Re: Various loose thoughts

> >
> >
> >Isn't accusative usually connected with movement towards

(allative),

> >and isn't *bhi, which has a locative sense, usually connected with
> >the locative (German bei + dative, < locative?); in which case it
> >would be -i + mi, -u + mi, which looks nice to those who think

there

> >existed locatives in both -i and -u (and I know you're not one of
> >them).
> >On the other hand, perhaps -m + bh- > -m-; I don't know whether

there

> >is a better explanation of the BSl-Germ-(and Ubian) -m- for -bh-

in

> >dat.loc. pl.?
>
> What's Ubian?

Several hundred fragments, found in the Rhineland in the once Ubian
territory, of votive inscriptions, therefore full of dat.pl.'s of
tribal names, vacillating between -ims and -abus (by memory!),
treated at length by Kuhn and Vennemann.

> Indeed the attractiveness of -m + bh- > -m(m)- is another
> aspect that I forgot to make explicit favouring an origin in
> acc. + *bhi. Since I don't believe that the adposition *bhi
> has anything whatsoever to do with the (athematic) plural
> oblique marker *-bhi-, it doesn't automatically follow that
> where Balto-Slavic has -m- in the one, it should have -m- in
> the other.
>
> The adposition *-bhi may have been combined in PIE with
> several cases, as so often. The Balto-Slavic forms of the
> ins.sg., however, exclude the locative (we would have gotten
> *-e:i-bhi, *-o:u-bhi, or, if -bhi came in lieu of -i, that
> would have given *-ei-bhi, *-ou-bhi, same result).
>
> The Greek forms with instrumental meaning (biê-phi "with
> force", etc.) seem to be based on the pure stem, and so do
> indeed the pronominal forms in Slavic, which I had
> overlooked, têmI, imI, simI, cêmI etc., which are derived
> from the "absolutives" *toi-, *ei-, *k^ei-, *kWoi- (cf. Skt.
> ay-am (*ei), nominative sg., and a reflex of *kWoi or *kWei
> as nominative in another language but I have forgotten
> which... See also the Vedic instrumentals te:-na, ke:-na) +
> *-mi. Perhaps the Balto-Slavic i- and u-stem ins. singulars
> can be explained in the same way (although the "absolutive"
> [i.e. vocative] of i- and u-stems ends in *-ei, *-eu, not
> *-i, *-u). For the later and exclusively Slavic o-stem
> ins.sg. in -omI, -UmI, both the accent (barytone), and the
> raising of *-a- to *-u- before *-m in N. Slavic make an
> origin in acc. + -mI the most likely solution.

-mI transferred analogically, you mean, not from *bhi? But if so,
where's the need for the -m of the accusative?


I can't object to the phonology, but I still think acc. + *bhi is
jarring. If an instrumental should be based on a local case, it must
be one that 'stands still', cf. the use of 'with' etc. for comitative
and instrumental both; and comitative comes the closest to locative,
out of the local cases, semantically speaking.

Torsten