Re: [tied] PIE athematic neuters

From: Miguel Carrasquer
Message: 43858
Date: 2006-03-15

On Wed, 15 Mar 2006 11:45:38 +0100, Piotr Gasiorowski
<gpiotr@...> wrote:

>On 2006-03-15 10:08, Miguel Carrasquer wrote:
>
>> According to my interpretation, due to underlying length in
>> the neuter form:
>>
>> *kí:rd > *k^é:r(d)
>> *ki:rdás > *k^r.dés
>
>So you have shortening to zero in the weak allomorph -- that looks more
>like a stipulation than something that could be proved with independent
>evidence.

I don't remember now which evidence specifically drove me to
conclude that the zero grade of **i: and **u: was zero, as
opposed to **a: which has stress-attracting é in the
zero-grade. It was probably the ablaut o ~ 0 of words like
*pónto:Hs, *pn.tHós (amphidynamic), which I could only
explain as coming from underlying **u:, not **a:. The words
with e: ~ 0 ablaut, such as *k^é:r(d) ~ *kr.d- nicely fit
the pattern, and cases like Vedic má:rjmi, mr.jánti now
become regular.

>> (Actually, the oblique is more like *k^r.diyós, which leads
>> me to believe that we're not dealing with a root noun,
>
>I suspect the same on other grounds: *k^red-, found in fossilised
>univerbations, _might_ reflect a forgotten locative, presupposing
>*k^er-d-. Of course the long-vowel problem remains.
>
>> and
>> the reconstruction could be something like:
>>
>> *kí:rd-in > *k^é:rd(r) > *k^é:r(d)
>> *ki:rd-ín-a:s > *k^@rdéyos > *k^@r.d@... > *k^r.d(i)yós
>> )
>
>Wouldn't something more orthodox be preferable, like a variant with the
>suffix *-ejo-, coll. *-ijah2 (Ck kardía, non-Att. kardíe: ~ kradíe:), so
>frequent in anatomical terminology, from which a secondary i-stem could
>easily have been extracted?

I want to explain e.g. Hitt. kir, kardiyas as being a single
paradigm, not the amalgamation of two.

>I realise full well you would derive *-ejo-
>from **-in-, but my point right now is that the suffix, whatever its
>ultimate etymology, is not necessarily an original part of the 'heart'
>lexeme (or 'bone', for that matter).
>
>>>The *h1we:su-s/*h1wosu type shows the difference clearly
>>>(though both are static and share the "weak" (e-grade) case forms.
>>
>>
>> Is this based on Irish fó "gut, Güte" vs. fíu "würdig"?
>
>Yes (also Wel. gwiw for *h1we:s-u-, and OIr. adv. feib 'in excellence'
>representing an old weak case, *h1wés-w-ei), plus traces of *h1wos-u in
>Anatolian (Luwian and Palaic), e.g. HLuw. neuter nom./acc. <wa:su->
>(also used as an adverb and occasionally substantivised).

But I wonder whether there is sufficient evidence to warrant
a reconstruction n. *h1wosu, mf. *h1wé:sus. In particular,
*h1we:su- is attested very poorly (Pokorny's Gaul. Vi:suri:x
besides Gmc. Wisuri:h seems arbitrary).

=======================
Miguel Carrasquer Vidal
mcv@...