From: Piotr Gasiorowski
Message: 10603
Date: 2001-10-25
--- In cybalist@..., george knysh <gknysh@...> wrote:
****GK: I'm sorry if I offended you. I didn't mean to. I like what
you write and the way you write it even if I don't always agree. I've
learned a great deal from you Piotr, and hope to learn more.*****
It's all right; I'm relatively thick-skinned.
*****GK: As for Petrov (1894-1969). I confess that I don't know a
great deal about him as a linguist beyond the two books that I read
("Skify. Mova i Etnos." "Etnogeneza Slovjan".) They both came out in
1972, and quickly became a bibliographical rarity. I understand that
these items were published by Kyiv's "Naukova Dumka" just weeks
before the "big Shcherbytskyj crackdown" on "original" Ukrainian
creativity. Petrov seems to have become a linguist rather late
(sometime in the 1950's.) Petrov had other interests also. He was the
Ukrainian literary critic Domontovych. (Same fellow different names).
And he had quite a career as a double agent in his youth and middle
age.******
Perhaps he went to conferences with a false beard and a false
identity :)). Anyway, calling him one of the leading Soviet Indo-
Europeanists was an overstatement, wasn't it?
>>PG: Closer connections can be demonstrated using shared
innovations, but not shared archaisms. That's the ABC of historical
linguistics.<<
*****GK: I understand (?). So that the Slavic "Bog(h)" rather than
the D word in Baltic akin to "Deus" indicates that their ancestor
population had closer contacts to Iranians than the ancestor
population of the Balts. Unless I'm confusing borrowing and
innovation...*****
You are, in this case. One has to be careful with shared words of
culture, since they diffuse easily and may prove to be borrowings
rather than shared inheritance. The Slavic meaning of *bogU is a
semantic calque from Iranian, one of a number of such items that
indicate close contact. Shared phonological and grammatical
innovations prove that Baltic and Slavic are rather closely related
in the genetic sense.
*****GK: But then when the "shared archaisms" are exceedingly
numerous (and exclusive) that tells you something too doesn't it?
They could also have been in Thracian of course but there's only
about 180 extant words of that...(nearly 60% shared with Baltic,
10% with Iranic, 10% with Indo-Aryan).*****
No, shared archaisms prove _nothing_ about genetic subgroupings. This
may be counterintuitive, since we associate similarity with
relatedness, but it isn't the only area where our intuition leads to
illusions. "Phylogenetic linguistics" uses essentially the same
methods that are employed in evolutionary biology or studies of
manuscript filiation. Imagine three languages, A, B and C, with A and
B descending from a more recent common ancestor (let's call it Proto-
AB) than the protolanguage of the whole group (Proto-ABC). Let's
imagine that C has changed rather slowly since its split from the AB
branch, and that B is also rather conservative, but A has for some
reasons undergone a number of innovative changes since its separation
from B. As a result, A and B are more closely related in the genetic
sense (they derive from Proto-AB, which is not ancestral to C), but B
and C may appear far more similar than A and B. This happens more
often than you might think. Note that in such a case what B and C
have in common (excluding loans and parallel developments) is shared
archaisms. What A and B have in common to the exclusion of C, on the
other hand, is a unique sequence of innovations that occured during
the history of proto-AB between the primary split that gave rise to
it and the next split that separated A and B. There will be no
similar sequence for the pairs {A, C} and {B, C}.
*****GK: And here ancient hydronyms can be very useful indeed. The
basin of the Sejm r. is a well-known ancient contact zone between
Baltic and Iranic. But given your loan words statement perhaps one
should say "BaltoSlavicScythian" and Iranic. The "loan words"
indubitably acquired there didn't make it into the hinterland. The
fact is though that linguists have reconstituted these hydronyms
thanks to Baltic terms. There was territorial contiguity between non-
Royal "Scythians" and the areas known to be then occupied by
populations which evolved into Balts, and the areas which were then
Thracian (if one includes Geto-Dacian) and the areas where Iranic
could be postulated (still, even if the "Royals" were
acculturating).******
There is a well-known study by Toporov and Trubachev (1962) on the
hydronymy of the Dnieper basin, and several later articles by
Trubachev, where he claims, quite convincingly, that several Iranian
words did make it into Baltic independently of Slavic. The contact
zone you mention looks real enough, but the scale and intensity of
contacts was evidently more limited than in the case of Iranian and
Slavic.
Piotr