Miguel:
> The preterite, of course, is not built on the present
> participle.
Of course, you'd probably say it's built on a "preterite participle".
Whatever. My point is that endings like *-pa and *-ya aren't "participle"
endings to begin with in Uralic. They are modal endings attached to
verb stems that can mark nouns in the nude, or mark verbs with pronominal
endings attached. That doesn't mean that the 3ps derives from a noun
or a participle.
At best, the action noun and an endingless 3ps, both with seperate
origins, had merged. I see little distinguishing an action noun
from the 3ps aorist in Mid IE except for *-a. So with that possibility
unaccounted for, there is no guarantee that your view is the right one.
The verb could just as well have always been a verb without it ever being
a participle or noun or what have you.
In fact, if Proto-Boreal had developped an alergy against words ending
in most consonants except nasals and *-t and fought against it by adding
a vowel as I've proposed long ago to explain Uralic's syllable structure,
then naturally action nouns formed with "participle" endings together
with 3ps verbs that previously had ended in *-a would merge and be
identical. This is probably why the 3ps developped *-sa out of a
pre-existing demonstrative to avoid confusion.
> "Having become" endingless? I see no reason whatsoever to
> think there ever was an ending here.
There was *-sa in Uralic.
= gLeN