No Subject

[owner-idn@ops.ietf.org]

Tue, 23 Oct 2001 09:15:50 -0400
Date: Tue, 23 Oct 2001 09:15:50 -0400 (EDT)
From: Jungshik Shin <jshin@pantheon.yale.edu>
To: Kent Karlsson <kentk@md.chalmers.se>
cc: <idn@ops.ietf.org>
Subject: Re: Hangul Jamo consonant or vowel clusters (was: Re: [idn]
call
Sender: owner-idn@ops.ietf.org
Precedence: bulk

On Tue, 23 Oct 2001, Kent Karlsson wrote:

> ----- Original Message -----
> From: "Martin Duerst" <duerst@w3.org>
> ...
> > At 19:05 01/10/22 +0200, Karlsson Kent - keka wrote:
> >
> > >And still it is not sufficient, since historic Hangul
> > >Jamo aren't covered by the precomposed Hangul syllable
> > >characters. I have also indications that many useful
> > >consonant (and vowel?) clusters aren't covered as separate
> > >(Jamo) characters, but have to be constructed from multiple
> > >'more basic' (non-cluster) Hangu Jamo characters.
> >
> > Now that's new to me. Can you provide some evidence
> > (e.g. some scans)?
>
> I don't have any scans, but a quote from the public Unicode mailing
list.
> In http://www.unicode.org/mail-arch/unicode-ml/y2001-m10/0253.html:
> :Unicode Hangul Jamo block has about 220 Hangul Jamos used in modern
Korean
> :and middle Korean. Why are there so many? Because instead of encoding
only
> :17 consonants and 11 vowels (one can go further down the
'reductionist'
> :road and say that only several base consonants and a few base vowels
> :along with 'diacritical' elements are all necessary to
represent/encode
> :Hangul. See p. 143 of Sampson), Unicode encodes all forms of
'composite
> :Jamos' (consonant clusters, 'double' consonants to represent tense
> :consonants, complex diphtongs, etc). However, this is not sufficient
> :and Microsoft Word 2002/Windows XP/MS IE 6.0 comes with truetype
fonts
> :that can be used by 'Uniscribe' to combine dynamically Hangul Jamos
> :beyond a simple I.C + M. V. (+ F.C.) model (I.C. = initial consonant,
> :M.V. = medial vowel, F.C= final consonant) as described in Unicode
3.0
> :..........
> :
> :  Jungshik Shin
>
> I assume that this functionality is not included just for fun...

  Yes, they're necessary for the full representation of Middle
Korean, sometimes modern Korean (see below) and perhaps future Korean.


> > I use to write my last name as a Hangul syllable as a joke,
> > but as the 'rst' combination doesn't turn up in as a final
> > cluster, I have to do this using graphics. So I'm familliar
> > with the problem. But I assumed that all actual combinations
> > that have been used historically are actually available
> > in the Unicode Jamo block.

  Well, you seem to give too much trust in the way Korean standard body
came up with the list of Hangul Jamos to include in U+1100 block.
In early 1990's when ISO and Unicode asked them to provided the list,
what they did was to ask some Korean linguists to come up with the list
in a very short notice. If they had had a better understanding of the
standard-making-procedure, they would have taken longer time to come up
with a better/complete list. Needless to say, their list was incomplete.
How incomplete was the list? A few years ago, I even found a vowel
missing
which is frequently used by modern Korean speakers although it's not a
part of the official orthography. Later I found that the standard body
of DPRK(North Korea) proposed to ISO that this vowel along with several
others be included in U+1100 block.

  Microsoft developers in Korea, after consulting Korean linguists,
determined that  MS Windows XP/Office XP had to support a lot more Jamos
than listed in U+1100 block. Therefore, they have to support consonant
clusters and complex vowels  by combining existing Jamos.

   Jungshik Shin

P.S. If you can read Korean, you may refer
to http://www.korean.go.kr where you can find some articles posted the
issue.