Re: [tied] Morphology (1/20)

From: Glen Gordon
Message: 14339
Date: 2002-08-15

>I will present here (in approximately 20 chapters) my current views on
>(Pre-)PIE morphology. Comments are, of
>course, welcome.

I notice that Miguel is slowly adopting some of my views. It
will only be a matter of time before he is fully assimilated.

We seem to agree in the order of certain processes. We
agree that the pronominal endings were originally formed by
affixing personal pronouns to the verb stem, although
I date this particular event to a time well in the past
(c.11,000 BCE) since other Nostratic languages such as
Uralic, EskimoAleut, Altaic and even Dravidian all share
this same feature. Thus it involves Proto-Eurasiatic more
than it does Old or Mid IndoEuropean whose pronominal
endings were long since fused to the verb stem.

We also agree that final *-t was then softened to *-c
(c.8000 BCE in Proto-IndoTyrrhenian) and then finally
to *-s (in Old IE), although you continue to insist on a
labial-obsessed and, as such, typologically unsound
version of this thinking: *-tw > *-cw > *-sw (???)

Finally, *-t to the 3ps and 3pp (quite clearly related to
the demonstrative *to-) followed by a later indicative
marker *-i must surely have been attached after the above
two events. Again we thankfully agree on these general
points, if not in the details of how this came to be.

This is where we strongly disagree:
First of all, the *-en-/*-on- seen in the 3pp was not
originally a plural marker at all. It was merely a pronominal
marker. It was only later in Mid IE, by analogy, treated as
a plural marker in pronominal endings. At that time, the
3ps was *-e (technically "unmarked" for thematic verbs
which had *-e- before all pronominal endings anyways) and
the 3pp plural was *-ene. Eventually, *-ene came to mark the
verbal plural _in general_, resulting in MIE variants of 1pp
*-mes and 2pp *-te, namley *-mene (> *-men) and *-tene
(> *-ten). This is why we see no trace whatsoever of *-en- being
used as a plural elsewhere. It's restricted to pronominal endings
where it had always been.

By the way, a likelier reason that *-te lacks the expected
plural marker *-s as we find in 1pp *-mes may lie in the
likelihood that the plural *-s comes from IndoTyrrhenian
*-c (and earlier *-t). One may notice that an original *-tit
might have been prone to be avoided if there was an avoidance
of a repetition of the same sound in the same syllable. In
fact, this is what we find since there is a strong avoidance
for syllables such as **nen-, **pep-, **ses-, etc. We may also
note that there is an alternation in neighbouring
Proto-Boreal (Uralic, Eskimo-Aleut) of a plural marker in
*-i and one in *-t. Thus, we might suspect that a similar thing
was happening in the 1pp and 2pp, once *-mit and *-ti (not **-tit)
in Proto-Steppe, becoming in IndoTyrrhenian *-mec (> *-mes) and
*-te (> *-te).

It's also clear that the dual formed at a late date, far
too late in fact to be of much concern for any previous
stages of Indo-European. Reconstructing a dual person in
previous layers is misguided.

The rest of Miguel's post consists of an exploration of
Indo-European's daughter languages and while interesting
it is irrelevant to the study of Pre-IE. Rather,
reconstructed PIE (which itself is based on these languages
already) should suffice unless there is reason to believe
that the accepted reconstruction is erroneous.

- gLeN

Send and receive Hotmail on your mobile device: