When Magical Thinking Works - It's Not a Secret
In 1959, Evon Z. Vogt and Ray Hyman wrote "Water Witching".
It was a fascinating and objective review of the practice of water witching -using a forked stick to find or indicate an underground source of water.
The book was reprinted in 1979 and 2000, and despite our technological advances, it appears that dowsing is still with us.
I almost passed this book over -who in their right mind could believe that dowsing has any merit?
But I would have missed a fascinating examination about human decision making and the utility of "magical thought".
In their chapter, entitled "Water Witching as Magical Divination", the authors argue that when an individual is faced with a one-shot decision, uncertainty and the need to make a decision, water witching makes sense as measure of psychological control over the state of nature -even if the practice is completely unscientific and unsupported by evidence.
When we need to make a decision, a decision which is not going to be repeated, modern decision theory offers us thin gruel, focussed as it is on the long run or similar situated choices.
"The key feature of this picture [of rational decision making] is that we assume that the decision he is making is one that is to be repeated over a large number of trials.
And, in this way, it becomes meaningful to speak of his average gain from repeated events over the long run.
But what happens to our hypothetically rational man his decision is restricted to one event? ...
Water witching from the scientist's point of view is "irrational" divination; he is evaluating a large series of outcomes from water witching.
But from the viewpoint of an individual decision it is perhaps meaningless to call water witching "irrational".
At least, if we do label it, we must be careful to specify precisely in what way we are making the transition from the institutional [long run] view to the individual situation."
The author's conclusion is based on their scientific observations that dowsers have no better than chance of being correct in locating water and that the regions in which there is high dowsing activity could be characterized as Pascal's Wager with respect to water - it is better to take a chance, any chance, that water will be found than do nothing.
A modern decision theorist, if informed that dowsing is no better than guessing about water, would recommend flipping a coin between the two method -assuming the costs of dowsing and guessing were roughly equal. (They appear to be, since according to Vogt and Hyman, view dowsers make a living from their "talents".)
I am reminded of a philosophy talk, which I was at, in which a famous philosopher declared that modern decision theory was irrational.
His example was this.
His wife was tied to the train tracks and only if he could warn the train, coming from the east or west we know not which, to brake could his wife be saved.
The train needed to brake 100 yards from his wife, a distance our philosopher could reach in 10 minutes. He is told that a train is coming, from the east or west we know not which, and the train will be 100 yards away in 10 minutes.
The famous philosopher opined that modern decision theory dictated that he must stride of in one direction, not look back and hope for that the choice was correct.
While, it was clear to him, that he would run a little bit east, backtrack and go west, trying to cover all bases.
But our famous philosopher was clearly irrational -you have ten minutes to achieve one goal, the consequences of running around and ignoring what you have to do are severe. You will fail.
You may go in the wrong direction, but going back and forth and in no direction is completely wrong.
Decision theory doesn't give comforting results for uncomfortable situations.
![Reblog this post [with Zemanta]](http://img.zemanta.com/reblog_e.png?x-id=a191e27b-82f7-409d-9681-3a23f5335bfb)
Comments
'Tyler and Alex, why don't you install a decent commenting system like disqus, for example? You could then add value to your very best readers.
'The typepad commenting system is fairly primitive, doesn't allow threading, and requires people to return directly to the blog to find out if anyone has responded to their point.'
- Your comment here, http://www.marginalrevolution.com/marginalrevolution/2009/01/the-show-so-far.html#comments
I have yet to see a blog as good for comments threading as was Glendale Litera BBS back in the late 80s, early 90s via 14k dialup. Litera had about 100 lurkers and 12 writers. Exchanges were ZIPd files and read/written in Simple Little Mail Reader (SLMR). Good comments today are ignored unless the blog owner is smart enough to pull them out of the mire by reposting them.
Any publication, paper or online, is only as good as its readers.
- http://govtwork.home.att.net/
I don't use email. There's a Feedback page on my website.
Posted by: Joel | January 8, 2009 5:05 PM