Re: Indo-European family tree (was Re: Celtic and Afro-Asiatic?)

Hallo!

Thomas Wier wrote:

> Joerg wrote:
> > > > Third, the age they assign to Proto-Indo-European is impossible.
> > > > Any archaeologist will tell you that the wheel wasn't invented
> > > > yet 8000 years ago.  Yet, a PIE word for `wheel' is reconstructed
> > > > with as much certainty as is possible in this discipline.  And also
> > > > words for `yoke', `wagon', `carry by wagon', etc.  This means that
> > > > Proto-Indo-European can hardly be older than 6000 years.
>
> I have always been skeptical of this argument.  First, it makes the
> assumption that to have a notion of "wheel" you have to have the
> technology to produce the kinds of wheels that came into existence
> approximately ca 5-6k years BP.  That this is clearly questionable is
> attested by the fact that in Mesoamerica, where as of 1517 no culture
> had developed large wheels used for transportation whose linguistic
> reconstruction is *kwekwlos, but they did have wheeled toys.  Moreover
> you don't even need wheel-technology to have a notion of "circle".
> It is e.g. not altogether clear to me whether the common IE metaphor "wheel of
> heaven" = "sun" comes from analogizing wheels to the sun, or the sun
> to wheels.   Thus the terminus ante quem suggested is not so certain
> at all, and may indeed be spurious.
And among closely related languages, words can spread in a fashion
that the resulting forms resemble cognates inherited from the common
ancestor.  The most famous example are the Algonquian `fire-water'
words; another example is French _ordinateur_ : Spanish _ordenador_,
both meaning `computer' and having exactly the forms expected from
reflexes of a Latin word *ordinator which of course didn't exist.
It is possible that the words from which PIE *kWekWlos is reconstructed
have a similar history.  And then, the oldest wheels unearthed by
archaeologists need not be the oldest wheels ever made.

> This does not suggest that people like Colin Renfrew are right, only that
> the people using this argument against him do not have anything
> like a clinching argument that they think they have.
You are of course right.

> > The idea behind the Anatolian origin hypothesis is that PIE was the
> > language of the first Neolithic farmers of central and eastern Europe,
> > who are archaeologically known to have spread across the area between
> > 5500 and 5000 BC.  The hypothesis further assumes that they hailed
> > from Anatolia - not because the Anatolian Neolithic is culturally
> > particularly close to the central European (it isn't), but because
> > it seemed the only place they could have come from.
>
> Yes and no. Agriculture in its most ancient form supposedly
> developed first in the Konya plain in Anatolia, not in Mesopotamia.
> And I think you're discounting the ridiculousness of the alternative
> hypothesis based on Voelkerwanderungen.  People like Maria Gimbutas
> had a very naive notion about the relationship between migrations
> and the kind of language contact that would arise,
True...

>       since she thought
> migrations automatically replaced language and other social customs
> into addition to the obvious changes of material culture.  There
> are many obvious counterexamples (the Norman conquest, Mugal India,
> Manchu China) in which invaders become assimilated into the indigenous
> population, rather than do the assimilating.
On the other hand, we have cases such as the Roman Empire, the
Indo-Aryans and the Anglo-Saxons.

>       (I wrote a whole thesis
> on this subject some years ago, so I can send you the document if
> you wish.)  The basic appeal of Renfrew's theory is that language
> replacement becomes for him a function of population pressures:
> the farmers are able to outproduce the hunter-gatherers in food
> resources, and as a result are able to expand their population
> base much more quickly.  And the more quickly the population base
> expands in a given community, the more likely some members of that
> community will leave, taking all their social and technological
> capacities with them, thus founding new, related communities and
> speaking, presumably, the same or a related language.
Yes.  It is indeed very likely that the first neolithic farmers
of Europe took their language with them and ousted most of the
pre-IE languages, just as English ousted most of the indigenous
languages from North America.  Yet, a language spread associated
with the Battle Axe culture or whatever is not impossible.

> > Now, with the
> > discovery of the Black Sea flood, which happened between 5600 and
> > 5500 BC, there is an alternative to Anatolia, namely what are now
> > the shallow northern extensions of the Black Sea.
>
> This is a genuine alternative, but it doesn't really help identify
> where a given language community was located.
True.  We don't *know* that the Central European Neolithic originated
from where is now the Bay of Odessa - it's just one of several
possibilities, and we know *nothing* yet about the culture of that
area now submerged beneath the Black Sea.

> > But the problem remains that those first Neolithic farmers of Europe
> > did not use wheeled vehicles (at least, not a shred of evidence
> > has been found for them), and probably also had not domesticated
> > the horse.
>
> These objections do not have the weight that you think they do.
> (See above.)
True.

>       Concerning the horse, John Colarusso wrote an interesting
> article a few years ago in a festschrift for my professor Howard Aronson
> saying that the IE word for "horse" and IIRC "dog" were eerily similar
> to analogs in NW-Caucasian languages, where all seem similar to some
> root for "quick".
I have no problem with NWC loanwords in PIE.

> > There is ample evidence for both wheeled vehicles and
> > domesticated horses at least in Common IE (by which I mean IE after
> > Anatolian and Tocharian split off).
>
> Not highly relevant (see above).
>
> > Another problem with the Anatolian origin hypothesis is that the
> > migration pathways become rather absurd.  There is lots of movement
> > around the Black and Caspian Seas required, and it is not very
> > likely that Anatolian simply stayed at home.
>
> This is circular reasoning:  whether they "stayed at home" is precisely
> the question at hand.  And you have said yourself, and this is true,
> that the Black Sea was much smaller then.  The smaller size of the
> sea would have made movement around it much easier than is currently
> possible.
And then they could have moved by boat.

> > The Hittites and
> > their relatives evidently were newcomers on the Anatolian scene,
> > having displaced or assimilated speakers of Hattic, Hurrian and
> > other patently unrelated languages.
>
> This is a genuine objection:  if Renfrew is right, we need to
> explain why there are all these other languages in Anatolia in
> historical times.  I will say, though, that I think at least
> some of these ethnoi were not autochthonous: the NE-Caucasian
> languages are believed to have originally been spoken *south*
> of the Caucasus,
Yes.

>       somewhere in modern Iran, not north, and this
> interestingly coheres with Diakonoff and Starostin's thesis that
> Hurrian-Urartian are related to the NE-Caucasian languages.  I must
> stress that this link is by no means widely accepted, especially
> among Caucasologists, but if it's true, it might suggest that
> Hurrian and Urartian are not urfolk.  The same might well
> be said about the Hattic-speakers; they seem, afterall, to be
> related to NW-Caucasian languages outside Anatolia.  Maybe they
> migrated into Anatolia too? Who knows, really, but the whole
> question is so unclear and backed up by so little empirical
> evidence that one cannot make any certain judgements about
> whether these other groups are original or not.
Very true.  All the languages of ancient Anatolia seem to have
relatives outside Anatolia.  The question is who moved into Anatolia
or who moved out of it.  We don't know.

> > Next (though this is not a particularly strong argument), IE and
> > Uralic look uncannily similar for "unrelated" languages.  If PIE
> > was spoken north of the Black Sea, it becomes easier to account
> > for these similarities by assuming either a distant genealogical
> > relationship or prolonged contact.
>
> These similarities I think have been vastly overstated;  besides
> a few potential pronominal forms that look like cognates, there
> are a mere handful of lexical items that look related.
True; and many of them could be early borrowings.

>       This is
> less than 100 or so proposed cognates linking Hurrian to NE-Caucasian,
> and even that link, as I said, is widely doubted (by e.g. Johanna Nichols).
> It is also the case that no less an authority than Eric Hamp, whom
> I see whenever he's not spending his time in southern France, Wales
> or Albania (he's a Balkanist, afterall), has said that there are
> aspects of Kartvelian that also look similar and that maybe IE
> may be most closely related to that family.
Yes; the similarity between IE and Kartvelian didn't escape me.
Actually, they are about as great as those between IE and Uralic,
and I am undecided whether Uralic or Kartvelian is the next-closest
kin of IE.

>        And note that Kartvelian
> is, and probably always has been, spoken *south* of the Caucasus,
> not far from the Urartian monuments in Armenia.  (There are some
> thousands of Georgian speakers north of the Caucasus now, but
> they moved there in the second half of the 20th century.)
Yes.

>       So
> these putative Uralic similarities are no more convincing to me
> than any other branch of Nostratic/Eurastiac/(insert long range
> comparison grouping here).  Why not say that IE was spoken in
> far eastern Siberia? Afterall, Eskimo-Aleut are held by some to
> be distantly related to PIE.
According to Fortescue (and his arguments look good to me), Uralic,
Yukaghir, Chukotko-Kamchatkan and Eskimo-Aleut form a stock he named
"Uralo-Siberian".  He gives sound correspondences and 95 cognate
sets.  Proto-Uralo-Siberian would have been spoken in northern
Central Asia according to him - which is quite a long way from
the Black Sea.

>       I am being facetious, of course, but
> the point of my sarcasm is to introduce a sufficient amount of
> skepticism into this debate.  You can say that PIE was spoken
> around the Black Sea, but anyone who posits this or any other
> homeland necessarily has some castles in the air to build, because
> the amount of rock holding the castle up is currently very limited.
>
> (What's my pet theory?  There were at least *two* waves of
> Indo-Europeanization;  one following the agricultural expansion
> into Europe, and another involving Voelkerwanderungen out of the
> steppes. The Urheimat under this conception would have been in
> parts now inundated in the Black Sea, but close to Anatolia,
> perhaps in modern Bulgaria.  But this is pure speculation, based
> on little more than a geographic center of gravity, and I don't
> take this pet theory too seriously.)
Actually, my pet theory is essentially the same, though I place
the Urheimat of "Macro-IE" in what is now the Bay of Odessa.
IE proper resulted from the second wave while the first-wave
"Macro-IE" languages disappeared from the scene, though some may
still have been spoken in Roman times.  Old Albic is meant to
represent such a "Macro-IE" language.

Greetings,

Jörg.

Reply