"centum"/"satem" "exceptions" [was Re: Northwest IE attributes]

[Miguel Carrasquer Vidal]

Stanley Friesen <sarima at friesen.net> wrote:

>At 01:53 AM 1/23/00 +0000, Miguel Carrasquer Vidal wrote:

>>**n'akut-.  The problem is the fate of stressed **i and **u, for
>>which we can hypothesize spontaneous diphthongization to *ei, *eu
>>(unlikely, I'd say, but an interesting possibility to account for
>>possible Pre-PIE long *i: and *u:), or loss as in the case of the
>>Slavic jers (with, as in Slavic, occasional retention to avoid
>>excessive consonant clusters, e.g. **CiC > *C^C, but **CiCC >
>>*C^eCC, likewise for **CuC > *CwC, **CuCC > *CweCC).

>Why is it necessary to go this way?  IMHO, there are sufficient instances
>of 'i' and 'u' in PIE the do *not* alternate with ablaut variants such as
>'eu' and 'ei' to suggest the inheritance of those vowels from the Pre-PIE
>stage, at least in some environments.

I'm not so sure.  There are certainly "loose" *i's and *u's among
the pronouns (e.g. *tu/*tu:), in affixes like *-i (dat/loc,
"present tense"), but IMHO anomalously few among common
nouns/adjectives and verbs.  For instance, I don't think
Benveniste's root-theory allows *i and *u in the V position (or
does it?).  It would be nice to have statistics.

Surely the existence of *kw, *k^ etc. suggests that some high
vowels were lost, passing their front- or backness to the
adjacent consonants (is there another explanation?).  Coupled
with the comparative rarity of non-zero grade *i and *u, I think
it's clear that *something* happened to the high vowels at some
stage of Pre-PIE.  The question is exactly what.  In which
environments (phonological or morphological) did *i and *u
survive as such?  What happened when they didn't?
In itself, the loss/decimation/weakening of *i and *u is not a
strange phenomenon.  Besides the Slavic case, there is also
Tocharian, where they merge [also *e] as <d> (oops, that's
a-umlaut), and some accounts of Afro-Asiatic vocalism (such as it
is) also imply *i, *u > *@ (or viceversa!).

On the subject of non-velar labialized/palatalized consonants, I
was wondering: since in Greek *pj > pt, could not such old
chestnuts as <ptolis>, <ptolemos> be derived from palatalized *p^
(*p^lH-).  I know Baltic <pilis> "city" is in itself no
supportive evidence (-il- [-ir-] is the normal Baltic
development, even tough Baltic and Slavic offer anomalous cases
of *ul, *ur which might be worth investigating) and Skt. pu:r-
might be seen as counterevidence (but p- is a labial after all),
but I still would regard *p^lH- as a neater solution than e.g.
Beekes' *tplH- (CIEL, p. 190).

Another thing to look at might be cases of Gmc. /i/ for expected
/e/ and viceversa.  *<fiskaz> "fish" because of *-sk^-?

=======================
Miguel Carrasquer Vidal
mcv at wxs.nl