FOM: my non-constructive proof

Ayan Mahalanobis ayanm at
Sat Jun 10 05:09:28 EDT 2000

From: Martin Davis <martin at>
To: Fred Richman <richman at>; FOM <fom at>
Sent: Friday, June 09, 2000 11:37 AM
Subject: Re: FOM: my non-constructive proof

> Of course the construction of an example provides additional information;
> one needn't be a constructivist to enjoy constructive proofs.

I seem to disagree with this sweeping statement. For example one can
"construct" a non-measurable set or Banach-Taraski paradox, (one sphere
transformed into two spheres) assuming axiom of choice. I doubt if these
examples provide any additional information. Do you think these are
constructive proofs? If not (well, I am assuming here) then we need to draw
some line between the "actual" constructive proof and "constructive" proof.
How do you think to achieve that?

> "challenge" has to do with whether a constructivist would really doubt or
> deny the truth of my conclusion.

Technically he can't deny the truth of the conclusion but of course he can
doubt. Suppose one proves a theorem for the abelian group, further suppose
there is no known counterexample to that for non-abelian groups. How will
one see the theorem in context of groups? I guess, constructive
mathematicians look at a classical theorem in the same way, unless he can
find some weak counterexample to that.


Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.

More information about the FOM mailing list