On 2006-01-25 16:31, tgpedersen wrote:
> There's no way around assuming that the two are examples of the same
> type of morphological construct, therefore, if a *ni-zd- is something
> you sit down on, then an *o-zd- is something you sit on, not something
> that sits on something else. And on branches not much else sits but
> birds.
Well, there's always the possibility that *hosdos (Hitt. hasd-wer-
'branch' requires *h2 or *h3) contains an indivisible root *hosd- (the
structure of the Hittite word would be more consistent with such a
solution). The connection between birds and nests is undeniable. Birds
_make_ nests, after all; they don't merely sit in them. Sanskrit even
has the athematic compound <ni-s.ad-> 'sitting motionless', with
agentive *-sed-, which is just what we would expect to underlie *nisdos.
There is no such evident connection in the case of the 'branch' word,
though of course I admit it as a possibility.
Piotr