Minor complications (such as palatalised
variants) apart, we have the following:
PIE
*k/*kW *k^ *s
Sanskrit
k s'
s (s. in the "ruki" environment)
Avestan
k
s h (s before -n- or next
to a stop, s^ by "ruki")
Lithuanian
k
s^ s (s^ in the "ruki"
environment)
Albanian
k
th gj-,-h- (and other odd
reflexes)
Armenian
kh-,-y- s h,-kh (s
before stops)
OCSl
k
s s (x or s^ in
the "ruki" environment)
As you can see, in most branches there is a
clear difference between an "original *k", "original *k^" and "original *s".
Even in Slavic, where PIE *k^ and *s usually merged as Slavic *s, the
merger did not take place in the "ruki" environment (after *i, *u, *r or a
dorsal stop), where PIE *s changes to Slavic *x before back vowels and *s^
before front vowels, while PIE *k^ is reflected as Slavic *s.
Piotr
----- Original Message -----
Sent: Friday, August 03, 2001 1:09 PM
Subject: [tied] Re: Satem shift
In what way does a satem branch language show an "original k"
(outside of the etymological dictionaries, of course) and how would I tell the
difference from an "original *s"?