On Wed, 11 Feb 2004 20:25:04 +0000, P&G <
petegray@...> wrote:
>This, I believe raises other problems. (I've pinched the examples from
>Sszemerenyi.)
>(a) What of forms without thematic vowel, such as (in Greek) eupáto:r,
>dysáno:r, áfro:n, dó:to:r, daímo:n, génos, and the forms with -eh2 such as
>forá?
>(b) What of those where the -o- grade takes the accent, such as fóros, fó:r,
>aidó:s, gónu?
There is more than one origin for /o/-grade. The way I see it [but compare
Jens' answer the same questions], we have at least the following cases:
* "original" /ó/ (< **á:), stressed, in strong forms of the root. It
alternates with stressed *é in the weak forms, e.g. Nom. **pá:d-z > *pód-z
> *pó:d-s Acc. **pá:d-m > *pod-m, Gen. **pa:d-ás > **pádas > *péd-s, etc.
This includes your example NA *g^a:nu(n) > *gónu(r), Gen. *g^a:núna:s >
*g^ánuwos > *g^énwos; also (**bhá:r-s >) phó:r; and for instance a form
like *wá:d-an > *wód-r, *wa:dána:s > *wédnos.
* collective /ó/, stressed. In forms like (**wad-á:n > *udór-h2 >) *udó:r,
Gen. **wada:nás > *udéns. Perhaps your example aidó:s (*h2isd-ós-h2)
belongs here originally (the Gen. aidó(s)os > aidoûs is then analogical).
* post-tonic /o/, unstressed. In my view this is most often the result of a
special lengthening of posttonic (svarita) syllables in pre-PIE. For
instance, proterodynamic roots with the suffix *-men- alternated
(svarita-lengthening notated as â):
N. *h2ák^-mân-z > *h2ák^mo:n
A. *h2ák^-mân-m > *h2ák^monm.
G. *h2ak^-mán-âs > *h2(a)k^ménos
This corresponds to hysterodynamic:
N. *pah3i-mán-z > *poh3imé:n
A. *pah3i-mán-m > *poh3iménm.
G. *pah3i-man-ás > *p(o)h3imnés,
without lengthening.
Your examples eupáto:r (secondary), dysáno:r (secondary), áfro:n, daímo:n,
génos fall into this category. Perhaps dó:to:r too, if *déh3-tor-.
* The lengthening did not take place if the preceding syllable was long
(-V:C) or heavy (-VCC), as in *wódr. (see above), (**pú:nt-ax-s >)
*pónto:h2s, Acc. (**pú:nt-ax-m >) **pónt&2m, Gen. (**pu:ntáx-âs >)
*pn.th2ós, or (**kárx-san >) *kér&2sr, (**karx-sán-a:s >) *kr.h2snós.
In the non-neuters, posttonic /o:/ (from the Szemerényi-lengthening of an
unstressed schwa) here alternates with zero in the accusative (where no
lengthening *-s follows), so that's another type of */o/ < **/&:/. If
dó:to:r is *dóh3-tor, it falls into this category.
* Another case of lengthening resulting in /o/ or /ó/ is found in the
thematic vowel, which is /e/ before voiceless consonants (or end of word),
and /o/ (< **a:) before voiced consonants. The classic example is the
thematic imperfect *-om, *-es, *-et, *-omes, *-ete, *-ont.
* The /ó/ in the perfect singular is another case of lengthening. In
Hittite, which has no reduplication, it alternates with /e/ in the plural
(of the hi-conjugation), as expected. The alternation /o/ ~ zero in the
PIE perfect is harder to explain, but perhaps it's due to the reduplication
(the reduplication syllable has the /e/-grade).
* Finally, there is the *R-prefix/infix. It accounts for the *o in the
causative, and in thematic forms like *bhorós ~ *bhóros. I do not totally
exclude possible different origins for at least part of the thematic
formations, such as a conceivable ancient vrddhi-formation **bhar- >
**bhá:r- + -a- > *bhor-o-, besides younger vrddhi *bhér- > *bhe:r- + -o-,
but certainly for those forms in *-mn-ós > *-n-ós ~ *-m-ós and laryngeal
loss, Jens has made a convincing case that they represent the infix.
Jens can explain better than I can the details of these cases of o-grade.
All I can add as a personal opinion is that I think the lengthened /i:/
(before CV-, not CCV-) in the reduplication syllable of the Sankrit
reduplicated (i.e. causative) aorist is also an effect of the R-infix
(behaving here more like, but not quite like a standard laryngeal).
>(c) Why should the initial combination gn- be a problem in nouns (leading to
>o-gn) when it is not in verbs?
Refresh my memory. Which noun had o-gn-?
=======================
Miguel Carrasquer Vidal
mcv@...