I think that the “irreducible complexity” argument of Michael Behe is quite interesting; suggesting, as I understand it, that natural selection based on the accumulation of mutations cannot explain the evolution of systems composed of multiple interacting parts, where the removal or malfunction of one part impairs or eliminates the function of the whole. However, the conclusion he draws, i.e. that there has therefore to be a designer, is absolutely flawed. A more profitable conclusion would be that since conventional natural selection, which is an established and proven fact (the more moderate ID people accept this), may not be enough on its own to explain the evolution of complex traits (which is a debatable point), we need to understand better how natural selection, and possibly other evolutionary mechanisms, work in highly complex systems. That is, we should aim to improve the theory of natural selection, or replace it with a better one, but without reference to an imaginary intelligence. All science is “work in progress”. Just as physics has progressed from one partially successful paradigm to more successful, but still incomplete, paradigms, without invoking an escape clause in the form of a directive intelligence, so to should, and can, biology.