Joined: May 2002
An interesting paragraph from a recent article that continues the debate between Behe and Shanks & Joplin, regarding the (chemical) specificity necessary for ICness vs. "simple interactive" systems or what-have-you.
Niall Shanks and Karl Joplin, "Behe, Biochemistry, and the Invisible Hand," Philo, Volume 4, Number 1.
Regarding blood-clotting & specificity, they write:
Behe bolsters his attack on the BZ reaction with a truly bizarre argument derived from the fact that the reagents in the BZ reaction have a wide variety of uses—in Behe’s terminology, they have low specificity. For example, one ingredient, sodium bromate, is a general purpose oxidizing agent, and ingredients other than the ones we mentioned can be substituted. In our reaction, we mentioned the use of cerium ions, but iron or manganese ions will work just as well. He points out that the reaction is easy to set up and runs over a wide range of concentrations.21
If this is the case, then mousetraps are not irreducibly complex either. The steel used in their construction has a wide range of uses, as does the wood used for the base. You can substitute plastic for wood, and any number of metals for the spring and hammer. Mousetraps are easy to make (which is why they are cheap) and will work with metals manifesting a wide range of tensile strengths. But the fact that they are easy to make does not mean they assemble just by chance. Mousetraps need a maker just as much as the BZ system needs chemical mechanisms governed by the laws of chemistry. Either the BZ system is an irreducibly complex system, or the complexity of mousetraps is not a model for irreducible complexity. Take your pick, for you cannot have it both ways.
This matter is made all the more acute because crucial components of Behe’s own examples of irreducible complexity have multiple uses and lack substrate specificity (interact with a wide variety of substrates). For example, plasminogen (a component of the irreducibly complex blood-clotting cascade) has been documented to play a role in a wide variety of physiological processes, ranging from tissue remodeling, cell migration, embryonic development, and angiogenesis, as well as wound healing.22 And though Behe tells us that plasmin (the activated form of plasminogen), “. . . acts as scissors specifically to cut up fibrin clots,”23 we learn in one of the very papers he cites that, “Plasmin has a relatively low substrate specificity and is known to degrade several common extracellular-matrix glycoproteins in vitro.”24
22. See Thomas H. Bugge, Keith W. Kombrinck, Matthew J. Flick, Cynthia C. Daugherty, Mary J. Danton, Jay L. Degan, “Loss of Fibrinogen Rescues Mice for the Pleiotropic Effects of Plasminogen Deficiency,” Cell 87 (1996): 709–19.
23. See Behe, Darwin’s Black Box, 88.
24. See Bugge, et al., “Loss of Fibrinogen Rescues Mice,” 709.
Blood-clotting comes up again here:
In our original article we pointed to the gene coding for the protein p53. Lab mice have been created in which this gene has been knocked out. In support of our claims about the existence of redundancy in biochemical systems, we pointed out that, though this protein was involved in a number of important biochemical and biological processes, its removal did not result in a catastrophic disruption of the developmental process. There was redundancy, and other proteins could conspire to do the work of the missing protein.
Behe acknowledges this case, but draws his reader’s attention to the blood-clotting cascade originally discussed in his book:
Yet contrast this case [p53] with that of mice in which the gene for either fibrinogen, tissue factor, or prothrombin has been knocked out. . . . The loss of any one of those proteins prevents clot formation—the clotting cascade is broken. Thus Shanks and Joplin’s concept of redundant complexity does not apply to all biochemical systems.41
Again, suppose this point is right. What is its relevance when the role of redundant complexity lies in its ability to account, in natural, evolutionary terms, for the origins of irreducible complexity? And origins, as Behe points out, is the central issue. Loss of functional genes reduces redundancy to yield an irreducibly complex system. All Behe’s example shows is that further losses at this point can catastrophically disrupt the system.
We also think, however, that Behe has oversold the irreducible complexity of the blood-clotting cascade. The cascade itself has features that manifest redundant complexity. The real situation is thus more complex than Behe’s carefully massaged description would lead you to believe. Plasminogen deficient (Plg-/-)—hence plasmin deficient—mice have been studied. As noted earlier, plasmin is needed for clot degradation (it cuts up the fibrin), yet:
Plasmin is probably one member of a team of carefully regulated and specialized matrix-degrading enzymes, including serine-, metallo-, and other classes of proteases, which together serve in matrix remodeling and cellular reorganization of wound fields. . . . However, despite slow progress in wound repair, wounds in Plg-/- mice eventually resolve with an outcome that is generally comparable to that of control mice. Thus an interesting and unresolved question is what protease(s) contributes to fibrin clearance in the absence of Plg?42
Behe cited this very paper, so we must assume that he, too, knows that parts of his clotting-cascade are redundantly complex. In this case, healing delayed is not healing denied!
41. See Behe, “Self-Organization and Irreducible Complexity,” 161.
42. See Bugge, et al. “Loss of Fibrogen Rescues Mice,” 717.
...Bugge et al. rides again! (This was a paper which Doolittle misread, or at least oversimplified, in his Boston review article, which gave Behe an opportunity to dodge the real issue, namely how Doolittle has been able to predict the presence of blood-clotting proteins in various species (with simpler systems, no less) unless Doolittle's model for the evolution of blood-clotting has significant validity.)
Brief commentary on Shanks & Joplin: while they have introduced the useful notion of "redundant complexity", and in the above 2001 Philo paper have tied the concept to the "scaffolding" model for the origin of IC (i.e., reduce redundancy and you end up with IC), I don't think that they have a general solution to the origin of IC unless they incorporate cooption/change of function into their analysis. I can only think of a few examples where "loss of scaffolding" explains the origin of an IC system, but many where cooption of a part/system to a new function explains it. Perhaps more importantly, the processes are not mutually exclusive and so in some cases both processes might operate in succession.