Re: THEORY: A possible Proto-World phonology

> Date:         Wed, 28 Jun 2000 09:19:14 CDT
> From: Danny Wier <dawier@...>
>
> Okay.  Let's suppose, as the Bible and other religious texts say, that all
> the world's languages came from one common ancestor.  I compared the
> reconstructed phonology for Nostratic, all three kinds, and Starostin's
> reconstructed phonemes for Proto-North Caucasian (thanx for the Ubykh
> list!).  I also remember vaguely the possible inventory of Proto-Salishan
> and even less of Proto-Na-Dene.  A little about Proto-Sino-Tibetan as well.
>
> It is my opinion that languages become phonemically simpler but
> syntactically more complex as languages age.  After comparing large
> consonant phonologies (which tend to have small vowel inventories), I
> decided to take a wild guess at what might result from a speculation of the
> very roots of human speech, many millennia ago...
>
> Of course this is far from exact science.  This is radical theory and I'm
> not a big time linguist by any means.  At least this could be used as the
> basis of a conlang (like my "Nine", or "Ini" which Nicole came up with, but
> in the opposite direction).
>
> So here is a possible "Proto-World" (or "Proto-Language", "Proto-Human",
> "Proto-Earth" etc.) consonant and vowel system:
Positing that all human languages descend from a single recent
ancestor is an excellent base for creating a conlang.

However, most researchers seem to agree that humans have been using
languages of exactly the same type as the ones we record now for at
least thirty thousand, more probably a hundred thousand years.

So if there is a single ancestor, it is almost certainly totally
unrecoverable. As in, not one single root or morpheme, and nothing
about its typological properties or phonology can be reconstructed.

I'm not sure I agree about the overall tendency to trade off phonemic
complexity for syntactical --- there certainly are exceptions. And if
there is an overall trend now, it cannot have been constant during the
huge prehistory of language: that would mean that the first language
had one-word sentences and one thousand different consonants.

Digression on reconstructive success:

The huge dispersion of Indo-European --- even huger if you include its
Nostratic cousins --- does give the impression that if we could but go
one 'layer' deeper, we must find relationships over even larger areas,
and that the whole world could be covered in a few easy steps.

However, the spectacular results of IE comparative linguistics do not
quite seem to be repeatable in other parts of the world. The existence
of long-distance recoverable relationships may well be a result of the
start of agriculture, dating on the order of ten thousand years ago
--- and the concomitant invention of writing.

Arguably, the reason why IE can be reconstructed so well is that there
are many surviving branches, and that many of the branches have
written sources or can otherwise be reconstructed back to a few
thousand years ago.

If the original speakers of Nostratic had stayed where they were,
their languages would presumably have changed just as much. But only
two or three of them would have survived. I doubt that PIE could be
reconstructed in the detail it has been starting from Neapolitan,
Delhi Hindi and Copenhagen Danish, with no written sources.

(One of the newest theories is that the breeding ground of Nostratic
was on the marshes around the dried-out Black Sea back in 9000 BP or
so --- and that the splitup happened when the Straits opened and
people fled up the nearest river valley).

The point is, most of the rest of the world seems to be like that.
Geographically close languages are often provably related, and it is
even possible to argue for larger groups, but the sheer mass of data
needed to reconstruct back to something at the level of Nostratic is
missing.

Add to that that the relatives of Nostratic were probably spoken in
areas close by --- and when the expansion started, most of them got
swamped. But some people do see Etruscan as possibly branching off
below Nostratic.

Other ancient languages of the Middle East, e.g., Sumerian, have no
discernible relation to anything Nostratic. If we could go back to
7000 BP, we would probably find lots of languages related to both
Nostratic and Sumerian, and perhaps be able to reconstruct enough of
their parents to find that they were ultimately related.

So if someone should prove that Etruscan, Sumerian, Luwian, even
Basque are related to Nostratic, it would increase the time depth of
the family and be an impressive scientific achievement, but it would
not really bring us closer to an ultimate ancestor --- all the
languages are still from the same small area.

Lars Mathiesen (U of Copenhagen CS Dep) <thorinn@...> (Humour NOT marked)