shin in every sense
but the purely glyphic.
> A too broad unification for
> characters that already are not immediately identifiable by their apparent glyph
> identity will just create a nightmare.
Every case is different.
> If we unifiy Phoenician with Hebrew too early, it will become nearly impossible
> to introduce new vowels or newer left-to-right layouts, because the Hebrew
> script will become too complex to handle correctly with these additions.
What new vowels? And nobody proposes unifying *anything* L2R with the 22CWSAs.
> Let's keep Hebrew clean with only modern Hebrew and traditional pointed
> Hebrew...
We already have non-Hebrew languages using the Hebrew script according to their
own conventions, and not always in Square either.
> The simple one-to-one mapping will still be possible for
> the most direct ancestors of Hebrew, but this will not work with lots of
> Phoenitic branches from which Hebrew is not an ancestor or child.
Name them, other than Greek, Indic, Arabic, and Syriac which are already handled.
-- Do NOT stray from the path! John Cowan <jcowan@reutershealth.com> --Gandalf http://www.ccil.org/~cowan
This archive was generated by hypermail 2.1.5 : Fri Apr 30 2004 - 12:31:42 EDT