Many nonbelievers may assign the probability that God exists a value of
0. In that case, the expected utilities of belief and unbelief if God exists
will not be infinitely positive or negative after all. The wager argument
will only be effective for agents who think it's an open question whether
there is a God or not.
Can we treat belief as just
another action we can make pragmatic decisions about? In general, we do not
decide what to believe; we just find ourselves believing or not believing
things. Pascal is well aware of this point, though. His recommendation is not to
snap your fingers and start believing, but rather to place yourself in
situations that make it more likely that you will come to believe. (Go to
church, mingle with religious friends, etc.)
Should we choose beliefs on pragmatic grounds? Or should we only decide what to believe
based on the evidence? (Or on some other grounds?)
It is possible that the marginal utility of moments of happiness diminishes
over time. But
summing an infinite number of units of decreasing size does not necessarily lead
to an infinite value: it depends on how fast the size is decreasing. So it's
possible that the total value of an infinite number of days of happiness is
still finite! (Who would
have thought you'd need calculus to figure out what the utility of believing in
God is? But it could be the case!)
It is possible that one could place a negative value on happiness (maybe a
masochist would not find the prospect of eternal bliss enticing, and would
prefer eternal torment). (Notice that although this shows that there may be some
individuals for whom the wager argument is not applicable, it does not show that
there is any problem with the argument in general.)