Re: Terkunan > Trekunan?
From: | Henrik Theiling <theiling@...> |
Date: | Tuesday, May 6, 2008, 19:42 |
Hi!
Mark J. Reed writes:
> My first thought is that it's odd to have those sort of constraints in
> a lang that allows initial mf-.
Hehe. But Dutch allows quite wild initial clusters like [sXR] and
still has exactly the same tendency to not like r + consonant,
inserting schwas.
> Is that m vocalic?
Yes, unless a vowel precedes:
mperi [m=peri] 'empire'
u' mperi [Umperi] 'an empire'
> That's also a lot of compensatory strategies to throw into the mix. A
> timeline of when each was dominant might be in order?
I suppose I need to come up with that, yes. Currently, the GMP is
ordered, but there are no actual time frames assigned.
> But based on your account of Sardinian, it does at least seem plausible.
I should add that Southern Sardinian also removes rC, but it does it a
bit differently, namely by swapping it to Cr in place, not backwards
across a vowel:
VL cardum > Sard. kaDru 'thistle'
Terkunan, after this proposed change, would probably have 'kradu'.
**Henrik