> Is there any sense to the notion that Hittite, Tocharian and perhaps
> Greek were always Centum, while West European IE (and perhaps Greek)
> desatemised? Recall that Albanian preserves the three-way split of
> 'palatal', 'plain velar' and labiovelar, so a Proto-Satem was
> potentially desatemisable.
Danish has depalatalized, with the exception of Vendsyssel and
Bornholm, while Swedish and Norwegian moved on in that direction.
Every conversation between speakers of those two groups takes place
across a kentum/satem dialect line, so to speak, and a least in the
case of the pair of Standard Danish and the 'big city' variety,
Rønne-fint, of Bornholmsk (Danish kør& "drive", Rønne-fint c^øra),
understanding isn't disturbed. That kentum/satem divide is much
overrated. 'Kentumisation' and satemisation could have occurred
everywhere there was a pair of competing IE dialects.
Torsten