Script Continuums (Was: Re: Greek glyphs)

Date: Mon Feb 28 2005 - 13:25:09 CST

  • Next message: Brian Gould: "Using java mail to send email with subject in chinese (from Unicode)"


    I am indeed familiar with your excellent and important font. Thank you for all
    your work on that for Greek users.

    It is exactly the two (presently unavoidable) drawbacks of your font, that I'm
    trying to solve for the archaic letterforms: that they can only be used in the
    non-standardized, problem-strewn PUA, or selected manually in a page layout app.

    My goal is standardized, universal access in Web pages.

    In looking at all of the ancient (really pre-modern) writing systems, I
    believe the essential thing that needs to be done is to approach them not as
    "scripts", but as "script continuums".

    Organizationally, for ease of use (in encoding and changing variants), and for
    data identity (of variants as a conceptual sub-catagory of the main standard
    form), the only way to properly implement a "script continuum" in Unicode, I
    believe, is with a "grid" rather than a "line".

    That is, a script like standard Greek (plus a few additional archaic
    "characterhoods") are assigned actual codepoints (forming a linear layout, x
    axis), and then some means is used to make selections along the spectrum of
    variants (layed out in a grid, selecting along the y axis.)

    Any attempt to forceably and artificial standardize a complex ancient script
    continuum like cuneiform, for example, destroys much of the purpose of
    encoding in the first place (academic study of the variants themselves). At
    the harshest critique, using an artificially standardized script on a highly
    divergent sample -- rather than simply transliterating -- approaches the
    kitsch of using an "ancient look" font to set the mood.

    The best available approach for this presently, I've concluded, is a series of
    fonts, all using the same standard codepoints. But this could easy mean 50 or
    more different fonts for a script continuum widely spread over time and place.
    Awkward, not distributable as a widely installed standard -- and striking like
    the 1995 Internet where you had to switch fonts to go from Latin to Cyrillic
    for example.

    And amusingly/annoyingly, some people, having previously dismissed the
    alternate forms as un-encodeable "merely glyph variants", now start getting
    mad at me because I am putting "different scripts" on the same codepoints!
    Just as there is no objectively definable line between a language and a
    dialect, there is no objectively definable line between a "script" and a "variant".

    Which leaves us with a shaded grid. A spectrum. A continuum.

    That is the essential and unavoidable nature of ancient scripts..

    I believe Unicode has simply failed to provide (or allow use of) any
    appropriate architecture for encorporating this reality into Unicode.


    David J. Perry wrote:

    > Have you taken a look at the most recent version of my Cardo font? It
    > contains, ASAIK, every extant variant of Greek letters. I have made use
    > of the OT stylistic alternates feature to provide access to these glyphs
    > while using the standard Greek alphabet Unicode values. You need Adobe
    > InDesign to see this feature in action. (Mellel also supports OT but

    > widely supported in a couple of years. The Greek variants also have PUA
    > values which can be used until OT support is more widespread. (Relying
    > on the PUA is a bad idea, of course, but it's all that most users have
    > right now.)

    This archive was generated by hypermail 2.1.5 : Mon Feb 28 2005 - 13:11:55 CST