Re: *a/*a: ablaut

From: Piotr Gasiorowski
Message: 52859
Date: 2008-02-12

On 2008-02-12 21:11, alexandru_mg3 wrote:

>
> 1. Do you accept 'the loss of laryngeals before mediae in Indo-
> Iranian"?
>
> like in:
> Skt. vájra-
> Skt. pajrá
> etc..

No, it's Lubotsky's private theory, not accepted by the majority of
IEists. In roots like *bHag- even LIV has *a. The reconstruction
*bHah2g- only creates problems instead of solving them. Lubotsky
performs some acrobatic tricks to explain away the short /a/ of
<bHájati>, while the simplest solution (*bHeg-/*bHog-/*bHg- with the
e-grade coloured to *a by the following *g) explains all the attested
forms without invoking analogy or any other "special effect". Forms like
<bHa:jáyati> and <babHá:ja> have /a:/ from the expected o-grade via
Brugmann's Law (*bHog-éje/o-, *bHe-bHóg-e). Just forget the laryngeal
and let regular sound changes do their job.

Incidentally, this verb has *e as its fundamental vocalism; it isn't one
of the *a:/a roots. Mixing it up with *wa(:)g^- etc. is wrong.

Piotr