Re: Dumezil

On 2008-10-12 22:53, Edgard Bikelis wrote:

>
>
> Hm, as I see, the suffix is the same and in the same grade: flamen from
> *...-me:n, like pater from *ph2te:r, but can also be from -m.n. If the
> root is in the zero-grade, then the full grade of the suffix is more
> probable, like poimé:n, so all three (flamen, brahmán and poimé:n) are
> nomina agentis.

The zero-grade of the suffix (*-mn.) is unlikely in an animate noun, so
the reconstruction *-mé:n is quite safe.

> Now the root. Without any initial guess, f- in latin can be from *bh
> (fero:) *dh (fores) *gWh (formus). la: from 'lH'. From latin that's all,
> I guess. Pokorny gives *bhl&gh-s-men to sufflamen, but does it have to
> do to the priest? like... flamini non fas bigam agere ; ). And why this
> *-s- exactly?
>
> The vedic indeed points to *bhleg'h. The problem here, then, seems to be
> the a: from flamen. If it was just bh.lg'hmén, it should be like
> **fu:lmen or something. But if there is such an animal like an original
> *a ; ), *bhlag'hmen *flahmen (like, say, hanser), fla:men, ecce. I think
> that is the original proposition.

We have no other examples of *g^Hm in Latin, but given the behaviour of
other stops, one would expect phonation to be neutralised before nasals
in inherited words, with *-k^m-, *-g^m- and *-g^Hm- all yielding <-gm->
(=[-Nm-]?) as in <segmen>. We know, on the other hand, that *-VCsm- >
-V:m-, hence presumably the idea that there should be an *s in the
cluster, even if there is a fundamental *a in the root -- in other
words, any *bHlaC-s-mé:n would do. A connection with <brahmán-> looks
tantalising, but there are so many imaginable sources of either form
(there could be no *a and no *s but a laryngeal in the Lat. word, and
the Skt. one could contain anything from *mreg^H- to *bHlag^H-), that
chance agreement is quite possible. The only thing that matches for sure
is the suffix *-men-.

Piotr