Image Credit: Simon Jackson

Over at the excellent magazine Culture 11, Joe Carter has taken an opportunity to offer up Stephen Unwin’s use of Bayesian probabilities to demonstrate that it is more likely than not that God exists. The whole idea has a somewhat pseudo-intellectual gloss, but it comes with a whole host of undefined terms that unfounded assumptions that, frankly, make the entire argument meaningless.

Let’s start with the basic concept, as Carter describes it:

By applying Bayesian probabilities, a statistical method devised by 18th-century Presbyterian minister and mathematician Thomas Bayes, Urwin attempts to determine the probability of God’s existence. Since 50-50 represents the mathematically neutral, “maximum ignorance” position, Unwin begins with a 50 percent probability that God exists and then applies it to the following modified Bayesian theorem:

This formula is then used to ascertain the probability of God’s existence by considering particular lines of evidence and assigning those evidences a value “D” (which stands for “Divine Indicator”). As a quantity, D is broken down as follows:

• Evidence which is quantified as a D level of 10 is much more likely to be produced if God exists.

• Evidence with a D level of 2 is moderately more likely to be produced if God exists.

• Evidence with a D level of 1 is neutral–that is, equally likely as evidence of God’s existence or not.

• Evidence with a D level of 0.5 is moderately more likely to be produced if God does not exist.

• Evidence with a D level of 0.1 is much more likely to be produced if God does not exist.

Both Unwin and Carter then apply D values to the following lines of evidence:

• Recognition of Goodness {Unwin: 10; Carter: 10}

• Existence of Moral Evil {Unwin: 0.5; Carter: 1}

• Existence of Natural Evil {Unwin: 0.1; Carter: 1}

• Intra-natural Miracles (defined as prayers being “answered”) {Unwin: 2; Carter: 2}

• Extra-natural Miracles (defined as direct intervention of God in nature contrary to natural law) {Unwin: 1; Carter: 2}

• Evidence from Religious Experience {Unwin: 2; Carter: 2}

From these values and this formula, Unwin arrives at a probability of God’s existence at 67%. Carter arrives at a probability of 99%.

A Problem of Definition

Right away, just from the very fact that two theists arrive at very different numbers demonstrates that obvious flaw in Unwin’s methodology: it is far too subjective. It does not provide any reason or justification to use the particular terms. Moreover, this method is not testable as a consequence of its subjectivity. It essentially invites anyone, theist and atheist alike, to plug in meaningless numbers and arrive at a meaningless result.

For starters, let’s take a look at the six factors that Unwin uses:

1. Recognition of Goodness

Both Carter and Unwin use the existence of goodness in the world as very strong evidence that God exists. This no doubt would have surprised many Christian theologians, such as Augustine of Hippo and John Calvin, who argued that as a consequence of Original Sin, there was no goodness in the world at all, period, except for the Incarnation of Christ. This would have equally been a surprise to philosophers such as Aristotle and Socrates, both of whom argued that morality and goodness existent independently of the divine. So what justification is there for assigning this such a high “D” value?

2. Existence of Moral Evil

Carter and Unwin weigh this differently–Carter puts it as “neutral” and Unwin as making it less likely for there to be a God. But this doesn’t make sense at all. Presuming that Carter and Unwin’s notion of morality is based upon some type of free will, then the existence of moral evil should have the same effect on the existence of God as moral goodness, right? The fact that Unwin and Carter weigh them differently suggests to me that they are less concerned about morality as evidence of the existence of God than they are about arguing the reverse–that is, arguing that morality is impossible without God. It seems to defeat the spirit of what they’re attempting to accomplish here.

3. Existence of Natural Evil

“Natural Evil” refers, of course, to earthquakes, volcanos, storms, etc. Unwin puts this in the “God less likely” pile, while Carter puts it in “neutral.” Again, there’s no real justification provided for either assertion. Still, it’s worth mentioning that David Hume, among others, would definitely agree with Unwin. However, there are definitely strains of Christian thought which see “natural evil” as a punishment for Original Sin and therefore perfectly compatible with God. Again–the subjectivity here shows.

The list goes on with the other three factors. “Miracles” is an ill-defined term, allowing for any value you want to be plugged in. “Religious experience” is again hard to test. There are those who argue that the universality of religious experience points to there being a God, but it’s just as easy to say that the accounts of St. Paul, Mohammed, Buddha, and Joseph Smith are to wildly different for them all to have been experiencing one God, which makes it more likely that such “experiences” are hardwired in our genes, rather than providing a probability of God’s existence.

What’s Missing Here?

It’s also worth noting that there’s no real justification for including only these six factors in considering the existence of God. What about considering evidence of spontaneous order? Surely spontaneous order, especially in the biochemical world, cuts against God. At least, a personal, creator God, right? What about evidence of moral evil done in the name of God? What about the fact that there are so many different religions?

The fact of the matter is that the factors Unwin and Carter use to consider the probability of God are pretty carefully framed so that it’s actually fairly difficult to argue against the probability of God. That’s just bad science, bad theology, and bad philosophy. If you’re going to consider probabilities, you can’t only consider those factors that lead to the conclusion you already want.

Who is this God, anyway?

As we’ve seen, the fundamental flaw with Unwin’s construction of the “probability of God” is that it is entirely too subjective. The numbers are subjective, the factors are subjective, and the quantification of those factors are subjective. There’s nothing testable or reliable about this method at all. There’s simply no point to it.

The reason for the subjective and unreliable nature of this method goes to the heart of the problem with many attempts at the philosophy of religion–an unwillingness to define God. Who is God? What is his nature? What are his definable attributes? These are the things that all too many believers are unwilling to do. But until God is defined, it is impossible to determine whether God exists. This is because until God is defined, it is impossible to say what constitutes evidence for God’s existence and what doesn’t. The net result is you have a bunch of people yelling at each other in circles, making arguments for and against God without ever knowing what, exactly, they are arguing about.