[Egyptian] Auto-discard notification

[Simon Schweitzer]

Hi all,

only a remark concerning the relevance of grouping.


> a) right or wrong, the egyptological community *wants* a « close enough » representation of the original text. An automated layout is simply going to alienate them.
> I’m not speaking of corpus linguists here, but of almost all the people I have interacted with concerning JSesh, and they are probably a good sample. Usually, I have a hard time explaining them they are not doing a facsimile.

And the control characters horizontal and vertical grouping would 
increase the acceptance of the Unicode encoding by the Egyptological 
community.

>
> b) there *is* information in sign positioning. In particular, quadrants limits are more likely to indicate word separation. In some cases, sign positioning can hint at possible alternative readings
> (I have a specific example in P. Turing 1880, where Gardiner’s reading D37:A1 could be replaced by D37:t with a better meaning. A « flat » rendering, like « D37-A1 » would have made the alternative rendering
> much more difficult to make.
>
> c) and of course, in monumental texts, sign positioning can be really significant - our colleagues working on Ptolemaic temples will probably be quite vehement about it. For instance, in the spelling pt:HH:tA for Ptḥ.
>

At this point, I cannot agree. In my opinion, complex combinations in 
Ptolemaic times should be addressed as signs not as grouping of signs 
like the sign E232 for anx-mi-ra or Di-anx-mi-ra. So, why not a special 
sign for nw:mw or for pt:HH:ta? And the highly developed encoding in the 
Graeco-Roman temples are always based on the sign level and not on the 
grouping level. In other words: there is a relevance which sign is used 
for a specific phonetic value. But I do not know any examples (in the 
ancient texts or listed by the colleagues who are interested in the 
Ptolemaic writing system) that the grouping of signs is relevant so that 
there is a difference between A:B and A-B.
By the way, if I remember correctly there is not only nw:mw for m-Xnw 
but also nw-mw. (But I cannot find the reference at the moment, sorry.)
Therefore, I am not convinced that these arguments could persuade all 
"suspicious" colleagues of the relevance of the grouping. But anyway, we 
should have these control characters because of the "political" 
arguments mentioned above by Serge in section a).

Best regards,

Simon