Re: conlan/natlang coincidences

On Tue, Jun 17, 2003 at 01:26:12PM -0400, James W wrote:
> This got me wondering about how we all approach the specific
> creation of words in conlangs, and how closely we try to relate them
> (or not!) to natlangs.
It depends on the type of vocabulary, of course.  Both of my current
projects are a priori; Methkaeki's roots come from enciphered words
from natlangs, while Okaikiar's roots are generated randomly by
computer.  In both cases, I rarely have conicidental similarities to
natural language words with which I'm familiar.

I'm becoming disenchanted with both approaches, though; they're too,
I don't know, easy.  Productive.  I can always pull a brand new word
out of an orifice whenever I need it, no effort required. If
the encipherment comes out unpronouncible, or the random word isn't
mellifluous, I just pick a different natural root or fire the
randomizer again.

I would really like to derive a language in steps from a protolanguage.
The protolanguage could still have randomly-generated roots, but there'd
be a small, fixed number of them, probably all monosyllabic, and I'd have
to come up with ways they may have been combined to represent new ideas
over time, as well as sets of sound changes, morphology changes
(separate turning into inflections, like "verb-did" -> "verbed" in English),
etc.  It would be intense, difficult, time-consuming - and very satisfying.

Ah, well.

-Mark

Reply