The way I've read it, the Germanic
weak verb was formed by adding a post-positional verbal auxiliary after
the bare verb, which in time, fully cliticized. Apparently, this auxiliary
verb is the ancestor of English 'do', descended from PIE dhē.
(AHD3-Pokorny), dheh1 (EIEC), 'to
put', 'to place' (this root is fascinating in its own right).
A post-positional
verbal auxiliary? This seems strange to me, but then, I am unaware of all the
details on how the verbs in other branches of the IE family, past and present,
can behave.
I also wonder if there is something
of the ancient Germanic substratum at work -- not so much in the choice of the
verbal auxiliary, but in how it was used. I've always thought that the
Germanic innovated vis-a-vis weak verbs as a response to the constraints on new
verb formation imposed by the ablaut system. How can you turn a noun
into a verb without an easy way to immediately inflect it? The innovation is
eminently logical; I wonder why Germanic's sisters didn't think of it
themselves.
Mark.