On Wed, 31 Oct 2001 09:47:19 -0000, "Sergejus Tarasovas"
<
S.Tarasovas@...> wrote:
>--- In cybalist@..., "Piotr Gasiorowski" <gpiotr@...> wrote:
>> I absolutely agree. It's neither possible nor desirable to explain
>_all_ Slavic word-initial *x's as Iranianisms, and most cases of *xr-
>must be variants of *skr- (whether "expressive" or due to a not-
>completely-successful sound change (resembling the things that
>happened to *sk- in Germanic). I only singled out *xromU for special
>treatment because it's hard to resist the spectacular comparison with
>Skt. sra:ma and Germanic lam-. Wouldn't it be lovely if Slavic
>*xromU 'lame' and *lomiti 'break' proved to be ultimately related via
>*(s)lom-?
>>
>
>It's also tempting to derive, eg, *xod- :*s^Id- 'walk,go' from *sed-
> 'sit' > 'move sitting in a vagon', and, if I recall, other (few)
>examples of possible *s- (not only *sk-) > *x- do exist.
The usual explanation is that <xodU>, <xoditi>, <s^IdU> have developed
from *sed- in combination with RUKI-generating prefixes such as u-,
pri- (u-xoditi "to (go) sit outside" -> "to go out"; pri-xoditi "to
(go) sit inside" -> "to come in", and then by analogy xoditi "to go").
However, one would expect <xaditi> (like <saditi> "to make sit, set" <
*sod-ei-, by Winter's law *so:d-ei-). The short vowel rather suggests
a causative *Sodh-ei-, from a root *Sedh-. Now, with a bit of special
pleading, there is a PIE root *sWe(i)-dh-, *sWe(i)-m- etc. "to move"
(in Germanic also "to swim"), which might fit (see IEW 1041
*swe(:)(i)- and 1046 *swem-). This presupposes that */sW/ gives
Slavic *x, as in <s^est'> "six".
I have of course suggested that the *x in the loc.pl. (*-sW-i ~
*-sW-u) and the 2sg.pres.them.act. -es^I (*-e-sWi) have the same
explanation. I would also want the dem. pronoun *so ~ *to- to have
been originally *sWo- ~ *to-, but of course Slavic offers no evidence
for that, because the nom.sg.masc. has been regularized to <tU>. Or
does it? Slavic is known to have used postfixed pronouns as a kind of
definite article (e.g. OCS rodU-sI > rodosI, rabU-tU > rabotU, and of
course novU-jI > novyj). In Bräuer I, $102, I find the following
interesting remark:
"Man findet ein solches -ch- auch in zahlreichen familiären Bildungen
und in Hyperkoristika als Ersatz regelmässiger Endungen: c^. brach
"Bruder, Kamerad" zu bratr, ar. ljachU "Pole" < *le,chU zu
*le,de^ninU, p. c^. Stach zu Stanislav, Mach zu Matve^j, Pech zu Petr,
Vach zu Václav u.a."
Any chance this might represent postfixed nom.sg. *s(W)o?