Re: [TIED] E-lim-mi-nate the negative...

From: Glen Gordon
Message: 2496
Date: 2000-05-23

> IE thematic adjectives are thought by many to be of the same origin >as
>genitives, especially given the Hittite form of thematic genitives >in -as
>(*-os).

Actually this is exactly something I was starting to contemplate after I
sent the last message. Eery. I'm relieved that you confirm the validity of
this idea. It would have to be a misseperation of genitive *-os as *-o- +
nominative *-s, yes? I'm starting to lean towards this.

It would make sense for the fact that the genitive and nominative were
always similar in form and that after the establishment of the nominative
*-s, an adjective marked in the genitive *-es/*-os next to an animate
nominative noun in *-s would appear to "agree" with each other in case,
causing some confusion about these "genitive adjectives" and about proper IE
grammar.

>[...]*wlkWos is initially stressed ('acrostatic') [...] but this may >be a
>relatively late shift (cf. contrasts like t�mos 'slice', tom�s >'cutting,
>sharp'; it seems that accent retraction could be used to >distinguish
>thematic nouns from adjectives).

... As is mentioned in T.Burrows "The Sanskrit Language" that I love so
dearly. The accent would still have to be explained if one is to attribute
the thematic ending to a misanalysis of a previous genitive ending. Hmmm...

According to Burrows, he suggests that there was an adjective (I presume
**wlkWo's) that is the basis of the noun "wolf". What would **wlkWo's mean?
What does the root **welkW- signify?

It would be tempting to associate the perceived accentuation changes of
*-r/*-e:r with this same puzzle of the thematic nouns but I haven't figured
things out enough yet. I'll keep yo'll posted.

> To me, *p@...:r/*p@... is practically the same pattern as
>*po:t->s/*ped-�s, [...] This contrasts with 'flexion ferm�e' [...]

We agree although I would seperate this flexion ferm�e into two different
developments: one of regularisation of stress (fixed initial) and that of
derivation (fixed final).

> Your derivation of *-e:r from **-e:n raises a number of questions.
> >There is no *r/*n alternation in "animate" (i.e. non-neuter) >paradigms.

Yes, I know. I don't claim "that *r was somehow borrowed from inanimate
heteroclita". That wouldn't work, of course. I think that the inanimate *-r
and the animate *-e:n were once very different suffixes. In Sanskrit
however, we see a blunt association between the two. They are merely
distinguished by accent at this point and have become part of a larger
pattern of "semantic" accentuation (eg. the accentuation of the /-man/
ending which is one of many reincarnations of this suffix).

From start to finish, here's what I propose. The *-r ending was originally
Mid IE *-en. Throughout all the cases, there was no alternation in the
beginning. The alternation began after the adoption of the penultimate
accent, which caused the accent to shift from syllable to syllable, and
after a regular phonetic change of final *n to *r, which is the direct cause
of that heteroclitic declension. This caused pairs like *we't:en/*wet:ene's
to become *we't:er/wet:ene's.

This alternation would not arise in animate agent nouns with *-e'ne for a
very obvious phonetic reason. There was never a final *n throughout its
paradigm and an alternation would not result. And so, time would pass and
the final vowels would eventually drop. This last change would obscure the
reasons for the accentuation difference between inanimate *-er/*-en- (later
*-r/*-n-) and animate *-e'n (later *-e:n, from compensatory lengthening
through loss of nominative *-s).

The suffix *-e:r is a later suffix, derived from *-e:n. It seems to have
been _inspired_ by (not directly taken from) the nominative of words with
inanimate *-r. An association of accent between the inanimate *-r via its
non-nominoaccusative alternate *-n- had already been assumed for the animate
ending *-e:n. The new heteroclitic alternation in the inanimate started to
bring confusion about the animate suffix that was incorrectly associated to
it, producing the alternate *-e:r. Of course, the suffix would have no
alternation in non-nominative cases since this didn't exist in the animate
paradigm anyway.

So, two suffixes eventually existed interchangeably side by side, the
original *-e:n and the later *-e:r. The former is directly connectable to
Etruscan -na which is used to denote peoples or groups.

As for the ultimate reason for inanimate nouns in *-r (or Mid IE *-en), I
still can't find any promising *-r correlations in Etruscan. They would show
up with /-n/. What I do know is that words with this *-r are found without
it in Uralic cognates, as in *wete "water". There are also some Altaic words
that would almost appear to me to employ this *-n ending like IE (a hunch).

>Also, even if the older form had been eliminated from paradigms, I >would
>expect it to leave some traces in derivationally related words; >but all
>derivatives from *p@... that I know have *r, not *n, and the >same is
>true of other -er animates (from well-known family terms to >agent nouns in
>*-te:r/*-to:r; note that NEUTERS in *-t@..., such as *ei->t@... 'path, route'
>are heteroclitic!).

Sheesh, Piotr. I know that! The word *p@... is a later word for the very
reason that it has no *-e:n ending, whether in the nominative or any of its
derivatives. Either the *-e:r suffix was used or *-e:n, not both at the same
time within a single paradigm. Anyways, it's not the original word for
"father" that's for certain.

Neither are many of the terms for family members that end in *-te:r. This
was a later addition, perhaps added to words that originally were without
this ending at all. One might suspect an earlier Mid IE **deuge "daughter"
(Etruscan s'ec) or **berahwe "brother" (Etruscan ruva)? A kite, you say?
Perhaps.

- gLeN

________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com