From: fournet.arnaud
Message: 56497
Date: 2008-04-03
----- Original Message -----
From: Patrick Ryan
> > >
> > > Chinese xu4 "raw silk, thread" < AF *sneH1-
> > > Baxter *snjaH
> > >
> > > Chinese xiu4 "embroider" < AF *syewH1
> > > Baxter *sjuwH
> > >
> > > Chinese xiu4 "sleeve" < AF *sluH-(bh)
> > > Baxter *zluwH-
> > >
> > > PIE loanwords with H1 pharyngeal unvoiced
> > > causing QuSheng.
> > >
> > > H2 causes ShangSheng.
> > > As in AF kuH2on "dog" > quan3.
> > > Baxter *kuHen
> > > Cf. PAA kvHvn "dog"
> > >
> > > Arnaud
> > >
> > ==========
> >
> > I give what Baxter 1992 has reconstructed.
> > I suggest better reconstruction for Proto-Chinese
> > on the basis of what PIE is.
> >
> > Arnaud
> >
> > ===========
To what does Baxter attribute the difference in outcome of ShangSheng and
QuSheng?
Patrick
***
According to Haudricourt's theory
originally based on Vietnamese
ShangSheng words had a final -?-
QuSheng words had a final -s-
Baxter (and others See Yahontov in Starostin's database)
automatically adds -?- or -s- to proto-forms.
I think this automaticity is probably flawed,
IE LWs show that
some QuSheng are from H1 /pharyngeal unvoiced/
It means a change in proto-Chinese :
s > H1 > QuSheng
some ShangSheng are from /velar voiced spirant/
And H1 = -?- is falsified by Chinese.
Arnaud
==============