On the Rationality of Theism: A Critique of Pascal’s Wager
Bliss Chapman
Word Count: 1316
In this paper, I will critique Pascal’s argument for the rationality of theism. In §1, I will outline Pascal’s argument for the rationality of theism. In §2, I will critique the idea that Pascal’s utility calculations provide any causal link between his premises and his conclusion that belief in God is the rational course of action. Finally, in §3, I will detail anticipated objections and responses to my criticism.
Pascal’s Wager
Before beginning a discussion of Pascal’s argument, we must establish a common framework of terminology. To borrow from Green in Engaging Philosophy, “we shall define a divine being as one having all possible perfections. If omnipotence is a perfection then, on the present definition, God has it; if omnibenevolence is a perfection then, on our definition, God has it.” Furthermore, a theist is someone who believes God exists, an agnostic is neutral, and an atheist believes God does not exist.
Now, we can begin to examine Pascal’s argument for the rationality of theism. Pascal suggests that a rational thinker should consider the consequences of their choices and seek to optimize the outcomes of these choices for the greatest expected utility, that is, the subjective value of the outcome multiplied by the probability of its occurrence. When applying this decision-making model to the question of belief in God, Pascal argues that theism is in accordance with reason because even if the probability of God’s existence is small, the reward’s value is many magnitudes larger than any other outcome. Pascal’s formal argument for the rationality of theism can be structured as follows:
(A1) One should always choose the action with the greatest expected utility.
(A2) Belief in God has greater expected utility than non-belief.
(A3) So, one ought to believe in God.
I will now explore some common objections to this argument. First, how can a rational being simply choose to believe in God? Second, why would God seek to reward belief that is motivated by a self-interested calculation of expected reward? Pascal anticipated these objections and suggested that immersing yourself in theistic culture can, over the course of many years, bury even the most severe of doubts. Moreover, even if the initial seed of belief is motivated by personal reward, Pascal argues that this impure impetus fades away in view of a lifetime of religious righteousness and virtue.
By Pascal’s argument, theism is in accordance with reason.
2. Against Pascal’s Argument for the Rationality of Theism
I will now critique the idea that theism is in accordance with reason through premise (A2) of Pascal’s argument: “Belief in God has greater expected utility than non-belief.” Central to my discussion of premise (A2) will be the notion that human knowledge about a potential God is inherently incomplete. By definition, religious belief lies beyond the realm of scientific exploration and philosophical reasoning.
In order to reject Pascal’s conclusion that theism is in accordance with reason, we must push back on some aspect of his expected utility calculation. In his argument, Pascal notes that even if the probability of God’s existence is small, the reward’s value is many magnitudes larger than any other outcome because the consequence of your belief is eternal bliss in the kingdom of heaven. However, this calculation belies two assumptions. Namely, that the God you believe in seeks to honor piety by sending you to heaven and that this God is the same God you have been worshipping.
With incomplete information about what a potential God seeks to reward, any expected value calculation is necessarily speculative. How does a pious theist know the omnipotent and omniscient God they seek to enthuse takes any interest in their actions? How do they know their God isn’t vindictive or that they aren’t praying to the devil? How do they know their God isn’t one of many Gods each with their own heavens and own demands for proper worship?
Recall that expected utility is the subjective value of your decision’s consequences multiplied by the probability of its occurrence. To illustrate, Pascal’s original expected value calculation looked something like this:
However, because religious belief is speculative by nature, it could just as easily look like this:
As illustrated in this calculation, belief in Zeus has a higher expected utility than not believing in Zeus, but by Pascal’s own argument that does not imply that a rational thinker should believe in Zeus. Recall Pascal’s argument:
(A1) One should always choose the action with the greatest expected utility.
(A2) Belief in God has greater expected utility than non-belief.
(A3) So, one ought to believe in God.
According to Pascal’s argument, a rational thinker living in Ancient Greece and considering whether or not to believe in Zeus should choose belief because it has a higher expected utility than non-belief. However, because this thinker isn’t aware of the Spaghetti Monster, they are not choosing the action with the greatest expected utility.
As humans considering the question of God, our input space of potential actions is incomplete and perhaps unknowable, rendering Pascal’s decision-making model a useless tool in logical argument. (A2) may always be true in that belief in a particular god has greater expected utility than non-belief, but that implies nothing about whether belief in that god will result in the greatest expected utility and is therefore the rational course of action.
3. A Response: Reformulation for Conditions of Uncertainty
In this section, I address a possible way in which Pascal could respond to my criticism that theism is not necessarily in accordance with reason. Pascal could respond by acknowledging that (A2), “Belief in God has greater expected utility than non-belief” may not necessarily imply the conclusion of (A3), “So, one ought to believe in God”, but that doesn’t mean the entire expected utility decision-making framework is useless in speculative domains. For example, Pascal’s argument could be reformulated as follows:
(A1) One should always choose the action with the greatest expected utility.
(A2) Belief in a god(s) has greater expected utility than non-belief.
(A3) So, one ought to believe in a god(s).
Restructuring the argument to allow for a rational agent to have incomplete knowledge of their environment provides a weaker but more flexible argument for non-specific religious belief over non-belief.
However, because any expected utility calculation is dependent on the subjective value of a decision’s consequences, work will still need to be done to show that belief in any god(s), or lack thereof, has a higher expected utility than non-belief. For example, if we imagine a future in which we know with absolute certainty that no god exists, it will be a difficult task to show that belief in god has greater expected utility than non-belief. Furthermore, is it still reasonable to expect that immersing yourself in theistic culture can overcome absolute knowledge that no god exists?
I wager that Pascal’s argument requires the same conditions of uncertainty that made expected utility calculations an insufficient causal link between premise (A2) and (A3). Without some further argument to show that belief in a god has greater expected utility than non-belief irrespective of the existence of god, there is no reason to accept the revised version of Pascal’s argument.