> 2. The retraction of stress-accent across the Germanic family to
the
> initial syllable (presumably, the unaccented prefixes were added
> later).
>
According to Piotr, Polish once had initial stress. So did Celtic and
Latin. That makes initial stress cover a huge area in North Central
Europe: Celtic, Germanic, Italic (when they hadn't left for Italy),
Polish, Czech, Hungarian. North of that, Finnish, Estonian and
Latvian. Latvian, at least in the north of the country, is spoken on
old Baltic-Finnic (Livonian) territory. Which makes that conservative
IE dialect, Lithuanian look like an intrusion. In other words, the
initial stress must come from a language that preceded IE.
Germanic obviously shifted to initial stress after Verner and Grimm,
but those two changes can be dated to sometime in the two centuries
(B)CE where Germanic expanded from the southeast into new
territories. These might then be where Germanic acquired initial
stress from a substrate (and that habit would then have to migrate
back to the place of Germanic origin).
The question is whether initial stress is original in Baltic Finnic
and Hungarian? I've seen some consonant-changing phenomena in Finnish
ascribed to a process similar to Verner, but how would that make
sense in a language with fixed initial stress?
Torsten