Michael Shermer on 'Magical Thinking'

http://www.terceracultura.net/tc/en/?p=234


"Science doesn’t come naturally, because it requires additional cognitive steps, beyond just the basic pattern-seeking, pattern-connecting, connecting the dots behavior we all do that’s known as superstitious thinking. This is what I call “patternicity” (in “How We Believe”), the tendency to find meaningful patterns in random noise. Now, sometimes patterns are real and sometimes they’re not. And we make two kinds of errors. A Type 1 error, which is a false positive: you think the pattern is real, and it turns out it’s not; and a Type 2: you think the pattern’s not real, and it turns out to be real. So let’s take an example. Think of our selves back in a Paleolithic environment, in a dangerous world, and there’s a rustle in the grass. Is it the wind, or is it a dangerous predator? Well, you make a Type 1 error and you think, well, it’s probably a predator, so I’ve got to move away and be cautious, and it turns out it’s just the wind. There there’s no harm. That doesn’t take a big investment of energy. But if you make a Type 2 error and you say it’s just the wind and it turns out it’s a dangerous predator and you stand there and become lunch, you’re now taken out of the gene pool.

So I’m arguing that there is a natural selection for: “Just assume all patterns are real, and operate accordingly”. And of course, that’s just magical thinking. That’s association learning, that’s just connecting A to B and assuming there’s always a connection."

-- Michael Shermer