The most conspicuous feature of the concept of IC is not so much its ambiguity, but the discrepancy between what it seems to promise and what it eventually delivers, as far as testable empirical claims are concerned. On first reading Behes argument, the unsuspecting reader may be left with the impression that Behe really sticks his neck out and presents evolutionists with a clear empirical challenge. However, this apparent rigour of the IC concept as an objective criterion for design, which arguably makes it appealing to anti-evolutionists, evaporates upon closer inspection. Under the weak interpretation, the concept describes a well-known phenomenon in the living world that is unproblematic for evolutionary theory. Under the strong interpretation, IC systems would indeed confront evolutionary theory with serious problems, but Behe has not given us an inkling of how we could ever demonstrate whether a system qualifies as IC in this sense. Indeed, it would require ruling out any conceivable evolutionary history, and would thus amount to showing that no part or precursor of the system in question is able to perform any other function, in any other situation and at any time.
This allows for an interesting bait-and-switch strategy, which one could summarize as follows: First, present evidence for weak IC in the living world, then pretend that strong IC has been demonstrated and continue to equate IC with unevolvability. If challenged on empirical grounds, jump back to the weak version and claim that your critics are misrepresenting your argument. Switch the IC claim to subsystems and assembly of components, keep raising the standards of evidence, and reassert that all this directly follows from the simple objective criterion of IC. Finally, when really pressed against the wall, give up this particular system and quickly find a new one. Repeat the circle ad libitum.
