Joel Bradford Klammer:

I heard from a friend that Marilyn vos Savant (chick with ludicrous IQ) attempted to refute the proof of Fermat's Last Theorem and failed miserably, and that, furthermore, there was discussion of this in sci.math. Is there a sci.math archive somewhere where I can read about this? Or can anyone just fill me in? Much thanks.

David T. Ose:

Savant wrote a book in the summer of 1993, about Fermat's last theorem and Wiles work (I have forgotten the exact title, sorry). There is a review of her book in the most recent copy of The American Mathematical Monthly. Interesting and entertaining reading. The reviewers say that, while she does a very credible job explaining some of the ideas of modern mathematics, most of her book's criticism of Wiles and of modern mathematics result from her own lack of understanding of both mathematics and mathematicians. And her deceitful use of several prominent mathematicians' names to give credibility to her book was highly unethical at best, say the reviewers. But check out the review.

Tim Chow:

Yes, that should give you a representative view. There is one thing that I would like to add, though. Vos Savant's most crucial argument is that "if we reject a hyperbolic method of squaring the circle, then we should also reject a hyperbolic method of proving Fermat's Last Theorem." Most mathematicians are content to dismiss this as nonsense; I prefer to regard it as an argument that *would* be valid except that it's founded on a fundamental misconception.

More precisely, vos Savant appears to regard Euclidean geometry and hyperbolic geometry as ALTERNATIVE AXIOMATIC SYSTEMS FOR *ALL* OF MATHEMATICS. If we accept this (false) premise, her argument runs approximately as follows. The problem of squaring the circle calls for a proof of a certain statement from the axioms of Euclidean geometry. Bolyai found a proof from the axioms of hyperbolic geometry, but this doesn't count. What Wiles has done is to find a proof of Fermat's Last Theorem from the axioms of hyperbolic geometry, but this doesn't count, because the original problem calls for a proof from the axioms of Euclidean geometry.

This argument is shot through with misconceptions, but it does have a certain logic to it. I've thrashed this out on sci.math before, and not everyone agrees with my interpretation of vos Savant, but I'm convinced that the above is what she was thinking, and it's helped me understand where she's coming from.

David T. Ose:

I would contest your statement that Wiles found a "proof of Fermat's Last Theorem from the axioms of hyperbolic geometry." The Shimura- Tanyama-Weil conjectures have little to do with hyperbolic geometry, although they can be written in the language of hyperbolic geometry, they are not dependent upon them. This was Vos Savant's biggest misunderstanding.

Dik T. Winter:

What I understand of Wiles' proof is that it is even not geometric. (I do not have a copy, but there was somebody who gave an introduction about it, so I have a minimal understanding.) Perhaps she was misled by the use of the term elliptic curves (that has nothing to do with elliptic geometry)?

RKinnamon:

I doubt it if Marilyn was "misled" . . . It is still uncertain that Wiles proof will hold up. I have vos Savant's book and all she says is that because of the methods Wiles used there may be errors and it may take a while to verify. And remember, she wrote the book BEFORE Wiles admitted his error a year ago December.

Aaron Bergman:

She said that [Wiles's proof] was wrong because he used some hyperbolic geometry, which shows a complete non-understanding of how the proof works, or even how axiomatic systems are used. And that error [found by Wiles and later corrected] was completely unrelated to anything she said.

Benjamin J. Tilly:

I have read the book. That book is a joke. If you check, you will find that she argues that since:

i) You cannot square the circle, and
ii) In hyperbolic geometry you can square the circle,
that
iii) hyperbolic geometry cannot be valid.

Thereby "overturning" over a century of well-established mathematics. (The error is that the theorem involved in (i) only related to EUCLIDEAN geometry, and says nothing about what you can or cannot do with NON-Euclidean geometry. The situation is like my telling you something about my dog Bob at one time, then something about my friend Bob at another, and then having you conclude that I have lied to you since the two comments do not jive with each other.)

In fact she goes on to argue that there is a possibility of disproving Einstein's theory of general relativity! All of which is complete hogwash. Oh, and she has yet to admit any mistake, although a variety of people have written to her, the mathematicians that she referred to having disputed her comments, and a number of mathematical publications (the latest being Math Monthly) having published highly critical articles.


Return to Menu
Go On to Part 2: Walt Whitman Weighs In

Сайт управляется системой uCoz