IAP Political Forum

In another thread I said this:

Actually I agreed that a cloud shape is absolutely complex enough.  My point was that the shape could not be specified independently from the event prior to the actual stipulated occurance.  It was not detatchable in that we cannot say that we regularly observe clouds in the shape of countries. The marker for design requires complexity and specification but the cloud represents only complexity.

To which barney responded:


In the interest of maintaining the original context of the other thread, I answer this here.

As far as we know there are three modes of expanaition.  Two involve material mechanisms and include necessity driven by one or more of the physical laws and chance contingency based on random events.  A second type of contingency is purposeful and we call it design.  The question is can one detect the mode of causation by observing the characteristics of the item or event?  I claim you can by looking for clear markers for design.  One such marker is "Specified Complexity".

In the context of Specified Complexity, a specification is objectively defined in terms of a probability density of a region that includes the event or item in question and similar events that include a common definition that uniquely and directly identifies those events in that region but who's specification is detachable from the event itself.  I have described this to barney several times but he continues to pretend he hasn't received a definition.

In the cloud example, unless we tighten up the rejection region to include cloud shapes that precisely match the exact shape of Austria, we have no objective way of defining what is "Like Austria" and therefore no way to define the rejection region.  This problem arises because the specification can only be made in terms of the event itself because the specification is not detachable from the event.

So in this narrative, I have defined both design the process, and specification a potential attribute of the product of the process and have set them apart from each other.  

Barney's third point is that evolution is a process that mimics design is quite bold.  Experimental biology has confirmed that evolution is quite proficient at breaking and mixing up functional components, but it has utterly failed to generate even the first steps in generating any new multipart components.  There is no indication that evolutionary processes can generate even one new protein binding site and many theoretical reasons why it should not.

Human adaptive immune system has whole pathways with new components absent in many animals, not to mention plants and microbes.

I think that this "nothing new" claim is about he equivalent to comparing Hollywood movie industry with old Greek theaters. The "nothing new" seems to be a very subjective matter. 

Yes, most uniquely identifiable organisms contain pathways, processes and components that are also unique.  The point of this discussion is to peer into empirical science to see if it can tell us how this is the case.  Evolutionary theory supposes that modification and natural selection accounts for these unique features.  If this is the case then we should be observing these processes in action in the process of accomplishing at least partial steps in this progression.  In addition we should be able to describe theoretically how these processes, when allowed to proceed after accounting for a theoretical maximum number of steps could possibly accomplish what the premise claims is accomplished.  The reality is that no material mechanism has ever been found that is capable of generating the kind of unique complex and specified systems that we observe when looking into biological systems.

What we find instead is that these processes do the opposite.  They break function and sometimes in the process of breaking function, they disrupt an otherwise effective attack strategy of a competitor that is applying selection pressure.  When the selection pressure is removed, the broken gene reverts back.

Thus the question of how these unique differences came into being is not answered by evolutionary processes. 


You are now changing the claim.  We clearly see a wide degree of diversity, that is not in question.  I don not claim that there is "nothing new", rather the reality is that by empirical studies (observation and experimentation) we find that evolution has not generated specified complexity, evolution has not generated steps toward specified complexity. and evolutionary processes seem to be missing a host of presumed but unobserved processes that might otherwise generate specified complexity.  In addition theoretical analysis from probability and information theory provide the basis to understand why evolutionary processes have failed.  Finally observation of design processes confirm that design can account for observed diversity.

Before I make a fool out of myself, what is the difference between complexity and specified complexity?

tejtej, I did not intend to draw you into this discussion midway, and was not expecting that you would have an interest in this.  I am happy to explain the premise and will return later to do so.

In this context, we are attempting to differentiate necessity and chance contingency from guided contingency (design).  Chance events behave according to the rules of probability and statistics.  Necessity can be viewed as a chance event with probability of one.  Therefore this concept of complexity when expressed as probability provides a mechanisms to potentially differentiate chance events (including necessity with probability = 1) from design. 

The degree of complexity in this context can be viewed as the probability that a particular event could have occurred by chance.  Highly probable events can easily be accounted for by chance while improbable events become progressively more unlikely to have been a result of some chance process. 

So complexity in this context is simply a measure of the probability that the event in question might have occurred by chance.

Since we know and observe that complex events do often occur by chance, complexity alone cannot be used to accurately differentiate necessity and chance from design.  By observation, we are able to note that while complex but otherwise uninteresting things occur often by chance processes, at the same time we note that complex unique events do not occur by chance but often do by design.  What we need is an objective way to distinguish between these two kinds of complex events.  Specification is an objective method to differentiate and categorize complex events. 

An item or event is specified if one can identify a region of equal probability in the set of all possible outcomes that includes the event in question around the event such that all the events contained by this region share a common precise identifier or description that is also detachable from, that is independently specifiable from the event in question.  That is without knowledge of the event can one define a specification that fits. 

In the example of a cloud in the shape of Austria, the event while very complex, is not specifiable, among other reasons, because without the event we cannot define how closely the shape must match the shape of Austria.  We need the event in question to define the relative degree of fit allowed.  Now it would be specified if the cloud precisely matched the exact shape of Austria and was an exact scale model so to speak.  Se we don't need the cloud to identify the conditions required to specify a scale model of Austria.  I would very much like to hear of an event like that. 

Is human adaptive immune system complex enough?

I don't know.  The objective test for Specified Complexity is a explicit form of the Fisherian probability analysis for identifying chance events.  If you want to make the case that it is complex and specified, I am happy to entertain the idea.

The thing is, if complexity can be measured by some kind of probability, it would help if actual values for one complex and one not complex enough example from molecular biology would be provided.

I would think that chance and natural law are both part of the same domain. Random events at some level are driven by natural laws. The stray photon that alters DNA has itself a causal history. You can't assess probabilities without adequate background knowlege. You certainly can't work backward from a specific function and determine that its causal history is so improbable as to require some sort of unknown intervention by an unknown agent. Can you account for all the complex chemical interactions on a global scale that can lead to the vast number of possible functional proteins? Process of elimination at this level doesn't work.


"We" need a better definition of purposeful. At what level of intent on a scale from the blind operation of natural selection to coral to a spider spinning a web to a bird building a nest to a chimp fishing for termites with his favorite stick to Ken Ham building a Creation Museum do these actions become purposeful? 

I'm not a big fan of deterministic universe. Takes away the fun of living.

Yes, and this is why the test for Specified Complexity is set up the way it is so the we have an objective rather than subjective or a relative indicator. 

First we identify the members of the sample space that contains the event and then we identify as many chance mechanisms as possible that could have generated the event in question.  From this we identify probability distribution in the sample space and the probability of that event occurring by chance.  We will compare the probability of event E,  P(E) with the opportunities available to cause the event or any of the other events in sample space that are equally or more specified.  If there are sufficient opportunities to account for this event E or any similarly specified event then chance accounts for the event otherwise we can infer that another process was involved.   If you are following me and don't have any questions, I will continue next post.

Numbers please.

While I agree that an objective measure of complexity would be a great thing, I think that so far all examples we discussed were the type of:

- not complex enough (say sickle cell anemia)
- to complex to be a result of chance event (flagellum)

Unless probability values are attached, this is a very subjective evaluation of how complex this examples are.

Not really. That's only from the perspective of looking back. There is an almost limitless potential for future fun-ness. That's what keeps me going, anyway.

I was trying to point out the futility of using an eliminative filter such as the one RF proposes to prove that what happened didn't happen particularly since what happened didn't need to happen in the way that it happened (?). Any event is so improbable at some level as to be almost mathematically impossible. Anyway, since the work goes on as far as finding evolutionary mechanisms, there is insufficient background knowlege to assess the probabilities to apply this filter for specified complexity. I don't see the utility of it (and apparently neither do any working research biologists). At what level can you apply it? Certainly not in the observed fossil transitions from reptile to mammal and the hominid sequence. Does it account for novel functions such as the observed evolution of the nylase enzyme? It uses artificially constricted parameters to set a boundary which doesn't really appear to exist in the natural world in order to default the results to concepts of intelligence and design based on faulty analogies of human technologies.

And that's the end of my speech (for about the 40th time). Thanks for listening.

The argument is fundamentally silly. Complexity = God is a worthless explanation, as it quickly collapses in on itself due to using the trait to explain the trait under investigation while simultaneously having no evidence for the non-explanation in the first place.

If God must have done X because it's too complex to have arisen naturally, then the same applies also to God, since is equally complex if not more so. Therefore, he's a shitty rationale for the existence of all complexity in the first place. Of course, you will just use your special pleading to put God outside the scope of the problem, as theists always do.

You are also repeating the bogus nonsense about evolution not providing any evidence. Complete lie.

Sure you can.  You can identify the number of permutations possible and identify probability distribution, without specific information.




 If it is true that necessity, chance and design are the only three modes and since we have a good definition of necessity and a good definition of chance, everything else must be design.

The actual numbers depend on the event and permutations and probability distribution.  Fo numbers we need a specific situation.  Dice games work nicely to illustrate the process.  But, you may pick a specific case.

Scripto, using an eliminative filter it is not futile unless you are clairvoyant that something happened a particular way.


And yet improbable events do happen all the time.  This is why probability analysis must address replication resources to ensure all the opportunities were considered and specification resources to ensure that the event was unique and distinguishable from all the other unlikely but incoherent results chance events generate.  It is interesting that as this work continues to find evolutionary mechanisms in the last 20 years no new mechanisms have been identified.



Nonsense.  Probability analysis does not require complete background information.


History does provide us observed fossil transitions.  There is nothing to observe. It is nothing but presupposition.  There is no way to empirically confirm it. 


Yes, it can and does demonstrate that frame shift that breaks usefull function and infrequently exploits new selection pressure with the modified protein is not beyond chance occurrence and existing evolutionary processes.


In statistics and probability there is nothing artificial about objectively identifying and accounting for probabilistic resources and comparing those to the probability distribution of the events in question.  This is not an arbitrary process.

I did pick specific cases, as seen from quotes. Sickle call anemia and flagellum.

Indeed.  I avoid making this argument. But that doesn't stop you from falsly reframing the premise.


The trait is used to explain the cause not the trait itself.  The evidence is that millions and millions of items and events with the same traits are known to be designed while no items with the traits has ever been adequately and specifically explained by materialistic mechanisms and no materialistic mechanism has ever been observed generating the subject traits.


Non-sequiter I don't make this claim.


A designer is outside the scope of the problem.  We are looking at causation.  We are not looking for the cause of the cause.   This is not a special pleading because I do not ask those who argue for chance and necessity to explain how material causes chance and necestity to work and how it causes material to exist in the first place.


Go ahaed and start a new thread and prove that I lie about what evolution can and can't accomplish and what has and has not been observed with respect to evolutionary processes.  If it is as you say, it should be easy since in science lies are easy to expose.  Be careful though, empirical science requires observation, tests and repeatable results.  Historical artifacts provide clues about what might have been but they are not evidence about how those fossils came to be the way they are.

I'm not sure what you are asking for then.  You seem to be in agreement (without doing the math) that advent of the sickle trait is not very complex does not contain Specified Complexity and therefore chance events can account for it and that flagellum does exibit Specified Complexity.  Do you suggest that advent of the sickle gene by evolutionary process is an event that should be considered specified and complex?  Do you suggest that you know of specific evolutionary processes that could have built out flagellum from a bacteria gene pool that totally lacked flagellum?

Or are you simply asking that we go through the math for these examples?

That's it. I want values of probability for these two examples. Numbers that will determine if these two events occured by chance or not.


No I am not. I want to see the math for these two.

OK, the advent of Sickle Cell trait in hemoglobin first. The Sickle Cell gene is one base pair different from the far more common typical human Hb gene at position 6 of the 146 amino acids.  At position six in sickle trait the gene codes for valine instead of glutamic acid. One chance hypothesis for this event is that Hb at some time in the past underwent a single chance substitution of a base pair in one or more persons.  We know from empirical analysis that this kind of substitution occurs about once for every 100,000,000 base pairs.  This gene contains 146*3 or 438 base pairs (not relevant for this analysis) and at that slot there were three errors possible. By this hypothesis, the combined odds of this mechanism is 1/300,000,000.  Next let's consider the number of replication resources available to derive this event.  For this we take the estimate that there have been about 120,000,000,000 human births (~4,000,000,000 5,000 years ago) and perhaps three times as many miscarriages. Thus without considering the specification resources we already have a 99.998% probability (better since I have approximated) that this event could have occurred by chance over 5,000 years ago. By this we conclude that this event was not complex and did occur by chance.

What I was attempting to say was that the probability calculations for the filter are based on the assumption that the target function was the only function that could arise. Not only are there redundant proteins with the same potential function (Cytochrome C) there have been independently evolved observed functions (Lactase tolerance).

As I understand it, small changes in regulatory genes can result in rather large changes in phenotype. I believe that is a relatively new field.

I said there is insufficient background knowlege.

Not true. The fossil record fits rather nicely with the molecular and genetic evidence. It is also amenable to prediction, witness the findings of cetacean intermediates as predicted in the Indus river valley and that croc-fishy thing titika rosea, not to mention the hominid record. Fulfilled predictions put it in the realm of empiricism. 

Duplication and frame shift. It retained its original function and added a new one. The idea that all mutation results in broken function is wrong. And since there are numerous other evolved bacteria feeding on synthetic products (PCB,trinitrotoluene) these mutations aren't all that infrequent.

Demski's filter does not account for the observed data. Setting the parameters is arbritrary and the definitions are inadequate. Surely, if it had any utility, someone, somewhere would be using it. It's been more than 10 years.

The filter considers a rejection region of similar events and then identifies all other regions in sample space with equal or less probability distribution that are of equal specificity.  In this way all the events in all these regions are considered.  Your claim is false.



We observe that there are different and redundant proteins but we have no information about how they came to exist.  You accept presupposition that evolutionary processes generated these proteins but it is not empirically validated and furthermore we have never identified any processes that can accomplish what you assume has happened.



This was discovered in the 1970's, and the processes are the same mutation process that have been known for some time.



You are splitting hairs.  Probability theory includes theorems that cover unknowns.  If one possible chance hypothesis is identified, all other must honor the number of permutations and degrees of freedom so far identified.



Actually, it fits very poorly.  Also it would tell us only that some organisms are similar to others.  It cannot and does not informs us at all how that came to be.  We are discussing here the how.



Thus far experiment has failed to confirm duplication and frame shift so even it remains speculation.  Perhaps you have new information on this?



Point mutations (including frame shift) and duplications are quit frequent, you are correct.  At the same time, they don't generate much new information and seem limited to monoprotein function such as enzymes too.  Given the starting point and mode, they are similar to the sickle cell trait complexity example above.
 


On the contrary, the fact that it accounts for the data very well is the power of it as a filter.  You turn a blind eye to those who have used the filter.

And now an example that is too complex to happen by chance.

The problem is that creationists tend to misrepresent evolution as some giant implausible saltation event. It couldn't be farther from the truth. If it were the case, then yes, it would be unlikely and the complexity would argue against "evolution" of that form, but reality defeats that assumption.

Evolution is unguided, but non-random algorithm. It's a slow build-up from A to F, not a direct jump immediately. Given the times involved and the slow process of build-up wherein you can lock-in changes, it's very unlikely at all and complexity doesn't negate it.

But positing "design" as the necessity because it's complex is a non-answer anyway, since God is far more or at least equally as complex. It doesn't answer the problem to use the very trait your looking to answer to answer itself, especially when the "answer" is effectively an unknown variable.

That may be true if you have a defined alternative event to match the probabilities against. In fact, the very definition of design using the filter depends on eliminating all other considerations. This requires more background knowledge than we have. Even if the filter worked, without a firm definition of design, there is no reason to conclude that specified complexity automatically defaults to design when the alternatives can include unknown or as yet to be discovered processes. Intelligent design, whatever that might be, would only be one of many possibilities.


I can't see what I can't find. Some independent corroboration, please.

Probability studies are ideal for situations where background information is sparse.  With background knowledge, we don't need probability to predict cause.

 
this is why it is important to identify the modes of causation.

1. Necessity -- driven by physical law
2. Unconstrained contingency - Pure Chance
3. Unintentional constrained contingency - for example mutation and natural selection
4. Intentional constrained contingency - design

Perhaps you can improve on this model by adding a fifth mode.  The SC filter eliminates option one through three. If there are no others then it has to be design.


Perhaps you should open your eyes first.  SC is an objective but otherwise unaltered application of the Fisherian probability model.  You will find that Fisher's probability mode is widely accepted.

You miss my point. Design events will change as the background information changes. It has no definitive existence on its own. Certainly not one outside of natural processes.

Not without some empirical evidence for design it isn't.


That's not my question. What is the utility of the eliminative filter, what biologists are using it and at what level can they apply it?

micro biology is just macro biology zoomed in.

In other words, humans behaviour is that of a virus...zoom in on Phoenix, AZ from space (google earth) and you'll see the random paths and arteries of our existence in the form of roads and bldgs.

The fact that we still sit in traffic hours a day still baffles me.

bump

Right, sorry for neglecting this.  You asked to look at the origin of flagellum as the second example.  Flagellum is a cellular propulsion system that makes use of a rotary turbine pump connected to a drive shaft complete with thrust washers, bearings, and seals.  The shaft is connected to a universal joint and a whiplike propeller.

Here again is a good site with detailed models, movies and animations. 

www.npn.jst.go.jp

Despite numerous claims to the contrary it is not known how flagellum could have evolved by any of the known, observed and cataloged evolutionary processes.  If evolution does ever explain the origin of flagellum it will only be through addition of processes that are not currently understood and validated.

The components making up the flagellum contain multiple parts and many of these parts in turn contain multiple protein subcomponents fitted together by binding sites.  Here is a diagram of the high level parts.

(http://www.researchintelligentdesign.org/images/b/b4/3Dbacflag.jpg)

The entire structure is built bottom up by a series of precisely timed developmental control, inventory identification, differentiation and management, component transport, positioning jigs, and measuring devices so that in total more than 120 different proteins are involved.

Each of these proteins have very specific structures, shapes, and chemical affinities exposed in order to identify the components it interacts with and reject those that it should not.

Given the number and size of each protein and the number of amino acids involved and available to construct each together with detailed studies of the number of 3-D shapes and chemical binding sites available one can estimate the total number of permutations and therefore the pool of possible components that could be generated.  Instead let's simplify this and presume that all these components are present at one time or another and available in the group of bacteria in time past to enable these components to come together.  Despite the fact that well over 75% of these proteins have no known source, let's stipulate they do.

To further simplify the analysis let's focus on the motor, the stator, the shaft and the propeller.  These four components are required of all rotary motion devices in order to serve the primary function of locomotion.  If any primary component of any of these subsystems is missing, malformed or misplaced the primary function cannot work.  The number of proteins involved in assembly and use of these components is about 70 and the number directly involve in the components is about 34.  These 34 components must be constructed, molded into the correct shape, delivered and placed in the right location at the right time in order to for the the rotary drive system to function.   Furthermore all of the 500 or so other protein components present in the bacteria must be excluded from the system to prevent malfunction.

In order to understand how this might have been constructed by materialistic mechanisms (or not) we need to identify all the pathways to this device that we can.   One that comes immediately to mind is that it was derived by an all in one combination of these parts in a single evolutionary step.  A second option is that motor and stator was a preexisting subcomponent (of 15 proteins) that simultaneously modified to allow attachment of the shaft assembly.  In this way the motor assembly would previously be assembled as a combinatorial object in one step. 

I will pause.  Perhaps you can identify other pathways?

A post from RF is not showing.  I was gonna quote it, but the graphics most likely wouldn't quote...

I see it in replies. Unfortunately, I am waiting for something else.

Below was my response but I paused prior to going through the math.  You seem to be under-whelmed by it and also have no other pathways to offer, so I will add to it below.




Right, sorry for neglecting this.  You asked to look at the origin of flagellum as the second example.  Flagellum is a cellular propulsion system that makes use of a rotary turbine pump connected to a drive shaft complete with thrust washers, bearings, and seals.  The shaft is connected to a universal joint and a whiplike propeller.

Here again is a good site with detailed models, movies and animations. 

www.npn.jst.go.jp

Despite numerous claims to the contrary it is not known how flagellum could have evolved by any of the known, observed and cataloged evolutionary processes.  If evolution does ever explain the origin of flagellum it will only be through addition of processes that are not currently understood and validated.

The components making up the flagellum contain multiple parts and many of these parts in turn contain multiple protein subcomponents fitted together by binding sites.  Here is a diagram of the high level parts.

(http://www.researchintelligentdesign.org/images/b/b4/3Dbacflag.jpg)

The entire structure is built bottom up by a series of precisely timed developmental control, inventory identification, differentiation and management, component transport, positioning jigs, and measuring devices so that in total more than 120 different proteins are involved.

Each of these proteins have very specific structures, shapes, and chemical affinities exposed in order to identify the components it interacts with and reject those that it should not.

Given the number and size of each protein and the number of amino acids involved and available to construct each together with detailed studies of the number of 3-D shapes and chemical binding sites available one can estimate the total number of permutations and therefore the pool of possible components that could be generated.  Instead let's simplify this and presume that all these components are present at one time or another and available in the group of bacteria in time past to enable these components to come together.  Despite the fact that well over 75% of these proteins have no known source, let's stipulate they do.

To further simplify the analysis let's focus on the motor, the stator, the shaft and the propeller.  These four components are required of all rotary motion devices in order to serve the primary function of locomotion.  If any primary component of any of these subsystems is missing, malformed or misplaced the primary function cannot work.  The number of proteins involved in assembly and use of these components is about 70 and the number directly involve in the components is about 34.  These 34 components must be constructed, molded into the correct shape, delivered and placed in the right location at the right time in order to for the the rotary drive system to function.   Furthermore all of the 500 or so other protein components present in the bacteria must be excluded from the system to prevent malfunction.

In order to understand how this might have been constructed by materialistic mechanisms (or not) we need to identify all the pathways to this device that we can.   One that comes immediately to mind is that it was derived by an all in one combination of these parts in a single evolutionary step.  A second option is that motor and stator was a preexisting subcomponent (of 15 proteins) that simultaneously modified to allow attachment of the shaft assembly.  In this way the motor assembly would previously be assembled as a combinatorial object in one step. 

I will pause.  Perhaps you can identify other pathways?

Continuing:

As combinatorial objects, one must choose, position and assemble them all at once.  Each of these is a distinct intermediate step in the process and any of the other steps could interfere without some constraint that would preserve order and prevent alternative configurations or placements or assemblies.

Beginning with selection, the 34 components must be selected from the 500 components contained by a typical bacteria (actually 4300 but again I will simplify).  Let's also assume that for each of these 34 proteins there are 2 suitable alternatives  The odds that these 34 correct components are selected and all others rejected are then (2/500)^34 * 34! since the order is not important  or 8.73*10^-44.  However examination of the flagellum structure has over 20,000 proteins in the filament, and 26 in each of the three rings, 6 in the proximal rod, 120 in the universal joint etc.  Let's use 10 subunits for each of the 34 pieces.  Redoing the math gives us (2/500)^340 * 340! or 2.55*10^-101.  Now we must consider the configuration problem (position and assemble) and then combine the two for a final probability score by this hypothesis.  The math required for this step is quite complex since we are determining the fraction that would result in actual function divided by all the different ways these components could have been assembled after taking consideration of the fact that some function must be preserved.  Dembski performed this math by using perturbation tolerance using Sterling's formula.  After simplifying it for the very small numbers we are dealing he gets p(pertub)~~K^(q-r)N q is the tolerance factor and r is the identity factor.  With N the total number of elements in the assembly (>20,000), one gets a number about 10^-3000 for flagellum.  With numbers this small we need not consider the number of probabilistic resources or specification resources since even the largest combination of the two of them can't exceed 10^150 the universal probability bound.  From these numbers (10^-101 and 10^-3000) as compared to the limit of resources and specification of 10^150, chance is ruled out.  Remember, I did not even consider origination of these elements.

Given in particular the large number of alternative configurations for the 20,000 plus sub-components in flagellum it should be easier to understand why biological self-assembly systems make use of the control, management, transport and configuration proteins to ensure that these components are built according to the intended function.

Now at this point one could object that my hypothesis is limiting, and so again I will ask anyone with an interest to propose an alternative pathway.

Below was my response but I paused prior to going through the math.  You seem to be under-whelmed by it and also have no other pathways to offer, so I will add to it below.




Right, sorry for neglecting this.  You asked to look at the origin of flagellum as the second example.  Flagellum is a cellular propulsion system that makes use of a rotary turbine pump connected to a drive shaft complete with thrust washers, bearings, and seals.  The shaft is connected to a universal joint and a whiplike propeller.

Here again is a good site with detailed models, movies and animations. 

www.npn.jst.go.jp

Despite numerous claims to the contrary it is not known how flagellum could have evolved by any of the known, observed and cataloged evolutionary processes.  If evolution does ever explain the origin of flagellum it will only be through addition of processes that are not currently understood and validated.

The components making up the flagellum contain multiple parts and many of these parts in turn contain multiple protein subcomponents fitted together by binding sites.  Here is a diagram of the high level parts.

(http://www.researchintelligentdesign.org/images/b/b4/3Dbacflag.jpg)

The entire structure is built bottom up by a series of precisely timed developmental control, inventory identification, differentiation and management, component transport, positioning jigs, and measuring devices so that in total more than 120 different proteins are involved.

Each of these proteins have very specific structures, shapes, and chemical affinities exposed in order to identify the components it interacts with and reject those that it should not.

Given the number and size of each protein and the number of amino acids involved and available to construct each together with detailed studies of the number of 3-D shapes and chemical binding sites available one can estimate the total number of permutations and therefore the pool of possible components that could be generated.  Instead let's simplify this and presume that all these components are present at one time or another and available in the group of bacteria in time past to enable these components to come together.  Despite the fact that well over 75% of these proteins have no known source, let's stipulate they do.

To further simplify the analysis let's focus on the motor, the stator, the shaft and the propeller.  These four components are required of all rotary motion devices in order to serve the primary function of locomotion.  If any primary component of any of these subsystems is missing, malformed or misplaced the primary function cannot work.  The number of proteins involved in assembly and use of these components is about 70 and the number directly involve in the components is about 34.  These 34 components must be constructed, molded into the correct shape, delivered and placed in the right location at the right time in order to for the the rotary drive system to function.   Furthermore all of the 500 or so other protein components present in the bacteria must be excluded from the system to prevent malfunction.

In order to understand how this might have been constructed by materialistic mechanisms (or not) we need to identify all the pathways to this device that we can.   One that comes immediately to mind is that it was derived by an all in one combination of these parts in a single evolutionary step.  A second option is that motor and stator was a preexisting subcomponent (of 15 proteins) that simultaneously modified to allow attachment of the shaft assembly.  In this way the motor assembly would previously be assembled as a combinatorial object in one step. 

I will pause.  Perhaps you can identify other pathways?

Continuing:

As combinatorial objects, one must choose, position and assemble them all at once.  Each of these is a distinct intermediate step in the process and any of the other steps could interfere without some constraint that would preserve order and prevent alternative configurations or placements or assemblies.

Beginning with selection, the 34 components must be selected from the 500 components contained by a typical bacteria (actually 4300 but again I will simplify).  Let's also assume that for each of these 34 proteins there are 2 suitable alternatives  The odds that these 34 correct components are selected and all others rejected are then (2/500)^34 * 34! since the order is not important  or 8.73*10^-44.  However examination of the flagellum structure has over 20,000 proteins in the filament, and 26 in each of the three rings, 6 in the proximal rod, 120 in the universal joint etc.  Let's use 10 subunits for each of the 34 pieces.  Redoing the math gives us (2/500)^340 * 340! or 2.55*10^-101.  Now we must consider the configuration problem (position and assemble) and then combine the two for a final probability score by this hypothesis.  The math required for this step is quite complex since we are determining the fraction that would result in actual function divided by all the different ways these components could have been assembled after taking consideration of the fact that some function must be preserved.  Dembski performed this math by using perturbation tolerance using Sterling's formula.  After simplifying it for the very small numbers we are dealing he gets p(pertub)~~K^(q-r)N q is the tolerance factor and r is the identity factor.  With N the total number of elements in the assembly (>20,000), one gets a number about 10^-3000 for flagellum.  With numbers this small we need not consider the number of probabilistic resources or specification resources since even the largest combination of the two of them can't exceed 10^150 the universal probability bound.  From these numbers (10^-101 and 10^-3000) as compared to the limit of resources and specification of 10^150, chance is ruled out.  Remember, I did not even consider origination of these elements.

Given in particular the large number of alternative configurations for the 20,000 plus sub-components in flagellum it should be easier to understand why biological self-assembly systems make use of the control, management, transport and configuration proteins to ensure that these components are built according to the intended function.

Now at this point one could object that my hypothesis is limiting, and so again I will ask anyone with an interest to propose an alternative pathway.

I'll avoid other comments and will try to focus just on your calculations. So,

flagellum:


Sickle cell anemia


2 points:

1. Your text for flagellum: Let's use 10 subunits for each of the 34 pieces. For sickle cell example you counted just 1 subunit, eventough mutated form has 2 subunits in every hemoglobin and a whole lot of hemoglobins bind together.


2. I have some problems with calculations. For sickle cell anemia, you calculated probability 1/300,000,000, estimated 4,000,000,000 human births over the last 5,000 years, 99.998% probability that this event could have occurred by chance over 5,000 years ago.

I can't get the 99.998%, I don't know how you got it. But anyway, for flagellum I can't see the matching numbers. Now, E. coli which flagellum you use lives in human guts. Let's assume that all human ever born had it (120,000,000,000 humans ever born - your number). Lets assume that every human gut has 10^8 E.coli, average historic live span of human is, I don't know, 30 years. E. coli replicates every 30 minutes. So, the number of all E. coli in all humans that ever lived is 6 311 520 000 000 000 000 000 000 (then again, flagellum is evolutionary much older, how about we recalculate it for the last billion years instead?). If you adjust probability from point 1, what is the recalculated probability for sickle cell anemia and flagellum?

Right.  In building out flagellum I used an average of 10 subunits per protein (the filament is 20,000+ while a few parts have five. Clearly 10 is conservative) a correct approach would be to take each one separate but that would have complicated the math and resulted in a number that makes the odds much worse.  In the case of sickle hemoglobin the hypothesis is completely distinct.  It involves the single substitution of one base pair for another.  The subunits don't figure into the formula.



Right, I did not calculate the number of opportunities for flagellum to arise.  I did not because the complexity results were so low that we could have taken every atom in the universe to be the count for bacteria and the vibration frequency of atoms as the replication rate and the entire 13 billion years that this universe is said to be old and use that for the number of opportunities (10^150) and still not even make a dent in the probability numbers from the complexity step.


I agree there has been a large number of bacteria produced on this earth  but compared to 10^-3000 it is insignificant.  I don't think we need to make an adjustment for sickle trait.

I agree that with the flagellum numbers, mine were guesses and they can be improved.  I don't think it will change the conclusion though. 

I am aware that HbS example is simpler than flagellum. But HbS is a part of multimolecular structure and flagellum is a multimolecular structure. You can't calculate with number subunits in just one of these 2 examples. You can include subunits in both or work just with a number of different elements.

I understand your point and agree that hemoglobin is a five part structure. 

The reason that the two calculations are different is because of the chance hypotheses I chose.  The HbS hypothesis deals with origination of the mutation which involves substitution of base pairs.  Origination of HbS began with the stipulation that Hb and the entire hemoglobin existed previously so we don't consider localization and assembly.

In the Flagellum example I stipulated origination and therefore skipped past this part to focus instead on localization, placement and configuration of the complete proteins I stipulated.  This portion of the problem requires consideration of the correct number of subunits.  If I were to include the origination problem in flagellum then the odds get far far worse for chance occurrence. 

While HbS from Hb is entirely consistent with evolutionary processes and does not calculate out as a Specified Complex event, in order for material processes to explain flagellum, we need an additional suite of evolutionary processes.  Flagellum does calculate out as being SC.

and would it be too much to ask to have him. show his work.  no doubt we will see him create a number that proves his point.

In short, you use different assumptions and methodology for both. This is so obviously wrong.

Is there a way to use the exactly the same approach to calculate probability for both cases? For example, evolutionary approach is that components in both cases did exist previously and are changing with the same mechanism (mutation, selection).

We are cleaver people and can find multiple approaches to nearly any problem, but it does not make one correct and another not.  In this case though they are different because the circumstances are different not because the approach is different.  You'll have to show me how it is obviously wrong because it is far from obvious to me.


I used the same approach in both cases but made different stipulations.  In sickle hemoglobin, the starting point was that Hb and hemoglobin exists.  The question under consideration was the generation of HbS in humans and whether or not this change had the character of exhibiting specified complexity. 

Are you suggesting that we allow that humans exist but that Hb and hemoglobin did not?  If we agree that the starting point for HbS is Hb then I believe my analysis was correct.  I need not consider localization or configuration because those processes were fixed prior to the HbS event.  In otherwords at the point that HbS occurred the probability of localization, placement, configuration etc. are all 1.0 and therefore don't change the analysis.

In the case of flagellum we we considering the development of flagellum in bacteria that lacked flagellum.  We were considering the very first flagellum.  In this example I skipped past the questions of mutation of genes required to generate the individual proteins despite that it would have made the numbers even less favorable for material mechanisms.  In HbS I addressed the origination problem and calculated the probabilities.  In flagellum, I gave the origination step a probability of 1.0 though in reality the odds would likely be much closer to zero.

Now in your statement you say that evolution presumes that components previously existed and I have granted that in the case of flagellum.  I did not require that these components mutate from some previous configuration.  Had I required this more realistic case probability would drop substantially.

For flagellum localization, placement and configuration were not given and that is where I focused my attention.  So the approach was the same but the circumstances and assumptions were different thus different parts had different probability results. 

reposting this:


We are cleaver people and can find multiple approaches to nearly any problem, but it does not make one correct and another not.  In this case though they are different because the circumstances are different not because the approach is different.  You'll have to show me how it is obviously wrong because it is far from obvious to me.


I used the same approach in both cases but made different stipulations.  In sickle hemoglobin, the starting point was that Hb and hemoglobin exists.  The question under consideration was the generation of HbS in humans and whether or not this change had the character of exhibiting specified complexity. 

Are you suggesting that we allow that humans exist but that Hb and hemoglobin did not?  If we agree that the starting point for HbS is Hb then I believe my analysis was correct.  I need not consider localization or configuration because those processes were fixed prior to the HbS event.  In otherwords at the point that HbS occurred the probability of localization, placement, configuration etc. are all 1.0 and therefore don't change the analysis.

In the case of flagellum we we considering the development of flagellum in bacteria that lacked flagellum.  We were considering the very first flagellum.  In this example I skipped past the questions of mutation of genes required to generate the individual proteins despite that it would have made the numbers even less favorable for material mechanisms.  In HbS I addressed the origination problem and calculated the probabilities.  In flagellum, I gave the origination step a probability of 1.0 though in reality the odds would likely be much closer to zero.

Now in your statement you say that evolution presumes that components previously existed and I have granted that in the case of flagellum.  I did not require that these components mutate from some previous configuration.  Had I required this more realistic case probability would drop substantially.

For flagellum localization, placement and configuration were not given and that is where I focused my attention.  So the approach was the same but the circumstances and assumptions were different thus different parts had different probability results. 

In HbS the approach was: we had something to start with, it has changed.

In flagellum the approach was: we had nothing, what are the odds for flagellum in such and such a number of building blocks to appear.

In first example, you use evolutionary approach, in the second you don't. In second, you assume what you are trying to prove - that there was nothing evolution could work on to finish with flagellum.

Aren't you finding the chance of what it would take to get to a final point, as if the existing structure was aimed for by the organism?  The nature of the existing structures are required for whatever development will take place. It doesn't happen in a vacuum.

The "chance" that something happens is based on the events preceding. Not that you can expect a heads if 100 tails happened before (it's still 50/50), but if the coin has been bent from many tosses and has 3 sides, then there are different possibilities available.

Agreed.


No, with flagellum we had something to start with too.  We had bacteria with 500 proteins of which contained all the proteins required to construct flagellum.  Each of these proteins was involved in some other functional system.  Then by the only process I can think of (as a combitorial object) an evolutionary change occured to cause the flagellum to be constructed by this change.   I am willing to entertain any evolutionary pathway you can conceive of to cause the appropriate events to occur.  Please suggest an alternative.  This is where I puased previously because I suspected that this is where we will have difficulty in coming to terms with my calculation methods.  At the time you did not object but now you do, and that is ok.  So please suggest an alternative.


I disagree.  In both cases I use the same approach.  In the first case of HbS the pathway selected is simple base pair substitution.  In the second case we have no actual or verified pathway so I stipulated one, namely an all in one combinatorial mechanism.  I stipulated this because I cannot conceive of any other way this system could be constructed stepwise because the components as configured don't have function apart from each other.

Again I ask you to propose an alternate pathway and we can put the numbers to it.  Realize that as we become more precise about a pathway we also should revise the stipulations I used to simplify the math.

this is still an argument from ignorance. even if you found that evolution as we know it can't account for the pathways, you don't get to insert a god into the gap.
btw, since you seem happy at the 500 mark, where is the evidence that your designer did something?
do you actually have to have evidence to posit a theory?
show us the POSITIVE act of design within the pathway. where did it occur and how?

This is an elimiative filter.  It argues nothing from ignorance it does eliminate chance and necessity as possible explanations for events.  With those modes eliminated the only one left is design unless you can posit an additional mode.

Correct that we don't insert God.  We can infer design but we cannot infer who or what the designer was.


That is a question for another thread.  This one is on specification and complexity.

OK, include the possibility for components of the flagellum that are similar to each other that they evolved trough gene duplication.

Include the possibility of horizontal gene transfer.

ADDED: Include the possibility that HbS is evolutionary older than Hb.

ADDED EVEN LATER: Consider that components of flagellum might not originate from pathways with very different functions. For example, outer flagellum from cytoplasmic skeleton and motor from systems for transfer of proteins trough the membrane, based on conformation change (=movement).

I have already stipulated that the components all exist with probability of 1.0.  If I understand correctly gene duplication and horizontal gene transfers are mechanisms that deal with origination.  Any relaxation from 1.0 reduces the odds for material mechanisms.


Since by this mode the odds for chance and necessity are dramatically reduced, and we could do the calculations, but the simpler mechanaismpreviously suggested is the favored chance mechanism for HbS.   In a sense this suggestion looks at origination of Hb(X) from non-Hb


Yes, and this hypothesis adds boundary conditions to the localization and configuration permutations.  It will improve the odds.  However evolutionary theory requires function for each component.  The motor has function on its own as a proton pump (I doubt it would make a for a protein pump given the configuration).  The protein injector that delivers proteins for filament construction also has clear alternate function.  I can think of no function for the shaft, the universal joint or the fillament.    You suggested one for the filament but didn't follow it. Once we have subdivided the system into these alternate systems and have identified a use for each we can rework the numbers.  I don't think the results will change much relative to the 10^150 universal limit.

What do you mean by origination? Gene duplication and horizontal gene transfer are mechanism for increasing the number of genes in an organism for evolution to work on under different selection pressure as the original gene and/or gene in organism that provided material for horizontal transfer. Anyway, if you agree that there was material to start with, then this material follows the HbS story.

For sickle cell anemia, you calculated probability 1/300,000,000, estimated 4,000,000,000 human births over the last 5,000 years, 99.998% probability that this event could have occurred by chance over 5,000 years ago. So considering that for E. coli generation time is 30 minutes, mutation rate higher, number of organisms higher, how long would it take for that kind of single mutation?

How long would it take for 100 proteins of 200 aa each to change 50% of their sequence?


Where did that number came from?

I mean the process by which the required proteins came to exist.


I think I understand.


A single selectable substitution should occur in very little time. In less than a day.


Each of the steps in the pathway must be functional and selectable or the mutations will not be retained.  Given  that experimental and theoretical protein analysis indicates that fewer than 1 in 10^55 roughly 400-500 base pair protein configurations are functional (Axe, 2003) I would predict that 50% drift could never occur.  We can go through the math if you like but it appears from Axe's analysis that functional proteins may differ each other by an average of 6-20 base pairs depending on the lengths considered.


It is the upper bound of all the macro-events that could ever have occured in this universe.  It is the number of atoms in the universe times the number of atomic collisions per second times the number of seconds this universe is old.  It places a reasonable upper bound on what chance can explain.

False. One argument is genetic drift. The other is that selection is not the same in all environments and that it can change. In case of E. coli, or any microorganism, each species is characterised by a number of strains with very different sources of energy it takes from the media. I am talking about wild strains, for which even dysfunctions or big differences in whole metabolic pathways don't prevent them from surviving. There was a reason behind sequencing whole genomes of several strains of E. coli. Human gut is a pretty stable environment, so why do you thing there is such a variability of strains?


Why do you assume that all transitions forms are subject only to drift, surely some are also under positive selective pressure. About low proportion of functional proteins, I have seen too many alignments of proteins with the same function and different aa sequence to believe it. It would made my job easier if conserved regions would be that frequent.


So the only reason why this limit was selected because it is a high number? Such a "reasonable upper bound" is still just a guess.

Of benign or mostly benign alterations including for example base pair substitutions that give redundant coding etc.  These changes also don't contribute to new forms and funtion.  If I am wrong on this point, please correct me.  I would be very interested in a detailed explanation of how genetic drift accomplishes this.  Then if you would offer an example of an actual pathway to new protein function by genetic drift it would help me to understand what you are talking about and it would help us decide if this pathway can be considered actual.


I don't think my statement about function and selction disregards this reality.  It was a high level statement.  Your elaboration adds clairity.  Once mutation disables a metabolic pathway, the dysfunctional proteins are subject to gentic drift by random mutations.  Any single mutation of a nonfunctional gene won't generally be retained it is subject to further drift since it contributes no function.


Given that evolutionary processes are proficient at breaking function and in bacteria, gene transfer can move funtional systems, it makes sense that multiple metabolic pathways exist and that several are disabled.

In any case the primary focus of this portion of the discussion deals with propsing an evolutionary pathway other than by combitorial objects.  We can quibble about what is and what isn't theoretically in play but we need a stepwise complete hypothetical pathway in order to test for Specified Complexity.  Can we derive such a pathway by genetic drift.



I don't think I make this assumption.  Again the focus here is to identify a plausible pathway.



Given that a 150 aa sequence has 20^150 or 1.4*10^195 permutations, by Axe's analysis we still have well over 10^100 functional combinations so it should not be a surprise that you observe alternate alignments that are functional.  The issue is whether or not there are functional pathways from one to the other observed proteins.  The data indicate there are not