Re: SUO: Core Meanings and the SUO
Doug,
I had intended to get around to your note of Nov 26 for some
time, and I finally made the time to address it. It raises
some fundamental issues about differences in meaning that
arise in different languages and cultures:
>I think there is more of a difference that you allow between meaning
>variations in natural languages vs. meaning variations in programming
>languages.
I would state the distinction differently: the differences
between languages (natural or artificial) are much less
significant than the differences between cultures. There is
some correlation because historically languages and cultures
have both been associated with specific regions of the earth.
But with the advances in speed of travel, communication, and
immigration, the correlations and associations of languages,
cultures, and locations have broken down or been reshuffled.
Typical example: Until the 19th century, the Japanese language
had a series of color terms that were based on traditional
vegetable dyes, which tend to be unsaturated pastel hues.
The European languages, however, had a series of terms for
saturated colors, which were based on the brightly colored
banners and shields of medieval heraldry. Even though each
European language used different words, they were calibrated
to the medieval standards.
When the new coal-tar dyes were introduced from Europe, the
Japanese adopted the European color standards, but named them
with words selected from the Japanese language. As a result,
the Japanese language now has two sets of color terms: the old
words, which only apply to artifacts made in the traditional
way, and the new words, which are calibrated to European
standards (which are themselves at least as old as the old
Japanese standards).
>I am thinking about the concept of prototypical vs.
>nonprototypical exemplars of categories, as demonstrated by
>Elizabeth Rosch.
That concept is older than Aristotle, who formulated two kinds
of definitions: the method of prototypes for his biological
classifications and the method of definition by genus and
differentiae for his syllogisms. Both techniques have been
used for over 2500 years, and various people have debated the
advantages and disadvantages of both and their relationships
again and again. Eleanor R. just happens to be one of the
more recent oversimplifiers.
>I don't think that variations in meaning between "file" in UNIX and
>"file" in DOS are of the same type as the variations in meaning between
>"bird" as applied to a robin and "bird" as applied to a kiwi. For one
>thing, in the former case (as you point out) we are talking about
>variations between different languages, while in the latter case we are
>talking about variations within English.
The correlation with language (natural or artificial) is purely
accidental. The culture of medieval heraldry established an
international (i.e. pan-European) standard, which still has
an influence on modern languages around the world.
> For another, the use of the
>"words" in question is quite different in the respective contexts.
The same could be said of every word or term in every language
(natural or artificial): The critical distinctions used to
define them are determined by use.
There is a lot more to be said about all these issues.
John Sowa