Re: IS PIE * DERU EXCLUSIVELY INDO-EUROPEAN ?

On 2008-01-27 23:38, fournet.arnaud wrote:

> OK, that's one of them, "Western" *dak^ru ~ "Eastern" *h2ak^ru-.
>
> ===========
> Thank you for this unexpected support.
> ===================

Huh, but it's the only one. I wouldn't dare to speculate about a prefix
on the basis of a single example that may be due to, say, reanalysis of
a set phrase or some such syntagmatic unit (the "nickname" phenomenon)

> > arbor = doru
>
> Any particular reason for connecting these two, other than the meaning?
> =================
> The same reason as for dakru = akru
> ================

The conditions are noth the same here. *h2ak^ru- and *dak^ru- are
identical except for the onset, while <arbor> and *...oru are not
directly relatable.

> The 'arse' word is to be reconstructed as *h1orsos. But medial *-rs-
> gives Lat. -rr-, so <dorsum>, whatever its etymology, reflects *-rt-t-
> rather than *-rs-.
> =============
> ok,
> so the form should be HorT- (T unknown status )
> Hence hort-to and t?-Hort-to

We know its status. It's just *s. The cluster behaves like *-rs-, not
like *-rt-t-, in Greek and Hittite.

> Why should it be H1 ?

Hitt. a:rras, with no <h>.

> I disagree with this statement "not supported".
> And it sounds strange : you agreed with D-akru = Hakru

I agreed that the reconstructions were correct so you have ONE example
of a word with or without an initial *d-. I don't agree this is evidence
for a suffix. English has a non-etymological /n-/ in <newt> (~ <eft>)
and <nickname> and LACKS an expected /n/ in <apron> and <adder>, but
this is not sufficient evidence for positing an n-prefix in English and
compare it with something in Georgian or Amharic.

Piotr