Friday, July 15, 2011

Peak Fallacy: A Follow-Up on Nature Paper Proving A = A

In my previous post I discussed a paper published in the leading journal Nature on protein evolution. In spite of the scientific evidence showing the evolution of proteins is unlikely, this paper is used as an apologetic by evolutionists for why said evolution is actually no big problem. The paper uses a somewhat circuitous method to arrive at its conclusion that protein evolution occurs early and often and that the findings are yet more “novel evidence of the common ancestry of life.” These conclusions are false and are based on a naïve and circular analysis. This is not easy to understand, however, because the analysis is circuitous. Here I will provide a simple explanation to help illustrate the fallacy.

Case 1: Two proteins that are different

Consider two proteins whose amino acid sequences are very different. You align and compare the sequences and find that fewer than 10% of their amino acids are identical. This is about what would be expected from two random sequences and we say that the distance between the proteins is large.

Now imagine that a mutation occurs, changing one of the amino acids in one of the proteins. Most likely the distance between the proteins remains unchanged. For instance, let’s say the amino acid mutated from alanine to valine. Meanwhile, the corresponding amino acid in the other protein is proline. So both before and after the mutation, the amino acids in the two proteins do not match, and the distance between the proteins remains unchanged.

There is a slight chance that the mutation could produce a match. The alanine could mutate to a proline. On the other hand, there is a slight chance that the mutation could eliminate a match. This would be the case if the original amino acid had been a proline. So there is a slight chance the mutation could either decrease or increase the distance. But these to events are unlikely, and in any case they average out. The bottom line is that for two distant proteins, a mutation is not likely to change the distance between them.

Case 2: Two proteins that are similar

Next consider two proteins whose amino acid sequences are very similar. Instead of 10%, let’say 90% of their amino acids are identical. In this case the distance between the proteins is small.

This time, a mutation most likely would increase the distance between the two proteins. This is because the mutation will probably change an amino acid that previously had matched the corresponding amino acid in the other protein. For instance, if corresponding amino acids in the two proteins are both alanine, and the mutation causes one of them to switch to valine, then the match is destroyed.

There is a slight chance that the mutation could produce no change to the distance, or even decrease the distance by producing a match, but in most cases the mutation increases the distance.

Conclusion: A mutation’s effect on distance depends on the distance itself

So the conclusion is that a mutation’s effect on distance depends on the distance itself. Specifically, for distant proteins, a mutation has little effect on distance. But for similar proteins, mutations usually increase distance.

This simple fact was a key result of the paper. In their research, the evolutionists performed a lengthy and circuitous analysis which demonstrated this straightforward relationship. It would be like a Rube Goldberg device that prints out: A = A.

Here is the figure in the paper showing the relationship. The average effect of mutations on the distance between the two proteins is plotted on the vertical axis. A small value close to zero means mutations will increase the distance. A value of one means that overall the mutations do not change the distance. These data are graphed versus the distance between the two proteins on the horizontal axis. As the solid line indicates, at short distances mutations will increase the distance. And at long distances mutations will not change the distance.

It may seem amazing that a leading journal would accept such meaningless research, but this is not too unusual for evolutionary papers. Starting with the “evolution is true” premise tends to blind one to objective analysis. And this is true for reviewers as well as authors.

This is a case where the classic criticism “it isn’t even wrong” applies. Researchers can make mistakes and an otherwise fruitful approach can produce erroneous results. But here there is no computational or scientific mistake. The evolutionists haven’t blundered. Rather, qua evolutionists they have followed their instincts and interpreted a meaningless tautology as powerful evidence for evolution. And of course evolutionists have welcomed it with open arms. The entire approach is absurd, but evolutionists are blind to their abuse of science.

Religion drives science, and it matters.

T: "CH doesn't 'discuss' science here. CH takes legitimate scientific research published in the primary literature, comes up with some goofy strawman version of what was actually done, then goes NUH UH!! at the top of his lungs."

Another evolution prediction proves true!

2. My initial reaction to the above comment was "why don't you ban this clown?". But then I realized that it is far better to let him make a fool of himself - you can't even mock him, because he's too stupid to understand it.

Good job.

3. Thanks Dr. Hunter,,, no wonder neo-Darwinists fight tooth and nail to keep any evidence against neo-Darwinism from being taught and discussed in schools,,, it can't withstand scrutiny!!!

Stephen Meyer - Functional Proteins And Information For Body Plans - video
http://www.metacafe.com/watch/4050681

The Case Against a Darwinian Origin of Protein Folds - Douglas Axe - 2010
Excerpt Pg. 11: "Based on analysis of the genomes of 447 bacterial species, the projected number of different domain structures per species averages 991. Comparing this to the number of pathways by which metabolic processes are carried out, which is around 263 for E. coli, provides a rough figure of three or four new domain folds being needed, on average, for every new metabolic pathway. In order to accomplish this successfully, an evolutionary search would need to be capable of locating sequences that amount to anything from one in 10^159 to one in 10^308 possibilities, something the neo-Darwinian model falls short of by a very wide margin."
http://bio-complexity.org/ojs/index.php/main/article/view/BIO-C.2010.1

What makes matters much worse for the materialist is that he will try to assert that existing functional proteins of one structure can easily mutate into other functional proteins, of a completely different structure or function, by pure chance. Yet once again the empirical evidence betrays the materialist. The proteins that are found in life are shown to be highly constrained in their ability to evolve into other proteins:

Dollo’s law, the symmetry of time, and the edge of evolution - Michael Behe - Oct 2009
Excerpt: Nature has recently published an interesting paper which places severe limits on Darwinian evolution.,,, A time-symmetric Dollo’s law turns the notion of “pre-adaptation” on its head. The law instead predicts something like “pre-sequestration”, where proteins that are currently being used for one complex purpose are very unlikely to be available for either reversion to past functions or future alternative uses.
http://www.evolutionnews.org/2009/10/dollos_law_the_symmetry_of_tim.html

Severe Limits to Darwinian Evolution: - Michael Behe - Oct. 2009
Excerpt: The immediate, obvious implication is that the 2009 results render problematic even pretty small changes in structure/function for all proteins — not just the ones he worked on.,,,Thanks to Thornton’s impressive work, we can now see that the limits to Darwinian evolution are more severe than even I had supposed.
http://www.evolutionnews.org/2009/10/severe_limits_to_darwinian_evo.html#more

Stability effects of mutations and protein evolvability. October 2009
Excerpt: The accepted paradigm that proteins can tolerate nearly any amino acid substitution has been replaced by the view that the deleterious effects of mutations, and especially their tendency to undermine the thermodynamic and kinetic stability of protein, is a major constraint on protein evolvability,,
http://www.ncbi.nlm.nih.gov/pubmed/19765975

The Evolutionary Accessibility of New Enzyme Functions: A Case Study from the Biotin Pathway - Ann K. Gauger and Douglas D. Axe - April 2011
Excerpt: We infer from the mutants examined that successful functional conversion would in this case require seven or more nucleotide substitutions. But evolutionary innovations requiring that many changes would be extraordinarily rare, becoming probable only on timescales much longer than the age of life on earth.

When Theory and Experiment Collide — April 16th, 2011 by Douglas Axe
Excerpt: Based on our experimental observations and on calculations we made using a published population model [3], we estimated that Darwin’s mechanism would need a truly staggering amount of time—a trillion trillion years or more—to accomplish the seemingly subtle change in enzyme function that we studied.
http://biologicinstitute.org/2011/04/16/when-theory-and-experiment-collide/

4. This following paper, and audio interview, shows that there is a severe 'fitness cost' for cells to carry 'transitional' proteins that have not achieved full functionality yet:

Reductive Evolution Can Prevent Populations from Taking Simple Adaptive Paths to High Fitness - May 2010
Excerpt: Despite the theoretical existence of this short adaptive path to high fitness, multiple independent lines grown in tryptophan-limiting liquid culture failed to take it. Instead, cells consistently acquired mutations that reduced expression of the double-mutant trpA gene. Our results show that competition between reductive and constructive paths may significantly decrease the likelihood that a particular constructive path will be taken.
http://bio-complexity.org/ojs/index.php/main/article/view/BIO-C.2010.2

Testing Evolution in the Lab With Biologic Institute's Ann Gauger - audio
http://www.idthefuture.com/2010/05/testing_evolution_in_the_lab_w.html

Even if evolution somehow managed to overcome these impossible hurdles for generating novel proteins by totally natural means, evolution would still face the monumental hurdles of generating complimentary protein/protein binding sites, in which the novel proteins would actually interact with each other in order to accomplish the specific tasks needed in a cell (it is estimated that there are least 10,000 different types of protein-protein binding sites in a 'simple' cell; Behe: Edge Of Evolution).

What does the recent hard evidence say about novel protein-protein binding site generation?

"The likelihood of developing two binding sites in a protein complex would be the square of the probability of developing one: a double CCC (chloroquine complexity cluster), 10^20 times 10^20, which is 10^40. There have likely been fewer than 10^40 cells in the entire world in the past 4 billion years, so the odds are against a single event of this variety (just 2 binding sites being generated by accident) in the history of life. It is biologically unreasonable."
Michael J. Behe PhD. (from page 146 of his book "Edge of Evolution")

Nature Paper,, Finds Darwinian Processes Lacking - Michael Behe - Oct. 2009
Excerpt: Now, thanks to the work of Bridgham et al (2009), even such apparently minor switches in structure and function (of a protein to its supposed ancestral form) are shown to be quite problematic. It seems Darwinian processes can’t manage to do even as much as I had thought. (which was 1 in 10^40 for just 2 binding sites)
http://www.evolutionnews.org/2009/10/nature_paper_finally_reaches_t.html

5. Thanks Dr. Hunter,,, no wonder neo-Darwinists fight tooth and nail to keep any evidence against neo-Darwinism from being taught and discussed in schools,,, it can't withstand scrutiny!!!

Stephen Meyer - Functional Proteins And Information For Body Plans - video
http://www.metacafe.com/watch/4050681

The Case Against a Darwinian Origin of Protein Folds - Douglas Axe - 2010
Excerpt Pg. 11: "Based on analysis of the genomes of 447 bacterial species, the projected number of different domain structures per species averages 991. Comparing this to the number of pathways by which metabolic processes are carried out, which is around 263 for E. coli, provides a rough figure of three or four new domain folds being needed, on average, for every new metabolic pathway. In order to accomplish this successfully, an evolutionary search would need to be capable of locating sequences that amount to anything from one in 10^159 to one in 10^308 possibilities, something the neo-Darwinian model falls short of by a very wide margin."
http://bio-complexity.org/ojs/index.php/main/article/view/BIO-C.2010.1

What makes matters much worse for the materialist is that he will try to assert that existing functional proteins of one structure can easily mutate into other functional proteins, of a completely different structure or function, by pure chance. Yet once again the empirical evidence betrays the materialist. The proteins that are found in life are shown to be highly constrained in their ability to evolve into other proteins:

Dollo’s law, the symmetry of time, and the edge of evolution - Michael Behe - Oct 2009
Excerpt: Nature has recently published an interesting paper which places severe limits on Darwinian evolution.,,, A time-symmetric Dollo’s law turns the notion of “pre-adaptation” on its head. The law instead predicts something like “pre-sequestration”, where proteins that are currently being used for one complex purpose are very unlikely to be available for either reversion to past functions or future alternative uses.
http://www.evolutionnews.org/2009/10/dollos_law_the_symmetry_of_tim.html

Severe Limits to Darwinian Evolution: - Michael Behe - Oct. 2009
Excerpt: The immediate, obvious implication is that the 2009 results render problematic even pretty small changes in structure/function for all proteins — not just the ones he worked on.,,,Thanks to Thornton’s impressive work, we can now see that the limits to Darwinian evolution are more severe than even I had supposed.
http://www.evolutionnews.org/2009/10/severe_limits_to_darwinian_evo.html#more

Stability effects of mutations and protein evolvability. October 2009
Excerpt: The accepted paradigm that proteins can tolerate nearly any amino acid substitution has been replaced by the view that the deleterious effects of mutations, and especially their tendency to undermine the thermodynamic and kinetic stability of protein, is a major constraint on protein evolvability,,
http://www.ncbi.nlm.nih.gov/pubmed/19765975

The Evolutionary Accessibility of New Enzyme Functions: A Case Study from the Biotin Pathway - Ann K. Gauger and Douglas D. Axe - April 2011
Excerpt: We infer from the mutants examined that successful functional conversion would in this case require seven or more nucleotide substitutions. But evolutionary innovations requiring that many changes would be extraordinarily rare, becoming probable only on timescales much longer than the age of life on earth.

When Theory and Experiment Collide — April 16th, 2011 by Douglas Axe
Excerpt: Based on our experimental observations and on calculations we made using a published population model [3], we estimated that Darwin’s mechanism would need a truly staggering amount of time—a trillion trillion years or more—to accomplish the seemingly subtle change in enzyme function that we studied.
http://biologicinstitute.org/2011/04/16/when-theory-and-experiment-collide/

6. Thanks Dr, Hunter. I can't figure out why anyone would even waste their time doing a study like this. They should have been able to figure that out without doing a study. The results are obvious!

I don't see how anyone can claim this is evidence for protein evolution. How foolish!

What evolutionists have to explain is how the proteins got there to begin with. Once they exist, mutations will move them either towards each other or away from each other, but what does that prove?

Thorton claims this was an evolutionary prediction. How can you lose with predictions like that? The closer proteins are to each other the more likely mutations are to move them away from each other. Duh!

We'll be impressed when they can predict or explain the original origin of proteins.

7. I admire Dr Hunter's persistent efforts to clarify what he thinks is wrong about the Nature paper by Povolotskaya & Kondrashov. He may have valid points, but I can't tell, because the full text off the paper is behind a paywall, and as much as I find the topic of interest, I am not motivated to pay \$32 to open that paywall and find out what the authors themselves had to say.

The fact that the paper has been cited in ongoing work, especially in studies of how epistatic interactions might have constrained pathways of protein evolution, indicates that it has been useful to other scientists. Maybe all of these people are in error, considering that they assume, as a working hypothesis, that proteins did evolve, and were not individually designed by a demiurge:

Reciprocal Sign Epistasis between Frequently Experimentally Evolved Adaptive Mutations Causes a Rugged Fitness Landscape

Initial Mutations Direct Alternative Pathways of Protein Evolution

Pervasive Cryptic Epistasis in Molecular Evolution

8. Pedant said:

"He may have valid points, but I can't tell, because the full text off the paper is behind a paywall, and as much as I find the topic of interest, I am not motivated to pay \$32 to open that paywall and find out what the authors themselves had to say."
===

Funny, when I once said the same thing here in the comments section about a paywall and was demonized and vilified by the resident trash talking Neanderthals. But I'm sure you'll get a free pass on it by your fellow parishioners.

'smirk'

9. Actually, Nature and Science are well known for publishing findings that seem obvious but have never been formally expressed. THis finding is simple, but not at all tautologous (if they said "proteins that are more different are more different" I would see the point, but they don't), and as Pedant points out other researchers have found it useful. THere is no logical or experimental flaw in this paper (as there are in many other Nature/Science papers). so what's the problem?

10. I have access to the paper and I will take time to read it and comment later. (Pedant, feel free to email me and I will shoot you a PDF of the paper.)

But right off the bat I can tell that Cornelius's analysis is not even wrong (to quote his newly favorite phrase), even in its second revision. Not only does he oversimplify the model at hand, he also makes errors as he analyzes his toy model.

The toy model, discussed in the post under Cases 1 and 2, is that of a random walk through protein sequences. Two proteins start at initial positions and make random steps. It is well known that a singe random walker moves away from the point of origin, on average: the distance squared is proportional to the number of steps taken. The same is true for two random walkers: the square of the distance between them grows, on average, no matter what the initial distance was.

This contradicts Cornelius's conclusion in Case 1 (the distance should stay the same).

11. Addendum. In a finite space, the average distance would reach a limit. However, with 22 amino acids, two randomly picked proteins would have an average overlap of 4.5%. If you start at an overlap of 10% they would still drift apart.

12. Like others I don't have access to the full paper, already just in reading the abstract it's clear that the paper covers more ground than CH indicates.

But even his explanation of the chart above is incomplete (and frankly not terribly clear) - there's another plot on the chart that he doesn't even explain. And of course he completely skirts over one of the key findings in the abstract: "We show that ancient proteins are still diverging from each other, indicating an ongoing expansion of the protein sequence universe.:

As others point out CH has a history of reviewing a paper like this (which is often fire-walled) with his own unique perspective. That's fine of course and he is entitled to, but we must always remember CH has some very distinct confirmation biases of his own which color and distort his analysis (even if he won't admit they are they exist, rest assured they are; CH, like the rest of us, is not immune to them even if he would like to think he is).

But again, if CH thinks the paper is inadequate or wrong, why not write to the editors of Nature? Why does he limit his voice to this obscure little barely-read blog?

13. As mentioned above, CH has taken another well written and interesting bit of science research and butchered it beyond recognition. The paper itself is about the total available protein sequence space, and how the divergence patterns we see today indicate that space is much bigger than proteins have been able to explore in the 3.5BY since LUCA. The analogy made in the paper is to the physical expansion of the universe as first proposed by Edwin Hubble. The expansion is inferred from the fact that more distant galaxies are receding from us faster than closer ones. The researchers applied statistical analysis to protein sequences and came up with a similar result, that the protein space universe is still expanding

diagram of expanding protein space

The rest of CH's 'recap' is his standard boilerplate anti-science blither, not worth reading.

14. Note that Cornelius hasn't actually addressed the comments made in response to his original post. Instead, Cornelius is attempting to attack evolutionary theory by suggesting it's not justifiable based on observation alone.

However, to again quote Deutsch…

…experimental testing is useless in the absence of a good explanation.

This includes all fields of science, not just evolution. Only gravitationalist would find the specific motions of objects meaningful, etc. Furthermore, all observations themselves are theory laden. Only realists would find the particular patterns of photons that strike our retina meaningful. It could be that there are no such things as photons and that everything is merely facets of our internal selves. Only heliocentrist find observations of the night sky meaningful. After all it could be that we're surrounded by a giant planetarium that merely presents a highly elaborate simulation of a heliocentric solar system, etc.

The majority of scientific philosophers accept Popper's explanation for scientific progress. New theories are not inferred from observations, but are hypotheses based on conjecture and which are discarded based on refutation. They also accept that progress made in this manner is found to be reliable. However, most scientific philosophers do not see why this should be the case. It cannot be justified due to the problem of induction.

This is what Cornelius' underlying argument is appealing to here. He's just disguised it by presenting an incomplete picture of the paper and claiming it's a tautology.

Deutsch has an entire chapter in his book, The Fabric of Reality, in which he addresses just this sort of argument, in detail.

Philosophers today yearn for this missing justification. They no longer believe that induction would provide it, yet they have an induction-shaped gap in their scheme of things, just as religious people who have lost their faith suffer from a 'God-shaped gap' in their scheme of things. But in my opinion there is little difference between having an X-shaped gap in one's scheme of things and believing in X. Hence to fit in with the more sophisticated conception of the problem of induction, I wish to redefine the term 'inductivist' to mean someone who believes that the invalidity of inductive justification is a problem for the foundations of science. In other words, an inductivist believes that there is a gap which must be filled, if not by a principle of induction then by something else. Some inductivists do not mind being so designated. Others do, so I shall call them crypto-inductivists.

In case it's not clear, Cornelius is essentially arguing that we cannot use induction to justify evolutionary theory. The results of the experiment are just observations, which cannot justify the conclusion outside the context of an explanation. But he then goes on to say that, unless we can use induction, then evolution isn't justified.

However, this objection is, in itself, a form of inductive reasoning which he claims is invalid. Nor has he disclosed a proposed solution for the problem of induction, despite numerous and direct requests for him to do so.

So, again, Cornelius must be smuggling in some assumption in his argument. Otherwise, this is yet another example of hand waving over a theory that he personally disagrees with.

I'd not that Deutsch goes on to elaborate on this via a hypothetical dialog between himself and a crypto-inductivist in his book. The following blog post presents a variation of this dialog in the form of the author and an a-unicornist.

15. The relationship of the probability of an actual substitution to increase distance is linear with distance. Why specifically linear and what is the meaning of the slope of the blue and green lines?

16. I see a lot of blather, bombast and bitterness from the critics here, but not much refutation of Dr. Hunter's arguments. I guess calling him stupid is the best they can do.

I see a lot of blather, bombast and bitterness from the critics here, but not much refutation of Dr. Hunter's arguments.

Since his 'argument' was an almost comical strawman misrepresentation of the actual paper, why do you think his nonsense needs refuting?

I guess calling him stupid is the best they can do.

While his arguments are pretty stupid we don't think CH himself is stupid at all. Possibly mentally unbalanced, certainly lacking the honesty and scruples usually found in most Christians, but not stupid.

18. Evey Solara:

The relationship of the probability of an actual substitution to increase distance is linear with distance. Why specifically linear…?

Excellent question. Thanks to the kindness of oleg, I now have the full text of the paper, and can give you a quick answer by quoting what the authors wrote:

“The linear relationship between Nt/Na and D suggests that, in contrast to the accelerating expansion of physical space, the expansion of proteins in sequence space is relatively constant (Fig. 3).”

To understand that quotation, a longer answer is needed that unpacks some of the considerable amount of conceptual baggage in the paper. Note that the y-axis values Nt/Na are not probabilities, but ratios of two computed values:

Nt = numbers of substitutions in a group of protein sequences from “sister” organisms that moved those sequences towards a reference sequence.

Na= numbers of substitutions that moved the test sequences away from the reference sequence.

(I interpret the terms “towards” and “away” to mean “more similar” and “less similar.”)

…and what is the meaning of the slope of the blue and green lines?

That’s something I haven’t figured out yet, because I haven’t fully unpacked the figure legend, but I’ll quote it for your reference:

“Figure 3. The rate of expansion of the protein sequence universe. Nt/Na values for sites where the derived or the ancestral amino acid was identical to the reference amino acid (blue) and for all type-1 and type-2 sites (green), including those where the directionality of the substitution was determined from the BLOSUM62 matrix. TheNt/Na ratio for fourfold-synonymous sites is shown in orange. Least-squares regression with correlation coefficient R is shown for the Nt/Na values calculated without the BLOSUM62 correction(blue). Error bars, s.e.m.”

As you can see, there are several concepts in there that need explaining. I hope that some of the more scientifically literate commenters here will provide further enlightenment.

19. Perhaps it has something to do with protein structure/function biophysics and epistasis.

20. RR: "I see a lot of blather, bombast and bitterness from the critics here, but not much refutation of Dr. Hunter's arguments. I guess calling him stupid is the best they can do."

I don't think anybody here thinks CH is stupid by any means, not given his previous scientific work and qualifications. Neither is anybody "bitter" (about what exactly?). I think if anything people feel rather sorry for CH in that he is blind to his own confirmation biases (we all have them). To me his work is the product of cognitive dissonance trying to resolve itself. I often think if only he would cast his critical and skeptical eye on his religious beliefs in the same way he does on evolution - the outcome would be very interesting!

As to the above post, as others have pointed out CH has not really made an argument, and the original piece he is quoting from appears to be far more complex than CH makes out (and again there is more data on that chart than CH has acknowledge or explained).

21. Evey Solara:

Perhaps it has something to do with protein structure/function biophysics and epistasis.

Exactly. Povolotskaya & Kondrashov discuss epistatic constraints extensively, and other authors have picked up on this, as I noted in an earlier post.

22. Over the weekend, I carefully read the paper by Povolotskaya and Kondrashov. Their work has a sound basis. We can turn Cornelius's profound misunderstanding of the article into a teachable moment and learn some amazing science.

I will split the discussion into several comments. In the next comment, I will use the toy model in which all protein configurations are allowed. We will deal with the complications due to the rugged landscape later.

23. This comment has been removed by the author.

24. Toy model: flat fitness landscape

Let us begin with a toy model where any random mutation is allowed. Proteins randomly walk in the configuration space. This model is far from realistic, but it will allow us to introduce the concept of equilibrium, which is quite important to the work described in the paper.

Proteins start from the same configuration (common ancestor) and undergo random mutations. Would proteins tend to diverge from each other in this case and how would the rate of divergence depend on the distance? Cornelius seems to think that closely related proteins should diverge fast and more distant relatives less so. He offers a simple explanation based on considering just two proteins. But the answer is not so simple.

What Cornelius forgets is that we are dealing not with two proteins but with many. Suppose that the proteins have been mutating for a long time and have uniformly filled the configuration space. (For simplicity, think that there is one protein in every configuration.) If two proteins have the same amino acid in a given configuration, any change to that amino acid will increase the distance between them. Say, two sequences NLKIY and NLKIG will diverge further if we replace K in one of them. In contrast, shrinking the distance between any two proteins with different amino acids requires a very specific change, in this case, Y to G. However, NLKIY is not the only available protein configuration in the ensemble. Any protein NLKI* can change into NLKIG by mutating * into G. This factor, missing from Cornelius's analysis, makes the observed numbers of proteins mutating to and away from a reference equal.

One can perform an actual calculation using combinatorics and it shows that the to and away numbers are indeed equal for any distance between proteins. There is, however, a much simpler way to see that the rates should be equal when proteins had a chance to access all of the configuration space and thus reached an equilibrium distribution. Let's look at protein pairs with distance D. If the away rate (distance change from D to D+1) is greater than the to rate (distance change from D+1 to D) then after some time the number of pairs with distance D will decrease. That means that the distribution will have changed, which contradicts the assumption of equilibrium. So the rates must be equal. This is known as the principle of detailed balance in statistical physics.

Yellow data points in Figure 3 show precisely such a case, where mutations are readily available and proteins have explored all of the available space. The to and away rates are equal to each other for all distances.

A different result obtains if proteins have not yet reached equilibrium and are still diverging away from their initial state. Then away rates exceed to rates for all distances.
So contrary to Cornelius's half-baked analysis, it is not at all obvious that similar proteins tend to diverge no matter what. Once they have explored all of the protein space available to them, we will observe equal numbers of proteins mutating to and away from each other, whatever the distance between them. This is guaranteed by the concept of detailed balance.

In the next comment we will add the complications arising from a fitness landscape. We will see that the concept of equilibrium and detailed balance will once again be useful.

25. This comment has been removed by the author.

26. A simple analogy: diffusion

If you find the above discussion too technical, here is a simple analogy. As any analogy, it is imperfect but it serves to illustrate the key difference between equilibrium and out-of-equilibrium situations.

You and a bunch of friends are dropped off in the middle of an island. To explore the island, your friends start random walks. Every second, each of them, independently, makes a random step in an arbitrary direction. You stay at the origin and observe the process. (You could also do a random walk, but the conclusion will be the same.)

If you select a given radius R, draw a circle of that radius and count the number of people crossing the line in the directions to and away from you. What would be the result? It is easy to get the wrong answer by following Cornelius's approach and looking at just one pair of people: you and a friend. At any given moment, a friend's step can be toward or away from you with equal probabilities. Does that mean the number of people crossing the line is going to be equal? Not necessarily.

The process we are looking at is known as diffusion. By taking random steps, your friends are diffusing across the island. Even though their steps are random, there should clearly be an overall outward flux of people at the beginning. People are spreading out. How can we understand that? That's actually quite simple. Although people just inside and just outside the line have equal probabilities of crossing it, there are initially more people on the inside than on the outside, so the outward flux is greater. (Technically speaking, the overall flux is in the direction opposite to that of the gradient of concentration.)

As time goes on, the concentration of your friends throughout the island will even out and become uniform. Then the outward and inward fluxes will become equal.

Let's summarize. Initially, when your friends are starting to spread out, the outward flux of people outweighs the inward flux at any radius. After a while, your friends will have spread out all over the island and their distribution will reach equilibrium. In equilibrium, the outward and inward fluxes are the same at any radius. This is a key idea behind the work of Povolotskaya and Kondrashov. Cornelius completely missed it.

27. Ah - real science!

With regard to diffision, does it help to think in terms of the law of large numbers?

http://en.wikipedia.org/wiki/Law_of_large_numbers

Each step can randomly take you closer to or further from the point of origion (think in X / Y axis terms). And for larger numbers, getting increased absolute distance from the point of origion should become harder and harder?

28. Oleg

nice analysis but...

1. Does cell really "explore protein landscape" by spending massive resources on producing trillions of proteins and than spending more resources on dismantling useless ones.

2. This "exploring" is likely to produce harmful protein which will destruct some cell process.

29. Producing a functional protein with a different shape from a functional protein would require multiple mutations of the gene producing the protein. All of the proteins formed during the "transition" would not be beneficial or they would still have their original function but not add any selective advantage. Protein evolution on the scale required for universal common descent has huge mathematical improbability problems.

30. Oleg,

I'm not so sure how relevant the concept of equilibrium (or "detailed balance") is for understanding the paper. It seems to me that Cornelius' argument is correct that distance tends to increase for proteins that are sufficiently similar, while distance is preserved for sufficiently dissimilar proteins.

In a flat fitness landscape, equilibrium occurs for proteins with very high D-values, to the right of the highest D-values in figure 3.

In fact, we can model the system with a diffusion approximation to a Markov model. The Markov model is:

There are N amino acids per protein. i AA's are identical between the same protein of two species (N-i are therefore different). There are K different AA's. Then given a change in one species, the probability that i -> i+1 is (1-i/N)/K, and the probability that i->i-1 is i/N.

According to the diffusion approximation, the stationary distribution (equilibrium) peaks sharply around D=K/(K+1) for large N, or about D=95%. At that point Nt/Na=1. If you extrapolate the straight line in figure 3, this 95% is almost exactly where the line crosses Nt/Na=1.

So, it's no big surprise that Nt/Na goes up in figure 3, since the D-values are all below equilibrium. Isn't that what Cornelius meant to say here? That's how I read it anyway.

31. Rich,

No, it is not enough to consider a single random walker. This is the crucial mistake of Cornelius. Any given random walker has equal chances of stepping closer to you and farther away. (This is particularly obvious in one dimension.) However, if the concentration of walkers decreases with the distance, more walkers will be crossing any given line outward than inward.

32. Neal,

As I explained, we are dealing with a toy model that ignores fitness. We will get to it in later comments. Stay tuned.

33. troy,

You are repeating Hunter's error by considering just two proteins instead of an ensemble. The authors counted the total number of moves toward or away from a reference, not whether two specific proteins would get closer of farther away.

Here is the combinatorial calculation that I mentioned above.

Assume first a state of equilibrium that is reached after a sufficiently long exploration of the protein space. For simplicity, there is exactly one protein in every possible state. For proteins of length L with N types of amino acids, the number of proteins that have M amino acids in the same places as a reference is
P_M = (N-1)^{L-M} L!/[M!(L-M)!].
The first factor comes from having a choice of N-1 AAs that differ from the reference. The second factor is the number of ways to choose M overlapping sites.

We now randomly mutate one AA in every protein. To mutate toward the reference, we must choose one of the L-M non-overlapping AAs and change it to the right one. The probability of that is Q_t = [(L-M)/L] (1/N). Thus the expectation number for proteins moving toward the reference is
N_t = P_M Q_t = (N-1)^{L-M} (L-1)!/[M!(L-M-1)!].

To get from overlap M+1 to M, we must take one of the M+1 overlaping AAs and change it to one of the N-1 states. The probability of that is Q_a = [(M+1)/L](N-1)/N. The expectation number of proteins moving away from the reference is
N_a = P_{M+1} Q_a = (N-1)^{L-M} (L-1)!/[M!(L-M-1)!].
Clearly, N_t/N_a = 1 in equilibrium.

If the system has been evolving from a state with all overlapping AAs and has not yet reached equilibrium then the ensemble will be skewed toward larger overlaps, which means that the ratio P_M/P_{M+1} will be smaller than it is in equilibrium. Accordingly, N_t/N_a will be less than 1.

FIgure 3 in the paper (reproduced in Cornelius's post) shows two sets of data. One, in yellow, is for proteins that have an easy time mutating and have reached equilibrium. They have N_t/N_a = 1 for all distances. The other, in green and blue, has N_t/N_a < 1. These proteins have not yet reached equilibrium.

The word "equilibrium" is mentioned explicitly in the paper. Here is an excerpt from the bottom of the first page (right column).

If N_t/N_a < 1 then the sister sequences are evolving away from the reference in sequence space, and, conversely, if N_t/N_a > 1 then the sister sequences are evolving towards the reference. If N_t/N_a = 1 then the distance in sequence space between them is at an evolutionary equilibrium.

34. Oleg:

You are repeating Hunter's error by considering just two proteins instead of an ensemble. The authors counted the total number of moves toward or away from a reference, not whether two specific proteins would get closer of farther away.

Perhaps you're right, but you haven't convinced me yet. I still think the comparison is between two proteins. Look at the first paragraph on p923: "D, the protein distance between the ancestral state of the sister sequences [the ancestral state being sequence 1] and the reference sequence [# 2]". In other words, the comparison is between 2 proteins (ancestral of sisters and reference) with a given distance D.

The question then is whether mutations from the sisters' ancestor (towards the sisters) tend to be away from the reference (Nt/Na<1) or not (Nt/Na>=1) and how this depends on D. In figure 3, D-values have been binned (within 0.05) to be able to calculate standard errors, since there a very few pairs (ref-ancestor) with a unique D-value.

Assume first a state of equilibrium that is reached after a sufficiently long exploration of the protein space. For simplicity, there is exactly one protein in every possible state.

That doesn't seem like a reasonable assumption at all. Given the vastness of the sequence space, there is no way that the sisters' recent ancestor has explored much of the space since divergence from the reference.

I still think it's remarkable that my calculation (D_equilibrium = K/(K+1), where K = your N = # AA's = 20, hence D_eq = 95%) fits the data so well.

What I don't understand is what is meant by the "fourfold synonymous sites" in the legend to figure 3. Are we talking about a sub-sequence here, of length 4 (small N = your L)? Maybe that might explain my confusion, assuming I'm confused.

35. troy,

It helps to think carefully about the definitions of N_t and N_a. These numbers do not characterize individual pairs of proteins. To quote the paper, these are

the numbers of substitutions in the sister sequences that moved the sequence away from (N_a) and towards (N_t) the reference sequence.

It took me a while to figure it out because the authors made the analogy with the universe expansion, a deterministic process that works for any pair of galaxies. But in this case it is really about ensembles, not individual pairs of proteins. Take the time to read and absorb the paper.

36. Tedford the Idiot said...

Producing a functional protein with a different shape from a functional protein would require multiple mutations of the gene producing the protein. All of the proteins formed during the "transition" would not be beneficial or they would still have their original function but not add any selective advantage.

Mr. Fat-headed Pastor still doesn't understand that 'beneficial' is determined with reference to the effects of the environment that 'transition' is in. What happens when the environment is constantly changing Mr. Fat-headed Pastor?

Protein evolution on the scale required for universal common descent has huge mathematical improbability problems.

Or Mr. Fat-headed Pastor has huge scientific and mathematical illiteracy problems.

I know which way I'd bet.

37. Oleg,

Thanks for the advice, but I did take some time to read and absorb the paper (not too much though, I do have other stuff to do). It may be helpful to think in terms of ensembles when doing the theoretical model calculations, but the data refer to pairs of proteins: each data point refers to a pair of proteins - the relatively recent ancestor (A) of two sister proteins (S1 and S2) and a reference protein (R) that shares a common ancestor with A. The comparison of S1 and S2 with A tells us something about Nt and Na, and since S1 and S2 have diverged recently, it's safe to assume that R hasn't changed in the mean time.

Therefore, we can compare A-R pairs, and see in what direction A has mutated to S1/S2 compared to R, and related this to the D between A and R.

Where do you suppose my summary goes wrong?

38. Until the two protein sequences are very distant from each other (as much as D of 80) almost all random mutations (~95%) increase this distance. Then the graph goes through a phase transition and by a D of about 95 the mutations are just as likely to increase D as they are to decrease it.

39. troy: The comparison of S1 and S2 with A tells us something about Nt and Na

No, it doesn't. N_t and N_a are obtained by summing over pairs where proteins move to and away. Comparison of protein pairs provides a single increment for N_t or N_a (or sometimes neither). The summation is done over an ensemble of proteins.

40. troy and oleg,

Check a bit further down page 923:

"We applied this approach to alignments obtained from 572 clusters of orthologous groups (COGs) that have been previously inferred to have been present in the LUCA encompassing a total of 836 different bacterial and archaeal genomes. These COGs included 28 quadruplets of closely related genomes yielding a total of 13,589,431 cluster-reference alignments (Methods)."

The word "quadruplet" also appears in Methods repeatedly.

41. Heh. Nothing clears out a room full of Creationists quicker than the introduction of scientific details.

42. Single origin vs. multiple origins in the island analogy

Before we move on to more complicated fitness landscapes (ours is flat at the moment), let us discuss another important point that appears in the article. Povolotskaya and Kondrashov state that their data provide further evidence for common ancestry. What is this claim based on? Here is a quick explanation.

In our latest toy model, random walkers were unloaded on an island at a single location and dispersed from there. As you watched them aimlessly wander about, you observed that, initially, they moving away from you, on average. Although the walkers could make steps in any direction, the influx was dominated by the outflux at all distances, except at the boundaries of the island. Your version of Figure 3 (see Cornelius's post) will have the ratio of influx to influx trending up and reaching 1 when the distance reaches the size of the island. Later, as the walkers disperse through the island, the outflux decreases and the influx increases until they become equal. Once equilibrium sets in, you cannot detect any trend. The ratio of the influx and the outflux is 1 for any distance.

But let us change the initial conditions and put several groups of walkers at different spots around the island. How would the flux distribution look in that case? Very different! Suppose the island has diameter D and the typical distance between the spots with walkers is R. As you watch your friends disperse, the initial picture would remain the same as it was in the previous situation up to the distance R. The outflux would dominate over the influx because your friends are spreading out, while other groups have not reached the vicinity of your spot. Beyond distance R, people will be moving preferably away from their points of origin but those directions are different from your to and away directions, so they will average out. Beyond distance R, you will see no net flux. So this time, your version of Figure 3 will be quite different: the ratio of the influx to the outflux will reach 1 at distance R to the adjacent groups, rather than at the island size D. Of course, eventually the walkers dispersing from different spots will overlap and make the distribution uniform. In equilibrium, the influx will equal the outflux and your graph will be flat as the yellow data points in Figure 3.

Here is the take-home message. If your walkers have not yet dispersed uniformly through the island, you can learn quite a bit from the statistics of their movements such as that in Figure 3 of the paper. If they started from a single point of origin, you should observe that the outflux dominates the influx at all distances except the very edge. For multiple origins, the influx will catch up with the outflux at a smaller distance. That distance will tell you how far another point of origin was.

The data of Povolotskaya and Kondrashov show the former trend: the outflux dominates the influx for all distances between proteins. That enables them to say that (a) the proteins have not yet reached equilibrium and (b) they are dispersing from a single point of origin.

43. Oleg:

No, it doesn't. N_t and N_a are obtained by summing over pairs where proteins move to and away. Comparison of protein pairs provides a single increment for N_t or N_a (or sometimes neither). The summation is done over an ensemble of proteins.

Duh? Yes the numbers presented are averages (+/- se), taken over "ensembles" if you like. In fact, for a given COG and a given 0.05 range of D-values, the Nt are summed and the Na are summed over all ancestor-reference pairs. The ratio sum_Nt/sum_Na thus obtained is then averaged over all COGs within that same bin of D-values.

Be that as it may, a neutral model (the one I presented upthread) predicts that Nt/Na=1 for D=95%, which seems to fit the data in figure 3 very well. And it supports Cornelius' "intuitive" argument, based on comparing pairs of proteins (ancestor of sisters vs. reference), that for low D-values one expects Nt/Na<1 whereas Nt/Na=1 for sufficiently large D-values. The reasoning based on (expected values of) pairs of proteins seems valid, since the data of the paper are based on exactly such (averages of) pair-wise comparisons.

44. Pedant:

The word "quadruplet" also appears in Methods repeatedly.

Yeah. It seems to refer to 28 quadruplets of closely related species (from a total of 836 species). If I understand it correctly, this allowed them to estimate the ancestral state of a COG for a sister pair of species (from the quadruplet) using two outgroup species (from that same quadruplet). This yields better estimates of ancestral states than based on a single outgroup per pair of sister species.

The idea is that the ancestral state of a given site in the two sister species is identical to the state of that site in the outgroup species. This fails if all three species have a different amino acid at that site because then any of the three could be the ancestral state. With two outgroups it's (more) unlikely that all four amino acids will differ from each other.

45. troy,

Tell me what you think of the yellow data points in Figure 3. How would you explain them? And why are they there?

46. Oleg,

The yellow points refer to Nt/Na ratios of nucleotides (not amino acids) at fourfold-synonymous sites, e.g. the third nucleotide in TC* Serine codons.

They all reach equilibrium (Nt=Na) fairly quickly, beyond about D=15%. They are there to show that the method is not in principle incapable of demonstrating convergence, e.g. due to systematic underestimation of Nt/Na.

What I don't understand is why they level off at D>15%. Do you have an idea?

47. Of course I do. Read my comment and follow the combinatorial calculation. There are two points to which you should pay close attention.

1. I set out to compute two numbers: N_a, the number of proteins whose overlap with a reference increases from M to M+1; N_t, the number of proteins whose overlap decreases from M+1 to M. Once again, note that the latter is from M+1 to M, not from M to M-1. This is crucial.

2. These numbers depend not only on the probabilities with which a single protein increases its overlap from M to M+1 or decreases it from M+1 to M. (I call these probabilities Q_t and Q_a.) The numbers also depend on how many proteins there are in the initial states, namely states with overlaps M and M+1. These are numbers P_M and P_{M+1}. There is no reason to assume that P_M and P_{M+1} are equal.

48. I follow your calculations, Oleg (except I'm missing a factor 1/N in your Nt and Na), but what do they predict about the D-value at which Nt=Na?

49. In equilibrium, N_t=N_a for all distances.

50. Of course it is. But my question was, how do you account for the fact that Nt<Na when D<15% for the nucleotides in fourfold-synonymous sites?

51. What we are witnessing here with evolutionary replies is a prime example of not being able to see the forest because of the trees.

From the Sequence Space article, they say, "The need to maintain the structural and functional integrity of an evolving protein severely restricts the repertoire of acceptable amino-acid substitutions."

Good observation!

However, what makes this improbability, a certainty? .... Assuming evolution to be a fact.

The article continues, "We formulate a computational approach to study the rate of divergence of distant protein sequences and measure this rate for ancient proteins, those that were present in the last universal common ancestor. We show that ancient proteins are still diverging from each other, indicating an ongoing expansion of the protein sequence universe. The slow rate of this divergence is imposed by the sparseness of functional protein sequences in sequence space and the ruggedness of the protein fitness landscape: approximately 98 per cent of sites cannot accept an amino-acid substitution at any given moment but a vast majority of all sites MAY EVENTUALLY be permitted to evolve when other, compensatory, changes occur. Thus, approximately 3.5 x 10(9) yr has not been enough to reach the limit of divergent evolution of proteins, and for most proteins the limit of sequence similarity imposed by common function may not exceed that of random sequences."

How do they know ancient proteins are still diverging from the last universal common ancestor?

Because evolution is an ASSUMED fact and we have more proteins now than in some less complex organism that is assumed to be like the UNKNOWN last universal common ancestor.

Why haven't we reached the limit of divergence? With evolution being an assumed fact, the slow rate of observed divergence is explained by neatly stowing it in the "need more time" evolutionary bucket.

52. They have not yet reached complete equilibrium.

At the beginning of the diffusion process, all proteins are in the same state, so P_M is only nonzero for M=L. At the first step, there is only outflux from L to L-1. N_t/N_a is undefined for any distance. After L steps, all P_M are nonzero and you can defined the in and out fluxes. The in flux is still dwarfed by the out flux for any distance.

After a long time, the numbers will equilibrate and the in and out fluxes will become equal. In the interim, as equilibrium is approached, you will have some distances where the in flux is still dominated by the out flux.

When I have a chance, I will reproduce this numerically and post a figure.

53. While I'm glad Oleg is unpacking the details of the paper, we can take a step back and put things into perspective.

Cornelius' argument is really just a variant of the typical creationist objection against transitional fossils. We do not observe common ancestry in the relationship between fossils. We explain the very specific way fossils are related using common ancestry.

To use an analogy, when someone admits a photograph into evidence, the photo isn't an observation of some event occurring. Instead, that the event depicted actually occurred is the best explanation for the existence of the photograph.

However, given advances in digital imaging technology, it could be that the photograph was forged. Expert witness might be brought in to analyze the photograph for signs of tampering. Who had access to equipment necessary to perform such a forgery? Did they have the knowledge to use that equipment? What motive would they have? Etc.

However, it could be that some yet to be discovered party with some yet to be discovered motive and some yet to be discovered new technology forged the photo in a way that is undetectable by the expert witness. For example, It could be that they used a supercomputer to re-dither the entire photo so it appeared seamless. Or perhaps thousands of frame of moving video. We'd be none the wiser. Nor can we rule this out with 100% certainty.

We could say the same regarding eye witness testimony of clear and direct observations of events. Given advanced enough technology, such things could be forged as well by manipulating photons in real time or modifying the electrical impulses they are translated to before they reach the brain. It could be that some unknown party with some unknown motive and unknown technology actually performed such a forgery. We can't rule this out with 100% certainty either. However, given our current knowledge, our best explanation of such clear observations is that the event observed actually occurred.

Yet this is exactly the sort of objection that Cornelius is making.

Just as we do not observe common ancestry in relationship between fossils, we do not observe common ancestry in the particular relationship between proteins. His A = A is just a red-herring. Instead, the concussion of the paper is that common ancestry is the best explanation the particular relationships we observe.

However, if all of biology was carefully and intentionally designed by a designer, then why do we observe the particular relationships found in the paper? Clearly, I can't be an accident since that's clearly counter to the core claim of ID. Nor is this particular pattern a necessarily outcome for an abstract designer who has not defined limitations.

So, who is the forger? What is the motivation? How was the forgery performed?

The best explanation we've heard from Cornelius is evolutionary thought is essentially supernatural collateral damage due to a supposed battle between cosmic good and evil that's supposedly ragging in our age. We've all been duped by forces of evil. But this hinges on the assumption that apocalyptic theodicy is true divine relation.

How does Cornelius, or anyone else, know what represents true divine revelation, if anything? Specifically, were does Cornelius put divine revelation on the traditional hierarchy of deduction, induction and philosophy?

54. I wrote: Specifically, were does Cornelius put divine revelation on the traditional hierarchy of deduction, induction and philosophy?

In case it's not clear...

Since Cornelius' underlying objection is based on the fact that we cannot use induction to justify evolutionary theory, then he certainly cannot use induction to justify what is true revelation either.

As such, Cornelius must be smuggling in some assumption in his argument, which he shares with his audience.

55. Scott: "As such, Cornelius must be smuggling in some assumption in his argument, which he shares with his audience. "

Absolutely. I've always been puzzled why CH is so quick to identify everybody elses metaphysical assumptions, yet cannot acknowledge the possibility of his own - even though he admits that he thinks it is more than probable that the Christian god is the "designer" (and has stated this on more than one occasion). For some reason he thinks he is immune to metaphysical influences, but has never explained how or why they would be so. I'd be interested to know by what process or mechanism he can be sure his own religious and metaphysical beliefs do not impact his science. Asserting it doesn't make it so.

But CH seems curiously absent from these conversations nowadays; although he is regularly posting, but for some reason no longer chooses to respond to participate in the combox discussions.

56. Cornelius Hunter seems to object to the claim that the finding of the paper is evidence for common descent and I will try to clarify as far as I understand it what this objection might be.

In the diffusion model common descent means that the diffusion started from a single point. In order to demonstrate that you can theoretically predict the distribution and then compare it to the experimental data. The distribution would be the sequence differences of the proteins. This is old stuff that has been done. What the guys in the paper did is that they calculated some kind of derivative of this distribution (the fluxes) and compared these. Now the question is how significant this is?

Take for example data that fits to a function f(x) = x^2 reasonably well. Then the derivative of the data should fit to f(x) = 2x. But if we find that this is true is this really new evidence that the original function was correct?

57. Scott, flying squirrels have been observed to glide from tree to tree. They can go airborne for short periods, so why not assume that they could fly to great heights with eagles?

Science is all about testing assumptions and then ACCEPTING the results without bias. So we run real time tests on flying squirrels and determine that their flying ability is limited to gliding short distances.

Evolution is like this flying squirrel. When we see what evolution can do in real time observation, we see limited change. Period. Everything else is speculation. Evolutionists must bring in all the theological baggage to prop up their speculation.

If I see a turtle on top of a fence post while I'm canoeing down a river on a nice sunny day, I assume someone put it there, because I know the ability of turtles to climb. We know the ability of proteins to evolve is greatly constrained. That's what we see. Everything else is speculation and assumption.

58. Neal Tedford: "If I see a turtle on top of a fence post while I'm canoeing down a river on a nice sunny day, I assume someone put it there, because I know the ability of turtles to climb." (emphasis mine)

Well, you don't know the ability of turtles to climb.

Snapping Turtle Climbs a Fence

Climbing Turtles

or if you want to get really specific;

Turtles Climbing fences

This is a perfect illustration of the creationist mindset: "If I can't immediately imagine an explanation for something, I assume an outside agency must have interfered; no matter how little I know about the subject or how sparsely I've researched it, if at all." (and believe me, it does not take long to type "turtle climbing" into a search bar)

Your statement would have been more accurately stated:

"If I see a turtle on top of a fence post while I'm canoeing down a river on a nice sunny day, I assume someone put it there, because I don't know how else it would have gotten up there, and I can't be bothered to spend 10 seconds investigating alternate explanations"

59. Derick Childress said...

Neal Tedford: "If I see a turtle on top of a fence post while I'm canoeing down a river on a nice sunny day, I assume someone put it there, because I know the ability of turtles to climb." (emphasis mine)

Well, you don't know the ability of turtles to climb.

Snapping Turtle Climbs a Fence

Climbing Turtles

or if you want to get really specific;

Turtles Climbing fences

This is a perfect illustration of the creationist mindset: "If I can't immediately imagine an explanation for something, I assume an outside agency must have interfered; no matter how little I know about the subject or how sparsely I've researched it, if at all." (and believe me, it does not take long to type "turtle climbing" into a search bar)

Your statement would have been more accurately stated:

"If I see a turtle on top of a fence post while I'm canoeing down a river on a nice sunny day, I assume someone put it there, because I don't know how else it would have gotten up there, and I can't be bothered to spend 10 seconds investigating alternate explanations"

LOL! Another Tedford the Idiot OWN GOAL!

60. Neal Tedford: "Science is all about testing assumptions and then ACCEPTING the results without bias."

Oh, how I wish I could bottle and sell irony. I'd be a millionaire from Neal alone.

61. Derick, in your zeal to prove me wrong, you showed turtles climbing a chain link fence.

Not what I said. Read it again.

62. Thorton: "LOL! Another Tedford the Idiot OWN GOAL!"

Sometimes in the summer when all my favorite sitcoms are in reruns, I peruse through the archives here and read Neal's comments for amusement. My particular favorites are the ones where he uses the existence/nonexistence of things to make a point without an ounce of effort to see if those things exist or not.

One of my favorite threads: here and here.

Neal, what a friend you have in Google.

63. Oh, and this one too.

....And what do you know, that's the very same thread where you said:

Neal Tedford: "Old earth and Young earth interpretations of Genesis 1 both existed hundreds of years before Darwin."

http://goo.gl/eeHuT

Neal, that was eight months ago and I've been asking you ever since to either name those theologians who had an old earth interpretation of Genesis 1 in that period, or retract that claim. Which is it?

64. Derick Childress said...

Thorton: "LOL! Another Tedford the Idiot OWN GOAL!"

Sometimes in the summer when all my favorite sitcoms are in reruns, I peruse through the archives here and read Neal's comments for amusement. My particular favorites are the ones where he uses the existence/nonexistence of things to make a point without an ounce of effort to see if those things exist or not.

One of the things that makes Tedford so funny is that the blustering windbag has no idea just how clueless he is. A poster boy for Dunning-Kruger if ever there was one. Tedford just *knows* he's right with no training, no research - just blurts out whatever idiocy crosses his teeny Fundy mind that day.

Like I said before - this blog is good for at least one FSTDT quote a week.

65. Derick, yes, someone did manage to mount a outboard motor on a van! You got me on that one! Intelligent designers can be very creative.

The ultimate point of that previous discussion was concerning nested hierarchies. Both you and Zachriel were schooled in the fact that there does not exist within nature a unique and objective nested hierarchy. While you were looking back did you happen to recall that you were going to do some research on the sea squirt? How's that research coming along?

--

Regarding this current discussion, as far as turtles climbing onto fence posts on their own accord I'm still waiting for your google search for that one. Your previous links were fun, but they failed the criteria I set.

66. Neal: "Derick, in your zeal to prove me wrong, you showed turtles climbing a chain link fence."

Neal, do you not know how Google works? I showed that a simple search finds examples of not just turtles climbing chain link fences, but bricks, screen doors, plastic boxes, and various other objects. And I wasn't trying to 'prove you wrong', I was pointing out the erroneous way in which you ascribe agency; the flaws in your thinking that keep you from investigating alternate explanations for a given phenomena.

"Not what I said. Read it again."

Ok then. Let's all read it again. What you said was:

Neal: "Neal Tedford: "If I see a turtle on top of a fence post while I'm canoeing down a river on a nice sunny day, I assume someone put it there, because I know the ability of turtles to climb."

Apparently you didn't know that turtles can climb fences. Obviously, if a turtle could climb over a fence, it could climb on to a fence. (and juuuuust in case your objection was over something stupid like whether the fencepost was made out of metal or wood, I think that if turtles can climb up slippery, wet vertical rock walls, concrete posts, and trees, I don't think it's out of the question that one could manage an old wooden post.

In other words, when you come across something unusual, instead of doing even the slightest amount of research or inquiry, you ascribe agency to it. And then, when it's pointed out to you that that's often a fallacious way of thinking, instead of conceding a simple point, you embarrass yourself by trying to squirm out of your own words.

Your original comment, and the subsequent (unsuccessful) attempt to save face are perfect illustrations of the creationist mindset. Keep 'em coming.

Oh, and by the way, if you see a turtle stuck on top of a piece of wood, climbing isn't the only explanation either.

67. Neal: Scott, flying squirrels have been observed to glide from tree to tree. They can go airborne for short periods, so why not assume that they could fly to great heights with eagles?

Neal,

it's unclear that we can have a reasonable discussion on this topic without first addressing a number of assumptions which form the foundation of your approach to problem solving.

In fact, it's not even clear that you're aware of these underlying assumptions or actually interested in solving problems in the first place.

For example, your question is actually quite vague since it's unclear what you mean by 'assume' and 'observed.' Perhaps you mean to ask: could one justify the conclusion that flying squirrels can fly with eagles based on a series of past observations alone?

If so, my response would be "No." This is due what is known as the problem of induction. Specifically, Hume showed us that inductive reasoning is not a justifiable means to gaining knowledge.

However, I'd note that this isn't just applicable to flying squirrels - it's applicable to all scientific claims, including that the earth exists in a heliocentric solar system, that objects follow laws of physics, etc. As such, the claim that we justify scientific conclusions based on inductive reassigning in science is actually a myth.

Instead, we justify conclusions based on explanations.

So, to return to your question, we lack an explanation as to how a flying squirrel would propel itself to the extent necessary to fly with eagles. Certainly, there could be some un-conceived explanation which might result in flying squirrels soaring to the heights of eagles on their own power in the future. Or it might have occurred in the past. But none exist at the moment.

In other words, we lack a conceived explanation for how this would occur.

Note that we haven't ruled this out with 100% certainty. Instead, our current best explanation as to why we haven't observed a flying squirrel fly to the the height of an eagle is that flying squirrels lack the necessarily propulsion and flight control to do so.

68. Neal Tedford: "Derick, yes, someone did manage to mount a outboard motor on a van! You got me on that one!"

Not that one Neal, but on essentially every example you put forth in that thread.

"Intelligent designers can be very creative."

Yes they can. They can mix and match parts, destroying the ability to make anything even remotely resembling an objective nested hierarchy, like I showed with the iPod line. Intelligent Designers are usually not constrained such that their designs would fall into a hierarchical pattern, and most often don't

"Both you and Zachriel were schooled in the fact that there does not exist within nature a unique and objective nested hierarchy"

Funny. Schooled by the guy who claimed for months that iPods could be organized into an objective, best fit nested hierarchy? Even when shown dozens of times (using pretty pictures and charts, no less) that they couldn't?
I don't remember ever claiming that a perfect nested hierarchy could be reconstructed for all organisms. Could you provide a link to where I said that? I do remember saying that there was an overall pattern.

"While you were looking back did you happen to recall that you were going to do some research on the sea squirt? How's that research coming along?"

I don't ever recall making any claims regarding sea squirts. Refresh my memory by posting a link.

And since we're trying to close out old conversations, I have been asking you about the names of those theologians for months now. Please answer that question, either by providing the names, or retracting the claim. (You don't even have to say "I was wrong when I made that statement," merely something like "That statement was unsubstantiated."

Or you can squirm and evade. Actually, I'm just as amused by your rationalizations for saying stupid things as I am when you say stupid things in the first place.

69. Derick, let's close out the discussion on the old earth theologians. After a closer look, I feel it is not substantiated that Christian theologians were "old earth" creationists. Some did not hold to a literal 24 hour 6 day creation, but that can't be taken by default to mean they were old earth creationists either. So, I'll retract that.

I take all views (old and young earth) with a grain of salt. It seems that many in both camps are convinced beyond a doubt. What I see, is that the Bible doesn't give a date of creation, so why force it into a dogma? The important thing for me is to see that the universe had a beginning and the evidence points to a Creator just as the Bible says.

Now its your turn. Since you insist on getting a bit silly on the turtle on the fence post illustration. Show me a video of a turtle climbing unto a fence post. While canoeing I've seen plenty of turtles on rocks, branches, fallen trees, logs, and stumps. Of course, a fence and post could be constructed to accommodate turtles, but my simple illustration was dealing with real life. I actually had a wooden horse fence in mind, but kept the description brief. But in true evolutionary form you must bet on incomplete evidence (snapping turtles climbing a chain link fence = climbing onto a fence post with a bit of imagination). One must imagine it, rather than see it in reality.

The evolutionary formula seems to be:

Incomplete evidence + imagination + theology = confirmation of evolution as a settled fact

Regarding sea squirts, you said, "The short answer is that I'm not an expert on sea squirts, so I'm still researching this. "

http://darwins-god.blogspot.com/2011/01/new-genes-putting-theory-before.html

70. Tedford the Idiot said...

Now its your turn. Since you insist on getting a bit silly on the turtle on the fence post illustration.

LOL! Keep squirming Tedford, it's hilarious! Simple fact is, you made a dumb claim (turtles can't climb) based on your own ignorance and had it come back to bite you on the butt. It's same combination of scientific ignorance and arrogance you exhibit in pretty much every one of your posts. Remember how you thought 'Mitochondrial Eve' data meant there was only one woman alive in the past?

The Tedford the Idiot formula seems to be

Woeful ignorance + ignoring scientific evidence + no reasoning ability + Fundy Creationist beliefs = GAWDDIDIT.

Show me a video of a turtle climbing unto a fence post.

Show us a video of GAWD poofing animal 'kinds' into existence. Show us a video of Noah's Flood.

71. Scott said, "Instead, our current best explanation as to why we haven't observed a flying squirrel fly to the the height of an eagle is that flying squirrels lack the necessarily propulsion and flight control to do so."

--

...and our current best explanation as to why we haven't observed origin of life or large change evolution is that natural processes alone lack the necessarily mechanisms to do so. Bottom line.

72. Tedford the Idiot said...

Scott said, "Instead, our current best explanation as to why we haven't observed a flying squirrel fly to the the height of an eagle is that flying squirrels lack the necessarily propulsion and flight control to do so."

--

...and our current best explanation as to why we haven't observed origin of life or large change evolution is that natural processes alone lack the necessarily mechanisms to do so.

No idiot. The reason why we haven't observed origin of life or large change evolution in real time is because such events take way longer, millions of years, for them to happen.

The mechanisms have been empirically observed to work, have been demonstrated to be sufficient, and there is 150+ years of positive evidence the same mechanisms have been working for the last 3 billion years.

Bottom line.

Real bottom line is you're a scientifically ignorant blowhard with a religious agenda to push. At least your relentless stupidity is entertaining.