[Biblical-languages] Unicode proposals for apparatus symbols, Biblical Hebrew

Peter_Constable at sil.org Peter_Constable at sil.org
Thu Jun 26 13:09:07 EDT 2003


Peter Kirk wrote on 06/26/2003 09:11:40 AM:

> I am disappointed that you have taken this proposal so far without 
> discussing it with this list, or with a former colleague like myself. 
> Indeed I would suggest that this makes a mockery of your answers: "2a. 
> Has contact been made to members of the user community? Yes 2b. With 
> whom? Biblical Hebrew researchers, content providers of materials 
> related to Biblical Hebrew". At least it seems that this significant 
> community of such people, of whom you are well aware, has been kept in 
> the dark, while the once sensible proposals you were working on a year 
> or so ago have been turned into something which is at least to me quite 
> unacceptable.

This was discussed with several members of this list, as the list of 
proposers on the proposal indicates. I'm sorry you weren't included in the 
discussion; it's just the way it happened, and I didn't have time to make 
sure a wider audience reviewed it.

As for the merits of the proposed solution...

> Is changing the 
> existing combining classes really unthinkable?

Yes. That absolutely will not happen. That is an absolute dead end. I 
raised the issue on the unicoRe list back in December, and the opinion 
that came out was that we'd need to propose alternate vowel characters. 
This was discussed in recent weeks among a group of content providers and 
implementers that included all the people listed on the proposal, plus 
some others. This is not just my idea, and not without considerable prior 
discussion.



> The objection that 
> existing documents will lose normalisation stability is meaningless as 
> anything normalised under the old scheme is still valid text under the 
> new one and vice versa

But it is no longer normalized. And that is a serious problem for many 
implementers of Unicode. I don't like it, but UTC absolutely will not 
consider changing combining classes.



> and your current proposal will completely 
> invalidate a significant corpus which is already in existence

It doesn't invalidate it; what it would mean is that it is no longer 
encoded in what is considered best practice for Biblical Hebrew. And major 
content providers (SBL, Logos and eBHS) and implementers (SBL, SIL, MS) 
were prepared to deal with the need to recode and reimplement.


> A case for changes can well 
> be made on the grounds that this is "a clear and evident mistake" 

I wish it were so. But, that is not a possibility.



> But there are alternative solutions to the problem which, though messy, 
> are not as bad as your proposal for a complete new set of characters;
> 
> 1) For the vowels... this issue could be solved by adding a non-printing 

> base character representing an omitted or "phantom" consonant.

Yes, that is another possibility, and could still be proposed.


> This is 
> also required in some Ketib/Qere cases (Qere without Ketib) where vowel 
> points are found without consonants e.g. at Ruth 3:17.

I'm in Vancouver, and don't have BHS with me, so I can't look at this, but 
if it's a case of points over/under a space between other consonants, that 
would have to be a distinct character from what you propose above -- 
another that I have been planning to propose.


> So this would 
> require one extra character to be added to Unicode; or could some 
> existing character be used e.g. FEFF or 200D?

No; those would have the wrong properties, introducing a word boundary 
where none should exist.


Another suggestion that has been raised in current discussion elsewhere is 
to simply have a character that is notionally ZERO WIDTH CANONICAL 
REORDERING INHIBITER. It's a kludge, but it would work, and would provide 
a solution for both multiple vowels and meteg.


> 2) For meteg/silluq: The simple solution here is to add characters for 
> the comparatively rare right meteg and mid-positioned meteg, with 
> compatibility equivalence to regular meteg.

Yes, that is another possible proposal that can yet be considered.



> Alternatively, the "normalisation 
> override" character approach could work here for the right meteg, though 

> not for the mid-positioned one.

Re the third position for hataf vowels, those of use working on 
implementations have already agreed on a solution, proposed by John 
Hudson, using ZWNJ to give meteg left of a vowel (medial is default, being 
most common).


 
> 3) For holam and shin/sin dot: This tidying up part of the proposal 
> makes sense if the other parts of the proposal are accepted but not if 
> the other parts are withdrawn. The new order would of course be an 
> improvement if used for all Hebrew, but not if applied only to biblical 
> Hebrew in the way you propose. Meanwhile the processing speed objection 
> should be disregarded as the processing time required to sort a list of 
> up to six characters is simply trivial;

It is not trivial if it has to be done hundreds of times while rendering a 
screen of text.



> Anyway, the proposal doesn't fix this problem for modern Hebrew texts 
> and so an efficient solution will still be required.

The problem isn't as much of a concern for modern Hebrew texts.


> As for usability, 
> users can type in the characters in any order and the result is still 
> valid Unicode

But it's unlikely that rendering systems and fonts will make any order 
display correctly.


> I would urge you to withdraw this proposal and prepare a new one on the 
> lines I suggest here, in consultation with this list and with others.

Again, this proposal has been considered by many and represents a joint 
consensus among content providers and implementers; and the issues are 
still being discussed, with alternatives such as you have mentioned being 
considered.


 
> I note that this proposal does not deal with some other characters 
> needed for biblical Hebrew, specifically inverted nun (SIL Ezra 078, 
> M-C/WTS N]8) and lower dot (SIL Ezra 179, M-C/WTS 53). I hope these 
> issues have not been forgotten.

No. I did not have time to prepare such a proposal before this past UTC, 
and intend to prepare such a proposal. It would not have been a good idea 
to combine the two into one.


 
- Peter


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

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



More information about the Biblical-languages mailing list