MacNeill's criticism of Behe's new book
by BilboAt the risk of being a blog hog, I'm starting another post, this time on Allen MacNeill's objections to Behe's arguments in his book, The Edge of Evolution, which Dr. MacNeill posted on a previous post. If I have time, I'll post Eric's reply. If not, I'll get to it tomorrow, or someone else can paste it in.
#
Eric wrote:
"The critical question, which is also raised in Behe's new book, is whether there exists a smooth an unbroken path of viable intermediates that evolution might find between two forms."
This problem has been around in evolutionary biology for over 70 years. As I have pointed out elsewhere, the germ of this problem is contained in R. A. Fisher's Fundamental Theorem of Natural Selection, which Fisher derived and published in his 1930 book, The Genetical Theory of Natural Selection.
According to Fisher's theorem, "The rate of increase in fitness of any organism at any time is equal to its genetic variance in fitness at that time." What this essentially means is that the rate of evolution by natural selection in a given population is a direct function of the amount of variation present in that population; more variation"“faster natural selection, less variation"“slower natural selection.
However, natural selection has the effect of reducing variation in populations. Indeed, according to Fisher's mathematical analysis of the effects of selection, if natural selection is intense enough, it eliminates variation altogether (a condition referred to as "fixation").
What this means is that, in the absence of some process that generates new variation, natural selection is an automatically self-limiting process.
Sewall Wright realized this, and integrated it into his "adaptive landscape" model of fitness (the kind of "landscape" that Behe refers to in his new book). Wright pointed out that, if selection were not "tempered" by some other process, it would have the effect of virtually eliminating variation in populations, and that this would bring evolution by natural selection screeching to a halt. To be precise, it would reduce the rate of natural selection until it exactly equaled the rate of production of new genetic mutations (i.e. very, very slow, as such mutations happen extremely rarely).
Recognizing that evolution does not seem to have slowed to a stop (i.e. variation and change over time still exists within every known population of real organisms), some mechanism must account for the ability of subpopulations to move from one adaptive peak to another in an adaptive landscape, a process that Fisher's theoretical models seemed to imply was impossible.
The mechanism that Wright proposed was originally called the "Sewall Wright effect," but is now much more often referred to as genetic drift. According to Wright's "shifting balance" theory of evolution, genetic drift allows the allele frequencies in small populations to randomly drift off of their adaptive maxima, thereby making possible transitions to new adaptive peaks without having to "go downhill" (i.e. pass through a "trough" of maladaptation).
In other words, purely random processes happening in small, isolated populations can "solve" the paradox presented by Fisher's Fundamental Theorem. Wright's mechanism has been tested empirically for decades, and not only has it been shown to conform to the observed evidence, it is reliable enough for evolutionary biologists to calculate predicted allele frequencies as a function of population size and to use these to test hypotheses about rates of change in allele frequencies in natural populations.
So Behe's supposed "problem" for evolution by natural selection was identified and solved long before he was born, a fact about which he seems blissfully ignorant. Furthermore, the solution that Wright proposed to Fisher's paradox depends upon precisely those elements so abhorred by most IDers: random processes (including random mutations) in small populations, combined with natural selection. Please note that Wright's solution is not the standard ID strawman of RM + NS. Rather, it is a process by which random processes (i.e. non-directed changes in allele frequencies in small populations) interact with non-random processes (i.e. differential reproductive success), with the outcome being precisely the kind of evolutionary change that Behe asserts is not possible by those mechanisms.
Now, it is not entirely surprising that a biochemist wouldn't know some basic history and methodology of evolutionary biology; it isn't in his training. However, it does seem somewhat presumptuous for Behe to make arguments about processes that he clearly does not understand, and particularly missing basic principles that were worked out over half a century ago by two of the most brilliant mathematical biologists of all time. That is, of course, a risk that anyone takes when dabbling in a field outside their area of expertise, and that is why most people do so with the greatest reticence, lest they look like idiots.
This is also why Darwin and many other scientists make it a practice to vet their ideas with their opponents, rather than their supporters. This is because their opponents will generally have a much clearer idea of where the flaws in their arguments are, and will therefore allow them to correct those flaws before they present them to the public. I have done this with my own research, and found the comments and criticisms of my opponents much more valuable than praise from people who uncritically support my ideas.
Has Behe done this? Take a look at his acknowledgements (always the most revealing part of any public presentation). Has he run his ideas by people who oppose them, and who therefore could be counted on to find the flaws in his arguments? Or has he mostly relied upon the judgement of people whose expertise (or lack thereof) and opinions match his own? I believe the public record speaks for itself in that regard.
Comment by Allen_MacNeill "” June 15, 2007 @ 8:49 am
June 16th, 2007 at 5:48 pm
At the Discovery Institute event filmed by C-Span this past Wednesday, Behe discuss R.A. Fisher's work.
Fisher's work has not been ignored by ID proponents, most notably William Dembski, Walter ReMine, and Stanley Salthe. A discussion of Salthe's take on Fisher is here: University indoctrination program launched, but one professor sees the light.
Behe made passing reference to this observation by Salthe (without mentioning Salthe explicitly).
Comment by Salvador T. Cordova — June 16, 2007 @ 5:48 pm
June 16th, 2007 at 6:03 pm
Great work Bilbo, I think the ID community can really benefit from the wisdom of somebody with a reputation for academic rigour and great teaching. It's good that you have the courage to bring Allen's work to your reader's attention. I think most TT readers would enjoy reading Allen's blog "Evolution List" - just google for it.
What did you think of Allen's methodological criticism of the kind of peer review that the more recent texts are getting. It used to be that ID critics could claim that ID papers had no peer review; this fortunately is no longer the case, however it seems that the review is still rather limited.
Care to comment?
Comment by salimfadhley — June 16, 2007 @ 6:03 pm
June 16th, 2007 at 6:16 pm
Even among evolutionary biologists Fisher's Fundamental Theorem (FFT, not Fast Fourier Transform this time) has created a lot of confusion. Fisher tended to be rather obscure in his writings. Check out this wonderful paper by Alan Grafen to understand some of the controversy (and to learn some interesting details about the life and work of that remarkable man Fisher):
http://users.ox.ac.uk/~grafen/...
It seems that MacNeill doesn't fully get FFT either, or maybe he simplifies:
No it doesn't. For example, if there is overdominance (heterozygotes having higher fitness than homozygotes), there will be variation in equilibrium. It is also important to realize that Fisher's model makes a lot of simplifying assumptions, such as no mutation and constant selection coefficients. Relax these assumptions and we're in a different "ball park" (an alien expression to soccer-loving Europeans, but I like it).
Comment by Raevmo — June 16, 2007 @ 6:16 pm
June 16th, 2007 at 7:16 pm
Behe's comment on Fisher will hopefully be on C-Span (which unfortunately takes months). All this to say, I think Behe is aware of Fisher's work and had taken interest in it. Fisher, like Haldane, have the distint privilege of being equally admired by ID proponents and opponents…
Unfortunately, I can't recall from memory what Behe had to say specifically on Fisher's ideas. I guess we'll have to wait for the C-Span coverage to come out.
Comment by Salvador T. Cordova — June 16, 2007 @ 7:16 pm
June 16th, 2007 at 9:47 pm
Raevmo wrote:
"…if there is overdominance (heterozygotes having higher fitness than homozygotes), there will be variation in equilibrium."
Indeed, and as you suggested, I did simplify Fisher's theorem. In my analysis, I considered only the simplest version of the theorem, as it applies to single genes with only two alleles, having simple Mendelian dominance, and in which either the dominant allele or its recessive allele (when homozygous) are either more or less fit than the alternative. That is, I did not consider those cases in which these conditions are not met (such as heterozygote superiority). There are other cases like this, including balanced polymorphism, etc. (see any introductory evolution text for majors for a complete analysis).
However, even with respect to the simplest version of FFT, Behe seems unaware that Wright's adaptive landscape (AL) formalism does not match his misperception of it. Wright pointed out (in the 1930s, BTW) that changes in phenotype that result from selection can easily have the effect of changing the AL, for example by changing the environment in which other organisms are undergoing selection. This is simply another example of coevolution, which several authors (e.g. David Wolpert) have pointed out can do what W. Dembski alleges is impossible: produce genuinely new genetic (and therefore phenotypic) information.
Furthermore, no one (including Behe) has yet commented on Wright's solution to Fisher's paradox: that is, non-adaptive drift as a means of moving from one adaptive peak in an AL to another. As we now know (thanks to Kimura, Ohta, Crow, and others), the amount of non-adaptive (i.e. adaptively "neutral") evolution that is going on is immensely greater than the amount of adaptive evolution (i.e. that is the result of natural selection). Therefore, the kind of drift that Wright proposed as a solution to Fisher's paradox is the predominant process in evolution as far as we can tell.
What this means for ID, however, is pretty devastating. ID theory is entirely focussed on adaptations; that is, on the characteristics of organisms that appear to have some "purpose" in the life of organisms. What is becoming abundantly clear is that the vast majority of evolution that has been and is going on is non-adaptive, and therefore irrelevent to ID theory.
Indeed, this suggests that a major revision of evolutionary theory is currently taking place, a revision that relegates Darwinian evolution (i.e. evolution that produces adaptations via selection) to a relatively minor role compared with the essentially random processes that produce the overwhelming amount of "non-adaptive" genetic code embedded in the genomes of eukaryotes. This revolution in evolutionary theory is similar to the quantum revolution in physics; rather than confirming that the universe is deterministic (and therefore at least arguably "designed"), at its deepest levels there is only chaos.
The same is becoming increasingly obvious in evolutionary theory: rather than Darwinian natural selection producing most of what we see in the living world, the seemingly stable, seemingly adaptive world of biology is actually just the spindrift on a seething ocean of random change. Selection doesn't give us a pan-adaptationist biosphere; it is instead apparently the only thing that prevents the random processes that dominate at the level of molecules from tearing the biosphere apart.
Comment by Allen_MacNeill — June 16, 2007 @ 9:47 pm
June 17th, 2007 at 12:25 am
Allen –
The problem with the "landscape" idea, even allowing for genetic drift to move a little off of the plateua, is that fitness landscapes over large scales of computational algorithsm are ALWAYS impassable by the nature of computational systems. I have a paper that I'm working on discussing the problem in-depth, and if you're interested I'd love to have your input on my ideas.
To put it simply, just notice that the places where genetic algorithms are known to work are all in non-computationally universal systems. Feed-forward circuits, not feedback circuits, domain-specific languages, not universal languages (or universal languages, but solving problems that are simple enough that universality is not needed), that sort of thing. I explain in my paper some of the theoretical reasons why this is the case.
Comment by johnnyb — June 17, 2007 @ 12:25 am
June 17th, 2007 at 7:07 am
johnnyb wrote:
Hi Jon,
Does your paper account for the fact that in the biological world, an organism's fitness landscape will change over time? I can think of several things that will cause this to happen:
1. Climatic cycles.
2. Catastrophes (volcanic eruptions, asteroid impacts, etc.).
3. Changes in the other species in the ecosystem.
4. Changes in the species itself.
5. Changes of location (for mobile species).
It seems to me that any arguments against evolution based on fitness landscapes will have to take changing landscapes into account.
Comment by keiths — June 17, 2007 @ 7:07 am
June 17th, 2007 at 2:35 pm
"Has he run his ideas by people who oppose them" (?)
OK, this is getting ridiculous. When's the last time (any Darwinist) has submitted their work to a Creationist or IDT journal?
Comment by Ben Z — June 17, 2007 @ 2:35 pm
June 17th, 2007 at 2:37 pm
Further, did you run your objection by Dembski? I'm guessing not?
…what's that I predict you'll say? He wouldn't listen? …OK, now you know why Behe doesn't do it.
*takes a look at the acknowledgments of several Darwinian books I have in my room* …that's funny, I don't really see any from non-Darwinists.
Comment by Ben Z — June 17, 2007 @ 2:37 pm
June 17th, 2007 at 6:07 pm
I ran out of time on the computer when I started this topic. I wanted to paste Eric's response to Dr. MacNeill, which I'll do now. I enjoy good debates, and I think we might have one here, if we stay on topic.
eric Says:
Comment by eric "” June 15, 2007 @ 11:47 am
Comment by Bilbo — June 17, 2007 @ 6:07 pm
June 17th, 2007 at 6:24 pm
Allen (to Raevmo):
Who told you that?
Comment by Joy — June 17, 2007 @ 6:24 pm
June 17th, 2007 at 11:21 pm
Allen MacNeill wrote:
This is a quite extraordinary claim. And it is rationally unwarranted, for it is based on a misconception about quantum mechanics:
Allen MacNeill also wrote:
This is a very tendentious claim. Life on Earth has existed for billions of years all through the eons during which these 'random' forces of destruction have operated. Nothing humans have designed can even remotely match that track-record for a combination of complexity and resilience.
Also the word 'random' used to qualify the word 'processes' does not entail that the processes so qualified are undesigned, any more than does the word 'lottery', when used to describe certain processes. A state lottery, for instance, is a designed process. So too is Russian roulette, shuffling a pack of cards, and manufacturing a gaming console that relies on a mechanism for random number generation.
Genetic mutations are physical events presumably governed by physical laws. The basic physical order those laws describe is 'fine-tuned' for life to an astounding degree, and was so structured prior to and quite independently of life itself.
If this basic physical order was the outcome of a random process, then that's certainly news to me. Similarly, if the genetic code was generated by random processes, that's news to me.
It is instructively amusing to see naturalists swing back and forth between Infinite Chaos (a la Linde and Susskind) and Eternal Impersonally Determined Cosmic Self-Instantiating Equations/Laws in their desperate need to avoid positing a transcendent Mind as the ontologically and explanatorily ultimate reality.
I expose their trilemma here. Notice that both the Infinite Chaos idea and the Eternal Equation/Laws idea are as invisible and as essentially metaphysical as God. The only difference is that they don't lend themselves to religious experience as much as God does. At least, I know God has done that more than once.
Comment by stunney — June 17, 2007 @ 11:21 pm
June 17th, 2007 at 11:30 pm
keiths:
I don't believe a changing fitness landscape would affect it at all, because it deals more with internal self-consistency than anything else. If it could be affected by a changing fitness landscape, it would have to be a landscape that was completely, drastically (and drastically is in fact a requirement) modified from one generation to the next, to exactly match the organism's needs for the moment, and be exactly on time for each genetic change. The only place I know of something like that happening is in intelligently-designed factories
I.e. I don't think anyone would think that this was a biologically plausible scenario, unless someone physically constructed a factory to deliver the environmental changes exactly coordinated with the desired genetic changes step-by-step. Note that when I say that the changes must be drastic, I really mean it — it must be a single-generation coordinated environmental and genetic change, where the environment in the new generation had almost no continuity whatsoever with the environment in the previous generation.
Comment by johnnyb — June 17, 2007 @ 11:30 pm
June 18th, 2007 at 6:47 am
Allen MacNeill wrote:
"In other words, purely random processes happening in small, isolated populations can "solve" the paradox presented by Fisher's Fundamental Theorem. …….So Behe's supposed "problem" for evolution by natural selection was identified and solved long before he was born, a fact about which he seems blissfully ignorant."
It seems to me this doesn't address Behe's main point about the "incoherence" of random genetic change, regardless of how sophisticatedly the theory of the process is modeled. This incoherence (with respect to particular traits that are needed for organisms to complexify and adapt as they are observed to have done over geologic time) is shared by environmental change, luck, etc. Random genetic drift may drive a species off local maxima, and environmental and internal changes may change the AL with time, but this process required each successive adaptive peak (or continuous change from one to another state) be on the average directional. Not only directional, but accomplished with no chasms of fitness. This still needs to be explained - how existing exquisitely intricate biological structures and systems arose bit by bit from an almost entirely incoherent process.
This requires that all the steps in a complicated genetic structural change coding for a complex biological system in an existing organism were either in the right direction for a complex combination of fitness factors, or close to neutral (no "chasms" where the only way to get from A to C was through B where B was very deleterious). If they were close to neutral they still had to be directional in terms of being in the direction of achieving the final genetic structure. Does this sound like something happening from an almost totally random process?
These biological systems aren't random over all aspects of genetic structure and corresponding function to the organism - they are specifically tailored to the needs of the organism given its already existing basic structure and environment. For example the dolphin suite of special adaptations for aquatic life, especially the echolocation system. If all the evolutionary processes except selection are truly random then the probabilities given population sizes and available number of generations produce the "limits of evolution" described by Behe.
To cite the adaptive landscape stasis solution by Wright is a theoretical objection and may explain some microevolution, but to say it explains macroevolution seems to be an act of faith. Anyway, it avoids addressing Behe's main arguments from evidence, in particular with the malaria protozoan. It needs to be explained why all of these MET processes working with vast populations and tens of thousands of generations did not allow this organism to come up with really innovative structural/biochemical adaptations to overcome drugs (other than a couple of simple changes), and any to overcome sickle cell. Also why his extrapolation to human (and other animal) population genetics was invalid.
Comment by magnan — June 18, 2007 @ 6:47 am
June 18th, 2007 at 11:51 am
Hi, Allen_Macneill,
There, fixed that for you.
Comment by AnaxagorasRules — June 18, 2007 @ 11:51 am
June 18th, 2007 at 12:34 pm
How is that a problem? Isn't that what ends up happening with most species?
Why would there be a guarantee? If a species is marooned on an island of viability and the fitness landscape changes, what will happen?
What has happened, historically, to nearly every species that has ever existed?
Comment by JAM — June 18, 2007 @ 12:34 pm
June 18th, 2007 at 1:14 pm
Allen,
I have to mention, you came to mind when I saw Behe talking about Spandrels. You were the one who introduced me to Lewontin and Gould's writings on the Spandrels of St. Mark. That is still one of my favorite evolutionary essays.
Thank you for introducing it to me.
Salvador
Comment by Salvador T. Cordova — June 18, 2007 @ 1:14 pm
June 18th, 2007 at 3:10 pm
JAM wrote:
Is the answer to your question dependent in any significant way on a resolution of the species problem?
It is probably true that most of the things that have ever lived on Earth are now dead. But so what? Generalized physical immortality strikes me as incompatible with finite physical resources. It would also be a bad idea for other reasons, as I explain here and here.
But I am always 'all ears' when it comes to proposed specifications of a complete science for alternative worlds that would contain complex and intelligent life-forms and be demonstrably superior to the actual universe in some sense of 'superior'.
Do you happen to have one such, by any chance?
Comment by stunney — June 18, 2007 @ 3:10 pm
June 18th, 2007 at 3:18 pm
No.
Comment by JAM — June 18, 2007 @ 3:18 pm
June 18th, 2007 at 6:32 pm
JAM and stunney –
Just to point out — many of the estimates for the number of species that have gone extinct is based on Darwinian assumptions — i.e. that there must be innumerable transitional forms between each species we see in the fossil record. I'm still searching for a number of species that have gone extinct that we have verifiable evidence that they existed in the first place, but have not been able to find it.
Comment by johnnyb — June 18, 2007 @ 6:32 pm
June 18th, 2007 at 8:25 pm
Allan McNeill,
By "non-adaptive" drift you mean "purely random" don't you? So if Darwinism requires "purely random" mutations to search sequence space before an adoptive mutation can occur then you have made Behe's argument for him.
When you said that RM+NS was a strawman, I thought you were referring to the argument that Dawkins often makes that Darwinism is non-random. In fact, it appears you were referring to the fact that it is sometimes purely random with natural selection playing no role. I find that surprising.
Comment by Jehu — June 18, 2007 @ 8:25 pm
June 18th, 2007 at 10:32 pm
How can one come up with a numerical estimate if one starts with the assumption that there must be innumerable transitional forms?
What's your estimate?
And how does that counter my point–which is that getting stuck on an 'island' is predicted, not a problem?
Comment by JAM — June 18, 2007 @ 10:32 pm
June 19th, 2007 at 12:38 am
"How can one come up with a numerical estimate if one starts with the assumption that there must be innumerable transitional forms?"
You have to make assumptions somewhere. That is the assumption that Darwinists make.
"What's your estimate?"
I haven't the slightest clue. As I said, I'm looking for the data that says how many extinct species are in the fossil record. I'm a YEC, so my view is that the fossil record is a fairly complete snapshot of creatures at the time of the flood, excluding mammals (they float when drowned).
"And how does that counter my point"“which is that getting stuck on an 'island' is predicted, not a problem?"
I actually wasn't following your conversation, just noted that one of you mentioned something about the number of species that are extinct, and most people aren't aware of the assumptions that underlie those numbers.
Comment by johnnyb — June 19, 2007 @ 12:38 am
June 19th, 2007 at 1:27 am
I don't know, as Darwin once said, "the number of intermediate and transitional links, between all living and extinct species, must have been inconceivably great." I have seen estimates by Darwinists up to 4 billion extinct species.
A better estimate is one based on actual fossil evidence. Here is the famous quote from David Raup that puts that actual number of fossil species at 250,000.
"Well, we are now about 120 years after Darwin and the knowledge of the fossil record has been greatly expanded. We now have a quarter of a million fossil species but the situation hasn't changed much. The record of evolution is still surprisingly jerky and, ironically, we have even fewer examples of evolutionary transitions than we had in Darwin's time. By this I mean that some of the classic cases of darwinian change in the fossil record, such as the evolution of the horse in North America, have had to be discarded or modified as a result of more detailed information — what appeared to be a nice simple progression when relatively few data were available now appear to be much more complex and much less gradualistic. So Darwin's problem has not been alleviated in the last 120 years and we still have a record which does show change but one that can hardly be looked upon as the most reasonable consequence of natural selection."
David Raup, "Conflicts between Darwin and Paleontology", Field Museum of Natural History Bulletin Jan. 1979, Vol. 50 No. 1 p. 22-29
It is a problem if you want to explain the evolution of organisms through random mutation and natural selection.
Comment by Jehu — June 19, 2007 @ 1:27 am
June 19th, 2007 at 11:53 am
I didn't mention a number, Johnny. I pointed out that the existence of these islands isn't the deal-killer Eric is selling it as being. It's very common.
I would think that an actual Darwinist would come up with much lower numbers. You'd need loads of nonDarwinian evolution to get numbers that large.
Why would the number of fossils yield an estimate instead of a minimum? For example, we know that passenger pigeons existed in huge numbers, but are now extinct. Has anyone found a single fossil?
Well, then, it's a good thing that as a modern biologist, I don't limit my thinking to those mechanisms, isn't it?
Comment by JAM — June 19, 2007 @ 11:53 am
June 19th, 2007 at 12:47 pm
Jam
Darwin was not an "actual Darwinist"? Wow.
You would expect the fossils to be at the very least a randomized statistical sample of organisms that have existed. Therefore, we would expect see to constant gradualism in the fossil record consistent with an "inconceivably great" number of transitions. Instead the fossil record shows abrupt appearance followed by stasis.
Is the other mechanism you rely upon neutral drift?
Comment by Jehu — June 19, 2007 @ 12:47 pm
June 19th, 2007 at 2:04 pm
Don't be silly; reread your comment. The quote from Darwin was about forms between species, not the species themselves. You vaguely attributed the number 4 billion to "Darwinists."
Please don't attribute expectations to me that I don't have, Jehu. It's not nice. I expect the set of fossils to be highly biased and unrepresentative. You also didn't address my point that it represents a concrete minimum, not an estimate, because we can't have fossils from animals that didn't exist.
We? Speak for yourself.
One of them, particularly when it comes to speciation. Are you aware that Darwin said nothing about randomness or mutations, so the phrase "random mutation" can't credibly be offered as Darwinism?
Comment by JAM — June 19, 2007 @ 2:04 pm
June 19th, 2007 at 2:49 pm
Mutations were not known of when Darwin published his famous book. Darwin focused on visible changes. I don't know if he thought they were random in nature or not. In any case the term Darwinism has come to encompass more than "random mutations."
Comment by Bradford — June 19, 2007 @ 2:49 pm
June 19th, 2007 at 4:00 pm
The "inconceivably great" number Darwin gave is clearly larger than 4 billion. Nonetheless, the 4 billion number is a high estimate from George Gaylord Simpson. Are you going to try and claim that G.G. Simpson was not a Darwinist?
Sure, it is a concrete minimum, but the pattern and distribution is absolutely not consistent with gradualism.
So you believe that there are non-random mutations? Are you a Lamarckist?
Comment by Jehu — June 19, 2007 @ 4:00 pm
June 19th, 2007 at 4:17 pm
I'm an IDist who believes non-random mutations constituting adaptative responses (most notably in rapidly reproducing unicellular oganisms) could constitute evidence of design.
Comment by Bradford — June 19, 2007 @ 4:17 pm
June 19th, 2007 at 4:28 pm
Correct.
Do we now view them as "random in nature," or do we view them as "random with respect to fitness" If they are random in nature, how can mutational hotspots exist?
As for what Darwin thought, I suggest that you read his books. Do they even contain the word "random"
"In any case the term Darwinism has come to encompass more than "random mutations.""
Only when used as a polemic weapon.
But it wasn't in reference to species, Jehu.
He obviously wasn't, because if you read the Web page you linked to, you'll see that his opinions differed from Darwin's on important matters.
Where did I write anything about gradualism? I merely pointed out that getting "stuck on an island of viability" is commonplace.
Jehu: Is the other mechanism you rely upon neutral drift?
1) Yes, belief doesn't enter into it.
2) No, I'm a geneticist who knows about things like hotspots, not a Lamarckist. Are you aware that "random" in this context only relates to fitness?
3) Is this your idea of some brilliant NIGYYSOB maneuver? Are you familiar with the fallacy of false dichotomy?
Comment by JAM — June 19, 2007 @ 4:28 pm
June 19th, 2007 at 4:32 pm
Jehu:
Why is 4 billion clearly smaller than inconceivably great? Can you conceive of 4 billion? Look, unlike theology, science actually makes progress. What's the point of quoting an ancient estimate that probably has been improved upon a lot in the mean time? Dead give away projection of creationists, whose heroes' arguments haven't progressed much for a thousand years.
Comment by Raevmo — June 19, 2007 @ 4:32 pm
June 19th, 2007 at 4:47 pm
Darwin didn't use the terms 'random' or 'mutation'.
These are the equivalents from NDT to his terms 'natural' and 'variation'.
Certainly Darwin talked about 'nature' providing the 'variation' upon which Natural Selection was to act.
These variations, he admitted, were 'random with respect to fitness' at the very least, in that he discussed Natural Selection's action of culling the deleterious, preserving the beneficial, and ignoring the neutral. If the variations were not random with respect to fitness there would not be good, bad and neutral.
Comment by Pez — June 19, 2007 @ 4:47 pm
June 19th, 2007 at 4:59 pm
Bradford:
What do you mean exactly with non-random mutations, and how is this evidence of design? I know that scientists design non-random mutations, but I presume you mean non-random mutations somehow caused by Another Designer.
Comment by Raevmo — June 19, 2007 @ 4:59 pm
June 19th, 2007 at 5:42 pm
Pez:
"If the variations were not random with respect to fitness there would not be good, bad and neutral."
Sure there would. There could be biases one way or the other that were insufficient to eliminate the beneficial or deleterious classes.
Comment by JAM — June 19, 2007 @ 5:42 pm
June 19th, 2007 at 6:00 pm
Mutations that are specified with respect to environmental stimulii (as opposed to causes of mutations themselves) would be non-random. Mutations are nothing more than leakage through cellular repair systems. Adjustments of such systems can effect the rate of mutations and their location within genomes.
Comment by Bradford — June 19, 2007 @ 6:00 pm
June 19th, 2007 at 6:03 pm
It was in regard to transitional forms, Simpson's 4 billion reference was in regards to species. If we are talking about the fossil record the distinction makes no difference.
He was a Darwinist in any normal sense of the word.
And you raised the issue with the rhetorical question, "What has happened, historically, to nearly every species that has ever existed?" Such a question makes relative the issue of the number of expected transitional species versus the actual fossil record.
Why don't you just come out and explain your fancy high-powered theory of non-random evolution instead of making me guess?
Comment by Jehu — June 19, 2007 @ 6:03 pm
June 19th, 2007 at 6:08 pm
Bradford:
Yes indeed. That sounds very interesting, but where does the design come in?
Comment by Raevmo — June 19, 2007 @ 6:08 pm
June 19th, 2007 at 6:31 pm
That's easily approachable experimentally. How do you propose to test the predictions of this hypothesis?
And I second Raevmo's question: where's the design?
Comment by JAM — June 19, 2007 @ 6:31 pm
June 19th, 2007 at 6:41 pm
We were talking about fitness/viability islands, remember?
How could you accurately describe him as a "Darwinist" if he disagreed with Darwin on such a fundamental point? Can you point me to a scientific discipline in which the scientists within it refer to each other as followers of a person? Do they often use any personal labels at all? If you can't, why not just admit that it is a cheap polemic device?
As an illustration, here's a case in which labels are being used as a joke:
Mudher A, Lovestone S. (2002). Alzheimer's disease-do tauists and baptists finally shake hands? Trends Neurosci 25(1):22-6.
My question wasn't about transitional species (I love how you are still trying to pretend that the Darwin quote was about species instead of forms). It had to do with dead-end species, which are the norm.
What does the question, "Are you aware that "random" in this context only relates to fitness?" mean in this context?
Comment by JAM — June 19, 2007 @ 6:41 pm
June 19th, 2007 at 7:00 pm
You also raised the issue of the number of extinct species.
In normal usage a Darwinist is one who believes in the advancement of species by natural selection, therefore, Simpson is Darwinist.
Do you consider Stephen J. Gould to not be a Darwinist as well?
And how did you establish that dead end species are the norm?
This transitional form versus species is a red herring that is not material to the issue.
Who knows? If your theory of evolution is so powerful why don't just say what it is instead of hedging?
Comment by Jehu — June 19, 2007 @ 7:00 pm
June 19th, 2007 at 9:14 pm
It's been my view that a good place to look for design is a selection anomaly. Of course if every adaptation must be attributed the random NS designer then we can cease wasting each other's time. I was sent an e-mail by a friend some months ago referring to a study indicating an adaptation involving a unicellular organism and two distinct mutations encompassing specified codons. The probability of this occurring by chance were extremely remote. Selection is a directional indicator but the mutations it acts on are assumed to be random. Separate mutations occurring in association with each other and collectively yielding an adaptive response would be evidence of a non-random process and an underlying design mechanism.
Comment by Bradford — June 19, 2007 @ 9:14 pm
June 20th, 2007 at 10:20 am
I'm not sure what JAM has in mind by "deal-killer" but it doesn't seem to fit the context of my exchange with Allen_MacNeill.
To put it very simply (I hope not too simply), Allen was presenting the rate at which variations are produced as the key factor limiting evolution.
I responded that mere quantities and amounts of variation are an insufficient characterization. I pointed out that there is a more fundamental issue that limits evolution, i.e. whether there is a gradual, accessible path that undirected evolution can follow. One could be trapped on an island of viability.
Allen responded essentially that this is not a problem and that it was solved long ago. This response was the note copied to start the current thread. He also appears to criticize Behe for failing to recognize this.
He goes on to discuss why it's reasonable that Behe wouldn't understand that this is not a problem, etc., etc.
My response (copied by Bilbo) points out that Allen has misunderstood the question and the issue that was raised. I go on to point out that taxonomic islands of viability are real and recognized. Neither increased rate of variation nor "drift" gets one off such an island.
You are right about the fact that extinction is a normal result if your island sinks. I hope you recognize that this is a point of agreement. Your tone suggested you might have thought I would disagree. And BTW, thanks for your support and agreement regarding this issue of islands of viability as real limits to undirected evolutionary change.
To return to my original point, when considering change to a system, it is not enough to consider the speed at which one makes attempts. One also must consider the limits of the terrain created by the non-localized commitments in the present design of the system. In other words, how big is your island?
In general, we should not expect that localized undirected variations are effective for revising non-localized conventions of the system. As I mentioned in another thread:
To get a practical feel for the implications of change in a system, ask any experienced programmer about what it is like trying to change any large established body of code when the change involves a non-localized convention of the system.
Comment by eric — June 20, 2007 @ 10:20 am
June 20th, 2007 at 11:59 am
Their mere existence is the point. The ratio of extinct to extant species was just icing on the cake.
The term is only the norm for polemicists. It is not normally used by real scientists who deal in producing new, real evidence.
No. I wouldn't call anyone a Darwinist unless he believed in Darwin's writing over the evidence. Do you have any evidence to suggest that was the case for Gould?
Natural history screams it to us.
Then why did you change "form" to "species"
Comment by JAM — June 20, 2007 @ 11:59 am
June 20th, 2007 at 12:10 pm
That's fine, but the scientific method involves putting forth a clear hypothesis and testing its predictions.
I'd say that about those who are unwilling to advance and empirically test their own hypotheses, like Behe.
What organism, what gene, and what mutations? Why be coy?
Were sequential and recombination models assessed? This is another huge problem with Behe's book.
The randomness is ONLY with respect to fitness. Are you trying to claim that this is only an assumption, that it has never been tested?
I don't see why. I do see that when asked to make predictions from a design hypothesis, you proposed to test a prediction of modern evolutionary theory instead. What's the source of your reluctance? Why is Behe reluctant?
If the environments changed back and forth, are you sure that evolutionary theory could not produce a nonrandom mechanism? Wouldn't a far more intelligent design be a molecular switch between the wild type and mutant alleles?
Comment by JAM — June 20, 2007 @ 12:10 pm
June 20th, 2007 at 12:25 pm
And I'm pointing out that getting trapped is the norm over history.
But change in the fitness landscape can. Could that be why Behe assumed a static landscape?
I do. Do you also realize that as competing species go extinct, the islands representing the still-viable ones are relatively rising?
We only disagree in the implications.
They are limits to change in most species, but they are not universal. Again, this makes clear predictions: as species go extinct, the "sea level" of your landscape goes down, creating "land bridges" between the islands.
But one also must consider that the terrain changes over time. Would you agree that Behe's assumption of a static landscape is a huge problem with his book?
But the fact that you mentioned it doesn't make it true. We are discussing biology, not apologetics, so evidence is the gold standard.
Your proposal isn't practical at all, given that the nature of the two systems is completely different. To get a practical feel for this fundamental difference, ask any experienced programmer about whether he writes large chunks of code that are similar–but never identical–that have partially overlapping functions, not separate or fully overlapping.
This is the fundamental nature of biological complexity. It is amazing, but it bears no resemblance to designed systems.
Comment by JAM — June 20, 2007 @ 12:25 pm
June 20th, 2007 at 1:40 pm
JAM
This is a point you keep making but refuse to support with evidence. When I discuss the evidence for extinct species you squirm, play word games, and then complain that it has nothing to do with fitness/viability islands. It is really getting pretty lame.
Because the point is not that fundamental. Here is a quote from Gould himself.
"The modern theory of evolution does not require gradual change. It in fact, the operation of Darwinian processes should yield exactly what we see in the fossil record. It is gradualism that we must reject, not Darwinism "
Not off the top of my head but Darwinism has a special cultic quality to it not found in other fields. Check out this article by Gould's co-author Nile Eldriged called Confessions of a Darwinist.
Again, this is the point you refuse to support.
Because, they are interchangeable in this context. This is an example you being more interested in word games than supporting your argument with evidence.
Comment by Jehu — June 20, 2007 @ 1:40 pm
June 20th, 2007 at 5:06 pm
Here's a derivation of 23 - 46 as a value for the extinct/living ratio:
http://www.bio.vu.nl/thb/cours...
Here are lists of threatened and extinct species:
http://en.wikipedia.org/wiki/I...
http://www.answers.com/topic/l...
You haven't introduced any evidence, Jehu. I just did.
And the lame one is the one who offers no evidence.
It wasn't about Gould, it was about George Gaylord Simpson. Talk about lame!
It's used tongue-in-cheek. One case does not justify pretending that it is the norm.
No. Reread the quote YOU introduced. If the links (not "forms" as I had incorrectly written) are between species, the writer is drawing a clear distinction between links and species.
Comment by JAM — June 20, 2007 @ 5:06 pm
June 20th, 2007 at 5:43 pm
If environmental changes occurred with regularity a molecular switch would be a straightforward coping mechanism consistent with an evolutionary process. Any test of a design hypothesis (beyond an origins scenario) must involve modern evolutionary theory because it is not consistent with imputing teleology through built-in response mechanisms able to signal foresight. Answering Behe's IC with co-option requires pathways with enough specificity to test them.
I would have pointed to different biological systems than did Behe in order to focus on design. More basic underlying issues for an evolutionary process involve mechanisms able to generate an initial genome (an origins matter) and mechanisms able to maintain the integrity of a genome if a minimally functional one existed (both an origins and post origins matter). My focus is on the latter and the question raised is could a minimally functional genome exist without an inclusion of adaquate genomic repair mechanisms within it. Although not strictly necessary to make my point I'll advance the view that genomic repair mechanisms are irreducibly complex. Multiple repair pathways would be needed at the outset and each pathway would entail a set of enzymes.
Genomic integrity needs to be assured before an evolutionary process takes place. Genomic decay is an inevitable process arrested only by repair functions. Adaptation within an evolutionary process requires that the fixation of selected changes outpace genomic decay. But would this occur without repair pathways already in place? If not why would this dynamic not simultaneously exclude the possibility of repair pathways themselves becoming fixed?
Comment by Bradford — June 20, 2007 @ 5:43 pm
June 20th, 2007 at 6:52 pm
But just pointing isn't science. You and Behe have to test your own hypotheses. Why didn't you answer any of the questions I asked in the paragraphs you quoted?
Isn't that almost trivially easy to define and test in E. coli?
All of them? If not, which ones specifically in E. coli?
But just pointing isn't science! You have to test your own hypotheses.
Well, let's just test your hypothesis that adequate repair mechanisms are required. Simply define "adequate" for E. coli, and we should be able to put your hypothesis to the test in no time! (I hope you're not going to go circular here.)
Again, that's easy to test in the lab with E. coli. Are you game?
Good question–let's see if it does! Aren't you excited by the prospect of putting your hypothesis to the test and seeing it pass with flying colors?
Comment by JAM — June 20, 2007 @ 6:52 pm
June 20th, 2007 at 6:59 pm
I fail to see how a ratio of 23:46 extinct to living species supports your argument. But then you haven't ever really explained your position.
I introduced evidence a long time ago "“ that is that there are only about 250,000 fossil species.
No it is not used tongue-in-cheek, nor does it necessarily mean that you follow Darwin the person. Rather it means that you subscribe to Darwin's views on evolution through natural selection. Here is another case. LYNN MARGULIS: "I am a Darwinist"
Darwin said that an "infinite number" of "fine transitional forms" connect all past and present species. According to Darwin, to be considered a new species an organism only needs to differ from its former state in a uniform but "extremely slight degree." So whether a fossil is a link, transitional form, or distinct species is really trivial to the debate we are having here. Whether you define something as a transitional form or a species, it is still subject to the same fitness landscape.
Comment by Jehu — June 20, 2007 @ 6:59 pm
June 20th, 2007 at 7:05 pm
Yes, although a less complex organism would more realistically mimick natural history.
Comment by Bradford — June 20, 2007 @ 7:05 pm
June 20th, 2007 at 7:26 pm
So what are we waiting for?
Comment by JAM — June 20, 2007 @ 7:26 pm
June 20th, 2007 at 9:36 pm
There is an old and common idea that all we need is an environmental change to open up a land bridge for morphological change. However, this does not take into account the internal limitations of the organism as a complex information system that is less free to change.
The evidence says that over the course of time there has been a loss of morphological plasticity.
Here is Kawano's "diagram of macro evolution emphasizing the process of diminishing morphological plasticity as evolution proceeds from higher to lower taxa".
This fundamental limitation is not about change in the environment, and so it cannot be opened up by further changes in the environment. The loss of plasticity is internal to the organisms themselves.
Consequently, "the magnitude of evolutionary effects on morphology decreased with time".
In short, over time the outer limits of the islands of successful variation have decreased. The islands became smaller and smaller.
The nature of the two system is not "completely different". Quite the contrary (see also my next post).
In particular, for the purpose of comparison I am making (which is quite different from the question you seem to be asking about the two), the two are quite comparable. I am specifying …
"To get a practical feel for the implications of change in a system, …"
As the code base of a large complex information system expands, and especially as the non-localized conventions and internal interdependencies increase, this restricts the freedom to change. If you try to change a non-localized convention with a departure in any one place, you necessarily introduce inconsistency.
That is simply the nature of undirected change in a system with non-localized conventions and interdependencies. Designers work around this by making coordinated non-localized sets of changes that replace old conventions and interdependencies with new, revised ones.
Comment by eric — June 20, 2007 @ 9:36 pm
June 20th, 2007 at 9:41 pm
from http://research.microsoft.com/...
Comment by eric — June 20, 2007 @ 9:41 pm
June 20th, 2007 at 10:09 pm
Let's see. From the origin of the phylum Chordata.
Vertebrates to fish to reptiles to theropods to birds to penguins. Vertebrates to fish to reptiles to mammals to whales. Consider also that humans evolved in just the last few millions of years. These are incredible diversifications from primitive chordates.
Comment by Zachriel — June 20, 2007 @ 10:09 pm
June 21st, 2007 at 12:22 pm
Obviously, other factors enter into it. Are you trying to claim that the level of competition (both inter- and intraspecies) isn't constantly changing, and represents sea level in this metaphor?
But you utterly, completely misrepresented the nature of biological complexity with your computer programming analogy. If you can't offer a single instance of similar, nonidentical programming modules with only partially-overlapping functions, why not at least admit that your metaphor is flawed instead of moving the goalposts?
Isn't this type of complexity far more tolerant of change than designed complexity?
If the evidence says that, cite the evidence, Eric. Quotes aren't evidence.
His diagram isn't evidence either. If you know of evidence, cite it. If you don't, don't try to palm off opinion as evidence, OK?
As a biologist, am I supposed to be awed by your labeling of Kawano as a "Japanese biologist"
But if you're going to argue that it's like a computer program, you need to show me several programs that feature the defining theme of biological complexity: similar, nonidentical components that have partially-overlapping functions.
Truncating your own claim to delete the false part of it and pointing out that we use computers to study biology only emphasizes that your analogy has no basis in real evidence.
But speaking of computers, have you ever bothered to look at the sequence evidence? You can generate trees on the fly in seconds.
Do you predict that we will lose the origins of the phyla through computer analysis of the sequence evidence?
Assertion isn't argument. Your next post did nothing at all to rebut my point, which is that the partially-redundant nature of biological complexity makes it tolerant of change. For example, starting with your assumption, what proportion of mouse knockouts do you predict would be viable as homozygotes, with no apparent problems?
None of that has anything to do with the defining feature of biological complexity, does it? It's about homologous, yet nonidentical, components and assemblies with only partially overlapping functions.
Can you point to a single example of this in designed objects, outside of those that were designed specifically to model biological systems, of course?
Comment by JAM — June 21, 2007 @ 12:22 pm
June 21st, 2007 at 12:50 pm
Eric,
Would you mind explaining the evidence in this paper in terms of your hypothesis regarding resistance to change?
http://www.plosone.org/article...
Remember, quotes aren't evidence.
Comment by JAM — June 21, 2007 @ 12:50 pm
June 21st, 2007 at 5:36 pm
I haven't read Beh