Re: [tied] PIE *HRHV > Pre-Latin HRVH > Latin RV: (*h1rh1-om-eh2 >

On 2007-03-19 04:07, alexandru_mg3 wrote:

> For the reconstruction with an initial laryngeal of the 'nose'-root
> see:
> (see Lubotsky - Leiden)
> "IE form: Hn(e)h2s- "

The initial laryngeal is guaranteed by Old Indic data (lengthening of
the final vowel of the first member in compounds). I don't believe in
such CeHC/CHeC/CHC- ablaut, however. There is no problem whatsoever if
one accepts PIE *a and the possibility of *a:/*a (Lubotsky doesn't but
many other people do). The word is therefore *Hna:s-/Hnas-.

> The hint here is Latin *na:rem < Pre-Latin Acc. *hneh2sm < PIE Acc.
> *hnh2esm- versus the Lat. Acc. salem < PIE Acc. *sh2elm (for a
> similar construction)
>
> So we have a laryngeal methathesis in this case *HR/*HV > *HR/*VH
>
> ===========================================================
> Rule: PIE *HRHV- > Pre-Latin *HRVH- > Latin RV:
> ===========================================================
>
> So applying the rule above we can derived Ro:ma in a regular way:
>
> PIE *h1rh1-om-eh2 > Pre-Latin. *h1roh1m-eh2 > Ro:ma
>
> with -om- initially and not with -o:m-
>
> (so the second issue 'why -o:m-?' dissapears too)

Even accepting the above paradigm of the 'nose' word for the sake of the
argument, it will get you this kind of variation only in a consonantal stem!

Piotr