Unfalsifiable?

Dr. Novella:

Intelligent design (ID), according to the Discovery Institute, is defined as follows:

“Intelligent design holds that certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.”

The primary scientific criticism of ID is that it is not a legitimate scientific theory, but rather a transparent attempt at recasting religious faith (creationism) in scientific-sounding jargon. But ID lacks the minimal criteria to be considered science. ID proponents, of course, reject this argument because the entire purpose of ID is to masquerade creationism as a scientific theory.

Much of the discussion on this question focuses on the specific point of whether or not ID can be falsified - can it theoretically be proven false by scientific evidence. ID proponents say yes, scientists generally say no. While I agree that the answer is mostly no, the more precise answer is that it depends - it depends on exactly how ID is being formulated and practiced. I contend that in practice, ID proponents have rendered ID unfalsifiable while playing with semantics in order to pretend that it can be falsified.

In the most recent blog entry of the Discovery Institute, Jonathan Wells responds to Francis Collins on this very issue. Wells writes:

More surprising is the fact that Collins is here citing experimental evidence against a theory he maintains is unscientific because it is not open to experimental testing. In claiming that evidence from gene duplication disproves ID, Collins is demonstrating that ID can be tested with scientific evidence. Either ID is unscientific, in which case evidence is irrelevant; or evidence can be cited against it, in which case ID is scientific. Collins can’t have it both ways.

It is actually Wells who is trying to have it both ways - citing evidence against ID to demonstrate that it is falsifiable without acknowledging that it has been falsified. This is the game that they play. Pretending ID can be falsified, but then always keeping just out of reach of scientific evidence so that in practice it can never be falsified. There are actually several problems with ID that render it unscientific.

...