|
Author
|
Topic: Still Spinning Just Fine: A Response to Ken Miller
|
William A. Dembski
Member
Member # 7
|
posted 17. February 2003 17:03
When I read Ken Miller's contribution to the volume I'm editing with Michael Ruse (Debating Design: From Darwin to DNA, Cambridge University Press, forthcoming 2004), I expected I'd have till the actual publication date next year to respond to it. But since Miller's contribution has now officially appeared on his website (http://www.millerandlevine.com/km/evol/design2/article.html -- it is titled "The Flagellum Unspun: The Collapse of 'Irreducible Complexity'"), I decided to comment on it at this time. For my response go here.
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 17. February 2003 18:13
There are severe problems with basically every paragraph of Bill Dembski's latest, I will focus on just one:
Dembski writes in Still Spinning Just Fine that:
quote:
The Argument from Personal Incredulity:
Miller claims that the problem with anti-evolutionists like Michael Behe and me is a failure of imagination -- that we personally cannot "imagine how evolutionary mechanisms might have produced a certain species, organ, or structure." He then emphasizes that such claims are "personal," merely pointing up the limitations of those who make them. Let's get real. The problem is not that we in the intelligent design community, whom Miller incorrectly calls "anti-evolutionists," just can't imagine how those systems arose. The problem is that Ken Miller and the entire biological community haven't figured out how those systems arose. It's not a question of personal incredulity but of global disciplinary failure (the discipline here being biology) and gross theoretical inadequacy (the theory here being Darwin's).
However, when confronted with rather a lot of literature on the origin and evolution of the immune system (one of Behe's originally identified IC systems from Darwin's Black Box), plus lots of evidence of precursors in organisms without the full system, not a single IDist was able to defend the previous statements of Behe and Dembski along the lines of "the entire biological community ha[s]n't figured out how those systems arose." A few of the more sophisticated ID-friendlies appeared to agree that the gradual evolution of the "IC" immune system was a perfectly reasonable idea supported by a fair number of observations and peer-reviewed articles.
This was all done right here on ISCID just a few months ago:
Organisms using GAs vs. Organisms being built by GAs
And yes, this will keep getting brought up as long as Dembski keeps repeating the same false line that biologists are clueless about how complex multipart systems can originate.
As for what things like the Type III secretion system do and do not prove (as well as for citations of important bits of evidence that Dembski failed to deal with, things like the Exb homologs of the flagellum motor proteins, and the archaeal flagellum--Type IV secretion system homologies), there's not much point in repeating the evidence yet again, so I've been accumulating a list of the relevant links here:
Antievolution.org resource thread on the prokaryote flagella (there's more than one kind of flagellum, durnit!!! )
[various typo edits]
[ 17. February 2003, 18:22: Message edited by: yersinia ]
IP: Logged
|
|
William A. Dembski
Member
Member # 7
|
posted 17. February 2003 21:49
Briefly Yersinia:
**Miller's paper as well as my response focus on the bacterial flagellum, not the immune system, not the eukaryotic flagellum, not the archaeal flagellum, not any number of other systems.
**Finding further systems besides the TTSS related to or embedded in the bacterial flagellum does not affect my argument. The point is to find a continuous Darwinian trajectory, not isolated islands. Your gap-fillers are islands, not bridges between them.
** Even if the trajectories are continuous, it still remains to be shown that they are Darwinian. I'm entirely comfortable with the evolution of the immune system, for instance. In fact, I'm comfortable with full common descent. My argument, always, is that evolutionary pathways to irreducibly complex systems, even if they exist, are non-Darwinian.
**Stop with the literature bombing and for once offer a coherent argument in your own words. Why should anyone not already converted to Darwinism believe that a continuous Darwinian trajectory brought about the bacterial flagellum? Because there are subsystems that can serve other functions? But how were they coordinated? By natural selection? But natural selection is a trial-and-error mechanism with no foresight.
**No doubt you know the answers to these questions and have searched PubMed for all the relevant references (the page you developed on the antievolution website is very impressive indeed). So do us a favor and put it all together in a coherent argument using your own words. My guess is that you won't be able to do it. Indeed, I've found a striking parallel between your inability to formulate coherent arguments in the service of natural selection and the inability of natural selection to coordinate random variations in producing irreducible complexity. That parallel is not coincidental. Nonetheless, I welcome you to prove me wrong. In fact, if you can formulate the argument in an 8000 word paper, I may be able to convince Miller to withdraw his paper from the Cambridge Press volume I'm editing and substitute yours -- just between you and me Miller's paper was a disappointment. Indeed, one expects better from Darwinism's top debater. Perhaps your are ready to displace him in, dare I say it, an evolutionary competition.
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 17. February 2003 22:25
quote:
Briefly Yersinia:
**Miller's paper as well as my response focus on the bacterial flagellum, not the immune system, not the eukaryotic flagellum, not the archaeal flagellum, not any number of other systems.
So what? You are making an argument about all 'IC' systems, of which the flagellum is just one (particularly ancient) example. You and Behe are (or at least, were -- you seem to be changing your tune) making an argument that IC means something with respect to evolvability.
You said,
quote:
The problem is that Ken Miller and the entire biological community haven't figured out how those [IC] systems arose. It's not a question of personal incredulity but of global disciplinary failure (the discipline here being biology) and gross theoretical inadequacy (the theory here being Darwin's).
(Dembski's italics)
You can't make a statement like this and then just just pick the IC systems that you think support your point and ignore the ones that don't. If your IC argument is wrong about the immune system (and blood-clotting and middle-ear bones and PCP, atrazine, and 2,4-dnt degradation...), then you have no grounds on which to argue for "global disciplinary failure".
[ 17. February 2003, 22:26: Message edited by: yersinia ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 17. February 2003 22:32
Dr. Dembski,
As for your other points, developing an essay that not only
(1) proposed a more detailed scenario for the evolution of the flagellum but
(2) dealt with all of the usual wrongheaded objections and confusions ("evolution can't produce information", all of the various conflicting definitions of IC, etc.) and furthermore
(3) explained all the relevant basics of biochemistry and phylogeny (the phylogeny being rather unresolved at the moment) that would need to be explained in order for anyone but biochemists to understand it and
(4) also laid out all the groundwork for what evolution actually regularly does (e.g., cooption, which you summarily dismissed as ideology in your flagellum article despite tons of peer-reviewed literature documenting in detail its occurence and mechanisms of occurence).
...would be a massive task. I do mean to do it at some point, people are free to follow the links in the antievolution.org flagellum thread to get some idea of how it would go.
However, given your complete lack of engagement with the immune system literature (and the long, very coherant argument in Matt Inlay's immune system article), I doubt that a flagellum article would get anything more than the usual brush-off. So I'm not in a hurry.
However, perhaps I can put together something and post it.
[added in PS: don't hold your breath however, it won't be this week]
[2nd edit: Reciprocal link to: ARN discussion
[ 17. February 2003, 23:03: Message edited by: yersinia ]
IP: Logged
|
|
Evan
Member
Member # 164
|
posted 17. February 2003 23:09
William Dembski writes,
quote:
I'm entirely comfortable with the evolution of the immune system, for instance. In fact, I'm comfortable with full common descent. My argument, always, is that evolutionary pathways to irreducibly complex systems, even if they exist, are non-Darwinian.
Dr. Dembski, this is a very interesting statement.
Early in my participation on ISCID I started several threads on a synthesis of evolution and design.
A) One centerpiece of my hypothesis was that if one assumes that the intelligent designer has access to genetic activity in all organisms at all times (or at least some organisms at some times - one or the other seems to be implicitly given in your theory), then that designer could manipulate genetic material so that slow, planned changes in organisms would not only follow a path of common descent, but that every birth of every organism would have nothing exceptional about it. That is, following some of your ideas in ID Coming Clean, I proposed that every single “design event” would be non-miraculous (in the sense that you describe in that essay) and therefore indistinguishable from a naturally occurring event. Only the aggregate result would indicate design - only an evolutionary pathway in which the sum sequence of births (and therefore discrete genetic events) leading from organism A to organism B, perhaps many generations later, would be the sign that design had occurred, even though many (perhaps very many) smaller, undetectable design events would have occurred along the way.
B) Without meaning to put words in your mouth, it seems like you are saying something similar here: that even if eventually biologists map out a fairly complete evolutionary pathway for the immune system or the middle ear or the flagellum, your position would be that intelligent guidance would have been necessary for that to happen.
So I have two questions:
1) Are you in general agreement with paragraph A above, or any part of it, even as a potential hypothesis for ID?, and
2) Are you in general agreement with the statement in paragraph B above, or any part of it.
I realize that you might not want to answer these questions directly, but I would appreciate any comments you might like to make. It seems like there is a key issue concerning the implementation of design that might benefit from being elucidated here.
Thanks very much,
Evan
[ 17. February 2003, 23:15: Message edited by: Evan ]
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 18. February 2003 01:38
Dear Bill
Others have already pointed out the many flaws in your arguments but I would like to respond to your comments to Yersinia.
You comment that your arguments are wrt the flagellum but Yersinia's comments about the immune system are highly relevant since they show how apparantly IC systems can arise through Darwinian pathways.
You refer to Yersinia's references as 'gap fillers' but that would make your argument 'ID of the gaps' would it not? You find 'evidence' for ID in the gaps yet to be filled by science but fail to provide either for any alternative nor for any reason to believe why these gaps may not be filled. Perhaps if ID could provide us with its own pathways or mechanisms but if I understand your more recent arguments correctly you do not believe that ID should be held to such expectations/standards?
Since Darwinian pathways to IC systems could be envisioned, your statement that you consider your argument to be that these pathways need to be non-Darwinian remain not only without much empirical support but seem to be actually contradicted.
You may consider Yersinia's use of scientific references to be 'literature bombing' but Yersinia has shown how science seems to contradict much of your ideas. One could hardly object to someone actually supporting their viewpoints with actual data?...
Yersinia has provided us with far more of a coherent argument than I have seen from the pro-ID side.
Just between you and me, your response to Miller's paper has been a disappointment. I was hoping that since your RAPID conference you would be working on actual empirical evidence. So far it seems to me that all that has happened is a redefinition of IC to artificially prohibit Darwinian pathways by insisting on that the function remains the same. However the data suggest that this is not necessarily how evolution works.
Your words thus sound particularly hollow in the light of the arguments presented here. As Yersinia points out so aptly, you cannot just chose the IC systems of your liking and ignore all the other IC systems which fail the expectations.
So far science has shown likely pathways to explain the data, ID has remained helpless at the sidelines.
From ICness to 'conservation of CSI' (aka the 2nd law of thermodynamics) it seems that ID has a little to offer.
Your attitudes so far seem to further undermine your tentative 'arguments'. Lets focus on addressing Yersinia's excellent observations and less on meaningless posturing? I am sure that the moderator would feel similarly about maintaining focus.
I have to state that Yersinia's and Miller's statements seem far more coherent than anything I have seen in opposition of it.
I would love to explore btw the extent of your '4th law' or should we say '2nd law' of information "conservation" as well but now that I have emerged from the private discussion forums into the public forums, that may have to wait until we have flushed out the issues of ICness and their relevance or irrelevance to Intelligent Design.
On ARN Charlie D mirrors my sentiments when he states that
quote:
My question is, if there is, as now Dembski admits, no logical difficulty to the naturalistic evolution of any IC system, and it all boils down to biologists not "having figured out" how such systems actually evolved, what is it that really makes IC systems special, in any meaningful theoretical sense? To me, this final concession by Dembski & C is what ultimately has killed the significance of IC (and the IC-derived ID inference), far more than all the new accumulated evidence for flagellar evolution (easily available here, thanks to Nic's patience), the immune system's (here), etc, have.
Is ID really a science stopper as suggested by RBH on the same thread?
[ 18. February 2003, 02:09: Message edited by: Pim van Meurs ]
IP: Logged
|
|
Rex Kerr
Member
Member # 632
|
posted 18. February 2003 04:57
Dembski writes, in his response to Miller:
quote:
Bottom line: Calculate the probability of getting a flagellum by stochastic (and that includes Darwinian) means any way you like, but do calculate it. All such calculations to date have fallen well below my universal probability bound of 10^(-150).
Certainly!
Based off of information in this description of the flagellum (found thanks to Yersinia), we see that the flagellum consists of at least:
Propeller: FliD (cap); FliC (filament); FlgL,K (joint); FlgE (hook)
Drive shaft: FlgG (joint); FlgB,C,F (main shaft); FlgH,I (bushing)
Rotor: FliF,G,M,N
Stator: MotA,MotB
Transport apparatus: FlhA,B, FliH,I,O,P,Q,R
First I will lay out an evolutionary scenario:
- A pore duplicates in an early bacterium and becomes associated with an enzyme whose products provide a fitness advantage when transported outside the cell.
- This pore acquires several associated subunits to regulate what is exported through it.
- One of these associated subunits develops the ability to bind proteins with a specific export signal sequence; other proteins pick up the signal sequence. [Transport apparatus done.]
- The bacterium acquires an outer membrane, and the pore picks up an extra layer.
- The bacterium acquires a cell wall, and the pore picks up yet another layer. [Bushing done]
- A new subunit adds a lining to the inside of the pore and reduces interactions between other proteins and the pore. reducing clogging.
- This subunit extends throughout all three layers. [Shaft done]
- A subunit that binds proteins of other cells is added to the cap of the export system, allowing the system to inject proteins into other cells. [Cap done]
- A buffer layer is added under the cap, making contact with other cells mechanically easier.
- The buffer filament is added underneath the cap, making contact with other cells mechanically easier. [Filament done]
- Motor proteins associate with the inner edge of the system, generating twitches that facilitate the mechanical rupturing of the target cell membrane.
- The motor proteins duplicate and diverge into structural-reinforcement and force-generating components, allowing complete rotational motion (thanks to the slipperiness of the pore-lining subunits). [Stator done]
- The pore-lining subunits duplicate and diverge such that friction is reduced with the pore subunits that form holes in the cell wall and outer membrane. [Rotor addressed]
- A subunit that adds a kink under the filament is generated, transforming the injection system into a functional motor. [Hook done.]
Note that we have a functional intermediate at every step that I have listed.
This produces a system that is as complex as the filament, and meets any specification that one would generate for the bacterial flagellum. Thus, if this system does not exceed the universal probability bound, we cannot use the explanatory filter to infer design of the actual bacterial flagellum, as the probability that something will fit the specification will be greater than the U.P.B..
Let's reduce these to the following actions:
Dupl_A -- duplicate a gene A.
Domain_A+X -- gene A picks up an existing domain X.
Bind_A->B -- novel specific interaction between A and B (with A changing)
Sig_A+Q -- gene A picks up signal sequence Q.
Coev -- coevolve an existing system via optimization of fitness.
Mech -- transmit mechanical force between components.
In each case, a number in parentheses after a gene or whatever indicate how many choices we have. Now let's (very roughly) estimate probabilities; most of these are a semi-informed guess, hopefully on the conservative side:
p( Dupl_A(n) ) = n/10^12 ; bacterial conjugation is really messy, and we have bacteriophages in the wild, and so on, so I will guess that the rate of duplication for any given gene is about one in a trillion.
p( Domain_A+X(n) ) = n/10^6 ; assuming we're already duplicating, or otherwise moving around, the chance of hitting a specific domain with a few amino acids as spacers is a few times 1/genome size.
p( Bind_A(n)->B(m) ) = n*m/10^18 ; immunology suggests that you can get good specificity with trying out about a billion different peptide configurations, and I'll assume that the frequency of (potentially junk) insertions into a single gene is 1/10^9 (1000x less frequent that point mutations). [Note: if we're already moving around, the probability changes to n*m/10^9.]
p( Sig_A+Q(n) ) = (n+1)/10^6 ; n=0 means a novel sequence, which we have to generate. Specificity tends to be less than 4 a.a., or less than 1/10000 in known signal sequences...but even if you're moving the chance that you'll get extra amino acids stuck on you in the right spot is not guaranteed. n>0 can be dropped into like adding a domain.
p( Un___ ) = 1/10^6 ; chance to lose a function ___.
p( Coev ) -- 1/10^14 per pairwise mutation (mutation rate 1/10^6 per gene, times 1/100 success rate of improving the interaction). With billions of bacteria having millions of generations, we get coev for free, so we'll assume it at the end of every step with survival value.
p( Mech ) -- 1/10^6 ; this is actually pretty easy, but to get usable motion may be tricky.
Now let's see how we're doing:
- Dupl_A(1) , Domain_A+Z(1) ; p = 1/10^18 (duplicate pore, associate with existing enzyme with existing binding domain).
- Dupl_B(1) , Bind_B->A ; p = 1/10^21 [repeat] (add associated subunits; will be lower with successive subunits since there will be more targets, and may be some protein binding domains we can use instead of novel interactions).
- Bind_B(10)->Q(1000) , Dupl_C(10) , Sig_C+Q(0) ; p = 1/10^31 (get export system going with a random signal sequence)
- Dupl_D(1) , Bind_D->A ; p = 1/10^21 (Stick outer membrane pores to existing pore)
- Dupl_D'(1) , Bind_D'->A ; p = 1/10^21 (Same for cell wall)
- Dupl_E(100) , Seq_E+Q(10) , Bind_E->A ; p = 1/10^24 (Lining, which needs signal sequence)
- Bind_E->D ; p = 1/10^18 (Extending lining)
- Dupl_F(10) , Bind_F->E , Bind_F->?(10) ; p = 1/10^28 (Duplicate something that has an export sequence and let it bind to something outside and the lining.)
- Dupl_G(10) , Bind_G->E , Bind_G->F ; p = 1/10^29 (Buffer between cap and lining)
- Bind_G->G ; p = 1/10^18 (Convert buffer into filament)
- Dupl_H(10) , Bind_H->A , Mech ; p = 1/10^26 (Add motor)
- Dupl_I(1) , Bind_I->?[10] ; p = 1/10^20 (Duplicate motor and have one half bind cellular structural components)
- Dupl_J(1) , Bind_J->E , Unbind_J->D,D' ; p = 1/10^27 (Covert lining to a lower-friction system.)
- Dupl_K(10) , Bind_K->J , Bind_K->G ; p = 1/10^29 (Add hook)
We might multiply all these probabilities together and generate 10^-331 or somesuch.
But this isn't the correct calculation, because there is selection at each step. Recent estimates are that there are about 10^30 bacteria on the planet. We'll assume that our new and improved bacteria can take over one billionth of the planet each stage, so we'll allow 10^21 bacteria to provide our replicational resources for this problem. Generation times tend to be under one day, so in only ten million years or so, we have 10^31 replicational resources.
But that means that the hardest step in our step-by-step approach would be expected to happen in only ten million years! The entire pathway could easily take place in a hundred million.
The only calculation we are left with is the probability of such a pathway existing--that there is a functional intermediate at each of those stages. Now, we see examples of virtually every one of those steps being advantageous in some system, but maybe at each stage there's only a 1/1000 chance that anything advantageous would actually arise from the scenario I've described. That gives a probability for the whole mess being a working scenario of 1/10^42, well above the Universal Probability Bound--even when you take out the replicational resources that I've already used up with stepwise evolution.
So, here is a calculation that falls well within the probability bounds. It would be nice to get more accurate numbers on each of the transformations I described above, but then it would be nice to get more accurate numbers on the perturbation tolerances that Dembski used in NFL.
I have met the bottom line. Perhaps now we can discuss whether Dembski's numbers or mine are more instructive.
Edited to add a link to estimate of number of bacteria.
Added in edit: it looks like I forgot to include the full derivation of the rotor, but it's the same kind of thing. You get the picture: duplicate, modify, splice, etc., with functional intermediates as frequently as possible.
Note also that in my scheme, the flagellum doesn't function to transduce force until the very last step! It goes from being a generic secretory system to a cell-cell transporter to a mechanical-attack system to a flagellum. Of course, one can imagine many other such pathways.
[ 18. February 2003, 05:11: Message edited by: Rex Kerr ]
IP: Logged
|
|
Moderator
Administrator
Member # 1
|
posted 18. February 2003 06:44
Just a reminder to participants in those thread: please keep a civil tone (I think Rex has done the best job so far) and stay on focus. The issue in this thread is the bacterial flagellum, not, for example, whether ID is a science stopper as Pim implies.
The point, as is made time and time again on this board, is that Brainstorms will not be used to make sweeping polemical shots to the groin. Massive complaining is also frowned upon: this applies to all parties involved in the discussion.
IP: Logged
|
|
William A. Dembski
Member
Member # 7
|
posted 18. February 2003 09:48
Rex:
There's not much agreement between our camps, but perhaps we are agreed that spontaneous generation probabilities for IC systems, whether at the level of individual proteins or higher level modular subsystems have probabilities below my universal probability bound. So the question is (A) whether we can describe a continuous Darwinian trajectory issuing in an IC system like the flagellum and (B) whether it is described with sufficient causal specificity so that we can have confidence that it can actually build the system in question, and (C) whether the trajectory's probability falls above the universal probability bound.
Unfortunately, your scenario doesn't even get past point (A). Indeed, there's no reason to think that there's a selective advantage going from one of your steps to the next. Now I'm sure you can tell a story about why there might be a selective advantage. But let's be clear that handwaving just-so stories don't become rigorous scientific reconstructions because one dubious evolutionary step is replaced by multiple dubious evolutionary steps or because specious selective pressures are invented to glue contiguous steps in the evolutionary scenario together.
Call me a stick in the mud, but it seems to me that all you've done with your 14 steps is break down an assembly problem based on a preexisting design and then list some assembly instructions in order. And then for good measure you invoke natural selection because this is all supposed to be happening under the guidance of natural selection, or as you put it more simply, "because there is selection at each step." Yes, selection always operates. But there needs to be a selective advantage in going from one step to the next, and there's no reason to think that obtains in your scenario (e.g., where's the improved fitness of going from step 1 to step 2; indeed how is such improved fitness at all ascertainable given the total lack of specificity in the additional subunits you invoke?) .
Let me emphasize that we are still not past point (A). Your scenario collapses at point (B). Take your step 2: "This pore acquires several associated subunits to regulate what is exported through it." Oh, really. How many subunits? And what exactly do they do? And how do they contribute to the fitness of the organism? As for the probability you calculate for the scenario, it is literally pulled out of the air. With my probabilities you can at least find experimental and theoretical justification in work by people like Robert Sauer and Hubert Yockey.
IP: Logged
|
|
VonRSmith
Member
Member # 671
|
posted 18. February 2003 12:27
Ladies and gentlemen, and children of all ages:
This is my first post here, and, hiding shamelessly behind the excuse of being a neophyte to this forum, I am going to take the extremely impolitic step of publicly taking issue with the moderator, specifically concerning the subject of this thread:
"Just a reminder to participants in those thread: please keep a civil tone (I think Rex has done the best job so far) and stay on focus. The issue in this thread is the bacterial flagellum, not, for example, whether ID is a science stopper as Pim implies."
As I read this thread, this statement is incorrect. The subject of discussion is Dembski's reply to Miller's latest article. As I see it, Dembski is submitting both Miller's article and his response for consideration in the message that starts this thread, so any topic broached, any claim or argument made in either essay should be considered fair game. I wasn't aware that there was a specific "topic question" to which posters were to confine themselves. Have I, in my inexperience with this forum, missed something?
Both the Miller article and Dembski's response do indeed broach and discuss the broader issue of whether or not IC poses problems for mainstream evolutionary research programs, so I'm not sure why the moderator objects to these issues coming up in the thread.
Dembski may claim later in this thread that the "focus" of both Miller's article and his are the flagellum rather than other systems, and that is true in the sense that it is the most important example under discussion; however, I read both of them as using the flagellum as an example of the merits (or lack thereof) of arguments from IC in general, not as a stand-alone argument upon which the merits of evolutionary biology hinge crucially.
Also, Miller *does* write in his article about other apparently IC systems, such as the clotting system, the Krebs cycle, and flagella in archaeobacteria; he even goes so far as to suggest that each of these other examples bears upon the issue of whether or not the eubacterial flagellum can be safely regarded as IC. So I don't understand why Dembski or the moderator would object to these other examples being called into issue in the discussion here.
It is my view that a moderator should hold all participants in a discussion to certain standards. These include staying on topic and keeping the tone civil. They also include holding all participants, especially the one who introduces a new topic, answerable to any and all material submitted for discussion. It was my understanding as I read the top of this thread, and apparently that of other posters as well, that Dembski was submitting the entirety of Miller's paper, and his response to it, for review and discussion, not just the bits about the flagellum specifically. I hope the moderator will not allow Dembski to *subsequently* limit the scope of the discussion and leave out issues raised by the initial material.
I also hope that the moderator will err on the side of inclusiveness and tolerance in moderating this post. I hope to see a public explanation of why anything I have written above is off-base, or how my impression of the situation is mistaken.
IP: Logged
|
|
Pim van Meurs
Member
Member # 541
|
posted 18. February 2003 12:51
Dear Von,
You express quite well some of my feelings. As far as I can tell the focus of this thread is the response of Dembski to Miller and the bacterial flagellum is but a part of the issue. I certainly support the moderator's suggestion to remain civil and focused but as Von I am somewhat surprised by the moderator's suggestion about what the topic of the thread should be.
To return the focus let me quote from Dembski's response
quote:
What's more, the function of transporting proteins has as little directly to do with the function of rotary propulsion as a toothpick has to do with a mousetrap. So discovering the supportive function of transporting proteins tells us precisely nothing about how Darwinian processes might have put together a rotary propulsion machine."
I think Dembski has missed the point here namely that if evolution reuses available pathways for new functions Darwinian processes and pathways are indeed not beyond our reach in explaining the origins of IC systems. By focusing on terms such as basic function etc Dembski may have attempted to 'Darwinian proof' IC systems but at the same time he has indicated the greatest weakness in such an approach.
Showing that the flagellum is not only involved in movement but also in exporting proteins does show how an original function may have been co-opted by a new function without providing a problem for Darwinian evolution. If we were to accept the limited definition of IC as proposed by Dembski we have artificially limited our ability to understand its origins. If ICness can be circumvented so easily by non-basic functions then it cannot be argued that no Darwinian pathways could exist.
So in the end ICness is not an argument which has eliminated evolutionary mechanisms in one grand swoop and thus arguments based on IC remain individual examples of "we don't know yet" although some IC systems such as the immune system do seem to be quite well understood in terms of Darwinian evolution.
If the argument is that we should focus only on IC elements for which science has yet to find the detailed pathways and that we are only allowed to pathways which preserve the 'original'/basic function then ICness seems to have lost its original attraction.
Thus when Dembski states that quote:
The scientific literature shows a complete absence of concrete, causally detailed proposals for how coevolution and co-option might actually produce irreducibly complex biochemical systems
, he ignores the various proposed evolutionary pathways (theoretical as well as actual) for IC systems. Dembski may consider them to be wishful speculations but science is doing more than wishing, it provides for testable hypotheses, works hard on finding out more about the unknown and does not sit back under the presumption that science may never know how/what happened.
So all the objections about IC systems and Darwinian pathways seem to have come down to "show detailed pathways" but as I read Dembski, ICness is NOT per definition unreachable by Darwinian mechanisms, but for specific examples such as the flagellum we do not know the specific details yet. Dembski wonders "But making the transition from existing function to novel function is the hard part. How does one get from functional pieces that are selectable in terms of their individual functions to a system that consists of those pieces and exhibits a novel function?" and this is exactly where the various examples of how in theory this could be done are relevant.
[ 18. February 2003, 13:00: Message edited by: Pim van Meurs ]
IP: Logged
|
|
Moderator
Administrator
Member # 1
|
posted 18. February 2003 13:35
Von,
Welcome to Brainstorms. In the future, do not address the moderators publicly. Please send all correspondence to moderator@iscid.org
Please also note that all further public correspondence will be ignored and deleted.
First though, let me explain my often imperfect actions. The moderation was mostly made in response to Pim. I felt as though his post was taking the thread down a hostile road. I would not have said anything had it not been for his post. However, having decided to remind Pim to remain civil, I also decided to make the warning more inclusive (i.e. all the posters in this thread, including Dembski).
Let me also put the moderating in context. Brainstorms is not meant to be a general discussion forum. The rules that apply in other forums have little bearing on the ethos that we are trying to develop at this forum. For example, it is common, due to the nature of the internet (i.e. no immediate physical response), for people to take potshots at each other. When people do this at Brainstorms, it quickly reveals that they are not all that interested in having civil discussions or participating in constructive dialogue (or, they are too habitually reactive and have trouble entering into a different mindset). Brainstorms moderators are constantly on the look-out for people that consistently lead discussions towards hostility.
Having said all this, let me also say that I have not read the Dembski essay, and probably erred in moderating based on what I deemed to be the thread's domain. It seemed, on a glance, that the thread's domain was the bacterial flaggela, and this was confirmed by the topic starter in a complaint. Perhaps I was wrong...
So, in conclusion, what I want is for people to respect each other, but by all means, if you feel that you have insightful criticism that bears directly on the thread, add it. I don't want to restrict discussion, I only want it to stay focused and civil. That was the purpose of my warning, and that is all I ask the participants of this thread to adhere to.
It seems that fortune gives too much to to many, enough to no one.
That is a great slogan for forum moderators!
[ 18. February 2003, 13:39: Message edited by: Moderator ]
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 18. February 2003 13:56
I have several web pages essays (up for over a year) that address several points that are relevant to this thread. Here is a portion from one of them:
Before beginning a scientific analysis of this scenario, it is worth considering the larger philosophical picture. Many people labor under the impression that IC means "something that even a hefty dose of Darwinian imagination cannot possibly explain." That is, many seem to think that if they (or anyone else) can invent a vague scenario for how the flagellum (or something like it) could have possibly evolved, all the issues brought to the forefront by an IC analysis disappear. But keep in mind that whenever you are dealing with a machine, it is always going to be possible to imagine the various parts existing without the machine, as long as you keep your explanation vague and are free to imagine simpler states with imaginary selective benefits and ad hoc functions. Consider how Ken Miller explains the origin of the bacterial flagellum:
quote:
But Brown's Mr. Miller contends that the ID argument completely misunderstands how Darwinian evolution works. The various parts of the flagellum weren't destined to serve together as a motor when they first appeared piecemeal in the deep history of bacteria. Instead, the parts served different and separate purposes originally. Only later did they happen to come together to form a motor. Remnants of that process are still visible today in various lines of bacteria, he says. Yersinia pestis, the species that causes bubonic plague, produces a complex structure with 10 proteins closely related to the ones found in bacterial flagella. The deadly bug doesn't use those proteins to move; instead, they help Y. pestis inject its toxins into the cells of its host. Drawing on Mr. Behe's favorite analogy, Mr. Miller says that the various parts of a mousetrap have uses even on their own. Three out of the five components form a handy tie clip. Two of the five can serve as a clipboard. Nature is opportunistic, he says, cobbling together different molecules and reactions for entirely new purposes.
What is interesting about this logic is that we already know that the mousetrap was intelligently designed. We also know that it did not first exist as a clipboard, then a tie clip. Thus, while it is logically possible to see the mousetrap as Miller does, that is, as a modified clipboard and tie clip, such perceptions are not tied to history nor the origin of the mousetrap. Thus, coming up with imaginary accounts that tap into our ability to imagine cooptional origins, by itself, is rather meaningless. If we can successfully come up with such explanations where they are known to be false(the mousetrap), how do we know that our ability to do likewise with things like the flagellum are not also inherently flawed?
[snip]
Thus, it would seem that we'd need to establish two things for the plausibility of the EFM hypothesis.
1. Would "any ol' transporter" really do? That is, could we take the framework of any ol' transporter and put the type of flesh on it that is exhibited by the bacterial flagellum? Or would most of these transporters be dead-ends in the sense that their transport mechanism entails a constraint that would prevent the evolution of something like the flagellum-as-we-know-it? Again, this latter point seems to be the case since none of the other transport systems evolved something comparable to flagellum among eubacteria.
2. Is there any reason to think the type III export system, complete with the ancestors of flhA, flhB, fliR, fliQ, fliP, fliI and others, existed as a "cooptable part." Thus far, the answer is no, as there are good reasons to think the type III system evolved from pre-existing flagella.
a. The bacterial flagellum is found in both mesophilic and thermophilic bacteria, gram-positive and gram-negative, high GC and low GC content bacteria, and spirochetes. Type III systems seem to be restricted to a few gram-negative bacteria. That is, if we look at the sequenced genomes from the various groups cited above, we can find the genes for the bacterial flagellum but not the type III system genes.
b. Independent evidence suggests the type III system is recent. It is not only restricted to gram-negative bacteria, but to animal and plant pathogens. In fact, the function of the system depends on intimate contact with these multicellular organisms. This all indicates this system arose after plants and animals appeared. In fact, the type III genes of plant pathogens are more similar to their own flagellar genes than the type III genes of animal pathogens. This has led some to propose that the type III system arose in plant pathogens and then spread to animal pathogens by horizontal transfer.
c. When we look at the type III system its genes are commonly clustered and found on large virulence plasmids. When they are in the chromosome, their GC content is typically lower than the GC content of the surrounding genome. In other words, there is good reason to invoke horizontal transfer to explain type III distribution. In contrast, flagellar genes are usually split into three or more operons, they are not found on plasmids, and their GC content is the same as the surrounding genome. There is no evidence that the flagellum has been spread about by horizontal transfer.
d. It's much easier to envision the evolution of the type III system from flagella than vice versa. For starters, evidence has surfaced that the basal body of the flagellum already works to secrete proteins other than the flagellar proteins, including virulence factors. Thus, the basal body is already poised to evolve into a type III system from the start. Evolution apparently would only have to duplicate and tweak the type III virulence protein secretion activity already existing in flagella. . In my opinion, this view is far more parsimonious than to propose that something like the type III system evolved long ago, was lost by all bacteria but gram-negative animal/plant pathogens and then was used to evolve the flagellum so that horizontal transfer could spread flagella far and wide (despite the lack of evidence for such transfer).
Thus, it should not be surprising that the scientific opinion has been converging on the notion that the export machinery evolved from the flagellar machinery [5-7].
Is there any evidence that supports transporting this system, or something like it, back in time? The type III system is one of at least four different bacterial protein transport systems. And it appears to be the most complex of the bunch. The key here is that the type III/flagellar cytoplasmic export system does not show clear homology with any of these other transport systems. But we also know that evolution builds on and modifies what already exists rather than create de novo. Thus, if these other transport systems were already in place (and they probably were), why didn't evolution simply build on one of the simpler versions rather than create a whole new method of protein secretion de novo? The type III-from-flagellum scenario better fits with what we know about evolution - that it uses what already exists rather than inventing de novo. Thus, not only is there no evidence to support putting this transporter (or something closely homologous) back in pre-flagella days, there is reason to think it wouldn't be there.
[ 18. February 2003, 13:57: Message edited by: Mike Gene ]
IP: Logged
|
|
gedanken
Member
Member # 594
|
posted 18. February 2003 17:39
Dr Dembski said:
quote:
There's not much agreement between our camps, but perhaps we are agreed that spontaneous generation probabilities for IC systems, whether at the level of individual proteins or higher level modular subsystems have probabilities below my universal probability bound. So the question is (A) whether we can describe a continuous Darwinian trajectory issuing in an IC system like the flagellum and (B) whether it is described with sufficient causal specificity so that we can have confidence that it can actually build the system in question, and (C) whether the trajectory's probability falls above the universal probability bound.
Dr Dembski,
First I must point out that I am not an expert on evolution, or even one who has studied it extensively. My experience is in viewing such discussions, and I am like yourself one who has studied mathematics (though in my case in other related disciplines that relate to information and communication). My mathematics background comes from studying physics, engineering, computer science, and analysis of algorithms subjects that relate to issues that relate to this discussion. So even though I lack in some credentials on subjects involved, I think that my arguments should fairly be examined for their validity. I just didn’t want anyone to think I was arguing from authority on any biological issue.
Your “explanatory filter’s” use of the Universal Probability Bound (UPB) is only (potentially) meaningful if one can “sweep the field clean” of possible explanations. This is according to your own descriptions.
Your argument given above seems to hinge on whether scientists (or presenter) has sufficiently well characterized the known science as to conclusively show how the flagellum could have developed. This is your argument in essence with regard to steps “A” and “B” as you described, and then you regard a question “C” about the probability being above the UPB.
A first difficulty that I think must be mounted is to define probability in this context sufficiently clearly.
A reasonable model of events, from a probability view (not necessarily one that can be connected directly to the known science) is to consider a Markov sequence. This is, of course, only applicable if we consider a sequence of discrete steps in the sequence of events. We might fairly make an approximation of the evolutionary theory (as somewhat of an abstraction) by considering the Markov sequence of events as the population (or individuals) pass through states. Thus we can consider a set of initial states at some level of sequence in evolutionary sequence. And from those initial states, we can consider evolutionary developments as fanning out to different possibilities (in a tree-like pattern) to varying states of the form of the organism (or of the genetic material or the proteins in question -- I’ll let biologists deal in specifics of relevance).
Now since a Markov model considers the fan out of possible states to a given later state in the tree, there is also a fan in of possible states wherein multiple starting states could transition to a given biological state of evolution. In other words there may be multiple intermediate states that can have a possibly describable probability of transitioning to a given state, for example the current state of development of the flagellum (or some other intermediate state in question).
The problem here in the definition of probability here is made clear. What starting state is being referenced when one speaks of the “probability” of the trajectory? Have we clearly defined the set of starting states over which the “trajectory”(s) can have fanned into the final state in question? (To accumulate the probability of the final state, one does not of course just compute a single pathway, but rather the aggregate of all possible pathways that result in the final state.) And if we have clearly defined those starting states, what is the relevance of missing information on other possible starting states?
The problem with “sweeping the field clean” is that one must show that one knows the initial states sufficiently well to characterize them in probability terms. Then one must know all the possible transitions so as to characterize them in probability terms -- at least an upper bound of all these terms.
Now consider another aspect -- the probability of making a mistake in the analysis of these other terms, such as to have not properly understood the upper bound.
The point I would make is that the explanatory filter, for example, to draw any useful conclusion from the low upper bound of probabilities calculated in the steps just described, is sensitive to any error in the analysis.
If an error in the analysis itself occurs at greater than the UPB, then the “known” probability cannot be claimed to be as low as the UPB. In fact since the UPB is chosen so low, the probability of the event is actually more closely related to the probability of an error in the forgoing analysis, than upon the result of the probability analysis.
Combine this with the lack of clarity of what are the initial states (indeed the lack of knowledge of potential initial states) from which the “trajectory” is to be calculated, and our knowledge of the probability of the event (of the flagellum construction) is limited by probabilities in our knowledge of the processes, rather than the probability calculation of those processes.
I know that you have argued that all science results are subject to revision when new information is learned. But this is the essence of argument to the “best explanation”. In normal scientific analysis, the previous concept is often only overturned when one finds an explanation that has higher posterior probability calculated by known pathway through a Markov-Chain-like sequence. In other words the postulated cause has the posterior probability estimated in some way, even if this is done in descriptive terms rather than rigorous mathematical application of Bayes rule. (In other words there are descriptive techniques that correspond to some extent to Bayes rule without calculating the detailed probabilities, and work in an analogous form.)
We must examine two different general problems in the fan-in probability to the flagellum (and some variations if the problem considered is varied somewhat).
1) First problem, as mentioned above, is the problem of decision of initial state (state set) with regard to probability of structure in the class. Dr. Dembski asks for a transition probability from some possibly identified prior state set (the set over which all fan-in probabilities by all pathways must be summed by proper Markov techniques -- to the extent that the problem can be modeled as a discrete-state problem). But ask this question of absolutely any configuration of a biological system, or in fact of any physical system. When one looks backward in successively proceeding state transitions (each of which is of a transition probability less than 1.0) one will eventually accumulate enough state transitions to be below the UPB. So finding some prior state from which the final state is based as a probability below the UPB is not sufficient to be meaningful as a difficulty for evolutionary theory. For example a sequence of more than 500 transitions, each of which is about or below 0.5, will result in such a probability of the final result below UPB of 10^-150. When dealing with an initial state set of multiple starting states, and multiple pathways, one would have to expand the depth of the pathway to reach this final state with probability below UPB, but there is little doubt that a sufficiently early state can be found in which the “trajectory’ by all the pathways that can be estimated with maxima will be below the UPB.
Due to this problem, the claim that the “trajectory” from initial state set to the final state is of low probability (below UPB) is not in itself remarkable. The issue for the study of evolution must then be whether there are or are not Markov state sequences which have pairwise state transitions that are of reasonably high probability. This is precisely what the study of the various possible pathways to the flagellum described by biologists posting in this thread are giving -- possible pairwise state transitions in which the probability is reasonable.
1A) A variation of the first problem is to consider the final state as a “class” of states, rather than a single or narrowly focused state. Such a case would be had by considering a specification of a “rotary propulsion device” as the final state, rather than the existing bacterial flagellum as the final state for the “trajectory” probability calculation. The problem here is that every outcome in evolution that will have a “rotary propulsion device” (as the destination class) would need to be examined. In this case (as well as 2A below) we have the problem that not only have we not identified the set of all possible starting states that could reasonably have an evolutionary pathway to some state containing the described characteristics (specification), we most certainly have difficulty in evaluating the maximal probability of the pathways to those final states -- having great difficulty even imagining all possible biological organisms that could contain a “rotary propulsion device”. The issue of “probabilistic resources” is apparent here. To consider this issue is simply to make the probability of reaching a state in the specification even higher, and does not help Dr. Dembski’s position on the flagellum.
2) The second is the difficulty, mentioned above, of assuring that missing pathways and intermediates have not been left out of the transition matrix. Addition of any transition paths will only increase the probability of reaching the final state. Errors (missing intermediates and transition probability) will for the most part not significantly effect the probability otherwise, if it is calculated as low as the UPB. So the issue of how likely it is that a pathway is missed is central to calculating the probability of the “trajectory”. This is so precisely because the calculation proposed at the given point is claimed to be below the UPB, and having such a low probability in sum over paths makes the existing pathways have extremely low probability, and thus the addition of any mid-level probability pathway will swamp the total. Sensitivity of the result to missing information makes this aspect in essence an “argument from ignorance” form.
2A) Once again, consider targets in a state class rather than a single final state (such as all possible biological organisms that could contain a “rotary propulsion device”). Once again (as in 1A), the probability only increases, and does not improve the argument proposed by Dr. Dembski.
The bottom line is that the claim that the probability of a “trajectory” to the flagellum is below the UPB is sensitive to any and all missing information. Thus the claim is based on the form of an “argument from ignorance”.
The giving of possible pathways to the flagellum does not provide a “solution” to the problem of the flagellum. What it does is to demonstrate that the field has not been “swept clean” of pathways. Changing the subject to incompleteness of knowledge is a diversion, because all scientific knowledge is incomplete at some level, and the potential pathways described are demonstrations of the failure to “sweep clean”. They are furthermore indication of the sensitivity of the probability analysis to future knowledge or missing knowledge. This is true not only when considering the probability of construction of the flagellum as part of the “explanatory filter”, it is equally true when simply making a claim that the probability could be expected to be below the UPB (because they are the same case -- I am just showing that the discussion does not have to be related to the explanatory filter which was not mentioned.)
[ 18. February 2003, 18:35: Message edited by: gedanken ]
IP: Logged
|
|
|
Printer-friendly view of this topic