Tuesday, January 20, 2009

On Superstition

It seems that superstition is not merely a factor of human irrationality, though a large proportion of it is. Superstition appears to be primarily a result of the human inability to intuitively comprehend probabilities, problems separating causality from correlation and the human ability to learn from experience.

The curious result of this is the learning that particular actions cause low probability events to happen. Say for example the waving of a tree branch was quickly followed by rainfall during a drought. Waving it again at a time in the future caused similar rainfall in a drought. Sold! Clearly, the probability of rainfall during a drought...twice over...is quite small. Learning occurs, creating an association between branch waving and rainfall.

Unfortunately, there was no way to verify that the correlation was actually causality. To do so would require lots of branch waving at random times during a drought and actually checking whether it would reliably cause rainfall. In fact, even then the causality would be tenuous unless the exact mechanism of rainfall causation could be established. Superstition thus serves as a handy shortcut in "comprehending" such matters, then getting on with life. This is not to say that superstition is necessarily good, or even useful, but it is just a human tendency given the somewhat inadequate nature of the human mind.

No comments: