The End of Pascal’s Wager? (2006)
Richard Carrier
In “Is Atheism a Safe Bet?” Amy Sayers responds to my essay “The End of Pascal’s Wager: Only Nontheists Go to Heaven” (2002), but she does not understand my argument. Her first objection is that:
Richard Carrier has not ended the wager in all forms, as he claims, because he assumes that the nature of Heaven is completely different from what biblical Christianity teaches. So, even if Carrier’s argument works if all his assumptions are true, it does not work if the biblical Christian’s assumptions are true.
This misses the very logic of Pascal’s wager, and ignores my opening paragraph, where I point out the truth that “we do not know whose assumptions are correct, and we therefore cannot exclude the assumptions on which this argument is based.” This is crucial, for in all forms of Pascal’s wager it is precisely such theological assumptions that we are wagering on. Therefore, Pascal’s wager is logically invalid if you must already assume its conclusion, e.g. that biblical Christianity is true (at least in the sense Sayers means, since not all Christians agree what “biblical Christianity” actually consists of). Therefore, in no form of Pascal’s wager can we grant the assumptions that Sayers wants. Instead, it is the very point of the wager that we are betting on whose assumptions are correct.
By definition, every version of Pascal’s wager holds that certain assumptions will stand a better chance of paying off without our knowing whose assumptions are best supported by the evidence. For if we can establish a set of assumptions (e.g. that biblical Christianity is true) on the evidence, we don’t need the wager. Hence the wager can only operate as an argument if we don’t know from the evidence whose assumptions are correct. After all, if the evidence supports biblical Christianity better than any alternative, then why would we need to wager on it? Hence Pascal’s wager is not an epistemic argument, but an argument from utility. It does not argue that Christianity is probably true. Rather, it argues that it is more useful to believe regardless of whether Christianity is true. That is what distinguishes Pascal’s wager from all other arguments (including several advanced by Pascal himself).[1]
Once we understand what Pascal’s wager logically argues, it becomes clear that I have just as much right to bet on my assumptions as anyone else has to place their bet on any form of biblical Christianity. That is, in fact, the entire point of my argument. But when we compare the two bets, we see that both pay off the same way. Therefore, we cannot decide between them: neither is a better bet. So no matter how you formulate it, no version of Pascal’s wager can produce the conclusion that we should believe in God. This does not mean that there are not or cannot be other arguments that support belief in God, but if such arguments were available, the wager would no longer apply—for we would no longer be betting on who was right, as we would know who was right (at least to some degree warranting belief).
Sayers’ second objection commits the same fallacy as the first: she objects to my “definition” of morally good even though that definition is precisely what I would be “betting” on if faced with Pascal’s wager. But her second objection also ignores what I say in my opening paragraph, that “to escape the logic” of my argument “requires theists to commit to abandoning several of their cherished assumptions about God or Heaven,” including their assumptions about what is “good.” Perhaps it is not obvious to Sayers how redefining “good” in a way that avoids the force of my argument entails “abandoning cherished assumptions” about God or Heaven. But obvious or not, that’s what such a redefinition entails.
Sayers is correct that some theists believe that it is not goodness that earns one a place in Heaven. Yet these same theists also believe that it would be unjust to punish a good person and reward a bad one. Therefore, to believe that God would do such a thing requires them to abandon at least one cherished belief about what it means for God to be “good.” He cannot be just if he does not punish and reward according to what is actually deserved. That is why accepting the saving forgiveness of Jesus requires a genuine repentance of one’s sinful nature: hence it is not about doing good deeds, but it certainly is about being a good person. Therefore, my assumptions do not in fact differ from hers. Someone who remains wicked while insincerely accepting Jesus merely out of self-interest is not a true Christian and won’t receive God’s grace—unless God is not “good” by any definition widely cherished by Christians. Therefore, either it is true that only nontheists go to Heaven or God is not “good” by any popular conception of the term. And even if we jettison every popular definition of “good” in order to make it possible for the believer’s bet to pay off, the wager still fails because my bet pays as well as the believer’s, leaving us at a stalemate.
Furthermore, Sayers suggests that God must reveal himself in order to convey what he wants from us (i.e. how to pass his test), but this is a non sequitur. God could reveal this to us through our natural moral intuition, or through his secret inspiration of the world’s cultures—neither of which requires him to be anything but hidden. And we do not have to believe in God in order to believe we evolved a natural moral intuition and then developed cultures that have explored and arrived at better and better ideas of what is good for us.[2] Moreover, that we must learn on our own what is right is precisely what God is testing on my bet, since God could not openly teach us without destroying his work (by causing an increase in conversions for insincere reasons). Therefore, it is logically impossible for God to reveal himself on my assumptions. Though this might mean many are not saved, nevertheless even fewer would be saved if God was not hidden, and God is compelled by unassailable logic to choose the lesser of two evils. Of course, if you don’t like this kind of God, you could propose the God Sayers is betting on, and thus avoid my assumptions, but that bet pays no better than mine. Or you could simply emend my bet to include reincarnation in alternate universes: so those not saved get to try again and again until they learn. This would indeed be the best kind of God. But that would just be another bet in which atheism pays as well as belief in God, proving that no bet on belief is certain to be the best bet, and therefore any version of Pascal’s wager fails.
Finally, when Sayers asks “What is the standard?” she is again missing the point: we can bet on any standard we want to. So long as that bet pays as well as any other bet, no one can claim any other bet pays better. Thus, Pascal’s wager is not an effective argument for belief in God. That is what my article argues. Though my argument does not require it, I do believe that my standard of moral goodness is the most rationally defensible, simply because it is the standard by which I judge others. Thus, unless God is less just and merciful than I am, he must judge me the same way. And as my opening paragraph observes, for Christians to believe that I am more just and merciful than God would require them to abandon yet another cherished belief about their God. But we need not go that far, for biblical Christianity already asserts that I shall be judged as I judge (Matthew 7:2). Therefore, Sayers cannot object to my standard, since it actually is the standard affirmed within her own biblical Christianity!
[1] See the Stanford Encyclopedia of Philosophy entry on Pascal’s Wager for a thorough discussion.
[2] I argue this very point in my book Sense and Goodness without God (2005). Contrary to Sayers, I never asserted that people are truly good only when they are good for no reward. Rather, I said that you are not truly good if you “behave as if good out of fear or selfish interests, not out of courage or compassion or a sense of personal integrity.” This does not entail believing that there is no reward for being good, nor even that we should be good without concern for its potential rewards. In fact, I agree (and explicitly argue in my book) that “a person can pursue goodness while being self-interested and without being selfish.” Hence my dichotomy is not between being good out of self-interest or being good for some other reason, but between being good and merely pretending to be good. We can be good precisely because of its rewards, so long as we are in fact actually good and not merely pretending to be. That is the distinction I make in my article, not the one Sayers alleges.