Re: The palatal sham :) (Re: [tied] Re: Albanian (1))

> > The root *mad- (thus LIV) of Skt. madati 'intoxicate', Lat.
> > madeo 'be wet' forms Skt. pf. mamá:da. The root *bhag- (thus LIV)
> of
> > Skt. bhájati, Gk. aor. éphagon forms Skt. pf. babhá:ga. Both forms
> > have Skt. /-a:-/ from *-o-.
> >
> > The root *kan- 'sing' of Lat. cano, OIr. canaid 'sing' forms Toch.B
> > keme 'melody' from *konmo-s.
> >
>
> Yes, but the question was: examples of ablaut a/o/zero _within the
> same language_ (or better _within a paradigm of a single root in one
> language_); all examples of the same root having (reconstructed
> PIE) /a/ in one language and (reconstructed PIE /o/) in another
> language, could be explained as the /a/-occurrence being borrowed
> from a language which didn't develop the original pre-PIE /a/ to /o/,
> and the /o/-occurrence being "normal" inheritance within the language
> in question.
> *kan- is one of the roots that Celtic, Italic, Germanic mysteriosly
> share, as Kuhn points out; now suppose the /a/ in those three IE
> dialects is due to the fact that it was borrowed from an IE
> substrate, eg. 'Old European', which is so full of /a/'s?

The evidence of comparative linguistics is, if anything, stronger when it
operates between languages than within a single branch. How could a form
a *kan- with reflex of o-vocalism in Tocharian be due to the "fact" that
the root is a borrowing from Old European? If its presence in Tocharian
is enough to secure its PIE age, why would a presumed *ken- (which must be
what you imagine) take a-vocalism on Old European soil if most other roots
do not do any such thing? And comparably with *mad- and *bhag- whose
presence in Indo-Iranian should give you cause for some concern.

Jens