--- In
cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@...> wrote:
> > Please consult the Lubotsky's article "Against a Proto-Indo-
> > European phoneme *a", available on
>
> This link should work better
>
>
> https://openaccess.leidenuniv.nl/handle/1887/2662
Having now read this article, I think the author makes good points.
Most interestingly to me, his explanations for the apparent a-
vocalism in e.g. *pag-, *swad-, *bhag-, *yag-, *mad-, *kad-, etc. as
containing the 'h2' laryngeal (probably /x/) suggest to me that these
were extended roots:
*pex-g- > *pa(:)g-
*swex-d- > *swa(:)d-
*bhex-g- > *bha(:)g-
*yex-g- > *ya(:)g-
*mex-d- > *ma(:)d-
*kex-d- > *ka(:)d-
As is well known, *-d- and *-g- are common root extensions. What
this means is that the words in question are more morphologically
complex than can be seen at first glance. A common word like
*swéxdus 'sweet', then, can be seen as an originally deverbal
adjective: *swex-d-u-.
- Rob