Hi!
I can read that *a didn't really exist in PIE, except in onomatopeic
or emphatic words (e.g. ah!), and it developed in daughter languages
from *e,*o influenced by laryngeal sounds (and by borrowings, I
suppose). But then I see reconstructed words with *a, like e.g. below
*h2ag^H(e)n-. So, how it really is? I guess the issue is not really
settled yet...
> >Many linguists prefer to link Germanic *dagas to Old Indic ahar
(Gen. ahnas) 'day' < h2ag^H-r, *h2ag^H(e)n-. The initial *d in
Germanic has been explained as "stolen" from a preceding
demonstrative pronoun (*tod h2ag^Hr 'that day' > *..d-h2ag^Hr >
*dHagHr > Germanic *dag-).
/Jerzy