CREDIBILITY

Pascal recognized that the metaphysical proofs of God were based2 on logical implication so remote from ordinary reasoning that they have little persuasive power against scepticism. He was specially concerned to argue a case for belief that would not rely on the authority of church doctors, nor upon the witness of the faithful, since these were already discredited by the sceptic. So he invented his wager. Modern decision theory requires an exhaustive list of the possible hypotheses of the way the world is, an inventory of possible decisions plus the different benefits from making a decision in all the possible various states of the world: from this the analyst can determine the decision most likely to pay off best. But Pascal ruled out observations of experimental data, since he would not consider reports of miracles either: this is the case of decision-making in face of uncertainty when no

experiments are possible. To solve the problem he correctly used three separate arguments, called in the jargon, ‘dominance’, ‘expectation’, and ‘dominating expectation’. Dominance applies when one course of action would be better, no matter what the world is like: there is more utility in preferring course A-1 than any of the other actions: then course A-l is said to dominate. To bet that God does not exist, and to live as if that is what the world is like, will bring damnation if the bet turns out wrong. Since salvation is infinitely better than damnation, the dominance rule directs the bet in God’s favour or rather in favour of living a life that is reckoned to win salvation-an important difference. Similarly for ‘expectation’ and ‘dominating expectations’, all three arguments indicate the decision to live by the rules of religion. All three arguments are valid in the sense that the conclusions follow from the premises. It is the premises that are hard to defend.