Atheist, Theist, or Agnostic? Sean Carroll's Failure to Resolve the Theism/Naturalism Debate

With his mastery of modern natural science and his deep understanding of the theistic arguments for the existence of God, Sean Carroll--a theoretical physicist at Caltech who has debated theists like William Lane Craig--would seem to be in the best position to resolve the debate between theism and naturalism.  In his best-selling book The Big Picture: On the Origins of Life, Meaning, and the Universe Itself (Dutton, 2016), he tries and fails to do this.  He fails because of his intellectual honesty--he explains both sides of the debate so clearly that his readers can see that Carroll's arguments for atheistic naturalism cannot completely refute the arguments for theistic supernaturalism. 

He also fails to draw the right conclusion from this--that Charles Darwin was correct in seeing that modern evolutionary science can be plausibly seen as compatible with either theism or atheism, so that a man like Darwin, who sees the good arguments on both sides, can reasonably choose to be neither a theist nor an atheist but an agnostic.

Earlier this month, I posted my assessment of the debates between Carroll, Craig, and Lawrence Krauss over the cosmological arguments for theism and naturalism.  Here is a video of Carroll lecturing about some of what he says in The Big Picture, although he does not include here his reasoning about morality and religion.

A crucial claim for Carroll's position is that scientific reasoning and all forms of empirical reasoning can be understood as Bayesian reasoning (named after the Reverend Thomas Bayes, who developed a theory of probability in the 18th century).  Bayesian reasoning is a way of calculating our degrees of belief in cases of ignorance or uncertainty.  We must start with some initial degree of belief about what propositions are plausible, which statisticians call a prior credence.  So we might start with some intuitive feeling that God's existence is very likely or very unlikely.  We must then make predictions about what outcomes are likely to happen if our idea is true.  If those predictions often come true, then we can increase our credence in the idea.  If those predictions often don't come true, then we must lower our credence in the idea. 

We must also see that evidence favoring one alternative disfavors the other alternative.  So any evidence that raises the likelihood of God existing must lower the likelihood that God does not exist.  So, in principle, as we gather more evidence and alter the likelihoods of alternative propositions being true, we should move towards agreement.

Oddly, however, even as Carroll applies this Bayesian procedure to the debate between naturalists and theists, he admits that it doesn't work:

"Remember that there are two parts to Bayesian reasoning coming up with prior credences before any evidence is in, and then figuring out the likelihood of obtaining various kinds of information under the competing ideas.  When it comes to God, both of these steps are enormously problematic.  But we don't have any choice" (145).

The first step is problematic, because both sides in this debate start with biases that are so strong--either a preference for theism or a preference for atheism--that it's difficult for either side to be open to evidence contradicting their position.

The second step is also problematic, because God is "a notoriously slippery notion," and therefore theism is so badly defined that it does not make clear predictions about what we should expect to be likely in the world if God exists (145, 148).  So, for example, if God is all-powerful, all-good, and all-benevolent, then we might predict that evil would not exist, or that good people would always prosper, and bad people would suffer.  If so, then the existence of evil in the world would lower the likelihood of God's existence.  But, of course, theists can always respond by arguing that God has good reasons to allow evil in the world, perhaps because he wants to allow humans free will, which includes the freedom to make bad choices.

If God exists, we might predict that He would reveal His existence to human beings, and so the fact that human beings think about God should count as some evidence for His existence.  And, indeed, Christian theologians like John Calvin have said that human beings have an intuitive sense of the divine.  But then atheists can argue that human beliefs about the divine in different religious traditions are so contradictory that this should be seen as evidence that there is no divine revelation of God's existence.

Carroll says that the fine-tuning of the universe is "the best argument we have for God's existence" (303).  If God exists, and if this God cares for human beings, then we might predict that He would design a universe that would be precisely tuned for human life, and therefore the fine-tuning of the universe would seem to increase the likelihood of God's existence. 

But, then, as we saw in the previous post on Carroll's debate with Craig, Carroll can argue that the evidence does not show a universe precisely tuned for human life: throughout most of the history of the universe, there has been no human life; and even now, human life seems to be confined to one tiny spot in the universe.  Moreover, we can anticipate that in the remote future, all life will disappear, and the universe will become a dark and empty expanse of entropic disorder.  This looks like evidence for a universe that has not been fine-tuned by God for human life, which makes God's existence unlikely.

The theist can argue that an observable universe with a hundred billion galaxies is exactly what one might predict if God exists.  But the naturalist would object that this is not the cosmology of the Bible, and that no theist predicted a universe with a hundred billion galaxies until this was discovered in the 20th century by astronomers with modern telescopes.

Strangely, then, even as he insists that Bayesian reasoning is the only way to rationally adjudicate the debate over theism and atheism, Carroll admits that it really doesn't work.

To me, this suggests that Darwin's position is more reasonable than Carroll's.  As I have indicated in a previous post, Darwin's religious views, as stated in his Autobiography and in his correspondence, can be summarized as five points:

(1) It is possible to affirm both theism and evolution--to be a theistic evolutionist--and people like Charles Kingsley and Asa Gray show this. 

(2) Darwin thinks his views of this issue should be of no special concern for others, because each person must make up his own mind based on his personal weighing of the pertinent arguments and evidence.

(3) Darwin finds this issue so mentally challenging that he fluctuates in his thinking, and he cannot come to any final conclusion.

(4) In all of that fluctuation, Darwin has never seen any good reasons to be an atheist, in the sense of denying the existence of God.

(5) And yet, Darwin has seen good reasons, especially in his later years of life, to be an agnostic, in the sense of being in such a state of ignorance that human reason is incapable of providing sufficient rational grounds to justify either the belief that God exists or the belief that God does not exist.

Despite Carroll's claim that the astounding increase in our scientific understanding of nature over the past one hundred years has strengthened the case for atheistic naturalism, I see nothing in Carroll's book that refutes Darwin's claim that human beings remain in such a state of ignorance that they cannot resolve the debate over theism and atheism, and that it is therefore reasonable to be an agnostic.