Sorry, I gave the wrong example, but it is
also connected with the RUKI rule, or more particularly with the divergent
development of *-k^s- in Baltic (*-s^-) and Slavic (*-s-). Compare Lith. as^is
and OCSl. osI < *ak^s-i-s 'axle'. The PIE numeral "6" was *swek^s- (with
variants), so the RUKI environment is there!
Interestingly, Indic and Iranian also treat
this cluster differently (Skt. -ks.- versus Iranian -s^-). Apparently, the
cluster *Cs^ (where *C stands for the "Proto-Satem" development of *k^ and *s^
for the RUKI sibilant) was affected by branch-specific assimilatory
changes.
Piotr
----- Original Message -----
Sent: Sunday, July 29, 2001 9:03 PM
Subject: [tied] Re: Baltic *gud3-
> Slavic RUKI is blocked preconsonantally. there are numerous
examples like *pIrstU : Lith. pirs^tas 'finger' or *s^estI : Lith. s^es^tas
'six'.
>
> Piotr
>
Uh... Hm... Lith. s^e~s^tas
'sixth', s^es^i` 'six' are products of RUKI-rulish developments? But there are
no /r/,/u/,/k/,/i/ or their allophones in
there?
Sergei