If you read almost any science blog other than mine, you're probably aware of Brown University biologist Ken Miller's smackdown of Intelligent Design (ID) shill Casey Luskin, posted on Carl Zimmer's Loom: part 1, part 2, and part 3.

At issue is the tired old concept of irreducible complexity, and it's amazing that after all this time, many ID advocates don't understand what the original point of arguing irreducible complexity was. ID advocate Michael Behe, in various publications including his book Darwin's Black Box basically argued that there are molecular systems inside of cells that, even in principle could not have been produced by evolution - systems like the bacterial flagellum and the blood clotting cascade. Such systems, according to Behe, are irreducibly complex - they need all of their parts in order to function, and if you're missing any parts, you have a non-functional system. Thus, without all of the parts there is nothing functional for natural selection to act on.

In other words, the only way evolution could produce a system like the blood clotting cascade would be to have all of the relevant genes suddenly appear at one time by mutation - an event improbable to the point of impossibility (which is one thing ID advocates and evolutionary biologists agree on).

One major rebuttal to Behe's argument is that the examples of irreducible complex systems he cites aren't really irreducibly complex: evolution didn't have to produce the mammalian blood clotting cascade in one fell swoop; you can find in nature examples of clotting systems that work just fine with many fewer proteins, suggesting that, yes, it is possible to produce the mammalian blood clotting system via step by step evolution.

But here's where Casey Luskin gets confused:

Arguments about the irreducibly complex core aside (see Part 1), the entire land-dwelling vertebrate system might still be irreducibly complex, for the fact that some vertebrates lack factor XII (called Hageman Factor) in their blood-clotting cascade, or even lack the entire intrinsic pathway, in no way implies that humans (and other land-dwelling vertebrates) don't require all of these components for their blood to clot.

As Miller points out, Luskin has the argument ass-backwards. The fact that a particular clotting factor is now absolutely essential in humans does not at all prove that such a system could not have evolved. In the distant past, one of our jawless fish ancestors probably had a simple clotting system (just like lampreys today), and it's not difficult to imagine how more components could have been added to the system one by one over time.

These new blood clotting components would have been merely beneficial at first but essential later. It's not difficult for a gene to become essential with some simple rewiring, and we can find many examples in the genomes of present-day organisms. Two homologous genes can perform similar functions, which means that if one gene becomes damaged, the other can take up the slack. Over time, these two genes can specialize. That means instead of a backup copy of a key gene performing multiple functions, you have just two key genes, each fulfilling a more specialized function, with no backup copy. In fact, there is strong evidence that gene duplication and functional specialization is responsible for much of the evolution of the blood clotting cascade in vertebrates.

This is why Factor V is essential for blood clotting in humans but not in lampreys: Factor V never appeared in the evolutionary lineage leading to lampreys, so lampreys aren't bothered by its absence, while in most other vertebrates Factor V has become essential after a long period of evolutionary tweaking.

Remember, irreducible complexity was supposed to be an in principle argument against evolution - the claim was that the human blood clotting cascade could never have evolved step by step via 'intermediate'* systems with fewer components, but the fact that lampreys survive just fine with a perfectly functional intermediate system means that the in principle argument of irreducible complexity is wrong: intermediate systems are possible, and they exist in nature. The fact that humans can't survive with an intermediate system is, in spite of what Luskin thinks, irrelevant.

If ID advocates can't even get their own side's arguments straight, what does that say about their ability to coherently critique evolution?

*I'm using the term intermediate in a somewhat sloppy way - the lamprey clotting system isn't necessarily an intermediate step to anything, nor is the human clotting cascade a 'final' system.