>
>
> > Yup, tough customers, those Celts. BTW, where did the pattern
spread
> from, and what was the pattern of stress before?
>
> It was originally (perhaps until about 1200) free, as in Russian
(and as in the conservative northern dialects of Kashubian), then
initial as still in Czech (and as in southern Kashubian plus some
dialects of southern Poland). The evolution from initial to penult
stress was along a rather familiar path: initial stress is often
accompanied by a rhythmically alternating pattern of secondary
stresses like this: Swsw, Swwsw, Swswsw (where s = strong, w = weak,
S = primary), and since one of the functions of stress is to mark
edges of words or phrases (e.g. clitic groups), the last secondary
stress tends to be stronger than others. Gradually, it came to be
perceived as primary in Polish, so that the above patterns became
swSw, swwSw, swswSw (this is what we have in standard Modern Polish).
No shift was needed in disyllabic words, of course, and in
trisyllabic ones the conflict between Sww and wSw was resolved in
favour of the latter.
>
> Piotr
There are a few similar examples in Danish: fórudsætning >
foruds´ætning "precondition" (especially helpful in definite pl.:
foruds´ætningerne), the anecdote of that pronunciation is that it
came from the speech of politicians from Jutland. Did penultimate
stress originate in any particular part of Poland?
Torsten