Apr 19 2012

Dear E. O. Wilson: Please retire or stick to ants

Tonight I went to a talk at Seattle Town Hall by E. O. Wilson, one of the most famous evolutionary biologists still alive today. I admit I went for two different reasons. One, Wilson is super famous and also very old, and I wanted a chance to see him speak because another chance might not come. But two, I saw that the topic was how group selection shaped human evolution, and I wanted to see what controversial arguments he would make.

Controversial because Wilson has recently been stirring the pot by trumpeting group selection and saying kin selection has been debunked. I don’t want to rehash the whole event, but Carl Zimmer has a good summary in the New York Times. The basic thing you need to know is that most biologists consider group selection to only occur in very rare and specific circumstances, and that selection usually takes place at the level of the individual or the gene.

But you wouldn’t know that from the talk. Wilson asserted that his controversial Nature paper definitively overturned kin selection theory and that “no one” responded to his critique of kin selection. This set off a red flag in my head, because I definitely remembered reading criticism of the paper at least in the blogosphere. I grabbed my phone and instantly dug up this critique by Jerry Coyne and this one by Richard Dawkins.

But maybe he meant a published critique. So I googled “response to Nowak 2010″ and instantly found a list of papers also published in Nature criticizing his paper:

Abbot, P., Abe, J., Alcock, J., Alizon, S., Alpedrinha, J., Andersson, M., Andre, J., van Baalen, M., Balloux, F., Balshine, S., Barton, N., Beukeboom, L., Biernaskie, J., Bilde, T., Borgia, G., Breed, M., Brown, S., Bshary, R., Buckling, A., Burley, N., Burton-Chellew, M., Cant, M., Chapuisat, M., Charnov, E., Clutton-Brock, T., Cockburn, A., Cole, B., Colegrave, N., Cosmides, L., Couzin, I., Coyne, J., Creel, S., Crespi, B., Curry, R., Dall, S., Day, T., Dickinson, J., Dugatkin, L., Mouden, C., Emlen, S., Evans, J., Ferriere, R., Field, J., Foitzik, S., Foster, K., Foster, W., Fox, C., Gadau, J., Gandon, S., Gardner, A., Gardner, M., Getty, T., Goodisman, M., Grafen, A., Grosberg, R., Grozinger, C., Gouyon, P., Gwynne, D., Harvey, P., Hatchwell, B., Heinze, J., Helantera, H., Helms, K., Hill, K., Jiricny, N., Johnstone, R., Kacelnik, A., Kiers, E., Kokko, H., Komdeur, J., Korb, J., Kronauer, D., Kümmerli, R., Lehmann, L., Linksvayer, T., Lion, S., Lyon, B., Marshall, J., McElreath, R., Michalakis, Y., Michod, R., Mock, D., Monnin, T., Montgomerie, R., Moore, A., Mueller, U., Noë, R., Okasha, S., Pamilo, P., Parker, G., Pedersen, J., Pen, I., Pfennig, D., Queller, D., Rankin, D., Reece, S., Reeve, H., Reuter, M., Roberts, G., Robson, S., Roze, D., Rousset, F., Rueppell, O., Sachs, J., Santorelli, L., Schmid-Hempel, P., Schwarz, M., Scott-Phillips, T., Shellmann-Sherman, J., Sherman, P., Shuker, D., Smith, J., Spagna, J., Strassmann, B., Suarez, A., Sundström, L., Taborsky, M., Taylor, P., Thompson, G., Tooby, J., Tsutsui, N., Tsuji, K., Turillazzi, S., Úbeda, F., Vargo, E., Voelkl, B., Wenseleers, T., West, S., West-Eberhard, M., Westneat, D., Wiernasz, D., Wild, G., Wrangham, R., Young, A., Zeh, D., Zeh, J., & Zink, A. (2011). Inclusive fitness theory and eusocialityNature, 471 (7339) DOI: 10.1038/nature09831

Boomsma, J., Beekman, M., Cornwallis, C., Griffin, A., Holman, L., Hughes, W., Keller, L., Oldroyd, B., & Ratnieks, F. (2011). Only full-sibling families evolved eusociality Nature, 471 (7339) DOI: 10.1038/nature09832

Strassmann, J., Page, R., Robinson, G., & Seeley, T. (2011). Kin selection and eusociality Nature, 471 (7339) DOI:10.1038/nature09833

Ferriere, R., & Michod, R. (2011). Inclusive fitness in evolution Nature, 471 (7339) DOI: 10.1038/nature09834

Herre, E., & Wcislo, W. (2011). In defence of inclusive fitness theory Nature, 471 (7339) DOI:10.1038/nature09835

Yeah, and he said “no one” responded. And it’s not just that Wilson is out of the loop – he came off as being purposefully disingenuous. Not only did he publish a response to the responses (Nowak, M., Tarnita, C., & Wilson, E. (2011). Nowak et al. reply Nature, 471 (7339) DOI: 10.1038/nature09836), but during the Q&A he changed his story and said that people did respond but they were 1. Wrong and 2. In the minority. Even though 1. He never explained why their critiques were incorrect and 2. The vast majority of biologists disagree with his views of group selection and the authors of the critiques weren’t random nobodies; they were very important and accomplished researchers.

I want to give Wilson the benefit of the doubt. Maybe when he said “no one responded” he meant “no one responded in a way that we think invalidates our hypothesis.” But even then, the rest of his talk was incredibly sloppy. He asserted that human eusociality evolved via group selection, but didn’t offer a shred of evidence the whole time. No proposed mechanism, no genetic evidence, nothing. He just waved the Wand of Group Selection and asserted it happened. He asserted that humans first ate cooked meat by scavenging carcasses from wildfires. That’s one hypothesis among many, but he presented it as a known truth and gave no evidence or citation for it. He asserted that eusociality only evolved recently but again gave absolutely no evidence as to why he thought so. I mean, maybe he’s right, but eusociality isn’t exactly something that fossilizes well, so it could have possibly existed in past species. At least put some sort of qualifier or explanation of your reasoning out there.

When someone in the Q&A asked him to explain why people disagree with group selection so much, he didn’t explain the objections or why he thinks kin selection was wrong. He instead stated that his paper was reviewed by a mathematician from Harvard and that it got into the prestigious journal Nature. Therefore it is right, or something. Here’s an alternative hypothesis: Your paper got published in Nature because you’re insanely famous and it was incredibly controversial, which Nature eats up. Nature is more about prestige and sexy topics than good science nowadays. Its retraction rate has increased ten fold in the last ten years when the number of papers published in all journals has only increased by 44%.

Look, I’m not a priori against group selection. Maybe Wilson is right and group selection is applicable in more situations that we currently think. But I’m not going to accept it until he presents compelling evidence, which he utterly failed to do. You can’t just say “Harvard” and “Nature” and leave it at that.

The most irritating thing about the night was that this was a talk given to an educated general public. These people are smart enough to appreciate science and know Wilson is a famous scientist, so they’re going to believe whatever he says. On the way out people were raving about how interesting the talk was. But he presented none of the controversy, no evidence, no reasoning, no citations, no qualifiers…nothing. I understand that a talk to the general public isn’t going to get into extreme detail, but asserting your incredibly controversial ideas as scientific fact is incredibly dangerous. This talk reminded me more of stuff I’ve seen from creationists and climate denialists than scientists.

Honestly, I left feeling bad for him. E. O. Wilson made huge advances to evolutionary biology, sociobiology, and conservation. “Huge advances” is an understatement. But tonight he went outside his expertise and left science behind, and it was kind of embarrassing. I would have loved for him to give an hour long talk about ants instead.


Skip to comment form

  1. 1
    Donovan of NH

    Hey Jen,

    You said group selection was “rare.” I was under the impression that it just didn’t happen; couldn’t, really. Stretch my mind as I might, I just can’t think of any mechanism that would allow for group selection. What rare cases are there?

    And I hate explanations like scavenging fire victims to learn about cooked meat for human evolution. It makes early hominids seem like blank slates fresh from the dust of Eden. There are much better explanations.

  2. 2
    Adam G

    Aw man, I leave town for 2 days and EO shows up to give an awesomely bad talk? Just my luck, trainwrecky talks are my favorite. At this rate the next time I go away Luc Montagnier will come to town to deliver a paean to homeopathy :-/

  3. 3

    Well, I can think of one phenomenon that can be attributed to group selection: the rise of humans who clearly benefit from being in groups.
    Other than that, I’m not a scientist and havn’t run in to any examples. And the rise of humans is a pretty spectacular one time event.
    So yeah, sounds like Wilson was way off.

  4. 4

    So he’s unscientific because he didn’t footnote a popular talk, but instead pushed his own views?


    If I’m the kind of guy who gets along with others, will my family benefit? Yes. Kin selection to group selection in one jump.

    Now remember, your refutation can’t depend on anything but your own footnoted footnotes.

    And argumentum ad auctoritatem doesn’t count…

  5. 5

    It sort of depends on who you ask. D. S. Wilson will say that group selection / multi-level selection (MLS) does occur and that it is different than individual selection of any kind, including kin selection (and he’ll claim, falsely, that nearly everyone agrees with him on this). The most well-known and easily understood example of this is probably the ‘hay stack’ model, if you want to read more about it.

    On the other hand, my opinion (following the work of people like Stuart West, Andy Gardner, and Steven Franks) is that group selection / MLS is equivalent mathematically to kin selection, and so saying that “group selection” occurs is a trivially true statement devoid of useful content. The contention is that because the models are equivalent, but kin selection models are easier to use and provide more biological insight, the focus on group selection is useless at best and possibly harmful. I’ve written about this a few times, most recently here if you want to find out about this side of it.

  6. 6

    Heh. It was just two days ago that my cosmology teacher said “If something is written in Nature, you can assume that it’s wrong”.

  7. 7
    Chris Lawson

    You look to be trolling here, but I’ll make one point anyway in the optimistic hope that you have a genuine interest: if kin selection was merely a subset of group selection, as you imply, then there would be a much smaller controversy here. But E.O. Wilson claims that kin selection cannot explain eusociality. It is his critics who are saying that Wilson’s formulation of group selection is just a mathematical variation on kin selection.

    And you complain about Jen being hypocritical about footnotes when (i) she herself said she didn’t expect footnotes in a public talk, and (ii) she provided several references in the text to support her statements. By this measure alone, you are an obnoxious bullshitter.

  8. 8
    Chris Lawson


    That’s not what group selection means. If group selection just meant “groups of organisms with these genes do better”, then it would not be controversial. Wikipedia has a decent article here.

  9. 9
    Chris Lawson

    Sorry, mucked up the href. Google is your friend.

  10. 10
    Chris Lawson


    I agree with you that I can’t really see a mechanism that would allow group selection to work and although I haven’t really gone into the maths, I find Wilson’s arguments unconvincing. (I thought about checking out the maths myself, but couldn’t summon the enthusiasm — especially as half the argument seems to be about what each mathematical term stands for.)

    Having said that, I don’t feel we can completely write off group selection. There are still a lot of explanatory gaps in evolutionary theory, and I am especially intrigued by the possibilities of finding evolutionary forces that might be invisible to us on the generation-by-generation scale but very important on the geological scale. (Rather like relativistic effects that are immeasurably small in our day to day Newtonian awareness, but extremely important when you start looking across the universe.)

    As I say, this is not really me agreeing with Wilson’s group selection (or anyone else’s for that matter).

  11. 11
    Chris Lawson

    It really irks me that Wilson has been draping himself in the heresy flag. It’s cheap rhetoric, especially as the worst he has had to suffer is reading people disagreeing with him. It’s also appallingly inaccurate because, as Wilson would remember, until the 1960s group selection was one of the dominant concepts in evolutionary theory.

    Wilson isn’t really overthrowing an outdated model, he’s trying to re-establish an even older one.

  12. 12

    Theoretically, groups selection might happen, but the circumstances for it to be of any importance whatsoever are so specific… I don’t really see how it could have played an important role in human evolution.

  13. 13
    1000 Needles

    Congratulations on beating your writer’s block and coming back swinging!

  14. 14

    “So he’s unscientific because he didn’t footnote a popular talk, but instead pushed his own views?”
    No, because he didn’t provide evidence and didn’t address objections of his colleagues.
    I’m not a biologist at all, but that point of JMcC was very hard to miss.
    So congratulations for achieving exactly that.

  15. 15

    I enjoyed your writing on this, Steven. One distinction that I think needs to be made is the difference between kin selection and inclusive fitness theory. Inclusive fitness theory partitions the fitness of individuals into two components: 1) direct fitness is the fitness of an individual from personal reproduction (how many kids do you make?), and 2) indirect fitness is the fitness of an individual from the reproduction of other individuals with shared alleles (usually kin). Kin selection is natural selection that increases the indirect component of inclusive fitness by increasing the fitness of kin. Group/ML selection is mathematically equivalent to inclusive fitness theory, but not necessarily kin selection.

    This is where Wilson’s arguments for group selection bother me; at first I assumed he was convinced and mistaken, but he’s starting to come off as disingenuous by confusing kin selection and inclusive fitness theory. For example, in his most recent book, he writes that inclusive fitness theory is grounded in kin selection (“The Social Conquest of Earth,” last chapter). The only way I can think of that this is true is to say that Hamilton (as I recall) came up with the idea of kin selection before generalizing to inclusive fitness theory — otherwise, conceptually, kin selection is a special (and empirically well-supported) subset of inclusive fitness theory. Perhaps I’m giving Wilson too much credit, but I think he knows better.

  16. 16

    Great post, Jen!

  17. 17
    Reginald Selkirk

    and that it got into the prestigious journal Nature.

    So did Jacques Benveniste’s 1988 paper on homeopathy. Woo hoo.
    E.O. Wilson is a very likable guy. He regularly gets selected to lists of “Most (adjective) Atheists” even though he has identified himself as a “Provisional Deist,” i.e. he isn’t an atheist at all.

  18. 18

    Hey Brad,

    You’re right that more attention needs to be paid to the difference, and I’m guilty of using the terms interchangeably when they’re not; I’d like to thank you for pointing that out. When I’m paying attention and being precise, I try to use inclusive fitness, as you mention, as the quantity (direct + indirect fitness) being maximized by selection, and kin selection as the name for the process by which the indirect fitness aspect of inclusive fitness is maximized (though it has also been used in a broader sense of the process of maximizing the totality of inclusive fitness; see the discussion in West et al. 2007, p. 418). The clearest single sentence about this that I’ve seen recently is: “Maynard Smith (1964) coined the phrase ‘kin
    selection’ to describe the process of natural selection operating through indirect fitness effects” (p. 1021).

    Thus, I have to disagree that MLS isn’t equivalent to kin selection. Group selection, like kin selection, is the *process* that is proposed to maximize inclusive fitness. (“Group selection and kin selection are simply different approaches to describing the same biological process”; West, El Mouden, and Gardner, 2011. Frank, 2011 and West, Gardner, and Wild 2011 cover the math of comparing MLS and kin selection as processes, going by way of the Price equation).

    Regarding Hamilton, I’m not entirely clear on how all that went down. Re-reading for this comment, the feeling I get is that Hamilton conceived of inclusive fitness and tried to show how natural selection maximizes it; then, from Maynard Smith and others (including, later, Hamilton as well?) we got the term kin selection and the analysis of the dynamics of maximizing inclusive fitness. I kind of gave up untangling the history after that, though. If you have any insight into that, I’d love to hear it!

  19. 19

    Hey Steven,

    I suspect you’re probably right regarding Hamilton’s thought process. Unfortunately, I don’t have the details despite my inordinate fondness for Hamilton and his work. The stories I’ve been told about his character are charming. He pursued his graduate work on the problem of altruism despite (well-intentioned, I’m sure) warnings from his professors that such a project could turn out very, very badly.

    You make an excellent point that kin selection and MLS are both processes; inclusive fitness is just a division of fitness components, so my describing MLS and inclusive fitness theory as being mathematically equivalent was incorrect. I believe that your description of MLS as being mathematically equivalent to the process of maximizing inclusive fitness is accurate (and probably the best way of stating the relationship, actually). But I must protest that kin selection does not encompass all of the processes for maximizing indirect fitness, unless you define it as such. The way in which I define kin selection is as selection on the indirect fitness component in such a way as to maximize the spread of an individual’s alleles (direct fitness), plus the alleles found in kin, which are specifically alleles that are shared because they are identical by descent with the alleles of the individual (I think the last part is where our thoughts are different). But the indirect component of inclusive fitness need not only include kin; green-beard effects, e.g., occur when individuals recognize shared alleles in non-kin. It’s a subtle distinction, but I think that this is where the claim that MLS is superior gets weaseled into the debate – when you leave out other indirect fitness components besides kin selection, the mathematical equivalence is lost.

    There is a recent (and awesome!) article in PNAS by David Queller, who specifies two additional forms of selection that act on indirect fitness: kith selection and kind selection .



  20. 20

    Oops, the link didn’t show up. Anyway, the title of Queller’s paper is “Expanded social fitness and Hamilton’s rule for kin, kith, and kind”

  21. 21

    I wish you could have been there. There was other assorted wackiness I didn’t include in the post, like some weird slide about how since dinosaurs didn’t have big brains and look like humanoid lizards then they weren’t eusocial (?!?!). And he went on this weird tangent about the complexity of art. I have no idea how it related to the rest of his talk.

  22. 22
    Azkyroth Drinked the Grammar Too :)

    I thought “eusocial” referred to the social structure seen in bees and ants. Aren’t there important differences with human social behavior?

  23. 23
    Greg Laden

    Very interesting.

    For a couple of semesters, I taught in the lecture hall next to Ed’s. He would often end lecture early and we would patiently listen to the applause from the enraptured students, then I would end my lecture mentioning something about the applause and my class would end with laughter.

    One really interesting feature of his class: He would get to the end of his lecture notes and end. This would be several days before the end of semester. But when he was done, he was done and there would be no more class meetings.

  24. 24

    Thanks for the discussion, Brad. I don’t want to turn Jen’s blog into a sprawling discussion on the subtleties of defining kin selection, so I’ll leave it here. I think that there’s more to chat about here, though, and if you ever want to wander over to my blog or find me on Twitter, I’d be happy to hash it out at length. :-)

    And thanks for the Queller link! I remember seeing that paper come out and then I forgot to read it. I’ll have to put it in the pile now (which, of course, means that I’ll get to it by some time in 2014…*sigh*).

  25. 25
    Richard Dawkins

    Congratulations, Jen, on this excellent article. Wilson has long been a hero of mine, and I am deeply saddened by this nonsense he is now talking, and the high-handed way he simply ignores the massive objections to his new-found enthusiasm for group selection and the way he dismisses kin selection. Kin selection is not something extra, over and above standard neo-Darwinism. It is logically entailed by standard neo-Darwinism. You might as well say you are in favour of Euclidean geometry EXCEPT Pythagoras’ Theorem. You might as well waste your time measuring lots of right angle triangles to see whether Pythagoras got it right.

    Richard Dawkins

  26. 26
    Donovan of NH

    I do agree it’s be best not to run this discussion too long. I study ecology, so I am familiar with the different modes of selection even if I draw a blank on the proper terms.

    I do see that no examples of group selection have popped up, so I will assume that they are not common claims. I’m going to have to track down Wilson’s examples and see if group selection makes sense. I don’t see how it could, but he deserves my ear.

  27. 27

    I just returned from Wilson’s talk in San Francisco. I’m not a biologist, just a relatively well informed civilian who’s done some reading. I enjoyed Wilson’s talk myself, and have a couple of thoughts about this thread:

    He definitely was fairly dismissive of the response to the Nature article, but in SF at least he was more specific about what he thought hadn’t been responded too, in particular the mathematical argument. That mathematical discussion is out of my league and I’ll leave it to others to decide if it’s been answered or not.

    I think it’s a little harsh to judge how much supporting evidence he gave in a hour long talk for a general audience on a book tour (assuming the Seattle event was similar to the one in SF). I would reserve that judgement for the contents of the book he was glossing over in the talk.

    He kept mentioning group selection, but he also kept coming back to the term multi-level selection. I haven’t read the book yet myself, but I frankly find his ideas interesting. Kin selection seemed so compelling when I first read about the haploid-diploid case, but it sounds much less compelling with other eusocial species, particularly humans. I’m skeptical of building the evolution of human society on something as thin sounding as kin selection.

    Given what is known about human cognitive biases, the amount of criticism Wilson is getting would be expected whether he is correct or not. There are a lot of people with years of career invested in current theories and they will be hard to convince. If I’m hearing what Wilson is saying correctly, I might put it differently. Calling “group selection” seems to be calling up the defenses.

    In addition to multi-level, he mentioned superorganisms quite a bit. Something like the queen is the core of the colony superorganism and the workers are external expressions of the phenotype.

    I’m always suspicious of when we get too attached to our words for things over observing the things themselves. I like to point out to people that species don’t actually exist in nature. The whole of taxonomy is to help us talk sensibly about the diversity of nature, but it’s a shorthand for something more complex which is actually out there. Just like nature doesn’t need the concepts of photon and electron to make lightning, the boundaries of our terminology is a convenience for our thinking, not for the phenomena themselves.

    I don’t know if Wilson would agree with me, but what I’ve been thinking ever since hearing him speak is that the boundary of the biological concept of the “individual” or “organism” is also one of our conveniences rather than a hard boundary which actually exists in nature. What would our explanations sound like if we tried to express them in language which allowed the “individual” to extend up through insect colonies to loose cultural groups of humans? Multi-level is important. He definitely talked about human nature being divided between being individuals and members of a group. (eusocial insects are a bit more “all-in” on that one.) All of the individual selection pressures apply, but they are affected by our membership in something larger than ourselves.

    With Richard Dawkins posting here, I’m a little reluctant to toss around the term “extended phenotype”, because I doubt my grasp of the concept is up to sufficient level for any argument, but exactly where does the extension of the extended phenotype end? If the term “group” is so reviled, is there another term for the sorts of eusocial associations which Wilson is suggesting might be objects of selection? Could the “extended phenotype” include other individual organisms?

    A human organism is already “more complicated than that” with five pounds or so of assorted bacteria without which we can’t survive. A human organism is also incapable of actually surviving as as functioning human without being associated with other human organisms, just like it’s hard for human liver cells to survive removed from the human body.

    Maybe “group selection” is a lousy term, but it seems like it might be fruitful to look for something more going on which can only be seen with a shift in perspective from the billiard-ball-physics view of individual organisms.

  28. 28

    “Please retire or stick to ants”?

    Wilson may be wrong, but that doesn’t give you the right to be rude.

  29. 29

    Respected scientists misleading the public is worse.

  30. 30

    I for one don’t wish Dr. Wilson continue giving talks and having an opinion even if it is wrong. I have learned understanding why some concepts are wrong as much as understanding the correct concepts.

    Telling people to shut up is a bad habit. It is ok to kindly point a mistaken fact or faulty reasoning. Another thing is to tell Dr. Wilson that he needs to shut up after a lifetime of achievements and often thankless and penniless work.

    Talk about privilege.

  31. 31

    (please ignore above)

    I for one do wish Dr. Wilson to continue giving his talks and having an opinion even if he is wrong. I have learned some concepts from understanding why they are wrong as much as understanding the correct concepts in the first place.

    Telling people to shut up is a bad habit. It is ok to kindly point a mistaken fact or faulty reasoning. Another thing is to tell Dr. Wilson that he needs to shut up after a lifetime of achievements and often thankless and penniless work.

    Talk about privilege.


  32. 32
    Lucas Black

    This is going to show my total ignorance on the subject (I arrived here following links) because I am not an evolutionary biologist. What is the difference between kin selection and group selection? Are they just different names for different types of groups or is it something deeper that that?

    If it is something deeper what is this difference?

    Please keep it simple, I spend over 40hours a week in a dark room looking down a microscope, I don’t get much chance for reading.

  33. 33

    The standard view of natural selection is that individual organisms live or die, hence pass on their genes or not, due to natural selection forces. So behaviors which preserve their own lives and reproduction would be selected for. Kin selection is sort of a minor exception which says behaviors which preserve siblings and cousins can also be selected for, since a portion of the genes are shared.
    Group selection is the hypothesis that natural selection can operate on whole groups as a unit rather than only on the individual organisms.

  34. 34
    F [i'm not here, i'm gone]

    That would then be the privilege to spew whatever that would get a less famous and accomplished scientist labeled a crank? Or the privilege to lie about the acceptance of one’s theory and the existence of criticism thereof? Or the privilege to dismiss such criticism and not respond to it in any meaningful way?

    Because that’s how science is done, right?

    It might do to remove your head from your fifth point of contact before ascribing wherein does privilege lie.

  35. 35
    Adam Rutherford

    While I don’t disagree with any of the sentiments written about EO Wilson necessarily, I feel obliged to contradict the basis of your slur at the journal Nature.

    Full disclosure: I am employed by Nature and have been for 10 years. So you may factor that in my response, though my aim is to correct a factual error.

    ‘Its retraction rate has increased ten fold in the last ten years when its number of papers published has only increased by 44%.’

    If you follow the link to the NYT article, it states that ‘the journal Nature reported that published retractions had increased tenfold over the past decade, while the number of published papers had increased by just 44 percent.’ It is fairly clear that this is referring to ALL published papers (in the Web of Science)

    If you follow the link that this article includes, it is the source, a news feature in Nature

    which states ‘In the early 2000s, only about 30 retraction notices appeared annually. This year, the Web of Science is on track to index more than 400 (see ‘Rise of the retractions’) — even though the total number of papers published has risen by only 44% over the past decade.’ The graphic confirms that, showing data from Web of Science and PubMed notices on retractions.

    Your assertion that ‘Nature is more about prestige and sexy topics than good science nowadays.’ is, I believe, an opinion that is demonstrably false. However I’m not willing to discuss this. But it does look like you have misread and thus fabricated support for your opinion (notably with the word ‘its’ rather than ‘it reports that’).I have no data to hand on what Nature’s actual retraction rate is, I am merely correcting what you have written based on what you have linked to apparently in support.

    Best wishes

    Adam Rutherford

  36. 36
    Reginald Selkirk

    The first amendment gives the right to be rude. If you don’t like free speech you can go suck on a lemon.

  37. 37
    M can help you with that.

    Do these describe a difference in process, or a difference in model? I’m also not a biologist, so I have a bit more trouble working through the descriptions of group vs. kin selection — but it almost looks (to borrow a metaphor from my field) that the difference is a shift in frame of reference rather than descriptions of different processes. Either it can be a matter of evolutionary advantage for some individuals to give up having offspring to help genetically-related members of the group, or it can be a matter of evolutionary advantage for an individual to have offspring who will help spread the first individual’s genes through fertile offspring instead of heading off on their own. I see the first described as group selection, the latter as kin selection; is this a mistranslation into non-biologist-speak of two descriptions of kin selection? Because it seems that the same process is involved, but described in terms of the “fitness” of two different generations. Unless there’s a claim of “fitness” as an inherent property of some sort (which I’ve never seen in serious discussions of evolution), it sounds like these are just two different models/descriptions of the same sequence of events. I could understand the “kin selection” model being more mathematically convenient; is there an entirely different process which is described as “group selection”?

  38. 38
    Azkyroth Drinked the Grammar Too :)

    Oh fuck off, you smarmy little shit. :)

  39. 39
    Azkyroth Drinked the Grammar Too :)

    Wait, you’re calling an honest mistake based on a misparsing of an awkward phrasing a “slur” and equating misreading with “fabrication?”

    That’s not very intellectually honest of you….

  40. 40
    Azkyroth Drinked the Grammar Too :)

    Also, the “you’re wrong and I’m not willing to discuss this” part is certainly going a long way towards dispelling negative opinions of Nature.

  41. 41

    To get the idea of kin selection, I think it helps to shift the focus from individual to the gene. At the bottom level, the organism can be viewed as a vehicle for genes to make it into future generations. Genes which produce behavioral traits which insure the survival and reproduction of the individual organism will be selected for. When a species adopts a strategy (adopt via genetics, not intention) of remaining in family related groups rather than scattering as individuals, then a gene which produces a behavior which helps the offspring of close relatives, who very likely also carry the same gene, will also be selected for, even if the behavior isn’t necessarily advantageous for the individual organism. Defense of the nest to the point of death might make no sense from an individual viewpoint if the individual does not itself have offspring in the nest, but it might make sense from a gene-centric point of view, since there is a high likelihood of other instances in the nest of the same gene which produces the behavior. In that case, the defense to the point of death might help insure that the other copies of the gene survive.

    Group selection would be the idea that the selection process is working on the survival of whole groups of individuals, such as whole colonies, rather than individual carriers of the genes.

    It’s ironic that hymenoptera, bees/ants/wasps, while the easiest place to make a strong case for kin selection due to the high relatedness, is also, due to the tight organisation, the easiest place to make a case for a super-organism which might be selected. So while it’s the easiest place for Wilson to argue for his group selection, it’s also the strongest place for kin-selection theory to argue that the idea of group selection isn’t necessary.

    I think the main problem in the controversy is that Wilson is currently going around presenting the issue as if group selection has won the day and, as he said last night, “it’s time to clean the wreckage of kin-selection off of the road”. While I think it’s always interesting to attempt a re-think of settled theories to see if some new insight can be gained from a fresh perspective, Wilson is not helping his case with his offhand dismissal of much of current evolutionary biology as irrelevant.

  42. 42
    Adam Rutherford

    No, as I said, I am correcting the statement of fact which was erroneous, and was written in support of a disparaging criticism, which I think can fairly be called a slur. Perhaps it was an honest mistake (and kudos for linking in order to make accountability easy), but it still constitutes a fabrication, unless I am more mistaken than God. I’m not judgmental or grouchy about it, but I think it’s important to correct this type of chinese whisper.

    As for your second remark, criticism of Nature is absolutely fair game, and has its place. I’m merely declaring that it’s not here, at least not for me.

  43. 43

    Slightly off the main topic, but I was interested to see the link to the piece about cooking and human evolution. I wrote something about this recently here. In short, I describe reasons for supposing that our experience of enjoyment when eating cooked foods might be an evolved adaptation. Its quite far removed from my own field, but I am curious to know if somebody thinks there is anything worth investigating in my suggestion.

  44. 44
    Chris Lawson

    No, Adam, you should know that calling something a fabrication is equivalent to calling it fraud, especially in scientific circles. I checked your reference and you had every right to correct Jen’s mistake, but you should retract the “fabrication” statement rather than dig yourself a deeper hole.

  45. 45
    Adam Rutherford

    That was not my intention, and if that is the read meaning, I do indeed withdraw the word ‘fabrication’. Apologies Jen, I was using the word fabricate in the sense of creating, not as in to deliberately mislead. I amend the sentence to:
    ‘But it does look like you have misread and thus generated support for your opinion (notably with the word ‘its’ rather than ‘it reports that’)’.

  46. 46

    Thanks Steven! Take care.

  47. 47
    M can help you with that.

    It still looks like different descriptions of the same process to my non-biologist self. As I understand it, “selection” is mostly an abstraction for “stuff that happens that affects which genes get passed on”; it the difference between kin and group selection as simple as describing the “selection” model as acting on different subsets of a species (individual vs. immediate family vs. broad somewhat-related group)? I can imagine how there could be some pretty divergent patterns in terms of actually making use of the models, but do “selection acts on individuals” and “selection acts on groups” refer to different ideas about how evolution happens or different models of a process it’s convenient to call “selection”? It’s my physics-affected brain, again — I can simultaneously say that a heliocentric model of the solar system is much easier to fit to the messy reality and that a geocentric model can be useful for certain contexts as long as it’s done right (even if the heliocentric model is in a very real sense “more right”). Is it the same sort of thing, where group selection (by modeling selection as a group-level phenomenon) just becomes an unnecessarily messy and convoluted way to describe a process that’s usually more effective to model on a kin level?

  48. 48

    M Groesbeck: “Is it the same sort of thing, where group selection … just becomes an unnecessarily messy and convoluted way to describe a process that’s usually more effective to model on a kin level?”

    That’s pretty much the standard view in mainstream biology.

    Wilson is saying that there are aspects of what happens at the eusocial level of organisation which he doesn’t think are sufficiently explained by building up from the “atomic” view of kin selection.
    I haven’t read his book yet, but I’m curious to see what he has to say about.

    It’s not like he’s some young upstart, Wilson gave what is usually cited as the standard definition of Eusociality back in 1971, but the way he is pushing has most biologists puzzled.

  49. 49

    If anyone missed my scathing review of “Social Conquest,” please take a peak here http://www.amazon.com/gp/cdp/member-reviews/A2MBWOEN5B5PC9/ref=cm_cr_pr_auth_rev?ie=UTF8&sort_by=MostRecentReview

    And then hurry over for some followup here: http://www.apesantsandancestors.com/

    Not to be missed

    John Wylie

  50. 50

    Just so long as the good doctor doesn’t turn to writing more fiction in his retirement. Anthill plodded. I couldn’t tell you if it plodded to any particular destination because neither myself, nor the two or three friends (entomologists all) that I passed it on to, could get in to it, let alone finish it.

  51. 51

    “If I’m the kind of guy who gets along with others, will my family benefit? Yes. Kin selection to group selection in one jump.”

    No. Not even close. There is nothing remotely related to group selection in that statement.

    For it to even resemble selection of any kind, you would have to assert that you being a nice guy entails the rest of your family getting laid more often, and subsequently pushing out more kids.

    That would still make it a somewhat bizarre instance of kin selection, since what you’d actually be increasing is the odds of some of your non-fixed genes making it into new generations.

    I don’t believe I’ve ever seen a group selection proponent cite an example of the phenomenon that did not demonstrate that said proponent doesn’t even know what group selection is.

  52. 52

    Slur and fabrication, really? Because I read an article in a way that it is very easy to interpret it? Wow.

    I edited my post to be more accurate about the 44% factoid, but your way of dealing with this issue doesn’t make Nature look any better. Congrats.

  53. 53
    John Wylie


    No, his novel is the key to his depraved Social Darwinist ant-myth about human nature. In his Social Dominance book he reigns it all back to seem more reasonable. -John

  54. 54
    Adam Rutherford

    Slur is a word meaning a disparaging comment. You wrote: ‘Nature is more about prestige and sexy topics than good science nowadays.’ So I’m not sure how that doesn’t qualify as a slur. As for ‘fabricate’, I explained in the responses that my intended meaning was as ‘create’, rather to invent with intended deceit. If that was not clear, I apologise, as I presume it was an error in comprehension.

    My intention was not to create a fuss, nor to upset, merely to correct said error. If I have done Nature a diservice in doing so, then that is also unfortunate.

  55. 55
    John Wylie

    Richard, check out my blog response to Wilson’s book. By the way do you remember Chris Wilcox from Oxford days? He is my best friend in Washington – and I met out at the first HBES meeting, although I’m sure you don’t remember me. http://www.apesantsandancestors.com/

    – John Wylie

  56. 56

    You haven’t Adam. You have been quite congenial.

  57. 57
    Midnight Rambler

    Your assertion that ‘Nature is more about prestige and sexy topics than good science nowadays.’ is, I believe, an opinion that is demonstrably false.

    Misquotes of the statistics aside, I’d say that this opinion is demonstrably true. The “reappearing phasmid wings” paper from a few years ago is another prime example.

    BTW Jen, the post is still incorrect – it now says that Nature’s retractions have increased tenfold, while all papers increased 44%. The number of retractions in the original source refers to all journals as well.

  58. 58
    Midnight Rambler

    Kin selection seemed so compelling when I first read about the haploid-diploid case, but it sounds much less compelling with other eusocial species, particularly humans.

    Humans are not eusocial, full stop. Also, there are very, very few examples of eusocial organisms outside of haplodiploid Hymenoptera (and only a few independent origins at all outside of bees), which is part of the evidence for kin selection.

  59. 59

    Ok, then that’s either another area where Wilson is stepping out of the mainstream, or at least an area where he wasn’t being very clear. He seemed to be at least implying eusociality in humans. The more I dig in to this, the easier it is getting to understand biologists frustration with Wilson’s current behavior. I could understand questioning some of the assumptions of mainstream theories in search of new insights, but Wilson is acting as if previous theories just got run over by a bus and there is no need to even bother questioning them anymore. Very strange.

  60. 60
    Roedy Green

    A hive of bees is not the same sort of beast as a group of frogs. In some ways a hive behaves as a single creature, with disjointed cells.

    What is happening with bees is not really group evolution even though it involves the fate of groups of bees. It is about the genetics of the queen. The rest of the hive could be thought of as auxilliary cells.

    This fight might be resolved by careful vocabulary.

  61. 61
    B Taylor

    I have similar feelings of embarrassment when I see Richard Dawkins Inc. holding forth on religion and theology, areas he knows very little about. There’s a lot to be said for sticking to one’s field…

  62. 62
    John Wylie

    To B. Tayor

    If you feel that way you might be interested in in the thinking of someone who does stick to their field – ie. me. Check out my site — http://www.apesantsandancestors.com/


  63. 63

    What’s with the agist title of this post? It’s a turn-off, and distracts from the excellent content.

    An analogous blog post substituting gender, race, etc. for age [i.e. "Dear Jen McCreight: Please go back to the kitchen or stick to genetics"], would generate a major s**t-storm. The aged are no more deserving of dismissal or derision for that characteristic than other marginalized groups.

  64. 64
    Caroline Packard

    The mathematical biologist Martin Nowak has a new book providing a new mathematical formula defining the conditions under which group selection could have led to selection for human cooperation in groups. He contends that these conditions could easily have existed in the environment in which we evolved, and that previous mathematical “proofs” that group selection can’t work make unnecessarily constrained assumptions.

    Would one of you commenters who have said here that group selection can’t mathematically have been a significant selective force, and who is familiar with Nowak’s newly published thesis, please explain how his thesis is flawed? (If I remember correctly, Nowak is at Harvard or MIT, and he gave a recent video talk that’s posted on the Edge website.)

    Thanks for any kind help.

  65. 65

    Wowser, am I glad I stumbled in here today as I’m on p.91 of Wilson’s new book, The Social Conquest of Earth. And now, I have some additional views/topics to ponder as I continue my read plus some solid links to the controversies he’s created.

    I truly didn’t know this book had opened Pandoras Box. Thanks all for the astute comments.

    As a sidebar, I’ve always liked Wilson and think it’s too bad he’s going out with controversy instead of his usual aplomb.

  66. 66
    Claus Larsen


    I, too, am not comfortable with the demand for retirement of scientists with dissenting views.

    Wilson may be wrong, he may be right. His theory should definitely be scrutinized as rigorously as any other theory.

    But calling for someone’s retirement because of a scientific dispute is the opposite of what science is: A search for truth (or, rather, reality).

    If we followed your lead, we should demand by default that scientists who challenge current scientific theories “retire”. We would never see any scientific progress at all.

    Additionally, ageism is also discrimination. Please bear that in mind. Whatever Wilson is arguing, his age is irrelevant.

  67. 67
    The Tideliar

    Of course “religion” and “theology” are real areas of academic research. Right. Nice comparison, poe-troll.

  68. 68

    Adam Rutherford, I think your original post was ENTIRELY fair and reasonable, absolutely NO NEED to apologize for anything you said there.

  69. 69

    “Misquotes” of the statistics? When the statistics are incorrectly used as part of an argument, that is more than just “misquoting” them!

  70. 70

    The comparison is not apt. The difference is that (1) the aged are ALL of us (G-d willing) at some point or other and (2) it is well known that there is a tendency for people’s cognitive powers to decay toward the end of their life. (Still may not justify the remarks against Wilson in this case)

  71. 71
    Brat Diet

    I don’t even know how I ended up here, but I thought this post was good. I do not know who you are but certainly you’re going to a famous blogger if you aren’t already ;) Cheers!

  72. 72
    Malden Dentists

    Pretty section of content. I just stumbled upon your site and in accession capital to assert that I acquire actually enjoyed account your blog posts. Anyway I’ll be subscribing to your augment and even I achievement you access consistently fast.

  73. 73
    Stuart Mathieson

    I happen to think EO is right. Darwin thought the same and so does David Sloan Wilson and Elliot Sober. It’s the phenotype that is selected and that depends on how the genetic code is expressed which is partly environmental including the cultural environment. Now the cultural environment is clearly a group characteristic and can vary significantly with little discernible genetic variation.
    I suspect your hostility to EO is part of the widespread ideological distaste for him which seems to be a kind of neo Lyshenkoism. It seems to me to be a piece with the post modern Libertarianism widespread these days. One can understand the feminist and Gay communities hostility to historic appeals to biology but an explanation is one thing, a justification is something else again.
    In any case Hamilton’s appeal to kin doesn’t make sense in the Pleistocene era when trobal groups were relatively small and genetically close anyway. In such circumstances how would selection distinguish between group and kin loyalty? I think the probability of altruistic reciprocity would be significent and it explains in group out group preferences more generally. The appreciation for the power of group selection depends on appreciation of non zero sum theory where the aggregate goods of cooperation constitute social goods and social capital. W Edwards Deming had it pretty sussed.

  74. 74
    Charles Goodnight

    PLEASE PLEASE PLEASE learn about multilevel selection from those who do it. Coyne, Dawkins, Williams, Maynard-Smith are totally clueless, as are most of those who signed that paper in Nature. I refused to, and frankly I found the paper to be a bit embarrassing for our field. Before you malign multilevel selection read the authors that have actually done experiments and theory on the subject. Most of the literature is a hard slog, so I understand why you might be hesitant, but here is a list of a few authors:

    Goodnight (that is me)
    Heisler and Damuth

    A few things that become clear from this:

    Kin selection and multilevel selection fundamentally similar. They differ only in some esoteric mathematical issues (that nonetheless make kin selection virtually useless from an experimentalists perspective.)

    No study has ever reported the strength of kin selection. The reason is simple it cannot be done. This leads to the kin selection approach being an experimental dead end. It inevitably grinds to a halt with a statement like “and this is consistent with kin selection”. Of course it is also consistent with alien intervention, and any number of odd if unlikely explanations. Give us numbers!

    Multilevel selection has been embraced, and increasingly widely used in agricultural settings. For the same reasons discussed above, kin selection is NEVER used in agriculture.

    You can read the details here: (https://blog.uvm.edu/cgoodnig/2014/04/23/why-i-dont-like-kin-selection/)

    I agree that Nowak, Tarnita and Wilson and Wilson and Wilson probably went a bit overboard in their condemnation of kin selection, but they are not wrong, and the truth is they may be more right than we think if we also demand dynamical sufficiency.


    The bottom line is that Williams and Maynard-Smith did us a huge disservice by discarding group selection in the 1970s. Kin selection is inherently limited, and ultimately ends up either giving only vague answers, or being functionally useless. Multilevel selection (the modern incarnation of group selection) is the firm theoretical foundation we need to advance the field at this point.

Leave a Reply