Re: A natlang-independant project

On Thu, 13 Jul 2000, Joe Mondello wrote:

> badraic kov ra:
>
> > Not that this is a bad idea, but these first words you've listed
> >  seem a bit abstract. How would you picture "light" or "sound"?
> >  A pictoral of some sort might be misconstrued by a reader of your
> >  book. Just a point to keep in mind!
>
> I'd thought about this.  at first, i think a large variety of objects would
> be pictured under the heading for a word such as "light".  for example, I was
> thinking that, for light, I would probably picture a light bulb, the sun, a
> fire, a lightning bug, basically, anything that gives off light.  the single
> word would refer not only to the light (pictured as rays emanating from the
> sun/lightbulb/fire/lightning bug) but to light-giving objects themselves.
> sound would be pictured as "rays" emanating from a mouth, a radio, a teapot,
> the handset of a phone and would refer in kind to all of these objects.  I
> believe this is called holophrasis (?).  as the project evolved, I would
> either use compounding (e.g. "sound-mouth" for "talk" or "word") or the
> defining of words with existing words in the language, such as:
> taras - sound
> open - mouth
> helgu - taras open (speak)
Wow; I must have missed this the first time around. This is a neat
idea! This seems to be almost the opposite of a project that Mark Line
had going a few years ago. His project, Classical Yiklamu, was going
to be a "Russian Lawn" language. (A Russian Lawn as I understand it is
a large grassy field where the paved paths are laid out according to
where people had made paths as they traversed the space.) Basically,
he created a little program which assigned words to entries from the
WordNet. This set of words (all 90,000+ of them!) was the full lexicon
of Classical Yiklamu. This huge set of words was then culled according
to frequency of usage--if a word wasn't used, it was dropped. The end
result after usage figures have been tallied is thus a sort of
semantic Russian Lawn with the unused paths allowed to grow over.

Your project also seems to be a Russian Lawn in a more authentic
sense; you need to blaze semantic trails as you go along rather than
using what's already available. The challenge will be to avoid English
categorizations.

Keep us posted.

Dirk

--
Dirk Elzinga
dirk.elzinga@m.cc.utah.edu