Re: [tied] The phonetic value of PIE *h3 and the 'drink' root.

From: Jens Elmegaard Rasmussen
Message: 14127
Date: 2002-07-24

The root *kreH1y- is at the same time a "long-diphthong root" and an
example of the ERU/RU: structure. Roots ending in -VHy alternate in a
special way in the full grade, in that they lose the /y/ before
tautosyllabic consonants (i.e. word-final consonants or consonant
clusters). That makes the aorist *kreH1-t which is reflected by the Latin
pf. cre:vi:. - The zero grade of a root with CVRHy or CRVHy ends in CRHy
which is regularly metathesized to CRiH before consonants (but retained
before vowels; in like fashion *CRHw turns up as CRuH before consonants).
The zero-grade CriH- in turn causes new full grade to be back-formed as
CreyH- (handbooks record a Latv. kreju and express doubts as to the
authenticity of a Lith. krejù which would fit here). From such a structure
the old causative **kroH1y-éye- was reshaped to *kroiH1-éye- > krojiti.
The metathesis must have deep roots (although it must be younger than
ablaut proper which supplies the input), for the lexicalized instrument
noun *krei(H1)-thro-m seen in OIr. criathar, Lat. cri:brum, OE hri:der is
based on it. Metathesis may also be supported by the special structure of
the nasal present which, from in -yH and -Hi alike, is always with the -i-
before the infix and the laryngeal after it, as *kri-n-H- > Lat. cerno,
*dhi-n-H1- > OIr. denaid 'sucks'. That just may explain dojiti and the
base verb Skt. dháyati from *dhoiH1-éye-/*dhéiH1-e-ti instead of expected
*dhoH1y-éye-/*dhéH1y-e-ti (the latter in Latv. dêju and OHG ta:en).
While I claim to have settled most of the problems concerning the long
diphthong roots and the ERU/RU: alternation in the two first chapters of
my 1989 book on IE morphophonemics, I am not so foolish as to believe that
the last word has been said. Especially, I could only hint at the problems
of interaction between these structures and the funny o-vocalism of
causatives (and some other formations) which I was forced to diagnose as
an old consonant in the second half of the same book. Even the few
definitive solutions I did arrive at still strike many as appalling, even
on this list which is not generally for the meek.


On Wed, 24 Jul 2002, Sergejus Tarasovas wrote:

> -----Original Message-----
> From: Jens Elmegaard Rasmussen [mailto:jer@...]
> Sent: 2002 m. liepos 24 d. 02:51
> To:
> Subject: Re: [tied] The phonetic value of PIE *h3 and the 'drink' root.
> >> I do have Slavic krajI 'edge, region, land' looking very much like
> *kroH1y-o-s
> 'distinction'.
> How would you etymologize Slavic *krojiti 'to cut (cloth etc.)' then?
> Sergei