Smithsonian recently asked me to interview Harvard biologist E.O. Wilson about his new book, The Social Conquest of Earth. You can read the Q & A on their web site.
Wilson and I spoke for quite a while, covering wide range of subjects. One particularly interest part of the talk addressed his ongoing attack on a major aspect of modern evolutionary theory, known as inclusive fitness. I reported on his attacks–and the response of his critics–in the New York Times in 2010.I reported on his attacks–and the response of his critics–in the New York Times in 2010.I reported on his attacks–and the response of his critics–in the New York Times in 2010.
Basically, most evolutionary biologists believe that a great deal of behavior–including altruistic behavior–can be explained by the way genes get passed down among relatives. If you help your cousins, some of your genes will get transmitted even if you have no kids of your own. Wilson and his colleagues at Harvard, Martin Nowak and Corina Tarnita, argue instead that inclusive fitness doesn’t make mathematical sense and is unnecessary. Wilson holds that good old natural selection on individuals can explain a lot, and he also argues for selection on higher levels. Groups of organisms–human tribes, for example–can be selected for their group-level traits.
In the year and a half since my article, his critics have counterattacked. If you want to see why they think he’s wrong, I’d recommend reading five fiery posts by University of Chicago evolutionary biologist Jerry Coyne at his web site Why Evolution Is True: 1, 2, 3, 4, 5.
For those who have been following this scrum, I thought I would include a portion of the conversation that did not end up in the final version, focusing on Wilson’s response to critics. It goes pretty deep in the biological weeds, and it doesn’t read smoothly, because both of us would stop in mid-sentence to make ourselves clear to each other. (I’ve trimmed a few really incomprehensible dead ends.) But if you just can’t enough of the Hamilton Inequality, enjoy:
Q: When you presented your ideas in that Nature paper in 2010 with Novak and Tarnita, a number of scientists responded–over 150 scientists responded both in Nature and then in some other journals as well, taking issue with your argument. They said that inclusive fitness was, in fact, a very powerful and legitimate explanation. Had you anticipated that kind of response?
A: Yes. (Laughter) It’s just that the inclusive fitness theory had persisted as the correct and prevailing theory for almost four decades. And the ones who were animated most of course were those who were working in the field, trying to perfect it and using it to explain social behavior.
But there were two things wrong, and this is what I was pointing out over a period of four years, and finally came to a more definitive form in the Nature paper. First, as the mathematicians showed, the basic foundations of the inclusive fitness theory are unsound. The Hamilton Inequality does not work except in extreme conditions that scarcely exist on earth. And inclusive fitness is a phantom measure which seems intuitively right but which simply doesn’t exist in any form that could ever be measured.
And that being the case, then we should look more carefully at just what has been accomplished with inclusive fitness theory. Extremely little, in quantitative terms. And most of the applications that have been made–and they’re made over and over again to make up the main corpus of literature on inclusive fitness theory–applies to degrees of conflicts of societies, which vary inversely with the degree of relatedness among individuals. I’ve been able to show that there is a perfectly good set of alternative biological hypotheses to explain that. That’s in my part of this article, in Nature, but I went into much more detail in an earlier paper in Bioscience.
Even the applications of inclusive fitness theory are not necessarily the only ones that can be made. I have argued that ones from an individual-level or group-level selection–or the interaction between the two–provide superior fits. And further, the multi-level selection model allows us to explore and explain a great deal more phenomena. And I expect that to be shown as work goes on. I think it’s particularly relevant to the explanation of the social behavior in humans.
Q: Just to take one example that the critics raised, they talked about how inclusive fitness theory makes a prediction about sex allocation, about the investment in different sexes in the offspring. And they say this is something that inclusive fitness predicts and we’ve gone out and we’ve done a lot of tests to see if that’s true and they find these ratios in lots of animals as predicted by that theory. When they make that sort of argument, what’s your response?
A: It’s a little bit like Ptolemaic astronomy: epicycles will always give the exact results if you’re willing to add them. And in this case–I have pointed this out as well–there’s a flaw in the reasoning about the studies of investment, particularly in whether you invest more in males or females in the social insect societies.
If you have only one female who is queen in the colony, and if that queen has mated only once so that her offspring are that close, then you should see because of the implications of haploid/diploid, the way sex is determined in ants, bees, wasps. You should see a favoring of investment in new queens, over investment in males as measured by the amount of biomass. And that inequality does exist and it should be three to one investment in the weight. And that has been what is thought to be a very powerful argument.
However, this I believe has a major flaw in the reasoning. The colony wishes to make an investment in males versus females in numbers that would be most advantageous in having a female successfully mated, when they leave the nest to get mated, bees, ants, wasps. And therefore, the colony should be trying to get something closer to a one-to-one investment.
And since females are much bigger–they have to have all that fats and ovary and so on–and males are much smaller because in most of these social insects. All they have to do is find a female, deliver their sperm, and die. So the males are much smaller.
This means then that getting a one-to-one ratio in sex that is the same as you see throughout the rest of the animal kingdom, means that you will be having to invest much more in the females when you invest in males. And actually when you make that hypothesis, use that principle, which is the obvious one, then that comes closer to the actual figures we have in the biomass investment.
They [Wilson’s critics] may dispute that, but my point is that they did not by any means find a testing ground on which the old theory could stand or fall. It’s in my view a much simpler and more precise explanation to use the argument of one to one ratios of male and female.
Q: One other thing that critics have brought up is they claim that there are no new predictions in the argument that you and your colleagues set forth. Are there predictions that you can make from this different view of social evolution?
A: My book, The Social Conquest of Earth, is filled with them.
Q: What would be a couple predictions?
A: You mean in terms of group versus individual level selection? Yeah. What it is, is more of a…Rather than an a-prioristic application of group selection as it is the use of group selection to explain what actually happened. And that is, you piece together only by the close examination of the biology of the various stages leading up to advanced social behavior. That’s the best I can say right now.
Multi-level selection theory is undeveloped, essentially, most undeveloped because it has been almost abandoned due to the dominance of inclusive fitness theory. I think inclusive fitness theory has been rejected and now we have every reason to return to a multi level selection theory and develop it. And Martin Novak is beginning to do that to some extent and to develop a rejuvenated, multilevel selection theory.
So I think at this point multi-level selection now is open to development without the inhibition of inclusive fitness theory–may I use the expression the deadweight of inclusive fitness theory?—will, I believe, from this point on be developed and tested in a proper way.