On Mon, 13 Jan 2003 19:58:41 +0100, Piotr Gasiorowski
<
piotr.gasiorowski@...> wrote:
>> ...Slavic *-go is simply the ablative ending *-od (*-ot) after a vowel.
>
>> As I said before, since the Lithuanian pronominal genitive is historically an ablative too (G. tõ, kõ, s^iõ, jõ), the origin of the Slavic pronominal genitive should in the first place be sought in the
>PIE ablative.
>
>> The uncontracted ablative form of the pronouns survived into proto-Slavic (*is -> *eod) > *jI -> *eo, (*k^is, *k^eod) > *sI -> *seo(d), and also *too(d), *k(W)oo(d), etc. To block the hiatus, -w- (-v-) or -h- (-g-) was inserted, as also happens in Slavic before /o/ in the Anlaut, thus giving jego, sego, togo, kogo (Russian /jIvó/, /s'Ivó/, /tavó/, /kavó/).
>
>It all sounds very convincing. Just two questions:
>
>(1) Do you mean that *e-od, *to-od etc. are forms inherited from PIE? The evidence for the ablative is limited, by my impression is that it originally employed the *-sm(o)- "extension" like the dative and the locative (*e-sm-ed, *to-sm-ed, or thematic *-smo:d). On the other hand, if you assumed that the shorter forms are (Balto-)Slavic innovations, perhaps thesecond question might be easier to answer.
We do not find forms without *-sm(o)- in the dative and locative, but
in the ablative and instrumental *-sm(o)- appears to have been
"optional". Besides abl. tásma:d, Skt. has abl. tá:d, used
adverbially. Latin has abl. eo:. The ins. lacks *-sm(o)- in Skt.
(te:-na), Slav. (te:-mI), Lith. (tuõ) and Gothic (þe:). If we may
equate Hittite -ed(an)- with extra-Anatolian *-sm(o)-, we see more or
less the same pattern: DL ke:dani (*k^e-edan-i), Abl. ke:z(za)
(*k^e-ot-i ?), Ins. ke:t (*k^e-et) or ke:dand(a) (*k^e-edan-t).
>(2) Why do we find plain contraction (*-o:(d)) rather than hiatus-filling (*-oho(d) ~ *-owo(d)) in the *o-stem nouns?
There are a few Avestan spellings showing no contraction in the o-stem
ablative (as.a:at~ca:, vira:at~ca:, Beekes Gatha-Avestan, p. 53). The
lack of contraction in Slavic may be due to the fact that the pronouns
have short monosyllabic stems. Perhaps more importantly, a part of
them had uncontracted -eo- (*eod, *k^eod) which would obviously have
been more resistant to contraction than the -oo- of the o-stems and
the other "thematic" pronouns (*to-od, *kWo-od, etc.). Maybe the
Proto-Slavic forms *to-o, *ko-o were analogical after *je-o, *se-o and
were used alonside (and eventually instead of) normally contracted
*to:, *ko: ( = Lith. tõ, kõ). If Slovenian (and Serbo-Croatian) jega,
toga etc. are ancient, they would represent analogical (cf. Latin eo:
for *eo, perhaps also Lith. jõ, s^iõ) or contaminated (*too ~ *to: >
*ta => to(g)a) by-forms of the pronominal ablative.
Amusingly, what I'm saying then is that such genitives as Rus. novogo
(/nóv&v&/), Pol. nowego, Slov. novega are to be derived from something
like *neuooeo(o) [= novajego ~ novajega].
=======================
Miguel Carrasquer Vidal
mcv@...