Re: Primary Stem Formants: -*H, -*i/y, -*u/w

----- Original Message -----
From: "Piotr Gasiorowski" <gpiotr@...>
To: <cybalist@yahoogroups.com>
Sent: Monday, April 07, 2008 5:39 PM
Subject: Re: [tied] RE: Primary Stem Formants: -*H, -*i/y, -*u/w


> On 2008-04-07 23:32, Patrick Ryan wrote:
>
> > Many illogicalities which are tolerated to allow the theory to be
> > maintained: e.g. 'laryngals' color short vowels ([*e]) but not long
> > vowels
> > ([*ee]). I would even accept 'breaking' but _no_ effect? Totally
> > unlikely.
>
> Why?

***

Because [h] affects any following vowel by, at a minimum, lowering it. This
is physiology. That is, essentially, what *H2 is asserted to do.

Even in English, though the vocalic allophones are minimally distinct from
their pronunciation in other phonological environments.

My explanation is far more conformant with reality: once a following has
been lengthen by *H, *H(whatever subscript) has no further affect on the
vowel; and, eventually, disappears if not immediate, in which case the
lengthened vowel is a result of compensation.

***

> > Unless you do not consider pronouns 'roots', *me, *te, *so, *to, *ne,
> > *se,
> > *kWé/o, *yo.
>
> I said "lexical roots". Those listed above are grammatical morphemes
> ("function words").

***

That seems like quibbling to me. *te means 'you'; that is its function.

***

>
> > How does Pinault explain why the 'laryngeal' did not simply become schwa
> > (*&)?
>
> Pinault merely observes that a postconsonantal laryngeal coalesces with
> a following *j and eventually disappears, so that we get e.g.
> *h2arh3-je/o- 'to plough' > *h2arje/o- and *krouh2-jo- > Lith. krau~jas.
> The effect is no doubt due to the syllabification of *-CHj- as *-C,Hj-
> because of the greater sonority of *j.
>
> > I find that explanation highly unsatisfactory. Where is the
> > stress-accent in
> > *wr.H(1)-?
> >
> > My suggestion, *wréH(1)- is far simpler; and shows a stress-accent where
> > there ought to be one.
>

Pinault is fine, and he made be right, but it also might be that an
unextended form, *wer-, is preferred before -*ye/o?

> Your suggestion doesn't explain why /wre:/ appears only in Greek, where
> it is the regular outcome of both *wreh1 and *wr.h1, hence the
> possibility of generalizing /wre:/ as a secondary full grade in original
> derivatives of *werh1-. *CeRh1 and *CReh1 roots are confused in Greek to
> a larger extent than elsewhere because in addition to sharing the same
> zero grade they merge it with the full grade of *CReh1-.
>
> Piotr

***

But it does not appear only in Greek, I think.

What about Avestan <urva:ta>; is this not a metathesis from *wre:-tó ???

***

Patrick