Re: *a/*a: ablaut

--- In cybalist@yahoogroups.com, "Patrick Ryan" <proto-language@...>
wrote:

> I do not understand why <bhájati> needs to be reconstructed as *bho-

rather

> than *bhe-.
>
> More coloring?
>
> Why would a palatal <j> color anything to [a]?
>
> 'g'-coloring is a real 'slippery slope'.

Because in *bHag the *g is a back or low velar - the softening to /j/
is by the law of palatals, which might conceivably be as late as Indic.

Richard.