Home » Intelligent Design » The Miracle of Co-option

The Miracle of Co-option

Over at evolutionnews.org (http://www.evolutionnews.org/2006/06/whats_up_with_ronald_numbers_a_2.html) Casey Luskin comments on an article coauthored by Ronald Numbers for the Journal of Clinical Investigation (“Defending science education against intelligent design: a call to action”).

In this article we once again hear the bogus co-option argument as a proposed refutation of irreducible complexity: “…Michael Behe’s irreducible complexity argument ignores exaptation (co-option)…”

This claim of the refutation of IC by co-option is so ubiquitous that some people are actually starting to believe it. I therefore feel that it is my civic duty to refute this “refutation” of IC, which turns out to be a trivial exercise.

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.

  • Delicious
  • Facebook
  • Reddit
  • StumbleUpon
  • Twitter
  • RSS Feed

25 Responses to The Miracle of Co-option

  1. I’ve always thought it would be worthwhile for biologists (or biology students) to have some sort of engineering apprenticeship. By spending time with a design engineer (be it software, mechanical, electrical, or any other discipline), they would see first hand just what it takes to end up with a tightly integrated, functional system on the back end.

    It still astonishes me how so many in the origins community are so sloppy with their thinking regarding co-option. I particularly like your point about assembly mechanisms and instructions. I’ve commented before that Darwinians seem to be satisfied with a parts list (hypothetical at that), never paying mind to the arguably more important step of assembly – spatially and temporally.

  2. Angus Menuge does an excellent job of dealing with the co-option argument in Agents Under Fire.

    Darwinians seem to be satisfied with a parts list (hypothetical at that), never paying mind to the arguably more important step of assembly – spatially and temporally.

    So true. A bit of an eye here, a bit of an eye there, voila, the eye could easily have evolved.

    The Darwinians crack me up, they really do. It has to be a religion for them. I wish someone would do a “scientific” study on the religious nature of Darwinian belief. They always seems so amazed when someone doesn’t find their arguments compelling.

  3. Thanks Gil. I have spoken to a molecular biologist who works at a drug company that believes that our genes are pre programmed for cooption. He is a in the closet theistic evolutionist (he will lose his job if he comes out of the proverbial closet since the Chief scientist hates anything that remotely smells of design). He works on building drugs that kill viruses and believes that scientists are going to discover that the genes are anticipatory and what we call natural selection is nothing more than pre programmed intelligence. Therefore, he believes that cooption is a microcosm of change that occurs on the macro scales as well.

    Dan

  4. This is just a question from ignorance. Does the bacterium flagellum reproduce by mitosis? Whatever way it divides or reproduces how are two flagellums produced and you must now have a duplicate set of proteins that have to be assembled. You have one flagellum and then you have two. What is the process? It there a web link that might describe it?

    Thanks if anyone has an answer.

  5. Nice breakdown. Perhaps we could also add:

    5. It’s not simply enough that there be an assembly mechanism and assembly instructions. The assembly mechanism also has to be able to “read” the instructions. That means we need to already have in place some type of chemical “language”.

    Moreover, all this has got to be hereditable somehow.

  6. Check out my little technological evolution contest: http://www.uncommondescent.com.....chives/341. Co-option was singularly unimpressive, even with a $1000 cash prize dangled in front of the participants if they could come up with a really spectacular example of forward chaining co-option. Entry #2 is the one that won. –WmAD

  7. Co-option: another ‘just so’ story. Is it bedtime already?

  8. I understand it is hard to imagine how co-option might explain IC at the molecular “machinery” level. But IMO it is much easier to visualize at the ecological level. Once you can see how co-option might work there, it perhaps easier to accept that this line of reasoning might work at a lower level.

    As you know, there are many obligatory mutualistic partnerships between different species. Humans can’t live without certain micro-organisms and vice-versa, specialized plant-pollinator relationships, etc etc etc. These partnerships are IC: remove a partner-species and the system breaks down. Yet it is quite easy to see how co-option could have produced such a relationship. Before species A and B were mutualistic, Species A could perform task X and species B task Y. After the species got together it turned out that species A was better at task Y than species B and the latter was better at X than A was. Whereupon A lost ability X and B lost Y. Viola, the systemhas become IC.

    “Imagination,” “visualization,” and “seeing” are irrelevant. I presented an evidential and logical argument against a foundational precept of neo-Darwinism. If foundational precepts break down at the lowest level, the entire house of cards logically collapses. — GD

  9. I always found the notion of cooption absolutely ridiculous, and I agree with everything said above. In my opinion, the fundamental problem is the following: if we introduce cooption to overcome the problem of IC, that is the extremely low probability that an integrated system can come into existence by chance, all at a time, because function could not be present and therefore be selected in a simpler system, how probable is it that different parts be selected for different functions and then, for some miracle, happen to be just the specific parts that are necessary for a new, useful function? Why should that happen? And not only once, but millions of times (that is certainly a conservative evaluation of the number of IC machines in livig beings). We know that the evolution of an IC machine is theoretically possible, but utterly improbable. In the same way, cooption is theoretically possible, but even more unlikely than the evolution of an IC machine by mere chance.
    The important point is: both spontaneous evolution of an IC machine and its generation by cooption are processes which cannot in any way benefit from natural selection, and therefore they cannot be more likely than mere chance, that is to say that they cannot, indeed, be an explanation of any biological complexity.

  10. Thanks, Gil, for the good post. Unfortunately, we have not seen the last of the co-option nonsense. Irreducible complexity (if I may be allowed to distill it down for a moment) essentially challenges the two great pillars of the traditional evolutionary “mechanism”: (i) slight successive variations, and (ii) vast periods of time. I have argued that a retreat to the co-option position is a natural (indeed, almost necessary) reaction to Behe’s challenge.

    “. . . by robbing the Darwinist of the twin crutches of ‘slight successive variations’ and vast periods of time, irreducible complexity forces the intellectually honest evolutionist to retreat to something more abrupt, like Goldschmidt’s hopeful monster miraculously springing forth without precedent, or like the more intellectually palatable (but equally freakish) organism capable of traveling through nature, both temporally and spatially, miraculously co-opting parts for inclusion into a wonderfully integrated system.”

    http://www.evolutiondebate.info/ICReduced.pdf

  11. I think this animation of the flagellum assembly process has been posted here before, but perhaps some missed it: http://www.npn.jst.go.jp/movie.....embly.mpeg

  12. Russ,

    Thank you for the flagellum assembly process. It will take a lot of just so stories to back up how this happened.

  13. Gil,
    I agree with you, but the evidence from homology which undergirds co-option hypotheses
    was probably worth a mention: i.e., the similarities between the type-III secretory system and the flagellum. I’ve often noticed that the evidence presented for evolution is almost exclusively evidence for common descent rather than natural selection. That is, evolutionists point to extremely convincing evidence for natural relationships, and then simply assert that the step from a->b was accomplished via variation and selection. Co-option is just another example of this.

  14. 14

    Two species sidle up to each other, share jobs, and gradually each loses a function to the other.

    This explains how genomes build new information.

  15. “Two species sidle up to each other, share jobs, and gradually each loses a function to the other.

    This explains how genomes build new information.”

    Comment by avocationist — June 29, 2006 @ 11:30 pm

    LOL! Clearly you have no imagination! ;)

  16. I know everyone is going to yell at me, but I still have to ask. If a human eye is IC, then how is IC not an argument for spontaneous supernatural creation? If you just want to tell me where to read the answer to my question that would be great too.

  17. Co-option is fine- in an ID scenario. ;) Heck we observe designers co-opt parts from one functioning machine to use as parts for some other machine to get it to function.

    But in a “sheer dumb luck” scenario- forgetaboutit….

  18. gpuccio: “We know that the evolution of an IC machine is theoretically possible, but utterly improbable. In the same way, cooption is theoretically possible, but even more unlikely than the evolution of an IC machine by mere chance.”

    This reminds me of the No Free Lunch arguments of Dr. Dembski. Co-option basically says that the bacterial flagellum can evolve by RM/NS as provided we can start from something even more complex and unlikely. It really reduces to a front-loading argument.

    The Type III secretory system argument, to me, is analagous to saying that swimming across the Atlantic ocean suddenly becomes a tractable problem because we know we can rest at Bermuda on the way.

  19. If a human eye is IC, then how is IC not an argument for spontaneous supernatural creation?

    If a camera is IC, then how is IC not an argument for spontaneous supernatural creation of all cameras?

  20. a human eye is IC, then how is IC not an argument for spontaneous supernatural creation?

    Front loaded pre-programmed evolution, much like embryonic ontogeny. The embryo is not fully functional, likewise many of it’s parts, but the organs still get built. Thus organs with no immediate selective or immediate fuction can be formed in anticipation of future funciton.

    Salvador

  21. 21

    I think that one of the worst flaws of the Kitzmiller v. Dover opinion was that Judge Jones swallowed the idea of “exaptation” hook, line, and sinker. Here is an excerpt from the opinion:

    “As expert testimony revealed, the qualification on what is meant by “irreducible complexity” renders it meaningless as a criticism of evolution. (3:40 (Miller)). In fact, the theory of evolution proffers exaptation as a well-recognized, well-documented explanation for how systems with multiple parts could have evolved through natural means. Exaptation means that some precursor of the subject system had a different, selectable function before experiencing the change or addition that resulted in the subject system with its present function (16:146-48 (Padian)).” — pages 74-75 of Kitzmiller v. Dover opinion,
    http://media.ljworld.com/pdf/2.....er_342.pdf

    In my opinion, Judge Jones should have refused to hear any expert witness testimony at all because the expert witnesses could do little or nothing to illuminate the motives of the school board or the reactions of the community’s citizens. In Edwards v. Aguillard, the creation-science case, the district court judge refused to hear a “Monday morning battle of the experts” and the Supreme Court said that this refusal was proper.

    The Darwinists say ID is unscientific because it implies supernatural causation, but I cannot imagine evolution actually occurring without supernatural causation.

  22. …the theory of evolution proffers exaptation as a well-recognized, well-documented explanation for how systems with multiple parts could have evolved through natural means.

    Exaptation is superfluous.

    Although it was quickly rejected by biologists on theoretical and empirical grounds, “irreducible complexity” has remained the main staple of ID Creationism. Ironically, this argument was just recently delivered a fatal blow in the prestigious science journal Nature, where a computer simulation based entirely on evolutionary principles (undirected random mutation and selection) was shown to be able to generate “irreducibly complex” outputs. here

    Notice how imagination (co-option scenarios) suffices for an explanation until UM+S can come to the rescue. I rather think this simulation shows how IC can be generated based on ID principles.

  23. Given the absurdity of co-option (aka just so story) argument, I’m tempted to come up with my own just so story to refute it. Perhaps that will qualify as science :) Bill perhaps we should have a competition.

  24. [...] In his post, David links to my UD essay, The Miracle of Co-option, in support of the claim that there are indeed serious scientific and logical problems with certain “fertile and unifying” Darwinian principles. These icons link to social bookmarking sites where readers can share and discover new web pages. [...]

  25. [...] might also like to check out my essays on the obstacles presented by combinatorial explosion, and the willingness of Darwinists to accept storytelling as fact, with absolutely no analytical scrutiny. These icons link to social bookmarking sites where [...]

Leave a Reply