Re: Hebrew hataf vowels (was: About CGJ)

From: John Hudson (
Date: Thu Jul 24 2003 - 15:18:56 EDT

  • Next message: "Re: Hebrew hataf vowels (was: About CGJ)"

    At 11:46 AM 7/24/2003, Peter Kirk wrote:

    >I'm glad to hear it. But such things need to be cross-platform. They
    >should also be public*, because that is the only way to make them
    >cross-platform and because that way we can all be sure that all expert
    >opinions have been taken into account. So probably Unicode is the
    >appropriate forum for discussions and for formalising these things. On
    >this issue there seems to be a serious lack of input from Jewish and
    >Israeli scholars. I just received a critique of Ezra SIL from an Israeli
    >source which would probably not have been necessary if he and others like
    >him had been consulted earlier.
    >One of the specific issues he brought up was this one: how do you
    >distinguish the holam-waw vowel combination from the consonant waw
    >followed by the vowel holam? They are clearly visually distinct in BHS and
    >other printed Hebrew Bibles, see Genesis 4:13, contrast words 4 and 5 in
    >BHS. And they are clearly semantically distinct. On a related issue, how
    >do you encode holam above the right side of aleph, as in the very common
    >Hebrew word for "head", see Genesis 3:15 12th word? This is another issue
    >on which different texts differ, and in nearly every verse as holam-waw is
    >very common. (Consonant waw with holam is not very common, but it is not
    >rare either.)

    These are display issues, not encoding issues, so I don't think Unicode is
    necessarily the correct body to determine how they should be handled. Much
    depends on the font format used, and how this interracts with the OS and
    application script engines. I'm pretty sure that most of what I've done in
    the OpenType lookups in SBL Hebrew could be done in a very similar fashion
    in an AAT for the Mac, or in a Graphite font, but this doesn't mean that
    another -- perhaps cruder, perhaps more sophisticated -- rendering system
    wouldn't require a completely different approach to achieve the same visual

    The way to encode all of the things you mention is pretty straight-forward,
    but in the case of the vav+holam vs holam+vav, in addition to two different
    precomposed glyphs, my font also contains a series of six lookups,
    including one that specifically handles Ha Shem, to make sure these render
    correctly in all circumstances. The positioning of the holam over the right
    side of alef and shin is handled contextually.

    There is a document currently available at
    that displays every sequence of consonant + mark(s) that occurs in the BHS
    text and the Westminster morphological database, with post-context
    consonants. This doesn't give a perfect representation of what happens in
    every circumstance (pre-context consonants affecting horizontal position of
    telisha gedola are missing; when holam is shifted left over shin, the
    document does not display additional contextual interraction with the shin
    dot), but it's a pretty thorough rendering test. This document and the
    manual will shortly have a permanent home at

    I'm anticipating that, as the SBL font goes into use, some updates will be
    necessary, but I'm pretty confident that most problems have been caught and
    corrected in pre-release testing. I spent a lot of time checking rendering
    of specific sequences against the printed BHS and BHL.

    John Hudson

    Tiro Typeworks
    Vancouver, BC

    The sight of James Cox from the BBC's World at One,
    interviewing Robin Oakley, CNN's man in Europe,
    surrounded by a scrum of furiously scribbling print
    journalists will stand for some time as the apogee of
    media cannibalism.
                             - Emma Brockes, at the EU summit

    This archive was generated by hypermail 2.1.5 : Thu Jul 24 2003 - 15:55:27 EDT