Thu 14th Aug 2008

Here’s an analogy. Can you guess what it’s of?

An Analogy

You're at a party and everyone's having a great time. The drinks are free and plentiful - people are laughing and talking and dancing. Suddenly a man bursts in:

‘Stop! Stop! Don't drink those drinks! You didn't just drink those did you? Oh that’s bad.’

He claims to be a scientist, and says he’s discovered that the drinks everyone's been downing have been poisoned. It gets worse: it’s a fast-acting poison, according to this stranger, and you could be incapacitated within three hours and dead within four. The only antidote is at a remote medical outpost three hours away. If you all leave now, you might just make it.

What do you do?

There are three choices here. You can:

a) Believe the ‘scientist’. He looks convincing, and now you think about it that drink did taste kind of funny. At best you’ll make it to the antidote in time and live. At worst, you’re going to look pretty bloody stupid. The entire party will have gone on a wild goose chase to a remote medical outpost in their party gear. When you all get back this stranger will probably have called his friends round, and they'll all be in hysterics. Embarrassing. Still – it beats death.

b) Disbelieve the ‘scientist’. No one trusts scientists anymore – not after BSE, and that thing where they tried to convince us we could fly through the air in big metal tubes (well, actually I suppose they got that bit right). There’s probably no poison at all – scientists are wrong all the time! Who the hell leaves poisoned drinks lying around at parties anyway? He’s probably not even a scientist either, and while you’re off desperately trying to reach the antidote he’ll be round your house nicking all your stuff. At best, you’ll make it quite clear that you’re no fool, save yourself a three-hour journey and keep all your stuff. At worst, death.

c) The third option is to wait for ‘evidence’. Surely that’s smart isn’t it? Otherwise you have to make an arbitrary choice over whether or not to trust a stranger. That choice isn’t helped when you ask for proof and the so-called scientist says things like ‘I’m 90% certain that the poison can be treated by this particular antidote’, or ‘My best guess is that you have three hours, but it could be two, or forty-eight.’ The problem with this option is simple. The evidence you need to prove to you that something is happening is that same something happening. By the time you get the evidence that you've been poisoned, the poison is already working, everyone is already dying and it’s too late to get the antidote. In fact, this isn’t a real choice at all - in terms of outcomes c) is exactly the same as b). In other words there are only two choices, believe or disbelieve the scientist.

So what’s the sensible course of action?