|
Author
|
Topic: Stephen Griffith: Irreducible Complexity
|
Moderator
Administrator
Member # 1
|
posted 25. May 2004 07:26
Irreducible Complexity
By Stephen Griffith
Abstract: Biochemist Michael Behe created quite a stir with the publication of Darwin's Black Box :The Biochemical Challenge to Evolution. In this book, Behe argues that relatively recent advances in biochemistry reveal the existence of hitherto unknown and unexpected biochemical machines and other complex systems within living organisms which exhibit a property Behe defines as "irreducible complexity". What makes the existence of such allegedly irreducibly complex features of living organisms so interesting and important is that, according to Behe, they (or, more precisely, their existence) simply cannot be explained by means of the various neodarwinian mechanisms available within the context of contemporary evolutionary theory, such as natural selection and chance mutation of genes. To put it another way, if Behe is right, neodarwinism, the currently accepted paradigm of evolutionary theory, will need to be abandoned or at least modified beyond recognition if we wish to provide a truly scientific explanation of these features. This in itself would be enough to agitate much of the scientific community, but an even greater source of agitation is that Behe eventually goes on to argue that the only possible explanation of the existence of irreducibly complex biological features of living organisms must be in terms of "intelligent design".
To read the entire paper, click here.
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 02. June 2004 05:09
Steven Griffith has presented a clear and fairly comprehensive review of Behe's arguments. The paper also purports to address many of the criticisms against using Behe's original definition of irreducible complexity as an indicator of design (or at least as an indicator of not-evolution). Unfortunately, on the latter point, the paper suffers from serious flaws.
The flaws all boil down to one thing. In each case, after the objection has been described, statements are made to the effect of, "But I can't really see how this would happen, and nobody's shoved contradictory evidence in my face, and it seems fanciful to me, so we can ignore this objection."
This is a flaw because it does not distinguish between an insufficiently imaginative and informed author, and a set of truly low-probability speculations about mechanisms of evolution.
I'm not going to look up any evidence, since one could spend weeks doing that, and that should be the author's responsibility at least initially. But I will sketch relatively plausible scenarios for each type of counterexample.
Scaffolding is perhaps most likely to occur when all the components needed are already available, but are simply misordered. For example, proteins that catalyze a wide variety of chemical reactions already exist, but their substrates are usually limited because of tightly regulated binding domains. But if we get a bunch of proteins that aren't so tightly regulated, we will create a wide variety of small molecules. Drug companies use vast libraries of small molecules to find new drugs now, so having a diversity of small molecules is a reasonable way to find a beneficial one. Thus, we could get a selective advantage (at substantial metabolic cost) simply by duplication of and accumulation of mutations in enzymes involved in existing biochemical reactions. However, once this advantage was achieved, the pressure would be to reduce the metabolic cost. Any mutation that colocalized the necessary enzymes would be selected for, as it would increase the local concentration of products and reactants. Likewise, mutations that selected for the proper reactants and products would be selected for. Those damaged proteins that were not involved in the correct synthesis pathway would be selected against. And so you'd end up with a new set of well-matched proteins devoted to the new synthesis pathway, each of which was essential--and this would be very deceptive, because it would be difficult to realize what cross-reactivity could have originally had some function. (Simple forms of this can be found in the literature.)
Cooption, if it occurs, would tend to lead to proteins that were similar in sequence being found in different complex systems. And there are sequence similarities all over the place--in a few cases the two proteins can even substitute for each other. So the rejection of cooption as "highly implausible" is highly suspect. At the very least, some other explanation for observed homologies is required, preferably with arguments as to why the homology could not have been caused by cooption. (Part of the objection to cooption is probably due to "common sense" developed while working with large mechanical objects, which tend to have tight tolerances. Biochemistry tends to have much looser tolerances and much wider cross-reactivity, which can really mess up one's physical intuitions.)
Incremental indispensibility simply notes that we may not have the simplest possible system that performs a function, even if it is irreducibly complex. That proteins can become dependent on each other is trivially demonstrated with mutations and suppressor mutations (a second mutation in a second protein that counteracts the defecit caused by the mutation in the first one--and you can find these in cases where the second protein was originally nonessential). There are plenty of cases where we might assume that the simplest possible system was quite complex (e.g. Eukaryotic RNA Polymerase II with 10+ subunits; splicing machinery with 5+ proteins), if we didn't have an example of a different, simpler system (e.g. Prokaryotic RNA Polymerase with four similar subunits; self-splicing RNA that doesn't require any splicing proteins at all). So it's not "highly implausible" that a minimal system can be elaborated into an irreducibly complex system (though the maximum extent of elaboration isn't clear at least to me), and "there's nothing simpler cause I can't think of it" has been repeatedly shown to be a bad form of argument in biology.
Reducible complexity is just an alternate way of looking at incremental indispensibility.
So at least for these three objections, the paper presented here hasn't presented what I would consider a compelling defense of IC.
IP: Logged
|
|
|
Printer-friendly view of this topic