Some systems are irreducibly complex (Talk.Origins)
- Some biochemical systems are irreducibly complex, meaning that the removal of any one part of the system destroys the system's function. Irreducible complexity rules out the possibility of a system having evolved, so it must be designed.
Source: Behe, Michael J. 1996. Darwin's Black Box, New York: The Free Press.
CreationWiki response: (Talk.Origins quotes in blue)
Irreducible complexity can evolve. It is defined as a system that loses its function if any one part is removed, so it only indicates that the system did not evolve by the addition of single parts with no change in function. That still leaves several evolutionary mechanisms:
- deletion of parts
- addition of multiple parts; for example, duplication of much or all of the system (Pennisi 2001)
- change of function
- addition of a second function to a part (Aharoni et al. 2004)
- gradual modification of parts
[note that the numbering is added for easy reference; T.O has them bulleted instead]
This confuses both (a) what Irreducible complexity says, and (b) what kind of evolutionary pathway such a thing must take. As for (a), it does not say it loses its function if _any_ one part is removed, but that it includes irreducibly complex parts. Let's look at Behe's own definition instead:
By irreducibly complex I mean a single system composed of several well-matched, interacting parts that contribute to the basic function, wherein the removal of any one of the parts causes the system to effectively cease functioning.
As you can see, when he is talking about the removal of "any one of the parts" he is talking specifically about the well-matched, interacting ones that contribute to the basic function. If X is an IC process, adding Y to improve it does not make the result non-IC. It just means that Y is not part of the basic function.
Now, let's look at their claimed list of ways that an IC system can evolve. (1) is silly. If you delete parts from a system, the subsystem that remains must have been there from the start. Saying that "I destroyed X but now have only Y" does not say how you got Y to begin with. Remember that biochemically not only do all the enzymes have to be there, but they all have to function in just the proper manner. I do not get a car just by removing the right parts from a jet liner, even if the jet liner has enough parts to make a car. I would then, after removing the parts, assemble them to get them in the right order. Thus (1) will not produce an irreducibly complex system. (2) is irrelevant. Duplicating system X just gives you two system X's. It does not result in a new system. (3) is also irrelevant, because it still does not tell you how the whole system got into place. Same with (4). (5) is incorrect, because first of all, this is contrary to what irreducible complexity is saying, and second, irreducible systems, though the parts can change, cannot come about on their own by such a process.
Okay, so let's move on:
All of these mechanisms have been observed in genetic mutations. In particular, deletions and gene duplications are fairly common (Dujon et al. 2004; Hooper and Berg 2003; Lynch and Conery 2000), and together they make irreducible complexity not only possible but expected. In fact, it was predicted by Nobel-prize-winning geneticist Hermann Muller almost a century ago (Muller 1918, 463-464). Muller referred to it as interlocking complexity (Muller 1939).
Because their premise is wrong, their conclusion is wrong. No one is claiming that deletions and duplications aren't common in the genome. But they are totally irrelevant.
Evolutionary origins of some irreducibly complex systems have been described in some detail. For example, the evolution of the Krebs citric acid cycle has been well studied; irreducibility is no obstacle to its formation (Meléndez-Hevia et al. 1996).
Obviously Talk.Origins is completely ignorant of both Behe's work AND the paper they are citing. First of all, Behe lists the Krebs cycle as NOT being irreducibly complex. So this whole paragraph is useless anyway -- Behe agrees with T.O and said so in his book. In addition to that, the paper itself does NOT talk about whether or not the Krebs cycle is irreducibly complex. What it talks about is the Krebs process occurring a step at a time, not the evolutionary pathway to get to the Krebs process. It would really help Talk.Origins case if they actually bothered to read (a) what they are arguing against, and (b) what papers they are citing in support of their position.
Even if irreducible complexity did prohibit Darwinian evolution, the conclusion of design does not follow. Other processes might have produced it. Irreducible complexity is an example of a failed argument from incredulity.
Talk.Origins continually confuses an argument from incredulity with an argument from design. The argument from design is "we know what design looks like, does this match our knowledge of design?" The argument from incredulity is "we don't know how this could have happened, it must be design." Michael Behe uses the former, while Talk.Origins continually claims that he and other ID'ers use the latter. The argument from IC to design is twofold: (1) it doesn't match Darwinism, and (2) it does match design. Talk.Origins here is simply plugging their ears at the second claim.
Irreducible complexity is poorly defined. It is defined in terms of parts, but it is far from obvious what a "part" is. Logically, the parts should be individual atoms, because they are the level of organization that does not get subdivided further in biochemistry, and they are the smallest level that biochemists consider in their analysis. Behe, however, considered sets of molecules to be individual parts, and he gave no indication of how he made his determinations.
Wow, Talk.Origins doesn't know what a part is. Usually parts are defined in terms of the designer and the tools available. When I design a program, I don't refer to the atoms of the computer individually, though it appears that Talk.Origins thinks that I should. The fact that biochemists consider them in their analysis is irrelevant, as computer designers also consider atoms in their analysis. It just happens to be completely irrelevant to the design that takes place on the computer. I think that Behe's use of gene products as parts is quite appropriate, since that is the level of detail that cellular functions usually operate at.
Systems that have been considered irreducibly complex might not be. For example:
- The mousetrap that Behe used as an example of irreducible complexity can be simplified by bending the holding arm slightly and removing the latch.
This involved redesigning other parts to perform the same function. There is still a latch, it is now not a separate piece but part of the hammer and it is not at all obvious that the design is any simpler.
- The bacterial flagellum is not irreducibly complex because it can lose many parts and still function, either as a simpler flagellum or a secretion system. Many proteins of the eukaryotic flagellum (also called a cilium or undulipodium) are known to be dispensable, because functional swimming flagella that lack these proteins are known to exist.
The number of non-sequiturs in this is amazing. (1) being able to lose a part does not make something not IC. Refer to the definition above. (2) if losing parts changes its function, then it would still classify it as IC. (3) the existence of flagella that lack some of the proteins is irrelevant if they are not part of what is irreducibly complex. (4) the existence of flagella that lack some of the proteins that are part of what is considered irreducibly complex is irrelevant if those flagella operate with a different basic mechanism as the flagella described by Behe. A contradiction of Behe's argument would require that the flagella operate on the same principles as the one Behe describes as well as remove some of the proteins that Behe describes as part of what is irreducibly complex.
- In spite of the complexity of Behe's protein transport example, there are other proteins for which no transport is necessary (see Ussery 1999 for references).
That is irrelevant.
- The immune system example that Behe includes is not irreducibly complex because the antibodies that mark invading cells for destruction might themselves hinder the function of those cells, allowing the system to function (albeit not as well) without the destroyer molecules of the complement system.
As a test, I encourage the Talk.Origins member making this claim to hinder his own immune system in the way proposed :)