Re: Constructive linguistics

Thomas Wier:

> Natural languages are *far* more complex and intricate than any conlang has ever been.
Yes, but so? Weather is far more complex and intricate than any system
for modeling it, yet we still do so and still achieve useful results.
Even if you assume (as I think you are) that a conlang is necessarily
*trying* to be a indistinguishable replica (or rather, exemplar) of a
natlang.

> If "constructive" or "creative" linguists has any place in academia, it would
> be from some kind of cross-disciplinary literary angle.
And thus is born a potential cottage industry in supplying conlangs to
new nations, novels, and D&D campaigns. ;-)


Kevin Athey:

"Consider for a moment a degree in art history with a consentration in
constructed languages.  That, I believe, is a far more reasonable
request to the academe."

Why necessarily art history, vs. applied art? (Instead of taking,
e.g., "oil painting 101" you'd be taking "creative morphosyntax 101"
;-))

However, that is plausibly viable - as a interdisciplinary art /
language crossover.


Gary Shannon:

"One possible exception would be in the area of auxlang
studies where it might be possible to test which
features of an auxlang make it most accessable, or
easy to learn for speakers of a large variety of
languages.  It might, in other words, be possible to
quantify candidate auxlangs by some objective measure
of "goodness", and in doing so, have more concrete
criteria for the design of future auxlangs.

Or, alternatively, it might be possible to demonstrate
that the whole auxlang movement is doomed to failure
(my own personal belief) and back up that claim with
objective studies of one sort or another."

*laugh* I agree with you, but do you really want to go into that,
seeing as last time I raised it it started a flamewar? :-P

Also, why not extend this to other (testable) goals? While I would
agree that aesthetics is no more testable in conlangs than in other
fields of art, one could theoretically test e.g. how well a particular
idea or worldview could be expressed with particular uses. Or how well
they stand up to noise while remaining recognizable. Etc.

(That last one, btw, would I think be marketable - e.g. to pilots,
military, and others who use static-y transmissions.)


Jörg Rhiemeier:

"Certainly.  Conlangs don't reveal more about natural languages
than model railroads reveal about real railroads.  This metaphor
of Jeffrey Henning is a good one.  They perhaps reveal more than
nothing, but certainly not much."

Why not liken it to model airplanes? These are used, e.g., in wind
tunnel simulations and other situations to hone one's understanding of
the processes the real thing will go through. This, too is extendable,
ne?

 - Sai