While passing Prof. Charles Fenster on campus a few weeks ago, I mentioned to him that I was reading The Social Conquest of Earth by E.O. Wilson. He suggested (I think facetiously) that I should write a review of it on my blog (at the time I did not have a blog).
E. O. Wilson’s made some rather abrasive and controversial claims in a widely criticized paper on the evolution of eusociality a couple of years ago. The book continues in the direction of alienating most evolutionary biologists. I think that explains Prof. Fenster’s tongue-in-cheek remark. But I’m taking him up on the suggestion, in part because I’ve decided after reading it– that this book does certainly deserves some critical reviewing. (Ideally, such a review would come from people with more scientific street cred and visibility than me, and appear someplace where someone might actually read it, but oh well.) The book has received many positive reviews from journalists who don’t realize that the ideas of the book are Wilson’s own rogue views which don’t reflect what most evolutionary biologists think. The rogue view I’m speaking of concerns modern group selection, or multi-level selection.
I’m also writing this because so many people I’ve heard from recently have fallen prey to the same misunderstanding as Wilson: that inclusive fitness and multi-level selection are alternative hypotheses for how social evolution works, rather than what they are: simply different modeling styles, different perspectives on the same biological process. So I have provided some links throughout to help people avoid this trap. For example, this recent back and forth in Nature summarizes the whole thing quite well in two pages.
Here’s an 18 minute video. There’s also this 1 hour entertaining talk where Peter Nonacs discusses alternative models to kin selection to give some more background.
I really wanted to like this book. It was a birthday gift to me from a dear friend, and I admire E. O. Wilson as a public figure. He’s made huge contributions to science, ecology, conservation biology, and the public’s appreciation of science. I was personally inspired by his friend and colleague Tom Eisner who had dinner with my friends and I once a week when I was an undergraduate, and told us adventure stories about the young E. O. Wilson. But alas, my feelings about this book are encapsulated well by a quote from a letter by Richard Dawkins to the journal Science regarding a 1969 article challenging the idea of the dance language of bees:
[The author] presumes to challenge findings of a great biologist. There is of course nothing wrong with this as such, but reliance on an elementary logical fallacy in doing so should not be encouraged in your journal.
The Social Conquest of Earth is a very ambitious book that challenges the findings of many great biologists. It is mainly about the evolution of humans, but Wilson also hopes to revolutionize the field by replacing some older general theories of social evolution with alternatives that are newer, sexier, and vaguer. The problem is that they are not new.
E. O. Wilson (above). A brilliant scientist who wrote a not-so brilliant book.
Wilson certainly made a big splash when he popularized the term “sociobiology” in his book of the same name, a label that today still means very different things to various people. (I’ve gotten the impression from my wife’s parents, who are both sociologists, that being called a “sociobiologist” is akin to being called a sexist, racist, and all-around bad person. The state of affairs that led to this mess is reviewed well in excellent the book The Blank State by Steven Pinker). To avoid such connotations and baggage, many “sociobiologists” started calling themselves “behavioral ecologists” instead. But you can only get so far by redefining old terms or inventing new ones. Rather than solving puzzles, it can just lead to more confusion. Perhaps much of the controversy around sociobiology could have been avoided if academics in the natural sciences and humanities didn’t have such different meanings for terms like “altruism”, “strategy”, and “preservation of the favoured races in the struggle for existence”. Oh well, hopefully that’s been cleared up by now.
The current controversy is Wilson’s rejection of inclusive fitness theory. W. D. Hamilton’s inclusive fitness theory is widely and rightfully regarded as the most important contribution to social evolution since Darwin. The idea was explained and popularized to a larger audience by Dawkin’s classic book The Selfish Gene and it has been the foundational framework for researchers studying social evolution. Natural selection typically occurs when genes cause the individuals carrying them to somehow have more offspring on average than others in a population. Genes that increase lifetime reproduction of the individual carrying them are in effect self-perpetuating, because any resulting offspring tend to have copies of that same gene. This is high school Biology 101.
Inclusive fitness simply involves the simple but profound fact that some genes are favored by natural selection through the exact same process occurring through relatives other than offspring. Genes perpetuate themselves when I take care of my own offspring. That’s obvious. But to a same or lesser degree, the same thing occurs when I take care of my offspring’s offspring or my sister or my sister’s offspring. This can explain why natural selection can lead to extreme reproductive altruism where organisms sacrifice their own reproduction to help related individuals (including offspring but also other relatives). It also means that individuals might compete less severely with relatives than with non-relatives, for example as discovered recently in plants.
There’s way more to it than that obviously, the theory has gathered evidence, theoretical extensions, and made predictions for more than 40 years. Inclusive fitness theory also helps explain and predict a heck of a lot of other things such as cooperative breeding, sex ratios, parent-offspring conflict, intragenomic conflict, genomic imprinting, microbial cooperation, etc and has applications as diverse as medicine and agriculture.
But Wilson considers this all a useless “misadventure” that utterly “failed”. Why? Because he claims we have better theories.
E. O. Wilson and D. S. Wilson (no relation) previously attempted to replace inclusive fitness theory with multi-level selection theory, which is an updated version of naïve and incorrect old version of group selection (the idea that animals live and die for the good of the group or to perpetuate the species). Multi-level selection theory however was later demonstrated again and again and again and again to be a rephrasing or reframing of the same process described by inclusive fitness. It’s a different modeling tool, not a competing theory describing a different reality (as Wilson sees it). Although it presents a useful new way of talking about social evolution, multi-level selection is mathematically equivalent, makes the same predictions, and differs mainly by having different definitions of key concepts like “altruism”. Most biologists don’t use it because it’s more vague and confusing than inclusive fitness theory.
E. O. Wilson then joined forces with Harvard mathematician Martin Nowak to write a second paper attempting to replace and discard inclusive fitness theory. This was also met with widespread criticism, including a critical response paper with 137 authors, representing the major experts in the field.
I feel a bit silly reviewing all this; it’s all been said before, more concisely and much more eloquently. For example here on Joan Strassman’s blog. The press reported this as a major controversy, but actually there’s a clear scientific consensus: relatedness is a key to altruism and indispensable factor in social evolution.
Following in the footsteps of other scientists with arguments rejected by their scientific community, E. O. Wilson now takes his argument to a popular audience who don’t have the necessary background to know any better. This is fine. Afterall, maybe the experts are all brainwashed and can’t see outside the box. But Wilson comes close to presenting his own fringe perspective as being the modern scientific consensus. I found that a bit dishonest. To clarify, it is dishonest to imply that there is a scientific consensus about something for which the scientific consensus is exactly the opposite. Wilson doesn’t do this explicitly, but he comes close. For example he states: “prior to the shift to group selection, the standard explanation … was inclusive fitness theory”. This supposed widespread paradigm “shift” he alludes to here never actually occurred.
Wilson’s own approach begins by classifying humans as “eusocial” together with ants, bees, wasps, and termites, and apart from other highly social primates like chimpanzees. I still don’t fully understand why he does this. Perhaps it to say something new by using old terms in new ways; that is certainly a theme of the book. He claims that eusociality in insects can be explained using natural selection, which he confusingly considers to be an alternative to the idea of kin selection (inclusive fitness theory). His argument is that sterile workers insects are not real individuals; they are just enslaved robot offspring extensions of the Queen’s own genes. Wilson somehow misses that he is just one tiny logical step away from inclusive fitness theory. If the workers are robot slaves to the Queen’s genes, then isn’t the Queen’s body itself just a robot slave to those very same genes? And now we could be reading The Selfish Gene. But instead Wilson uses this as an argument for what he calls individual selection as opposed to kin selection (somehow not realizing that this is purely semantic).
Wilson then suggests that pedigree relatedness is not important and that only a single gene in all the individuals and controlling a single cooperative trait is necessary to create indiscriminate cooperation and kickstart group selection (because the gene will favor copies of itself in nestmates). Again he does not realize that he is just reinventing a concept first described by Hamilton and later termed “Greenbeard” altruism, based on a thought experiment by Dawkins in The Selfish Gene (where Dawkins explains Hamilton’s idea by having the reader imagine if an organism with a distinctive trait, like a green beard always helped other organisms with green beards). This mechanism turned out to be very rare as predicted by Dawkins, because it’s too easy to cheat the system. Wilson misses this point.
Having solved eusociality in insects, Wilson turns to humans. Human “eusociality” by contrast cannot be explained by “natural selection” or “kin selection”. No, that can only come from group selection. The human tendencies toward cooperation and altruism are adaptations that evolved for the good of the group, and altruism here has nothing to do with genetic relatedness. Yes, really.
Wilson goes on to make the ill-advised claim that group selection (competition between human groups) generally leads to our virtuous traits and individual selection (competition within groups) gave us “sin”. He didn’t seem to notice that this strange statement classifies the Holocaust (indeed all wars and all genocides) as a “virtue”, not to mention the group-selected biases that lead to bullying, racism, sexism, and all those other mostly unproductive “-isms”. These weird attempts to build scary, rickety bridges between the humanities or ethics and the natural sciences are so common throughout the book that one becomes habituated after a while.
One of my other favorite blatant examples of the naturalistic fallacy is E. O. Wilson’s explanation for homosexuality. Homosexuals, he argues “may give advantages to the group” (by this he means that homosexuality would somehow lead to larger human groups and/or groups that split more often)… and that is why “their presence should be valued” because “a society that condemns homosexuality harms itself”.
It worries me that people tolerate moral arguments from nature like this. What if the proportion of gay people in a “group” is actually on average inversely correlated with the group’s size, growth, and popularity? And what if someone showed instead that societies that condemn rather than condone homosexuality were more likely to be larger or replace other societies? According to the above logic, that would mean we should not value homosexuality afterall. Oops. And what is the fastest growing religion? Because apparently that’s the one we should value the most.
Most of the facts that Wilson uses to support his theory are very familiar. Many of them could be used to support several alternative theories. He recycles evidence for kin selection as evidence for group selection (that’s what happens when two theories make the same predictions). For gene-culture coevolution, he uses the example of lactose tolerance in societies that do or do not drink cow milk. I’ve heard this example so many times it makes me wonder if there’s actually another clear example that exists.
E. O. Wilson also attacks Noam Chomsky’s theory of universal grammar. Universal grammar is, in the broadest sense, simply the fact that there is an innate cognitive instinct in humans to learn a language. The language instinct is why kids learn to speak a language almost effortlessly and why kids who grow up isolated without adults attempt to build their own language. This innate ability involves the drive to start learning words and how to place them in a proper order following certain unconscious rules (universal grammar). Like inclusive fitness theory and evolutionary theory more generally, the existence of universal grammar is pretty much universally accepted. Controversies mainly regard the form and extent of these innate rules. Previous to Chomsky (shown below), people largely considered human minds to be blank slates that just passively absorb language. B.F. Skinner argued that all parts of language were 100% learned. And presumably then, the only reasons chimpanzees raised by English people don’t speak English is because they simply lack the vocal apparatus or are not generally clever enough (but any innate intelligence would imply the slate is really not blank) or perhaps they have some special anti-language adaptation that just keeps their blank slate minds from learning syntax, recursion, and grammar as quickly as a human which also has a blank slate mind that’s just bigger with extra blank slate space.
But it’s not really fair to criticize people’s ideas on language pre-Chomsky. Like the world before Darwin, nobody really thought carefully or clearly about what now seems obvious to us. For example, today, we have no idea about how to speak clearly and scientifically about subjective consciousness (what is it? what is it for? what does it do?). In retrospect, perhaps it will be clear how confused we were.
But E. O. Wilson’s book states that B. F. Skinner was more correct than Chomsky about language!! Why? Because, according to a study described in the book: “the way that people actually think through an action scenario [was indeed universal]”, but “it was less than fully consistent across the languages they used in speech”.
Wilson even admits that “there does appear to be a biasing epigenetic rule for word order embedded in our deeper cognitive structure”. BUT because “its final products in grammar are highly flexible and learned” that makes Chomsky wrong. In fact, Wilson states that Chomsky “succeeded because, if for no other reason, he seldom suffered the indignity of being understood”.
Ouch. On the other hand, Wilson states that his own vague view on language evolution is supported by “recent mathematical models of gene-culture coevolution”.
Though E. O. Wilson is near retirement age, he still has the spirit of a young scientist trying to make a name for himself. He hopes to replace old, outdated theories with supposedly new, updated (maybe even revolutionary) versions. Out with inclusive fitness theory and in with multi-level selection theory. Out with universal grammar and language instincts and in with “epigenetic gene culture coevolution” and “prepared learning”. In his view, biology will unite with the social sciences and humanities to understand human nature through vague but techy-sounding concepts like social network theory and gene-culture coevolution. He does not use these terms to clarify anything in his readers’ minds, or make any testable predictions, but it certainly sounds sophisticated. Wilson is not afraid of bold statements. For example, stating his own “blind faith” that his theories might help bring about “a permanent paradise for human beings”.
I guess this is what some science writers do best. As Jared Diamond, describes on the back cover of the book: “a big but simple question” is solved with “magisterial knowledge of the sciences and humanities”. There are no statements of praise on the back of this book from evolutionary biologists however.
E. O. Wilson is undoubtably a great scientist and a great advocate for environmentalism. I think one take home point of this book is that science does not always rely on great scientists being right all the time. At least big dramatic statements make everyone re-examine all their assumptions. Being wrong is an important part of science; it’s what makes it work.
I kept thinking how bad this book would be for me if it was my first introduction to social evolution. However, there was one really useful tangential idea I got from the book– a reference to a 1977 paper I had not read on speciation and population fragmentation. I don’t know much about speciation. I thought, even if the rest of the book was mostly BS, this was pretty interesting and informative to me. Even if the rest of the book was not scientifically useful to me, this was this little gem. I cited the idea and the paper in a recent book chapter draft I was working on. When my advisor read it, he suggested I delete the whole paragraph and commented: “The idea behind this paper was debunked years ago.” I should not have been surprised.
Despite all my disagreements, I still enjoyed reading this book a lot. Martin Nowak has a similar overly ambitious book called Supercooperators, which is sure to piss off most people in the field. It also attacks inclusive fitness and attempts to reinvent the social evolution wheel by using mathematical models that are really cool conceptually, but pretty useless empirically. Nowak did have the excellent taste to include a chunk of text on vampire bats though! Can’t complain about that.
- Agreement and disagreement in social evolution: insight from David Queller (sociobiology.wordpress.com)
- ‘The Social Conquest of Earth,’ by Edward O. Wilson (nytimes.com)
- Eusociality? A Review of Edward O Wilson’s The Social Conquest of Earth (dinmerican.wordpress.com)
- A NYT profile of E. O. Wilson and his new book (whyevolutionistrue.wordpress.com)
One thought on “Review of “The Social Conquest of Earth” by E. O. Wilson”
“Richard Dawkins’s review of The Social Conquest of Earth (Prospect, Issue 195, 24th May 2012) makes little connection to the part he criticizes. The central issue in the book, which he urges others not to read, is the replacement of inclusive fitness theory (kin selection theory) by multilevel selection theory (ie, individual and group selection combined), with a new and major role assigned to group selection in the origin of advanced social behavior. The original formulation was made by Martin Nowak, Corina Tarnita, and myself in 2010 (Nature 466: 1057–1062). We demonstrated that while inclusive fitness theory sometimes works, its mathematical basis is unsound, and inclusive fitness itself is an unattainable phantom measure.”
can you explain what eo wilson means by the last comment regarding the math?