Re: First Post and Proto-Conlang rough sketch

Jason Monti wrote:
> Okay, I was thinking about instead of limiting compounds to
> O-grade+Zero-grade, to allowing for Zero+E-grade for a word-class distinction.
>
> For example, to follow the way English does things in the word "record":
>
> A "REcord" is something that you can play on a turntable,
>
> but
>
> I can "reCORD" someone's voice.
>
> So what I would like to do is, given two roots:
>
> *piet and *gerts:
>
> Nouns: piotgrts ['pjot,gr=ts] OR gortspit ['gorts,pit] (not ['gort,spIt])
>
> Verbs: pitgerts [pit'gerts] OR grtspiet [gr=ts'pjet]
>
> I'm also reading up on creole grammars to find out what kinds of things they
> do. I realize already that just the small level of morphology I have here is
> quite divergent from the fairly non morphological creoles, but at this stage
> in the proto-language, perhaps I would like to retain a few features, like
> if a creole from the ancient past has changed a bit over time giving rise to
> this strange e/o/0-grade system.
>
> Finally, I want to make this an active/stative language like PIE was at its
> earlier stages (at least, according to the FAQ on Zompist) such that
> inanimate objects cannot be the subjects of sentences but that a
> grammaticallized and eventually fused mass-noun particle will give rise to a
> "feminine" whence there was only "masculine" and "neuter" (active/stative).
>
> I'm still rather conflicted over making it SOV or SVO, but my knowledge of
> PIE (what little of it there is) and my experience with Japanese tell me
> that starting out with SOV seems the best way to go.
>
> Thoughts? Opinions?
>
I suppose it all depends on two things: how much like PIE you want it to
be, and how much work you want to put in on the background.  If I were
making "a creole from the ancient past" that had changed "giving rise to
this e/o/0 gradation", I would just make the creole, then put it through
sound changes to get the gradation.  It'll look a lot more realistic
that way!