Science and extraordinary events

Auditors have discovered a million dollars missing in the accounts of the venerable First Bank. What is the explanation?

(1) There has been no impropriety; small rounding errors of a few pennies have occurred many times over many years, which happened to add up to a million dollars.

(2) There was an embezzlement of a million dollars.

This is the situation science faces. A large difference might be explained by a long series of small changes, but at what point does the long series of small changes become harder to believe than one extraordinary event?

Science follows the principle that what is easiest to believe should be believed. Hume and others have argued that it would take an extraordinary amount of evidence to establish that an extraordinary event occurred. One result is that a large number of ordinary events have been taken to be more believable than one extraordinary event.

Evolution requires a very long and particular series of events to occur in which small changes lead to larger and larger changes that ultimately result in the biological diversity observed today. This has the advantage of not requiring the belief in any extraordinary event. But the sheer number and variety of particular changes over an extremely long time that would have to take place are harder to believe than any one extraordinary event.

It is reasonable to expect science to minimize the number and extent of extraordinary events but if extraordinary events are thereby eliminated, the result is something harder to believe than an extraordinary event. That goes against the principle that science should believe what is easiest to believe.

July 2014