Re: [tied] The full list of kentum words in Slavic by Goł

From: Rick McCallister
Message: 68521
Date: 2012-02-09

Torsten and anyone else who wishes to jump in:

You've answered some of my question. I'm aware of the palatal vs. velar dichotomy in IE. 
BUT The break must have happened either during or after Germanic and Greek broke off, right? 
c. 2000 BCE? 
And it likely spread from Indo-Iranian into neighboring Balto-Slavic, right?
Or am I too simplistic regarding this? 
After all, we do have palatalized vs. velarized /k/ in US English: NY <coffee> /kwaafiy/ vs. Southern California <California> /kyael@...@/
It is interesting the play between historic splits vs. Sprachbund in Scandinavian languages and in Ibero-Romance, how deep the splits really are but get papered over by proximity. 
If I were to speak Portuguese with João, I'm sure he'd understand everything I say but wonder what hillbilly taught me Portuguese. I imagine I'd say the same thing about his Spanish. 
Do "hillbillisms" work like that in Scandinavian languages, where a "good" Danish word is a "hillbilly" Swedish word, etc.?

From: Torsten <tgpedersen@...>
To: cybalist@yahoogroups.com
Sent: Thursday, February 9, 2012 11:48 AM
Subject: Re: [tied] The full list of kentum words in Slavic by Gołąb

 


--- In cybalist@yahoogroups.com, Rick McCallister <gabaroo6958@...> wrote:
>
> Torsten et al:
 
> I wonder about the idea of satemization at the earliest levels of
> Balto-Slavic --because it would, of necessity, crop up in Germanic
> --wouldn't it?

with loans - yes, otherwise no.

> You may wish to think about satemization more in terms of time than
> genetic relationships (branches). Given that Germanic (centum) is
> most closely related to Balto-Slavic and Greek (centum) is generally
> seen as part of a Graeco-Armeno-Indo-Iranian group, then perhaps
> satemization occurred sometime around 2,000 BCE, give or take 500
> years or so, shortly after Germanic moved away from Balto-Slavic and
> Greek from its congeners. One could perhaps see satemization as an
> adstrate in Balto-Slavic from Indo-Iranian, hence the incompleteness
> of the process. Now, it could well be that there was a later
> re-centumization from Germanic, Venetic, et al., but I don't buy the
> idea of a Balto-Slavic-Indo-Iranian node.
> Please correct me and explain if you disagree.

This is what I think happened:
PIE velar stops had two allophones, depending on some feature of the following phoneme (most likely some front/back or palatalization feature). The two allophones of the velar stops corresponded to (some prestage of) their later correspondents in the satem and centum languages, respectively. At some time the central, satem languages generalized the former allophones in their paradigms, whereupon the peripheral languages in protest(?) went to the other extreme, generalizing the latter allophones (this implies that plain velars did not exist, so they must be later loans).

Does that answer your question?

PIE
*k'/*k *g'/*g *gh'/*gh
*k/*kW *g/*gW *gh/*gWh

->

satem
*k' *g' *gh'
*k *g *gh

and

kentum
*k *g *gh
*kW *gW *gWh

cf.
Swedish
*hW- -> *v- vit, va(d), varför
Norwegian
*hW -> *kv-,*k- kvit, ka, koffor

Torsten