Re: Corrections to Glagolitic

From: Kenneth Whistler (
Date: Mon May 16 2005 - 16:18:49 CDT

  • Next message: Philippe Verdy: "Re: Virama based model - a note"

    For those who may not be completely familiar with the process,
    the Glagolitic script in the Unicode Standard has already reached
    the *published* stage, as part of the Unicode Standard, Version 4.1.
    Its corresponding encoding in ISO/IEC 10646 has reached the FDAM
    ballotting stage, which also means it is past the stage of
    technical change, and will be published imminently there, as well.

    What *that* means is that discussion of the formal character
    names is academic at this point -- they are no longer proposals,
    but are published and immutable.

    [Patrick knows this, of course... I am repeating this for the
    benefit of those reading the thread who may not be aware of
    the status.]

    That does not preclude appropriate future annotations of the
    character names and text, of course.

    The reason why Michael Everson attempted to focus on the question
    of what evidence might be brought to bear on the need to encode
    additional characters for Glagolitic is that that falls in the
    realm of the *possible*, for future changes.

    > >> Incidentally N2610R says that U+2C26 had as name YO in ISO 6861:1996,
    > >> this is not strictly true, it was JO. (I thought names were important
    > >> for cross-standard references).

    Not for the old TC46 character standards, which didn't have "names" in
    the same strict sense as SC2 character standards.

    > >
    > >
    > > *shrugs* THe comment in N2610R refers to the name, not the spelling, I
    > > suppose. All of this is water under the bridge.
    > >
    > [PA] Yes, it would have been nice to have a bit more regularity (Y/J)
    > that is all I'm trying to say. And are ISO names of no importance in
    > cross-standard references (in other words, should Unicode have kept some
    > of the ISO 6861 names ?).

    Unicode could not have "kept" the ISO 6861:1996 names, because those
    names are not well-formed, from the SC2 point of view.

    Just one example, involving U+2C09 GLAGOLITIC CAPITAL LETTER IZHE

    ISO 6861 shows the one form at G0 set 4/9 and gives its
    "Old Slavonic name" as <i, z-hacek, e>. It shows the other
    form at G1 set 2/5 and gives its "Old Slavonic name" as
    <i, z-hacek, e>. In other words, the "names" were identical, and
    refer to the traditional Slavonic name of the letter, using
    a particular Latin transliteration scheme. The scheme itself
    goes beyond allowable letters for inclusion in SC2 (and Unicode)
    character names, so it had to be adapted, in any case.

    I agree that it might have been possible to make some choices
    that resulted in a somewhat more consistent adaptation for
    the Unicode names of Glagolitic characters, but given the
    artificial constraints of using only A-Z in character names,
    the Unicode character names can't generally be relied on as
    transliterations, anyway. There are just too many arbitrary
    conventions that over the years have been applied in trying to
    mash names into the A-Z constraint.

    > > and we called one INITIAL because it tends to come first and because
    > > we needed to have unique names.
    > [PA] Do you have any source for this ?
    > What I have mentions two theories for the need of these two izhe : 1)
    > the two triangular izhe [U+2C0A] is apparently used mainly for the « i »
    > conjunction [i without the prothetic "y"/yod in areas other than Moravia
    > [Glagolitica, pp. 24-25], 2) it indicates the front i, while the vase
    > shaped Izhe [U+2C09] represents the central i (ɨ) [cf. Glagolitica, pp
    > 59-65]. Could it be that the initial should be interpreted as front here
    > ? In other words, could it be that the word describes here a phonetic
    > aspect rather than one related to the writing position ? Again, just
    > trying to understand.

    I have no information regarding the provenance of the "INITIAL" in
    the name of U+2C0A, to distinguish it from U+2C09. Whatever the
    actual distinction being made in text, it can be taken as simply
    distinguishing the two names, from the point of view of the
    Unicode Standard. Note that unlike Arabic positional presentation
    forms, the "INITIAL IZHE" in Glagolitic is not given any
    compatibility decomposition involving and "<initial>" tag, so
    there are no presuppositions made in default processing or
    normalization regarding the relationship between U+2C09 and

    If your concern is primarily how to correctly translate "INITIAL"
    into French for the French name of U+2C0A, my suggestion would be
    to let yourself be guided by your scholarship, unless Michael
    has any other insights into the origin of the name.


    This archive was generated by hypermail 2.1.5 : Mon May 16 2005 - 16:19:55 CDT