This fourth and final installment of a critique of Elliott Sober’s recent article entitled “What is Wrong With Intelligent Design?” will show some final problems Sober’s claim that ID is not testable because, he alleges, ID can always regress to a higher level of design. In Part I, I explained some problems with Sober’s history of ID, and in Part II, I explained how Sober eschews ad hoc explanations while ignoring how modern neo-Darwinism commonly invokes them. In Part III, I explained that Sober ignores the testable predictions of ID. In this final installment I will show that Sober is wrong to claim that ID is not testable because he bases his argument on the false claim that ID permits the possibility that a designer produced a universe where natural processes can produce novel specified complexity on their own.

Designer-Regress?
Sober objects to the prediction that ID creates certain types of complexity by claiming that if “purely physical antecedents” are shown to produce complexity, then ID proponents will just claim those “purely physical antecedents” were themselves designed. In other words, he thinks that ID proponents can always appeal to higher levels of design to save the theory. His example is the printing presses, a physical machine, which can reprint intelligently designed information. His implication is that we might treat nature like a printing press–asserting that Darwinian processes themselves were designed to create information. In his view, this makes ID an untestable theory because whenever real design is lacking, we’ll always appeal to higher, untestable levels of design.

But the printing press gives an inappropriate example because of course we know that printing presses are designed, and we do not find printing presses in nature. The question is not “can processes which we know are human-designed re-transmit information and complexity?” but rather “can processes we find at work in nature generate novel specified and complex information?”

Intelligent design is making real claims, not untestable speculations like Sober puts it. William Dembski explains this point:

According to the theory of intelligent design, the specified complexity exhibited in living forms convincingly demonstrates that blind natural processes could not by themselves have produced those forms but that their emergence also required the contribution of a designing intelligence. The design found in nature therefore exhibits that nature is incomplete. In other words, nature exhibits design that nature is unable to account for.” (Dembski, The Design Revolution, pg. 147, emphasis added)

That last sentence completely refutes Sober’s argument: ID states that “nature exhibits design that nature is unable to account for.” Sober’s false, untestable version of ID might be characterized as “nature can account for its design but there’s still a designer behind it all.” Rather ID is making a much stronger, and bolder claim, saying: “nature cannot account for its own design.” According to ID, the design in nature is real, and detectable, and not explainable by nature itself. This is an eminently testable claim, and again it seems that Sober attacked only a straw-man version of ID.