Hi Torsten,
there's no need for apologies, the way this forum is structured, I keep losing the thread ('verliere den Faden') myself constantly. All the information one gets via the email updates is that *somewhere* within the 100-something replies in a thread, someone has added something... I think there's maybe room for improvement.
Unfortunately, our university library does not seem to have a copy of your book in stock, I will check whether it is available via remote order. I will have to step up my game if I want to have a meaningful discussion on the subject...
I think I've heard someone flaunting the idea that in the path integral, one should somehow integrate over all geometries (though how exactly that works, I'm not sure), and R^4 then may be singled out as giving the dominant contribution due to the plethora of smoothness structures---but I'm unsure as to how viable this is, or whether I am remembering correctly.
Anyway, thanks for your reply,
Jochen