Out of the depths of psychology ...

Harvard psych prof Daniel Gilbert recently posted a piece about "The Vagaries of Religious Experience" at Edge. It seems mostly an attempt to refute William Paley's proto-intelligent design argument (though Paley was not, as Gilbert would have it, a "naturalist" -- he was an Anglican priest, theologian and philosopher) by arguing that "highly ordered phenomena can and do emerge from random processes." But the example he uses -- coin tosses -- seems somewhat bizarre. "If we toss a coin for long enough, we eventually observe some highly ordered strings such as 'head, head, head, head, head, head' or 'head, tail, head, tail, head, tail.'" What highly ordered phenomena are these supposed to be? No predictable sequence will ever emerge. Even if one grants the premise -- that "highly ordered phenomena can and do emerge from random processes" -- the point would be to demonstrate that a given phenomenon has emerged from such processes.
Gilbert says that "statisticians have sophisticated techniques that can help determine whether a particular pattern of coin flips is so unlikely that it (like Paley's watch) can only be explained by a non-random process. But research in psychology has shown that people have rather poor intuitions in this regard, and that they tend to mistake the products of random processes for the products of non-random processes but not the other way around." OK, so people are unlikely, using their intuition, to arrive at a correct determination of whteher something emerged out of randomness. So skip the intuition, and use those "sophisticated techniques that can help determine" it. Stephen D. Unwin, a theoretical physicist turned risk analyst, wrote a book a couple of years ago called The Probability of God. Unwin makes skillful use of one of those sophisticated techniques statisticians have: Bayes's Theorem. Gilbert should take a look at Unwin's book.