>>Your
> -ái > -ì seems to
> >be ad hoc -- why do we have sakai~ (with -ai~ showing the effect of
> >Saussure's law and having acute in part of the Z^emaitian
>> dialects (and
> >broken tone in the other)), not +sak-ì?
I'm sorry, I wrote in haste and didn't put it clear enough. I meant the
*verb* -- 2sg.praes. sakai~ 'you say' (inf. sakýti) < pre-Saussure *'sakái
(' indicates the place of the ictus) (cf. 3.praes. sa~ko with the originally
circumflexed ending). Standard Lithuanian reflexives (sakai~si) are
analogical and thus indecisive, since we find acute in Z^emaitian dialects
(both in non-reflexive and reflexive forms), and, actually, even in
Auks^taitian dialects (and Standard Lithuanian) the -ai~ of sakai~ 'you say'
differs from the "truly circumflexed" -ai~ of, say, sakai~ 'resin'
phonetically, the former being realized rather like [-ài]. So my point was:
why would we have -ì < *-ái in adjectives (gerì < *gerái), while we
definitely have -ai~ < *-ái in verbs? (the implicit answer being "because
there seems to be no rule "-ì < *-ái" in Lithuanian).
> It is curious that mobile paradigms end-stress the o-stem
> nom.pl. (sakai~). If I'm not mistaken, end-stress should
> belong in the sg. nom./voc./gen. [not the abl. > gen. of the
> o-stems] (what about the instrumental?), and in all the
> plural except nom. and acc.
I don't know.
> How did the pronouns get their -íe?
No idea. Does PIE *-ey (>*éi) seem absolutely impossible?
Sergei