glen gordon wrote:
>>But what's your basis for believing that all root
>>presents are secondary?
>
>
> When you mean "root presents" are you talking about
> athematic presents? The vast majority of presents
> are thematic, no? So it would be even more outrageous
> to claim that the vaster list of thematic stems are
> all secondary!
"Root presents" are those like *h1es-ti/*h1s-enti or
*gWHen-ti/*gWHn-enti, where the stem = the root, without any affixes
(excluding also the thematic vowel). The thematic presents in *-jé/ó-,
*-sk^é/ó- and *-é(j-)je-, while derived, must be historically old, given
the comparative evidence. The same is true of athematic presents
(including the root ones). The PIE status of simple thematic presents,
on the other hand, is dubious despite their being so numerous in Greek,
Sanskrit, Germanic etc. There are none such in Anatolian, and just a few
in Tocharian -- the branches that seem to have split off very early.
> Yes, but *-ye- is given to both noun and verb stems
> equally.
Yes, but is it the same *-je/o-? English has <-s> in plural nouns and in
the 3sg.pres. of verbs, but it isn't the same morpheme, whether
synchronically or etymologically. The nominal *-jó- forms adjectives
derived from thematic stems, so it makes sense to analyse it as *-e/o- +
*-é/ó-, like Jens does. One could toy with the possibility that the
verbal *-jé/ó- was likewise the result of "double thematisation" in
pre-PIE times, but I can't think of any independent motivation for such
an analysis. Crucially, the *-jé/ó- verbs correlate pairwise with
_athematic_ stems, not with simple thematic ones.
> We also have things like Hittite /newahh-/
> 'to make new', a factitive based on *newo-, even
> though *-h2- is normally applied to other verb stems,
> not adjectives. Seems to me like the earliest IE had
> a more liberal set of rules on derivation than you
> are suggesting above.
Why can't this *-h2- be a collective suffix, for example? *newa-h2 is
'the new', meant generically -- the collectve of *newo-m. This would in
fact account for the non-reduction of the vowel before the *h2, by
contrast to aor. *-eh1- vs. pres. *-h1-jé/ó- in statives. The second
vowel in *newah2(-je/o)- is underlyingly thematic, and so has to stay.
> So subjunctives correlate mostly with aorist stems
> in Tocharian, not presents? Did I get that right?
> I'll think about that some more.
That's right.
>>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, [...]
>
>
> So why not **sOwe:p- then? Why is *O doing an ad
> hoc dosey-doh?
It isn't ad hoc. In ordinary causatives there is an underlying short
vowel, lost pretonically, so that we get something like *O-mn- > *mOn- >
*mon-, as in *mon-éje/o-, In a long-vowel root the reduction still
leaves a short vowel which attracts the accent from the suffix:
*O-swép-. If my memory serves me right, Jens then proposes *swOép- >
*swoép- > swó:p-, but since compensatory lengthening is much more likely
when the lost tautosyllabic segment _follows_ the full vowel (cf.
English <night>, <folk>, or RP <card>, and of course *eH > V: in IE), I
think *O-swép- > *swéOp- > *swó:p- looks a bit more plausible. It's just
my personal opinion about a tiny fragment of Jens's hypothesis.
Piotr