tgpedersen wrote:
>>The -i- belongs to the suffix, not to the root, which loses its *w as
>>a result of cluster reduction: Gmc. *bwijo: > *bijo: (> OE be:o),
>
> etc.
> Which suffix is that? -ye/o-? Which is unstressed in fio:?
In Italic, the suffix is simply *-jé/ó-, the extremely common suffix of
present stems (the unextended root *bheuh2- had an aorist meaning). The
West Germanic form looks more like a reflex of (iterative?) *bHuh2-éje-,
but this may be so just because *-je- and *-eje- fell together in
Germanic as allomorphs governed by Siever's Law (*-j-/*-ij-). One may
add that, because of the existence of alternative roots for 'to be'
(*es-, *wes-) in Germanic, analogical restructuring is very common in
their paradigms. Forms like OHG bim (OE beom), bis(t) (not to mention
birum, birut) are not inherited but analogical.
>>(cf. *ph2w- > *fw- > f- in Goth. fo:n).
>
> *ph2w- > *fw- > f-, therefore *bhw- > *bw- > *b- ? Maybe it's a good
> thing you only wrote 'cf.'
What's wrong here? Think about it as a phonotactic filter: *w was
disallowed in syllable onsets after labial obstruents -- a rather common
kind of ban.
Piotr