Re: PIE */t/ > Hittite t, z / tt, zz

From: elmeras2000
Message: 38470
Date: 2005-06-09

--- In cybalist@yahoogroups.com, "Petusek" <petusek@...> wrote:
> In the course material I've got, PIE */t/ > Hitt. /t, z, tt, zz/.
However,
> it isn't clear from the chart what conditions made the PIE */t/
> /z/, for
> example.
>
> 1. There's a note that PIE */t/ > Hitt. /z/ when followed
by /y/, is that
> correct?

Yes: IE *-ti > HItt. -zi; 3rd person endings of verbs supply the
best know examples. A cluser *-ty- is present, e.g., in derivatives
in -zzis like hantezzis, gen. hantezziyas 'first' from an adver
hanti 'in front' + suff. *-tyo-.

> 2. There's a note that PIE */t/ > Hitt. /zz/ intervocalicly, is
that
> correct?

Not really. The unconditioned reflex of intervocalic *-t- is Hitt. -
tt-, but it may of course be assibilated before following /i/:
daskizzi 'takes repeatedly' from *-sk^é-ti.

> 3. If PIE */t/ > Hitt. /tt/, what were the conditions or
circumstances?

That's the basic form. It is lenited to /t/ after long vowels,
perhaps also between unaccented vowels.

Jens