2000.06.21 Science Needs You

Science Needs You - by Mike Holderness

Imagine there are four cards on a table in front of you, on which are written "A", "D", "4" and "7". You are told that each card has a number on one face and a letter on the other. You are also told that the rule is that a card with a vowel on one side has an even number on the other. Which card/cards do you have to turn over to find out whether the rule holds out? This is a version of the Wason test beloved of experimental psychologists. It would be a disservice to the discipline to reveal the answer while there may still be papers to publish about it. So we won't.

If you find it hard, don't worry. You're in good company. Over many repetitions of the test, designed to include a wide variety of subjects (though doubtless with a real-world skew towards readily available undergraduates), fewer than 10 percent get the right answer. You may still be a gifted human being, and you might even make a good lawyer or musician. There is, however, strong reason to doubt whether you should be a scientist.

Before dealing with that arrogant assertion, try another puzzle.

It was a dark and stormy night. A man was driving his son home. The car skidded off the road and hit tree, killing the father instantly. The son was seriously injured - but with the kind of luck necessary to make stories work, and ambulance was passing and took him to hospital. There, he was wheeled into the operation theatre, where the surgeon exclaimed: "I can't operate! That's my son!" What happened?

If you can't work this out, you should definitely never design an experiment. You are also disqualified from being a feminist. This is embarrassing to the majority of self-defined feminists, many of whom propose ever more bizarre explanations heading resolutely away from the answer: "The surgeon's his mother."

The stormy-night puzzle is a test of your ability to apply the 14th-century principle of Occam's razor, often translated as "Do not multiply entities unnecessarily." Or, in a 21st-century version, "Whatever happens, try all possible ways of explaining it in terms of things we already know about, before introducing aliens or dark matter."

The key to doing this well, or at all, is to throw off the shackles of established ways of thinking about known entities. The unstated assumption "a surgeon is a man" is a case in point. Huge amounts of bad science - especially, these days, bad evolutionary psychology and genetic determinism - are caused by accepting such received wisdom.

Another, subtler, case is the assumption that there is - outside criminal law - such a thing as evidence. Karl Popper's assertion that experiments can only falsify theories really works. And that's a hint about how to approach the Wason test and identify bad science.

A final puzzle. After your first day in a new job, you arrive late at an office party. You see one couple kissing. And a second couple not kissing, whose hands you can't see. And a third couple wearing matching wedding rings. And a fourth couple not wearing wedding rings. What further observations of these couples do you have to make to establish which of them are having extramarital affairs?

This is, of course, another trick. If you find the connection with the first puzzle obvious, drop whatever you're doing and take up science immediately. It needs you.