“…and, in the same manner…”

Analogies. They come up all the time; useful in teaching or explaining, perhaps essential to our way of viewing the world; and yet highly problematic when too much relied upon. In his summary of Behe’s argument Allen suggests intelligent design theorists have made a fatal mistake in their reasoning, and are presenting nothing but the poorest form of logical argument, an “argument from analogy”. Is this a fair criticism?

All of the examples of design Behe provides in pages 194-204 to support his definition and design detection algorithm are clearly and unambiguously designed because they are all designed by humans, and we all agree that humans can indeed design things. However, arguing that this somehow validates his definition/algorithm is simply an argument by analogy, and we have already concluded that this form of argument alone is logically specious.

I submit that this charge is itself specious; that the design hypothesis, while based on analogies in the same way all non-deductive reasoning must necessarily be, is nevertheless a valid inductive argument; fraught with the same pitfalls as other non-mathematical inductive arguments, but neither unsound nor of inferior logic.

Our reasoning is as follows: in all situations in which we have a causal history, the presence of complex specified information (or, in Behe’s case, IC systems) unequivocally entails intelligent agency. While intelligent agency is capable of producing CSI, no other causes have been shown to have that capability. The reasonable inference, then, in those situations where we observe CSI but do not have a causal history, is to infer design by intelligent agency as the best explanation.

Allen’s charge is grounded upon the “human” element in most examples of unequivocal design; he claims that this shared property moves the argument into the camp of logically specious transductive arguments. This claim is unwarranted. In any inductive argument there is a case to be made for potential dissimilarities between the set that is observed and that to which we are generalizing; indeed, simply by virtue of having been observed, the initial set will always have some common property the extended one does not. While we may take this as a warning of the uncertainty that accompanies any inductive inference, this does not in any way invalidate the argument.

Michael Behe addresses this briefly in his response to Kitzmiller:

Cellular machines and machines in our everyday world share a relevant property — their functional complexity, born of a purposeful arrangement of parts — and so inductive conclusions to design can be drawn on the basis of that shared property. To call an induction into doubt one has to show that dissimilarities make a relevant difference to the property one wishes to explain. Neither the judge nor the Darwinists he uncritically embraces have done that in respect to intelligent design.

Nor yet, as far as I can tell, has anyone here.