Home » Intelligent Design » Carl Zimmer on Irreducible Complexity

Carl Zimmer on Irreducible Complexity

Darwinist Carl Zimmer, who maintains a blog over at the Discover Magazine website recently posted this little diatribe on irreducible complexity.

Oh No! I’ve Seen the Impossible! My Eyes!
Ah, the things you learn from creationists…

If you’ve ever read about intelligent design (a k a “the progeny of creationism”), you’ve probably encountered their favorite buzz words, “irreducible complexity.” If you take a piece out of a complex biological system (like the cascade of blood-clotting proteins) and it fails to work, this is taken as evidence that the system could not have evolved. After all, without all the pieces in place, it couldn’t work.

Scientists have shown over and over again that this is a false argument. At the famous intelligent-design trial in Dover in 2005, Pennsylvania, for example, Brown biologist Ken Miller showed how dolphins and other species are missing various proteins found in our blood-clotting cascade, and they can still clot blood.

Three years later, the creationists are still trying to salvage irreducible complexity. This generally involves a bait-and-switch game. Today, for example, the Discovery Institute tells us that the evidence of dolphins does not touch the argument for irreducible complexity. See, what you have here are two different irreducibly complex systems, with one that just happens to have an extra part. Just think about bicycles…

“Bicycles have two wheels. Unicycles, having only one wheel, are missing an obvious component found on bicycles. Does this imply that you can remove one wheel from a bicycle and it will still function? Of course not. Try removing a wheel from a bike and you’ll quickly see that it requires two wheels to function. The fact that a unicycle lacks certain components of a bicycle does not mean that the bicycle is therefore not irreducibly complex.”

Of course not. No. It’s not as if five seconds of googling could turn up a bicycle that still functioned without both wheels…

You’ll need to click the link above to see the photos and Youtube he includes as “evidence” of functional bikes with only 1 wheel. I’ll get to that in a moment.

First, I just love how he vigorously hand waves IC away with “Scientists have shown over and over again that this is a false argument.” Oh really? Perhaps the good Dr. Z would be so kind as to provide a bibliography listing all the peer reviewed scientific research studies that provide the detailed, testable (and potentially falsifiable) biological models for any of the IC systems that Mike Behe described in his ground breaking book Darwin’s Black Box. (hint: its about a dozen years since Behe first published his book and made the claim that no such research existed — and 12 years later, that’s still the case!) And please don’t give us references to computer games like the famous Avida study by Lenski et.al. or the review article by Pallen and Matzke. Scientists have not shown us any such thing even once let alone “over and over again.”

Dr. Z is one to talk of “bait and switch” games. He references part 2 of this article at the Discovery Institute website, but fails to mention part 1.
Perhaps if he’d read both parts he’d know that the bait and switch is accomplished not by the DI, but by his fellow Darwinian Ken Miller, as aptly explained in the entire article.

Finally, he resorts to “disproving” (falsifying?) the bicycle analgy by providing a photo AND a bonus video of a “functional” bicycle with only 1 wheel. Not to be picky (alright I am being picky) but if you look closely at the photo you’ll notice it isn’t just the front wheel that’s missing from this bicycle, but the entire front wheel assmembly, including the handle bars and wheel frame. In other words, its a modified unicycle. So what’s the point here? That we could just add a wheel to the functioning unicycle and have a fully functional bicycle? To what will the wheel be attached? Well, what if we add just the wheel frame first? Well what’s that attached to? Hmmm….seeem that the leap from UNI to BIcycle requires a whole lot more than a simple 1 part at a time addition. Not much use to an extra wheel with no place to attach it. No good to have the place to attach it with no wheel. etc etc. Thanks, Dr. Z., for demonstrating once again that IC doesn’t come cheap! But of course, scientists have shown “over and over again” that it does.

The video is virtually irrelevent because all it proves is that you can do a wheelie with a bicycle. (we all knew that as kids!) As soon as that front comes down the rider will find out how non-functional his 1 wheeled bicycle is. The front wheel assembly will be quite detrimental without the wheel!

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

62 Responses to Carl Zimmer on Irreducible Complexity

  1. DonaldM wrote: “In other words, its a modified unicycle.”

    No, it is very obviously a modified bicycle. Many parts have been removed from a standard bicycle, making it into a marginally functional unicycle. But it is clearly NOT a unicycle that has been modified.

    Both Pharyngula and Panda’s Thumb have picked up on this latest gaffe by Casey, and are having a wonderful time discussing it.

  2. You’ll need to click the link above to see the photos and Youtube he includes as “evidence” of functional bikes with only 1 wheel

    OK, Dr. Zimmer starts his rebuttal by calling names, a strong indicator of desperation.

    Then he uses to rebut IC something that is, well, an invention. Repeat unicycles are inventions. Inventions are designed objects.

    But the really great kicker is that unicycles likely ARE DEVOLUTIONS!!!! IOW The bicycle actually came first, and the unicycle is based on the bicycle.

    Dr. Zimmer should make his handle Booji Boy. Are we not men?

  3. *zoooooooom!*

    That’s the sound of the argument going way over Zimmer’s head. In order to respond he’s forced to make a mockery of himself by twisting the major points with an over-simplified distortion.

    The unicycle and bicycle share components and have similar functionality but they are not the exact same systems. (Although Darwinists are fond of ignoring arguments and using word games instead. As if they can just as easily define reality to be their preference…) And notice that even the unicycle has an IC core set of components.

    There’s also a difference between components necessary for functionality and components necessary for efficiency. Even in that picture there are further components that could be removed without destroying functionality. The seat, several screws, foot pedals (just use the metal attachment on the crank), rubber tire (just ride on the metal clincher rim), rubber tread, and inner tube are all necessary for efficiency and usability but do not represent the IC core. Although I’d be curious if anyone could actually ride the unicycle without all that.

    I’d like to see Zimmer propose a gradual stepwise pathway between a unicycle and a full-fledged bicycle that does not drastically impair the usability and functionality of the unicycle before it becomes a bicycle. Or require imaginative reworking of components in order to make it appear “simple” (why should a frame be shaped that way?). Or presume that a bicycle could be accomplished in a couple steps by assuming a large set of components coming together at once. For example, some Darwinists seem to think all you need is a modified frame, handles, and a wheel. Ha! As if it were that simple.

    Obviously, an extension of the frame is doable since while usability suffers it does not completely inhibit functionality. But why should the frame be extended in such a manner that it would eventually allow for the attachment of handles? Why should the handles allow for ease of balance like in that video? Why have steering forks? Spokes? The hub, depending on the design, has 12 components just by itself. Even if we assume that the tire, spokes, ball bearings, inner tube, etc. all came into place why should we assume a steering tube that allows the bicycle to be useful at all?

    In sum, Darwinists like to ignore the hard questions of reality and focus on trivial examples which we ID proponents would expect in the first place.

  4. Paul –No, it is very obviously a modified bicycle. Many parts have been removed from a standard bicycle, making it into a marginally functional unicycle. But it is clearly NOT a unicycle that has been modified.

    Paul, I just want to emphasize my point in my previous post. What Zimmer is using as an example is a one-wheeled human-powered vehicle or a fully (not marginally) functional unicycle.

    Note also that Wiki goes on to describe unicycles as similar to, “but less complex than, bicycles” and then explains how they are likely the descendants of penny-farthing bicycles not vice-versa.

    D E V O

  5. Heh, sort of like the flagellum -> T3SS (presuming the historical narrative is true).

  6. OK, Dr. Zimmer starts his rebuttal by calling names, a strong indicator of desperation.

    ==

    And then you finish your post by saying, “Dr. Zimmer should make his handle Booji Boy. Are we not men?”

    So apparently it’s okay if you FINISH by calling names?

    ==

    Then he uses to rebut IC something that is, well, an invention. Repeat unicycles are inventions. Inventions are designed objects.

    ==

    When an ID proponent gives as an example of irreducible complexity something that is *known* to be a human invention, why is it unacceptable to use another human invention to show that it is not irreducibly complex? The fact that bicycles are the result of “intelligent design” is not under debate here. The discussion is about the bankruptcy of the “irreducible complexity” concept.

    Carl Zimmer is not arguing that bicycles evolved independently of human minds, and that’s not what his counter-example was intended to show. His intention was to demonstrate that a bicycle can be functional when some of its parts are missing.

    And I’ll add that Zimmer didn’t even need to use the obvious “exaptation” argument against ID here. For example, a bicycle missing one of its pedals is also perfectly useful as a pushed conveyance for cargo.

    ==

    But the really great kicker is that unicycles likely ARE DEVOLUTIONS!!!!

    ==

    A dodge. Zimmer’s argument has nothing to do with true unicycles. Luskin specifically said:

    “Try removing a wheel from a bike and you’ll quickly see that it requires two wheels to function.”

    Zimmer’s counter-example disproves this claim.

    Incidentally, the word “devolution” is being used here in a misunderstanding of evolutionary theory. A change in traits over time (more specifically in alleles in a population) is *evolution*. “Devolution” is a *kind* of evolution, not a process that runs contrary to it.

  7. Also, I wonder why they focus on the bike analogy instead of the blood-clotting cascade, which is the real issue at hand? Behe expressly limited his argument in Darwin’s Black Box to a specific sector of this biochemical pathway which is composed of an IC core set of components. All rebuttals I’ve seen so far focus on a different sector that Behe excluded from his discussion of the IC of the blood-clotting cascade.

    Here’s Behe’s testimony:

    # Q. — and Doolittle and Davidson, et al, to argue against the irreducible complexity of the blood clotting system. Do you agree with his assessment of those studies?

    # A. No, I do not.

    # Q. And you have some diagrams to explain this further, sir?

    # A. Yes, I do. This is a slide from Professor Miller’s presentation showing work from Jiang and Doolittle. And he also shows a diagram of the blood clotting cascade. And notice again, it’s a branched pathway with the intrinsic pathway and the extrinsic pathway. And Professor Miller makes the point that in DNA sequencing studies of something called a puffer fish, where the entire DNA of its genome was sequenced, and scientists looked for genes that might code for the first couple components of the intrinsic pathway, they were not found.
    And so Professor Miller demonstrated that by — if you could push to start the animation — Professor Miller demonstrated that by having those three components blanked out in white. Nonetheless, puffer fish have a functioning clotting system. And so Professor Miller argued that this is evidence against irreducible complexity.

    But I disagree. And the reason I disagree is that I made some careful distinctions in Darwin’s Black Box. I was very careful to specify exactly what I was talking about, and Professor Miller was not as careful in interpreting it. In Darwin’s Black Box, in the chapter on blood clotting cascade, I write that, a different difference is that the control pathway for blood clotting splits in two. Potentially then, there are two possible ways to trigger clotting.

    The relative importance of the two pathways in living organisms is still rather murky. Many experiments on blood clotting are hard to do. And I go on to explain why they must be murky. And then I continue on the next slide. Because of that uncertainty, I said, let’s, leaving aside the system before the fork in the pathway, where some details are less well-known, the blood clotting system fits the definition of irreducible complexity. And I noted that the components of the system beyond the fork in the pathway are fibrinogen, prothrombin, Stuart factor, and proaccelerin. So I was focusing on a particular part of the pathway, as I tried to make clear in Darwin’s Black Box. If we could go to the next slide. Those components that I was focusing on are down here at the lower parts of the pathway. And I also circled here, for illustration, the extrinsic pathway. It turns out that the pathway can be activated by either one of two directions. And so I concentrated on the parts that were close to the common point after the fork.

    So if you could, I think, advance one slide. If you concentrate on those components, a number of those components are ones which have been experimentally knocked out such as fibrinogen, prothrombin, and tissue factor. And if we go to the next slide, I have red arrows pointing to those components. And you see that they all fall in the area of the blood clotting cascade that I was specifically restricting my arguments to. And if you knock out those components, in fact, the blood clotting cascade is broken. So my discussion of irreducible complexity was, I tried to be precise, and my argument, my argument is experimentally supported.

    # Q. Now just by way of analogy to maybe help explain further. Would this be similar to, for example, a light having two switches, and the blood clotting system that you focus on would be the light, and these extrinsic and intrinsic pathways would be two separate switches to turn on the system?

    # A. That’s right. You might have two switches. If one switch was broke, you could still use the other one. So, yes, that’s a good analogy.

    # Q. So Dr. Miller is focusing on the light switch, and you were focusing on the light?

    # A. Pretty much, yes.

    # Q. I believe we have another slide that Dr. Miller used, I guess, to support his claim, which you have some difficulties with, is that correct?

    # A. Yes, that’s right. Professor Miller showed these two figures from Davidson, et al, and from Jiang, et al, Jiang and Doolittle, and said that the suggestions can be tested by detailed analysis of the clotting pathway components. But what I want to point out is that whenever you see branching diagrams like this, especially that have little names that you can’t recognize on them, one is talking about sequence comparisons, protein sequence comparisons, or DNA nucleotide sequence comparisons. As I indicated in my testimony yesterday, such sequence comparisons simply don’t speak to the question of whether random mutation and natural selection can build a system. For example, as I said yesterday, the sequences of the proteins in the type III secretory system and the bacterial flagellum are all well-known, but people still can’t figure out how such a thing could have been put together. The sequences of many components of the blood clotting cascade have been available for a while and were available to Russell Doolittle when he wrote his essay in the Boston Review. And they were still unhelpful in trying to figure out how Darwinian pathways could put together a complex system. And as we cited yesterday, in Professor Padian’s expert statement, he indicates that molecular sequence data simply can’t tell what an ancestral state was. He thinks fossil evidence is required. So my general point is that, while such data is interesting, and while such data to a non-expert in the field might look like it may explain something, if it’s asserted to explain something, nonetheless, such data is irrelevant to the question of whether the Darwinian mechanism of random mutation and natural selection can explain complex systems.

  8. sort of like the flagellum -> T3SS (presuming the historical narrative is true).

    Exactly. The rebuttal to the IC of the flagellum requires eukaryotes to have appeared before prokaryotes, which, if true, ought to falsify the ToE :-)

  9. Yellowshark,

    That was quick…let the word games and distortions begin!

    His intention was to demonstrate that a bicycle can be functional when some of its parts are missing.

    So he demonstrated the obvious that Behe and others–including Luskin–would not disagree with.

    Since you did not like this statement:

    “Try removing a wheel from a bike and you’ll quickly see that it requires two wheels to function.”

    Perhaps you can explain how a bicycle with a missing wheel is still a bicycle (or restated: able to function as a bicycle)?

  10. yellowshark

    A dodge. Zimmer’s argument has nothing to do with true unicycles. Luskin specifically said:

    “Try removing a wheel from a bike and you’ll quickly see that it requires two wheels to function.”

    Zimmer’s counter-example disproves this claim.

    No it doesn’t. First of all, Luskin didn’t specify which wheel. Sure, in the photo you can have a sort of makeshift unicycle…but that is hardly a functional BI-cycle. Suppose it was the rear wheel removed….what then? And as I pointed out, it wasn’t just the wheel removed, but the entire front wheel assembly…handle bars and all.

    Luskin is exactly correct…remove a wheel from a BI-cycle, and you no longer have a funtioning BI-cycle.

  11. TheYellowShark — OK, Dr. Zimmer starts his rebuttal by calling names, a strong indicator of desperation. . . .And then you finish your post by saying, “Dr. Zimmer should make his handle Booji Boy. Are we not men?” . . . So apparently it’s okay if you FINISH by calling names?

    Well, the idea is to start with your strongest argument and if you start with the name calling it’s a certain sign that you don’t have a very strong argument. It doesn’t really matter what you finish with after you make your point.

    But the suggestion to change his handle to Booji Boy is not really name calling . For instance if I said don’t listen to Dr. Zimmer because he’s a “Booji Boy” (or a creationist) that would be name calling and would reflect rather poorly on me I think.

    OTOH, if I said “Hey Doc, you are advocating devolution. You ought to call yourself Booji Boy” you’d have be be pretty sensitive to think you were being the subject of a dismissive or personal attack. I’m assuming, of course, you know who Booji Boy is.

    When an ID proponent gives as an example of irreducible complexity something that is *known* to be a human invention, why is it unacceptable to use another human invention to show that it is not irreducibly complex?

    I don’t find it unacceptable. I think it’s great that you do. Keep up the good work :-)

    A dodge. Zimmer’s argument has nothing to do with true unicycles

    It merely involves one-wheeled human-powered vehicles which are . . . ?

    Luskin specifically said: “Try removing a wheel from a bike and you’ll quickly see that it requires two wheels to function.”

    He obviously didn’t know the history of unicycles (and clearly was not thinking when he said that.)

    But let’s get to point I think he was trying to make: does IC exist in anything or is it a fallacy?

  12. 12

    His intention was to demonstrate that a bicycle can be functional when some of its parts are missing…So he demonstrated the obvious that Behe and others would not disagree with.

    ==

    Precisely. It IS obvious. Which is EXACTLY what makes Luskin’s claim (and the obvious refutation) so humourous.

    As DonaldM said, everyone who’s ever popped a wheelie knows this claim is patently untrue.

    ==

    Perhaps you can explain how a bicycle with a missing wheel is still a bicycle (or restated: able to function as a bicycle)?

    ==

    I don’t know why I’m required to explain this.

  13. I don’t know why I’m required to explain this.

    And thus you demonstrate your ignorance of the basic concepts surrounding Irreducible Complexity. Or perhaps you realize that by explaining this you defeat Zimmer’s argument, who demonstrates similar misunderstanding (although that’s being nice to say he merely misunderstands and is instead not purposefully distorting)?

    EDIT:

    BTW, this is not the first time I’ve written about unicycles:

    Let’s say you have a scrapyard and monkeys are randomly taking pieces (which might be IC in themselves) and recombining them. Objects that don’t function get junked by the monkey owners. So if we wanted a bike Directly ALL of the pieces involved would have to come together at once, since a partially completed object can’t serve as a “standard bike”.

    Here’s an analogy of an Indirect pathway for a bike. A unicycle, not a standard bike, comes together via a chance arrangement of wheel, nuts/bolts, tire, seat, and pedal. So it’s functional and won’t be junked. But the reason this pathway is Indirect is because although the unicycle is useful and functional it still isn’t a standard bike. So then by chance the frame is modified and a duplication event creates another wheel. Homologous structures like chains, gears, and handle bars are found on other objects tossed into the monkey junkyard. The monkey co-opt these features for use in their standard bike. So we have a model for how monkeys randomly put this standard bike together. But can we reasonably expect such an unlikely scenario to occur? Never mind that all the independent parts themselves require intelligence to create in the first place.

    Also, the T3SS would be like the unicycle and the flagellum like the standard bike in the analogy.

    What if some steps involved simplification – e.g. a component is lost, broken or simplified compared to a previous state in the pathway?

    ID proponents typically have no problem with destructive modifications that help an organism/creature under particular circumstances. We just do not see evidence for major constructive creative evolution via Darwinian processes (intelligent evolution is another matter).

    Darwinists usually try to argue that the T3SS predated the flagellum, but it’s more generally accepted that the flagellum came before the T3SS. Why?

    The flagellum appears on bacteria that are widely presumed to predate anything with a T3SS.

    The gist of it is that the T3SS appears on small number of bacteria that prey on eukaryotes. It’s a weapon used to inject toxins into the prey. In the meantime the flagellum appears on a large number of bacteria that don’t prey on eukaryotes. Thus saying that the T3SS predates the flagellum is like saying that anti-aircraft missiles predate aircraft. Non sequitur. Flagella were useful to bacteria before eukarotes appeared but a t3ss would be useless before then. The prey must precede the predator.

    The reasonable conclusion is that the T3SS destructively evolved (simplified for another usage) from the flagellum rather than the flagellum evolving from the T3SS. This scenario, which actually makes sense in the context of a tree of life beginning with bacteria, is also congruent with what we actually observe in nature today – useful things devolving from something more complex. Behe gives many examples of useful destructive evolution in his recent book “The Edge of Evolution”.
    ……
    The main point is that the unicycle is an island for minimal functionality. The parts that compose the final product can’t really do anything in themselves (yes…I’m sure there are OTHER creative functional uses for tires, but I hope you get the point). The monkeys may never create a unicycle in the first place, never mind the full tandem.

    Now from any island there may be some gradual modifications. Envision this island being surrounded by shark-infested water. In the distance is the island for the tandem bike. How to reach it? There are many stepping stones. But some stepping stones crumble when you step on them, an analogy for modifications that have negative selective pressure. For example, if the monkeys added an anchor to the unicycle that’d be a crumbling stepping stone. The question is…is there a series of stepping stones leading to the other island?

    For example, for a minimal unicycle you really don’t need a seat (although it obviously helps). You could even add a horn or other relatively minor features. You could even add stickers onto the bike, which serves as an analogy to “neutral mutations” since the stickers don’t help or hinder the functionality of the unicycle. But to get the full tandem bike requires several modifications in order to reach that new island of functionality. An extension to the frame throws the balance off. An extra wheel does nothing without an extended frame to attach to…never mind handle bars to steer (which are themselves useless without the other parts). Gears just add weights. A chain is useless without gears and the pedals being moved to the extension of the frame.

    Now a Darwinist might be able to sit down and conceive a series of unicycles that’d evolve into a functional tandem bike. But I’d imagine they’d be barely ridable. Remember, these modifications must either aid the functionality of the unicycle. If you were to start tacking on features relevant to the tandem bike it would not help. Darwinian processes don’t know of a future goal. All they know is that these additional features are causing negative selection pressure until the final goal is reached.
    ……
    Being IC does NOT equate to “X-structure cannot evolve in principle”. IC primarily deals with DIRECT Darwinian Pathways and always has. Behe has always stated that INDIRECT Darwinian pathways are another matter. Also, what is “plausible” is usually tied into positive selection pressure. Every single indirect gradual genetic change must be positively selected, which is what I highlighted with the bike example[yes, realize there are minor exceptions with neutral mutations and genetic drift and then a changing environment may positively select these couple mutations]. Essentially, there must be islands of functionality that are within leaping distance of each other (or only have 1 or 2 or 3 stepping stones to connect them). So, yes, there will be some IC structures where an indirect pathway is plausible. Unfortunately for Darwinists, the flagellum is not one of them.

  14. 14

    I don’t find it unacceptable. I think it’s great that you do. Keep up the good work :-)

    ==

    So you’re asking me to provide you with the undesigned evolutionary precursor of something that is known NOT to have evolved?

    Okay, I’ll get right on that…

    ==

    But let’s get to point I think he was trying to make: does IC exist in anything or is it a fallacy?

    ==

    I think IC is a poorly defined concept. As defined by Behe (“…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…”), evolution is perfectly capable of producing irreducibly complex systems–by producing scaffolding and then removing that scaffolding after the system is in place. This was a prediction of evolutionary theory made by Hermann Muller (some time in the 40s, as I recall).

    This form of IC most certainly DOES exist, but it is not a challenge for evolutionary theory.

    There is also a version of IC that evolution could NOT produce. That would be the one touch on by Darwin, in which there is absolutely NO (“gradual”) evolutionary path to an endpoint. (In the terminology of Richard Dawkins, there is no path to that point in Biomorph World, whether scaffolding is used or not.)

    I don’t believe that there are examples of this kind of IC in the natural world. There may or may not be (human-)designed examples of this. I haven’t given that matter much thought.

    But the burden of proof is always on the ID proponent who is making the claim “there is absolutely NO way this could have evolved”. They have not substantiated this claim.

  15. His intention was to demonstrate that a bicycle can be functional when some of its parts are missing

    That is something to thing about — what makes a bicycle IC? One of the hard parts finding the answer would be the inherent redundancy of the design (design hee hee) i.e. the two wheels.

  16. So you’re asking me to provide you with the undesigned evolutionary precursor of something that is known NOT to have evolved?

    I wasn’t asking you do anything. I was approving of the use of using designed objects to attempt to demonstrate evolution. That’s all :-)

    evolution is perfectly capable of producing irreducibly complex systems–by producing scaffolding and then removing that scaffolding after the system is in place.

    That’s the theory.

    There is also a version of IC that evolution could NOT produce. That would be the one touch on by Darwin, in which there is absolutely NO (”gradual”) evolutionary path to an endpoint.

    And whether the flagellum suffices is the debate.

    But Luskin’s point (I think) is that IC exists and is evidence of design, and (I think) we are in agreement there.

    With regard to the bicycle, maybe what we have stumbled onto is that something can be designed and not IC.

  17. I don’t believe that there are examples of this kind of IC in the natural world. There may or may not be (human-)designed examples of this.

    An IC machine cannot, by definition, be the result of a Direct Darwinian pathway. Direct means that the steps are selected for the improvement of the same function we find in the final machine. IC makes a direct Darwinian pathway impossible.

    So, only two possibilities are left: either sudden appearance of the complete machine (practically impossible for statistical considerations), or step by step selection for different functions, and with the target function COMPLETELY INACTIVE for natural selection. This is a point that Darwinists tend to bypass. The components specific to the functionality of a bicycle will be quite useless to the functionality of the unicycle until the IC core is assembled.

    In order to have positive selection the changes being made to a unicycle not only have to pull together a functional bicycle but they can’t have a negative effect that is worse than the positive of having the fully functional IC core set of bicycle components. It could be argued that some changes to a unicycle could be considered neutral mutations in that a skillful unicyclist could still ride the unwieldy contraption. The addition of the frame and handlebars “might” be helpful for balance, although that’s a question for a unicyclist since the addition of extra unbalanced weight would likely outweigh any limited benefit (I suppose you could argue for a scaffolding-type mutation that acts as a counter-weight). And there will be instances of co-opted components like ball-bearings, screws, etc. But that still leaves all the problems listed in comments #3 and #13 (never mind the ones I did not bother elaborating on) which cause negative selection.

  18. But the burden of proof is always on the ID proponent who is making the claim “there is absolutely NO way this could have evolved”.

    That’s what those in establishment want to us think, but it’s backwards. It’s those making a definitive claim about how something occurred who have the obligation to show conclusively that it happened that way.

  19. “That is something to thing about — what makes a bicycle IC?”

    The one who is riding it, ofcourse. The missing IC core is the skill required to ride a unicycle vs. the skill required to ride a regular bicycle to get from point a to point b (its sole purpose). A bicycle or unicycle does nothing by itself.

  20. he burden of proof is always on the ID proponent who is making the claim “there is absolutely NO way this could have evolved”

    Actually, while I agree with Tribune that Darwinists should provide positive evidence for their claims, for any biological object I would think it should be potentially possible to imagine a gradual pathway(s) by which it could come to be. I can imagine a long series of neutral and beneficial mutations and presume the easy co-option of various components, also presuming that major changes in functionality can always be achieved with a couple amino acid residues. But the question is whether my imagination is realistic, especially when the resources required for just a couple steps in a gradual pathway has been observed to be quite large.

    In relation to the flagellum this was discussed in Computers vs Darwinism?.

    EDIT:

    There is also a version of IC that evolution could NOT produce. That would be the one touch on by Darwin, in which there is absolutely NO (”gradual”) evolutionary path to an endpoint.

    It’s not just gradualism, but whether the pathway is Direct. And also whether all available Indirect stepwise pathways are composed of a good percentage of deleterious steps.

    But with this paragraph you illustrate that you do have an understanding of the problem, yet you insist that it cannot ever apply in biology despite being given examples of such.

  21. The one who is riding it, ofcourse. The missing IC core is the skill required to ride a unicycle vs. the skill required to ride a regular bicycle to get from point a to point b (its sole purpose). A bicycle or unicycle does nothing by itself.

    That’s like pointing out the flagellum does nothing without the organism it’s attached to.

  22. “That’s like pointing out the flagellum does nothing without the organism it’s attached to.”

    Thus, even with all its parts intact function seizes rendering it that much more irreducibly complex then previously thought.

  23. Patrick

    Actually, while I agree with Tribune that Darwinists should provide positive evidence for their claims, for any biological object I would think it should be hypothetically possible to imagine a pathway(s) by which it could come to be. I can imagine a long series of neutral and beneficial mutations and presume the easy co-option of various components, also presuming that changes in functionality can always be achieved with a couple amino acid residues. But the question is whether my imagination is realistic, especially when the resources required for just a couple steps in a pathway has been observed to be quite large.

    Good point, Patrick. The claim that ID says “there’s no way this could have evolved” is a misrepresentation. The actual claim is that there is no known Darwinian pathway that can account for this or that biological system. And further that the system bears characteristics we would normally associate with design.

    Now, to falsify design, all you have to do is find just ONE Darwinian pathway. But, (and here’s the point that so many Darwinians seem to miss), it has to be an actual pathway, not an imagined pathway. IC or ID isn’t falisfied in any scientific sense based on merely conceivable Darwinian pathways any more than any other hypothesis would be considered confirmed in the absence of any experimental data. Merely imagining an hypothesis doesn’t confirm it.

    Thus when Dr. Zimmer writes that scientists have “shown over and over again” that IC is false, all he means is that scientists who are also Darwinians have imagined all sorts of possible Darwinian pathways for all these supposed IC systems. What they haven’t done is move the imagined pathway to the lab and confirmed it with real detailed testable biological models.

    But, by keeping the terms in the camp of conceivability, evolution is made virtually unfalsifiable, because there’s always the possibility of some other imagined pathway. This is backwards. The burden of proof is not to show that there is NO Darwinian pathway…that isn’t possible. The burden is on the Darwinians to show which of these myriad of imagined pathways is actually true with actual detailed testable biological models.

  24. ab #22,

    Good point, although the existence of a super-system is not always a requirement for being IC.

    Anyway, it’s been fun being involved in this discussion but I have some work to do before the festivities begin tonight. So if any other admins are watching please clear the queue.

  25. 25

    …for any biological object I would think it should be potentially possible to imagine a gradual pathway(s) by which it could come to be. I can imagine a long series of neutral and beneficial mutations and presume the easy co-option of various components…
    ==

    Congratulations. You’re an “evolutionist”. And don’t forget that they need not be neutral or beneficial mutations. If the population is small enough, genetic drift can result in peak shifts.

    ==
    Actually, while I agree with Tribune that Darwinists should provide positive evidence for their claims…
    ==

    And indeed they have. To the point where evolution is accepted beyond all reasonable doubt within the scientific community.

    ==
    But the question is whether my imagination is realistic, especially when the resources required for just a couple steps in a gradual pathway has been observed to be quite large.
    ==

    Example?

  26. A watch has often been used as the example of complexity. At one time the gears had to be well-formed, with regular teeth that interlock just right. It was a good example of something that is probably still a fair argument that nature wouldn’t make.

    Miller’s response runs along the line of showing that there are no gears in the modern watch, so something else can fulfill that macro purpose of watchmaking. Thus, watches you find in deserted places could have evolved that way. Just because we don’t know all the scaffolding that has to go on, is no reason to jump to the conclusion that men had to make it.

    It’s silly to introduce “refutations” that were in the original concept. Basically, it comes with a blunt suggestion that you’re an idiot, or at least have not thought it through.

    For example, somewhere along the line, we’ve lost that flagellum gene, but locomotion is still possible, because look at our feet and legs. And hey if we didn’t have locomotion that way, since locomotion can come by way of secretory glands, it’s nothing for locomotion to come by the way of our sweat glands or nose.

  27. The claim that ID says “there’s no way this could have evolved” is a misrepresentation. The actual claim is that there is no known Darwinian pathway that can account for this or that biological system. And further that the system bears characteristics we would normally associate with design.

    Excellent points, DonaldM

  28. 28

    Luskin is exactly correct…remove a wheel from a BI-cycle, and you no longer have a funtioning BI-cycle.

    ==

    I hope you’re kidding me. How could *I* be the one accused of playing word games?

    Luskin said:
    “Unicycles, having only one wheel, are missing an obvious component found on bicycles. Does this imply that you can remove one wheel from a bicycle and it will still function? Of course not. Try removing a wheel from a bike and you’ll quickly see that it requires two wheels to function.”

    He is attempting to say that bicycles are irreducibly complex because they are not reducible to unicycles (which are functional) if one merely removes a single wheel.

    If the question were as you suppose, he might as well have said: “By definition, a bicycle has two wheels. Therefore, if you remove one wheel, it is no longer a bicycle. Bicycles are therefore irreducibly complex. QED.” A completely meaningless statement. That’s not how evolution works, and that’s not even how the concept of IC works.

  29. for any biological object I would think it should be potentially possible to imagine a gradual pathway(s) by which it could come to be. . . .Congratulations. You’re an “evolutionist”.

    Patrick, I gotta go with TheYellowShark on this one. Showing the ways it is possible for the flagellum to have evolved is a great work of imagination :-)

    And indeed they have. To the point where evolution is accepted beyond all reasonable doubt within the scientific community.

    Geocentricsm was accepted beyond all reasonable doubt the scientific community circa 1610. Being in an echo chamber isn’t good for anybody.

  30. 30

    And thus you demonstrate your ignorance of the basic concepts surrounding Irreducible Complexity. Or perhaps you realize that by explaining this you defeat Zimmer’s argument, who demonstrates similar misunderstanding (although that’s being nice to say he merely misunderstands and is instead not purposefully distorting)?

    ==

    No. The original question was:

    “Perhaps you can explain how a bicycle with a missing wheel is still a bicycle (or restated: able to function as a bicycle)?”

    That was not Luskin’s intention.

    And note that I have provided Behe’s definition of IC above, and it *does not require* that an evolutionary precursor function in the same manner as what it gives rise to. It merely requires that removal of one component causes “the system to effectively cease functioning…”. It does not specify the nature of the function, nor should it. Evolution doesn’t work this way. See “exaptation”.

    Therefore, according to Behe, a bicycle with one wheel removed *need not function as a bicycle*, nor could it _by definition_. This renders your interpretation of Luskin’s comments completely meaningless.

  31. “The fact that a unicycle lacks certain components of a bicycle does not mean that the bicycle is therefore not irreducibly complex.”

    How is that for twisted logic?

    A unicycle in its current form is IC because it requires far greater skill to control (break, turn etc…) and to perform the way a bicycle performs. A bicycle includes additional functions to compensate for the skill that would be required on a unicycle. Both are therefore IC as per the intent to get from point a to b as a method of human transportation.

  32. 32

    Geocentricsm was accepted beyond all reasonable doubt the scientific community circa 1610. Being in an echo chamber isn’t good for anybody.

    ==

    And intelligent design was accepted by the scientific community prior to 1859. Both ideas were replaced, and rightfully so.

  33. especially when the resources required for just a couple steps in a gradual pathway has been observed to be quite large.
    ==

    Example?

    First off, I’m speaking of observed evolution, not hypotheses. Off the top of my head, Malaria and HIV have bother been observed to use a huge number of resources in order to produce stuff within their expected range. This has been covered in EoE and on Behe’s Amazon blog. Dr. Lenski and his 20 years and 40,000+ generations of E.coli for citrate which is the equivalent of +-800,000 years of evolution in a species with an average generation of 20 years.. Again with E. coli, there is the lactose consumption experiment which involved 30,000+ generations which is the equivalent of +-600,000 years of evolution in a species with an average generation of 20 years. Nylonase, which was a a pre-existing esterase with B-lactam folds that had minimal nylon hydrolysis activity from the start. Pretty much every example trumpeted by Darwinists that I can think of involve a large number of replications.

    That was not Luskin’s intention.

    and

    He is attempting to say that bicycles are irreducibly complex because they are not reducible to unicycles (which are functional) if one merely removes a single wheel.

    Ah, so that’s how you are interpreting him. I did not read it that way but I can certainly see how you could read it that way. I’ll email Luskin to see if that was what he intended to say.

    Also, if a component that is part of the IC core is removed from a system then that system would either have to operate as a different system, a combination of several different systems with different functions, or potentially be completely non-functional for any purpose (unless you assert that something like a ball bearing, screw, or spring is a “system” in itself).

    I’ll also note that you’re resorting to nitpicking based solely upon a single sentence by Behe which does not explain the concept at length. These short definitions are a little better…

    “A system performing a given basic function is irreducibly complex if it includes a set of well-matched, mutually interacting, nonarbitrarily individuated parts such that each part in the set is indispensable to maintaining the system’s basic, and therefore original, function. The set of these indispensable parts is known as the irreducible core of the system.” (No Free Lunch, 285)

    “An irreducibly complex evolutionary pathway is one that contains one or more unselected steps (that is, one or more necessary-but-unselected mutations). The degree of irreducible complexity is the number of unselected steps in the pathway.” (A Response to Critics of Darwin’s Black Box, by Michael Behe, PCID, Volume 1.1, January February March, 2002)

    …but it’s funny how yet again you’re resorting to word games. I swear this past month has been the Word Games Winter Olympics for UD.

    And intelligent design was accepted by the scientific community prior to 1859. Both ideas were replaced, and rightfully so.

    Replaced due to the ignorance of the amazing information-based systems in biology (they thought they were mere blobs and OOL was easy!) and currently held onto with a death grip due to ideology.

  34. Yellowshark


    Patrick: for any biological object I would think it should be potentially possible to imagine a gradual pathway(s) by which it could come to be. I can imagine a long series of neutral and beneficial mutations and presume the easy co-option of various components…
    ==,

    Congratulations. You’re an “evolutionist”. And don’t forget that they need not be neutral or beneficial mutations. If the population is small enough, genetic drift can result in peak shifts.

    Interesting. All it takes to be an evolutionist (or Darwinist) is merely imagine. Well, all due respect to John Lennon, but imagine makes a great song, but doesn’t provide evidence in any scientific sense. Imagination can inspire scientific discovery without a doubt. But, imagination without experimental confirmation is scientifically empty.

    ==

    Patrick:
    Actually, while I agree with Tribune that Darwinists should provide positive evidence for their claims…
    ==

    And indeed they have. To the point where evolution is accepted beyond all reasonable doubt within the scientific community.

    Science has historically accepted all sorts of things beyond all reasonable doubt. That doesn’t tell us much. Sure there’s a mountain of data that one can take to be evidence for this or that aspect of evolution. The issue is how strong is the evidential value of that data. What seems to be the case with evolution is that a lot of the data claimed to be evidence for evolution loses its evidential starch when the assumption of evolution is removed. For example, homologies in biological systems are often cited as ‘strong’ evidence for common ancestry. But, they could also be ‘strong’ evidence for common design. There’s nothing in the data that allows us to scientifically exclude the one in favor of the other.

    What seems to be “accepted” within science beyond all reason is naturalism (whether philosophical or merely methodological – not that theres any real difference). When all data is looked at that way, then evolution, or something very much like it is the only game in town.

    Okay, its New Years Eve…everyone go party!! Happy New YEar. I won’t be posting again until next year!

  35. 35

    Off the top of my head, Malaria and HIV have bother been observed to use a huge number of resources in order to produce stuff within their expected range. This has been covered in EoE and on Behe’s Amazon blog. Dr. Lenski and his 20 years and 40,000+ generations of E.coli for citrate which is the equivalent of +-800,000 years of evolution in a species with an average generation of 20 years.. Again with E. coli, there is the lactose consumption experiment which involved 30,000+ generations which is the equivalent of +-600,000 years of evolution in a species with an average generation of 20 years. Nylonase, which was a a pre-existing esterase with B-lactam folds that had minimal nylon hydrolysis activity from the start. Pretty much every example trumpeted by Darwinists that I can think of involve a large number of replications.

    ==

    I’m not sure why this is a problem, even if I accept the numbers you are providing.

    You may be forgetting that evolution in the same manner as a parallel processor. We don’t need to solve one problem before we move onto the next. Lineages throughout the population are “working on the solutions to various problems”, and they share these solutions with each other by sexual reproduction, conjugation, etc.

    So we have plenty of time to work with. And much of the “heavy lifting” that was done was in the evolution of various gene toolkits, proteins and enzymes early in the history of life (when generation times were likely as short as modern bacteria).

    I have not read EoE, and only briefly glanced at Behe’s blog, but needless to say, I was not impressed with his demonstrably false claim that there has been no significant biochemical evolution in HIV, as discussed by ERV and others.

    Anyway, it’s New Year’s and I’m going out to have a drink. I suggest you do the same.

    I will address some posts tomorrow, but then I’ll be catching a train across Canada tomorrow night, so it will be a few days before I can address responses any further.

  36. Yes TheYellowShark, via intelligent design, ie design modifications, one can take an irreducibly complex structure and modify it to make another IC structure with a similar function.

    However not everyone that can ride a bicycle can ride a unicycle.

    That does nothing to demonstrate that undirected processes could/ would do such a thing.

    Ya see the debate isn’t that IC can’t “evolve”. It is about the MECHANISM- ie designed to evolve vs evolution via an accumulation of genetic accidents.

    Toolkit genes- more evidence for intelligent design in biological organisms.

    As for HIV, the change Behe missed, ONE in the billions of opportunities, is not that significant.

  37. And intelligent design was accepted by the scientific community prior to 1859.

    “All matter originates and exists only by virtue of a force which brings the particles of an atom to vibration and holds this minute solar system of the atom together . . . . We must assume behind this force the existence of a conscious and intelligent mind.” Max Planck, 20th century

    Both ideas were replaced, and rightfully so.

    Sheer dumb luck isn’t a viable replacement. And that is you have without ID…

  38. Geocentricsm was accepted beyond all reasonable doubt the scientific community circa 1610. Being in an echo chamber isn’t good for anybody . . .And intelligent design was accepted by the scientific community prior to 1859. Both ideas were replaced, and rightfully so.

    The point is don’t equate consensus with fact. The late Michael Crichton addressed how great a mistake is here.

    And do you really think accident can explain everything?

  39. Happy New Year, YellowShark and everybody else.

  40. Why doesn’t the majority of commenters, even long time contributors like tribune7, use the blockquote tag when they cite from other comments? It actually works, e.g. see Jerry at 37.

  41. sparc

    You wrote:

    Why doesn’t the majority of commenters, even long time contributors like tribune7, use the blockquote tag when they cite from other comments?

    I know this sounds dumb, but to be honest, although I am capable of programming in HTML code, I never realized that you could use HTML tags (such as blockquote) when entering in comments on this blog. The thought never crossed my mind. I’d always wondered how other people manage to indent and do italics and stuff. OK, I’m going to try an experiment.


    To indent with italics, type “

    “, and at the end of the idented section, finish with “

    “.

  42. OK, it worked, kind of.

  43. With me it comes to down to it being easier to type i that blockquote. :-)

    Also, aesthetics. I prefer italics for short quotes and blockquote for long quotes, especially if I’m starting my post with the quote.

  44. I emailed Luskin with this:

    I was hoping you could clarify a comment you made.

    You said:

    “Unicycles, having only one wheel, are missing an obvious component found on bicycles. Does this imply that you can remove one wheel from a bicycle and it will still function? Of course not. Try removing a wheel from a bike and you’ll quickly see that it requires two wheels to function.”

    The Darwinists are saying that “he[Luskin] is attempting to say that bicycles are irreducibly complex because they are not reducible to unicycles (which are functional) if one merely removes a single wheel.”

    Is that correct?

    Luskin’s response:

    Thanks for your e-mail. As usual, Darwinists are mis-stating / over-stating my argument and missed the point of my argument very badly

    My point with the bike/unicycle example is to illustrate the fallacy of Ken Miller’s argument re the blood clotting cascade (BCC). KEN MILLER effectively said that if system B (i.e. the jawed-fish BCC, or a unicycle) functions while missing component x (i.e. factors 12/12a/11, or a wheel) found in system A (i.e. the land-dwelling vertebrate BCC, or a bicycle), then system A is NOT irreducibly complex. My point is that this is not a valid argument, because:

    (1) System A and System B could be very different systems that have different functional requirements
    or
    (2) System A might have an irreducible core that does not include X, and therefore our findings about System B is irrelevant.

    (1) and/ or (2) could be true for the BCC, so Miller’s argument did not refute Behe’s argument.

    Regarding bikes and unicycles, bikes are irreducibly complex whether one has invented the unicycle or not. The reason we know this is because we can perform knockout experiments on a bike where we remove various parts, and find that the bike does not function. I argue that a bike is irreducibly complex with respect to:

    (a) Both its wheels
    (b) Frame
    (c) Steering mechanism
    (d) Motor mechanism

    …because if you knock out any one of those parts, the bike won’t function.

    So I’m not arguing that if bikes aren’t reducible to unicycles, then bikes are irreducibly complex. The irreducible complexity of bikes can be demonstrated quite separate from consideration of unicycles; unicycles are discussed simply as part of my example to show the fallacy of Miller’s argument. The reason we know bike’s are irreducibly complex is because the knockout experiment described above shows that they are. It has nothing to do with unicycles.

    Does that clarify? Thanks and happy new year.

    So, Yellowshark, if your comment was the essence of Zimmer’s argument then you both misread Luskin and put words into his mouth. Although, I will say that his original paragraph was poorly phrased since it was open to interpretation.

    You may be forgetting that evolution in the same manner as a parallel processor. We don’t need to solve one problem before we move onto the next.

    Of course.

    Lineages throughout the population are “working on the solutions to various problems”, and they share these solutions with each other by sexual reproduction, conjugation, etc.

    At least with HIV and malaria (oh, and the changes wrought in humans) these were not controlled experiments. These organisms are constantly being modified and attempting to solve any problem that may affect their survivability. So your objection holds little force, although you may appeal to “deep time” as a solution to anything.

    So we have plenty of time to work with.

    Perhaps that may be reasonably asserted for fast replicators. But what of higher organisms with longer generation time?

    And much of the “heavy lifting” that was done was in the evolution of various gene toolkits, proteins and enzymes early in the history of life (when generation times were likely as short as modern bacteria).

    That’s actually a design hypothesis, except you’d replace “evolution” with “design”.

    I have not read EoE, and only briefly glanced at Behe’s blog

    No surprise here.

    but needless to say, I was not impressed with his demonstrably false claim that there has been no significant biochemical evolution in HIV, as discussed by ERV and others.

    When writing the book Edge of Evolution Behe wrote that we should expect certain things within certain boundaries. He was not aware of any at the time so he put a “0″ in his graphics. Smith simply highlighted an example that Behe did not know about, but which fit within the expected “edge” for HIV, and if the book is ever revised I’m sure it’ll be added.

  45. Something that should be obvious is that while there is no way a unicycle could evolve into a bicycle it is possible for a bicycle to evolve (change via chance/accident/non-design) into a unicycle, with of course a loss of much function.

    Something else to consider is that while much function would be lost, this new, accidental unicycle would still have some advantages — such as taking up less space in the garage and being lighter.

    And again it gives me pleasure to no end to see objects of known design being used as examples of undirected evolution :-)

  46. 46

    However not everyone that can ride a bicycle can ride a unicycle. That does nothing to demonstrate that undirected processes could/ would do such a thing.

    ==

    …Umm, okay. Point taken?

    ==

    Ya see the debate isn’t that IC can’t “evolve”. It is about the MECHANISM- ie designed to evolve vs evolution via an accumulation of genetic accidents.

    ==

    Okay, so basically what you’re saying is that IC is a worthless concept in ID. If IC can evolve, then it cannot be used as a form of design detection.

    ==

    Toolkit genes- more evidence for intelligent design in biological organisms.

    ==

    How so?

    ==

    As for HIV, the change Behe missed, ONE in the billions of opportunities, is not that significant.

    ==

    Behe claimed that: “It still has the same number of genes that work in the same way. There is no new molecular machinery.”

    Both of these claims were disproved, and were in fact known to be false long before he made them.

  47. TheYellowShark

    You may be forgetting that evolution in the same manner as a parallel processor. We don’t need to solve one problem before we move onto the next.

    Your understanding of natural selection and population genetics is lacking deeply.

    Natural selection, you see, doesn’t get to select between individual alleles. It can only select whole organisms which contain millions of alleles – the good, the bad, and the ugly.

    Actually, it’s not. I do recommend that you actually learn something about evolutionary theory before you criticize it.

    Natural selection is the antithesis of “dumb luck”.

    I recommend YOU learn a bit more about what you think you’re championing.

    In the theory commonly called Random Mutation Plus Natural Selection the mutations are random with respect to fitness. Fitness enhancing mutations are exceedingly rare because life is already exquisitely well designed. So the vast majority of random mutations are in the range of immediately fatal to slightly deleterious.

    Because natural selection doesn’t operate on single mutations but instead must accept or reject all mutations because it operates on the entire organism not microscopic bits of the organism, it must accept the bad and the nearly neutral along with the good.

    This leads to things like Haldane’s Dilemma and Sanford’s Genetic Entropy.

    Unless very beneficial adaptations are front loaded into the genome in such a way that random mutuation finds them far more frequently than it takes to generate them from scratch then natural selection among species with obligatory sexual reproduction acts as a conservative force keeping the species from wandering too far off the reservation, so to speak, until the inevitable accumulation of nearly neutral (natural selection can’t “see” nearly neutral mutations) slightly deleterious mutations spells eventual extinction.

    The fossil record, I might remind you, is a record of abrupt appearance of new species, a long period of no significant change to the new species, followed by an abrupt disappearance. Exactly what a front loaded, pre-programmed phylogenetic progression should look like and nothing like what Darwinian gradualism should look like. Get a clue.

  48. 48

    Thanks sparc@40 for the blockquote tip. I wasn’t aware that those tags would work here.

    Tribune said:

    The point is don’t equate consensus with fact. The late Michael Crichton addressed how great a mistake is…

    I’m not. The evidence for evolution speaks for itself. And it was that evidence that convinced each and every one of those scientists as well.

    And do you really think accident can explain everything?

    No. See “natural selection”. It is not chance. It is not “accident”.

    (Sorry if this double-posts. Internet isn’t working too well.)

  49. 49

    Natural selection, you see, doesn’t get to select between individual alleles. It can only select whole organisms which contain millions of alleles – the good, the bad, and the ugly.

    This is not a refutation of anything that I’ve said. Stay on topic.

    Fitness enhancing mutations are exceedingly rare because life is already exquisitely well designed. So the vast majority of random mutations are in the range of immediately fatal to slightly deleterious.

    You meant to say “well-adapted”, not “well-designed”, I’m sure. And no. The “majority of mutations” are NEUTRAL. The *average* effect of random mutation, however, is slightly deleterious for the reason you mentioned above.

    But I’m sure you knew _exactly_ what you were talking about and you just misspoke.

    Because natural selection doesn’t operate on single mutations but instead must accept or reject all mutations because it operates on the entire organism not microscopic bits of the organism, it must accept the bad and the nearly neutral along with the good.

    Yes. Congratulations. Genetic hitchhiking, linkage disequilibrium, selective sweeps, etc.

    Still waiting for this to get to the refutation of the parallel processor analogy. (Hint: You are aware of the whole *population* part of “population genetics”, right?)

    The fossil record, I might remind you, is a record of abrupt appearance of new species, a long period of no significant change to the new species, followed by an abrupt disappearance.

    So says the creationist version of Stephen Jay Gould. You might keep in mind that even die-hard punctuationists don’t make this an absolutist claim. “Gradual” transitions between species exist. It was always an argument about which is the *dominant* pattern. And the jury’s still out.

    Exactly what a front loaded, pre-programmed phylogenetic progression should look like and nothing like what Darwinian gradualism should look like.

    Strange you talked up all that teams of genes and you haven’t even thought about what it means for species. How’s about you do a little reading? You might want to start with:
    Futuyma, D. J. 1987. On the role of species in anagenesis. The
    American Naturalist 130: 465-473.

    Enjoy your clue. I’m getting on the train now.

  50. DaveScot, this belongs on the front page of UD. Darwinists should also have a post-it note on their fridge for what you just explained. Although, maybe a post-it note might not “stick” that well.

  51. Ya see the debate isn’t that IC can’t “evolve”. It is about the MECHANISM- ie designed to evolve vs evolution via an accumulation of genetic accidents.

    ==

    Okay, so basically what you’re saying is that IC is a worthless concept in ID. If IC can evolve, then it cannot be used as a form of design detection.

    I take it you have reading comprehension issues.

    IC can “evolve” if it was DESIGNED to do so.

    Your position of accumulated genetic accidents is still very much in doubt pertaining to IC.

    Toolkit genes- more evidence for intelligent design in biological organisms.

    ==

    How so?

    1- We have direct observational evidence and experience with designers putting together the resources needed to implement the design.

    2- We have NEVER observed undirected processes doing so.

    As for HIV, the change Behe missed, ONE in the billions of opportunities, is not that significant.

    ==

    Behe claimed that: “It still has the same number of genes that work in the same way. There is no new molecular machinery.”

    You need to get updated. It appears you are stuck in the past.

    As for natural selection- it is the result of three processes, each with either a random component or entirely random.

    And in the end it doesn’t do much of anything in populations over 1000.

    Now in sexually reproducing populations there isn’t any guarantee that the most beneficial mutation will get passed on.

    Anything less than 50% and the math shows that unless there is a very high selective value it will get lost- no chance of remaining in the population.

    Yet your position requires that these genetic accidents not only become fixed but also accumulate.

    There isn’t any evidence for this leading to new body plans.

  52. Dave Scott

    Natural selection, you see, doesn’t get to select between individual alleles. It can only select whole organisms which contain millions of alleles – the good, the bad, and the ugly.

    Exactly, Dave. So often NS is discussed as if it only selected one allele at a time. But its quite possible that there are combinations of changes. What no one can predict is what the “winning” combination will be.

    Yellowshark

    Natural selection is the antithesis of “dumb luck”.

    NS isn’t really anything other than a descriptive phrase to talk about certain observations after the fact. It not that NS is the “antithesis” of dumb luck, its not really anything at all except a handy phrase to describe observations.

  53. And do you really think accident can explain everything? . . . No. See “natural selection”. It is not chance. It is not “accident”.

    Natural selection is an observable event but it can’t be shown that it does what proponents of evolution claim it does.

    Anyway, if it is a law, how do laws come about without design?

  54. tribune7-

    According to Hawking in “A Briefer History of Time” the laws “Just are (the way they are)”.

    And that is about all you will ever get from those who oppose ID.

  55. The evidence for evolution speaks for itself.

    “Evolution” is NOT being debated.

    Please read the following for YOUR clue:

    Biological Evolution: What is being debated

    Hint- the debate is about A) Origins and B) Mechanisms

  56. TheYellowShark,

    Essentially all you are doing is making assertions and assuming they are true. You are welcome to your speculative interpretations of some isolated facts but that is all they are is speculations. To come here and assert that your assertions are established is hubris. We have seen it all before and none have stood up when scrutinized.

    Also the natural selection is not random is a red herring argument and not an issue. It is a meaningless discussion. ID accepts NS but NS is very limited so the answer to the assertion that NS is not random is “So what!” We been down this path a hundred times at least. There is a random component to NS but mainly what is in play is the population gene pool and the environment. No big deal. The random part comes from how something is added to the gene pool. So let’s move on to something non trivial. The fact that you waste any time on this issue means you do not have anything of substance or else you would have trotted it out.

    There are no examples of gradual transitions in the fossil record, none, nada, zero, zilch, zip. We have yet to see one. We have been given the forrest animal to whale scenario but there are no gradual transitions there, only fossils millions of years apart and variations of marine mammals. There is also what is referred to the “crown jewel” of the fossil record and that is the reptile to mammal movement of three bones from the jaw to the ear. Again interesting but not an example of gradualism and ignores all other problems with the differences between mammals and reptiles. So have at it if you find this when you read again. We are still waiting for an example and as for the natural selection red herring, we have been down this path hundreds of times.

    I always looked at concepts such as scaffolding and co-option as things that were pulled out of one’s back pocket when something got sticky and there was no answer for what happened or exists. They are not quite all purpose tools but they often are handy to survive a debate. I am surprised you did not resort to the all purpose concept of emergence and that one can solve everything. Maybe you are waiting till things get tough to pull that one out. So essentially your assertions here are all begging the question arguments. You take what seems logical within a framework and then assert them without examining the assumptions upon which the framework is based. You beg the question.

    It will be interesting to see how long you stay once you come back because assertions don’t cut it. Just as heads up for you. We accept all forms of micro evolution so any changes to gene pools through the typical genetic processes is all well accepted even if a lot of it is in fluctuation as new information is discovered. The debate is over how novel complex functional capabilities arose in life given that the information to run such capabilities is immense and unlikely to happen by chance. So stick to that.

    I hope you enjoy your train ride.

  57. Short on time so I’ll only focus on one topic and ignore the unsubstantiated assertions:

    The “majority of mutations” are NEUTRAL.

    If you are saying that in order for a long stepwise pathway to be viable that the majority of steps/mutations must be neutral or beneficial (weakly, strongly, destructively, constructively; doesn’t matter), then I’d agree with you. But if you’re talking in general, then I have to wonder what you are speaking of.

    Most of the estimates I’ve seen for deleterious mutations are generally in the range of 90 to 99 percent, with the remaining percentage comprising weakly beneficial yet often destructive mutations, and then rarely beneficial constructive mutations. But if you’re only considering “beneficial” mutations in general, the ratio of deleterious-to-beneficial mutations is very roughly estimated to be one in one million by Gerrish and Lenski.

    I’ll also quickly comment on the Behe vs Smith disagreement. Honestly I wonder why Darwinists choose to highlight this viral examples, especially since it only confirms Behe’s own hypothesis. There’s better examples in vertebrates, which have far less resources to work with compared those fast replicators. Even then there’s not much to excited about. For example, I was reading about a change that required 3 amino acids which were encoded by 9 nucleotides each. So that’s a pathway with only 54 informational bits involved. Where exactly do examples like this cause a problem for ID when it’s all within previously stated expectations?

    so basically what you’re saying is that IC is a worthless concept in ID. If IC can evolve, then it cannot be used as a form of design detection.

    I’ll just quote myself from this very thread:

    “Being IC does NOT equate to “X-structure cannot evolve in principle”. IC primarily deals with DIRECT Darwinian Pathways and always has. Behe has always stated that INDIRECT Darwinian pathways are another matter.” The key is whether there exists indirect pathways for everything. For example, if a system is composed of 3 parts there’s a good chance that there is a viable indirect pathway.

  58. 58

    Patrick@44:

    Ken Miller’s reponse is up at the Loom.

    http://blogs.discovermagazine......en-miller/

  59. 59

    If you are saying that in order for a long stepwise pathway to be viable that the majority of steps/mutations must be neutral or beneficial (weakly, strongly, destructively, constructively; doesn’t matter), then I’d agree with you. But if you’re talking in general, then I have to wonder what you are speaking of. Most of the estimates I’ve seen [link here] for deleterious mutations are generally in the range of 90 to 99 percent…

    This is not at all what your link says. Your link states, quite clearly: “Our analysis suggests that ?95% of all nonsynonymous mutations that could contribute to polymorphism or divergence are deleterious”.

    Note the word “nonsynonymous”, and that they restrict themselves to “mutations that could contribute to polymorphism or divergence”.

    The fact that “most mutations are neutral” was one of the most important & surprising findings of biology in the last 50 years. It is the reason that genetic drift of neutral mutations forms the null hypothesis for molecular biology.

    Estimates sourced in Futuyma’s “Evolutionary Biology”, which I don’t have on me at the moment (I’m at Winnipeg Station) give approximately 80% of all mutations being neutral, 15% being deleterious and 5% being beneficial.

  60. Ugh, I just reread what I wrote. My poorly worded point was that “most mutations are neutral” would need to apply uniformly (in general) and thus specifically to pathways relevant to the problems at hand. As in, it does no good if 80% of ALL mutations as a whole are largely irrelevant to the scenarios/barriers facing Darwinism. More on this later when I have time. I’ll quickly note Futuyma’s early estimates were limited mostly to inferences from phylogenetic analyses of sequence data, inferences that are only as good as the assumptions they make about population demography and the neutrality of synonymous mutations. The data from experimental observations is limited at this time. It’s also not quite clear from experimental observations whether percentile distributions hold steady from virus to eukaryotes to all other higher creatures, never mind specific categories and pathways.

    EDIT:

    I’d also like to add that many amino acids at different parts of a protein are interchangeable and thus functionality remains unchanged for some nucleotide changes in a sequence. If that’s where the majority of neutral mutations are limited to then how does that help Neutral Theory?

  61. The YellowShark,

    Can you please provide a testable hypothesis pertaining to blind, undirected processes?

    I believe by attempting to do so you wil see the scientific vacuity of the anti-ID position.

  62. #61

    I’d add to this that the testable hypothesis must not consist of pathways less than 100 informational bits and only several beneficial and/or neutral steps, which we all agreed should be feasible for non-foresighted processes. I say this because that’s all Darwinists keep hyping while they ignore the real problems.

Leave a Reply