On 04-10-14 04:09, etherman23 wrote:
> *pibeti is an often discuss word because of the *b where a *p is
> expected. However that's not the point of this posting. This is a
> reduplicated form of *po:. According to the laryngeal theory this
> should be reconstructed as *peH3. The reduplicated form would then be
> *pipH3-eti. However this should result in **pipoti. Again, ignoring
> the issue of *b, why do we have an *e attested instead of *o?
Because morphological transparency took precedence over regular phonetic
processes and the value of the thematic vowel was analogically restored
(or the colouring was blocked in the first place). For a similar reason
English <starry> and <carry> have different vowels, though, given the
regular phonetic developments in English, they "should" rhyme: the free
form <star> has influenced the derivative. In PIE, the form of *-e-ti
was constant no matter if a laryngeal preceded it. You could say that
the cumulative influence of *bHer-e-ti, *gWm.-sk^e-ti, *mon-eje-ti,
etc., was sufficient to preserve the same vowel colour in *pi-ph3-e-ti.
Piotr