The most serious problem in Dembski’s account involves specified complexity. Organisms aren’t trying to match any “independently given pattern”: evolution has no goal, and the history of life isn’t trying to get anywhere. If building a sophisticated structure like an eye increases the number of children produced, evolution may well build an eye. But if destroying a sophisticated structure like the eye increases the number of children produced, evolution will just as happily destroy the eye. Species of fish and crustaceans that have moved into the total darkness of caves, where eyes are both unnecessary and costly, often have degenerate eyes, or eyes that begin to form only to be covered by skin—crazy contraptions that no intelligent agent would design. Despite all the loose talk about design and machines, organisms aren’t striving to realize some engineer’s blueprint; they’re striving (if they can be said to strive at all) only to have more offspring than the next fellow.
Another problem with Dembski’s arguments concerns the N.F.L. theorems. Recent work shows that these theorems don’t hold in the case of co-evolution, when two or more species evolve in response to one another. And most evolution is surely co-evolution. Organisms do not spend most of their time adapting to rocks; they are perpetually challenged by, and adapting to, a rapidly changing suite of viruses, parasites, predators, and prey. A theorem that doesn’t apply to these situations is a theorem whose relevance to biology is unclear. As it happens, David Wolpert, one of the authors of the N.F.L. theorems, recently denounced Dembski’s use of those theorems as “fatally informal and imprecise.” Dembski’s apparent response has been a tactical retreat. In 2002, Dembski triumphantly proclaimed, “The No Free Lunch theorems dash any hope of generating specified complexity via evolutionary algorithms.” Now he says, “I certainly never argued that the N.F.L. theorems provide a direct refutation of Darwinism.”
Those of us who have argued with I.D. in the past are used to such shifts of emphasis. But it’s striking that Dembski’s views on the history of life contradict Behe’s. Dembski believes that Darwinism is incapable of building anything interesting; Behe seems to believe that, given a cell, Darwinism might well have built you and me. Although proponents of I.D. routinely inflate the significance of minor squabbles among evolutionary biologists (did the peppered moth evolve dark color as a defense against birds or for other reasons?), they seldom acknowledge their own, often major differences of opinion. In the end, it’s hard to view intelligent design as a coherent movement in any but a political sense.
It’s also hard to view it as a real research program. Though people often picture science as a collection of clever theories, scientists are generally staunch pragmatists: to scientists, a good theory is one that inspires new experiments and provides unexpected insights into familiar phenomena. By this standard, Darwinism is one of the best theories in the history of science: it has produced countless important experiments (let’s re-create a natural species in the lab—yes, that’s been done) and sudden insight into once puzzling patterns (that’s why there are no native land mammals on oceanic islands). In the nearly ten years since the publication of Behe’s book, by contrast, I.D. has inspired no nontrivial experiments and has provided no surprising insights into biology. As the years pass, intelligent design looks less and less like the science it claimed to be and more and more like an extended exercise in polemics.
And yeah, that's because metaphysics isn't science.