>Following the Greek-Armenian-Slavic connection, one is tempted ...
How realistic is the idea of a Greek-Armenian-Slavic connection without I-I?
(a) Greek-Armenian-I-I connections are well established. This suggests
Slavic does not have such strong links to Greek as I-I does.
And
(b) Greek shows no sign of satemisation, although Armenian, I-I and Slavic
do. This suggests Greek separates off while I-I and Slavic are still in
contact.
Both of these would prima facie make a sprach-bund of Greek-Armenain-Slavic
without I-I rather unlikely, wouldn't they?
Peter