Back in the late eighties and early nineties there was a show on television called Macgyver. The premise of the show was that Macgyver would fashion from whatever materials lay at hand some ingenious device that would get him out of whatever jam he was in during that week's episode.
I think of Macgyver everytime I read how the Darwinians seek to explain how Michael Behe's irreducibly complex machines, like the bacterial flagellum, could have arisen purely by natural selection. The theory is called co-option (or exaptation) and advocates maintain that the several dozen protein parts of the flagellum were probably present in the cell serving other functions prior to the emergence of the flagellum. At some point these parts were "co-opted" by the cell, guided by natural selection, and incorporated into the flagellar motor. We might call this the Macgyver hypothesis.
Unfortunately for the Macgyver hypothesis, it's even more implausible, if that can be believed, than the tv show. GilDodgen at Uncommon Descent explains why:
1) In order for co-option to produce a bacterial flagellum (for example) all of the component parts must have been present at the same time and in roughly the same place, and all of them must have had other naturally-selectable, useful functions. There is no evidence whatsoever that this ever was the case, or that it ever even could have been the case.
2) The components would have to have been compatible with each other functionally. A bolt that is too large, too small, or that has threads that are too fine or too coarse to match those of a nut, cannot be combined with the nut to make a fastener. There is absolutely no evidence that this interface compatibility ever existed (between all those imaginary co-opted component parts), or that it even could have existed.
3) Even if all the parts are available at the same time and in the same place, and are functionally compatible, one can't just put them in a bag, shake them up, and have a motor fall out. An assembly mechanism is required, and that mechanism must be complete in every detail, otherwise incomplete or improper assembly will result, and no naturally-selectable function will be produced. The assembly mechanism thus represents yet another irreducibly complex hurdle.
4) Last, and perhaps most importantly, assembly instructions are required. Assembly must be timed and coordinated properly. And the assembly instructions must be complete in every detail, otherwise no function will result. This represents an additional irreducibly complex hurdle.
Co-option is a demonstrably fantastic story made up out of whole cloth, with absolutely no basis in evidence. And it doesn't withstand even the most trivial analytical scrutiny. There is not a shred of evidence that this process ever took place, or that it even could have taken place. Worst of all, it requires blind acceptance of the clearly miraculous.
There is a great irony here. This verifiably ridiculous co-option fantasy is presented as "science," while a straightforward and reasonable inference to design is labeled pseudoscience. The real state of affairs is precisely the reverse.
The problem for Darwinians is that for them, given their metaphysical assumptions about science, they have to maintain that any naturalistic explanation, no matter how far-fetched or improbable, is still preferable to an explanation that invokes intentionality. What they need to be reminded of, of course, is the fact that Macgyver was an intelligent designer.