Re: Stacking up on standard works

From: Tavi
Message: 69175
Date: 2012-04-01

--- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...> wrote:
>
> So what prevents you from demonstrating its application? On the one
> hand, you promise a lot; on the other, you sound very evasive when
asked
> to back up your claims.
>
I've already told you where and when to look.

> > I never said Vasco-Caucasian was a family but a MACRO-FAMILY, that
is, a
> > distant linguistic relationship.
>
> Linguistic taxa, like other clades defined on the basis of shared
> ancestry, are all of the same kind, irrespective of their time depth.
> They consist of a most recent common ancestor plus all its
descendants.
> There are no other valid *genetic* groupings. It doesn't matter what
you
> call them -- a branch, a family, a macro-family or a
super-mega-phylum.
> A family is nothing special or privileged -- it's just a conventional
> term for a linguistic clade whose wider relationships are obscure at
the
> current state of our knowledge. It's in fact our present ignorance
that
> defines it. If one day we demonstrate beyond reasonable doubt that
> Uralic and Indo-European (or Na-Dené and Sino-Tibetan, or whatever)
are
> related, we shall have combined two or more families into one (as has
> happened before in several cases). If Basque and (NE?) Caucasian
derive
> from a common ancestor, they are by definition members of the same
> clade. Either you can prove that by applying the comparative method in
> the same way you would apply it to any "family", or you can't. So you
> can't, can you?
>
Considering most Vasco-Caucasian languages have become extinct and only
survive in loanwords to other families such as IE, a direct
reconstruction of Proto-VC would very difficult if not impossible.
However, it looks like Starostin's PNC (actually Proto-NEC with
Proto-NWC fitted in) is a much older entity than commonly though, so
IMHO it constitutes a good approximation to the real PVC.

BTW, I don't think "Indo-Uralic" is a valid node. AFAIK, their
respective lexicons look so far apart for suporting a close
relationship. IMHO Altaic is a better candidate.