This one is a bit harder than my resurrection post with a similar title, for there is more than one reason. But I’d like to focus on what I think is the biggest one, one that was made clear by Richard Dawkins in The God Delusion. It’s an important topic since many think that the fine-tuning argument is the most compelling argument for God’s existence.
What is the argument? Here’s a summary from William Lane Craig taken from his debate with Lawrence Krauss:
In recent decades scientists have been stunned by the discovery that the initial conditions of our universe were fine-tuned for the existence of intelligent agents with a precision and delicacy that literally defy human comprehension. This fine-tuning is of two sorts. First, when the laws of nature are given mathematical expression, you find appearing in them certain constants, like the gravitational constant. These constants are not determined by the laws of nature. Second, in addition to these constants there are certain arbitrary quantities which are just put in as initial conditions on which the laws of nature operate, for example, the amount of entropy in the very early universe.
Now all of these constants and quantities fall into an extraordinarily narrow range of life-permitting values. Were these constants or quantities to be altered by even a hair’s breadth, the life-permitting balance would be destroyed and life would not exist. We now know that life-prohibiting universes are incomprehensibly more probable than any life-permitting universe.
Now there are three possible explanations of this extraordinary fine-tuning: physical necessity, chance, or design.
Now it can’t be due to physical necessity because, as I’ve said, the constants and quantities are independent of the laws of nature.
So maybe the fine-tuning is due to chance. After all, highly improbable events happen every day! But what serves to distinguish purely chance events from design is not simply high improbability but also the presence of an independently given pattern to which the event conforms. For example, in the movie Contact scientists are able to distinguish a signal from outer space from random noise, not simply due to its improbability but because of its conforming to the pattern of the prime numbers. The fine-tuning of the universe for intelligent agents exhibits just that combination of incomprehensible improbability and an independently given pattern that are the earmarks of design.
So, again, God’s existence is clearly more probable given the fine-tuning of the universe than it would have been without it.
What is Dawkins’s reply? He claims that chance can explain it. For he offers the tantalizing speculation popular with physicists like Lee Smolin that there are a huge (perhaps infinite) number of reproducing parallel universes, each with a different combination of initial conditions and physical constants. Those universes with black holes will reproduce more than others — these universes are also more likely to have life-permitting constants. If this is so, some universes are bound to be in the “Goldilocks range”. And Dawkins thinks that this hypothesis, though a bit extravagant and certainly speculative, is less extravagant than saying God did it, for God is an immensely complex being. And Krauss’s own response is completely analogous to the reason skeptics like Ehrman reject the resurrection of Jesus; Hume’s argument against miracles. I think that, at bottom, they are really the same response.
Let’s revisit Bayes’s Theorem:
P(G|E)/P(not-G|E) = P(G)/P(not-G) × P(E|G)/P(E|not-G)
P(G|E) is the probability of God’s existence given the evidence we are considering. P(not-G|E) is the probability of God not existing on the evidence before us. P(G) and P(not-G) is the probability of God existing or not existing just on background information; P(E|G) and P(E|not-G) is the probability of observing the evidence we do on God’s existence and God’s non-existence, that is, how much does G (or not-G) lead us to expect E, or how much does G (or not-G) explain E. What the formula says is that the ratio of the probability of God’s existence to his nonexistence is a combination of how well God leads us to expect the data as opposed to naturalism, and how probable God’s existence is just given our background information compared to naturalism.
What I think Dawkins and Krauss are saying is that even if God’s existence leads us to expect fine-tuning for life (God wants life somewhere in the universe), God’s prior probability P(G) is sooooooo low as to make God a non-starter, due to his complexity. That’s why Krauss thinks saying “God exists” is an “extraordinary claim” which requires “extraordinary evidence.” And that’s why Dawkins thinks even a wacky hypothesis like Smolin’s is better than God’s existence because, again, P(G) is soooooo low due to God’s complexity.
These are exactly the same reasons for rejecting miracles that I considered earlier, including the resurrection!! (See my post on that). So David Hume is at the bottom of all this!
Pretty much any evidence a theist might try to give for God’s existence will fall on deaf ears, for the same reasons evidence for the resurrection of Jesus will; these hypotheses have low prior probabilities.
Craig spends his time in his debates arguing for high values for P(E|G)/P(E|notG), and atheists aren’t buying because of low P(G)!
What to say about this? I think that Plantinga has given good reasons to reject Dawkins’s argument for God’s complexity implying a low P(G). Krauss doesn’t really argue for it as much as assume it. Maybe he thinks it’s obvious. God is, after all, a rather silly being.
I have written in this post about God’s alleged low prior probability. Suffice it to say that if P(G) isn’t low, the case for God is a serious one.