Re: Etruscan -na (was Re: [tied] Affects of ... etc.)

In a message dated 13/08/01 17:13:06 GMT Daylight Time,
glengordon01@... writes:

> I have strict sound *rules* that reach back into previous stages of
> IE that dictate to me what few possibilities exist in Mid or Old
> IE. In fact, these laws have been brought about by the _internal
> reconstruction_ you mention, by the unraveling of irregular systems
> into regular ones.

...

> With *-ame and *-ane to choose from as the only possibilities, *-ane
> is the more Occam-compliant solution since, as part of a larger
> MidIE declensional system with 3 genitives (*-ese, *ele and *-ane),
> accusative (*-m), ablative (*-eta) and locative particles
> (*dei, *bei), everything matches perfectly with Etruscan and
> demonstrates regular sound correspondances.

The value of doing internal reconstruction BEFORE doing comparative
reconstruction, as you know, is to make sure that we are dealing with
the oldest forms in the language to be examined. This involves study
of Etruscan *texts* in a historical context, not just looking at dead
glossaries, which as you and I both know are of variable quality,
especially on the internet.

I think these should not be recommended to people without a health
warning - it is no more than a few weeks ago that you and I both
expressed reservations about one of the better ones. (BTW, the word
/as/ does exist in Etruscan, and it and its derivatives are attested
more often than some words that are listed in glossaries. It's just
that, like thousands of other Etruscan words, and unlike /ais/, nobody
knows what /as/ means).

People should be aware that much of the 'evidence' that gets cited in
discussions relating to the affinities of Etruscan is based on words
that occur a handful of times, in contradictory forms, and the meaning
they are glossed with may or may not be correct. A similar problem
applies to morphological and other grammatical features.

No criticism of you, Glen, is implied in any of the above. Now on with
the quibbles.

If we look at late Etruscan, we find two "genitive" endings, /-s/ and
/-l/ (to simplify what actually is the case). This is an irregularity.
Examination of this problem leads us to the conclusion that the
endings were previously distinguished by having different functions,
because archaic Etruscan texts and Raetic texts both show /-ale/ being
used as a dative and not a genitive. It is not valid to link the late
Etruscan "genitive" /-l/ with a reconstructed genitive /*-ele/ in IE
without taking the apparently intervening *dative* stage in archaic
Etruscan into account, and somehow, explaining it. But, of course,
there may well be a perfectly good explanation for the problem fact.

That's a very minor point, of course. Don't get me wrong, you make a
good case for the endings /-s/ and /-l/ (in Lydian) being inherited
from PIE by regular sound correspondences. Etruscan sharing these
features therefore makes one justifiably suspect a connection between
Etruscan and IE.

Internal reconstruction deals "exclusively with the forms existing in
a single language, with no comparison involving forms from other
languages or dialects" (Trask). Forgive my scepticism, would you have
come up with a past tense ending of /-k:e/ for pre-IE in 8000BCE
without trying to make Etruscan part of the equation? If you did
reconstruct this initially from IE, can you take me through your
reasoning? Also, it seems strange that Etruscan should meanwhile
retain this ending for 8000 years.

> No, I'm assuming the more economical solution - that the *entire*
> declensional paradigm is inherited.

At the end of the day, comparative linguistics can only be done on
languages that are related. So, humour me, explain why:

a) Core vocabulary in Etruscan is not IE. Explain the numbers, and
numerical morphology, for instance.

b) Etruscan appears to have an agglutinative typology,

c) and does not inflect verbs for person or number.

If you do that, I promise to shut up.

> A completely absurd statement that ignores true relationships
> governed by rules and sound correspondances in favour of something
> hodgepodge.

While reconstructed languages come neatly pressed and in a plastic bag
marked "Occam's Laundry", real ones don't, they have buttons and
sleeves missing and have stains and patches. (With apologies to
whoever it is I am misquoting).

Ed. Robertson