Dude, if semitic languages could develop naturally, *anything* is possible.




> Rather than applying grammatical affixes to root stems, Feayran circumfixes
> its roots around its grammatical markers. In most cases this just looks like
> infixing, but in some roots, either the first or second component of the
> root is null. Thus, depending on the root, grammatical markings can appear
> as prefixes, infixes, or suffixes.

Aha!  That's true of a lot of infixes too, though.

>
> Examples:
>
> root - noun form - verb form
>
> *skaì - óaskaì (someone awake) - iváviskaì (I am awake)
> th*lme - thóalme (warmth) - thivávilme (I am warm)
> avalash* - avalashóa (hope) - avalashivávi (I am hopeful)
>
> Further complicating matters, the complexity of both root components
> varies--so, there is no "infixed markers fall after the first metric foot"
> or such pattern that I can see.

There's the problem.

Here's one solution that occurs to me.   The original root in the PL
for skai wasn't -skai but V-skai; the original root for th-lme was
th-lme; the original root for avalash- was sh-.  avala- is a prefix or
the element of a compound.  Then all you got to do is account for any
cruft before the first C, and come up with a way that V drops out word
initially (not hard).  The drawback is, depending how many of your
roots have something like avalash-, you could end up with a lot of
extra morphemes to account for.

I'd definitely analyze it as an infixing of the grammatical morphemes
rather than a circumfixing of the root, though.  I mean, unless you
want to.

The Unfolding of Language, by someone or other, is a pretty good
description of how some of the weirder bits of language (like the
semitic languages) evolve.  Might be a good resource if you can dig it
up.


> While I like the system, and I've had some simple conversations in the
> languages with learners to at least suggest that it's parseable, I'm totally
> at a lost as to how it could have developed diachronically, and without a
> protolanguage, I don't know how to go about related languages. I'm a little
> afraid I'll end up having to go in and gut my lexicon, redoing a lot of
> established words in an attempt to create some semblance of etymological
> feasibility--perhaps infixing roots grew out of old multi-word phrases?

Well, this could *be* the protolanguage.  There's nothing tht says
that a protolanguage has to be simple or even well understood.  I
mean, look at PIE.  If anything, most IE languages have simpler
grammars than PIE, for at least some values of "simple."



-- 
I have stretched ropes from steeple to steeple; garlands from window
to window; golden chains from star to star, and I dance.  --Arthur
Rimbaud