Wednesday, July 23, 2008

Climbing Mount Improbable as Metaphor

 
One of my postings, Good Science Writers: Richard Dawkins, has been re-posted on RichardDawkins.net. This doesn't happen very often—in fact this may be the very first time. I can't imagine why they would have selected this particular posting.

I mentioned that some of the Dawkins metaphors are misleading and I suggested that Climbing Mount Improbable was one example. That prompted a comment from Richard Dawkins so I replied on his website. In case anyone is interested, I'm reproducing it here.



Richard Dawkins asks,

I am interested in the suggestion that Climbing Mount Improbable might not be an ideal title.

Richard, we've been over this ground before but for the benefit of the lurkers let me explain why I think the metaphor is inappropriate.

To begin with, you use the Mt. Improbable image as a metaphor for evolution. This is misleading since evolution encompasses more than just adaptation. It would be difficult to apply the "Climbing Mt. Improbable" metaphor to the organization of our genome, for example, since it's clearly not well-designed and could never be characterized as the peak of an adaptive landscape.

But even as a metaphor for adaptation the image is less than perfect. Most readers will see the peak of Mt. Improbable as a goal of adaptation, implying that evolution somehow recognizes that there is an ultimate perfection that all organisms seek to achieve by reaching the summit. As you well know (I hope) there are very few (any?) species that are perfectly adapted to their environment. If this were true, adaptation would cease because the species resides on the summit of Mt. Improbable.

Thus, in the real world, species tend to move about in the foothills rather than attempt to scale the highest peak. As long as they are good enough to survive and reproduce that's all that's required.

Yes, some individuals within the population might acquire a mutation that makes them a little more fit but in most cases the selective advantage will be too small to make much of a difference. I don't believe there's any great pressure to get to the top of Mt. Improbable. That's why we usually don't see perfection in nature. And it explains why most organisms do not look as though they have been designed by some intelligent being. If anything the "design" looks more like a Rube Goldberg creation, and I doubt that anyone would say that those creations represent the peak of perfection.

I prefer a different view of evolution, one that emphasizes chance and accident [Evolution by Accident]. For me, the metaphor of "Climbing Mt. Improbable" is quite wrong as a metaphor of evolution.

Now, I understand that you disagree about the role of chance and accident. You say, for example, on page 326 of Climbing Mt. Improbable, "It is all the product of an unconscious Darwinian fine-tuning, whose intricate perfection we should not believe if it were not before our eyes" (referring to the evolution of figs and fig wasps). For someone who believes that such a description is characteristic of most evolution (adaptation) the "Climbing" metaphor may seem quite appropriate.

BTW, I agree with you that your case for adaptationism is much stronger in Climbing Mt. Improbable than in The Blind Watchmaker. I especially like the chapter you mention, The Museum of All Shells, where you discuss - among other things - the contrast between your view of evolution and the mutationist view. Ironically, you begin that discussion by pointing out that this is a sophisticated controversy, "... and Mount Improbable, even in its multiple-peaked version, isn't a powerful enough metaphor to explore it."


11 comments :

  1. My view of "Mt Improbable" is that it is a useful metaphor only in the context of explaining how things like the human eye or the bacterial flagellum (i.e. things that creationists love to pick on) could have evolved. I've never thought of the peak of Mt. Improbable as implying 'perfection' or as somehow 'greater' than the landscape.

    By contrast, I've always viewed the metaphor as Richard has explained it; that one cannot jump to the peak of the mountain (that is, obtain a complex eye) without first taking many tiny steps. It doesn't imply the human eye (or, indeed, the human genome) is better or perfect than any other; just that one can only explain it as the product of intermediate steps.

    I'm not even sure that 'Mt. Improbable' implies the change has to come about by natural selection rather than random genetic drift. The point is that however the peak is scaled, it is scaled only in small steps.

    ReplyDelete
  2. Re: your evolution by accident article:

    a simple way to say this is that evolution, while not a random process, is a stochastic one. Mutations might be random, but those that get fixed by natural selection are not.

    ReplyDelete
  3. It's been a long time since I read Mount Improbable, so I am probably imagining this ... I have in my head the idea that the metaphor included some comment about the peaks changing over time (some kind of plate tectonics metaphor), so that a species that manages to ascend that tallest peak may find, in geologic time, that it is now on a foothill or even a valley; there is no tallest peak, for long, because the environment shifts around you.

    ReplyDelete
  4. It was good to read your post on RD.net. I have never felt at ease with the metaphor, but, as I explain there, it is for different reasons.

    Steve Zara

    ReplyDelete
  5. olie says,

    a simple way to say this is that evolution, while not a random process, is a stochastic one. Mutations might be random, but those that get fixed by natural selection are not.

    And what about those mutations that are fixed by random genetic drift? Surely those are examples of accidental evolution?

    If a large percentage of total evolution is due to random genetic drift then it becomes quite reasonable to say that much of evolution is due to chance. No?

    That's the point. Those who emphasize the non-randomness of evolution are only thinking of adaptations. That's why they are called adaptationists.

    ReplyDelete
  6. "This doesn't happen very often—in fact this may be the very first time."

    You can lay the blame at me. I sent them the link. Here's the e-mail I sent them:

    "Hi -

    I thought you might be interested in this article on Larry Moran's blog Sandwalk.

    "Good Science Writers: Richard Dawkins."

    Normally, a blog post on Richard Dawkins wouldn't be noteworthy, except for the fact that Moran has publicly disagreed with Richard Dawkins about biological issues. Larry also met with Richard on a trip to Oxford with PZ Myers, and so they are at least cordial acquaintances.

    Mike"

    Whether you wish to thank or curse me is up to you, Larry.

    ReplyDelete
  7. I'd stop just short of saying that there is an intentional misreading of Climbing Mount Improbable in this post.

    The focus was on the distinction between jumping to the peak and taking a longer stroll up an accessible path.

    The "peaks" don't represent some degree of perfect adaptation, they simply represent the evolution of a complex feature that simply can't be reached by jumping up the cliff. Even the peaks are defined post hoc as was Mont Blanc in Watchmaker. Having reached a peak (or local maxima) says nothing about the idealism of the resultant construction, merely that it was reachable without saltation.

    When you write that the genome is "clearly not well-designed and could never be characterized as the peak of an adaptive landscape" I think you completely miss that it is very well adapted for the propagation of the genome, particularly the "parasitic" junk.

    ReplyDelete
  8. Larry: And what about those mutations that are fixed by random genetic drift? Surely those are examples of accidental evolution?

    ----------
    In my uneducated opinion: "yes and no"; "yes" in that they have no adaptive benefit but "no" in that, and forgive my technical illiteracy here, but it is my understanding that such mutations must be those that allow for the newly mutated genes to "be compatible with" the already present genes, no?

    Anyway, this is why I call it a stochastic process: randomness certainly plays a huge process (as that recently published experiment on the evolution of bacteria recently demonstrated) but genetic realities keep the process from being purely random.

    ReplyDelete
  9. malachi says,

    I'd stop just short of saying that there is an intentional misreading of Climbing Mount Improbable in this post.

    Thank-you, because if you had not stopped short you would have been completely wrong (instead of just partially wrong).

    The focus was on the distinction between jumping to the peak and taking a longer stroll up an accessible path.

    Correct, but irrelevant. I don't like the metaphor of Mount Improbable whether or not it is a gentle slope on the backside or a steep cliff on the front side.

    I think you missed the point.

    ReplyDelete
  10. And what about those mutations that are fixed by random genetic drift? Surely those are examples of accidental evolution?

    I'm not an evolutionary biologist by any means, but here's my take on this:

    Imagine a scenario where a mutation arises in a small population (or imagine a founder population). In this case, you're quite right; genetic drift will predominate, and the evolution of that population will be highly stochastic. However, the possible mutations that may occur and be fixed, even in very small populations, is still limited by natural selection.

    If, in a small population, mutations arise that produce an inverted retina, this mutation (that is, the genotype) may persist solely due to chance. However, the fact remains that the phenotype, the inverted retina, must be compatible with survival in the environment in which this small population finds itself.

    Similarly, mutations may arise in populations that attenuate or eliminate the function of particular enzymes (such as those we seen in human founder populations). Those mutations may predominate and spread throughout the population simply due to chance (either chance matings, or indeed the chance produced by the genetic isolation of the founder population in the first place). BUT, those mutations are still subject to natural selection; mutations in the enzyme ATR do not predominate, for instance, because they are incompatible with life (ATR ko is embryonic lethal). Even milder mutations, while they may arise and prosper by chance, are selected for or against based on whether they limit reproductive success. Granted, weak selection pressures may have minimal effect on gene pools, especially if the mutation in question does not severely limit reproductive success, but even in this case, the number of possible mutations that can become fixed is limited by selection.

    Genotypes may change through drift, and I have no doubt that drift is responsible for setting up patterns of alleles in the population. But you cannot escape the fact that natural selection acts on phenotype, not genotype. Genotypes may vary wildly, but only a select few phenotypes are compatible with life. That is selection.

    ReplyDelete
  11. I have to say it's many years since I read Climbing Mount Improbable but I don't remember taking the fitness peaks to be perfection, more like optimums in a landscape that was constantly changing anyway. The metaphor was to illustrate how even tall peaks that could not be reached in a single bound could be reached easily by many small steps if there was sufficient time.

    Again, if I remember correctly, although this may have been in another of his books, Dawkins was not claiming that selection was all that there was to evolution but that it was the most interesting part. As far as I remember, he did not deny that random genetic drift occurs or even that random mutations could become fixed in a population. But could the fixation of random mutations through genetic drift lead to the evolution of an eye or hand or brain, for example? In that sense, adaption is the more interesting process because it is capable of creating more complex biological structures.

    ReplyDelete