On 2008-11-09 23:20, tgpedersen wrote:
> That claim is impossible to fulfill. The corpus of Venetic
> inscriptions basically contains a few verbs of sacrificing and
> donating, as you very well know.
How convenient. No evidence, nothing to constrain you.
> Instead I have another criterion I'd
> like them to fulfill: present stem in -a- in both Germanic and Latin.
And this argumentum ad ignorantiam makes them Venetic of all things,
just like that? The root *skabH- has an /a/ also in Greek (aor.
éskapsa), and a long /a:/ in Lat. perf. sca:bi. The ablaut pattern
*wag^-/*wa:g^- 'break' (including perf. *we-wa:g^-e) is attested in
Greek and Tocharian. See Jasanoff (2003) for a detailed discussion.
There are some staunch a-deniers, but many mainstream IE-ists now admit
that *a/*a: is a legitimate PIE alternation beside *e/*o and *e/*e:.
There is accordingly nothing "foreign" about *skábH-e/o-/*(ske-)ska:bH-
> *skaBa-/*sko:B-.
> Very simple. That was three verbs.
Tres faciunt collegium.
> So it's a class of composite origin?
Possibly. Convergence of at least two originally different types (plus,
possibly, a few *Cah2C- or *Coh3C- roots), establishing a pattern to
which o-grade presents like *molh2-e/o- or *h2woks-e/o- were
assimilated. If you think your "Brugmannian lengthening" explanation
could work without invoking analogy, you're wrong. There could be no
original Brugmannian lengthening in such roots as *h2weks- (= *hweg-s-),
*perh3-, *melh2- or *h2anh1-, with a final consonant cluster.
> Why not 'just-so'?
Because it's just just-so. Not even crazy enough to be interesting.
Piotr