WM
unread,Feb 5, 2012, 3:35:28 AM2/5/12You do not have permission to delete messages in this group
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to
Das Kalenderblatt 120206
NN: Lieber Herr Mueckenheim,
Cantors Methode versagt fuer keine Auflistung von reellen Zahlen, wie
der einfache diagonale Beweis zeigt: Zu jeder Liste reeller Zahlen
kann man eine relle Zahl finden, die in der List nicht vorkommt.
Folglich kann es keine Bijektion zwischen den natuerlichen und den
rationalen Zahlen geben.
Es gibt immer wieder Einwaende gegen den Cantor Beweis, und ich
erhalte ab und zu 20-seitige Abhandlungen darueber, die zum Teil dann
auch noch die Heisenbergsche Unschaerferelation mit ins Spiel bringen.
Ich kann davon ruhigen Gewissens abraten.
Es hat sich bewaehrt, einfach auf den einfachen Beweis von Cantor und
seine Aussage zu verweisen, und zurueckzufragen: Wo ist der Fehler in
diesem Beweis? Und: Eine gegebene Liste von reellen Zahlen ist eine
gegebene Liste. Wenn Sie sie irgenwie veraendern, erhalten Sie am Ende
wieder eine gegebene Liste, fuer die das Diagonalargument greift.
Natuerlich koennen Sie bezweifeln, dass es eine fertige Menge aller
rellen Zahlen gibt. Niemand kann in ZFC die Konsistenz von ZFC
beweisen, vorausgesetzt, dass ZFC widerspruchsfrei ist.
Mit freundlichen Grüßen, NN
Diese Korrespondenz stammt aus 2003. Damals war ich leider noch nicht
in der Lage, meine Einwände präzise genug zu formulieren. Heute würde
ich antworten: Ein Widerspruchsbeweis wie der von Cantor muss in allen
Fällen funktionieren. Betrachtet man aber eine spezielle Cantor-Liste
wie die folgende
0,0
0,1
0,11
0,111
...
so erhält man bei Ersetzung von 0 durch 1 die Zahl 1/9 = 0,111, die
entweder konstruiert wird, indem 0,1 unendlich oft verlängert wird
(dann steht die "Diagonalzahl" d bereits in der Liste), oder die nicht
konstruierbar ist, weil keine aktual unendliche Menge von natürlichen
Exponenten existiert (dann unterscheidet sich die Diagonalzahl d nicht
von allen Listeneinträgen z_k. Die Annahme jedenfalls, dass die Folge
der Zeilenzahlen
(z_k) mit z_k = 10^-1 + 10^-2 + ... + 10^-k
ihren Grenzwert nicht enthält, während die Folge (d_k) der endlichen
Anfangsabschnitte
d_k = z_k
der Diagonalzahl d ihren Grenzwert 1/9 annimmt, ist eine zutiefst
unmathematische und würde wohl von keinem Mathematiker akzeptiert -
wüsste er nicht, dass dieser unmathematische Unfug die unabdingbare
Basis für das Funktionieren des Cantorschen Diagonalargumentes ist.
TJ: I was looking in the 2008 proceedings of a conference MACAS2 on
mathematics education. Wolfgang Mückenheim (WM) contributes with a
paper featuring his argument about the infinite binary tree, the one
which uses as an unproven premise that the union of the sets of paths
of finite subtrees is equal to the set of paths in the infinite tree.
I have written to WM and to the editors suggesting the need for
producing an erratum, which is obviously an appropriate action. Now,
in an erratum or a retraction, it is customary to give credit to the
discoverer of the counterexample or disproof.
Can anyone help with digging back into history and figuring out who
was first to point out to WM that the two sets of paths, the union of
sets of paths of finite subtrees, and the set of paths of the entire
tree, are actually not the same but (very) different. The name(s)
should be added to the erratum in an ackowledgment. I feel that I need
help with this, since I am unsure whether WM will be able to produce
the names by himself, and since it is difficult to find answers just
by searching back. I found some nice quotes dating from January 2007,
but it was in a marathon thread of 7873 posts, and I could not spend
time to dig back through the whole mess looking for earlier instances.
I suspect these are not the earliest yet.
I hope to get help also from the German groups.
WM: Frequently I have been asked why I had coined the word
matheology. Here I will explain it in an easily comprehensible way.
Men consist of molecules. Theologians recognize something that exists
independently and that remains even when the molecules have gone: the
soul. Mathelogians argue in a similar way. They recognize souls within
real numbers.
The complete infinite binary tree can be constructed using
countably many finite paths (each one connecting a node to the root
node), such that every node is there and no node is missing and every
finite path is there and no finite path is missing. I will call these
components of the tree the molecules of the tree. Matheologians now
recognize by far more in the tree, namely uncountably many paths of
infinite length. This recognition cannot be confirmed by looking at
the molecules of the tree. Therefore it must be something analogous to
the soul, which has been breathed into the purely mathematical
construction of molecules. My theological interpretation, however, is
not only based upon the close analogy between set theory and theology.
It is also suggested by the fact that the concept "soul of tree" has
been triggered by the work of a trusting anti-darwinist. But the
remarkable point is, that Cantor himself did not apply the soul-
concept. His diagonal argument of 1891 applied only "molecules",
namely digits (or more generally the terms W and M of a sequence)
existing at *finite* places.
[Tommy Jensen,"Help! with Infinite tree history", sci.math, 6. 12.
2009]
Gruß, WM