From: Peter Kirk (email@example.com)
Date: Mon Jul 28 2003 - 18:38:32 EDT
On 28/07/2003 14:16, John Cowan wrote:
>>This lends credence
>>to those of us who are BHS fans and would like to see a visible difference
>>holem-waw and waw-holem. The most reasonable means of achieving this is to
>>encode the holem before the waw when it is holem-waw.
>This argument is unsound. Encoding is essentially autonomous to either
>input methods or rendering methods, and it may demand things that would
>be very unintuitive to the uninstructed user who examines the encoding
>directly. There may be very good reasons for encoding holem-waw as
>other than a holem followed by a waw.
It is not entirely unsound, because in a case like this, where there is
a distinction in the text, the encoding must be such that that
distinction in the text is either explicitly encoded or can be
determined unambiguously (and preferably efficiently, especially for a
rendering algorithm) from the context. In this case the general
algorithm to determine which collocation of holam and vav is intended is
complex (requiring a recursive lookback potentially to the beginning of
the word) and not entirely unambiguous, although there is a simplified
algorithm which will account for all regularly spelled Hebrew words.
The issue of whether the distinction is a real and ancient one or one
introduced by relatively modern editors is entirely independent. It is
certainly older than BHS; for example, the special form of the holam vav
vowel (not at all like vav with a regularly placed holam) is clearly
seen in the facsimile of an 1889 Viennese Bible reproduced by
p.18, second line of text, third word from the left. Anyway, Unicode
should be able to make any distinction which is commonly made by modern
editors. If it were a criterion for inclusion in Unicode that a
character had been in use as a distinct character for centuries, the
standard would be a lot slimmer than it actually is.
-- Peter Kirk firstname.lastname@example.org http://web.onetel.net.uk/~peterkirk/
This archive was generated by hypermail 2.1.5 : Mon Jul 28 2003 - 19:07:09 EDT