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

On 2007-03-15 01:38, alexandru_mg3 wrote:

> a) You try to tell me that Skt. anánt is 'the most natural possible
> output of *h2n.h1ént ?

Why should I say anything so odd? What do you mean by "natural"? I'm
only saying that if the normal outcome of *Cn.HV- is CanV- (cf.
*tn.h2ú-), I would expect the same when the first consonant is a
laryngeal. Why not?

> 'The most natural' should be a-ant > *a:nt- and Such forms with
> long a: really exists see: a:ni:t [3sg.impf.act.]

Have you ever heard about the augment? In the 3sg.impf.act. form you
have the _full_ grade *h2an&1- (as in the present tense ániti <
*h2án&1-ti) contracted with the preterite marker *é-. The sequence *é
h2an- becomes *áHan- > *áan- > Skt. á:n-.

> So Both examples are constructed based on an IMPOSSIBLE TO PROVE OR
> TO DENY SUPPOSITION:
> "The hiatus that appeared is filled 'exactly' with the missing
> phonem: n or w"
> True or not : you cannot say that this is a general linguistic
> rule...this could be POSSIBLE.

It is not only possible but actually attested in practically all the IE
branches. Not only when a laryngeal is lost between a syllabic resonant
and a vowel, but also when a resonant becomes syllabic via Sievers' Law
(as in Skt. da:s'u[v]ás- from da:s'- + -vás-) and in Lindeman variants
such as *di[j]e:us or *k^u[w]o:n.

Piotr