In this episode, my brother Jon and I discuss my work on the evolution of kin recognition. Jon is a software engineer and likes to put my arguments in terms of debugging software. For many years, the mere finding of kin recognition in nature was taken as prima facie evidence of W. D. Hamilton’s theory of “inclusive fitness.” A large paradigm was built on the teleological assumption that kin recognition is evidence of the final cause of “inclusive fitness maximization.” A major anomaly to this paradigm called “Crozier’s paradox,” analogous to a software bug, suggested that kin recognition could not evolve for directing altruism to kin. When I finally resolved Crozier’s paradox almost 30 years after it first appeared, the implications were extremely disruptive. As Jon would put it, much of the “software” of social evolution came to depend on the assumptions that led to Crozier’s paradox. By questioning these assumptions, my theory implied that social evolutionists had misunderstood the adaptive basis of kin recognition, incorrectly tested Hamilton’s rule, and misinterpreted Darwinism. Particularly, social evolutionists had misinterpreted Darwin’s theory as teleological and tried to justify this teleology with generalized mathematical equations, like inclusive fitness or generalized versions of Hamilton’s rule. Jon and I discuss how that theorists rejected my work because it did not conform to their prior expectations about what “general theory” is supposed to be, even though it yielded novel predictions for the genetics and evolution of kin recognition that were upheld by 50 years of evidence. We end this podcast with a brief discussion of the differences between scientific peer review and open software forums that allows “bugs” to persist in science. This episode is essential listening for anyone who wants to know what is wrong with science today.
G. C. Williams (1966). Advocating teleology:
“For the same reason that it was once effective in the theological ‘argument from design,’ the structure of the vertebrate eye can be used as a dramatic illustration of biological adaptation and the necessity for believing that natural selection for effective vision must have operated throughout the history of the group.” (p. 6)
“Whenever I believe that an effect is produced as the function of an adaptation perfected by natural selection to serve that function, I will use terms appropriate to human artifice and conscious design. The designation of something as the means or mechanism for a certain goal or function or purpose will imply that the machinery involved was fashioned by selection for the goal attributed to it.” (p. 9)
“A frequently helpful but not infallible rule is to recognize adaptation in organic systems that show a clear analogy with human implements.” (p. 10)
E. O. Wilson Against Inclusive Fitness:
“The misadventure of inclusive-fitness theory originated in the belief that a single abstract formulation, in this case the Hamilton inequality, has implications that can be unpacked layer by layer to account for social evolution in ever-growing detail.” (p. 182)
“The evolutionary origin of any complex biological system can be reconstructed correctly only if viewed as the culmination of a history of stages tracked from start to finish. It begins with empirically known biological phenomena at each stage, if such is known, and it explores the range of phenomena that are theoretically possible. Each transition from one stage to the next requires different models, and each needs to be placed in its own context of potential cause and effect. This is the only way to arrive at deep meaning of advanced social evolution and the human condition itself. ” (p. 180)
Example of Confusion on Greenbeard Effect:
See Madgwick et al.’s (2019) paper on greenbeard genes. In this paper, the authors take my classification of association based aggregation and fusion (Gilbert 2017) without citing me. They then confuse greenbeard genes and kin recognition genes. I notice that the authors thank Andy Gardner for discussion. Andy Gardner is perfectly aware of the difference between greenbeard genes and kin recognition genes, which is confused in this paper (e.g., Gardner and West 2010). Andy Gardner was also the editor on my 2015 Am. Nat. paper (Gilbert 2015).
My Replies to Reviews of my Association Theory Paper:
Disclaimer: I have not published anyone else’s written review but instead included summaries in my own words. I have mentioned names only in cases that reviewers voluntarily revealed their identities. I mention the names of editors because they are well known and non-anonymous anyway. I share what happened because I think it paints a picture of overall bias. The biggest recurrent problem was that reviewers always rejected a paper that had a lot of data and evidence in support of it, and never even mentioned this fact.
I submitted this paper to Nature, Evolution, Journal of Evolutionary Biology, American Naturalist, Quarterly Review of Biology, and Frontiers in Ecology and Evolution, over the course of 2 years. It was reviewed at Nature and American Naturalist. I would say that I improved the paper after a first round of reviews at American Naturalist, where I was invited to resubmit. I published the version that I resubmitted to American Naturalist as a preprint in 2017.
After 2 years, submissions to 6 journals, and 2 reviews, I finally moved on to another project when the paper was desk rejected at Frontiers in Ecology and Evolution after I was invited to submit the paper. I even gave a talk at a conference about it, which was part of the invite, but they still didn’t review my paper. At that point, I decided to move on to my macroevolution project. I ended up being able to publish a much more speculative and challenging theory about macroevolution before I could publish this review paper. I think that speaks somewhat to the conservatism of the field of social evolution.
I really thought that it wouldn’t be too difficult to publish this paper. After all, I had already published a model that showed how the theory could apply. I had much evidence in support also.
I tried submitting the paper one more time in 2020 to PNAS but it was desk rejected. So I decided to leave it as a preprint, and publish something different. I think the preprint is still a pretty good rendition of the paper even though I had modified it a bit for the other submissions. Mostly the differences were in the way that I introduced the paper.
Nature (Reviewed, Rejected)
Feb 24, 2016
I first sent a version of this article to Nature. It was reviewed by two people, one who seemed to be an inclusive-fitness advocate and the other more of an inclusive-fitness critic. The inclusive fitness advocate misrepresented my paper as being about the “greenbeard effect,” even though I was explicit that it was about genetic kin recognition and asociation preference. The second reviewer argued that my paper lacked novelty, and that others had already made some of the distinctions necessary for my argument. They also said that my paper lacked “proofs” and “intuition,” which I took to mean that they thought it was not “mathematical enough.”
I found these same sorts of arguments in all the following reviews. People either misrepresented what I wrote to make the claim that it was not new, or said that it was not mathematical enough. Neither argument made no sense because (i) if my theory wasn’t new, it shouldn’t have made new predictions upheld by much evidence, which I reviewed; and (2) I had a mathematical model that was previously published. The point of this paper was to explain the broader implications of the model and review the evidence for its predictions.
Evolution (Desk Rejected)
Dec. 31, 2016
I submitted this article to Evolution and Mohamed Noor, the Editor, desk-rejected the paper. His rationale was that he had held to a standard since his first day as editor to not publish “verbal theory” papers. It was odd to me that a paper with new predictions and a discussion of evidence would not even be considered at Evolution even though it had been reviewed at Nature.
Journal of Evolutionary Biology (Desk Rejected)
Feb. 13, 2017
The editor Michael Ritchie rejected my paper on grounds that it makes bold claims but does not really deliver. Ritchie cited an anonymous deciding editor for the decision.
American Naturalist (Reviewed, Resubmitted, Rejected)
Dec. 15, 2017 (When I received the second review back)
After a first round of review in which some of the reviewers were partially complementary to my paper, the editorial board rejected my paper. Their main reason was that my framework was supposedly too similar to previous ones. The editors also said that my theory lacked “simplicity and power to proceed to further development.” One reviewer, J.B. André, said that kin recognition could evolve to gain automatic benefits of association without differential treatment. André did not have a mathematical model supporting his argument nor had anyone ever made that argument. André also argued that I should have cited Richard Michod, however I made clear in my letter that Michod never modeled association preference or genetic kin recognition. Nevertheless, André’s arguments seemed to influence the editor Bronstein who said that I had sold other authors short and that my ideas were not new or useful.
The three people who reviewed the paper, Peter Taylor, James Marshall, and J.B. André were all signatories on the Abott et al. 2011 reply to Nowak et al. 2010, suggesting that they are the sort of old-guard “inclusive fitness” types. None of them said anything about the data in support of my theory while they were rejecting my paper. Nor did they mention that I already developed a specialized model and applied it to a particular case, which shows that the theory had already proceeded to further development.
Quarterly Review of Biology (Desk Rejected)
Feb. 24, 2018 (Submission date)
After American Naturalist, I sent a slightly modified version of my paper to QRB. It sat there for four months before the editors got back to me on June 14, 2018 and said that they were not going to review it. Their excuse for taking so long was that one of them suffered an injury during field work. They said that they decided not to review the article in part because it had already been posted on a preprint server. They mentioned that a potential reviewer had read the preprint and had some reservations. The editor expressed concern that my paper had already been cited 20 times and that they were worried that a version published in QRB would not be cited much–the preprint absorbing the citations. Not only had my preprint paper not been cited at that point, however, their argument did not make sense. Why would a highly cited preprint be less likely to attract citations as a published paper? The editor then said that he thought the reason my article was important was that I added an association stage in the evolution of social groups. He then suggested either cutting the length or publishing the article as a monograph. He said he was open to resubmission of a shorter paper with a new title, which I was supposed to write to a professional biologist…not familiar with the specialized language of the field.
I did not end up resubmitting to QRB because I did not want to reduce the length of the manuscript very much. I also thought that it was unfair to not allow me to use field-specific language, or terms that I had defined in a glossary. It is difficult to speak of such an intricate subject without terms like “association preference,” “differential treatment,” etc. Without a the ability to use specialized language it is impossible to build coherent arguments. This is proven throughout biology and any field of science. What would physics be without “muons” and “quarks”?
(the Editor at the time was Daniel Dykhuizen and he said he read my article, the other section editor was James Thomson).
Frontiers in Ecology and Evolution (Invited to Give Talk and Paper, Talk Well Received, Paper Desk Rejected)
August 1, 2019 (Final Decision)
The last journal I submitted this paper to was Frontiers in Ecology and Evolution. I was invited to give a talk at a special section at the Evolution (2018) conference in Montpellier France by Christie Riehl and Mark Elgar. I was also invited to submit a paper for a special section on “Mechanisms of Recognition in Social Evolution” in Frontiers. I was excited to submit because I saw that the research topic was extremely relevant to my article. Particularly, it asked for papers that investigated diverse taxa and which seek common patterns and with treatment of the mathematical stability of recognition systems. This seemed perfect for my paper which covered an array of taxa and which had mathematical models to explain the evolution of mechanisms of genetic kin recognition.
I queried the editors of the journal when I gave my talk at the evolution conference to make sure that my article would fit their style. The person I talked to assured me that my article could be quite long and there was no set formatting style.
A lot of people came to my talk at the conference and a number of people came to talk to my afterward, all very complementary. This included several theorists who worked on other topics.
This is particularly what was posted about the research topic:
The ability to recognize individuals and communicate with fellow group members is crucial to the evolution of complex social behavior. At a minimum, cooperating individuals must produce and receive signals that identify themselves as appropriate partners, and collective behaviors require group members to communicate their intentions and synchronize their actions. But how do mechanisms of recognition and communication co-evolve with social behavior, and how do similar signaling abilities arise across animal lineages with vastly different sensory systems and cognitive capacities?
The ability to recognize individuals and communicate with fellow group members is crucial to the evolution of complex social behavior. At a minimum, cooperating individuals must produce and receive signals that identify themselves as appropriate partners, and collective behaviors require group members to communicate their intentions and synchronize their actions. But how do mechanisms of recognition and communication co-evolve with social behavior, and how do similar signaling abilities arise across animal lineages with vastly different sensory systems and cognitive capacities?
Mechanisms of communication and recognition, using different sensory modalities, have been investigated across diverse animal systems. Progress in our understanding of the roles of communication in social systems is uneven across both contexts and taxonomic divides, but perhaps more significantly, these studies are rarely synthesized to seek common patterns across taxa. Studies of social insects have a substantial focus on the role of chemical communication for individual identity, such as cuticular hydrocarbons for recognition systems and more specific pheromones for queen fertility. Studies of social vertebrates investigate the co-evolution of vocal complexity and social group size in primates and, more recently, the significance for young birds of learning family-specific vocalizations while still in the nest. Much of this research has focused on the relationship between signaling and social evolution, although there is an emerging interest in considering how selection acts on the receiver.
This Research Topic will bring together leading evolutionary theoreticians and empiricists to discuss the proximate mechanisms and selective pressures that favour the diversity of communication and recognition systems of social animals. With a wide range of taxa, extending from bacteria to insects to humans, and with mathematical treatments of the evolutionary stability of recognition systems, this reflection of our current understanding of the field will surely help shape future research directions.
My paper was then desk rejected on July 26, 2019 because it was said to be too long and more of a “conversation” than scientific piece. I thus emailed Christie Riehl as follows:
July 27. 2019
Dear Christie,
I got an email from Frontiers saying my manuscript will not be reviewed. The comments are so vague that I do not know whether I should resubmit this, or how I might revise it. I wonder if you or Mark could help me by telling me what you think about these issues mentioned in the email. I went to some effort to prepare this, so any comments would be appreciated.
There were basically three issues brought up, but I don’t see why these would prevent review:
(1) Length—the article is slightly over limit, but not by much. Some of the figures could be moved to a supplement.
(2) Copy right issues—One of my figures was redrawn from another journal but I could request permission or redraw it.
(3) Whether it is science or conversation—By it’s nature as a framework-altering advance, my article requires a new classification scheme and language, and a lengthy discussion of how the theory applies to natural systems (including how to measure its parameters, how to apply it to well-known biological examples, how the theory resolves anomalies of former theory, how its major predictions are upheld by large bodies of evidence, etc.). This type of “conversation” is expected of a major scientific advance. I thus see it as a false dichotomy between conversation and science (and not just because I have read Darwin).
Also, (3) is particularly bothersome because I *specifically* spoke to one of the Frontiers editors about this very issue at the meeting last summer. I asked him whether they would allow an article of abnormal length and in somewhat of an “unconventional” format. He said yes, that it was not problem. I therefore wonder if this is really the issue. If there is some other concern then please illuminate me.
I appreciate your time.
On August 1, 2019 I then got a letter back from Mark Elgar saying that Christie was unavailable. Elgar said that my manuscript addressed a far broader set of issues than specified in the research topic. What I showed in my manuscript was that in order to explain the evolution of genetic kin recognition systems it is necessary to expand and alter the theoretical approach. The new theory was merely the means to the end of addressing what was in the research topic. What my theory explained actually was perfectly within the topic and an example of something that was very integrative and got insights from a wide range of taxa. According to Elgar’s logic, however, if new theory is required to address the research topic then it should not be considered. This to me makes no sense from a scientific standpoint and merely reflects extreme conservatism.
Elgar then said that my arguments appeared similar to others and that I should shorten them. Again, this was very similar to reviews I got at American Naturalist and Nature. However, if my arguments are so similar to previous ones then why do they newly explain kin recognition and make so many novel predictions? Nobody who reviewed my paper ever addressed this question.
Elgar then said that he’d be happy to reconsider this manuscript if I paid the large publication charges, which I originally was not required to pay because it was an invited article. Being without funding I was not able to resubmit nor did I appreciate Elgar’s comments about my paper.
I was very disappointed in Elgar’s response because I had high hopes for him. In an article he published in 2015, Elgar called for more integration and attention paid to overlooked model organisms. Many of my insights came from focusing on oddballs like botryllid ascidians, hydroids, fungi, and slime molds not usually considered by social evolutionists (who typically focus on arthropods and vertebrates). I thought Elgar might thus be partial to my unique perspective and paper that integrated theory and empirical work.
In that paper, Elgar (2015) said, “…empirical insights…seem to be too frequently informed by taxonomically constrained traditions and, as a consequence, advances in the field are surprisingly patchy across different taxa.” My question is how we are supposed to have “advances” in the field if any analysis that requires a new theoretical advance is rejected because it is “outside the scope” of what is considered for publication?
PNAS (Desk Rejected)
Jan 30, 2020
In 2020 after I had made substantial progress on my macroevolution project, I submitted a version of this paper to PNAS. It was desk-rejected without explanation. At that point, I decided to just leave my preprint as it was and take a different approach.
Another Approach
Although I started biology accepting inclusive fitness taught in my textbooks, at some point after graduate school I realized that the way people had taken facts to support the theory of inclusive fitness was pseudoscientific. Particularly, they took the finding of kin recognition as evidence of inclusive fitness even though inclusive fitness never predicted kin recognition. However, I could see it as an honest mistake. To avoid offending people I decided to defer to biological evidence and not make a big deal about it. But the people in the field didn’t let me do that, so I had to finally call them out. I did this first in a short section in my 2020 paper on natural reward. Next, I plan to write a paper that really describes the problems with this field.