NetNews Usenet Archive 1993 #3/NN_1993_3.iso/spool/comp/std/internat/1312

Internet Message Format | 1993-01-25 | 4.8 KB

Path: sparky!uunet!usc!sol.ctr.columbia.edu!destroyer!gumby!yale!mintaka.lcs.mit.edu!ai-lab!wheat-chex!glenn
From: [e-mail] (Glenn A. Adams)
Newsgroups: comp.std.internat
Subject: Re: Alphabets
Date: 25 Jan 1993 15:12:14 GMT
Organization: MIT Artificial Intelligence Laboratory
Lines: 81
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]> <[e-mail]>
NNTP-Posting-Host: wheat-chex.ai.mit.edu

In article <[e-mail]> [e-mail] (Dik T. Winter) writes:

>"What is Unicode encoding?".  Scripts?  Writing system?

Unicode encodes scripts, and not writing systems (alphabets).

>It is not encoding the Latin script I think. Consider for instance the
>German writing systems that have been used.  Fraktur.  Is that a different
>font?

Fraktur is a different font typeface.

>Suetterlin.  Is that a different font?  Many would think the latter not
>predominantly derived from the symbols used in the Roman alphabet.  I see
>them as being more derived from the Germanic Runes.  Still, there is a 1<->1
>corespondence between the symbols in the Suetterlin script and the German
>version of the Latin script (I think).

I assume you refer to the written form developed by von Ludwig Suetterlin
(1865-1917).  I don't have any detailed information on it, so I can't say
for sure.  Without knowing any details, I would be willing to say it was
a distinct script to the extent that Suetterlin created new forms or even
borrowed forms from other scripts, perhaps modifying them in the process.

Aside from the issue of encoding utility, I would say that abstracting the
forms of two or more alphabets into a single script should take into account
historical derivation, formal similarity, and perhaps even functional
similarity, although I would give the much less priority than the former two
criteria.

>I think you should add that unification of different scripts is possible
>iff the scripts can be viewed as just being font changes (although the
>derivation of the scripts can be completely different).

I think you may be confusing "script" as I am using it with "handwriting
form" or possibly "written form."  Clearly the latter would be a matter of
only font changes, and nothing more.  However, I am using "script" in a
different way, namely, to capture the notion of a collection of abstract
symbols which tend to have a fairly clear historical relationship, which
still bear a fairly strong resemblence in form, and which are used to
represent fairly similar functions.  In addition, such a script -- as an
artificial construct -- can include elements which violate these criteria
to some degree.

The general process used in Unicode is to identify an alphabet (i.e., the
symbols used in a particular writing system) with some historically known
collection of symbols (a script), attempt to unify the alphabet with the
this collection, and, then, to the extent that the unification is successful
and doesn't interfere with basic processing tasks, replace the script with
the (unified) union of the original script and the forms of the new alphabet.
This produces a new script which may have forms in it that were not in the
original script; e.g., if you look at Unicode Latin blocks, you will find
elements like:  LATIN SMALL LETTER THORN, LATIN SMALL LETTER EXCLAMATION
MARK, LATIN SMALL LETTER BARRED LAMBDA, LATIN CAPITAL LETTER YOGH,
LATIN LETTER TWO BAR, LATIN CAPITAL LETTER TONE FIVE, and so on.  Clearly
none of these are members of the Roman alphabet, or are even close.  In
these cases, alphabets which were largely derived from the core symbols
of the Latin script, i.e., those in the Roman alphabet, were unified with
the Latin script, resulting in unification of some forms, and the addition of
other, novel forms which were innovations in the alphabets being unified.

Clearly, none of the different alphabets which share the collection of symbols
referred to as LATIN LETTERS in Unicode actually make use of *all* of the
symbols so identified; therefore, you can't say it is just a font shift.

>So while you can unify the Suetterlin and the Latin script, you can not unify
>Latin and Greek script although Latin is derived from Greek.

You could unify Latin and Greek if you want, but it would require radical
unification of both form and function.  And it wouldn't buy much as far
as encoding is concerned.  Keep in mind here that a SCRIPT in the Unicode
sense is largely an artificial engineering construct; and not adherent to
any "theory of scripts."  To my knowledge, there is no theory of scripts
anyway.  The problem with written language is that it is based almost
entirely on convention and historical accident; one can't articulate a
theory of writing on the basis of necessity (at least I wouldn't give
credence to any such theory).

Glenn Adams