In pure disciplines like mathematics, error may be completely eliminated; but where physical reality is concerned, error is an unavoidable hazard. It is an integral part of observation and perception. The scientific method is itself bound up with the process of experiment and observation, so it turns out that all information, including our most fundamental scientific understanding, is the result of interpretation. As Brownowski puts it, ‘All knowledge, all information between human beings can only be exchanged within a play of tolerance. And that is true whether the exchange is in science, or in literature, or in religion, or in politics25’.

If we must accept that error is an unavoidable and integral hazard, when dealing with interpretation of scientific evidence, we may wonder how it is possible to evaluate the validity of any proposition. A scientific answer is that we might consider its probability. If we can correctly assess the mathematical probability of each of the available alternatives, then we will surely have a good guide to discernment of the truth.

THE INEVITABLE COINCIDENCE

That power which erring men call chance. - John Milton

My father once tiled the kitchen floor. His original idea was to have a single background colour with an odd foreground colour thrown in here and there, at random. Before he stuck the tiles down he laid them out distributing the foreground tiles at random. When he looked at the arrangement, he decided that it did not look random enough. So he took up the foreground tiles and placed them out again at random. He repeated this process several times before he decided to abandon the idea for a formal pattern arrangement. It seemed impossible to arrive at a suitably pleasing and yet apparently random design.

This story highlights a common misconception with regard to the notion of randomness. The truth is that random groupings will often contain an element of form or pattern. Casinos would surely take less money if the nature of randomness were better understood. A tossed coin has an equal chance of landing heads or tails, even if the last ten throws have all landed heads; just as there is also an (almost) equal chance of red or black coming up at the roulette table. So, according to the rules of chance, randomly distributed tiles are free to constitute geometric patterns; tossed coins may continually land heads; and red numbers may continually occur at roulette. That this truth is poorly understood is continually demonstrated by the various punters who play the odds. After a succession of red wins at the roulette table, it is frankly surprising to see people gambling on black at the next turn as if that possibility were somehow more likely.

You might like to try the Notice Red test: tell yourself to notice all of the red items in the room; repeat the suggestion to yourself a few times; then have a good look around. You will be surprised how much red is in the room. Then try the Notice Yellow (or any other colour you like) test, and you will again be surprised how much yellow (or whatever) is in the room. On the next occasion you are amongst a large crowd of people, try telling yourself to notice items of red clothing. What a coincidence it will be that so many people will turn up, all wearing red, on the very day that you decide to conduct the test.