Re: [tied] Old Japanese Numbers (was: Why borrow 'seven'?)

From: Andy Howey
Message: 34267
Date: 2004-09-24

Perhaps this thread should be moved over to the Nostratic-L list.  This is well with the scope of that list.
 
Andy Howey 

petusek <petusek@...> wrote:

----- Original Message -----
From: "Richard Wordingham" <richard.wordingham@...>
To: <cybalist@yahoogroups.com>
Sent: Thursday, September 23, 2004 10:26 PM
Subject: [tied] Old Japanese Numbers (was: Why borrow 'seven'?)


>
> --- In cybalist@yahoogroups.com, enlil@... wrote:
> >
> > Exu:
> > > I have seen that. Kind of like organizing English as
> > >
> > > one
> > > two    three
> > > four    five
> > > six      seven
> > > eight   n-ine
> > > ten
> >
> > Not at all. Perhaps, considering that you goofed when it came
> to "shi",
> > you should listen for once. This is common knowledge and the vowel
> > harmony that Altaic languages in general are known for is evident
> in
> > the Old Japanese forms. The vowel alternations are predictable.
> None
> > of this pattern is random at all. To dismiss this is like
> dismissing
> > chemistry in favour of alchemy. Look again, Exu.
>
> Could you explain the systematic pattern, please.  I can't see one.
>
> Richard.

Let's list the Koguryo reconstructions offered by Starostin (in Altajskaja
problema i proisxozhdenie japonskogo jazyka, Moskva: Nauka):

(I'm using /?/ for schwa and /N/ for velar nasal here)

1    *pit?-
2    *puta-
3    *mi-
4    *d?-
5    *�t�-
6    *mu-
7    *n�n�-
8    *da-
9    *k?k?n?-
10    *t?w?-

In 1 & 2, /i/>/u/ and /?/>/a/
In 3 & 6, /i/>/u/
In 4 & 8, /?/>/a/

The regular change of /i/ to /u/ (numerals "1" & "2", "3" & "6") and /?/ to
/a/ (numerals "1" & "2", "4" & "8") is apparent. If one doesn't agree with
Starostin reconstructions, Old Japanese must be taken seriously (being
attested!):

1, 2    fit� >> futa (thus /i/>/u/ and /�/>/a/)
3, 6    mi >> mu (thus /i/>/u/, see above!)
4, 8    y� >> ya (thus /�/>/a/, see above!)

Let's exclude 5 & 10 for their incompatibility with the rule. As Glen
correctly mentions, Altaic languages are famous for their vowel harmonies.
Someone has claimed here that similarities among various numerals are often
of an alliterative character. The above, however, seems to be more than
that. To that extent I must disagree.

As for Altaic parallels (several, if right, excluding some of the above, I
could list a few possibilities (those who are interested in the most
promising set of regular correspondences among Altaic branches and the
reconstruction of the Proto-Altaic consonantism by Starostin and Vovin, send
me empty e-mails with "alt cons tab" in the subject field, and I'll send you
a table I'll have created in MS Excel :-)):

"1"    It is very tentative to compare P-Jap *pit? with Tk *bi^:r "1", Mo
*b�ri "all, each", MKor pi^r�s- "to begin"
"2"    *puta- is compared to Mkor pc^�k "pair" > mKor c^c^ak, cf. ipc^ak
"this side" (Ramstedt) and Tk *buc^uk "half"; it is fair to mention that
Miller (Japanese and other Altaic languages, 1971, Chicago & London, Chicago
University Press) speculates about unattested pJp *yuta- "2" > *puta under
the influence of *pit?- "1" and this hypothetical form would be compatible
with MKor turh and Tg *z^�w�r
"3"    There is no convincing etymology of OJap mi-/myi-. An evident cognate
is Koguryo *mi(l/r/t/) (Murayama/Lee/Miller) "3" (Miller's *mit would
correspond with Tk *�(:)c^ (different reconstr. by diff. auths.) "3". I
could also list Tungus examples, but they are too many, I'd rather make a
table of the some time and offer it to the Neugierig :-) It might be enough
that some people reconstruct P-Jap *n'u-
"4"    OJp y�- < pJp *d?- has been compared to Tg *duj-gin / Mo d�r-ben / Tk
*d�:rt (< PAlt *tu:r ~ *to:r or just *tV-)
"6"    OJp mu- might be compared to Tg *n'u-Nu-n "6" (Vovin's reconstr.)
"8"    No convincing etymology. Miller compares OJap ya- "8" (also
"several") with Tg z^abkun "8", but Blazek claims this numeral is an
innovation in Tungusic.

However, I'm a much fairer guy than someone might have ever expected, and
will list some of the extra-Altaic (and even extra-Nostratic) examples:

Other Nostratic:

"3" compared to Dravidian *mu:n_ "3" (tempting, uh?), originally perhaps
"protruding [finger]", Samoyed *n�kur etc.
"4" compared to Samoyed *tett?^ "4", FU *n'elj� "4" etc.
"6" compared to Samoyed *m?^ktut "6" (but it is etymbl. on the basis of
Samoyed *m?^ka� "back") etc.

Austronesian (=Aun.):

"1" Aun. *it'a? "1" (*p- = prefix) (Murayama, Kawamoto), Aun. *pi[t.]oN
"one-eyed" (Benedict)
"2" Aun. *pat'aN "pair" (Mur., Kaw.), OJp futa "2" < P-Tsouic *-pusa- "2
(years, nights etc.) & Ojp "20" < Aun. *pats123aN "pair" (Benedict both)
"4" OJp y�- < redupl. *y�y�- < Aun. (x2?)x2?pat "4"

Am I fair enough?

We have to decide which of the possibilities, I tried to list briefly above,
we shall take. As for me, vowel harmony is the easiest and most
straightforward solution to our problem, although I can admit their PAlt
origin and secondary analogical restructuring.

But folks, we must be careful, this really seems to be far beyond the edge
of what the topic of this list might have been intended for some
unimaginable time in its deepest, pastest Past! ;-) Just to remind you, if
it helps anybody, feel free to contact me, as far as the Altaic table is
concerned, ok?

Petusek