Re: Latin -idus as from dH- too

Richard, perhaps I could have written more clearly:

?+VoicelessC <M> VoicelessC+? -> VoicedC

I am not really dropping the glottalization.

However, the Tai reconstructions give me reason to think that I may be
wrong.

However, I still do not believe that PIE *H+*t -> *d.

Do you?


Patrick

***

----- Original Message -----
From: "Richard Wordingham" <richard@...>
To: <cybalist@yahoogroups.com>
Sent: Friday, March 14, 2008 11:29 PM
Subject: Re: [tied] Latin -idus as from dH- too


--- In cybalist@yahoogroups.com, "Patrick Ryan" <proto-language@...>
wrote:
>
> Richard, I cannot believe the pre-glottalization voiced stops.
>
> I would accept a metathesis (VoicelessC+? -> VoicedC).
>
> Could that work?

I had written:
> 2. The modern Tai voiced stops are generally reconstructed as
> deriveing from pre-glottalised stops. Pre-glottalisation is evidenced
> by pre-glottalised reflexes in a very few languages and by
> pre-glottalised stops and glottal stops having the same effect on tone
> in a few dialects.

It sould be very odd in the Tai case, for it would violate the
principle that tone splits are determined by the initial consonant.

But why drop the glottalisation as part of the metathesis? In some
Afroasiatic languages, it seems that the emphatics of Semitics
correspond to implosives or ejectives depending on the point of
articulation - voiced implosive at the front, voiceless ejectives at
the back.

Richard.