On Sat, 23 Feb 2008 03:41:24 -0600, "Patrick Ryan"
<
proto-language@...> wrote:
>One fact that I think should be held in mind when discussing the quality of
>the vowel of the reduplication syllable in PIE is that there were two
>patterns:
>
>1) perfect reduplication: *C'V-CVC
>
>2) intensive reduplication: *CV-C'V
>
>These two patterns may not have always been properly compartmentalized.
LIV recognizes the following reduplicating formations:
(1g) e-redupliziertes athematisches Präsens
(*dhé-dhoh1-/*dhé-dhh1- : C1é-R(o)/R(z)-)
(1h) i-redupliziertes athematisches Präsens
(*sti-stéh2-/*sti-sth2-' : C1i-R(é)/R(z)-)
(1i) i-redupliziertes thematisches Präsens (*si-sd-é/ó- :
C1i-R(z)-é-)
(2c) redupliziert-thematischer Aorist (*wé-wkW-e/o- :
C1é-R(z)-e-)
(3a) redupliziertes Perfekt (*bhe-bhóidh-/*bhe-bhidh-' :
C1e-R(ó)/R(z)-)
(5b) redupliziertes Desiderativ (*wi-wn.-sé/ó- :
C1i-R(z)-sé-)
(6a) redupliziertes Intensivum (*kWér-kWor-/*kWér-kWr-
(C1éC2-R(o)/R(z)-)
Category (1g) is, I believe, a phantom. Most of these verbs
are simply o-grade verbs, without a trace of reduplication,
which Jasanoff connects with the Hittite hi-conjugation
(o/e-Abalut). A case in point is the first such verb listed
in LIV, *bhedhh2-, where *bhé-bhodhh2-/*bhé-bhdhh2- is given
a the pre-form of unreduplicated Hitt. paddai, Latin fodio:
and Slavic bodoN, bosti. In Lithuanian bedù, bèsti, the
e-grade of the dual/plural was generalized, so that gets
assigned to plain thematic *bhédh2-e-.
The splitting up of reduplicating presents in LIV into three
categories also covers up some interesting variations in the
quality of the reduplicated vowel and/or the stress in the
present forms, as pointed out by Rasmussen ("Indo-European
-i- ~ -e-/-o-"). We have 3sg. juhó:ti vs. 3pl. júhvati
(*g^hú-g^hw-n.ti < *g^hé-g^hw-n.ti) and 3sg. sís.akti vs.
3pl. sás'cati (*si-sékW-ti vs. *sé-skW-n.ti).
Category (1i) certainly exists, but why?
Category (2c) in LIV is a simplification. In Vedic, where
this category is is perhaps best attested, the redupliaction
vowel is /i:/ [c.q. /u:/] (except before CC- in the root
when the vowel is shortened). Not all verbs show zero grade,
and not all forms are thematic.
The perfect (category 3a) incidentally also shows forms with
a long reduplication vowel /a:/ in Vedic. This happens in
more than 30 Vedic verbs (e.g. ca:-kan-, ca:-kl.p-, tu:-tu-,
etc.), only a few of which can be traced to verbs starting
with HC- (I can find 9 examples in LIV: *Hmelk^-, *HweRdh-,
*h1ger-, *h1lenghW-, *h2leks-, *h2merg^-, *h2weks-, *h2werg-
and *h2wers-).
=======================
Miguel Carrasquer Vidal
miguelc@...