Re: The English Phonotypic Alphabet

From: Kenneth Whistler (kenw@sybase.com)
Date: Mon Oct 02 2006 - 16:49:54 CST

  • Next message: Michael Everson: "Re: The English Phonotypic Alphabet"

    David Starner asked:

    > http://books.google.com/books?id=gV0KAAAAIAAJ&pg=PA2ws Alexander
    > John Ellis's English Phonotypic Alphabet. It's been used a couple
    > other books, though I don't know if it was ever used by any other
    > author. Has anyone looked at encoding it?

    "It" would not be encoded, as Latin-based phonetic alphabets are
    not encoded per se in Unicode.

    The English Phonotypic Alphabet is one of a number of attempts
    at developing a systematic phonetic alphabet, going back at least
    as far as Benjamin Franklin, during the 18th and 19th centuries,
    and culminating in the eventual standardization of IPA.

    Most of the non-A-to-Z characters in the English Phonotypic Alphabet
    shown in this book sample are already encoded as Latin letters
    in Unicode. The few that aren't would be perfectly appropriate
    as proposed additions to the Latin repertoire, if someone would
    care to do the research and write up the appropriate proposal.

    Note that there would be some casing issues for any proposal to
    complete the set of characters needed for the English Phonotypic
    Alphabet. Ellis was mixing up the need for an orthographic
    reform for English with the need for a technical phonetic alphabet.
    Thus the EPA maintains orthographic casing rules for English
    and requires both uppercase and lowercase forms for each
    letter. This is non-problematical in most instances, but in
    those instances where Ellis adopted an unusual letterform in
    lowercase, he apparently just invented fanciful uppercase forms.
    This might not be an issue for some case pairs where both lower-
    and uppercase are inventions, but for others the uppercase might
    not match what one would expect.

    For example, I would suggest U+1D8F, U+1D96, and U+1D99 for the
    a, i, and u with right hooks, respectively, in the EPA. But each
    of those has a fanciful uppercase form, and in the case of
    the a with right hook, looks too much like an uppercase A-macron.

    Also, the EPA definition uses a normal uppercase "A" as the uppercase
    of the rounded-top-a (= U+0251) for the ash sound (IPA U+00E6), but
    uses an invented rounded-top-A for as the uppercase of the normal
    roman a (= U+0061) for the a sound (IPA U+0061).

    Because of these kinds of oddities, any proposal for EPA letter
    additions would have to pay some attention to casing -- and my
    suggestion would be to eschew expectations of regular casing
    for EPA letters. Expecting all EPA casing relations to be
    encoded directly would lead to more confusing letter clonings
    likely to be rejected by the UTC.

    --Ken



    This archive was generated by hypermail 2.1.5 : Mon Oct 02 2006 - 16:57:09 CST