The NYRB reviews E. O. Wilson’s latest

June 11, 2012 • 10:52 am

The latest issue of The New York Review of Books contains an appraisal of Ed Wilson’s new book, The Social Conquest of Earth, by Steven Mithen, a professor of archaeology at the University of Reading.  Mithen seems a strange choice given that he’s not an evolutionary biologist, but it turns out that his expertise enabled him to catch some errors that might elude most biologists. His review, “How fit is E. O. Wilson’s evolution?“, is behind a paywall, but sufficiently industrious readers can get a copy from me.

Mithen begins by touting the importance of Wilson’s 1975 book, Sociobiology: The New Synthesis, as a landmark in how we think about the evolution of behavior.  He’s right.  But after extolling the book, he gives a taste of what is to come:

. . . how marvelous it felt a few months ago to have received an email from the man himself asking if he could reproduce a diagram relating to human evolution from one of my own publications in his forthcoming book, The Social Conquest of Earth. What an honor to then be invited to review the book for The New York Review. And how awfully disappointed I have been.

Wilson’s book, of course, is about the two pinnacles of social evolution—humans and “eusocial” insects (those species, like honeybees, that have a division of labor, castes, and sterile workers)—and how those groups became so successful.  One of the vehicles for that success, in Wilson’s view, was group selection, the differential reproduction of groups of individuals having different traits. (In this case, traits like altruism in humans or sterility in insects.) This puts him, at least in the insect case, at odds with the vast majority of other evolutionists, who see selection among groups as an ineffective way of evolving anything, and see no evidence that it has occurred in nature.  Rather, selection based on relatedness—kin selection—seems more plausible. I’ve written about this at length, and won’t reproduce my critique here. I’ll just remind you that when Wilson (along with Nowak and Tarnita) published their “group-selection” theory for the origin of eusociality in Nature in 2010, about 150 biologists, including nearly all the luminaries in the field of social evolution—wrote letters to the journal criticizing Nowak et al.’s assertion of the primacy of group selection over kin selection.

According to Mithen, Wilson’s group-selection arguments are also the centerpiece of his book (I haven’t yet read it).  Although he doesn’t pronounce on who gets the upper hand in the kin selection/group selection dispute (kin selection wins, by the way), he finds Wilson culpable for not even mentioning the extensive criticism, which apparently could have been included in the book:

Now, it is not the purpose of this review to pronounce upon the validity or otherwise of inclusive fitness theory and Wilson’s alternative theory of multilevel selection. Indeed, I would not presume to have the expertise to do so. Wilson develops his case by referring to scientific matters on which only experts can make judgment, such as the demise of the “haplodiploid hypothesis” and new mathematical work allegedly exploring inherent weaknesses in “Hamilton inequality.” I have, however, remained unimpressed by multilevel selective theory and persuaded by the weight of academic opinion in favor of inclusive fitness, dogma or otherwise.

My greater concern is about the responsibility of the scientist writing for the general reader, especially a scientist of Wilson’s academic reputation. Such readers, the type targeted by Wilson and his publisher, may never have heard of Nature and would be unlikely to consult endnotes. Such readers, owing to his failure to acknowledge the extent of opposition to his views, would be entirely misled into thinking that Wilson had indeed “demonstrated that inclusive fitness theory, often called kin selection theory, is both mathematically and biologically incorrect.”

I recognize that there might be an issue of timing: The Social Conquest of Earth may have been so far into production by the time that the 2011 issue of Nature was published that citation was impossible. I suspect not: it was a March issue and Wilson’s book cites several 2011 publications. Even if this was the case, Wilson would have been quite aware of the vast weight of academic opposition to his views, since he has been promoting them since 2005 at least. I cannot avoid the impression that the manner in which Wilson presents his views verges toward polemic rather than providing a responsible work of popular science.

Mithen’s expertise enabled him to spot some embarrassing factual errors in the treatment of human evolution (I can’t judge these issues):

I am tempted to think the same [i.e., the trend toward polemic] about Wilson’s characterization of the archaeological record for human evolution. While he correctly identifies the key themes of human evolution—big brains, bipedalism, control of fire, shift to a meatbased diet, adaptive flexibility—his account is marred by a succession of factual errors. “Spear points and arrowheads are among the earliest artifacts found in archaeological sites.” No, the earliest artifacts are from around 2.5 million years ago, but spear points are not made until a mere 250,000 years ago and arrowheads might have first been manufactured no longer ago than 20,000 years. “Archaeologists have found burials of massacred people to be a commonplace,” while “archaeological sites are strewn with the evidence of mass conflict.” No, both are quite rare, especially in prestate societies, and those that are known are difficult to interpret. “Homo erectus…was able to shape crude stone tools.” No, many of the hand axes made by Homo erectus are quite exquisite. “Axes and adzes [were] invented in the Neolithic” and “Neolithic toolmakers invented the concept of a hollow structure, with an outer and an inner surface.” No and no. These are elementary errors that could have been avoided by consulting any undergraduate textbook.

Elsewhere, the language Wilson employs provides a completely erroneous impression of the past. He remains wedded to antiquated phrases from a time when cultural evolution was envisaged as an inevitable progress toward Victorian values, as in the “dawn” of the Neolithic and the “ascent to civilization.” Wilson writes how Homo sapiens “slogged cautiously on foot” when dispersing from Africa; while this may have often been literally correct, the archaeological evidence—that Wilson goes on to accurately summarize—reflects an astonishingly rapid global dispersal with that of Australia at least involving the use of boats. On the next page, Homo sapiens have quickened their pace to become “skilled warriors” who outcompeted the Neanderthals. That phrase implies warfare and a distinct class of person within a tribalbased society specializing in combat: neither of these can find any supporting evidence in the Palaeolithic archaeological record.

Wilson’s factual inaccuracies and misrepresentations of the past are especially infuriating because in his own specialist field of insect evolution he meticulously attends to the data. . .

Finally, Mithen argues that Wilson’s treatment of human social evolution is superficial:

But what of human eusociality? While Wilson argues that the first three of his stages may be applicable to human evolution, he recognizes that the final two stages could have only happened in insects and other invertebrates. So the final section of his book returns to humans with a few short chapters that attempt, no less, to explain “What Is Human Nature?,” “How Culture Evolved,” “The Origins of Language,” “The Origins of Morality and Honor,” “The Origins of Religion,” and “The Origins of the Creative Arts,” before ending with “A New Enlightenment.”
Even for a double Pulitzer and Crafoord Prize winner such as Wilson this is too much to take on, especially as there have been several complete books published on each of these topics in recent years. There has been a great deal of work on these issues from an explicitly evolutionary perspective, some of which Wilson appears to be unaware of—such as that about the evolution of music and its relation to language. Hence one gets a taste of the issues involved but without much satisfaction, always sensing that one is engulfed by the enormity of the issues—perhaps rather like E.O. Wilson sampling the aphid excrement below the canopy of a rainforest.

Mithen concludes, as do I, that no matter how bad the book is, Wilson’s place in the pantheon of great biologists is secure.

It is often said that you are only as good as your last book. That too is now falsified: E.O. Wilson is far better than The Social Conquest of Earth. For me, he remains an intellectual hero.

Wilson is an engaging and admirable man who has done great work in conservation, in ant biology and systematics, and in synthesizing behavior and evolution. Pity that at the tail end of his career he’s engaged in such a debacle.

Another review of Wilson’s book, by James Fowler, a professor of medical genetics and political science at the University of California at San Diego, has just appeared in Nature. It’s more positive, but really is just a recounting of what’s in the book, though he does take Wilson to task for giving short shrift to inclusive fitness theory:

Many of Wilson’s ideas in this book will stand the test of time. However, he is perhaps a bit too assertive in the way he frames his theory. He is excessively critical of inclusive fitness theory, repeatedly claiming that it is “incorrect”, and saying that the literature on it has produced “meager” results. Yet inclusive fitness theory has prompted much empirical and theoretical investigation, with more than 1,000 articles published in the past 40 years. Albert Einstein, after all, didn’t disparage the numerous physics experiments showing that Isaac Newton’s simple formulae work remarkably well under specific conditions.

Wilson would, I am sure, object to this characterization on the grounds that inclusive fitness theory accounts for a much smaller subset of his own theory than Newton’s work does for Einstein’s. In fact, Wilson continually claims that inclusive fitness theory works only “under stringently narrow conditions”. But there is no empirical evidence for this.

37 thoughts on “The NYRB reviews E. O. Wilson’s latest

  1. On the reviewer, Steven Mithen:
    I have read his book, “After the Ice: A Global Human History, 20,000-5000 B.C.” (Harvard University Press, 2003). It’s a “popular” archaeology book that is written for an educated audience. His storytelling technique is a bit odd at first but it does work, and the book is packed with wonderful details, even if you’re not particularly into archaeology.

    Although E.O. Wilson’s contributions to entomology are obviously huge, I have never been a fan of his writings about humans. I gave up on him after “Biophilia”, which appears to be naive wishful thinking masquerading (not very well) as science.

    1. “After the Ice: A Global Human History, 20,000-5000 B.C.” (Harvard University Press, 2003)”

      Good book, I enjoyed it. Though it does get a little samey as he repeats a similar story for different continents. For a non-specialist audience it could have done with being half the length and not trying to be comprehensive.

  2. Evolution works by one simple rule: “Close enough is good enough”.

    Anything more than that is wishful thinking, i.e., religion.

    “Mother Nature” follows the “Cable Guy” motto: Let’s git ‘er done.

  3. The interesting question is – What logic is there is Wilson’s end of life polemic and crackpot ideas?

    The answers seem to include:
    – He is “backdoor” claiming his life’s work is the “secret” to human exceptionalism = eusociality
    – His moralizing agenda revealed a the end of a reader’s long pro-Wilson set of videos in a earlier post where Wilson claims the opposite of eusociality is “sin.”

  4. Dear Professor Coyne and readers of Why Evolution is True,
    Richard Dawkins has reviewed Wilson’s book in Prospect magazine.
    William Stewart

    1. Richard’s review was detailed in a dedicated post here on WEIT a few weeks ago. As you might expect, it is also on his own site RDFRS.

  5. There are more errors than just those. His section on war makes a few dubious claims that seem to go against a lot of what I’ve read by Steven Pinker and others on the subject.

    Wilson also thinks that philosophy is useless and that we will have a biological understanding of consciousness in about 20 years, this seems delusional to me because the hard problem of consciousness is not even close to be solved.

    The chapters on insects are interesting and well done at least..

    1. Wilson also thinks that philosophy is useless

      What do you know, seems Wilson are as incisive as the WEIT readership in some cases. (While no doubt more so in others and less so on paleontology.)

    2. The hard problem is not even close to being solved?

      On contrary consciousness is already essentially solved. We have the outline, we just need to fill in the details.

  6. “Spear points and arrowheads are among the earliest artifacts found in archaeological sites.” No, the earliest artifacts are from around 2.5 million years ago, but spear points are not made until a mere 250,000 years ago and arrowheads might have first been manufactured no longer ago than 20,000 years.

    Um, oh, wow. Even I, with a superficial interest in paleonthology, knew that spear poinst and arrow heads aren’t early. And while I haven’t latched on to the dating of the earliest spear points it is quite a memorable fact that arrows as a technology are as late as the last few ten thousand years.

    I mean, if you read some about human history, how can you not know this? Not one of Wilson’s interests, I take it.

  7. I was able to access the review without any particular problem. I proxied my connection through a campus system. Presumably the campus libary has a NYRB subscription that allow this kind of access.

    I mention this in case it is helpful to others.

  8. If anybody knows to what book(s) on the origins of the creative arts and music Mithen was referring, my interest is piqued.

    1. I’ve not read the article yet. But, the ‘I’m not going to get a balanced view here’ alarm bell rang upon noting the author: DS Wilson. I’ve been unimpressed with his ability to fairly treat the subject in the past. I suspect the picture of the consensus that he paints will not be the actual consensus position.

      1. Well, I found DS Wilson’s first misrepresentation in his second paragraph… DS Wilson implies that EO Wilson and Dawkins are equally outside the true consensus, which is complete BS. Sure, Dawkins’ views may not be in totally alignment with the consensus, but they are far, far closer to it than EO Wilson’s.

        It’s ironic too, that DS Wilson accuses Dawkins of unfairly borrowing the weight of the consensus to support his views and then distorts the consensus position to flog his own views. Dawkins freely admits in his review that authority is not a good way to support your argument. But, he also uses it to make an important point; that EO Wilson has ignored a huge amount of criticism.

        I remain unimpressed with DS Wilson’s scholarship. Reading his essays on group selection is like listening to a used car dealer talk about the cars on the lot.

  9. I’ve followed the controversy regarding the paper of Nowak and Wilson from the sidelines and have considered it a settled matter for a while now.
    Looking at the original paper in Nature this morning though, I just found out that the Templeton Foundation is acknowledged for providing funding. This only aggravates their case I’m afraid. Am I the only one who missed that in the first place?

  10. “One of the vehicles for that success, in Wilson’s view, was group selection, the differential reproduction of groups of individuals having different traits. (In this case, traits like altruism in humans or sterility in insects.) This puts him, at least in the insect case, at odds with the vast majority of other evolutionists, who see selection among groups as an ineffective way of evolving anything, and see no evidence that it has occurred in nature. Rather, selection based on relatedness—kin selection—seems more plausible.”

    In something like eusocial insects, how do you even distinguish the two? You’ve got kin groups in which reproduction is effectively a group-level rather than individual-level phenomenon.

    It seems to make more sense to think of this as group selection, to me (workers -don’t reproduce- so who gives a &*%^ about their reproductive interests? it’s just baffling, IMO, to study the fitness of sterile organisms), but the two explanations are ultimately… compatible. This is the basic motivation behind the model of selection presented by Nowak et al.

    That aside, I’ve no opinion on Wilson’s most recent book.

    1. There’s a lot of variation among social insect ‘societies’, and factors like group size, frequency and equality of colony fission, number and age(s) of reproductive females, number of males per queen, number of matings, rate of sporadic or regular egg-laying by unmated workers, type and distribution of food resources, winter die-off, interannual variability, pathogens and predators, etc…. can have distinct effects on what is being selected for.

      Disentangling all these effects is obviously not trivial, and we should expect it to take a while, but the number and social diversity of hymenopteran species (and other groups with origins of sociality) and relatively short generations means there’s a lot of statistical power to measure stuff.

      Given that it’s not only ‘queens’ that reproduce in all species that have them, and there are various continua between the honeybee ‘monarchy’ and more anarchic and democratic patterns, it’s not obvious that ‘groups’ even exist, as a class of competing and reproducing entities, but I think you’ll find that individuals and genes always do.

  11. And, because I can’t help myself… if you want to reject group selection you’re stuck with only allowing selection at the level of an individual gene. Or the level of an individual exon, codon, or nucleotide. If selection occurs at anything above the lowest possible level, it’s group selection.

    A fruit fly is a group. What is special about this particular level of grouping that leads us to give it a privileged status as the only legitimate focal point of evolutionary theory?

    1. Ah, no. The unit of selection is the gene, the level of selection is where the gene is expressed (i.e. the individual). The fruit fly ‘grouping’ is special because it’s the level that reproduces to put genes into the next generation.

      “If selection occurs at anything above the lowest possible level, it’s group selection.”
      – That’s very post-modern. Take a term that has a commonly accepted meaning and redefine it to support your argument.

      1. “Ah, no. The unit of selection is the gene, the level of selection is where the gene is expressed (i.e. the individual). The fruit fly ‘grouping’ is special because it’s the level that reproduces to put genes into the next generation.”

        That’s the basic classical view, yeah. However, it is at best a very rough simplification at this point. It relies on a “one gene, one protein, one phenotype” viewpoint that we know is flat-out wrong and on a simple -assumption- that the individual multicellular organism is the “right” level to look at.

        For instance, in the way you’ve phrased it here – the individual multicellular organism is the level that “reproduces to put genes into the next generation”… the next generation of -what-? Why, of individual organisms. You’re just begging the question.

        “- That’s very post-modern. Take a term that has a commonly accepted meaning and redefine it to support your argument.”

        🙂 Yeah, I guess it is silly to think that “group selection” ought to mean “selection among groups”.

        But, really, why should we just stick with a commonly accepted usage when it has false implications? The typical usage implies that multicellular organisms are fundamental, indivisible units, and that group selection is anomalous because it looks at, you know, groups. That just isn’t the case. A multicellular organism is a group (of organs, cells, etc., depending on which level of analysis you choose). There’s nothing remotely odd about looking at selection among groups. That’s what our normal, standard theory of selection does. We’re just not used to -thinking- of multicellular organisms as groups, whereas we are used to thinking of, e.g., a colony of ants as a group. That’s the difference, not that one is a group and the other isn’t; they’re both groups.

  12. I have read Wilson’s book.

    And I agree with Mithen’s critics, the book read as very agressive against kin-selection. I didn’t expect this to be so, disappointing to see an old respected professor being emotional about something that’s not very clear cut, or even wrong, something similar to Bart Ehrman’s defense of historicity of Jesus recently, or even Stephen Jay Gould’s emotional defense of punc-eq.

    Wilson definitely wants to leave something (I sense some jealousy to Dawkins) about human evolution in his old age, and he found this “group selection” might be his last hurrah (similar to Gould’s punc-eq).

    It was interesting read, bit and pieces about the human culture – but like Mithen said, it is not very clearly referenced. I got the feeling that Wilson try to get a lot of things under his seniority / well-known name. It is kind of selling short his own name.

    It is a pity.

    1. I’ve read EO Wilson’s book too. I found it a slightly difficult read; it wasn’t exactly well written, but it was fairly interesting.

      I find group selection to be more plausible than kin selection. How is an individual animal going to be able to calculate the degree of relatedness to another individual without a large and complex brain?

      Group selection sounds more plausible. Groups with genetically programmed altruism (not that we know if it is genetic) do better than other groups. Individuals with higher altruism do better reproductively within the group because they’ve survived with a disadvantage (like the peacock’s tail) and therefore must have superior genes. Ergo, they’re seen as better partners.

      1. The problem with group selection as an explanation for the evolution of altruism is surely that cheats prosper. A non altruisitc individual benefits from the altruism of others at no cost to itself and thereby sees its genes increase in frequency. This inevitably undermines the group level selection and pulls everything back to selfishness.
        Looking around I don’t see much evidence for your assertion that individuals with higher levels of altruism achieve better reproductive success – i am sure that is not the case in Humans!

  13. The comparison of human groups as “eusocial” is an interesting hypothesis. Wilson is hinting that we are evolutionary success because we become eusocial, in the same way ants and termites are successful because they became eusocial insect.

    This is interesting. He develops an idea of something like eusociality among intelligent mammals, sketchily. A deeper study comparing termites to naked-mole rats first will be an interesting first step.

    Rather than jumping directly to human.
    In human society, the “atoms” are very different, individually we are much more complex, and the replicator at work in human societies are cultural memes, no longer genetic, this will warrant a very different approach.

    And, many times it is clear that Wilson is out of his depth. And taking the cardinal sins of insisting things on without proper references.

    But, the hypothesis is interesting: human societies, are they eusocial groups? What are the rules? (can not be as simple as Wilson said: thing that good for the groups are the good things, that are good for individuals are the sins). This is a broad question which a lot of papers in anthropology, sociology, humanities are already written.

    Papers that Wilson does not seems to care.

    And the worst? All of this does not necessarily dependent on the acceptance of “group evolution”, his pet theory. Kin selection wins in insects, in human societies what’s needed is a modified kin-selection evolution, where the replicator is memetic, and horizontal transfers are common.

    Maybe the differences in levels make comparison of human societies to social insects eusocialities useless, but it could be a start.
    Some PhDs might go here.

Leave a Reply