Re: Ubykh suspicions (was brz reloaded!)

Paul Bennett wrote:

> Well, that's just it. How do you tell those two patterns apart?
>
> To me, /1/ + /C/ > /iC/ is isomorphic with /i/ + /k/ > /iC/ (absent
> any  other data), and that's the crux of the biscuit for Ubykh: only
> certain  vowel and consonant combinations (I think it's CV rather than
> VC) are  allowed. I lack an understanding of how it can be proven one
> way or the  other.
Well.

You need more data, that's how. Suppose you have some forms,

MEREG
MERIK

And you know that these decompose to merV-C, morphologically. Either
there are two merV forms and the C is varying phonologically, or vice
versa. Suppose that you are certain C is a number marker... now you have
two possibilities:

MEREG means something like "one tapestry", and MERIK is the same stem
but with a different number, like, "excess tapestries." Thus, you have
consonant triggering a vowel alternation.

Or, MEREK and MERIK have the same number, but, like, one means "moon"
and the other "tapestry." Thus, you have vowels triggering consonant
alternation.


--
The "Million Style Manual" is a set of sixty-four jade stones marked
with pieces of Chinese characters. It expresses the kung fu of the void,
as taught by P'an Ku's axe.

Shreyas Sampat
http://njyar.blogspot.com