Re: Etymology of 'daughter' (was: Rg Veda Older than Sanskrit)

On 2008-04-22 18:28, batinquo wrote:

> For 'star', he rules out the presence of the agent suffix -ter- on the
> basis of Jasanoff's "quite important observation that root nouns
> normally do not have doublets (parallel action or agent nouns) derived
> from the same root with any suffix"; since there is evidence for a
> root noun *h2oh1s- *h2eh1s- ('hearth', 'ashes' etc), then 'star'
> cannot be 'the burning thing'.
>
> This was news to me - is that argument really decisive?

_Any_ suffix? What about pairs like *h3re:g^- ~ *h3re:g^-on-? It's true
that the survival of an older and less productive type of derivative may
prevent the formation of a productive one (the way that the existence of
<cook> in English blocks the interpretation of <cooker> as a human
agent), but *h2oh1s-/*h2ah1s- was not necessarily an agent noun. If
*-ter- has a contrastive value in *h2(h1)s-te:r, what is the 'star' word
contrasted with? A black hole?

Piotr