From: Jens Elmegaard Rasmussen
Message: 12848
Date: 2002-03-25
On Mon, 25 Mar 2002, Piotr Gasiorowski wrote:
> I agree with you as regards the relation between *-es- stems and
> *-eto- adjectives, though I remain unconvinced that the "-th"
> abstracts derive generally from the latter. The required middle term
> (*-eto-) is postulated in your argument but unattested -- at least I
> can't think of a convincing example (*nemos and the like are nouns,
> not adjectives).
I never called *nemos anything but a substantive, nor do I need to. I also
quoted the Greek s-stem substantives báthos and báros (just as examples of
a type). The type in *-eto- forms adjectives, right, but substantives may
be formed from adjectives if you extend the line, as this language very
often does.
> For example, *newotah2(t)- (or whatever the correct
> reconstruction) goes with good ol' *newo- rather than *newes-/*neweto-
> (unknown to me).
The type *newo- is old, but not typical; in fact, I know only *néw-o- and
its antonym *sén-o- as examples of primary adjectives formed from root +
thematic vowel. And of course *néwo- is itself special by being a vrddhi
derivative from *nu 'now'. Therefore, I don't see what would oblige the
language to contain an abstract adjective-based stative noun of the
structure *néw-es-, especially if this is something made only from a very
old stock to which this item rather plainly does not belong.
> Examples like Lat. tempesta:s or senecta:s show
> *-tah2-t- combined with bare consonantal stems, and note Gk. barute:s
> (baru-te:t-) = baros (bar-es-) and batHute:s = batHos. Sanskrit has
> guruta: as well, which confirms my suspicion that the original suffix
> was simply *-tah2-.
I think you're winding back the clock here. There is cross-branch
agreement in IE that s-stems are used to form nouns denoting the state
described by an adjective, as Gk. kûdos : kudrós, thársos/thérsos :
thrasús, or Skt. ójas- : ugrá-, ám.has- : am.hú-. In this procedure the
adj.-forming suffixes are not retained in the s-stems, and so - I say -
they would not be expected to be present in their further derivatives
either. That adj.-forming suffixes keep entering derivatives made from
adjectives is only to be expected of the subsequent language development.
Incidentally, tempest:as is not based on an s-stem adj., but rather
reveals a living pragmatic association between the s-stem and the
substantivized *-etaH2 formation that ended up meaning the same.
> BTW, in addition to the formation being discussed we have *-osti- (as
> in Slavic and Hittite), also with *-o-.
I know, and I take that as a point of support, seeing that here the
allomorphs with -s- and -t- have been superimposed (contaminated?).
> To sum up, I think that *dHeubetos, if correctly reconstructed, is not
> an abstract noun ('depth') -- those are different in Albanian -- but a
> substantivised adjective ('the deep'). This should probably presuppose
> an *-es-neuter (*dHeubos 'depth'), as in your theory.
Isn't that just what I said? At least I agree fully, thanks for the
support.
Jens