Re: *-T + dh- > *-zdh- ? *-tó- < *-dhó-?

From: stlatos
Message: 48335
Date: 2007-04-18

--- In cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@...> wrote:
>
> Burrow: The Sanskrit Language, p 90

> and on the other hand some ancient forms
> testifying to the existence of z instead of d as in Iranian : dehí
> impv. 'give' beside daddhí, cf. Av. dazdi, and dhehí 'put', both
> with e out of earlier az according to the rule below. Either
> this is a case
> of dialectal divergence, or the type dehí (< dazdhí) represents the
> regular phonetical treatment which has been replaced in the majority
> of cases by new analogical formations.
> "
>
> Does the latter paragraph mean (could be interpreted to possibly mean)
> that in Skt. eg. the combination -d + dh- > -ddh- went via a PIIr
> stage like this: -d + dh- > -zdh- > -ddh-?

More like: either there are dialects with different rules
or ordering of rules that give these different forms

or

-d(h) + dh- > -ddh- > -zdh- by reg. rules then in most forms
(in Indic) z was replaced by d(h) by analogy from other forms
in the paradigm where the original C wasn't changed (before a
vowel instead of t/d/dh, for instance).

However, this doesn't seem likely to me. A syllabic "laryngeal"
is lost between syllables regularly in Iranian before Ch+t > Cdh
as in:

dhugYh-x-te:r dhidh-xY-toi
dhugYhx-te:r dhidhxY-toi
dhughx-te:r dhidhxY-toi
dhugh-te:r dhidh-toi
dhugh-dhe:r dhidh-dhoi
dhug-dhe:r dhid-dhoi
dhug-dhe:r dhiz-dhoi
etc.

but in Indic it is lost between syllables only in a few cases
after Ch+t > Cdh:

dhugYh-x-te:r dhidh-xY-toi dhidh-xY-dhi bhudh-tos
dhugYh-x-te:r dhidh-xY-toi dhidh-xY-dhi bhudh-dhos
(double C forms geminate, no longer 2 dif.) bhu<dh>os
dhugYh-x-te:r dhidhxY-toi dhidhxY-dhi bhu<dh>os
dhugYh-x-te:r dhidh-toi dhidh-dhi bhu<dh>os
dhugYh-x-te:r dhid-toi dhid-dhi bhu<dh>os
dhugYh-x-te:r dhit-toi dhid-dhi bhu<dh>os
dhugYh-x-te:r dhit-toi dhiz-dhi bhu<dh>os
dhugYh-x-te:r dhet-toi dhez-dhi (analogy)
etc.

This order shows that whether there was originally an X
between C's has no difference in the Iranian outcome, but
does change the Indic outcome (with those words that lose
X undergoing changes more similar to Iranian and most PIE
> modern languages.

This requires a very early split between the branches,
with many later same (or sim.) changes shared between them
due to close contact for a long period. I believe this is
true since it's also needed to explain differences in other
branches (such as Khowar which must be in close contact with
Armenian during several early changes in that branch).