The Moral Psychology Argument for Atheism

(Note: I am clearing out my backlog of draft blog posts but I am officially on a hiatus from blogging. What this means is that while you are free to comment on this post, readers should not expect engagement from me anytime soon.)

Updated 7-January-2022

John Jung Park's 2017 paper revives an old argument for atheism: errors in human moral judgment due to various psychological biases provides a reason to think God does not exist. Park's argument refers to "framing effects." Following another one of Park's papers, we may say that moral framing effects occur when 

morally irrelevant differences in the way a scenario is presented affect people’s moral judgments regarding that scenario.

Here's an example (taken from Park's 2021 paper):

Save vs. Kill: Petrinovich and O’Neill (1996, Exp. 1) reported that some participants indicated that they would do one act when the act was described in terms of how many people it would save (Save condition) but would not do that act when the act was described in terms of how many people it would kill (Kill condition), even though they had been told that the act would have both effects.

Returning to Park's argument for atheism, Park does not explicitly provide the logical form of his argument, so the following is my reconstruction.

(1) If there exists a God who is omnibenevolent, omniscient, and omnipotent, then there exists a God who is not an evil deceiver. [From the definition of omnibenevolence]

(2) If there exists a God who is not an evil deceiver, then no finite persons would be susceptible to unconscious errors such as making contradictory judgments due to framing effects.

(3) If there exists a God who is omnibenevolent, omniscient, and omnipotent, then there are no finite persons susceptible to unconscious errors such as making contradictory judgments due to framing effects. [From (1) and (2) by Hypothetical Syllogism]

(4) Some finite persons (i.e., humans) are or have been susceptible to unconscious moral errors such as making contradictory judgments due to framing effects. [empirical finding]

(5) No God (who is omnibenevolent, omniscient, and omnipotent) exists.

As I read him, Park offers the following reasons in defense of (2):

• Because God is omnibenevolent, God "would not want us to be susceptible to unconscious errors such as making contradictory judgments due to framing effects" (507).
• Because God is omnipotent, God "would have the power to not make us susceptible to the above psychological biases" (507).
• Because God is omniscient, God "would know how to make us without such biases" (507).

Park anticipates objections and responds to the following objections.

• Objection: If God made humans without psychological biases, that would undermine other important goods that God would have reason to bring about.
  Reply: "There is no greater good that the possibility of such errors allows for that would not be capable of being brought about without the given possibilities of error." 
• Objection: Having cognitive biases is necessary for free will.
  Reply: Having cognitive biases is not only unnecessary for free will, but in fact might undermine free will. 
• Objection: Having cognitive biases is necessary for making the world sufficiently challenging for moral testing.
  Reply: Such biases are not needed to test us to live an ethical life.
• Objection: Good such as soul-building cannot be achieved without cognitive biases. 
  Reply: Empirical data about the moral development of those with PhDs in ethics shows that even they are not immune to cognitive biases.

As I have reconstructed Park's argument, it is deductive in form, but as Park himself states, the support for the proposition represented by (2) is inductive. For this reason, perhaps the strongest version of Park's argument would be an evidential argument along the lines of Paul Draper's various evidential arguments against theism. Such an argument would run as follows.

1'. E (the role of psychological biases in erroneous moral judgments) is known to be true, i.e., Pr(E) is close to 1.

2'. Theism (T) is not intrinsically much more probable than the hypothesis of indifference (HI), i.e., Pr(|T|) is not much greater than Pr(|HI|).

3'. Pr(E | HI & B) >> Pr(E | T &B).

4'. Other evidence held equal, T is probably false, i.e., Pr(T | B & E) < 0.5. 

While there is excellent reason to affirm premise 2' in this revised formulation, note that it introduces the concept of intrinsic probability. Neither that concept nor how it applies to T and HI is addressed in Park's paper, but, again, I don't think that is a problem for the revised argument due to Draper's defense of his theory of intrinsic probability.

In any case, empirical data about the role which cognitive biases play in causing humans to reach erroneous moral judgments is another significant problem for theism.