Monday, March 7, 2011
I’ve been going through Richard Swinburne’s and Robin Collins’ versions of the teleological argument for my Philosophy of Religion students recently. A striking point has leapt out at me. Both arguments insist that some empirically discoverable features of the universe make it more likely than not that God was responsible for it. In Collins’ case it is the fine tuned constants and values in physics that make the world hospitable to life. For Swinburne, it is the fact that there is something rather than nothing, the matter in the world is uniform and subject to uniformities of succession (lawful behavior), and that the laws of nature are relatively simple and elegant.
In both cases, it is fair to attribute to these authors the view that were it not for the efforts of God, the universe would have a radically different state than we find it in. That is, the occurrence of physics in a world left to itself is exceedingly unlikely. Another way to put it is that the non-God augmented or default state of the world is to be nothingness, or chaos. Call this N) The default state of reality (without God) is nothingness or chaos.
It’s this last statement that I find to be extraordinary. I have always been incredulous about design arguments. But the more I ponder this idea, the more stultified I become. What I cannot fathom is how anyone might claim to argue with confidence that N is true. On what possible basis, aside from a prior (question begging) assumption that orderly worlds must come from God, could one claim to possess reasons that justify N. I certainly can understand that many people have very strong intutions in favor of N, and that it is a very appealing notion. But the majority of people have the strong intuition that they can control the physical world with their minds if they just believe hard enough.
N) is an intriguing claim, certainly. But I, for one, just don’t have the sort of confidence necessary in my intuitions to proclaim that they are reliable guides to what must be real or not real at the broadest, most cosmic level. I can’t fathom what sorts of arguments or reasons one might have for thinking that N) is true. And I certainly can’t fathom being sure enough about those grounds to rest the weight of a belief in an almighty, supernatural creator on it. Would you be willing to flip a coin to decide if you have cancer? And then would you be willing to claim certainty about the result if the flip goes in favor of cancer. In general, rational people proportion the strength of their conviction that some conclusion is true to the strength and reliability of their evidence. To believe (or disbelieve) while disregarding the quality and quantity of the evidence one has is the paradigm example of irrational.
I’m no expert of matters of probability, but here are a few more thoughts that occur to me here. On objectivist accounts of probability, the way one would gather the relevant evidence to evaluate the probability of X happening giving conditions C is to look at lots and lots of cases where C occurs and then determine the rates at which X is the outcome. If X happens 90% of the time when C obtains, then predicting or postulating X in some instance of C is a really good bet. But we can’t do anything like that in the teleological argument cases above. We aren’t able to look at lots and lots of universes that either have a divine designer and don’t, and then compare the rates at which the designed universe are orderly and lawlike. If we had the data sets, we might be able to argue—in general, we have found that when a divine designer is responsible for creating a universe, that universe is orderly in 92% of cases. Furthermore, we find that in universes with no divine designer, the odds are less than 3% of order occurring. We find ourselves in an orderly universe. So, all other things being equal, we conclude that our orderly universe was most likely the handiwork of a divine designer. Or something like that. That’s all absurd, of course, because we do not have, nor we will ever be able to acquire, that sort of data about universes, order, and Gods.
On subjectivist Bayesian accounts of probability, like Swinburne and Collins are invoking, it is the information that one has and the prior beliefs that one brings to the table that leads one to make estimations of the likelihood of some hypothesis being true given some observations. Those prior beliefs and expectations, whether they are accurate or based on anything in reality, will generate some subjective expectations. Given your priors, you will find some outcomes to be quite surprising and some other outcomes to be quite predictable. On this account, a medieval doctor, if you can call them that, would be quite shocked and incredulous to find out that the plague is caused by a bacterial infection of yersenia pestis (instead of, say, an evil demon possession). Nevermind that he’d be dead wrong in this, on Bayes’ theorem, if I’m understanding this right, he’d find that hypothesis to be outlandishly improbable.
So on these estimations of probability, anything, no matter how improbable, irrational, or false it is, can turn out to be excedingly improbable. Probability here is a function of the information and expectations you have, not necessarily the facts.
Now back to N). We certainly can’t agree that N) is exceedingly probable on the basis of the previous account of probability. We don’t have the information we’d need for that. Can we even see the way clear to agree with Swinburne and Collins that N) is improbable on the subjectivist Bayesian account? I can’t really imagine non-question begging considerations that would lead me to agree here either.
Posted by Matt McCormick at 7:54 PM