H1 and t??

[Glen Gordon]

Hello all, I goofed again.

They say "To err is human" but I can only suspect that someone has
slipped some hallucinogenics into my Ovaltine. I've re-read what Jens
has said about the "weak case-endings" and I realise now that I better
correct myself before the death threats start pouring in.

Let me re-illustrate, this time properly, the development of the
animate t-final stems with a pseudoword **kut (my vowel-final stem
illustration of **kuti still stands) and this time note the previously
neglected alternation of *t/*s in my paradigm example :(

                Pre-Pre-IE    Pre-IE                 IE
  Nominative    *kut          *kwet-se    *kwets     *kwets
  Accusative    *kut-im       *kwetm      *kwetm     *kwetm
  Dative                      *kwet-i:    *kweti:    *kwesi: (*kwesey)
  Genitive      *kut-isi      *kwetese    *kwete's   *kwese's
  Ablative      *kut-ita      *kweteta    *kwete'd   *kwese'd

Does that look right now? So **-t > *-s still works to explain much
however there seems to be an extra sound change lurking about that
I've overlooked. It occasionally wreaks havoc with the results. I
thought about a possible palatalisation of **t but that wasn't cutting
it.

Finally, I got it. There's an additional sibilization of intervocalic
MEDIAL **t. _That's_ why things are not working properly! Hence,
genitive *-us-os. Hence, feminine Nsg *-us-iH2. Hence, *le'wk-o:t-s
versus dative *luk-e's-ey. All these examples share the fact that **t
would be in an intervocalic medial position throughout its
pre-history.

This would explain alot more than just the quirky *t/*s alternation of
*-wo:ts. It would connect things like IE *ghesr and Uralic *ka"ti
"hand" together which could not have been due to any borrowing. They
aren't cognates per se but are both based on an earlier verb root
**git- and thus show a **-VtV- > *-VsV- sound rule in effect lending
further support for genetic relationship.

The word *nekwt- "night" derives then from **nug-t with substantive *t
that doesn't change because it isn't intervocalic.

This might also serve to explain those pesky endings in *-es that Jens
mentioned earlier alongside the supposed *-eH1- passive. Instead of
*-eH1-, we should reconstruct *-e:- which, due to accent rules and the
openness of the syllable, oscillates with *-e- (cf. *tu: versus
enclitic *twe). The *-es ending then is really *-e- (passive) + *-s
(from earlier **-t, the substantive, and hence a *t/*s alternation
in the declension).

Okay, that should do it now. I can only pray that makes better sense.

--------------------------------------------
Glen Gordon
glengordon01 at hotmail.com