| isherwood,
whether or not prediction is part of the "scientific method" it has very little to do with what counts as an adequate explanation.
If you disagree, could you explain the connection?
As to the necessity of repeatable experiments. I belive merely saying that an experiment must be reproducable is insufficient. If a purported "one shot" experiment can be explained by the theory that also explains other reproducable phenomenon, and the unlikeliness of the "one shot" experiment, then the one shot add scientific credit to the theory.
Take for example good ol' quantum tunneling. I'm sure you've all heard some physics crackpot claim that if they threw a ball at a wall an infinite amount of times, then it would, one some occaisions, go straight through the wall. Now say someone set up this experiment, and observered a ball going straight through the wall. This experiment cannot easily be repeated because of the unlikliness of the phenomenon, however, it is this unlikliness that provides support for quantum theory.
If you don't like quantum theory you can use any rare event (such as the planets' - or other celestial bodies' alignment) or any statistical governed theory (evolution, radioactive decay etc).
So if we can perform a "one shot" experiment that shows someone can read minds, do we adopt a parapsychology theory? No, the fact that it is a one shot experiment supports the falsehood of parapsychology because the theory cannot account for the unliklihood of the outcome. Simple probability theory however, can explain the outcome. |