Re: [tied] The centum-word.

On Mon, 02 Apr 2001 17:58:42 , "Glen Gordon"
<glengordon01@...> wrote:

>I feel that *-i: was derived from the misanalysis of an earlier *t:wai-x
>"two", with a !COLLECTIVE! marker *-x, as *t:wa-ix. This eventually produced
>*dwo-i: as well as the dual marker *-i: which was secondarily placed on the
>word for "twenty".
>Whether I say plural or collective, it hardly affects my theory on these
>underlying little pokemon etymons... so was Miguel protesting anything
>important in the end?

Yes. The collective is not a plural. Neither is it a dual. And I
don't even believe collective *-(e)h2 has anything to do with the
feminine.

The dual endings are *-oh3 ~ *-(e)h1, the collective ending is
*-(e)h2, and the feminine endings are *-ih2 ~ *-(e)h2(i) [< *-(y)(e)h2
?]. They all have different origins. For the dual, this is obvious:
*-ih2 (your *-ix, I presume), should, as I said, have given Greek -ia,
not -i:, in the n.du. The link between the collective and the
feminine is part of PIE orthodoxy, so to speak, but I have yet to see
a good explanation for such a thing. To quote myself from sci.lang:

>On 13 Mar 2001 14:06:54 GMT, Greg Lee <greg@...> wrote:
>
>>Rich Alderson <alderson+news@...> wrote:
>>...
>>> First, the feminine gender as known in the classical IE languages arose as a
>>> reinterpretation of a neuter collective: If the primary animal in a herd is
>>> the male--important for breeding--the rest of the herd are referred to _en
>>> masse_, and then this collective term gets reapplied to non-male individuals.
>>> Once a concept of "masculine/feminine" is introduced, other pairings which did
>>> not originally have the collective suffix also get reinterpreted.
>>...
>>
>>So cows and women are like indistinguishable inanimates because of a
>>coincidence of form between two case endings, involving a single sound?
>>Wow. Speaking as an outsider to IE studies, this really sounds like a
>>crackpot theory, to me.
>
>It's an imaginative, if unconvincing, attempt to explain the formal coincidence
>between the collective (*-[e]h2) and the feminine (*-ih2/*-eh2). As a typological
>parallel one may adduce Semitic, where one of the plural markers (*-a[:]t-)
>[which may be taken as marked "inanimate" plural, besides animate plural *-a[:]n-
>and suffixless "broken" plurals] also does double duty as the feminine marker.
>
>Another option would be to treat the coincidence as just that, a coincidence, and
>look for other ways a feminine may have arisen (there is no doubt the feminine *is*
>secondary in PIE). One rather common development is diminutive -> feminine (e.g.
>the diminutive is one of the ways of explicitly marking feminine gender in Dutch),
>and it is tempting to compare PIE diminutives such as *-ik-o- with the feminine
>suffix *-ih2 (if from **-ik). I was going to put this forward as my own idea,
>but I see Greenberg anticipitated me on p. 166 of his Eurasiatic book ("Indo-European
>and its closest relatives").

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