From: Arnaud Fournet
Message: 62208
Date: 2008-12-20
----- Original Message -----
From: <jdbengt@...>
To: <MTLR@yahoogroups.com>
>
> There is some confusion. We do not need sound correspondences in order to
> classify languages, i.e. formulate a model of a language family or
> macrofamily. The initial job of classification is done by the observation
> of
> similar grammatical and lexical morphemes. But after that the
> classification
> must be tested.
>
=========
I think we definitely need to make clean work on sound correspondences in
order to classify.
If we take the example of the so-called *me "P1" and *te "P2" system
which ultimately is the alpha and omega of the relationship between IE and
Uralic,
The situation is very far from being as clear as has been sold ever and ever
again these last years.
In Mordvin,
I know you is soda-te-n (n sometimes omitted)
you know me is soda-sa-ma-k
soda = know
te = P2 objet
-n = P1 subject
sa = present tense (an intrusive morpheme here)
ma = P1 objet
-k = P2 subject
On the surface, we have traces of *te and *me (= ma)
Now, there are a couple of questions,
there is a clear and consistent distinction between objet and subject,
in all the system for all persons,
as a matter of fact, P1 subject is always -n and P2 subject is -k (sometimes
replaced by -t)
It's worth noticing that Ungarian has en "P1"
The next hitch is -te- can only be from *-hte- not just *te
because in the intervocalic position, it should be voiced *soda-te-n should
have become *soda-de-n
as a matter of fact, when te or ta related to you-all 'P5', we have da/de
*soda-de-z they knew you-all.
The next indication is I knew you is soda-j-hte-n, precisely with -hte- (-j-
is the past morpheme).
In other words, the apparent coincidence of te in soda-te-n with PIE *te is
superficial
what we have is P5 *t-e/a and P2 *-ht-e/a
The form for P2 is longer
and this feature is true for P1 and P4 as well,
P1 when built with m- is always longer than P4 throughout URalic languages.
And the careful analysis of P3 versus P6 in Mordvin shows that P6 is -s-
(voiceable -z) but P3 is not voiceable -hs-
P3 is longer again than P6
the conclusion (my conclusion...) is that
P1 P2 P3 are derived from the plural forms
P5 te => P2 hte
P6 se => P3 hse
I have so far no way of proving that P1 was -hm-.
In other words the old system is Uralic
P1 -n- (PIE -m-)
P2 -k- (PIE -t-)
P4 -m- (PIE -nos-)
P5 -t- (PIE ywos)
The forms of the plural have invaded the forms of the singular in Uralic
and this trend has been reinforced by IE languages in contact with Uralic.
In other words, the pillar of the IE-URalic connection does not exist.
*m and *t for P1 and P2 in URalic is a myth resulting from easy-go-merry
analysis.
It's interesting to note that P1 *-n- is like many Amerind languages.
This is the place to look first for Uralic relatives.
A.
======