Re: [tied] Re: Satem shift

From: Piotr Gasiorowski
Message: 8675
Date: 2001-08-22

----- Original Message -----
From: tgpedersen@...
Sent: Wednesday, August 22, 2001 1:08 PM
Subject: [tied] Re: Satem shift

[Piotr:] >> Well, we all have our pet ideas, but just to broaden your perspective as regards alternative explanations, you might want read a good overview of the discipline. I recommend Hans Heinrich Hock's excellent book "Principles of Historical Linguistics" (1986).

[Torsten:] > Thank you for your concern for my general upbringing. May I recommend
to you Stephen Oppenheimer's excellent book "Eden in the East" to broaden your perspective as regards alternativer explanations?
[Piotr:] Well, the difference between Hock and Oppenheimer is that the former offers a balanced, objective and comprehensive overview of lasting contributions to a field of science (and so is good for that sector of your general upbringing, especially if you're truly interested in liguistics), while the latter concentrates on a highly personal and, to use a mild term, controversial hypothesis. "Eden in the East" has been selling well, I'm sure, but I can see no indication of its having had a tremendous impact on linguists or archaeologists. Thank you, I've learnt enough about Oppenheimer's arguments second-hand and feel no desire to be enlightened by him more directly. You might say that it is unscientific to dismiss something without reading it, to which my only reply is that life is too short for anyone to read everything and we must be selective.

[Piotr:] >> Now, since you claim that some IE dialects generalised *K, while others generalised *K^ in originally alternating paradigms, you also have to assume a balance of strength between the two variants. But then full regularisation on either side of the satem/centum divide becomes unlikely.
[Torsten:] Not understood. Please elucidate.

[Piotr:] I simply mean that a paradigm that can be regularised in two ways normally contains two frequently occurring allomorphs (say, *dje(:)u- and *diw-), each being a potential basis for analogical levelling. A rare and functionally constrained allomorph is not the stuff that new paradigms grow out from. If not retained in the original function, it can be either lexicalised (and so expelled from the original paradigm) or lost forever.
Now if a language has many "balanced" alternations, they are not likely to be regularised in one manner in daughter language A and in the other possible manner in daughter language B with much consistency. Why? Because analogy affects individual words rather than word classes and even originally similar patterns of inflection may develop along divergent paths in the same language. The Satem innovation just isn't messy enough to have resulted from such a process. The simplest theory about it, and one that is generally accepted by IEists, is that it was a purely phonetic process -- a context-free fronting of a set of stops. I know of no data suggesting that Satemisation began in pre-front-vowel environments or in any other specific context. If you have any such evidence, I'd like to see it.

Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.