If you read biology or medicine-related blogs you may have heard of creationist neurosurgeon Michael Egnor, who over the last few months has engaged in a debate, of sorts, with a number of bloggers over the theory of evolution. His latest post on the Discovery Institute web site was meant to provide an example where use of the design inference was of aid to biological research, using Watson and Crick's work on DNA.
Orac of Respectful Insolence addressed this matter already, and I was only going to comment on it, but I can't seem to get any posts written for this silly blog thing, I decided to turn the comment into a post for my own blog.
Many creationist arguments boil down to the inference "it looks designed, therefore it was designed." As is frequently pointed out, "looking designed" is a squirrelly thing to define - things like specified complexity and irreducible complexity turn out not to be as useful for inferring design as creationists claim. Egnor seems to take a different approach. Rather than trying to define "looks designed," he tries to argue that whatever it means for an organism to look designed, that's how biologists treat it, even evil evolutionists like Watson and Crick, and therefore, the assumption of design must be useful.
As Orac points out, however, that can apply to anything. In essence, what Egnor did is turn "methodological naturalism" on its head. Science assumes that phenomena can be explained without recourse to supernatural causes, that the universe is consistent, coherent, and logical. Egnor simply defines this as assuming design rather than assuming natural consistency, et voila, Jesus returns.
But let's look at the second half of the creationist inference, which I haven't seen as much commentary on lately: "Therefore, it was designed." The conclusion doesn't follow from the premise.
Many things in biology have the appearance of design. If you were to find some new species that hadn't been described before, and upon dissecting it, you found some new organ you hadn't seen in other creatures, one of the first questions you'd ask is "what's this thing do? What's it's function?" The creationist argues that this question implicitly assumes the organ was designed. So, that the organ has the appearance of design - that it has a function - is something that demands an explanation.
Of course, the theory of evolution provides one: structures that appear in animals are functional because the selection process favors functional structures and tends to eliminate costly but non-functional structures, where "functional" depends on the current environment and challenges faced by the animal (or a population of them).
This explanation can be tested. It has nontrivial implications that a biologist can look for. For example, an unusual organ in an animal would have to have a history. A biologist should be able to look at related animals and find similar structures (homology). Indeed, the appearance of a completely new organ would suggest that you've discovered not only a new organ, but some new group of animals.
Unfortunately for intelligent design advocates, there's no indication of how one gets from the "appearance of design" to "actual design" without simply asserting - without proof - that only actual design can produce the appearance of design. Creationists like Egnor apparently seem to think that its sufficient to describe how something looks designed, and then they think they've won a point.
How would they bridge this gap? Unless you catch them on a Sunday, most ID advocates won't tell you who the designer is, what methods it might have used, or what its motivations might be. There is, basically, no test you can do to determine if the Intelligent Designer is really there or not, and the ID advocates won't provide one.
So a biologist can certainly look at something in an organism and reasonably ask "what is this for?" without making any "design" assumptions. The question is just shorthand for "what functional advantage did this (and its antecedents) provide to the organism (and its ancestors)?"
Dr. Egnor should strive for a clearer definition of what it means to assume design in science, distinct from either assuming any systematic structure, or assuming some function. To most of us, that means an explanation which depends on finding and demonstrating a designer.