Biblical Hebrew Forum

[Peter_Constable AT sil.org]

Dear all:

As author of the proposal document Peter Kirk has mentioned, I guess I 
should respond. Please note that I do not do so defensively; I have 
discussed this issue with Peter prior to his posting, and this is not a 
matter of personal attack for me. I am merely wanting to work with 
interested parties to obtain a solution to an existing problem so that 
Biblical Hebrew texts can adequately be represented in Unicode.

The proposal document was no more than that: a proposal; a position that 
gets expressed in writing and put out in public for discussion and 
consideration. There has been enough response expressing concern over 
dividing Biblical and Modern Hebrew data and implementations that this 
proposal has a very low likelihood of being pursued.

There have been some suggestions that this proposal was part of a plot on 
the part of Microsoft to make existing solutions obsolete thereby forcing 
people to buy software upgrades. Nothing could be further from the truth. 
Forcing new upgrades by intentionally breaking existing solutions would 
cost MS *far* more in terms of negative PR and support costs than the 
revenue they would get. Lots of people, including me, have loved to hate 
MS, but this criticism simply doesn't wash.

The proposal to encode alternate characters for Biblical Hebrew was 
prepared because it seemed to me after discussion with others connected 
with the development of the Unicode and ISO 10646 standards to be the only 
option available. Basic agreement to the idea had been expressed by an 
Israeli involved in Hebrew encoding standards, and various parties 
involved in development of font/rendering implementations and corpora had 
expressed willingness to deal with the issues that solution would have 
created (it seemed worth it to them if that was what was needed to obtain 
a solution to the other existing problems). Of course, that group of 
people may not have been (and, in light of feedback, was not) 
representative of all those involved in development of BH software and 
corpora, let alone users; but input from such others has now been 
received, and is being given consideration to the extent that that 
proposal will almost certainly be abandoned.

That leaves the problem of what alternate solution to propose. All along, 
I have felt that the solution I would prefer to see is to correct the 
canonical combining classes of characters (non-technical: to correct the 
root cause of the existing problems). This is not a trivial matter, 
though. In particular, there is very strong oppostion from certain parties 
in industry involved in Internet standards to making any change of the 
kind that would be required (this is a general oppostion related to all of 
Unicode, not specifically related to Hebrew). The problem is that such a 
change could create serious problems for existing data; this includes not 
only Biblical Hebrew data, but also Modern Hebrew data that includes vowel 
points.

Peter Kirk (and others) have suggested an alternate solution involving 
insertion of an extra control character within vowel sequences, the effect 
of which is to ensure those sequences are preserved. There have been 
variations on this idea involving different Unicode control characters, 
some of which would lead to other serious problems. There is one that may 
work, though it needs further evaluation. As some have noted, this is a 
kludge, but unfortunately the only solution that wouldn't feel like a 
kludge is the one that would require overcoming strong resistence from 
Internet standardizers.

To conclude, the supposed threat Peter Kirk mentioned is not very real, 
though his effort has been useful in gaining input from a wider group of 
users and implementers, and is certainly appreciated.


- Peter


---------------------------------------------------------------------------
Peter Constable

Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485