Re: String name and Character Name

From: Hans Aberg (haberg@math.su.se)
Date: Fri Apr 22 2005 - 13:34:38 CST

  • Next message: Dean Snyder: "Re: String name and Character Name"

    At 12:15 -0600 2005/04/22, John H. Jenkins wrote:
    >Personally, I tend to think that the best long-term solutions
    >involve Unicode emphasizing that these are really semi-arbitrary
    >character identifiers and not names per se -- perhaps, if that's
    >possible, taking to referring to them as character identifiers and
    >not character names -- and having an official or semi-official list
    >of localized names for actual use in UIs and which are explicitly
    >guaranteed *not* to be stable.

    You express essentially what I propose, even though the word
    "identifier" is one possible synonym for "name": The intent with a
    name is to identify the individual. In this case, the individual is
    an abstract character (excluding those that are merely code number
    placeholders). If one wants the abstract character to be uniquely
    identifiable via its name, then the name should be unique, clearly.
    Since folks are fighting so much over getting these names
    linguistically correct, it might be wise to check how words in
    natural languages are used: Then the word usually identifies merely
    one property of the object it represents. Thus, in different
    languages, one may use different properties to identify an object.
    For example, the handle of a cup may in Swedish be called (when
    translated back to English) be called "ear", which sounds funny to an
    English user. But one can find examples, where this happens in
    reverse. The property to use in order to identify a word is quite
    arbitrary. For example, somebody claimed that the word "kangaroo" is
    the native word for "I do not know", evidently created by some
    communication problems; still, if this is true, few would argue
    finding a better, more accurate, name for that animal. One cannot
    expect a name of an abstract character to capture all its properties,
    just as one never expects the words of the natural languages people
    use all day long to do it.

    -- 
       Hans Aberg
    


    This archive was generated by hypermail 2.1.5 : Fri Apr 22 2005 - 13:36:05 CST