Thursday, February 12, 2009

Dawkins on Chance

 
I know I'm going to be accused of beating a dead horse but as Emile Zucherkandl and Linus Pauling said in 1965 ...
Some beating of dead horses may be ethical, when here and there they display unexpected twitches that look like life.
It is absolutely safe to say that if you meet somebody who claims not to believe in evolution, that person is ignorant, stupid or insane (or wicked, but I’d rather not consider that).

Richard Dawkins
Richard Dawkins has reviewed Jerry Coyne's new book Why Evolution is True in The Times Literary Supplement. The text of the review is posted on RichardDawkins.net [Heat the Hornet].

As you might have guessed, when an adaptationist reviews a book by a fellow adapationist you can expect heaps of praise. Dawkins does not disappoint.

One particular claim caught my eye since Dawkins has made it in the past. I know for a fact that others have pointed out to Dawkins the flaws in this claim. Here's what he says,
Coyne is right to identify the most widespread misunderstanding about Darwinism as the idea that, in evolution, “everything happens by chance”. This common claim is flat wrong – obviously wrong, transparently wrong, even to the meanest intelligence (a phrase that has me actively restraining myself). If evolution worked by chance, it obviously couldn’t work at all.
It's true that to say everything happens by chance is wrong. However, it is not true to say that, "If evolution worked by chance, it obviously couldn’t work at all."

Here's a quotation from the most popular textbook on evolution, Evolution by Douglas J. Futuyma. It's in Chapter 10—a chapter titled Random Genetic Drift: Evolution at Random.

Almost all factors are affected simultaneously by both chance (unpredictable) and nonrandom, or deterministic (predictable), factors.... So it is with evolution. As we will see in the next chapter, natural selection is a deterministic, nonrandom process. But at the same time, there are important random processes in evolution, including mutation and random fluctuations in the frequencies of alleles or haplotypes: the process of random genetic drift.

Genetic drift and natural selection are the two most important causes of allele substitution—that is of evolutionary change—in populations
Futuyma closes the chapter with a summary of the important points. The first two are ...
  1. The frequencies of alleles that differ little or not at all in their effect on organisms' fitness (neutral alleles) fluctuate at random. This process, called random genetic drift, reduces genetic variation and leads eventually to the random fixation of one allele and the loss of the other., unless it is countered by other processes, such as gene flow or mutation.
  2. Different alleles are fixed by chance in different populations.
Thus, according to the textbook, evolution by chance occurs in spite of the fact that Dawkins says, "If evolution worked by chance, it obviously couldn’t work at all."

Now, the only way to reconcile his statement is to assume that either Dawkins doesn't know about random genetic drift, or he uses a non-standard definition of "evolution" (or he is wicked, but I’d rather not consider that ).

I know that Dawkins has written about random genetic drift so I have to assume that he uses a definition of the word "evolution" that excludes it. Since he is using a non-standard definition of evolution, I think it would be wise of him to make this clear in his writing. He should have written something like ...
In my opinion, the only valid mechanism of evolution is evolution by natural selection and that is definitely not a chance process. If natural selection worked by chance it obviously couldn't work at all.


28 comments :

  1. Could not genetic drift be affected by sexual selection? Is there a random genetic drift and a non-random genetic drift?

    ReplyDelete
  2. Evolutionary biologists, including Larry and Richard Dawkins, seem to accept as a matter of principle that "natural selection" is NOT a "chance" process, while mutation, drift, and apparently everything else in biology IS a "chance" process. I don't get it.

    Can someone provide an abstract definition of "chance" that we can substitute into these arguments? For instance, if we substitute "chance" with "bad" or "stuff I don't like", then it appears that Larry and Richard like different stuff.

    ReplyDelete
  3. @NAL

    No, sexual selection is a form of natural selection, because genotypes differ in reproductive fitness. Genetic drift occurs when allele frequences change irrespective of fitness differences.

    @Arlin

    Chance processes are processes with an unpredictable outcome. The outcome of natural selection is dependent on fitness differences, and therefore predictable.

    ReplyDelete
  4. The problem is this reductive definition:

    "Allele substitution = evolutionary change"

    Evolutionary change does indeed depend, in part, on allele substitution, but it is not allele substitution. Allele substitution is an important part of the process of evolutionary change, that's all. The reduction of evolutionary change to allele substitution is, at best, distorting.

    ReplyDelete
  5. The reduction of evolutionary change to allele substitution is, at best, distorting.

    On the contrary, defining evolutionary change in this way makes evolution a nicely transparant concept. But I'll concede it is not complete (not including polyploidisation, etc).
    Perhaps you can supply us with an alternative definition?

    ReplyDelete
  6. Corneel, you offered that "chance processes are processes with an unpredictable outcome."

    OK, lets go through 2 implications of this definition.

    First, recall the starting observation that all evolutionary biologists agree that selection is a "non-chance process". From your definition, this means that all evolutionary biologists believe that the outcome of selection is not unpredictable. They believe that they can predict the outcome of selection. Do you believe that this is true?

    Second, lets suppose that I set up an evolution experiment in which some lab organisms are exposed them to new conditions. This kind of thing has been done many times over the years, and I don't recall anyone ever predicting the outcome before-hand. Therefore, according to your definition, evolution is a "chance process", right?

    ReplyDelete
  7. If genetic drift is a stochastic process, and genetic drift requires sex, and sex requires sexual selection, and sexual selection is a subset of natural selection, then mustn't natural selection also be a stochastic process?

    ReplyDelete
  8. I think Dawkins is just using a different definition of chance.

    A coin flip is pure chance, there's no real bias to either tails or heads, but for genetic drift it's not pure chance as real harmful stuff tends not to propagate.

    ReplyDelete
  9. Here we go again :)

    A rather narrow and excessively pedantic (and IMO, disingenuous) objection, particularly since it is quite clear that the meaning of “evolution” that he and Coyne are addressing is the one that all but 0.001% of the world’s population has in mind when they ponder the issue. It’s the same meaning of the word as it is used in high school (and lower level) biology text books (and which is about as far as it is appropriate to go at that level). It’s the concept that species have changed over time, thereby giving rise to new species, along with all of the possibly exciting or disturbing corollaries that are so implied. It’s the particular meaning of the word that refers to the concept that creationists find so objectionable.

    This definition of the word has a long history and it is this definition that concisely framed the issue that Darwin’s theory was addressing. It’s use in this sense predates our understanding of molecular genetics, and (thus) the more general and inclusive “standard” definition that you prefer it to have.

    It is not uncommon for a particular word to have somewhat different meanings depending on the context. “Theory” and “Fact”, particularly in the context of discussing evolution at the level of social acceptance and cultural impact, are a couple of other ones. Often a writer does not explicitly denote the definition they are using, and so the reader must evaluate the context to ensure they are not confused.

    In this regard, I am surprised that you failed to quickly discerned that Dawkins was using the term in the sense I described above, not the sense that you would like it to always be used. The clues that tipped me off were that the essay was explicitly about creationists misrepresenting the theory of evolution (said essay also including sections devoted to both the meaning of “fact” and “theory”), in a review of a book about the strength of evolutionary theory intended for a broad layman audience. I also am familiar with how Dawkins uses the term from reading his other writings, and I understand that he is often explicit when he uses intends an alternate meaning (e.g., in p 450 of “Ancestors Tale”, where he uses the term “evolution at the molecular level” to indicate he recognizes that most evolution at this level is random). So, unlike you, I was not confused by his message.

    If the reader understands the how the term is being used (and as long such use is legitimate, as it clearly is here), it is deceitful to argue against his points by measuring them against a different definition. (Unfortunately, that’s a tactic that creationists employ often: see Dawkins discussion of the meaning of “fact” and “theory” in his essay).

    When addressing the problem of acceptance of evolutionary theory in society at large, in a forum intended to reach non-specialists, it is not a defect to use the particular definition of a term as it is understood by 99.999% of the worlds population. When a 9th grade high school student (or his parent, or minister) asks “If we evolved from monkeys, why are there still monkeys?”, he is asking a very basic (but confused) question centered on the phenotypic differences that are apparent to a lay observer. It helps not at all to include in your answer to this question the subtleties of neutral changes in junk DNA, because those sorts of differences are the farthest thing from the questioner’s mind (and any discussion of that is more likely to result in someone who wants to reject evolution to seize upon it as evidence that confirms their misconception of how evolution works).

    ReplyDelete
  10. Now, the only way to reconcile his statement is to assume that either Dawkins doesn't know about random genetic drift, or he uses a non-standard definition of "evolution"

    Or that you are attacking a staw man of Dawkins (or perhaps that you think Futuyma is saying more than he really is). I don't see Dawkins denying that random processes occur, only that, overall, biological evolution is non-random. Nor do I see Futuyma saying anything that contradicts Dawkins -- he says there are random processes in evolution, but also that evolution, overall, is non-random -- exactly what Dawkins is saying.

    Are you sure the donkey isn't attached to strings, and the "twitching" isn't just you tugging on them?

    ReplyDelete
  11. @NAL

    If genetic drift is a stochastic process-Yes

    and genetic drift requires sex-YES

    sex requires sexual selection-NO, sexual selection presumes NONRANDOM mating, when individuals breed without regard to genotype mating is RANDOM

    and sexual selection is a subset of natural selection, then mustn't natural selection also be a stochastic process?-NOPE but it does work on averages, i.e. not every individual in a population with a "high fitness" genotype will survive and reproduce but on average some genotypes will do better than others because of some real advantage

    ReplyDelete
  12. So, for genetic drift, random mating is assumed?

    ReplyDelete
  13. They believe that they can predict the outcome of selection. Do you believe that this is true?

    Yes, provided you know the fitnesses of the genotypes (and I admit this is the tricky part ;-)

    Therefore, according to your definition, evolution is a "chance process", right?

    No, this is a practical limitation, not a real one. It is common practice to set up replicated populations in the experiments that you describe. Usually, they all show a similar response to selection, showing that the outcome is not determined by chance.

    ReplyDelete
  14. Divalent says.

    A rather narrow and excessively pedantic (and IMO, disingenuous) objection, particularly since it is quite clear that the meaning of “evolution” that he and Coyne are addressing is the one that all but 0.001% of the world’s population has in mind when they ponder the issue. It’s the same meaning of the word as it is used in high school (and lower level) biology text books (and which is about as far as it is appropriate to go at that level).

    Here's part of what Grade 12 students in Ontario are expected to know. You can see the complete biology curriculum at Evolution in Ontario Schools

    Understanding Basic Concepts
    By the end of this course, students will:
    – define the concept of speciation and explain the mechanisms of speciation;
    – describe, and put in historical and cultural context, some scientists’ contributions that have changed evolutionary concepts (e.g., describe the contributions – and the prevailing beliefs of their time – of Lyell, Malthus, Lamarck,Darwin, and Gould and Eldridge);
    – analyse evolutionary mechanisms (e.g., natural selection, sexual selection, genetic variation, genetic drift, artificial selection, biotechnology) and their effects on biodiversity and extinction (e.g., describe examples that illustrate current theories of evolution, such as the darkening over time, in polluted areas, of the pigment of the peppered moth, an example of industrial melanism);
    – explain, using examples, the process of adaptation of individual organisms to their environment (e.g., explain the significance of a short life cycle in the development of antibiotic-resistant bacteria populations).
    – formulate and weigh hypotheses that reflect the various perspectives that have influenced the development of the theory of evolution (e.g., apply different theoretical models for interpreting evidence).


    Divalent continues,

    It’s the concept that species have changed over time, thereby giving rise to new species, along with all of the possibly exciting or disturbing corollaries that are so implied.

    Much of that change is due to random genetic drift. That includes speciation.

    I guess you don't understand evolution as well as you think you do.

    I believe there are lots of example of species changing over time due to random genetic drift. Do you deny that this can happen?

    I surprised you think Richard Dawkins would deliberately use a bad high school level definition of evolution instead of what he thinks might be the scientific one. Do you really believe he's doing that?

    ReplyDelete
  15. Jay D says,

    Or that you are attacking a staw man of Dawkins (or perhaps that you think Futuyma is saying more than he really is). I don't see Dawkins denying that random processes occur, only that, overall, biological evolution is non-random. Nor do I see Futuyma saying anything that contradicts Dawkins -- he says there are random processes in evolution, but also that evolution, overall, is non-random -- exactly what Dawkins is saying.

    That's a very interesting way of looking at it.

    So, given that there are several different mechanisms of evolution, according to your logic it would be OK to say that evolution, overall, is not due to natural selection. Right?

    In other words, evolution overall is not entirely nonrandom.

    Do you see where that kind of logic leads?

    ReplyDelete
  16. NAL said...
    If genetic drift is a stochastic process, and genetic drift requires sex, and sex requires sexual selection, and sexual selection is a subset of natural selection, then mustn't natural selection also be a stochastic process?

    Genetic drift does not require sex, and sex does not require sexual selection. Neither does genetic drift require random mating (often it occurs with inbreeding).

    The main distinction with natural selection is that alleles do not increase or decrease in frequency because of their effects on the phenotype. With natural selection, they do. If you know which phenotype is the most adaptive, you know what´s going to happen.

    Aluchko said...
    A coin flip is pure chance, there's no real bias to either tails or heads, but for genetic drift it's not pure chance as real harmful stuff tends not to propagate.

    If real harmful stuff fails to propagate, it is called natural selection, not genetic drift.

    ReplyDelete
  17. So, given that there are several different mechanisms of evolution, according to your logic it would be OK to say that evolution, overall, is not due to natural selection. Right?

    In other words, evolution overall is not entirely nonrandom.

    Do you see where that kind of logic leads?


    Reductionist logic would seem to say that if all of reality ultimately reduces to completely random quantum events, it must follow that only random chaos exists at the macroscopic level. But... somehow that's not observed.

    ReplyDelete
  18. Your kidding, right Larry?

    From your article on Biology in Ontario schools: "Evolution is only covered specifically in Grade 12 Biology ... [T]his course is only taken by a small percentage of students in Ontario high schools."

    So a 12th grade biology course, which has an 11th grade biology course (which doesn't cover evolution) as a prerequisite, and which you lament is only taken by a small percentage of students in Ontario, is relevant how?

    Are you really suggesting that the majority of people think of your definition when they hear the word "evolution"? Even in Canada? And when there is discussion/debate over how evolution should be taught in the schools, that even a miniscule portion of the concern of either side regards random drift of mutations in junk DNA?

    If so, might I politely suggest you've had one too many meals in the faculty dining room. You don't need to go fishing into published teaching standards (as if everything in there is retained by students when they move on). Just open your eyes at what is is out there. If you still don't see it, then perhaps you need to go "creationist-slumming" with Jason Rosenhouse, have a beer in a working class bar, or maybe even listen to a little AM talk radio. Spend some time with people in a rural town who wish you to have a "Blest Day". Ask them what evolution means to them, and I'll fly over to Toronto and buy you a nice lunch at your faculty dinning room if even one of them so much as hints that drift of neutral mutations is part of their concept of evolution.

    BTW, in my state, the basic biology course (not AP Biology) is (unfortunately) taught at the 9th grade level (which is common in the US) and as written the standards related to evolution are focused almost exclusively on natural selection. (OTOH, unlike the Canadian standards, at least ours also brings evolution into the discussion of classification.)

    Thus, my guess is that a greater fraction of the population gets exposed to evolution in HS compared to canada, but at a much younger age (there is a huge difference between 9th and 12th graders) and at a simplifed level.

    IMO, neither my state nor Ontario's brings it up often enough; it should be there at the beginning and touched on throughout the year during classification, genetics, ecology, anatomy/morphology, behavior. (Imagine teaching geology where plate tectonics is merely a 2 week section in the middle of the year.)

    ReplyDelete
  19. Actually, in looking at the Ontario standards for teaching evolution, I have to agree that it is far superior, as it would appear to integrate the topic with just about every section of a year-long course. Which is as it should be.

    But unfortunately, in many HS text books designed for 9th graders in the US, and in many state standards, it is instead treated almost as some sort of special topic. That's a mistake because MET provides a framework for understanding so much of biology, which in turn helps demonstrate the strength of the theory. The inefficiencies of the Calvin cycle might be a tad esoteric for this purpose, but there are many other topics where the relevance is clear and easily understandable.

    ReplyDelete
  20. Divalent says,

    Your kidding, right Larry?

    From your article on Biology in Ontario schools: "Evolution is only covered specifically in Grade 12 Biology ... [T]his course is only taken by a small percentage of students in Ontario high schools."


    First, what "small percentage" means depends on the high school. I was talking to a former high school biology teacher last night and he told be that at his former high school about half the students took Grade 12 biology.

    This was similar to the numbers at the school my kids went to.

    Second, I was responding to this comment of yours ...

    A rather narrow and excessively pedantic (and IMO, disingenuous) objection, particularly since it is quite clear that the meaning of “evolution” that he and Coyne are addressing is the one that all but 0.001% of the world’s population has in mind when they ponder the issue. It’s the same meaning of the word as it is used in high school (and lower level) biology text books (and which is about as far as it is appropriate to go at that level).

    The point is that your claim is incorrect if you are referring to high schools in my province.

    Are you really suggesting that the majority of people think of your definition when they hear the word "evolution"? Even in Canada?

    No. But that's a problem that I'd like to change and one of the first steps is to get leading writers like Richard Dawkins to change their may of explaining evolution.

    Incidentally, this discussion isn't really relevant. Even if you define evolution as "descent with modification" the fact remains that most of it is due to random genetic drift and not natural selection.

    Ask them what evolution means to them, and I'll fly over to Toronto and buy you a nice lunch at your faculty dinning room if even one of them so much as hints that drift of neutral mutations is part of their concept of evolution.

    What exactly are you trying to say?

    Of course I know that the general public has a poor understanding of what evolution is.

    Are you saying that because of that I should give up trying to teach them?

    ReplyDelete
  21. @Corneel
    "Aluchko said...
    A coin flip is pure chance, there's no real bias to either tails or heads, but for genetic drift it's not pure chance as real harmful stuff tends not to propagate.

    If real harmful stuff fails to propagate, it is called natural selection, not genetic drift."

    That's kind of my point. Genetic drift is a subset of natural selection, when you get to the genetic drift stage you've already gone past pure chance in determining it's not being naturally selected. It's kinda like saying "this process is completely unbiased, except when its not". It's valid to consider just the group of genes that are disjoint from the set that are hit by natural selection and are truly random, but I don't think Dawkins is wrong when he thinks that just because some genes happen to not be hit by natural selection that they're still not in the realm of pure chance.

    ReplyDelete
  22. I cannot believe Larry Moran is still using the ridiculous word 'adaptationist'. It was silly when Gould and Lewontin coined it, and it is even sillier today.

    Obviously an enormous amount of evolutionary change is due to random genetic drift. I have emphasized this again and again in my books. I have gone so far as to say that the vast majority of evolutionary change, at a molecular level, is random. But that subset of evolution which gives rise to adaptation -- that's the interesting bit that biologists, apart from molecular biologists, actually see, can NOT be due to random drift. The only known mechanism capable of generating adaptation is natural selection. So, when we are talking about the evolution of legs that walk, wings that fly, eyes that see, claws that catch, noses that smell -- everything that tempts creationists to see what they think is design -- evolution is nonrandom.
    If Larry does not believe in adaptation, we have a real problem on our hands. But I've met him, and he is not that stupid.
    Richard Dawkins

    ReplyDelete
  23. Richard Dawkins says,

    I cannot believe Larry Moran is still using the ridiculous word 'adaptationist'. It was silly when Gould and Lewontin coined it, and it is even sillier today.

    Actually I think its a very useful word. It describes a subset of evolutionary biologists who's (almost) exclusive interest is evolution by natural selection. They're really not interested in any other kind of evolution. They frequently use the word "evolution" as if it were a synonym for "evolution by natural selection." They are quite comfortable calling themselves Darwinists.

    Most of the time this subset of scientists writes and speaks as though there was no other mechanism of evolution except adaptation by natural selection.

    Obviously an enormous amount of evolutionary change is due to random genetic drift. I have emphasized this again and again in my books. I have gone so far as to say that the vast majority of evolutionary change, at a molecular level, is random.

    Yes, you have said that, at least once. But let's not forget that all evolutionary change takes place at the molecular level—including change that occurs by natural selection.

    Adaptationists tend to write and speak as if there was a kind of change that does not occur at the molecular level—presumably visible phenotypes—and that change doesn't occur by random genetic drift.

    That's not correct. In fact it may be the case that most phenotypic variation is also fixed by random genetic drift and not by natural selection. Do you agree?

    But that subset of evolution which gives rise to adaptation -- that's the interesting bit that biologists, apart from molecular biologists, actually see, can NOT be due to random drift. The only known mechanism capable of generating adaptation is natural selection.

    That's not much of an argument, is it? What you are saying is that change that is due to natural selection is due to natural selection.

    The real question is how much change is due to natural selection. We agree that most evolutionary change is due to random genetic drift but is there a subset of evolutionary change that isn't?

    The trivial answer is yes. The subset that interests adaptationists qualifies. Does this subset of evolutionary change encompass all visible (to the naked eye) changes?

    Are most of the phenotypic differences between, say Asians and Africans, due to adaptation or drift? I've always been puzzled about the position of those biologists who are only interested in adaptation. Do you only get interested in something once it has been proven to be an adaption or do you sometimes get interested in things that aren't proven adaptations and then lose interest when they are shown to be due to drift?

    So, when we are talking about the evolution of legs that walk, wings that fly, eyes that see, claws that catch, noses that smell -- everything that tempts creationists to see what they think is design -- evolution is nonrandom.

    We all agree. It's quite permissible to say that evolution by adaptation couldn't possibly work if it was only due to chance. However, it's not permissible to say that evolution—at least the kind I'm interested in—could not be due to chance.

    ReplyDelete
  24. Larry, if we are going to get that boring and pedantic, of course we agree, and nothing that I have ever said, either in my review of Coyne's book, or here, or before, denies it.

    But it isn't just a question of whether you or I might find adaptation interesting. Jerry's book is a riposte to creationists, so let's ask what THEY find interesting. The answer, of course, is 'design'. In our scientific language, that means adaptation: the illusion of design.

    To refute creationists, we have to explain the evolution of adaptation, and that means we have to concentrate on natural selection as the important driving force of evolution. You and I know that, if you actually measure gene frequencies over generations, the majority of changes are random: fixations of adaptively neutral mutations. But adaptively neutral mutations are totally IRRELEVANT to the problem of design. They might as well not be mutations at all. You will only CONFUSE creationists, and give them spurious aid and comfort (something your hero Steve Gould spent much of his life doing), if you bang on about evolution being random.

    Creationists are forever traducing evolution as 'random'. They are forever saying things like "random evolution couldn't possibly produce something as complicated and efficient as an eye, or a bacterial flagellum". The best way to combat this foolish line of argument is to say, "No no, you are wrong, mutation may be random but natural selection is quintessentially NON-random." Can't you see how positively unhelpful your style of pedantry is?

    Richard

    ReplyDelete
  25. Richard says:

    To refute creationists, we have to explain the evolution of adaptation, and that means we have to concentrate on natural selection as the important driving force of evolution. You and I know that, if you actually measure gene frequencies over generations, the majority of changes are random: fixations of adaptively neutral mutations. But adaptively neutral mutations are totally IRRELEVANT to the problem of design. They might as well not be mutations at all.

    I'm an evolutionary biology layman but for what it's worth, I disagree with this assesment. Neutral mutations can play an extremely important role. Michael Behe has argued that it is impossible to evolve the bacterial flagella without every step being selectable. It may be possible to show such a pathway, but the job also becomes easier with neutral mutations. I'm most familiar with the importance of neutral mutations from a protein evolution perspective, so I'll provide such an example to illustrate the point:

    Bloom, J.D. et al. (2007) Neutral genetic drift can alter promiscuous protein functions, potentially aiding functional evolution. Biology Direct, 2:17 doi:10.1186/1745-6150-2-17.

    The authors conclude that:

    Overall, experiments have now demonstrated two clear mechanisms by which neutral genetic drift can aid in the evolution of protein functions. In the first mechanism, neutral genetic drift fixes a mutation that increases a protein's stability [24,25,55], thereby improving the protein's tolerance for subsequent mutations [26-28], some of which may confer new or improved functions [28]. In the second mechanism, which was the focus of this work and the recent study by Tawfik and coworkers [44], neutral genetic drift enhances a promiscuous protein function. This enhancement poises the protein to undergo adaptive evolution should a change in selection pressures make the promiscuous function beneficial at some point in the future.

    I also enjoyed this recent paper from Andreas Wagner:

    Wagner, A. (2008) Neutralism and selectionism: A network-based reconciliation. Nature Reviews Genetics 9, 965-974.

    the abstract is:

    Neutralism and selectionism are extremes of an explanatory spectrum for understanding patterns of molecular evolution and the emergence of evolutionary innovation. Although recent genome-scale data from protein-coding genes argue against neutralism, molecular engineering and protein evolution data argue that neutral mutations and mutational robustness are important for evolutionary innovation. Here I propose a reconciliation in which neutral mutations prepare the ground for later evolutionary adaptation. Key to this perspective is an explicit understanding of molecular phenotypes that has only become accessible in recent years.

    Cheers

    Steve

    ReplyDelete
  26. Richard Dawkins refers to the evolution of legs, claws and wings.

    Thinking of Sean B. Carroll's examples of such features originating with gene duplications, wasn't it the very neutrality of these duplications that allowed the genetic/epigenetic changes that followed to take place without compromising the viability of the species that carried the duplications? Such neutral mutations must form a large part of the biomolecular history of even those adaptations fixed by natural selection, since it is far too much to expect selectively advantageous mutations to constantly arise full-fledged in the evolutionary "nick of time."

    To the extent that creationist attitudes regarding the Cambrian "explosion," for example, are driven at least partially by the thought that this many body plan changes in such a relatively short time are evidence of outside guidance, it seems to me the concept that random drift accumulates a "bank" of mutations neutral at the time, but potentially adaptive in a changed environment, can be quite helpful.

    ReplyDelete
  27. Richard Dawkins says,

    To refute creationists, we have to explain the evolution of adaptation, and that means we have to concentrate on natural selection as the important driving force of evolution. You and I know that, if you actually measure gene frequencies over generations, the majority of changes are random: fixations of adaptively neutral mutations. But adaptively neutral mutations are totally IRRELEVANT to the problem of design. They might as well not be mutations at all. You will only CONFUSE creationists, and give them spurious aid and comfort (something your hero Steve Gould spent much of his life doing), if you bang on about evolution being random.

    First, I'm not obsessed with refuting creationists. I'm also very interested in educating non-scientists who accept evolution. I'm even interested in educating scientists who accept the wrong version of evolution! :-)

    Second, one way to educate everyone, including creationists, is to point out that much of life does not look designed. That's because many characteristics of living things are the results of accident and contingency and not of natural selection.

    I think it's important for the average person to understand that evolution is sloppy. Many of the results we see today are not something that even has the appearance of design. Unlike you, I'd prefer to battle creationists by discrediting the false metaphor of design instead of agreeing with them.

    Third, if promoting correct science means giving spurious aid and comfort to the bad guys then so be it. I hope you're not suggesting that people who disagree with you should keep their mouths shut just because the creationists might take advantage of dissent among evolutionary biologists?

    Sometimes it sure sounds like that's what you're saying, although when challenged you usually deny it. I suggest that you try to avoid bringing up that topic. It didn't work with Stephen Jay Gould and it won't work with me.

    ReplyDelete
  28. Richard writes
    Creationists are forever traducing evolution as 'random'. They are forever saying things like "random evolution couldn't possibly produce something as complicated and efficient as an eye, or a bacterial flagellum". The best way to combat this foolish line of argument is to say, "No no, you are wrong, mutation may be random but natural selection is quintessentially NON-random." Can't you see how positively unhelpful your style of pedantry is?

    As someone who studies the role of mutation in evolution (http://www.ncbi.nlm.nih.gov/pubmed?term=stoltzfus-a+mutation), what irks me and makes my life difficult is not what creationists say (I couldn't care less), but what my fellow evolutionists say, and particularly the way they "traduce" in the "randomness" of mutation, and perpetuate the "mindless pedantry" of neo-Darwinian doctrines about randomness.

    "Random" can imply many things: indeterminate, lacking purpose, unpredictable, uniform, and irrelevant.

    The aspects of the term "random" that can be applied to mutation also can be applied to selection. Scientists generally accept the doctrine that the future does not cause things to happen in the present, so "purpose" is ruled out as a cause. Selection and mutation are both accidental or "random" processes in this sense. Mutation is not uniform in outcome. Geneticists have known this for a century. Individual events of mutation are not predictable with an appreciable degree of certainty, but one could say the same thing about the unit events in "selection"-- which are births and deaths.

    At a conference many years ago I presented a "constructive neutral evolution" model of the evolution of RNA editing (Stoltzfus, 1999) that included non-uniform relevant factors that would be expected to have predictable effects, and thus was not "random" in the sense of uniform, irrelevant or unpredictable. Yet after the talk, one scientist raised her hand and asked "So, you're saying its all random?". What she meant was "Are you saying that RNA editing does not have a { God-given | adaptive } purpose?".

    ReplyDelete