Re: [tied] Morphology (13/20)

From: Glen Gordon
Message: 15069
Date: 2002-09-04

>At the most remote stage, pre-PIE may have been an ergative language, with
>a triptotic case marking ergative
>**-u, absolutive **-a, and genitive **-i. [...blah, blah, blah...]
>A collective, similar to the Semitic "broken plural", [...blah, blah,

Of course, Miguel is blindly patterning all of his
PreIE reconstruction on Semitic as if Semitic were
the likeliest language to be closely related to IE.
Uralic would be a sane man's choice but he chooses
Semitic. And isn't it funny that Semitic would go
unchanged while IE made leaps and bounds. "Triptotic"
is a pretty word but there is no basis for any of these
case markings for any stage of IE and a triple asterisk
would seem more justified than merely two.

Plus, how many times must we tell Miguel that the dual
developed late in IE? He strongly desires to reconstruct
it for PreIE and yet the fact remains that the dual
obviously wasn't even fully worked out by the time IE
split up, otherwise we would surely see consistency in
the dual paradigms. Misguided would be a kind word here.

So after endless paradigm examples, Miguel succesfully
makes PreIE look like quantum mechanics. Congrats! To
bad it doesn't explain anything.

And now for some sobriety...

The MIE paradigm must have consisted of at least the

singular plural
nominative NULL *-es
accusative *-em *-emes
genitive *-se
partitive *-te

Usage and origin of cases
Aside from the endingless nominative and the accusative
in *-m, nothing of the case system derives from a highly
remote ergative stage of Pre-IE. The other case endings
are affixed postpositions that attached themselves to
noun stems at a time when accusativity was in full swing.
In fact, because the case system is largely unrelatable
to Uralic or other related language groups, it's definite
that the non-nominoaccusative case endings were decided
upon only in IndoTyrrhenian, since only Tyrrhenian
languages like Etruscan share any non-nominoaccusative
case endings (like the *s-genitive) with IE. Before this,
they had to have been postpositions.

In MIE, the nominative case was used for more than just
the subject. It was also used for the vocative, and in
conjunction with postpositions, the dative and locative
as well.

The accusative case was used for definite objects.

The genitive was used to convey belonging as well as an
ablative sense of coming from something.

The partitive (later used as the "ablative") would have
been used to convey an indefinite object and to denote
something that is part of something larger.

Examining the PIE declensions properly, we see that the
development of a plural, let alone a dual, paradigm beyond
the nominative and accusative must have come late. In
other words, the usage of the plural appears to have been
restricted to subjects and direct objects from early on.

We notice that the Late IE plural is certainly marked by
*-es in the nominative which is probably ancient and
related to plurals in *-t found in Uralic and EskimoAleut
languages, just as the second person singular *-s(i) is
surely related to its equivalent pronoun *tu. The plural
marker has also been apparently secondarily strengthened
and made guna grade even though we should expect zero grade
for an unaccented suffix. (The reason for this is lucid:
It was to avoid merger between the plural in expected **-s
and the singular *-s, derived from the demonstrative *so-.)

The accusative plural *-ns consists of *-m (accusative
singular) plus the plural *-s (the expected and now
substantiated zero-grade form theorized above), as anyone
can see.

The genitive plural *-om (an entirely different suffix from
that of the singular in *-s, *-os or *-esyo) and the
plurals of other cases are merely a mishmash, loosely
created to fill in gaps that must have once existed. They
could not possibly be ancient and any attempt to fit such
endings into any Pre-IE theory is baseless.

So it's as simple as that. What else is there to know?
After that there are minor details...

Paradigmatic qualitative ablaut in thematic vowels
Of course, we've discussed the reasons for qualitative
ablaut in the thematic vowel of thematic verbs and nouns
such as *ekwo- before. We find *ekwos (nom.) but
*ekwesyo (gen.) and one wonders what the vowel terminating
the stem *ekwo- might have originally been to account for
this alternation. Well wonder no more because it was a

The schwa underlies the *e/*o ablaut and naturally at one
time, the paradigm was more regular:

early Late IE PIE
*ekw&:s *ekwos
*ekw&:m *ekwom
*ekw&syo *ekwesyo

The reason for the sporadic lengthening of schwa largely
centers around the same phenomenon as we find in English
("writ" vs. "rid"). The presence of voice in a consonant
causes lengthening of a preceding vowel. So, after the
formation of long and short schwas in early Late IE, the
short schwas were fronted to *e while the long ones were
reassigned to the mid-back position of *o.

Naturally, the only thing left to figure out is the
conundrum concerning the unexpected vocalism of *o in the
nominative that one would expect to have become **e because
of the voicelessness of nominative *-s. While Miguel opts
for creating more complexity and positing an otherwise
unwarranted phoneme *z, the unexpected lengthening of the
nominative should be trivial given the lengthening seen in
other athematic stems such as *kwo:ns (stem *kwon-). Is
it any small wonder then that expected *ekw&s would follow
the same pattern, producing *ekw&:s (and thus later

Problem totally solved without complexity.

Paradigmatic qualitative ablaut in stems
There are other examples of ablaut that have nothing to
do with the lengthening of early Late IE schwa such as
the example of *po:ts (nom.) and *pedos (gen). The reason
for this particular stem ablaut, as opposed to thematic
ablaut, lies in a simple MIE rule: Only schwa was allowed
as the vowel in unaccented syllables. As such, when
a syllable in Mid IE *a (> *o) was unaccented, it was
reduced to schwa. While schwa normally disappears in
Late IE, schwa was strengthened to *e to avoid an
assyllabic stem within the paradigm (ie: **pd-os). The
application of this paradigmatic exception is predictable
and a necessary corollary to the general rule:

Mid IE *& > Late IE NULL.

- gLeN

Chat with friends online, try MSN Messenger: