Evolution for Everyone

The newest effort to place individualism on an evolutionary foundation is led by Alan Grafen and Andy Gardner at the University of Oxford. Their stated purpose is to develop a mathematical framework whereby natural selection leads toward the optimization of individual fitness. They want to be able to say that virtually all adaptations are individual-level adaptations. This is a daunting challenge because it is constrained by previous versions of individualism that have already breathed their last.

First, as good evolutionary theorists, Alan and Andy acknowledge that group selection happens. It was Andy who said “Everyone agrees that group selection occurs” on the pages of Nature magazine (see part II).

Second, nobody can deny that examples such as cancer and meiotic drive, which reflect natural selection operating below the level of the individual organism, tend to decrease fitness at the level of the individual organism. The only way that Alan and Andy can achieve their goal is to say, somewhat lamely, that these examples aren’t very common.

Third, Alan and Andy do not contest the fact of major evolutionary transitions. They agree that individual organisms are highly regulated social groups of lower-level units that previously led a more fractious existence. They also concede that at least some present-day groups, such as social insect colonies, quality as superorganisms.

If Alan and Andy concede all of these points, then how do they achieve their own version of individualism? By calling everything that falls short of a fully formed superorganism an individual-level adaptation.

Readers interested in the details should consult both an article by Andy and Alan titled “Capturing the superorganism: a formal theory of group adaptation” and my newly published article with Elliott titled “Adaptation and Natural Selection Revisited“, both published in the Journal of Evolutionary Biology. Think of these articles as like the advanced level of a video game that requires that you pass through the lower levels before you are be able to become a master. Here I will make a few observations that can be appreciated without arcane knowledge.

First, Andy and Alan’s argument divorces the concept of adaptation from the concept of natural selection. In other words, according to them, a trait can count as an individual-level adaptation even though it was never influenced by individual-level selection. Their argument therefore violates Williams’ Principle. That doesn’t make it wrong, but evolutionists should think long and hard before they reject a principle that, until now, has been the most enduring part of Williams’ legacy.

Second, there is a robust alternative to portraying natural selection as a process that always leads toward the optimization of individual fitness. Natural selection can also be understood as the net effect of opposing forces, as we can see by returning to the example of sex ratio that has figured so importantly in the group selection debate (see part II). Suppose that within-group selection favors an even sex ratio and between-group selection favors a highly female-biased sex ratio (to maximize population growth), with only enough males produced to fertilize the females. Based on design principles, we can clearly delineate the optimum sex ratio that maximizes individual-level fitness and the ratio that optimizes group-level fitness. Now let’s say that individual and group selection operate to different degrees in different species. Individual selection is the exclusive force in some species, leading to an even sex ratio. Group selection is the exclusive force in other species, leading to an extreme female-biased sex ratio. And in still other species, both individual and group selection operate, leading to sex ratios somewhere between even and extremely female-biased (e.g., ¾ females). For the intermediate cases, it makes perfect sense to conclude that nothing is being optimized. The outcome of natural selection is the net effect of forces tending toward different optima. Alan and Andy resist this conclusion. They want to say that every outcome of natural selection, from an even sex ratio to a highly female-biased sex ratio, counts as an individual-level adaptation but that only a highly female-biased sex ratio counts as a group-level adaptation, though even here what evolves still counts as an individual-level adaptation. If this sounds mystifying to you, you’ll need to enter the next level to judge the intellectual battle between Alan and Andy vs. Elliott and David.

I end this installment with a final observation that can be appreciated without arcane knowledge. Whatever the verdict on Alan and Andy’s form of individualism, it is different than the previous forms. For people without arcane knowledge, the debate over what counts as a group-level adaptation seems continuous. It was controversial then and it’s controversial now. One reason that I have written this series is to demonstrate to a wide audience that the form of individualism being argued now is completely different and incompatible with the form that George Williams argued for, which is that group selection never happens. It is a very different beast that–as far as Elliott and I can tell–is also breathing its last gasp.

To be continued…


  1. #1 samc
    January 18, 2011

    One reason that I have written this series is to demonstrate to a wide audience…

    Rubbish, the only reason you wrote this epic autohagiography was to sneer at people you disagree with and show off what you regard as your wonderful insights. Don’t kid yourself this is educating others; you’re just showing off incessantly, unpleasantly, annoyingly. The lack of follow-up comments rather suggests few people take the time to wade this glacier of sticky porridge.

    Second, nobody can deny that examples such as cancer and meiotic drive, which reflect natural selection operating below the level of the individual organism, tend to decrease fitness at the level of the individual organism.

    Oh yes they can. Your failure to understand the underlying issues partly stems from your inability to understand that fitness is not a trait. Fitness is not inherited. Fitness is not an emergent property.

    You really don’t understand your subject very well, do you? Other than licking EO’s arse, you seem to think nobody, possibly even including Darwin, understood evolution and natural selection as well as you. You are so wrong. Fail.

  2. #2 benc
    September 1, 2011

    It’s nice that samc was able to keep such a reasonable, even-keeled tone, or I would have had difficulty taking his comment seriously.

    Thanks for these posts Dr. Wilson. I’m not yet sure how I feel about group selection, but this blog is a great place to learn about the issue and learn where to read about counter-arguments to get both sides of the issue.

New comments have been disabled.