Re: Cowboys on Horseback (Germanic)

From: Alexander Stolbov
Message: 214
Date: 1999-11-10

cybalist message #142cybalist: Odp: Cowboys on Horseback
[Alexander]
Maybe it is better to locate Germanic (with other Northern) in the Eastern block together with Balto-Slavic? Germanic with its *at the place of *k is not Satemic? And Iranian with the same *h is?
[Piotr]
No, whatever else Germanic is, it's NOT Satemic. And you are wrong: Iranian does NOT have *h from *k! In Iranian, the Proto-Aryan palatal turned into *s (the word satem illustrates that!). Iranian *h (or rather velar *x, which is more correct for early Iranian) developed from PIE *s. The Germanic development *k > *x (part of Grimm's Law) is a manner-of-articulation shift (like the accompanying changes *g > *k and *gh > *g), NOT an instance of palatalisation.
Now I see that I was wrong when comparing Germanic *h (*x) to Iranian *h (*x), sorry.
Still some doubts remain. It is not easy for me to give up thinking about Germanic and Balto-Slavic as close relatives (neither linguistic morphological nor archeological evidences allow this).
You wrote on Nov 6 :
 
<<The change *s > *x > *h (in Greek and Iranian, but also e.g. in Brythonic Celtic) is again a rather common process.>>
 
If so, could the following process take place (theoretically and practically)?
 
                    PIE             GBS            Germanic
 
                    *k            *s            *x (*h)
                    *g            *z            *k
                    *gh          *zh          *g
 
(GBS = Germano-Balto-Slavic)
[Piotr]
One essential feature of the Satemic languages is the loss of the labial element in the *kw, *gw, *ghw series. Since Germanic preserves the full labiovelar articulation, it simply cannot be Satemic. As in biology, phonological evolution is irreversible: once obliterated, a contrast cannot be reborn.
Agree, it can't (to say more exactly, the probability of such an event is too small to be discussed seriously). However, the obliteration can occur parallel, convergently (like *s > *h in Greek and Iranian). We can demonstrate it. The transition from Mycenian *kw to Classical Greek *p happend independently of the delabialisation in Aryan and in Balto-Slavic.
This argument works (and completely destroys all my logic constructions) only if it's proved that *kw > *k happened before the partition of Aryan and Balto-Slavic. Do such proofs exist?
 
Alexander Stolbov
http://siem.newmail.ru