Re: AA and IE

In a message dated 15/04/02 19:22:30 GMT Daylight Time,
glengordon01@... writes:

> >My own opinion is that maybe Afroasiatic is ultimately related
> >to Nostratic and DC, and maybe Kartvelian is more closely related
> >than it is, [...]
>
> Kartvelian is more closely related to what than what?

Sorry. I meant Kartvelian is more closely related to
Nostratic and DC than Afroasiatic is to Nostratic and DC.

> >but I am not convinced that Kartvelian should be regarded
> >as being more closely related to IE than some familes that
> >are usually classified as DC - unless you can come up with
> >a (central) Asian origin for it :-)
>
> Linguistics determines language relationships, not archaeology.

Genetic relationships are a function of sharing a common
ancestor. In order to posit a genetic relationship, you
not only need to come up with the linguistic evidence,
which I agree is primary, but also a plausible story for
how come the ancestral homeland and the location of the
daughter are in two different places. Traditionally the
Nostratic homeland has been placed in the Middle East to
accommodate Afroasiatic and Kartvelian. If we take these
two most problematic families out of the equation,
everything is much tidier and points to a central Asian
homeland. Or if Kartvelian had a central Asian origin
too.

> >There has even been made a suggestion that the name Udi is a
> >reflex of *quti".
>
> Interesting. How come it isn't on Starostin's site? How many
> words are reconstructed?

Dunno. I am just quoting from "HU as an EC language". They
give a handful of Sumerian words which have unusual
phonotactics and mention Proto-Tigridian as well, but don't
go into it any further.

> I doubt PreIE was anywhere near the Caucasus at any time, so
> as a consequence I doubt ND loans exist in PreIE except via an
> intermediate language like AbAd or one of its ancestral forms.

I was thinking of the other way round, because I have
seen a number of isoglosses in N/D that remind me of IE.
I don't see how AbAd can have been in contact with both IE
and HU/ND but at the same time direct contact between IE
and HU/ND can be ruled out.

> >Ok. Verbs in Nakh cannot be borrowed.
>
> There's no such thing as a language that can't borrow verbs.
> There are languages that might resist foreign influences by
> tending towards creating native terminology for new things
> (like Finnish or German) but there are no impenetrable language
> fortresses that avoid foreign influence altogether.

That's as may be, and Nakh is as far from impenetrable as
you can get, because it's full of all sorts of stuff. But
verbs can't be borrowed into Nakh. They *have to* be
borrowed as noun+"to do". The reason is probably that
verbs are sooooo irregular. What paradigm would you use?

> >The phonological structural preferences of nouns and verbs are
> >different - see Johanna Nicholls "Chechen" for further details.
>
> Perhaps but how does this prove borrowing as opposed to a million
> and one other reasons for the differences?

The million and one other reasons being...?

> >I mean that words for certain concepts tend to be
> >derivations or compounds of other words rather than
> >independent roots. Thus "aunt" is "mother's sister" or
> >"father's sister".
>
> Means nothing. It happens in Swedish (farmor, morfar, farbror,
> faster, etc) and we all know that Swedish is not a creole.

Ok. In Nakh, the degree of lexical economy is also
reflected in concepts like "cold" being expressed by
"like ice"; "green" is derived from "grass", "grey"
is "mouse colour" and so on. Even most of the
postpositions are derived from nouns, including the
amusing re-use of t'e ("penis") to mean "on top of".

> >No, but Semitic does have personal pronouns. For example,
> >the 3rd person singular masculine pronoun in Semitic tends
> >to feature /w/ prominently and the feminine /j/.
>
> Yes, I forgot about that connection. Okay, that could be an
> exception but I still am not sure how Semitic suffixes can end
> up as prefixes in Nakh. Something is bizarre there.

It could be a coincidence too. But if HU and ND are related,
we do have to explain why N and D have noun classes (after
a fashion, because they don't exist at all in some
Daghestanian languages either) and H and U don't.

> >There are strange aspects to the behaviour of noun classes in
> >Nakh that could point to an external pronominal origin. For example the
> >class marker used by J and V (i.e. /w/) classes in
> >the plural varies between /b/ and /d/ depending on whether the
> >noun is used in the 3rd person or not.
>
> That's inherited. Check out NaDene's obviative and the pronouns
> used there. I also believe that both of these pronouns exist in
> Nostratic and that they had something to do with word classes.

Can you explain this further please, Glen? I can see how
this might relate to obviation, but Nakh has no means of
expressing obviation. In fact, obviative expressions are
specifically prohibited. Are you saying that this is
because the native obviational pronouns have been reused
as class markers?

Referring to Starostin's reconstructions:

> And I don't
> know about anyone else but... well, why the hell would you use
> an otherwise unattested phoneme *z to reconstruct "I"??

The Lezgian languages?

> I have a solution. First, split ND and AbAd apart and let's never
> speak of "North Caucasian" again. Second, group Hattic with AbAd
> ("Abadha") and HurroUrartian with ND ("Caucasic").

We seem to be in complete agreement here. Is something
wrong?

> Now, let's
> try and get further to the truth here. Maybe something like:
>
>
> Abadha Caucasic
> ------ --------
> 1ps *së *ci (erg. *aci; gen. *ici)
> 2ps *wë *xu (erg. *axu; gen. *ixu)
>
> 1pp *txa *La
> 2pp *sywa *s'u

You wouldn't mind showing us how you have come to
these results, would you? There is a hell of a lot of
variation in the daughter languages, particularly in
Caucasic, and a lot of them look nothing like this.
Also, you make no distinction between inclusive and
exclusive 1pp.

> I think that Caucasic is closer to BuruYen
> than it is to Abadha and the 1ps is an interesting link between
> the two groups.

Maybe it is, but don't go dragging it into the
reconstruction or you'll be making the same mistake as
Starostin with his "North Caucasian".


Ed.