Re: 'lower case a' and 'script a' in unicode

From: James Kass (
Date: Thu Mar 24 2005 - 23:37:23 CST

  • Next message: Werner LEMBERG: "Re: 'lower case a' and 'script a' in unicode"

    John Hudson wrote,

    > This whole thread is happening precisely because the needs
    > of IPA transcription differ from conventions of general
    > purpose typesetting. This is not surprising, since IPA
    > attempts a precision of phonetic representation that is
    > well beyond that of the normal alphabet of any Latin script
    > language. One of the ways in which IPA does this is to
    > assign a distinctive meaning to forms that in general
    > purpose typography are considered purely stylistic variants
    > of the same letter. As a result, IPA transcription requires
    > fonts that do not follow typical typographic conventions in
    > the distinction between roman and italic styles. In simple
    > terms, an italic IPA font needs to be something akin to an
    > obliqued version of the roman, rather than a distinct style
    > of lettering. Ergo, general purpose multilingual typesetting
    > cannot be said to include IPA transcription, which is a
    > specialised technical kind of typesetting, not very
    > different from e.g. mathematical typesetting in the way that
    > it assigns distinct meaning to stylistic variants of
    > letterforms.

    You lost me at the "Ergo".

    If we exclude dictionaries, lexicons, and all manner of linguistic work
    from general purpose multilingual typography, the purpose ceases to be

    The analogy with mathematical typesetting is interesting, but mathematical
    typesetting involves complex layout which is beyond the (present) scope
    of plain text. It is this complex layout requirement which would tend
    to remove math typesetting from anything considered "general purpose".
    IPA has no such stricture, however.

    Of course, math typesetting also assigns "a distinctive meaning to forms
    that in general purpose typography are considered purely stylistic variants
    of the same letter". Unicode has handled this by assigning distinct
    characters to those distinctive meanings. Not only for the math
    alpha-numerics, but also for various technical symbols found in the
    letterlike symbols range. Likewise, Unicode has assigned separate

    At the character level, there is no confusion between the two -- the
    binary strings are different. This thread is a font issue.

    Ergo, whether one considers IPA transcription to be part of the general
    purpose scheme of multilingual things or not, the proper way to retain
    any desired distinction remains to choose an appropriate font. With
    the caveat that somebody, someplace, will be displaying your carefully
    constructed page with "Comic Sans".

    Best regards,

    James Kass

    This archive was generated by hypermail 2.1.5 : Thu Mar 24 2005 - 23:38:01 CST