Re: Evidence for Nostratic? (was Re: Proto-Uralic?)

[Hopefully this hasnt been answered yet...*grumbles about sporadic internet
hours*]

From: "Rob Haden" <magwich78@...>
> >The distinction between s-stems and thematic stems survived quite well;
> >they still were distinct classes in Classical Latin at least 3000 years
> >after the breakup of PIE (e.g. corpus, gen. corporis < *corpos-is).
>
> But you also have Latin genus (earlier *genos), gen. generis < *genes-is.
> Why wasn't it *genoris?
Mainly because regular Latin sound changes in medial syllables reduce short
vowels before /r/ to /e/.  Secondarily because the stem should be in *corpes-,
not _corpos-_ to begin with--_corporis_ is the anomaly, not _generis_.

Apparently what it was is all short vowels in medial syllables reduce to a weak
vowel that became /e/ before double consonants and /r/, /o/ then /u/ before /w/
or dark /l/, and /i/ before single consonants and /N/.

The exceptions to this being that original /i/ remained /i/ before double
consonants, and, more relevantly, that /o/ > /i/ is nearly always restored to
/o/ when analogy permits it.  The Latin -oris-type stem are an anomaly caused by
this, as the normal stem is -es- (seen in Greek, Sanskrit, Slavic, Hittite, and
Germanic):

    << Neut s-stems in -er-, like _genus, generis_, reflect *-es-;
       s-stems in -or- like _temporis_ to _tempus_  are secondary
       hypercorrect forms, and provide clear (if indirect) evidence
       of a period when /e/ and /o/ were in alternation in accord with
       [the rules above].  That is, a creation like _tempora_ is only
       possible if Romans could think that the stem-vowel -e- of
       intermediate *tempera was a weakened form of the vowel seen in
       the nom. *tempos. >>
           --New Comparative Grammar of Greek and Latin, §67.

    *Muke!
--
http://www.frath.net/