i-verbs in Baltic and Slavic

From: Miguel Carrasquer
Message: 44648
Date: 2006-05-22

I was thinking a bit about the i-verbs ("causatives-
iteratives") in Balto-Slavic.

The IE reconstruction of the present personal forms is
pretty straightforward: we have a root (generally o-grade or
zero-grade) followed by *-éi- and thematic endings.

The expected Slavic outcome would have been:

*-éi-o: > *-IjoN
*-éi-esi > *-Ijes^I
*-éi-et(i) > *-Ijet(I)
*-éi-o-mes > *-IjemU (via *-éi-e-mos)
*-éi-e-te(s) > *-Ijete
*-éi-ont(i) > *-IjoNt(I)

What we in fact have is:

-joN
-i~s^I
-i~t(I)
-i~mU
-i~te
-eNt(I)

But in such a common verbal category I do not find irregular
contraction (*ijV > *i~) very surprising.

So if we have irregular contraction in Slavic, perhaps we
can have it in Lithuanian too. The paradigm:

-au
-ai
-o
-ome
-ote
*-aN (=> ptc.praes. nom. pl)

(similarly Latvian -u, -i, -a, -a:m, -a:t)

might be explained by irregular loss of -j-, _after_ the
generalization of /a/ as the thematic vowel in East Baltic.
So, leaving out the 1st and 2nd person sg., we have:

-ej-a-t > -ea > -a~
-ej-a-me: > -eame: > -a~me
-ej-a-te: > -eate: > -a~te
-ej-a-nt > -eant > -a~nt

I have the vague feeling I've seen this explanation
somewhere else. Is that correct?

That leaves the infinitive, where, despite the wildly
different (i~ vs. a~) development in the present forms,
Slavic and Baltic show remarkable agreement down to the
(acute) intonation: Slavic -i"ti, Lith. -ýti. Where can
this acute come from? There are a few possibilities in
Slavic, but the Lithuanian form can, or so I think, only be
explained as *-iH- (certainly not *-ei- or *-eiH-).

This takes me back to an old idea of mine, that the
causative-iterative is an old compund of verbal root
(showing o-grade with "Rasmussen infix", which must have
some meaning) plus the verb "to make" [ = Hittite iyami,
iyezzi etc.] This would also explain the intercalation of
the preverb pV- [Hitt. piyami, piyezzi] in Vedic
causatives-iteratives such as dha:-páya "cause to put",
jña-páya "cause to know", etc.

The verbal root in question is given in LIV as *Hyeh1-
"werfen". I see no unsurmountable problems in assuming the
verb was grammaticalized as a causative marker, and in the
process metathesized to *(H)eih1-e- (and besides, the
Hittite form can easily be from original *Heih1-e-,
postponing the metathesis of the independent verb to a later
stage ancestral to Greek ie:mi and Latin iacio:). The first
laryngeal is lost in the composition, the second one is
taken care of by the thematic vowel and the syllable break,
so: *R-weid Héih1-e-ti > *woidéyeti "he makes see".

The second laryngeal does surface in the Balto-Slavic
infinitive, which is athematic and stressed on the ending
*-té(:)i, leaving the causative marker in zero grade
*-(H)ih1-. That would explain the Balto-Slavic acute.

=======================
Miguel Carrasquer Vidal
mcv@...