Re: [tied] Re: IE prefix "*s"

From: Piotr Gasiorowski
Message: 28655
Date: 2003-12-19

19-12-03 20:05, Daniel Baum wrote:

> We have Latin super, Greek hyper, Sanskrit upari and Gothic ufar. On the
> face of it the first two are identical, as are the second two. However,
> the "spiritus asper" is obligatory in Attic Greek on all words beginning
> with /u/, so there is no way of telling whether the Greek form
> originally had an /s/ or not.
> Does anyone have an opinion on the matter? Do other Greek dialects shine
> any light on the problem, bearing in mind that some of them have lost
> the "spritus asper" altogether?
> As I am new here, I hope this has not already come up (I searched the
> list archive), and that I am not diverting this apparently very long and
> varied thread too far from its topic.

Welcome, Daniel,

As far as I know, no Greek dialect distinguished etymological *u- from
*su-. The non-psilotic dialects (such as Attic) always had /hu-/ in both
lexical sets, and the psilotic dialects consistently dropped their
aitches and always had /u-/. The *s of *(s)upo- and *(s)uper- is
believed to be a case of "*s mobile". However, if we simply count the
votes, it turns out that Lat. sup- gets no real support (or shall I say
upport?) from elsewhere and may represent an Italic idiosyncrasy, while
*s-less forms occur rather widely. In Celtic we also get *wo-, *wer- ~
*wor- < *upo-, *uper-. I would therefore risk the opinion that (Attic)
Gk. hupo, huper are more likely to reflect *upo-, *uper-, impossible as
it is to rule out protoforms with *su-.