[tied] Re: PIE i- and u-stems again

--- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...> wrote:

>
> On 2007-02-15 23:02, Miguel Carrasquer Vidal wrote:
>
> >>> There are no n-stems derived from thematics
> >> What is then the relationship of catus/cato: ?
> >
> > One is thematic, the other is an athematic n-stem. The
> > n-stem is not built on the thematic form: that's impossible.
> >
> > You *can* build thematics on consonant-stems: happened all
> > the time.
>
> Well, so what's the relationship between Gk. gnátHos 'jaw' and
> gnátHo:n 'chubby' or híppos and hippó:n 'stable, posting-house'?
> It's quite clear to me that Hoffmann formations in *-h3on- are not
> restricted to Tocharian, and that's one clear source of n-stems
> derived from thematics.

And if the Hofmann suffix should somehow be identical to IE *en as a
suffix (as in Oscan) / old genitive à la F-U, then an n-stem built on
a thematic stem would correspond exactly to

Mordvin
moda "earth"
moda-so "in the ground"
moda-so-s´ "the one in the ground"

PIE (all wrong, but close to the intention)
*wVd- "water"
*wVd-en -> *wVd-er "in the water"
*wVd-en-s "the one in the water"
The last one is certainly untrue in this particular case; I hope the
idea is right anyway.

> The origin of the catus/cato: type is debatable, but one distinct
> possibility (Birgit Olsen's idea, I think) is that at least some
> "individualising/definite" nasal stems corresponding to thematic
> (and other) adjectives were originally -n(t)- participles of stative
> verbs, derived in turn from adjectives (*X-h1-on(t)- '[singled out
> as] being X').

Cf. the Hittite animate -ant-.

> The process is a complex one but again it is eventually capable of
> transforming thematics into nasal stems.

Lars Brink in "PIE Feature Synchronization and Verner's law" in Jens'
Festschrift tries to nail the pesky -d- of Goth. tamida from accented
*-tó by positing a conceptual syllable division after the dental, thus
accounting for Eng. dead vs. death. Now suppose the suffix was
*-én-/*-ndV´-, then the latter variant might develop into *-n|t- by
new syllable division (and if at the same time something like *dhó-
generalized as *-tó-).


Torsten