So those poor test subjects tried to do their best (and yes, I think they were led to think, though not explicitly, that the levers had, or should have had effect on the result), in order not to be embarrassed and lose face.

My own hypothesis would be that pattern-recognition is a fundamental part of how the brain works, not just in terms of how we gain conscious knowledge about the world, but actually in how we learn all sorts of things. Riding a bicycle (or walking!) involves finding and getting used to the right pattern of movements. I suspect the early stages of learning to talk have a lot to do with children learning that they get particular reactions when they repeat particular sounds (Mumumumumum… is sure to get a reaction whether the baby knows it refers to his/her mother or not). I'd guess that pattern recognition is built into us at a far deeper level than the one on which we do scientific induction. Mind you, I don't to what extent that hypothesis has been tested.

Everything I have studied about the brain (keeping in mind that I am not a full-fledged neuroscientist) points to this kind of situation. If you're interested in reading more, Vernon Mountcastle proposed a common modular architecture throughout the human cortex in 1978. More recently, based on Mountacstle's idea, Jeff Hawkins postulated a memory-predicition theory about how all parts of our cortex basically store spatio-temporal patterns and use these stored associations to predict what will happen next in any given situation. Our larger cortex essentially allows us humans to store much more abstract and complicated patterns (ie: language, culture, scientific reasoning, religion etc) than other animals. I highly recommend reading Hawkins' book On Intelligence.

At work, when I first took over my current position from a co-worker whose intellect I greatly admired, I found irregularities that some would have called "mistakes." He's no longer around to ask. I steadfastly stuck to a presumption that anything odd probably had a sound, logical basis. (This guy was very invested in his work) In the face of evidence to the contrary I trusted his work and tried to figure out what it was he was tying to accomplish. In the end, I changed the way things are done because it was simply wrong.
When you start out presuming there's order, you end up looking for explanations, any explanations and, by golly, you WILL find them.

I'm guessing it probably worked well for our prehistoric ancestors. "Good superstitions" (which resulted in a survival advantage and may have contained a grain of truth; the two are not necessarily correlated) would be reinforced by natural selection and, over the years, as knowledge advanced, refined until eventually a rational explanation emerged. "Bad superstitions" (which resulted in a survival disadvantage) would be rapidly selected against, and so would perish along with their believers.

What has changed is that it's a lot harder for us today to get killed accidentally than it would have been for our cave-man ancestors. "Indifferent superstitions" (which would not affect your survival chances one way or the other whether you believed and acted upon them or not) are no longer selected against, and have a chance to thrive.

There seems to be a sort of breakdown even in the possibility of a bridgehead lexical understanding between atheists and the religious (which disturbs me because they are big and strong and rich and bullies. so there.) We can't agree on a common conceptualisation/provenance/domain of any of the following words really, as far as I can see: fact, evidence, theory, cause, logic, argument, proof, fallacy, contradiction, existence, matter, religion, morals, history, meaning, people, superstition, rationality and probability, just off the top of my head. It's like two perfect strangers who speak different languages trying to communicate over a bad land line through a defunct exchange.

So here's my question: shall we all just run screaming from the building? I can see how dialogue, discussion and debate would be important for atheists who have been feeling isolated, or agnostics who aren't sure, or even the doubting religious. However, I just don't see the point in carrying out any further conversation with committed believers, and I find that deeply shocking somehow. It's impenetrable.

How do people get to a post-superstition stage? I was trying to remember things I was superstitious about, and I couldn't (probably proving that I have so many superstitions that I can't keep them straight.) How does one communicate across gaps of surrealism? Sorry for the rambling on, but I just cannot get a handle on this.

We hung up this mobile above her changing table. After a while she seemed to have noticed that she could make the mobile move. She was repeating a pattern of behaviour that went like this:
1. Kick legs for a while
2. Pause
3. Wave arms
Upon the arm-waving stage, her hand would usually hit the mobile at some point and it would move.

Eventually I think she must have figured out that the first two stages were unnecessary. Still, I think this tells you something. My own hypothesis would be that pattern-recognition is a fundamental part of how the brain works, not just in terms of how we gain conscious knowledge about the world, but actually in how we learn all sorts of things. Riding a bicycle (or walking!) involves finding and getting used to the right pattern of movements. I suspect the early stages of learning to talk have a lot to do with children learning that they get particular reactions when they repeat particular sounds (Mumumumumum... is sure to get a reaction whether the baby knows it refers to his/her mother or not). I'd guess that pattern recognition is built into us at a far deeper level than the one on which we do scientific induction. Mind you, I don't to what extent that hypothesis has been tested.

I'm not even sure how you could go about testing the idea, without hitting an ethical wall. Children raised by other humans (parents or otherwise) would learn the expectation from their parents -- there's no way around it, really; proper childcare involves reacting to what the children do -- and you can't leave a bunch of babies alone to see what happens when they grow up, because they wouldn't.

...the subjects were led to believe that relationships existed between their actions and the number of points scored, while presumably humans didn't start off believing that about the universe.

I don't necessarily agree. Yes, no one told early humans that they could control the workings of the cosmos through their actions, but I'd argue that humans instinctively believe this; it doesn't have to be taught. If a primitive group of people performed some action and noticed that a good result happened soon afterward, it's a natural tendency to think that the two are somehow connected. If they repeated the behavior and, by chance, the good result repeated also, this association would only be reinforced.

And once a superstition of this kind gets going, people can be amazingly resilient in overlooking or explaining away the times when it didn't work. ("The rain dance would have worked, if only I had taken one more step to the left!!")

Last comment off topic (promise!), but that's encouraging to hear (that you like football, not that the networks won't broadcast it). If you want to discuss "footy", there's an email address on my blog.