Re: Why grammar is so complex a subject

On 12/28/05, Henrik Theiling <theiling@...> wrote:

> Gary Shannon <fiziwig@...> writes:
> >...  Therefore conlangs should not be "designed", they should be
> > "used into existence." ...
>
> Hmm?  Why (not)?  They are conlangs.  Of course, you may very well use
> a different approach, but since grammar is a model for simplifying the
> chaos, it seems like a very feasible approach to describing a lang and
> thus, a conlang.
The approach Gary suggests is probably a good one for
naturalistic artlangs.  What also works well is a method
involving a priori design of some aspects of the language
and allowing other aspects to emerge gradually from actual use
of the language.  In fact, I suspect that  pure a priori design
may be impossible if you want a language complete enough
to use -- you probably can't foresee every
issue at the beginning, so some grammatical and
semantic issues can only be resolved
after you've started using the language.  Some of the
problems that emerge at this stage would require,
for their optimal solution, fundamental changes that would
retroactively make much of the corpus ungramatical,
so you may prefer a less optimal solution that allows
the existing corpus to stand (and doesn't require you
to re-learn basic aspects of the language).  This
is the approach I'm taking with gjâ-zym-byn; for the
first few years it was all designed a priori, but once
I started using it extensively, I ruled out any changes
that would make the existing corpus obsolete or
require me to relearn aspects of the language
I'm already fluent with.

--
Jim Henry
http://www.pobox.com/~jimhenry/gzb/gzb.htm
...Mind the gmail Reply-to: field

Reply