Richard Dawkins on Design Detection as Good Science

In The Ancestor's Tale, Richard Dawkins calls design detection a "genuinely important question":

[I]t is possible to imagine validly using some version of the argument from design, or the argument from irreducible complexity. Future visitors from outer space, who mount archaeological digs of our planet, will surely find ways to distinguish designed machines such as planes and microphones, from evolved machines such as bat wings and ears. It is an interesting exercise to think about how they will make that distinction. They may face some tricky judgements in the messy overlap between natural evolution and human design.

 …

Given that the illusion of design conjured by Darwinian natural selection is so breathtakingly powerful, how do we in practice, distinguish its products from deliberately designed artifacts? Jacques Monod began his Chance and Necessity in similar terms. Could there be genuinely persuasive examples of irreducible complexity in nature: complex organization made up of many parts, the loss of any one of which would be fatal to the whole? If so, might this suggest genuine design by a superior intelligence, say from an older and more highly evolved civilisation on another planet?

It is possible that an example of such a thing might eventually be discovered. pp. 549-550. 

After discussing Francis Crick's theory of Directed Panspermia (seeding life on earth from outer space), and saying it raises interesting questions, he noted:

Crick was talking about superhuman, not supernatural, design. The difference really matters. On Crick's world view, superhuman designers of bacteria, or of the means to seed Earth with them, would themselves have originally evolved by some local equivalent of Darwinian selection on their own planet. p. 552.

Dawkins gives no serious attempt to explain how we would determine the identity of the designer scientifically. He does not say what a scientist is to do if she detects design but cannot identify the designer. He does not give clear criteria for distinguishing designed from non-designed objects. But he is clear that the question is "important" and gives no indication or reason why it would not be scientific.

For a video of Dawkins saying much the same thing to Ben Stein, go here.