Slavonic imperfect; IE subjunctive

Jens Elmegaard Rasmussen jer at cphling.dk
Tue Apr 13 23:22:10 UTC 1999


On Sun, 11 Apr 1999, Miguel Carrasquer Vidal wrote:

[...]
>
> The doubly thematic subjunctive is only Greek and Indo-Iranian, I
> believe, and is in my opinion analogical.  [...]

You are right about analogical, but not about the geographic limitation. A
form such as *bhereet(i), the subj. to go with ind. *bhereti or inj.
*bheret must be of relatively late (if probably still pre-PIE) making,
for the general rule that the "thematic vowel" becomes /o/ before anything
voiced, including vowels, is not obeyed. If *deywe/o- + dative sg.
*-ey, dual *-e, and nom.pl. *-es becomes *deywo:y (better *deywo::y?),
*deywo: and *deywo:s, the o-timbre must have come from the thematic vowel,
wherefore one expects *bhere/o- + e/o + t(i) to yield **bhero:ti, but the
form is *bhereeti (yes, with disyllabic -ee- reflected by Gathic). The
formation with *-ee- is also found in Latin (fut. lege:s) and perhaps
Armenian, if Birgit Olsen's derivation of the morpheme -ich- from *-e:ty,
based on the use of the 3sg *-eeti as a new stem, is correct. The same
formation accounts for the fact that Albanian has no umlaut in the
subjunctive, while the prs. indicative shows fronting umlaut in the 2.3.sg
and 2pl, i.e. in the places that had a front thematic vowel in IE; since
e: goes to o in Albanian (no doubt via a:), it is no wonder that this
category as no umlaut. I would of course have liked to see the long-vowel
subjunctive in more languages, but I think we have enough to accept it as
PIE.

Jens



More information about the Indo-european mailing list