Over in the ID Quiz thread at Uncommon Descent, Mark Frank asks;
“Here’s quiz on ID for you ID proponents:
On page 21 of “Specification: The Pattern That Signifies Intelligence” William Dembksi defines the context dependent specified complexity ofgiven
as
Consider the context of the bacterial flagellum.
1. What is
2. What is the function?
3. How is
4. What is
5. How isestimated?
6.is meant to be a probability. Under what conditions might the answer exceed 1?”
given 
as ![-log_2[M\cdot N\cdot \varphi_S(T)\cdot P(T|H)]](http://theskepticalzone.com/wp/wp-content/ql-cache/quicklatex.com-e8afc353618be80af64fbcec7564a345_l3.png "Rendered by QuickLaTeX.com")
?
estimated?
is meant to be a probability. Under what conditions might the answer exceed 1?”This seems a fair question, asking IDists to use their own proposed methodology to detect design, empirically.
But instead of heartily embracing this opportunity to show the power of ID, he is set upon.
Joe moans that
“Really Mark? Then tell us how to determine the probability wrt unguided evolution. If you know the answers to your questions you should be able to do that.”
Sorry Joe, it’s your methodology, do your own work, if you can.
KirosFocus has a length moan from large numbers and incredulity, finishing with;
“We cannot stop such from those tactics but we can expose them and red ring fence off those who insist on such tactics.”
STOP. RIGHT. THERE.
If you can’t employ your own design detection methodology it is no ones fault but your own.
Either the methodology is unusable,
OR non of you are smart enough to use it,
OR you’re all secretly supports of Darwinian evolution and are making ID look as ridiculous as possible.
All Mark Frank has asked is for you to employ your own methodology. You can’t. ID is intellectually bankrupt, and only addressing the required calculations will change that.
[Edit: symbols latexed by Lizzie]
Dembski defines P(T|H) as the probability of getting the Target (say a bacterial flagellum) given H, “the relevant chance hypothesis that takes into account Darwinian and other material mechanisms.”
How do you think Dembksi has taken “into account Darwinian and other material mechanisms” in the above example?
There are no ‘molecules’ here. Only permutations of proteins
of flagellum
I think (I don’t have access to his book)that would be the 4,289 proteins , though he doesn’t mention it as ” Darwinian and other material mechanisms”:
coldcoffee,
*blink*
Read Mike’s comment:
where have I referred to molecules?
If you doubt whether the number of proteins quoted by Demski are correct, search for Eubacteria species at http://www.genome.ad.jp
E.Coli has Genome Size( bp :4,639,221), Proteins : 4,289
If Dembski’s calculation does not take account of evolutionary mechanisms (which it doesn’t), what is it actually calculating the chance of?
That’s right, a single-event, grand assembly of the flagellum by randomly picking a huge collection of proteins from the E coli genome.
Is that how evolution works? No.
Of what possible relevance is Dembski’s calculation to evolution then? That’s right, none whatsoever. We can infer absolutely nothing using Dembski’s calculation.
Still completely irrelevant. Why is Dembski calculating the odds of a single-event, grand assembly of the flagellum by randomly picking a collection of proteins from the E coli genome? That’s not how evolution works. It has nothing to do with evolution. There’s no natural selection, no functional intermediate steps of any kind.
Could you calculate the probability of flagellum assuming Natural selection?
No. Neither can Dembski. What’s the use of such calculations then?
He’s trying to rule out whether the flagellum evolved so that he can infer design. But he doesn’t actually calculate the odds that it evolved. How can he then rule out that it evolved and infer design? He obviously can’t.
Unless you can show natural selection affects the flagellum probability and is above the universal probability bound of
, how can you reject his calculations ?
Dembski is the one who needs to show that natural selection(among other things) doesn’t significantly affect the outcome, he’s trying to rule it out. It’s that simple, you can’t rule something out you don’t take account of. He doesn’t even have a way of roughly estimating the odds of the flagellum evolving with natural selection. He’s totally clueless about how natural selection affects the outcome. He doesn’t know to what extend it affects it, whether it becomes only slightly more probable or significantly more probable.
If Dembski wants to rule out an evolutionary process, he needs to come up with some kind of way to represent it. But his calculation has nothing to do with an evolutionary process. It only calculates that single-event grand assembly by randomly picking E coli proteins. That’s obviously not how evolution works. There are no intermediate steps with different functions, no trial and error, no natural selection.
If Dembski doesn’t know how how natural selection affects the outcome, how can he rule it out? How can he be sure he’s not scoring a false positive?
coldcoffee,
Nobody claims bacterial flagella evolved in a single tornado-in-a-junkyard event so the calculation is irrelevant and meaningless.
How will you know that natural selection will change the probability, and to what extent will it change the probability of flagellum evolution if you don’t calculate the effect of Natural selection? Since you are all evolutionists, wouldn’t it be better to show how Dembski is wrong in his calculations and prove him wrong emphatically ?
Dembski is the one claiming to have a methodology for reliable distinguishing between design and evolution. But he doesn’t have a method of estimating the probability of “evolution”, so he has no clue whether he’s scoring false positives. Demski needs to remedy this, he’s the one claiming he can reliably distinguish design from evolution.
Tell me coldcoffee, how do we know Demski isn’t scoring false positives? How can he honestly say he’s ruling something out he has no goddamn clue how affects his outcome?
Remember, Dembski is the one advancing this methodology for “detecting design”. He’s the one who claims to have a method that can do it, can tell us whether something evolved or was designed. So how can he be sure he’s actually doing that, and not just scoring false positives?
coldcoffee, this calculation rules out a specific chance hypothesis: throw together a random protein sequence and see if it is functional. Do I even need to tell you that no one suggests that this is how functional proteins come about? You have to add natural selection to the process.
There are simple models that show that. Take a smooth landscape and search for its maxima using (1) a random walk and (2) a random walk with feedback. In the latter case, a random step is accepted if it leads up the hill; otherwise the walker takes a step back.
The purely random search takes an amount of time that grows exponentially with the size of the system. The search with feedback takes an amount of time that scales as a power of the system. Big difference.
Monte Carlo methods of simulating physical systems in thermal equilibrium are based on the advantage provided by feedback.
As Dembski acknowledges. Then doesn’t do.
He doesn’t follow his own formula.
Or coldcoffee could just look at any one of the innumerable teaching simulations used in biology courses. In all of them selection changes the outcome a lot.
Or you could just consider what happens when you have a bacterial infection and don’t take the antibiotic as regularly as you should. Is there an increased probability that an antibiotic-resistant variant of your bacterium will take over and cause a relapse in your condition?
Or you could ask animal breeders why they bother to select from the best animals in the herd. If it did not change the probabilities, it would not be worth doing.
There are also equations for the outcome of natural selection in simple models. It’s called population genetics.
coldcoffee,
Proteins are molecules.
olegt,
And recombination. End-joined fragments, modular repeats, etc, mine the space of functional modules already accessed by pre-existing protein segments. It’s not the entire space that is being ‘searched’, but largely that already occupied by functional sequence. The same motifs appear over and over, and they didn’t need to evolve from scratch.
Anticipating cc’s next response, the earliest protein was not necessarily a complex or lengthy molecule.
coldcoffee,
Cold Coffee, evolution isn’t the odds of a certain hand, its the odds of all winning hands. The odds of any specific configuration of 2 shuffled decks of cards is (52 x 2)!
That’s above the UPB.
What do you think that means:
Cards can’t be shuffled?
They must have been placed that way?
You might want to rethink probability?
Yes. Based on the evidence we have the probability is 1.0
More to the point, it’s the probability that in any set of deals, there will be winning hands.
What ID does is retroactively calcuate the odds against the evening’s winner.
It’s the lottery fallacy.
And yes, there have been instances where lottery employees weighted the balls, increased the winnings of co-conspirators, and got caught. Bayesian analysis works. Dembsky knows this and ignores it.
Technically, tthe probability computed for P(T|H) in a card game example is not the probability of all winning hands, it’s the probability of all hands whose value is as great as the one we got, or greater.
The difficulty with using P(T|H) is that we have to take natural selection as well as the deal into account. So if the player throws away 3 cards and takes 3 new ones, in hopes of improving their hand, you cannot any longer just use the probabilities for a single deal of 5 cards.
In the natural selection case, there are millions of years of redealing and replicating cards. And unless you can show that the winning hands are extremely improbable after all that, you can’t use the probability calculation to make a Cheating Inference.
Calculating the number of strings of length L with K choices per position isn’t even K^L. One also has to divide K^L by all the permutations of each and every letter that appears more than once in the string. Otherwise one is asserting implicitly that a given molecule becomes a different molecule when it is in a different position in the string.
But even worse, there is no temperature dependence in the probability calculations of ID/creationists. That screams loudly that they don’t believe in chemistry and physics; or that they simply don’t care about the strong, temperature dependent interactions among molecules.
To ID/creationists, atoms and molecules are all simply inert objects in an ideal gas of such objects once they have decided that a complex assembly of molecules is designed. Crystals are not designed; therefore they assemble “naturally.” But the complex molecules of life have to spring suddenly and fully complete out of an ideal gas of inert molecules.
Somewhere along the chains of complexities in the assemblies of atoms and molecules the laws of physics and chemistry stop and “semiotic theory” or some kind of “information” derived from “intelligence” has to start pushing atoms and molecules around into a specified configuration.
ID/creationists all “know” this because the logarithm to base 2 of K^L is greater than or equal to 500 for complex molecules.
Furthermore, they assert that it is our responsibility to calculate probabilities in the way that they do.
In real theoretical chemistry there are huge computer programs running on supercomputers that calculate and predict molecular assemblies. None of these programs look anything like K^L calculations.
A K^L calculation can be done on a simple “four-banger” calculator by a middle school student; that is the mathematical level of sophistication at which ID/creationism operates.
=> About using other methods for evolution probability :
I think we have already gone down that road. No program can simulate evolution. Eg: Avida is supposedly pro evolution, Mendel’s accountant is pro ID. When Avida and Mendel’s parameters are matched, there is net loss of genetic information despite selection, so those programs don’t really explain evolution.
No. Dembski is talking about specified complexity, so in case of a card of deck, it would equate to specific series of hands. I just found that the cards analogy has already been discussed at UD way back in 2007 :
You are looking at protein at molecular level ? If they are so complex, how can you justify stochastic process leading to formation of specific protein structure ?
No. Mark is referring to molecules. See his later comments about molecules and temperature.
coldcoffee,
You’ve just Hoyled/ Tornado in a junkyard / spontaneously generated. Do you believe all evolutionary forms come into existence de novo? Is this what evolution claims?
That is YEC claim, not ID’s claim
It’s IDs claim too unless you accept discarding and replacing cards heading towards better configurations.
Why are you still asserting stochastic processes in molecular formation? This indicates you have absolutely no knowledge of even the most basic notions in chemistry; which is no surprise because ID/creationists to a person know nothing of even high school level science.
Complex molecules evolve.
Check out the story of the winners of the 2013 Nobel Prize in Chemistry. It is about work that goes back to 1967.
We are discussing evolution which is a stochastic process ( and natural selection). ID claims evolution is not possible by pure stochastic process. Designer is needed. I have no idea why you are bringing in molecules and super computers and claiming I have no notion of chemistry. What are you trying to say – that evolution is not stochastic?
You keep forgetting that protein molecules mostly evolve from closely related ancestral proteins. So there’s a protein in an organism doing some function, it gets slightly mutated here and there, and the environment determines the fitness of the carrier. If the changes to the protein lead to higher fitness, well you know the rest. That’s how most proteins evolve, that’s how they got to be how they are, that’s how they will change in the future. The individual mutation events are stochastic, in the sense that we can’t predict where in the sequence of amino acids the protein will change, and we can’t predict with certainty what kind of mutation will happen (deletion vs insertion, transversion vs transition), although the nature of the biochemistry (for example DNA methylation) and the structure of the genetic code introduces certain biases such that some outcomes are more likely on average than others. This constitutes a random sampling of the nearby sequence space of the protein. If the environment is such that one of these mutations is beneficial wrt fitness of the carrier organism, we get natural selection in action.
Ultimately the environment determines what happens. Some mutations might be deleterious or neutral in many different environments, so they are either lost to purifying selection or drift. Or if the population is small, drift might actually fix a neutral or weakly deleterious mutation. Such a mutation can be a potentiating mutation, that will later make another mutation beneficial. But again, the environment is what determines this. We cannot unambigously state that some given mutation will always be deleterious or beneficial, it depends. It could be changes to a regulatory protein such that it activates under different conditions. Maybe it can now digest citrate under anaerobic conditions. That’s what happened in the Lenski E coli experiments.
Or it could be changes to an enzyme so it now slightly accepts another related substrate. There could be a gene duplication to the enzyme and it could mutate further. If the previous substrate is running out but the new one exists in large quantities, such a mutation might be beneficial and eventually lead to the formation of a new enzyme through subsequent mutations. Studies documenting such changes through evolution exist:
Reconstruction of Ancestral Metabolic Enzymes Reveals Molecular Mechanisms Underlying Evolutionary Innovation through Gene Duplication
“Abstract
Gene duplications are believed to facilitate evolutionary innovation. However, the mechanisms shaping the fate of duplicated genes remain heavily debated because the molecular processes and evolutionary forces involved are difficult to reconstruct. Here, we study a large family of fungal glucosidase genes that underwent several duplication events. We reconstruct all key ancestral enzymes and show that the very first preduplication enzyme was primarily active on maltose-like substrates, with trace activity for isomaltose-like sugars. Structural analysis and activity measurements on resurrected and present-day enzymes suggest that both activities cannot be fully optimized in a single enzyme. However, gene duplications repeatedly spawned daughter genes in which mutations optimized either isomaltase or maltase activity. Interestingly, similar shifts in enzyme activity were reached multiple times via different evolutionary routes. Together, our results provide a detailed picture of the molecular mechanisms that drove divergence of these duplicated enzymes and show that whereas the classic models of dosage, sub-, and neofunctionalization are helpful to conceptualize the implications of gene duplication, the three mechanisms co-occur and intertwine.“
coldcoffee,
Yes. They. Are.
Temperature is relevant to protein folding. So, for that matter, are pH, solute concentration, presence of lipids/water, and so on. Because they’re molecules. They aren’t just combinatorial ‘strings’.
coldcoffee,
What prevents a stochastic process from generating a functional protein sequence over generations of recombination, duplication, extension and selection?
Yes, of course. Protein is Macro molecule. My full reply was:
No. Mark[Mike] is referring to molecules (May be I should have said No Mike means molecule, not protein)
Look at his comments. I don’t think he is referring to protein as molecule:
I am aware of protein folding. Tertiary protein folding are needed to create new Cambrian animal forms. Axe has studied the probability of folding and this study is a bone of contention between ID and Evolutionists.
Would it be possible for you to summarise this work, and explain what you think it means in relation to evolution?
Give references that this is true and that this is an issue.
For new Cambrian animals to form new proteins are required to form new organs and cell type. Proteins capable of functioning require new folds to perform the function. Douglas Axe examined studied these:
Please refer to Darwin’s doubt Chapter 10 for full experiment’s impact. I don’t have the epub, only the book so I can’t type the details. Essentially Axe concludes that new protein folds can’t not arise by random mutations. He concluded that for a sequence of 150 amino acid long protein, only 1 in 1074 sequences will be capable of folding into a stable protein.
So now you have quoted Douglas Axe making the same claim. That just means there are now two people who need to substantiate the claim with a reference.
I just looked in Darwin’s doubt chapter 10, it says
“Axe knew that as new life-forms arose during the history of life—in events such as the Cambrian explosion—many new proteins must also have arisen. New animals typically have new organs and cell types, and new cell types often call for new proteins to service them. In some cases new proteins, while functionally new, would perform their different functions with essentially the same fold or tertiary structure as earlier proteins. But more often, proteins capable of performing new functions require new folds to perform these functions. That means that explosions of new life-forms must have involved bursts of new protein folds as well.”
No references are given for any of the claims contained here. ZERO. They’re simply blindly asserted.
That statement is simply false on its face. Many simulations of known evolutionary mechanisms have been written. A simple Google search will provide you with many links.
Zachriel dismantled the nonsensical Mendel’s Account in this discussion thread.
Further, known evolutionary mechanisms have been demonstrated to increase information in a population. For one direct demonstration, see Thomas Schneider’s ev.
So the odds have changed from one in 10^500 to one in 1074.
He’s actually wrong about this. But it’s interesting to see how the astronomical numbers change when you switch from tornado in a junkyard to testing adjacent space.
But it’s really not that difficult to evolve a new protein. Axe’s work centers around polishing an already optimized protein.
Are you implying Douglas Axe is not to be trusted and that he based his study on non-existing facts? I don’t have Douglas Axe’s references, but will see if I can get his original paper references.
Douglas Axe is not to be trusted when it comes to thinking and drawing conclusions.
His other remarkable work is demonstrating that it’s unlikely for one existing protein to evolve into another existing protein.
No one has ever suggested that anything like this has ever happened.
You mean the whole protein fold and Axe’s experiment to determine probability of protein folds is bogus?
coldcoffee,
All that fundamentally happened in the late Cambrian was exploration of multicellular animal and plant phenotypes (most organisms aren’t animals or plants). These would be mediated by changes in developmental switching, only rarely by anything one could term a ‘new fold’. Most protein motifs have homologues in many other proteins, indicating extensive modular swapping. I don’t have a clear idea of how a ‘new fold’ might distinguish itself, and I doubt you do either.
Well I’m certainly not going to take his word for it, no.
At best Axe’s work has nothing to do with how evolution works.
Exactly right. In fact, phylogenetic studies have shown that a significant majority of all enzymatic functions known in all of life (over 30.000 different enzymes) all ultimately reduce to a relatively small set of 276 protein superfamilies from which they all ultimately derive.
Exploring the Evolution of Novel Enzyme Functions within Structurally Defined Protein Superfamilies
My bold.
Oh, I see. And if that was shown to be false, would that change your mind about Axe’s work?