>> That depends what *dh and *gh stand for. We still don't know.
>
> Um, what do you mean by that? Don't they stand for an aspirated "d"
> amd an aspirated "g"? Or is there another side of this I'm missing?
They were written that way when they were first reconstructed, because it
was thought that Sanskrit preserved PIE better than other languages. Since
then we have realised that Sanskrit is highly innovative in both morphology
and phonology. So all we have is the reflexes in the various languages, and
we have to work out what kind of sound could explain all those reflexes.
But our spelling has got stuck; *dh, *bh, *gh are going to be with us
forever, whatever they actually were.
Some people have suggested fricatives.
Some have suggested voiced stops with optional aspiration (plain voiced are
then glottalised)
Some have suggested voiceless aspirates (to make typological sense of PIE)
One of the problems is that almost all known languages with voiced aspirates
have voiceless aspirates, and almost all known languages with aspirates of
any kind have the phoneme /h/; but PIE shows neither of those certainly.
What we have reconstructed is a rather unlikely system
Several language groups (Hittite, Tocharian, Iranian, Celtic, Slavic,
Baltic) show no difference between *dh and *d, or *gh and *g.
Amongst those that do distinguish them only Sanskrit (not even Iranian)
shows voiced aspirates.
Germanic and Armenian show voicing, Latin shows variation in voicing; and
Greek and Oscan show no voicing.
Only Greek and Sanskrit show aspiration (though Italic can show the reflex
/h/).
So with this variety, it is no wonder that different people make different
suggestions.
Peter