Atheism and Theism as Model Choices
One way to think about the comparative rationality of atheism and theism is to treat atheism as the rational default. After all, atheism postulates no God, theism does, and so theism bears the burden of proofs, especially because God’s existence is extraordinary or random postulate.
That’s the wrong way to think about the rationality of theistic belief. Following Michael Rea’s important book, World Without Design, I think of atheism and theism more like frameworks or research projects for making philosophical judgments. Many of the arguments for both positions involve claims that the other side sees as non-starters, and whose premises ultimately suppose that atheism or theism is true. What’s more, the frameworks are sufficiently different that they don’t even affirm the same phenomena to be explained, such as the existence of libertarian free will. In this way, we might even see theism and atheism as distinct philosophical paradigms, rather than simple competing propositions.
Atheism and theism are therefore competing frameworks for explaining certain kinds of “big” phenomena, like existence, order, morality, freedom, and consciousness. Both postulate an ultimate ground for everything. Atheism tends to ground everything in particles and fields. Theism tends to ground everything in a single agent, usually a perfect agent.
For reasons Mike Huemer outlines, simplicity does not settle fundamental philosophical problems, and I think that extends to atheism and theism. Which view is simpler? Postulating a single perfect agent that grounds all kinds of phenomena? Or particles and fields? On the one hand, atheism usually says that there is one kind of thing (matter/energy), whereas theism has two kinds of thing (matter/energy and mentality). On the other hand, theism has one grounding object, whereas atheism has many grounding objects. So which is simpler? I think theism is, but it’s not obvious.
So then how do we assess the rationality of theism? We engage in model-based reasoning. List shared phenomena and determine whether they are better explained by theism or atheism. Then list the unshared phenomena, assess whether you affirm or reject them, and then determine whether they are better explained by theism or atheism.
To illustrate, here are some agreed upon phenomena: contingent existence, order in the universe, consciousness, empirical knowledge, suffering, religious experience.
And here are some (usually) disagreed upon phenomena: necessary existence, libertarian free will, objective moral facts, a priori knowledge, universals, objective life meaning.
Here’s my balance of evidence.
Agreed-upon phenomena favoring theism: contingent existence, order in the universe, consciousness, empirical knowledge.
Agreed-upon phenomena favoring atheism: suffering.
Agreed-upon phenomena that may break even: religious experience (some is consistent with theism, but the totality may be inconsistent with theism).
Disagreed-upon phenomena favoring theism: necessary existence, libertarian free will, objective moral facts, a priori knowledge, real universals, objective life meaning.
Disagreed-upon phenomena favoring atheism: ? [atheists belief in less stuff, so it’s perhaps not significant that this category is empty, but feel free to fill in the gaps]
For me, the theistic model explains more phenomena that theists and atheists agree upon. I admit that suffering favors atheism, but not so much that it overrides the explanatory power of the other factors, especially order in the universe. I think the probability of suffering on atheism is like 99.99%, and on theism, it is well under 50%. [The skeptical theist reply convinces me that suffering is not very improbable on theism, since we don’t know God’s reasons. Plus, if everyone goes to heaven, every life has infinite worth, so suffering is always an infinitely small portion of one’s life.] But I think the probability of order on theism is 99%, and order on atheism is something like 10^-10. [Those are the odds the cosmic fine-tuning argument gives us, and that’s being generous to atheism.]
I think I have good arguments that the disagreed upon phenomena are real, and I think it is clear that these phenomena make more sense on theism, but I understand that atheists will not be as moved by these arguments.
So from my epistemic point of view, theism is supremely rational as a modeling choice. It’s a way of looking at the world that makes a whole lot of phenomena seem natural, coherent, and unsurprising vis-a-vis the main intellectual alternative of atheism. I think theistic argumentation is sufficiently strong that atheists will have trouble assigning theism a very low probability, but I don’t think that atheism is an unreasonable view. It’s not something that has to be proven. It’s a modeling choice, and I get why people choose it.
But atheists should see theism as a modeling choice, and not treat theism as irrational because it is not subject to proof.