> Does this mean any post-tonic /e/ is impossible ?
> Arnaud
It means that we have e:/o: alternation depending on the location of
accent. For example, the agent suffix *-té:r has the posttonic allomorph
*-to:r. Similarly, accented *-mé:n alternates with posttonic *-mo:n.
=========
Sometimes you have stranges idea about phonology.
I'm talking about short /e/.
not /e:/ versus /o:/
Arnaud
=========
> *kuH2-on is a derivative of *kuH2.8 "to bite".
> Some kind of a participle.
Any evidence for *k^uh2- 'bite'? And why is the weak stem of the 'dog'
word *k^un- rather than *k^uh2n-?
Piotr
=======
I guess k_H2 is better to account for LAtin ca-nis
than kun.
Arnaud
===============