Re: Character Identity and Font Selection

From: Jukka K. Korpela (jkorpela@cs.tut.fi)
Date: Wed Jun 08 2011 - 15:03:42 CDT

  • Next message: Andrew West: "Re: Character Identity and Font Selection"

    8.6.2011 14:21, Otto Stolz wrote:

    > • Many IPA symbols are unified with small Latin, or Greek, letters

    That’s somewhat odd because it makes IPA symbols belong to different
    scripts. But that’s basically just odd, not problematic.

    > e. g., the IPA Open Front Unrounded Vowel, and the Latin Small
    > Letter A are unified as *U+0061*. As the Latin Small Letter A
    > covers a wide range of glyph designs, including some like the
    > IPA Open *Back* Unrounded Vowel (e. g. in the Comic Sans MS font),
    > you may get the wrong glyph (belonging to some other character!)
    > if you display IPA data in a font not designed for IPA.

    I wouldn’t go that far. There are large fonts that cover IPA without
    being particularly designed “for IPA.” And while some fonts indeed are
    unsuitable for presenting IPA, partly due to the shape of “a,” there are
    many fonts that are unsuitable for presenting many types of text. The
    glyph for “a” in Comic Sans is as such no more wrong than Comic Sans in
    general; but it implies that Comic Sans is unsuitable for IPA. This is
    evident for an even more obvious reason: Comic Sans lacks IPA extensions.

    Using IPA, you should consistently use one font for all IPA text, if at
    all possible. Otherwise there is a risk of gross stylistic mismatch and
    even misunderstandings, as the IPA symbols are rather special.

    >If IPA characters cannot be dis-unified from
    > Latin, and Greek, characters, eventually the text-processing, and
    > the rendering, software should solve the problem via language data,
    > as outlined above; i. e., IPA should be handled as a ‘language’,
    > in its own right.

    That would be an odd move, for any normal meaning of “language.” There
    is already a lot of confusion around that word; let’s not add to it. IPA
    text is a presentation of a human language and specifically designed for
    presenting _any_ language, or at least a very wide range of languages.
    English written in IPA is still English, German written in IPA is still
    German, and a piece of multilingual IPA text still contains words in
    different languages. It is a special writing system, or orthography. It
    would be a category error in principle and greatly confusing in practice
    to call it a language.

    IPA is no more a language than the Latin alphabet is a language.
    (Actually less, since the original version of the Latin alphabet was
    developed for writing a particular language; IPA wasn’t.)

    Whatever you mean by handling IPA as a “language,” please find another
    word for the concept.

    -- 
    Yucca, http://www.cs.tut.fi/~jkorpela/
    


    This archive was generated by hypermail 2.1.5 : Wed Jun 08 2011 - 15:06:05 CDT