Joan, I am a little confused by your response which seems to be out of order. It seems that I wrote:Meteg to the right does not actually need an extra character, because if CGJ is used to override canonical equivalence and reordering of vowel sequences, the mechanism is already in place to use it in exactly the same way for sequences of vowels and meteg.
<snip>
Meteg to the right does not actually need an extra character, because if
CGJ is used to override canonical equivalence and reordering of vowel
sequences, the mechanism is already in place to use it in exactly the
same way for sequences of vowels and meteg.
and you replied
Peter, I don't think CGJ alone will work, since Meteg in the canonical order must always come after the vowel. If the currently defined meteg is the one on the left of all vowels except hatafs,and central to hatafs, then vowel-CGJ-hataf could be left of the hatafs, leaving us with a need for a right meteg to be defined.
Or am I misunderstanding?
My proposal would be to encode as follows:
Meteg to the left (standard): vowel - meteg, or its canonical equivalent meteg - vowel
Meteg to the right: meteg - CGJ - vowel
Meteg in the centre (hataf vowels only): either vowel - ZWJ - meteg or vowel - CGJ - meteg
Or we could vary this by doing what John Hudson has proposed, for hataf vowels only:
Meteg to the left: vowel - ZWNJ - meteg
Meteg to the right: meteg - CGJ - vowel
Meteg in the centre (commonest): vowel - meteg, or its canonical equivalent meteg - vowel
But in the latter case we would end up with a possibly undesirable ZWNJ between two combining marks - by the way in just six places in the Hebrew Bible (according to some data which you sent me last year).
If we take either of these options, we don't need a special right meteg.
-- Peter Kirk [EMAIL PROTECTED] http://web.onetel.net.uk/~peterkirk/
