--- In
cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@...>
wrote:
> Hypothetical example: what is the "true vowel" of a Semitic root?
The
> question is meaningless since Semitic ablauts. Now why do we say
that
> that majority of verbs in IE have /e/ as stem vowel, when it can
only
> be derived for the present stem (etc)? Should we rather say: it
seems
> PIE ablauted?
> That's why personally I prefer to use Møller's notation *C-C-
(etc.)
> for standard *CeC-. Or I should be loyal to the proposal I once
made
> and write -A- for "the ablaut vowel" (which is different
> from "real /a/", which mostly entered the language in loans), thus
> *CAC-.
Exactly! Why didn't we say it so clearly much earlier? If Semitic
really has regularized its morphology completely as far as the
vowels are concerned, as I quoted Grande and in part Diakonoff to
say, and as it's being confirmed here, then Semitic is a language
without vowels in its lexicon, but with a variety (not many, but
still a variety) of vowels in its morphology. Its lexical vowel
system would then not just be of an extreme paucity as in IE, but
even more extreme by being completely empty.
I am not aware that the vowels of Semitic morphology can be
explained by phonetic rules from one single underlying vowel
phoneme, let alone from zero, but that may be just due to lack of
knowledge on my part. I have seen and once tried to read
Kurylowicz's book L'apophonie en sémitique, and the remaining
impression has been that the actual vowels of Semitic morphology
were not derived by anything resembling real phonetic laws. That
makes Semitic a language (group) with no vowels in it lexicon, and a
mysterious but standardized vocalism in its morphology.
For IE we have seen that it *is* widely possible to derive the
vowels of different morphological classes and their variations by
means of blindly operating phonetic changes. The unexplained residue
is now so small that there can be no reasonable doubt that the
vocalism is indeed rule-governed. For most words - not only verbs -
one can derive the actual forms from structures of great regularity
which have no other vowels than /e/. One can have different opinions
about the rest.
I can see two main reasons for the difference between IE and Semitic
which are apparently both true. On one hand, Semitic morphology has
completed its job by carrying through analogical changes without
significant residue. On the other hand the actual phonetic changes
that befell the vowels in Semitic have not been found yet. The
reason for the second part would be of course that analogy has been
carried to extremes so that one cannot see where a process of change
began and what caused it. Still, since the morphology also operates
with consonantal markers, it should be expected that it is possible
to explain at least some morphological vowels by their correlation
to consonantal material. But that is for others, I'm happy to say.
Jens