Hebrew: inconsitencies

From: Arno Schmitt (arno@zedat.fu-berlin.de)
Date: Wed Jun 02 1999 - 17:10:38 EDT

Jonathan Rosenne schrieb:
> The placement of the Hebrew points is a complex matter, and the Holam is
> part of this complexity. In addition to the placement on the right of the
> following letter, it is also sometimes elided.
> At 11:41 01/06/99 -0700, Peter_Constable@sil.org wrote:
> >

> >There is another question that needs to be asked in this discussion: how
> >should the "right holem" and "left holem" be encoded. I'd like to see both
> >Jony and Arno answer this question for the set of examples Arno provided:
> They are both encoded as Holam. This is just a placement issue, which is a
> matter of typography and not of encoding. If you want, you can look at it
> as an automatic shaping for Hebrew.
> I wonder why, if you (Arno) believe so strongly that the finals should be
> automatically shaped, you want shape variants encoded in this case.
I wrote earlier:
I have no idea how many people have ever wondered why there is a
single free space in the middle of the Hebrew block: 05BA
There used to be both holams, 05B9 and 05BA.
This was simplified to one holam, because: if the program knows
rules of Hebrew spelling well (rules like: sin, alef, waw
the holam that phonologically "belongs" to the preceding
shin "attracts" the holam that phonologically "belongs" to the
following consonant), and knows what the preceding and following
signs are, the program finds the right form of (place for) the
But this is true of the five _final_ _shape_ of the consonants in
question as well.

I think we have to live with this __inconsistency__.
But there is no excuse for software developers,
- not to give us proper placing of the holam,
- not to provide us with proper shaping of Hebrew consonants (as
is done for Arabic and Indic scripts) - for thus to lazy to type
final forms where they regularly occur.

So, I can give no straightforward answer.
Interestingly, Jony did not recognize "Aoax" (eagle-owl),
as alef (glottal stop), right holam (o), furtive patah (a), het
Unicode decided that furtive patah, signaling /a/ BEFORE the
letter under which it stands, and patah, signaling /a/ AFTER the
letter under which it stands.
So I have to input for /'oax/ "'oxa" -- the argument being: a
vowel signs are keyboarded (stand in the code stream) AFTER the
vowel with which they are WRITTEN (not before or after according
to when they are SPOKEN).
This being so, the holam (the single holam admitted by Unicode),
must be keyboarded after the letter with which it is written.
Aoax is keyboarded/coded as Aoxa (letter, vowel, letter vowel)
roAsh must then be keyboarded/coded as rAosh, because the holam
sits on the alef.

Since Unicode decided that there is only one patah
AND decided that there is only one holam,
the obviously want us to keyboard holam according to when it is
and patah according where it is WRITTEN.
Right, Jony?

I think I understand you twisted logic, but I would prefer a
consitent way.
And logically, there are only five different SHAPES to the
consonants in question, just like in Arabic.


This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:46 EDT