Re: Measuring a writing system "economy"/"accuracy"

From: David Starner (
Date: Tue Jun 28 2005 - 23:51:20 CDT

  • Next message: John Hudson: "Re: Measuring a writing system "economy"/"accuracy""

    On 6/28/05, Eric Muller <> wrote:
    > Has there been attempts to measure the "economy"/"accuracy" of writing
    > systems? If so, what measures have been used? And what are the results?
    > I suppose that IPA could be described as an efficient and accurate
    > writing system for any language, if one consider only the signs needed
    > to write a particular language.

    The IPA records the exact sounds, not the underlying forms. In many
    languages, it is better to use the unmodified form for the base; for
    example, in English, "eat", "eatten", but eat is often pronounced with
    an aspirated t and eatten with a tap instead of a t. Likewise, the s
    in cats is pronounced s and the s in dogs is pronounced z. Accurate
    invokes a question of accurate to what.

    The correspondence of symbols to the IPA is not a good measure of
    efficient. It's the number of symbols, or time taken to write the
    symbols or the space taken to write or print the symbols that would
    need to be measured. In some languages, recording syllables only or
    consonants primarily is more efficient than recording vowels and
    consonants. The logographic Chinese system doesn't primarily record
    sounds at all. Encoding a language in such a way that underspecifies
    the sounds can be efficient in many ways; English uses fewer letters
    than a phonetic scheme would need, and other languages use shorter
    words than they would if they marked all the sounds. To speak of
    efficiency, you need to talk about efficient how, and it's doubtful
    that the IPA will be significantly efficient for any modern language
    in most ways.

    This archive was generated by hypermail 2.1.5 : Tue Jun 28 2005 - 23:53:11 CDT