Re: [tied] 1sg. -o: [was Re: IE Thematic Vowel Rule]

Miguel Carrasquer wrote:

> For me, the dual in -o: is clearly *-oh3. The *h3 has a rounding
> efect on the locative suffix *-i in *-oi(H)-h3-i > *-ojju(m) (> Grk.
> -oiiun), and the Sanskrit alternation -a: ~ -a:u is also best
> explained by a rounded laryngeal.

We've been through this before. The contracted perfects with -a:u <
*-a:(H)-a < *-oH-e show the same development. For me, this amounts to a
demonstration of the "intrusive" character of the -u/v.

> Skt. GL -ayo:s (*-oyHous) also
> reqires the presence of a laryngeal, as do the thematic neuter ending
> *-oiH and the athematic ending *-iH.

The _neuter_ dual ending is a separate issue. The lengthened final in
non-neuter *i- and *u-stems may be analogous to the thematic *-o:.

> The odd form is laryngeal-less -e (actually: *-ye, as shown by Greek
> osse < *okW-ye). My explanation is that it comes from a lengthened
> variant (in posttonic position) of the dual ending *-iH > *-i:H >
> *-ye:H, with loss of the laryngeal after the long vowel, and
> subsequent shortening of unstressed *-e:.

I must say I like Jens's elegant derivation of animate duals from
pre-PIE *h3, retained internally in some PIE case forms but vocalised to
*-e in _word-final_ position (or deleted after the insertion of *e as a
prop vowel). Of course, it's very much an open problem.

Piotr