Saturday, June 25, 2011

Forgotten Details of a Lost Universal

An objection I heard recently is that a good argument must contain premises that are more probable than their negations. But if one has two premises that are each 51% probable, then the conclusion can have only a 25.5% probability of being true.

It depends on whether the -probabilities- (technically, the underlying random variables) of the premises are in any way dependent on each other. Only if their probabilities are independent of each other would that calculation be correct.

If valid, the objection's argument would work against any two probability-independent premises including its own.

But the more basic question is how could one know the probabilities of -any- basic philosophical premise in the first place. What about the probabilities of the premises that imply the truth of that probability statement itself or the premises underlying any claims about the relation between a way of construing such probability and the existence of God.