Re: [tied] PIE athematic neuters

On 2006-03-10 01:03, Miguel Carrasquer wrote:

> My views on the length issue are somewhat diffrent, as you
> know. I regard forms like *doru(r) or *wodr as having
> underlying **a: or **u: vocalism, while original **i: yields
> *e:, which has the advantage that the two types become
> fundamentally one. Posttonic lengthening (after light roots
> only) explains the vocalism of the proterodynamic type
> (*h2ák^-ma:n *h2ak^-mán-a:s => *h2ákmon-, *h2k^ménos), while
> the amphidynamic type is a special development of roots
> containg long *u: and *i: (e.g. if the water root is from
> *u:d, not *wa:d-: *ú:d-an ~ *u:d-án-a:s ~ *u:d-án-a(i) =>
> *wódr, *udnós, *udén(i), with accent shift in the oblique if
> a long vowel follows). I have explained this better and
> more at length the past...

My current view (very tentative, but I'm still thinking) is that the
*-r/n(t)- neuters are in fact old -nt- neuter participles (hitherto,
essentially, a lost tribe since there is little other evidence of such a
formation). If so, they should somehow fit into the verb system. I can
only make sense of the type represented by *potr. (Hitt. pattar)
'feather, wing' if I assume that a Narten-verb neuter participle
*pe:t-ent- received contrastive root accent in the nom./acc.
(distinguishing it from both animate *pe:t-ént-s 'flying, flier' and the
collective *pe:t-ént-h2), so early that the pretonic vowel retained its
length. After the operation of the usual ablaut rules (including regular
accent shifts) we should get *pé:tr., gen. pétn(o)s, coll. *péto:r, etc.
What remains to be explained is the "substitution" of *o for the
expected long vowel. If it reflects overlength, this means that the
nom./acc. was affected by some kind of vr.ddhi, and if so, the same
process may be responsible for the alternating strong-case forms
*h1we:s-u-s/*h1wos-u, for the length in neuters like *k^e:r(d),
*wa:st-u, *g^e:r&2s (Gk. ge:ras 'old age'). I'm not sure under what
conditions the expected *o: was shortened or why *e: occasionally
preserved its length in similar contexts.

> Now I too have been wondering about the peculiarities of the
> neuter in this respect, and I've had a suspicion for quite
> some time now: what if it's not a peculiarity of the neuters
> that they have heavy stems, but that it's a peculiarity of
> heavy stems that they are neuter? The nominative and
> accusative endings *-z and *-m are asyllabic and the only
> ones to be so, and a soundlaw that drops them after a heavy
> / long vowel syllable is not unthinkable.

It seems to me that the quantitative anomaly in question is restricted
to the nom./acc. of neuters; there's nothing funny about the other
cases. In other words, it looks like a process which affects not the
stem as such, but only an individual case form.

Piotr