> > > And BTW how come the very early PIE compounds *ni-zd- "nest", *o-
> zd-
> > > "branch" both have to do with birds, not just "sit"?
> >
> > The analysis of *h(2/3)osd- as *ho-sd- is a little suspect: it
> > presupposes the existence of a hypothetical PIE preposition with
> an
> > uncertain meaning, reflected as *ho- in this word and not doing
> much
> > other useful work elsewhere.
>
> Greek o-kello "rush ashore", I believe. Why the laryngeal?
>
"run a ship aground", actually.
Another quote from P.'s article, cognate set 20 "sit, set, west, nest"
"
The homophonous [of 'west'] word xi: 'roost, settle' does not have a
separate entry in the dictionary, but it is stated that xi: "west;
roost" can also be written with 'tree' + qi: "wife":, E(arly) M(iddle)
C(hinese) tsHej. The modern reading qi: for 'roost', which lacks
ancient authority, may be a spelling pronunciation based on this form
of the graph.
"
Unfortunately I am not equipped to determine whether the 'tree'
component of this Chinese character is a phonemic, thus with no
semantic function. Otherwise, it's odd that there should be a semantic
connection with "tree", unless it is the 'sitting-branch' variant, PIE
*o-zd-, appearing in Chinese too (with some prefix?, cf. the EMC forms
s^ej vs. tsHej.
Torsten