Mackie on Pascal’s Wager

I’ve never been a fan of Pascal’s Wager.  But there’s a bit more that one might say for it than is often supposed.  For example, the objections J. L. Mackie raises against it in his classic defense of atheism The Miracle of Theism, though important, are not fatal.  Let’s take a look at the argument, at Mackie’s objections, and at how a defender of Pascal might reply to them.

The wager

Pascal begins with the assumption that unaided reason cannot establish one way or the other whether God exists.  I think he is quite wrong about that, since I hold that several of the traditional arguments for God’s existence are compelling.  But suppose, for the sake of argument, that Pascal is correct.  We still, he holds, must “wager” over whether God exists, either betting that he does or betting that he does not.  Yet how can reason decide what bet to make, if it cannot show whether it is theism or atheism that is more likely to be true?

In its simplest form, Pascal’s argument is this.  God either exists or he does not, and you can either bet that he does or bet that he does not.  Suppose you bet that he exists, and it turns out that he really does.  Then you will enjoy an infinite benefit, eternal life in heaven.  But suppose you bet that he exists and it turns out that he does not.  You will have been mistaken, but will have suffered no loss.  Of course, while someone who regards a devout and moral life to be of value in itself will agree with that, a more worldly person would not.  He would say that by mistakenly betting that God exists, he would deprive himself of worldly pleasures he could have enjoyed.  But even if one concedes this, Pascal holds, what one will have lost is still of relatively small value, and certainly of finite value.

Now suppose that one bets that God does not exist, and that in fact he does not.  Then, Pascal says, one will enjoy no gain from this.  Or, even if a worldly person suggests that he will have gained worldly pleasures from it, this would still be a relatively small gain, and certainly a finite gain.  But suppose that one bets that God does not exist and it turns out he is wrong – that God does in fact exist.  Then, says Pascal, he will suffer an infinite loss.  He will have lost out on the infinite reward of eternal life in heaven.

When we consider this cost-benefit analysis, concludes Pascal, we can see that the only rational wager to make is to bet that God exists.  Now, Pascal is aware that one cannot simply and suddenly make oneself believe in God, the way one might make the lights go on by flipping a switch.  But since it is reason that tells us to bet on God’s existence, the problem, he concludes, must be with our passions.  These are what prevent belief.  And they can be changed by throwing oneself into the religious life.  Doing so will gradually alter one’s passions, and in this way belief in God can be generated indirectly even though it cannot be produced directly by a simple act of will.

Mackie’s critique

Against all this, Mackie raises two main objections.  First, Pascal emphasizes that there is no affront to reason in his argument, and indeed that wagering that God exists is what reason dictates.  But this, says Mackie, is not the case, for Pascal’s advice to work up belief by way of molding one’s passions amounts to recommending self-deception.  Mackie notes that Pascal might respond by saying that what one is trying to work oneself into is really what amounts to a deeper wisdom or understanding.  But given Pascal’s own assumptions, argues Mackie, such a response would beg the question.  For whether belief in God does in fact reflect wisdom or understanding about how the world really is is precisely what Pascal acknowledges to be impossible to establish directly by rational arguments.

Second, says Mackie, Pascal’s argument can work only if the options we have to choose from are two, belief that God exists or the absence of such belief.  But in fact there are many more options than that.  We have to choose between Catholicism versus Protestantism, Christianity versus Islam or Hinduism, theism versus polytheism, and so on.  And once we realize that, we see that Pascal’s argument falls apart.  No cost-benefit analysis of the issue is going to give us anything like the crisp and clear advice he thinks it does.

Mackie’s second criticism overstates the case somewhat.  For not every religious view entails that one risks suffering an infinite loss by rejecting it.  Only religions that posit eternal damnation entail that.  And for purposes of Pascal’s reasoning, one need consider only religions of that kind, which narrows things down.  Still, Mackie’s basic point remains that there are more than just the two options considered by Pascal (since there is more than one religion that posits eternal damnation).

Are Mackie’s objections fatal?  It seems to me that that may depend on the epistemic situation of the person approaching Pascal’s wager scenario.  Suppose that, as far as you know, there really are no good rational grounds at all for preferring any one religion over another.  Given the evidence and argumentation available to you, none of them seems like a live option, any more than believing in elves or witches does.  In this case, Pascal’s Wager seems to have no value, for the reasons Mackie gives.  It cannot by itself give you a reason to opt for one among the variety of available religious options, and the exercise in artificially working up belief in one of them would seem to entail irrationally “suppressing one’s critical faculties,” as Mackie puts it (The Miracle of Theism, p. 202).  In short, as a strategy for rationally persuading the most unsympathetic sort of agnostic or atheist, Pascal’s Wager appears to fail.

Can it be salvaged?

However, suppose one is in a very different epistemic situation.  Suppose, for example, that one is not entirely certain that the arguments for God’s existence, Jesus’s resurrection, and other elements of Christian doctrine are correct, but still judges them to be strong and thinks that Christianity is at least very plausible.  Suppose that one considers further that among these doctrines is the teaching on original sin, according to which our rational and moral faculties have been damaged in such a way that it is much less easy for us to see the truth, or to even want to see it, than it would have been had we not suffered original sin’s effects.  Then one might judge that it may be that while he regards the evidences for Christianity to be strong, the reason he nevertheless remains uncertain is due to the damage his intellect and will have suffered as a result of original sin. 

His situation would be comparable to someone who judges that he is suffering from chronic delusions and hallucinations, like John Nash as portrayed by Russell Crowe in the movie version of A Beautiful Mind.  Nash has good reasons for holding that some of things he is inclined to believe and thinks that he sees are illusory.  Yet he finds he nevertheless cannot help but continue to see these things and be drawn to these paranoid beliefs.  Since, overall, the most plausible interpretation of the situation is that these nagging beliefs and experiences are delusional, he decides to refuse to take them seriously and to keep ignoring them until they go away, or at least until they have less attraction for him.  This is not contrary to reason, but rather precisely a way to restore reason to its proper functioning.

Similarly, the potential religious believer in my scenario judges that he has good reason to think that Christianity really is true, even though he is also nevertheless uncertain about it.  And he also judges that he has good reason to suspect that his lingering doubts may be due to the weaknesses of his intellect and will that are among the effects of original sin.  Suppose, then, that he appeals to something like Pascal’s Wager as a way of resolving the doubts.  He judges that Christianity is plausible enough that he would suffer little or no loss if he believed in it but turned out to be mistaken, and little or no benefit if he disbelieved in it and turned out to be correct.  And he also judges it plausible that the potential reward for believing would be infinite, and the potential loss for disbelief also infinite.  So, he wagers that Christianity is true.

Like Nash in A Beautiful Mind, he resolves to ignore any nagging doubts to the contrary, throwing himself into the religious life and thereby molding his passions and cognitive inclinations until the doubts go away or at least become less troublesome.  And like Nash, he judges that this is in no way contrary to reason, but rather precisely a way of restoring reason to its proper functioning (given that the doubts are, he suspects, due to the lingering effects of original sin). 

In this sort of scenario, then, it’s not that the Wager by itself takes someone from initially finding God’s existence in no way likely, all the way to having a rational belief in God’s existence.  That, as I’ve agreed with Mackie, is not plausible.  Rather, in my imagined scenario, reason has already taken the person up to the threshold of a solid conviction that God exists, and the Wager simply pushes him over it. 

No doubt, even this attributes to reason a greater efficacy in deciding about theological matters than Pascal himself would have been willing to acknowledge.  But, tentatively, I judge it the most plausible way for the Pascalian to try to defend something like the Wager argument, at least against Mackie’s objections.  (And I don’t claim more for it than that.  Naturally, there is a larger literature on the argument that I do not pretend to have addressed here.)