1  Six things to do with a conjecture you can't prove

This applies to the simplest guesses of everyday life, and the profoundest guesses about the Universe, as much as it does to the most intractable conjectures of mathematics.

(1) Keep trying to prove it.
You might never get an answer, but your attempts can help to narrow down the possibilities, and close in on the quarry. Mathematics has always been a communal enterprise. Honourable failed attempts on the Big Conjectures have nearly always led to insights into other things, and even proved significant side results along the way. The story of thousands of years of attempts to solve the great construction problems of antiquity (ultimately shown impossible) is the story of large tracts of modern mathematics spawned in the process. Some beautiful mathematics has arisen out of ``failed'' endeavours. The recent success of Andrew Wiles in cracking Fermat's Last Theorem at last (see Issue 8.1 page 7) has actually encouraged many mathematicians to have a go at some of the Big Ones, like the Riemann Conjecture - the Holy Grail of mathematics (Issue 8.1 page 8).

(2) Try to disprove it.
Try to show that assuming it leads to something absurd, or try to find exceptions/counterexamples. The mental activity demanded here is often quite different to that in (1), but the two can serve each other. R. H. Bing said of the Poincaré conjecture that he would alternate, spending 2 weeks trying to prove it, then switching to 2 weeks trying to construct a counterexample. (See the article following this.)

(3) Prove that it would be implied by some other things you don't yet know are true.
The story of Fermat's Last Theorem (like that of the Four Colour Theorem) is full of heroes who succeeded in reducing the problem to verifying one thing or the next thing. Seek for a good forward position, a staging post, half-way house, a series of base camps for your assault on the summit. Think of a simpler problem whose solution would help with the harder problem. It would be OK if only ... It reduces to showing.... It will be sufficient to show... This strategy often surprises - suddenly you are nearly all the way there!

(4) Prove that it implies other things you would like to know.
If this then what? It would be wonderful if this were so, because then.... Pointing out the gains can load the problem with significance - make its solution more worthwhile and provide motivation for solving it. The conjectures in mathematics which have most prestige are those which not only have defeated the best for the longest, but whose solution would open the gates to other treasures.

(5) Generalize the problem.
Sometimes mathematicians do this to cover their embarrassment, and make the problem sound harder and more grand! But sometimes, indeed, the more general expression touches on some other field where progress is being made, and makes fruitful connections.

(6) Ignore it.
Consider first: maybe there is no great advantage in knowing it, and your isolated effort in proving it may lead to no great acclaim or new insights. On the other hand, it may be important for you symbolically - cracking it may give you great encouragement. It may be crucial for your self-esteem not to pass by, or give up. It may be an important step forward, a significant block along the way, and be worth substantial sacrifice of time and effort. Perhaps, like the really Big Conjectures, it lurks at the very heart of mathematics and its importance cannot be overestimated: it lures you on irresistably. The trouble is that taking on something too hard can seriously discourage you, dissipate your energies - even kill you. George Polya once had a young mathematician confide to him that he was working on the great Riemann hypothesis. `` I think about it every day when I wake up in the morning,'' he said. Polya sent him a reprint of a faulty proof that had been once been submitted by a mathematician who was convinced he'd solved it, together with a note: ``If you want to climb the Matterhorn1 you might first wish to go to Zermat where those who have tried are buried.''

-----------
The Riemann Hypothesis is not just a problem. It is the problem. It is the most important problem in pure mathematics. It's an indication of something extremely deep and fundamental that we cannot grasp.Enrico Bombieri

-----------

2  What if your conjecture gets killed?

Marcus du Santoy, of the Mathematical Institute, University of Oxford, and writer of the recent best-selling book The Music of the Primes, was challenged by an interviewer: ``The conjecture you have spent a lot of time on has recently failed. How do you face up to that?'' He responded:

``My student Luke Woodward came in and showed this example that disproved my conjecture about palindromic symmetry. He was a bit worried that I was going to be angry or something, but he was obviously really pleased with his work and he did this lovely build-up. He showed me some of the things he had been doing, and then a few more, and he kept this example to the last minute and then said: `Oh, look, I've found this!' And you know, you just have to accept what nature gives you. There is always a silver lining to every discovery, so I'm not too disappointed: there is clearly something going on there. We can't have calculated all these examples and seen all these beautiful symmetries without there being something going on.''


Footnotes:

1A mountain in Switzerland.


File translated from TEX by TTH, version 3.01.
On 2 May 2005, 19:19.

Previous Next Home