Piotr:
>You've got to be careful with such implicational universals
True enough.
>It's more common for linguists to insist that Neg V is the expected order
>in
>an SVO language, but that implies NOTHING about SOV. Dryer (1988)
>demolished
>the popular view (endorsed e.g. by Lehmann) that SOV implied V Neg. In his
>sample of 117 SOV languages (from 23 families) Neg V was found in exactly
>one-third of cases (39 languages).
Yes, but aren't you admitting then that V-Neg is still found in the
_majority_
of cases? You know me by now. I keep myself aware of the difference between
absolute truths and varying probabilities. It would seem to me that Occam
would go with V-Neg nonetheless given these stats. Granted 66.6% still opens
things up to alternatives. However it might be more telling, in relation to
Nostratic if we can stomach it, if we knew of the percentage found in
Uralic,
Altaic, EskimoAleut and CKam languages specifically. It would seem to me,
admittedly impressionistically of course, that a tendency for V-Neg is
stronger here, particularly the further back in time we reconstruct these
languages.
I also feel confident that the relative particle was once placed after the
clause and I suspect that it might explain the origin of the conjunctions
*kWe and *ma which seem to eerily parallel Uralic's interrogatives in *k-
and
*m-.
Back to *-syo, btw, I mentioned that I analyze the suffix as *-s [genitive]
plus *yo-, an indefinite relative particle. It's postfixed in contrast
with the pattern of a relative particle _before_ its clause.
Also, the reason why I mentioned the subjunctive having a laryngeal *-h-
was because I'm suspecting lately that the subjunctive originates from an
entire clause-containing sentence structure of an original form:
*/SUBJ, OBJ kW& SUBJclause OBJclause VERBclause, *?& VERB/
The verb within the clause (VERBclause) would have been a bare stem,
terminating only in thematic vowel if applicable. The sentence is that
of a simple relative construction such as "Bob thinks that Jim is wrong",
which would be translated here into the pattern:
"Bob that Jim wrong being thus thinks".
So, as time goes on, the endingless clausal verb might have merged with the
following locative particle *?& (> *?e) which seperated the clausal verb
from
the main verb, producing a new "subjunctive" of the form *VERB-h&-. As
already discussed in another thread, *& becomes both *e or *o depending on
voicing of the following phoneme. This pattern also works for if-then
sentences, btw:
"If Mary goes, Yuri will cry."
becomes: "Yuri that Mary going thus will cry."
or rather: "That Mary going, Yuri thus will cry."
This idea might relate to word order because the immediate predecessor of
such a construction as "Yuri that Mary going thus will cry." could
conceivably have been "Yuri Mary going that (thus) will cry." with the
relative particle _after_ the clause and before the main verb.
This then is a nifty possible transition from "typical" SOV to the
less-than-majority pattern we see in PIE.
- gLeN
_________________________________________________________________
The new MSN 8: smart spam protection and 2 months FREE*
http://join.msn.com/?page=features/junkmail