On Thu, 27 Nov 2003, Harald Hammarstrom wrote:
>
> I don't know about PIE but for PSem the "standard eplanation" is
> without (linguistic) evidence and contradicted by e.g Ar. baqarun
> 'cow/cows (collective)' vs. baqaratun 'a cow'....
This actually is the point (or should have been): The feminines are
individuals singled out of a collective mass. The group is unmarked, but
the remarkable attention drawn to a single item within it gets a
morphological marking. The example you give will appear to illustrate this
perfectly. Indo-European is not all that sophisticated, but simply uses
the collective itself as an animate singular. However, the basic
association looks the same.
Jens