Re: [tied]

From: Miguel Carrasquer
Message: 38206
Date: 2005-06-01

On Tue, 31 May 2005 21:55:27 +0000, elmeras2000
<jer@...> wrote:

>--- In cybalist@yahoogroups.com, "Pavel A. da Mek" <a.da_mek0@...>
>wrote:
>> > May I ask about the reason for
>> > the reconstruction of the nominative marker as *-z?
>>
>> Because of the rule that the thematic vowel is in the o-grade
>before voiced sounds and in the e-grade otherwise.
>
>How nice!
>
>> > Is this just my theory,
>> > or have you have had similar thoughts yourself?
>>
>> I read it here on cybalist and found it convincing.
>
>That's even nicer, thank you!
>
>> (Moreover, such orthography helps to clearly distinguish
>nominative ending
>> from genitive ending, plural ending and ending-less s-stem forms.)
>
>It was designed to do just that. It feels *very* nice to be
>understood.
>
>> > If it is my teaching, I will like to make it clear
>> > that I see no reason to posit *-z for the stage of PIE
>> > that we reconstruct on comparative basis,
>>
>> Yes, it is the matter of the internal reconstruction rather than
>the comparative one.
>>
>> > I merely see a need to derive some sibilants from *-z at an
>earlier stage.
>>
>> Is it only the matter of the thematic nominative
>> or are there other instances of supposed *z?
>
>I have, with less persuasion however, suggested the same for the -s-
>of the sigmatic aorist which also lengthens. And I have suggested
>there is this kind of sibilant behind the /s/ of the neuter root
>noun *me:ms- 'meat' also because the Sanskrit form má:s has the same
>shape as the m-stem nominatives, Skt. ks.á:s 'earth', Avest.
>ziiå 'winter'. I have also assumed it for *ters- 'dry' because of
>the long *-e:- of Celtic *ti:ros 'land' which I took to be a
>restructuring of *te:rs as an s-stem *te:r-os.

The question is whether the o-grade of the thematic vowel
(o-stem nom.sg. **-V-z > *-o-s) and the lengthening (C-stem
nom.sg. **-VC-z > *-V:Cs) are the _same_ phenomenon (and a
third phenomenon is the loss of **-z after (most)
resonants).

I completely agree that the o-grade of the thematic vowel in
the nominative sg. indeed shows that the nom.sg. marker was
once **-z. This allows the rule for /e/ vs. /o/ to be
formulated in simple terms of voiceless (or silence) vs.
voiced at the relevant stage.

The lengthening is another issue, however. We know that the
same lengthening is also triggered in the same context by
*-h2 (**-VC-h2 > *-V:C(&2)), which suggests that the
relevant phonological feature is +fricative rather than
+voiced. If so, then both **s and **z could have triggered
'Szemerény lengthening'.

The loss of the sibilant in the nom.sg. after the resonants
/n/, /r/, /y/ and perhaps /l/ can perhaps also be attributed
to the former voicedness of the nom.sg. ending. That would
explain the nom.sg. [**-Vy-z > ] *-o:y of *oy-stems versus
the instr. pl. of o-stems [**-oy-s > ] *-o:ys, both with
lengthening, but with loss of the former voiced sibilant in
one form, and retention of the voiceless sibilant in the
other.

=======================
Miguel Carrasquer Vidal
mcv@...