Re: [tied] IE thematic presents and the origin of their thematic vo

glen gordon wrote:

> If 'non-suffixed' here means 'athematic' then we're
> on the same wavelength. That similarity is what
> makes me feel that the athematic stems are originally
> aorist and that originally, inherent duratives were
> given *-e-. That seems to make the root aorist the
> unmarked part of this earlier system, interestingly
> enough.

But what's your basis for believing that all root presents are
secondary? There're quite a few of them and they're well in evidence
everywhere, including Anatolian. It seems more likely to me that root
presents and root aorists are equally ancient. If a durative verb had to
be derived from an aorist, the most common morphological markers were
*-je/o- or *-sk^e/o-, or a nasal infix. On the other hand, aorist stems
could be derived from root presents by adding *-s- and lengthening the
underlying root vowel.

As regards the subjunctive theory of the origin of the plain thematic
present, it's instructive to look at Tocharian: only three verbs of that
class correspond to simple thematics outside Tocharian (and Anatolian);
the others mostly function as subjunctives in Tocharian itself, and
typically correspond to reconstructible PIE root verbs (mostly aorists).
In other words, inherited subjunctives of this type outnumber the
formally identical simple thematic presents in Tocharian. This squares
well with the hypothesis that the shift from subjunctive to durative
meaning took place gradually after the IE breakup.

> The exact origin of Narten presents still stumps
> the hell out of me. However, I wouldn't call the
> sigmatic aorists "Narten alternations". There, the
> aorist's vowel has been lengthened by the same
> sound change that has caused lengthening in the
> nominatives in *-s. 'Clipping', a corollary of
> Syncope. It works for all fricatives, like *-x, too.

How does that explain the pattern *wé:g^H-s-t vs. wég^H-s-n.t? It
doesn't seem to be any different from *ste:u-ti vs. *stew-n.ti .

>>My suspicion is that the barytone thematic type
>>(*bHér-e-) is the subjunctive of the lost
>>pre-sigmatic Narten aorist, [...]
>
>
> I think saying "pre-sigmatic" here is untenable.
>
> I link the sigmatic aorist with Tyrrhenian *-as-e,
> seen in Etruscan (-asa), EteoCypriot (-as(a)i) and
> Minoan (-asi). I similarly link the IE *n-infix
> with Tyrrhenian *-an-e.
>
> Even if you don't appreciate the connections I make
> between IE and Tyrrhenian and wish to take the
> stricter internal reconstruction route, you still
> have to contend with these tasty grammatical
> parallels:
>
> *-no- <=> *-n-
> *-to- <=> *-s-
>
> Nope, I don't think there ever was a 'pre-sigmatic'
> stage in IE, unless we're going back many millenia.

"Pre-sigmatic" here only means root aorists -- a class which was on the
decline, giving way to the transparently aoristic *-s-stems. Root
aorists surely existed till the end of PIE, and afterwards.

> What this looks like to me is that *-he- has been
> added in the subjunctive *stew-he-, originally with
> accent on the suffix, therefore explaining the
> shortening of the once unstressed root vowel.

Yes, I also assume an originally accented suffix, the only difference
between us being that I don't see much evidence for the *h1. Not that I
can exclude it absolutely, but I think Ockham is on my side until such
evidence can be presented.

> And if we are really talking about plain ol' *-e-
> in the subjunctive, why would we theorize
> *swo:p-eye- > *swo:p-ye- and yet also *ste:w-e- >
> *stew-e- then?

If one accept's Jens's infix theory, pretonic **O-swe:p- becomes
**O-swep-, then the remaining full vowel attracts the accent; next, the
still consonantal *O gets metathesised and eventually vocalised,
coalescing with the full vowel and causing it to lengthen. All is
regular here. In the case of *ste:w-é- we have the usual shortening of
the pretonic long vowel and the retraction of accent, but no lengthening
by contraction, hence the outcome *stéw-e-.

Piotr