Re: Questions about Hungarian

On 02 May 2004 John Cowan <cowan@...> wrote:

> Finnish is interesting here:  8 and 9 are "two less than *deksan*"
> and "one less than *deksan*", but "deksan" (which is obviously IE)
> is not the Finnish for 10.
  Finnish '1' is _yksi_ in nominative, but it's genitive stem is
_yhde-_ and the same is for '2' _kaksi / kahde-_. Therefore the
_de_ part of _deksan_ belongs to the stem of '1' and '2' (the
voicing is due to the old gradation process). Thus we have an
ending _-ksan_.
  And if we study the Finnish case system, we find that _-ksi_ is
the suffix for translative. It should had an older form
_*-ksa/ksä_. Note that the Finnish translative compares to the
instrumental case suffix of other FU languages. The last _-n_ seems
to be an inetymological development due to the ending of
_seitsemän_ '6' and _kymmenen_ '10'.
  Therefore _kahdeksan_ and _yhdeksan_ originally meant '[we must
complement this number] with 1/2 [to reach 10]'. Probably the '10'
was dropped in Proto-Baltic-Finnic (PF).

  But back to the problematics of the FU '10'. Actually Hungarian
has three of it: (1) the ordinal _tíz_; (2) the ending _-(n)c_ in
_nyolc_ '8', _kilenc_ '9' and _harminc_ '30; (3) the ending
_-van/ven_ in _negyven_ '40', ..., _kilencven_ '90'.

  Hungarian _tíz_ is a cognate of Permic _das_ '10' and it's a
common Hungarian-Permic borrowing from Aryan _das'a_ (cf. Greek
deka). Ending _-(n)c_ also compares with Permic /m1s/ ending in
Zyrienish /k3kjam's/ '8' (< /k3k/ '2') and /3km's/ '9' (< /3t'ik/
'1'), therefore this old word *m1s' it should have meant 'ten' in
the time of Hungarian-Permic cohabitation. The suffix _-van/ven_
(and its cognates Vogulish _-man/pan_, Permic _-m1n_) is connected
with Finnish _moni_ 'many' (it's an early PFU borrowing from IE,
cf. English 'many').
  And there're another two FU roots for '10': one in Baltic-Finnic
(and Mordvin) as _*kümmen_, and one in Ob-Ugric--Lappish--Cheremiss
as _*loB-/loG-_.  Therefore the word for '10' is changed many times
in FU languages. The _*loB-/loG-_ seems to be the "original"
number, but it was replaced by the temporarily cohabiting Hungarian-
Permic group with two etymons of unknown origin (probably from an
extint forest Turkic or Palaeo-Siberian language), and a new
development is found in Baltic-Finnic group after the separation
(or before the joining) of the Lappish tribes. [The number for '10'
in Selkup-Samoyedic is /k2t/ and if we add FPU _*mone_ 'many' to
this with harmonic assimilation, *probably* we found the origin of
Baltic-Finnic _*kümmen_.]

  Thus the structure '1/2-minus-10' is constant in Uralic
languages, only the elements were changed when a base number was
replaced, or the parts were freezed in a temporal state as it's in
case of Hungarian.