Is that correct?

Allow me to give an illustration of what I mean.

Let’s say Gene X needs to change 6 nucleotides to form a new protein with a novel function that will create a selective advantage. Unfortunately for Gene X, until all 6 nucleotides are replaced there is zero selective advantage. So in a fixed fitness landscape, Gene X must change each of those nucleotides by a purely random walk. Behe would point out that the odds of this happening are prohibitive. But Chu-Carroll and company come and point out that the fitness landscape is dynamic, so a change in the enviroment could occur that allows a selective advantage for each of the needed six nucleotide changes in a gradual stepwise manner.

Is that really what Chu-Carrol and Company are arguing? Frankly, that seems like such an unpersuasive agrument that I suspect I am not understanding it properly.

Now THAT’S more like it!

Thank you, Magnan and Jaredl.

I particularly liked this from Magnan:

However many dimensions of different interacting fitness functions, and however this “landscape” changes with time for a species [i.e. M is looking at the dynamic case emphasised by Chu-Carroll, Patrick], for any particular reproductive fitness function the species can still be trapped at a local maximum, unable to get across the “valley” to the next, higher peak without an extremely improbable giant leap. The reason for this is that the physical genetic loci coding for different fitness functions or factors are generally uncorrelated with each other. Usually they are not even in the same gene. No matter how many other varied genetic changes affect the phenotype in varied ways, certain specific mutational or other genetic changes are needed to make the jump from phenotype structure A to elaborated structure B in time T as evidenced by the fossil record.

In short, if the biofunctrional states are sufficiently isolated, random walks cannot credibly move from one to another. That sufficiency of isolation comes in at oh, about 250 base pairs or so, up. As Meyer ands othershave long since documented, just to get tot he Cambrian life revolution, with dozens of new phyla and subphyla over maybe 10 MY on theusual projected chronology, you have to jump from genomes that may be on the order of 1 mb bp to say comparable to a modern arthropod at ~ 180 mn bp. Even if 90% of the increment is mere junk, that makes but little or no material difference tot he resulting isolation of functional states relative to random walks.

So, Behe has summarised a key point.

And, as for the issue of the dynamic nature of fitness landscapes, that still does not get around the issue of isolation in the configuration space that makes the lucky noise thesis incredible.

Jaredl then put the final knife-thrust in by going back tot he No Free Lunch work. Namely, either you don’t find the fit islands, or, you write in the information on how to migrate and where to find the new fit islands.

But of course, at popular or semipopular levels, misleading rhetoric is often effective – especially for those who don’t want to see their favoured view crumble into dust. Even at professional scientific levels, people are astonishingly prone to rhetorical traps. Not to mention, it is notorious that new paradigms triumph by generational replacement, not by persuasion.

GEM of TKI

Just try to program … time-varying/coevolving fitness landscapes and see if they produce solutions to interesting problems (i.e., produce specified complexity). You’ll find one of two things. Either you’ll get sludge because you didn’t adequately constrain how fitness landscapes vary with time in response to a changing environment, or you’ll get something interesting (specified complexity) because you carefully introduced constraints and thereby did design work that cannot be reduced to material mechanisms.

To be sure, fitness in biology varies with time. As organisms evolve and the environment changes, what the environment deems fit changes as well. But what exactly constrains the transition from one fitness landscape or function to the next? If there is no constraint, then we are in the position of Wolpert and Macready’s Theorem 2, with evolutionary algorithms proceeding independently of their progress to solution and thus unable to outperform blind search (which means that even with 3.5 billion years of evolution, it’s going to be vastly improbable that the evolutionary algorithm approaches a solution). Conveniently, [Darwinists don't] tell us what constrains the transitions. Presumably nature, unprogrammed and unguided, spontaneously gives rise to the right and needed transitions between successive fitness landscapes, thereby ensuring a form of complexity-increasing evolution. But that is precisely what needs to be explained.

[Dembski] show[s] [in No Free Lunch] that coevolving fitness landscapes are mathematically equivalent to evolution with respect to a fixed fitness landscape. The argument requires reconceptualizing the configuration space so that coevolving fitness, as it were, gets built into it. The upshot is that coevolution introduces no new mathematics and therefore no way out of the displacement problem. If displacement is a problem for evolution with respect to a single fitness landscape, then it remains a problem for coevolving fitness landscapes. (cite)

Chu-Carroll’s criticism has been dead for a few years.

Suppose species X is trapped a local minimum. The fitness function, being dynamic, can change; for instance X might migrate to a place with different weather conditions, hence a different fitness function. In the new fitness function, X is no longer trapped.

I found mostly prejudiced misinterpretations and invalid arguments, more than I can completely recount here. The Edge of Evolution is quite evidently directed at nonscientist readers and is simplified accordingly, unfortunately glossing over the fine points. So Chu pounces on every relatively simplified description of evolutionary theory as an indication of Behe’s supposed ignorance and stupidity. For instance, he claims that Behe says that mutations are always single point changes. This is absolutely ridiculous and of course deliberately insulting. Chu, look at Chapter 3 page 62 first paragraph. Another example: “…his (Behe’s) ignorance of any source of genetic diversity other than mutation.” Of course Behe is aware of other sources of variation. Recombination is supposed to be the major source other than mutation. Behe doesn’t mention recombination because mutation is still the major source of change to the genome, as admitted in many orthodox MET sources. Recombination mainly reshuffles alleles (different mutated versions of genes) during reproduction of sex cells in eukaryotic organisms.

Behe’s prime statistical example of the limits of Darwinian evolution with only random variation is the malaria parasite, and this is a protozoan eukaryote (plasmodium) in which meiotic recombination continually occurs. This example gives every advantage to random variations from all types of mutations and recombinational events in a huge population over millions of generations, but the limitations still applied.

Chu goes into a long diatribe over Behe’s use of the “fitness landscape” concept in his argument. It seems to me these criticisms are obfuscations and irrelevant to Behe’s thesis. However many dimensions of different interacting fitness functions, and however this “landscape” changes with time for a species, for any particular reproductive fitness function the species can still be trapped at a local maximum, unable to get across the “valley” to the next, higher peak without an extremely improbable giant leap. The reason for this is that the physical genetic loci coding for different fitness functions or factors are generally uncorrelated with each other. Usually they are not even in the same gene. No matter how many other varied genetic changes affect the phenotype in varied ways, certain specific mutational or other genetic changes are needed to make the jump from phenotype structure A to elaborated structure B in time T as evidenced by the fossil record. The probability of this occurring by accumulation of small random changes or by one giant (random) leap is a function of the total complexity of that particular genetic change, the likely presence of steps that are too deleterious to reproductive fitness to spread and fix in the population, and the number of generations. This is regardless of abstract models like the “fitness landscape”.

The malaria parasite drug resistance example (in addition to others) demonstrates these limitations in the living world, regardless of abstract models.

Chu then sets up a straw man and demolishes it by implying Behe doesn’t even account for the trillion or so malaria protozoa in each human individual with the disease, in estimating the total number of reproducing parasites subject to Darwinian evolution in the human population. Of course this is ridiculous – Behe clearly accounts for this in his calculations, as shown in numerous places in chapter 3.

After this travesty, Chu continues to use rhetoric rather than specific arguments and counterexamples, to somehow through any means destroy Behe’s malaria test case. He grudgingly admits some validity to Behe’s statistical estimates for the malaria parasite acquisition of chloroquine resistance, but claims the malaria example is still an “artificially inflated probability number based on the biochemistry of one specific organism”. He vaguely asserts without substantiation that for an organism like malaria these numbers just aren’t “compelling”. I guess we are supposed to take this on faith in his wisdom. He has plenty of rant and bluster, but doesn’t show any specific valid way these numbers were inflated, and he doesn’t show specifically why Behe’s application of these results to human population genetics is invalid. This did, however, make me wonder if this extrapolation was perhaps simplistic. The only factor I could identify that might be questionable in relating the basic population genetics of the two organisms in this way was genome mutation rate per individual per generation. This is fairly low for unicellular organisms with very short lifespans and generation times, but higher animals (metazoans) accumulate mutations in their germ (sex) cells over a much longer time for each individual, so their mutation rate per generation is much higher. To try to correct for this I found some published mutation rate estimates, which indicated that the ratio is a factor of about 10,000. I tried correcting for this and Behe’s numerical argument was not significantly affected. It’s in the noise compared to the other factors. Behe didn’t mention this aspect probably because it is beyond the detail level of the presentation.

Chu doesn’t even try to address the other related points made in the book, such as in chapter 7 on the failure of the malaria parasite over human history to have evolved any new cellular protein-protein interactions (binding sites).

Chu also makes the usual hand-waving general claim that the chances of producing any particular biological change is admittedly extremely small, but that the chances of producing at least something or anything adaptive is very high. As if this really explains anything. So if it looks like you were chosen by design it really is only the end of a long chance winnowing process. This is just a rhetorical ploy and carefully avoids trying to apply it to explain any particular evolutionary development sequence.

___________

Fritz Ward (who unlike some of the “reviewers” on the other side seems to have read the book . . .):

In this book Behe strikes off in a new direction from his previous work, ‘Darwin’s Black Box.’ Rather than simply explore cellular mechanisms that seem unlikely to arise from chance, Behe instead considers all the areas where evolution seems to function very well. For example, the rise of resistance among certain diseases, notably malaria, to synthetic drugs. Remarkable evolutionary pressures are at work in the struggle between humans and deadly pathogens. Humans who develop an immunity to maleria have a strong evolutionary advantage over those who don’t. Similarly, protozoan parasites which can avoid the drugs we use to combat them also have an evolutionary advantage. Indeed, this is common knowledge among all biologists and most of the literate public. Germ resistance of all kinds to drug treatments is the star example of evolution at work.

But what is not so commonly known is that random mutation has severe limits in how effectively it can cope with evolutionary pressure. Indeed, what Behe demonstrates in precise detail is that evolutionary mechanisms are for the most part destructive: a part of the DNA stand is destroyed or replaced with a less efficient coding and the result is a weaker organism, though one which can survive the “trench warfare” of survival with hostile organisms. Thus, for example, humans have developed sickle cell anemia to cope with malaria. This is hardly beneficial, in and of itself, but compared to malarial death, it is a very helpful mutation. Similarly, malaria can rapidly evolve resistance to some drugs, slowly to others (more changes are required, and hence far fewer resistant copies of the cell are likely) but the mutated genes that come from this battle for survival are not optimal. Indeed, like sickle cell anemia, they rapidly die out of the malarial population if not subjected to the pressure of deadly (for the parasite) toxins in the form of antimalarial drugs.

So, while malaria (and several other cases Behe examines) suggests the efficacy of random mutation, it also suggests limits to just how much it can accomplish. Indeed, Behe finds that even two or three simultaneous random changes in DNA sequencing is exceedingly unlikely, and more just about impossible. This is very important because it suggests real limits to the amount of random mutation that could happen among higher mammals. People mistakenly believe that time is the most important factor in allowing for evolutionary change but as Behe demonstrates, population, not time, is what determines successful mutations. Malaria, and even moreso HIV are extraordinarily effective at utilizing evolution. There are a lot of such organisms and they reproduce quickly. Humans, and indeed, all vertebrate and most invertibrate animals, do not. Even given the entire history of life on the planet, it is extremely unlikely that the random mutation proposition of evolution could account for a significant amount of the diversity we witness in the world around us.

____________

Compare 2 – 3 as an effective upper limit on mutations on a per stage basis, with the number of DNA shifts to get new cell or organ or body plan level functionality, and the 250-base pair limit on the Dembski type upper bound.

Design looks very serious as a contender on the merits — as opposed to the politics.

Okay, over to the readers of the book for their thoughts . . .

GEM of TKI

__________

The book’s subtitle speaks of the “limits of Darwinism.” Are you saying that Darwin’s theory is completely wrong?

Not at all. It is an excellent explanation for some features of life, but it has sharp limits. Darwin’s theory is an amalgam of several concepts: 1) random mutation, 2) natural selection, and 3) common descent. Common descent and natural selection are very well-supported. Random mutation isn’t. Random mutation is severely constrained. So the process which produced the elegant structures of life could not have been random.
__________

GEM of TKI

__________

How does intelligent design differ from the prevailing Darwinist view of evolution?

To a surprising extent prevailing evolutionary theory and intelligent design are harmonious. Both agree that the universe and life unfolded over vast ages; both agree that species could follow species in the common descent of life. They differ solely in the overriding role Darwinism ascribes to randomness. Intelligent design says that, while randomness does exist, its role in explaining the unfolding of life is quite limited.

How does intelligent design differ from creationism? What do you say to critics who charge that it is merely “creationism in disguise”?

Intelligent design theory is to creationism as the Big Bang theory is to the book of Genesis. Although both intelligent design and the Big Bang may be reminiscent of some religious ideas about the universe and life, they are both grounded on the empirical study of nature, not on holy books. The phrase “Let there be light” may be evocative of the Big Bang, but the Big Bang is science, not scripture. Intelligent design may be compatible with some religious concepts, but the astounding intricacy of cellular molecular machinery is hard scientific data . . . .

How does your view of intelligent design in biology fit with the findings and theories of cosmology and physics?

The conclusion of intelligent design in biology fits very well with unexpected results in the past few decades from physics and astronomy, which show that the universe, its laws, physical constants, and many details, are “fine-tuned” for life on earth. For example, if the charge on the electron or the properties of water were much different, life as we know it would be precluded. Biology has now discovered that the fine tuning of the universe for life actually extends into life. The term “consilience” denotes the situation where results from several scientific areas point in the same direction, reinforcing our confidence that the conclusion is correct. Biology has attained consilience with results from cosmology and physics . . . .

Why do you think there is such resistance within the scientific community to the idea of intelligent design?

Scientists are trained to think of the universe as a self-contained, self- explanatory system. Unexpected findings that go against that supposition can be disconcerting. When it was first proposed, the idea that the universe had a beginning in a big bang was strongly resisted by some scientists, because it pointed to a reality outside of the universe. Intelligent design of biology evokes even stronger reactions, perhaps because it challenges the supposition of a self-contained universe even more strongly.

____________

We need more like this, with dig-in details.

GEM of TKI

I did a web search on the title, and turned up this interview from Simon and Schuster’s book intro page.

Telling excerpts:

__________

What do you believe Darwinian evolutionary processes can actually do?

The Edge of Evolution asks the sober question, what is it reasonable to think Darwinian evolutionary processes can actually do? Unprecedented genetic data on humans and our microbial parasites (malaria, HIV, E. coli) now allow us to answer that question with some precision. The astonishing result is that, even under intense selective pressure, and given an astronomical number of opportunities, random mutation and natural selection yield only trivial, mostly degenerating changes. The bottom line: the major events that produced life on earth were not driven by random mutations . . . .

How does the book evolve from the failure of randomness to the conclusion of intelligent design? Aren’t there possible unintelligent evolutionary explanations other than Darwinism?

The new genetic results on humans and our parasites tell against not only Darwin’s theory, but against any unintelligent process. In their reciprocal evolutionary struggle, human and parasitic genomes could have been altered in nature by whatever unintelligent mechanism had the ability to help. Yet virtually nothing did. Because the categories of “intelligent” and “unintelligent” processes are mutually exclusive and exhaustive, ruling out unintelligent processes necessarily implicates intelligence.

What evidence speaks most clearly to the role of intelligent design in biology?

The elegance of the foundation of life — the cell. Charles Darwin and his contemporaries supposed the cell was a “simple globule of protoplasm,” a microscopic piece of Jell-O. They were wrong. Modern science reveals the cell is a sophisticated, automated, nano-scale factory. For example, the journal Nature marvels, “The cell’s macromolecular machines contain dozens or even hundreds of components. But unlike man made machines, which are built on assembly lines, these cellular machines assemble spontaneously from their …components. It is as though cars could be manufactured by merely tumbling their parts onto the factory floor.” . . . .

One criticism of ID has been that it makes no predictions, and thus is unscientific. Does The Edge of Evolution address this?

The Edge of Evolution is almost entirely concerned with the major, opposing predictions of Darwinism and ID. The most essential prediction of Darwinism is that, given an astronomical number of chances, unintelligent processes can make seemingly-designed systems, ones of the complexity of those found in the cell. ID specifically denies this, predicting that in the absence of intelligent input no such systems would develop. So Darwinism and ID make clear, opposite predictions of what we should find when we examine genetic results from a stupendous number of organisms that are under relentless pressure from natural selection. The recent genetic results are a stringent test. The results: 1) Darwinism’s prediction is falsified; 2) Design’s prediction is confirmed . . .

____________

Sounds worth the read . . .

GEM of TKI