> >> The usual pattern for this kind of verbs is short voweledYes, I agree. It seems to be related to the collective (later
> >> i-verb (e.g. skoc^iti, skoc^joN, skoc^itI) vs. long-voweled
> >> a/je-verb (skakati, skac^joN, skac^jetI).
> >So the *-j- comes from the verb stem?
> >When do you think the *-a- ending was added?
> The use of *-ah2- as an iterative ending is at least PIE.
> The lengthening of the root vowel is Balto-Slavic (PIE /o/Right. IIRC, /o(:)/ merged with /a(:)/ in Balto-Slavic, with /a/
> (> /a/) is lengthened to /a:/, not /o:/).
> The accentuation (lengthened a: and e: are circumflex, but i: andThis is off-topic, but what caused acutes in mobile paradigms to
> u: are acute) suggests that it happened before the Slavic soundlaw
> *ei > i: (which creates a circumflex /i:/) and/or before Meillet's
> law (which turns acutes in mobile paradigms into circumflexes).