00 voturi pozitive00 voturi negative

9 vizualizări4 paginiAn introduction to Fermat's Last Theorem and its history, by Andrew Granville

Jul 30, 2014

© © All Rights Reserved

TXT, PDF, TXT sau citiți online pe Scribd

An introduction to Fermat's Last Theorem and its history, by Andrew Granville

© All Rights Reserved

9 vizualizări

00 voturi pozitive00 voturi negative

An introduction to Fermat's Last Theorem and its history, by Andrew Granville

© All Rights Reserved

Sunteți pe pagina 1din 4

\baselineskip=14pt

% History of Fermat's Last Theorem

% by Andrew Granville

% A version edited by Ian Katz appeared in The Guardian

% (U.K.), on June 24

It was announced yesterday that the most famous

question in pure mathematics, Fermat's Last Theorem,

has finally been answered. Englishman Andrew Wiles,

of Princeton University, a Fellow of the Royal Society,

told an academic audience at the

Isaac Newton Institute in Cambridge, including

leading authorities from around the world,

that he had finally concluded one of the most

controversial chapters in scientific history.

The story of Fermat's Last Theorem can be traced back

to ancient Greek times and stems from one of the best-known results in

mathematics, Pythagoras' Theorem. That is, in any triangle

which contains a right-angle, the square of the length of the

hypoteneuse is equal to the sum of the squares of the lengths

of the other two sides. (In mathematical notation,

$z^2 = x^2 + y^2$, where $z$ is the length of the hypotenuse

and $x$ and $y$ are the lengths of the other two sides.)

This rule was extremely useful to

the ancient Greeks because it allowed them

to construct an accurate right-angle. The idea was simply to

cut pieces of rope of lengths $x, y$ and $z$ feet, pull the ropes

taut (holding the ends together) and, voil\`a, a right angle.

For this to be practical they needed $x, y$ and $z$ to each be whole

numbers; and the example $5^2=3^2+4^2$ sufficed for their purposes.

Other examples that they recorded included $13^2=5^2+12^2$ and

even $8161^2=4961^2+6480^2$.

One of the great intellectual masterpieces of the ancient Greek

world was Diophantus' {\sl Arithmetic}. This work, available in Latin

translation in the seveteenth century, was an important

inspiration for the scientific renaissance of that period, read by

Fermat, Descartes, Newton and others.

Fermat, a jurist from Toulouse, studied mathematics as a hobby. He

didn't formally publish his work but rather disseminated his ideas

in letters, challenging others to match and/or admire his understanding.

Fermat was evidently inspired by the {\sl Arithmetic} and made many

notes in the margin of his copy. After Fermat's death, his son

Samuel published these notes and amongst them was the following

tantalizing sentence, beside the description of

Pythagoras' Theorem: ``...it is impossible for a cube to be

written as a sum of two cubes or a fourth power to be written

as a sum of two fourth powers or, in general, for any number

which is a power greater than the second to be written as

a sum of two like powers.

I have a truly marvelous demonstration of this proposition which this

margin is too narrow to contain''.

In mathematical notation, one cannot

find whole numbers $x,y,z$ and $n$, with $n$ bigger than $2$, for which

$z^n = x^n + y^n$. Whether Fermat was

being overly optimistic about his `demonstration', we

shall probably never know, but his argument has not been reproduced

in the intervening three and a half centuries, despite

no shortage of effort to do so.

His remarkably simple, seductively simple, assertion is known as

{\sl Fermat's Last Theorem}. Substantial prizes have been offered

for its solution (the largest of which, the Wolfskehl prize,

was rendered trivial by the

German hyper-inflation of the twenties). The great mathematicians of history

have been in two minds as to whether to work on it. Ernst Kummer,

the German mathematician of the last century who did so much to

establish modern algebra, wrote that Fermat's Last Theorem is

``more of a joke than a pinnacle of science'', yet his

most important work originated in his failed attempts to prove it !

Fermat's Last Theorem is the question that mathematicians love to hate:

it is really only one example of an equation, and then one that isn't so

relevant in a general study of equations. On the other hand, there is no

denying its charm and simplicity. As a question it goes in and out of

favour, sometimes being dismissed as `irrelevant' and `unimportant',

at other times hailed as the most interesting equation around.

Many people have tried to give an unassailable `proof' that Fermat's Last Theor

em is true. It is a favourite of professional and

amateurs alike: and, every few years, the newspapers

report yet another purported proof, which is subsequently

found to be lacking in some way.

So why is Professor Wiles's attempt

different ? And why are there many of us, who are usually extremely skeptical

of such claims, prepared to believe

that Wiles could have succeeded where so many others have failed ?

To understand, it helps to have some perspective of the recent history

of the problem: \ Up until the last decade the question seemed unassailable.

To be sure, many interesting approaches

have been proposed, persuading us of the truth of

Fermat's Last Theorem, but without providing a proof that can be checked

in every case. One such `proof' is known to work for every

exponent $n$ up to four million: that is, for every number $n$ up to four

million there are no

whole numbers $x, y$ and $z$ for which $z^n=x^n+y^n$ has solutions.

However this method of proof involves an individual computer verification

for each $n$, and so will not give the result for {\bf all} exponents $n$.

In 1983, a young German mathematician, Gerd Faltings,

proved a result far, far beyond what anyone had thought provable.

Instead of looking only at the equations $z^n=x^n+y^n$ in isolation,

he studied a very wide class of equations, the so-called `{\sl curves}', and

discovered exactly when they could and could not have infinitely many solutions

in whole numbers. In particular,

Faltings almost proved Fermat's Last Theorem; he showed

that for any given $n>2$ there are only a few, that is finitely many,

whole number solutions to $z^n=x^n+y^n$ (not allowing `scaling'):

however, it seems

unlikely that his methods can be modified to actually show that there

are {\sl no} solutions. Faltings' work revolutionized the way that

people thought about solving equations, leading to the award of a

Fields' Medal, the mathematical equivalent of the Nobel Prize.

Wiles' approach to Fermat's Last Theorem comes, however, from a somewhat

different direction, with rather different types of ideas. It all

began in 1955, with a question posed by the Japanese mathematician

Yutaka Taniyama:

\ Could one explain the properties of {\sl elliptic curves},

equations of the form $y^2=x^3+ax+b$ with $a$ and $b$ given whole numbers,

in terms of a few well-chosen curves. That is, is there some very special

class of equations that in some way encapsulate everything there is

to know about our elliptic curves ? Taniyama was fairly specific about

these very special curves (the so-called {\sl modular curves})

and in 1968, Andr\'e Weil, brother

of the philosopher Simone Weil, and himself one of the leading mathematicians

of the century, made explicit which modular curve should describe which

elliptic curve. In 1971 the first significant proven evidence in favour of

this abstract understanding of equations was given by Goro Shimura,

a Japanese mathematician at Princeton University, who

showed that it works for a very special class of equations.

This somewhat esoteric proposed approach to understanding elliptic curves,

is now known as the Shimura-Taniyama-Weil conjecture.

There the matter

stood until 1986 when another young German researcher, this

time Gerhard Frey from Saarbr\"ucken, made the most surprising and

innovative link

between this very abstract conjecture and Fermat's Last Theorem.

What he realized was that if $c^n=a^n+b^n$ then it seemed unlikely that

one could understand the equation $y^2=x(x-a^n)(x+b^n)$ in the way

proposed by Taniyama. It took deep and difficult reasoning by Jean-Pierre

Serre in Paris, and Ken Ribet in Berkeley, California, to strengthen

Frey's original concept to the point that a counterexample to

Fermat's Last Theorem would directly contradict

the Shimura-Taniyama-Weil conjecture.

This is the point where Wiles enters the picture. Wiles, already

established as one of the deepest pure mathematicians of

his generation, has drawn together a vast array of techniques

to attack this question. Motivated by extraordinary new methods of

Victor Kolyvagin of the Sieklov Institute in Moscow, and Barry Mazur of

Harvard Univeristy, Wiles has

succeeded in establishing the Shimura-Taniyama-Weil conjecture

for an important class of examples, including those relevant to

proving Fermat's Last Theorem. His work can be viewed as a

blend of arithmetic and geometry, and has its origins way back in

Diophantus' Arithmetic. However he

employs the latest ideas from a score of different fields, from

the theories of $L$-functions, group schemes, crystalline cohomology,

Galois representations, modular forms,

deformation theory, Gorenstein rings, Euler systems and many others.

He uses, in an essential way, concepts due to mathematicians from Britain,

France, Germany, Italy, Japan, the United States, Canada, Russia and

Colombia; the culmination of work by many people around the world thinking

about very different questions. Few suspected that their work might have

this kind of application; and there are perhaps no more

than half-a-dozen people in the world who are capable of fully

understanding all the details of what Wiles has done.

Given the enormous complexity of this work it will take some time

to be certain that every detail Wiles relies upon,

from such an array of areas, is correct. However, leading experts have now

examined the essential ideas and there can be no question

that his work is a profound contribution which sheds

light on many important questions in pure mathematics. It is

a tour de force, and will stand as one of the scientific achievements

of the century. His work is not to be seen in isolation, but

rather as the culmination of much recent thinking in many directions.

There can be little doubt that over the next few years, mathematicians

will shorten and simplify Wiles' proof, so that it will be accessible

to a wider audience. (Wiles' proof, starting from scratch, would surely

be over a thousand pages long).

The story of this important discovery is a tribute to the deeper and

more abtruse levels of abstract understanding that mathematicians have long

claimed as essential. Many of us, while hailing Wiles' magnificent achievement,

yearn for Fermat to have been correct, and for the truly marvellous, and presuma

bly comparatively straightforward, proof to be recovered.

\end

## Mult mai mult decât documente.

Descoperiți tot ce are Scribd de oferit, inclusiv cărți și cărți audio de la editori majori.

Anulați oricând.