I intended to go to this talk, given by renowned philosopher of science Elliott Sober at the University of Chicago on April 13, but missed it due to other commitments. Given the content, I’m sure I would have participated in the discussion, but am now limited to posting it and making some post facto comments (the video below was first posted on Vimeo).
The talk itself is about 48 minutes long, with about an hour of questions and discussion. I hope some readers will listen to it and comment below, for I profoundly disagree with Sober’s thesis. It seems to me that he uses a lot of words to say something that is both trivial and almost meaningless. But, naive about philosophy as I am, perhaps I’m wrong. Here’s the video including the discussion afterwards:
What is Sober’s thesis? Here’s his abstract of his talk:
Can conclusions about God – for example, that he does not exist, or that he lacks this or that property if he does exist — be deduced from well-established scientific theories? For example, do established results in biology concerning mutation show that God never guides mutations? I answer this question by explaining what biologists mean (or should mean) when they say that mutations are unguided. I make use of the fact that evolutionary theory is a probabilistic theory; its truth does not entail that it is causally complete.
His point is that although there is scientific evidence that mutations are “random,” that doesn’t mean that God can’t affect the probability of a given mutation occurring. So science cannot rule out a theistic hand in evolution. The apparent disparity between random mutations and God-guided mutations comes from the difference between how scientists and theists view “random mutations.”
To scientists, as Sober points out, mutations are “random” because the probability of their occurrence does not depend on the likelihood of their being favorable. That doesn’t mean that the rate of mutation of a given gene can’t change with environmental circumstances, or that different genes always mutate at the same rate, but simply that whether a mutation would create a new form of a gene that replicates better than its alternative forms (i.e., has higher “fitness”) doesn’t increase if that mutation would be favorable. Mutations that induce antibiotic resistance in bacteria, for instance, don’t increase in frequency when the bacteria are exposed to antibiotics.
Scientific evidence supports this contention: we haven’t found a single case of a gene whose mutation rate depends on how favorable it may be. There is some evidence from bacteria that the general mutation rate of many genes may increase in times of stress, which may be an adaptive response of the genome to changed conditions, but we have no evidence that any gene’s chances of mutating to an adaptive as opposed to a maladaptive form increase when the environment changes. So the definition of “random” mutation I gave above still stands and is accepted by evolutionary biologists.
Sober claims, however, that this cannot rule out the theistic proposition that God can affect, at least occasionally, the mutation rate. This could, for example, have been God’s way of making humans moral, and facilitating our evolution in other ways. (Sober doesn’t say this, but that’s what theists would claim). He says that the scientific proposition “mutations are unguided” does not equate with the idea that “God never guides mutations.”
Why is that? At about 38 minutes in, Sober, using a quantum-mechanics analogy, claims that God could be a “hidden variable” that affects the probability of mutations occurring, even though they seem random to us. So, just like a coin appears to have a probability of 50% of coming up either heads or tails, but in reality its fate is completely determined by forces like the force and angle of the toss, local air currents, and the surface on which it lands, so in reality mutations could appear random but some of them could have been engineered by God. In other words, God could be a “supernatural hidden variable” in the process of mutation.
Sober, then, draws a distinction between the “scientific” view of mutation (that the mutation rate in is the same for a gene regardless of the environment it’s in, which isn’t precisely true, but it’s close enough), and the “theistic” view, which he presents as this (the slide below shows the view that theists deny):
So the theistic alternative, which Sober finds completely compatible with scientists’ view of mutations as random, is that the probability of a certain mutation given that most of them are caused naturally is not equal to the probability of the mutation given that most are caused naturally but some can also be caused by God’s intervention.
Now this can’t be widely true, because if God intervenes in at least a consistent way, we’d expect that mutations in general would occur more frequently when they were more beneficial. And that isn’t observed. (This, of course, presupposes a beneficent God who hastens the progress of selection.) Alternatively, you can say that God just intervenes rarely to cause a mutation to occur (I suspect this is Sober’s position, though he’s not clear about it), so that we’re unable to distinguish God-guided mutations from those that are both natural and “random.” Under this position, rare mutations would be the same as miracles; and presumably theists would argue that these miracles were more frequent in the evolutionary lineage that produced Homo sapiens. All of this justifies what Sober calls “interventionism”: the proposition that God sticks his hand into the natural world from time to time, affecting the course of events in a way that’s undetectable.
I am not sure why Sober is making this argument, which doesn’t really say anything new (“miracles could have happened rarely in the course of evolution”) but gives theists the comfort of knowing that a reputable philosopher supports their views. Here is his summary slide:
But why does a philosopher spend his time arguing that it is philosophically cogent to accept the existence of miracles in evolution? Indeed, if you take that tack, you can’t rule out the occurrence of miracles anywhere if they’re sufficiently rare and cryptic to masquerade as natural events. It could have been a miracle that God guided the asteroid that hit the earth and caused the dinosaurs to become extinct, thereby (supposedly) setting the evolutionary stage for the radiation of mammals. We have no way of knowing for sure. Is that good philosophy? Why not also argue that God affects not just mutations, but the course of evolution itself? After all, genetic drift, another process that can cause evolution and even override natural selection, is taken to be a “random process” (based on Mendelian segregation and limited family size), but why couldn’t God have affected which allele got into which offspring? Why didn’t Sober add that to his lecture, too? Lots of evolutionary phenomena could have occurred via God’s hand, but of course there’s not a jot of evidence for any of it, or of God himself.
Indeed, during the question session (at 1:12:50), a student asks Sober why we couldn’t rule out the possibility that the appearance of the Earth’s great age is a miracle too: that maybe God created the Earth 6000 years ago but made it look old. That, too, is a divine intervention that we cannot philosophically rule out as a logical possibility. To me, that question is a devastating critique of Sober’s argument. And he doesn’t give a good answer, responding only that we have to take the scientific evidence at face value and trust what it tells us about reality (which he doesn’t do when it comes to mutations), and adding that “We are not totally deluded in what we believe about empirical evidence.” But he says we might be completely deluded when it comes to the empirical evidence about mutations!
I’d appreciate readers’ takes—particularly if they know some philosophy and biology—about whether Sober is saying anything either profound or new. My opinion is that he isn’t.
Over at EvolutionBlog in February, Jason Rosenhouse, who knows about evolution, philosophy, and probability, took apart an earlier interview of Sober in Philosopher’s Magazine. After reading about Sober’s “divine mutations” argument, Jason gives a right-on response:
Notice just how tepid Sober’s conclusions are. Hypothesizing direct divine action in evolution is consistent with the science. Evolutionary theory cannot absolutely rule out such a thing. But for me those two statements I placed in bold face [ “I think maybe you could have a reason for thinking ] that God’s intervention is] wrong”, and “I see no reason to believe in these hidden variables”] really give the game away. Near the end he notes that there is no reason to believe in the sorts of hidden variables he discusses. Earlier he tells us that there could be good reasons for thinking that God does not intervene to affect the outcomes of coin tosses. Presumably he is talking about theological reasons, since he emphasizes that empirical data is silent on such questions. Certainly, given reasonable assumptions about the nature of God’s interactions with the world, it becomes implausible to think that God is intervening in coin tosses.
But the same objection applies to God influencing specific mutations. If he is going to micromanage at that level, then why not skip the bloodsport by creating ex nihilo, precisely as the Bible says He did? Why does God operate in a way that seems tailor made to fool us regarding how nature works, by using what seem like random processes as a way of covering His tracks?
Indeed. Sober might as well have expanded his talk beyond mutations to everything in the natural world: we can’t rule out rare, unobserved miracles in biology, physics, chemistry, or anything else.
But why posit that there is a God who affects these things when there’s not a scintilla of evidence for it? Why multiply hypotheses unnecessarily, unless Sober wants to give succor to theists? Indeed, why shouldn’t Sober give a talk saying that we can’t rule out that God occasionally influences the results of coin tossing because He wants a certain result? Is that a good subject for a philosophy talk? Or why wouldn’t Sober argue that the probability of mutations, or of any other miracle, is raised by the existence of a particular unicorn who remains hidden from view?
Why is Sober doing this? The philosophy seems pretty lame to me, unworthy of someone of Sober’s intelligence. Perhaps a clue comes at 1:37:00, when Sober says he wants evolutionists and theists to realize that there is no real conflict between them. Jason quotes Sober from his Philosopher’s Magazine interview:
And on the other hand, if [Sober] manages to persuade theists, he says, “maybe the reaction against evolution will be less extreme. Theists can accept evolution and believe in God. This compatibilist view is the least visible of the views out there in the public debate. Incompatibilism is the dominant idea whether you’re talking about creationists who reject evolution or evolutionists who reject theism. I want to insert this third idea into the discussion.”
In other words, accommodationism breeds bad philosophy. I was starting to gain some respect for philosophy, and now this—from one of the most respected philosophers of science in the business. It’s set me back. My response is that of Laplace to Napoleon’s supposed question about why there was no God in Laplace’s book on celestial mechanics: “I have no need of that hypothesis.”
(Note to atheists: see the discussion starting at 1:23:15 when Bob Richards asks Elliott if New Atheists have any good scientific arguments against God’s existence. Sober avers not, but I think that the absence of evidence for God, when there should be such evidence, is indeed empirical evidence against God, just as the absence of evidence militates against the existence of pink unicorns living in New World rain forests.)
UPDATE: Alert reader Sigmund notes in the comments below that Sober’s talk is part of a lecture series on “Debating Darwin” run by Bob Richards from the University of Chicago and Michael Ruse from Florida State University. The series website says this:
The Debating Darwin workshops are a series of lectures given by the most acclaimed historians and philosophers of science. These workshops, headed by Robert J. Richards of the University of Chicago and Michael Ruse of Florida State University are co-sponsored by the Fishbein Center for History of Science, the Office of the President, and the Templeton Foundation.
Ah, good old Templeton—always ready to hand out big bucks to those who purport to effect a consilience of science and religion.