Re: [tied] Sanskrit form a.s.tau

In most branches the numerals 5-10 are indeclinable (if preserved without secondary suffixation). The apparently dual ending of 'eight' was a lexical fossil; the word was not treated as a real dual sychronically. The Sanskrit plural endings are clearly secondary:

pan~ca : pan~ca-bHih.
s.at. : s.ad.-bHih.
as.t.a:(u) : as.t.a(:)-bHih.

It's clear that the case endings are simply cliticised to the "free" (nom./acc.) form of the numeral, which it retained in its archaic form, without any plural inflections.

Piotr

----- Original Message -----
From: <mmdesh@...>
To: <cybalist@yahoogroups.com>
Sent: Sunday, December 22, 2002 11:30 AM
Subject: [tied] Sanskrit form a.s.tau


> Hello friends,
>
> In Sanskrit, the nominative form for "eight" comes out to be
> a.s.tau (a dual-like appearance), while forms for instrumental
> etc. come out looking like a.s.taabhi.h (with regular plural
> endings). Is there an IE based explanation for this behavior?