> Indo-Iranian ava-, Slavic ovU, are generally derived from
>
the root *h2au- ~ *h2u-, which otherwise gives words and
> particles
largely denoting binary opposition (and..and,
> or..or, also, etc.: Ved. u
"also", va: "or", Av. uta "also",
> Grk.
aû "on the other hand", aûte "again", Lat. aut
"or",
> -ve "or", German auch "also",
etc.)
The connection of the Slavic demonstrative and these particles is very
possible, but inside Slavic o-vU is surely made of two pronominal stems (cf.
o-nU), and here the -vU expresses the 1st person deixis. Cf. Maced. kniga-va
"the book (near me)" vs. kniga-na "the book (near you)";
also Serb. evo meni, eto tebi, eno njemu "this is for me, that is for you,
that is for him".
> I'm familiar with the Celtic forms Bomhard
mentions under
> roots 555 (*ma, *mu demonstrative) and 524 (*mi,
*ma
> interrogative, relative). Not very convincing.
Yes, the attestation of *mo in IE is very week, and some problems exist
here (e.g. front vowels in Anatolian and Singhalese), but the common system
leads me to the conclusions I have made.
> This is the same variation
as in the Altaic 1st. person
> pronoun *bi, *min-, pl. [Mong-Tung]
*bu/*ba, *mun-/*man-.
I suppose that the very primary system
(proto-proto-... - probably existing even before Nostratic) did not distinguish
personal and demonstrative pronouns at all. It is not easy to find out what
meaning was expressed by vocalic variations, but shortly - front wowels seem to
express something near and restricted (for example, "me", "you
here", "he/she/it near us", "our group that is
present", "we" inclusive, "this", "here" etc.
). Like this, back vowels expressed something far, unrestricted or large
("we inclueding those who are absent", "we" exclusive,
"that" and so on). Firstly the forms with front vowels were
separated as personals, espesially in singular, but even they somtimes
(especially in Uralic, Mongolic, Tungus-Manchu) are used in demonstrative
meanings. Having no primarily expression of number in personal pronouns,
different languages formed it each in its own way. Sometimes, as in Mongolic and
Tungus, the stems with back vowels mostly were used (and so I think that these
Mong-Tung. forms are originally connected with Turkic demonstratives). But the
plurals could be made by adding a separate affix too, as we see in Turk. bi-z =
Mong. bi-de (? = Hitt., Goth. we-s, Lith. me-s). That's likely that in IE the
both ways were used as well. Thus, along with the agglutinative 1Pl. forms cited
above, we find Lat. nos, vos (with unexpected initial consonants, but surely
back vowels).
> The usual development *mw- > *b(w)- is reversed or
blocked
> by a following /n/.
Only in Terkic. In all other Altaic (and, further, Nostratic) groups
the suppletive w/bV : mV cannot be explained phonetically.
> The PA
accusative ending *-ba [I incorrectly gave that as
> *-b earlier] comes
from the pronoun *bu ~ *bwa, which in
> turn derives from the PN pronoun
*mu- [*ma-]. This is an
> interrogative in Afro-Asiatic, Uralic and
Kartvelian, and a
> demonstrative in Altaic and Kartvelian. I haven't
studied
> the particulars in enough detail to be sure, but my gut
>
feeling is that the interrogative usage is original and the
>
demonstrative secondary. In any case, it doesn't matter all
> that
much to the connection between the pronoun *mu-/*ma-
> and the accusative
ending *-m(w)a, which we can make in
> either case.
K.Ye.Maytinskaya ("Mestoimeniya v yazykax raznyx sistem",
Moskow, 1969) has investigated the typology of diachronical changes in
pronominal systems of different languages deeply, and her conclusion is that
there can be only a change of demonstratives into interrogatives ("Did you
see that?" > "What did you see?"). So the interrogative and
demonstrative pronouns with initial m-, if connected, must have a demonstrative
origin. And semantically, the development of a demonstrative into an accusative
marker seems to me much more probable than such a development of an
interrogative.