Re: Hebrew hataf vowels (was: About CGJ)

From: Peter Kirk (
Date: Thu Jul 24 2003 - 19:40:27 EDT

  • Next message: J Do: "Re: [OT?] LCD/LED Keyboard"

    On 24/07/2003 15:27, John Hudson wrote:

    > At 01:40 PM 7/24/2003, Peter Kirk wrote:
    >>> These are display issues, not encoding issues,...
    >> Not entirely. First I need to know what sequence of Unicode
    >> characters I should use to encode holam-waw and aleph with right
    >> holam. Garbage in, garbage out. Then I need to be sure that your
    >> sophisticated rendering system actually makes the required
    >> distinctions and is not confused by any rare cases.
    > In both these cases, the holam belongs to the preceding consonant, but
    > is contextually shifted onto the following alef, vav or shin. I wasn't
    > aware that this was in any doubt. ...

    Well, I was in little doubt that this was the correct thing to do, but...

    > ...So, for example the word for 'head' in Genesis 3:15 is encoded
    > resh + holam + zaqef qatan + alef + shin + shindot

    This is indeed how it is encoded at and, and in the text
    provided for Paratext 6.

    But you imply that holam vav should be encoded with the holam before the
    vav. I agree. But none of these three texts agree with us. They all
    three encode the word for "great" in Genesis 4:13 as:
    gimel dagesh qamats dalet merkha vav holam lamed

    The following word "my punishment" is encoded at and in Paratext 6 as:
    ayin hataf-patah vav holam nun hiriq tipeha yod

    i.e. holam-vav and vav with holam are encoded identically.

    In an attempt to make a visual distinction between holam vav and vav
    with holam, Mechon Mamre inserts ZWJ between vav and holam, thus:
    ayin hataf-patah vav ZWJ holam nun hiriq tipeha yod

    So it seems that there is a considerable measure of doubt over just what
    to do! You imply that it is best practice to encode holam-vav as holam
    followed by vav, perhaps with an intervening cantillation, i.e. "great"
    would be:
    gimel dagesh qamats dalet holam merkha vav lamed

    If this is indeed best practice, that needs to be clearly agreed and
    clearly stated so that text providers know what they should be doing.

    Both Ezra SIL and SBL Hebrew (the version which was temporarily on a
    website) with a released version of Uniscribe displays all of these
    combinations sensibly. Ezra SIL makes a clear graphical distinction
    between holam vav and vav holam, but SBL Hebrew does not.

    > I'm looking into this. The substitution should result in a clear
    > distinction, but there may be a problem in the lookups. I'll send you
    > a PDF offlist once I have confirmed that it is working properly.
    > Regards, John
    John, I put all this in a table in a Word document, I'll send this to
    you offlist.

    Peter Kirk

    This archive was generated by hypermail 2.1.5 : Thu Jul 24 2003 - 20:21:36 EDT