Re: Res: [tied] Etymology of Rome - h1rh1-em-/h1rh1-o:m-

From: Piotr Gasiorowski
Message: 47847
Date: 2007-03-14

On 2007-03-14 15:03, alexandru_mg3 wrote:

>
> I don't know if I could well present my point (with my English)=>
> To rsume: there are 2 points to clarify here:
> 1. Who really trigger the vocalisation of the Resonants and who
> doesn't.
> 2. Who cannot trigger the vocalization of resonants but can
> influences the value of the resulting output

Sorry, I don't understand either question. Could you rephrase them?

> My argument:
> The fact that in *tnh2-u we don't have the same output with *h2nh2-
> ti clearly indicates that the Laryngeal Has No Role To Trigger your
> Supposed Vocalisation of R, otherwise the output would be the same
> (if the laryngeal would have been the Role as "That one that
> closes/defines the Context".

Syllabic resonants in interconsonantal positions arose in PIE, while the
loss of laryngeals took place independently in the individual branches.
The _vocalisation_ of syllabic resonants (i.e. the appearance of a prop
vowel next to them or replacing them) is still later, hence its
different outcome in different contexts.

> a) so in h2nh2-ti the vocalisation n. > a is trigerred by -t- NOT
> by h2 and once we said this, we will logically complete that
> b) in tnh2-รบ there is NO Vocalisation of n because the Required
> Context is not there ...

What happened in the sequence *n.h2V in Indo-Iranian was as follows:

(1) The laryngeal was lost, which resulted in the rise of prevocalic
*/n./ in the sequence *n.V, which may have been pronounced as *n.nV with
a "linking" non-syllabic *[n] filling the hiatus between syllabic
segments (just as */iV/ was pronounced *[ijV]).

(2) The syllabic *n. developed as in other positions, yielding a nasal
vowel, eventually denasalised and ending up as /a/. The whole sequence
became /anV/.

> => once we agree that h2 has no Role to Trigger a R-vocalization:
>
> b) the value of the output nh2/V > an etc.. is secondary in this
> discussion: the main point is that: THE LARYNGEAL CANNOT TRIGGER BY
> ITS-SELF THE VOCALISATION OF THE RESONANT (IT CAN ONLY INFLUENCE THE
> VALUE OF THE OUTPUT)

It makes it syllabic in PIE -- that's what counts. Prevocalic sequences
like *CiH-, *CuH-, *Cr.H-, *Cn.H- etc. retained a syllabic resonant
after the loss of the laryngeal (and developed a transitional glide like
*Ci[j]-, *Cu[w]-, *Cr.[r]-, *Cn.[n]-). As regards the later vocalisation
of the resonant, it generally proceeded in the same way as in other
positions, though there are occasional divergences. For example in Latin
prevocalic syllabic liquids are vocalised with /a/ (from an epenthetic
"schwa secundum") rather than /o/, so a sequence like *Cr.HV- develops
into /CarV/).

> c) to come back to *h1rh1-o:m-eh2 we have the same thing : the h1
> CANNOT TRIGERRED ANY VOCALISATION OF R

It functioned as an obstruent consonant in PIE, so a resonant flanked by
two *h1's was _syllabic_ in PIE. It was vocalised later in
branch-specific ways.

> d) Finally regarding the Latin output of HrH/V => you need to show
> me another example HrH/V- in Latin before to trust you that:
> Latin ro: < PIE h1rh1-o: (Ro:ma < *h1rh1-o:m-eh2 ) is not possible
> in Latin.

Am I really obliged to prove that it's impossible? I should think it's
your duty to show that it's possible, e.g. by presenting transparent
examples of such a treatment of initial *HrH-. Bu the way, why long *o:
in the suffix? It looks unmotivated to me.

> P.S. Constructions like RHV > R.RV quoted on different books ... have
> Only the value of a Formula (But it's a confusing model, at least
> this is my opinion)

I hope my elucidation will make you feel less confused.

Best,

Piotr