Evolution for Everyone

So far I have paid homage to George Williams for his clear thinking about adaptation and natural selection, including his principle for evaluating whether a trait counts as a group-level adaptation (see part I). George was wrong when he made the empirical claim that “group-related adaptations do not, in fact, exist”, but it is thanks to his principle that we can say so. Providing the basis for proving oneself wrong is a high honor for a scientist.

Another enduring part of George’s legacy was his ability to speak in plain language. By his own account, he wrote Adaptation and Natural Selection to make the arcane knowledge that had developed within the field of population genetics accessible to a broader audience of biologists. His examples, such as flying fish obeying the laws of gravity and the distinction between a fleet herd of deer and a herd of fleet deer, had a parable-like quality that causes them to be remembered and recited to this day.

George’s gift for plain talk shows that the basic concept of adaptation can be easily understood, even though formal theory might also be important for working through the details. A mathematical model isn’t needed to decide that the wing-like fins of a flying fish are well adapted for gliding through the air. Even if we disagree about the agent or process that designed them (e.g., creationism vs. natural selection), we can agree that they are well designed for gliding and that this is very unlikely to have happened by chance. It’s equally easy to decide that the fins of the flying fish are individual-level adaptations, since they so clearly benefit the individual fish and not other fishes. What’s hard to understand?

Group-level adaptations are just as easy to understand, at least in many cases. We don’t need a mathematical model to decide that a termite mound or paper wasp nest is well designed, although models might well reveal design features (or failures of design) more subtle than meets the eye. In these cases, individuals are clearly building a structure that benefits not just themselves but a whole group. The same is true for the traits associated with human virtues that Darwin wrote about. Altruists and solid citizens generate benefits that are not confined to themselves, unlike flying fish fins. Group-level selection is required to explain the evolution of traits that benefit others or one’s group as a whole. What’s hard to understand?

Another simple point is that adaptations need to be evaluated on a trait-by-trait basis. For a human-designed implement such as a can opener, the parts that engage the can might be well designed while the parts that engage the hand might be poorly designed. For a species such as the flying fish, the fins might be well designed for the individual, the schooling behavior well designed for the group, and temperature tolerance poorly designed for global warming. The phenomenon of major evolutionary transitions, whereby groups become integrated in so many respects that they quality as superorganisms in their own right, is one of the most important developments in evolutionary thought–but it doesn’t change the fact that adaptations need to be evaluated on a trait-by-trait basis.

Williams’ Principle is important not only for the academic study of evolution but in everyday life. Cultural evolution is a multilevel process, no less than genetic evolution. What’s good for the individual is not necessarily good for the group. What’s good for one’s immediate relatives is not necessarily good for the clan. What’s good for the clan or political party is not necessarily good for the nation. What’s good for the nation is not necessarily good for the planet. The fact that adaptation at any level of a multi-tier hierarchy tends to undermine adaptation at higher levels is something that everyone needs to know about the evolutionary process. It shows that the concept of the invisible hand, whereby the pursuit of self-interest automatically results in the common good, is profoundly mistaken. Our lower-level strivings will continue to be part of our higher-level problems until all of us absorb the message of Williams’ Principle.

Comments

  1. #1 Ian Kemmish
    February 3, 2011

    “The fact that adaptation at any level of a multi-tier hierarchy tends to undermine adaptation at higher levels is something that everyone needs to know about the evolutionary process.”

    Sadly, this sentence, which appears to be the entire point of the post, bears no relation to what’s gone before. “Tends to” would likely imply in most readers’ minds that it happens more than half the time, so you should not be short of evidence so support this claim, which appears sweeping to a layman.

  2. #2 Ormond Otvos
    February 4, 2011

    “adaptations need to be evaluated on a trait-by-trait basis.”

    I see nothing in the article that implies, or proves, this statement. Quite the contrary, the implications are the opposite.

    We humans love to oversimplify. Efficient in some ways, blinding in others. Yes, it’s more difficult to keep multiple traits in concert causing emergent behaviors that can qualify as simple traits. But, don’t be lazy as a scientist.

  3. #3 Roy Nilles
    February 4, 2011

    The social group exists because the individuals benefit from the cultural lessons that only an ongoing group of them can provide. Have you ever written about the possibility (to me the probability) that all social species have a culture by necessity? To fit successfully within the group IS in the individual’s best self-interest.
    The fins designed for the flying fish were in effect approved by the group as part of the design perfection process. The genomic purpose if one concedes that genomes serve a purpose. And if one concedes the possibility of facilitated variation as a selective function.

  4. #4 Donald Forsdyke
    February 5, 2011

    For a preprint of an appraisal of George Williams (who, from time to time, visited us at Queen’s University in Kingston), please see:

    http://post.queensu.ca/~forsdyke/bioinf14.htm

  5. I see nothing in the article that implies, or proves, this statement. Quite the contrary, the implications are the opposite.

  6. #6 Jen Williams
    February 7, 2011

    ” concept of the invisible hand, whereby the pursuit of self-interest automatically results in the common good, is profoundly mistaken.”
    I agree that if all the needed ingredients to commit murder are there, the only thing holding back a human from going through the act is conscience; fear of getting caught, fear of feeling guilty. It seems to me that some humans lack this trait of conscience or become insane because they go ahead go through with it already knowing the consequences. This to me is an individualized trait- not governed soley by heredity or environment. This individual is not out for the common good of man- only out for him/herself. It almost a type of survival of the fittest mode- like it’s me or him type of thinking but not in a fight or flight circumstance for the perpetrator.

  7. #7 Mitch Diamond
    February 20, 2011

    Not sure if this is the right place to ask this, but…
    I’m reading DS and EO Wilson’s Evolution “for the Good of the Group. Two of the examples they use for group selection, the siphonophores and slime molds, don’t seem to support the group selection they espouse.

    In the case of the siphonophores they acknowledge that the organism, or as they call it, the colony, is genetically identical, and they want to make this an instance of group selection. The zooids, the component units of siphonophores, may have evolved from separate species to symbiotically cooperate but are now one organism sharing one genotype. How can this animal be an example of group selection when we’re dealing with one genetic individual? Wouldn’t group selection only work with non-identical genotypes?

    In the case of the slime molds, the formerly individual, single-cell amoebas form a multicellular organism with specialized cells, some of which give rise to reproductive spore cells and others that will intentionally die without reproducing. The Wilsons claim this is an example of group selection because some of the amoebas contribute to the fitness of others without any reproductive ability for themselves. This would be true if each amoeba was genetically different, but it would be nice if the Wilsons actually had that evidence and stated it. However, recent studies have suggested that, at least in some cases, the slime mold reproductive structures (slugs) are comprised of clones, i.e. are genetically identical. Even then, the Wilsons claim this is a case of group selection when, at the very least, it should be kin selection. (See Discovery of a large clonal patch of a social amoeba: implications for social evolution. http://www.ncbi.nlm.nih.gov/pubmed/19243508)

    Can someone explain to me how these are examples of group selection?

  8. #8 David Sloan Wilson
    February 20, 2011

    Very quickly, multilevel selection is a nested series of fitness comparisons:

    Among genes, within individuals (gene-level selection)
    Among individuals, within groups (individual-level selection)
    Among groups, within a metapopulation (group-level selection)
    And so on…

    When we apply these simple comparisons to clonal groups, it is clear that there is no individual-level selection (because there is no genetic variation within groups) and natural selection takes place entirely on the basis of group-level selection. These are examples of PURE group selection, not NO group selection.

    Empirically, however, it is not the case that slime mold groups are purely clonal, so there is some individual-level selection for group-level selection to contend with. Moreover, the balance between levels of selection is not static but itself evolves, as Mitch also observes. Cases of nearly pure group-level selection evolved from states in which individual-level selection was once stronger.

    Finally, there are plenty of empirical examples in which all levels of selection are operating strongly. In other words, when we make our nested series of fitness comparisons, we find genes succeeding compared to other genes within the same individuals, individuals succeeding compared to other individuals within the same group, and groups succeeding compared to other groups within the metapopulaton. All three levels of selection are significant and must be considered to know what evolves in the total population.

    What is definitely NOT the case is Williams’ claim that lower-level selection invariably trumps higher-level selection.

    d.

  9. #9 orjin krem
    March 21, 2011

    Empirically, however, it is not the case that slime mold groups are purely clonal, so there is some individual-level selection for group-level selection to contend with. Moreover, the balance between levels of selection is not static but itself evolves, as Mitch also observes. Cases of nearly pure group-level selection evolved from states in which individual-level selection was once stronger.

    Finally, there are plenty of empirical examples in which all levels of selection are operating strongly. In other words, when we make our nested series of fitness comparisons, we find genes succeeding compared to other genes within the same individuals, individuals succeeding compared to other individuals within the same group, and groups succeeding compared to other groups within the metapopulaton. All three levels of selection are significant and must be considered to know what evolves in the total population.

    What is definitely NOT the case is Williams’ claim that lower-level selection invariably trumps higher-level selection

  10. #10 joe
    April 19, 2011

    David, you wrote:

    “Very quickly, multilevel selection is a nested series of fitness comparisons:
    Among genes, within individuals (gene-level selection)
    Among individuals, within groups (individual-level selection)
    Among groups, within a metapopulation (group-level selection)
    And so on…”

    However, G.C. Willimas (“Domains, levels, challengens” 1992, p. 10) introduced a dualism that takes genes out of the material hierarchy of groups, individuals, cells, and chunks of DNA. He put them in a mutually exclusive and conceptually separate domain of selection he called the codical domain. Therefore, Williams was able to admit mutilevel selection in the material domain and maintain a gene’s eye view on the codical domain. Dawkins, in comparison, sorts genes into the material domain.

    How do you argue against Williams’s dualism? You have to if you want to attack the gene’s eye view, for as long as it resides in a mutually exclusive and separate domain of selection, it is inert to attacks from the multilevel selection perspective on the material domain.

Current ye@r *