Richard Dawkins published The Selfish Gene in 1976 and in 2017 the Royal Society conducted a popular poll that voted it the most inspiring science book of all time.
I don't think that book is a good representation of modern evolutionary thinking and I've expressed that opinion many times on this blog [The selfish gene vs the lucky allele] ["The Selfish Gene" turns 40] [The "selfish gene" is not a good metaphor to describe evolution] [Die, selfish gene, die!].
The issues are complicated—here's how Dawkins described his book in the 1989 edition.
The selfish gene theory is Darwin's theory, expressed in a way that Darwin did not choose but whose aptness, I should like to think, he would instantly have recognized and delighted in. It is in fact a logical outgrowth of orthodox neo-Darwinism, but expressed as a novel image. Rather than focus on the individual organism, it takes a gene's-eye view of nature. It is a different way of seeing, not a different theory.
My objections are that the book is far too adaptationist and too "Darwinian." It fails to convey the complications of evolution to the general public, including the dominance of chance, accident, and contingency. It misrepresents some views of evolution, especially punctuated equilibria. It focuses too much on genes (alleles) as the unit of selection (evolution?) at the expense of individuals, groups, and species. (Species sorting may not be important but it deserved more than passing dismissal.) It focuses too much on the evolution of animals, especially animal behavior and that feeds into the biases of the general public who know very little about the vast majority of species outside of the animal kingdom.
Nathaniel Comfort has just published an article in Nature expressing his own disappointment about The Selfish Gene [The Selfish Gene, 50 years on — it’s time for an update]. He also doesn't like the book but for entirely different reasons.
Both components of the modern synthesis focus on the gene as the sole determinant of heredity. Both halves reached essentially universal acceptance well before Dawkins came along — the spit and swagger of the field’s proponents keeping challenges and alternatives at bay.
To a historian, this universal acceptance sounds close to ideology. Fortunately, since The Selfish Gene was published, scientists’ understanding of the basic mechanisms that drive evolution has changed drastically.
For instance, neo-Darwinism teaches that evolution is a random walk through changing environments, with selection favouring whatever traits help an organism to survive. Today, many scholars argue that there has to be direction to evolution. In How Life Works (2023), for instance, journalist Philip Ball points out that in a single cell, there is an unfathomably large number of possible interactions between different kinds of molecule — and there are about 37 trillion cells in one human body. “Such a system,” Ball writes, “can only hope to be stable if, out of all this complexity, only a limited number of collective ways of being may emerge”. Living organisms are forced to evolve in those few directions.
Firstly, "scientists' understanding of the basic mechanisms" changed BEFORE The Selfish Gene was written. That change occurred in the late 1960s with the development of neutral theory and the recognition that random genetic drift was the dominant mechanism of evolution. Secondly, Philip Ball is dead wrong about all kinds of things, including agency. [Philip Ball doesn't understand sloppy genomes]
Comfort keeps digging deeper with an appeal to the extended evolutionary synthesis, the epigenetics myth, and the microbiome. (What percentage of all evolution over 3.5 billion years has anything to do with probiotics?)
For years, a group of biologists, philosophers and historians has been calling for an extended evolutionary synthesis that accounts for phenomena beyond this isolated window of transmission. Heredity operates at many levels at once. For example, there’s some evidence that epigenetic changes — molecular modifications to the genome that can alter gene expression without changing DNA — are heritable. Ecosystems involve many life forms that adapt to one another, creating stability and developing a kind of homeostasis — the self-regulating process through which balance is maintained — that confers resilience against catastrophic change. Your body consists of more microbial cells than cells derived from your parents’ genes. The microbiome doesn’t participate directly in your genetic heredity, but you can’t live without it.
Jerry Coyne didn't like the Nathaniel Comfort article [Nathaniel Comfort’s misguided take on The Selfish Gene at 50]. Coyne argues that there's nothing new in the so-called extended synthesis and he quotes Douglas Futuyma as support.
That's good but unfortunately Coyne gets so caught up in defending the Dawkins' view that he doesn't recognize the more legitimate criticisms that many people have raised. Let me quote from another article by Douglas Futuyma (Futuyma, 2023).
There is no question that by the 1960s, selection was thought to be by far the most important of the factors that influence allele frequency change, owing to diverse studies in ecological change. ... The neutral theory of molecular evolution, proposed by Kimura contrasted strikingly with the current thinking, and resulted in the neutralist-selectionist controversy that engendered a large, diverse range of research programs in evolutionary biology for the next decade or more. ... The outcome, as we now know, is that sequence variation within and among species includes both selected and neutral (or nearly neutral) sites with immense consequences for understanding genome variation and evolution. ... Of all the events in evolutionary biology that I have witnessed, this is the single one that has most changed our understanding revolutionary.
Futuyma goes on to say, "The only controversy that might be called a substantial 'paradigm shift' was the neutral theory of molecular evolution."
Futuyma, like Coyne, dismisses the so-called extended evolutionary synthesis as trivial or unimportant but, unlike Coyne or Dawkins he proclaims the development of neutral theory a revolutionary change that might even be called a paradigm shift. Coyne is right to question the silliness of Nathaniel Comfort's article (Philip Ball - really?) but Coyne is wrong to ignore the fact that Dawkins said nothing about the revolutionary change in evolutionary theory that occurred years before he published The Selfish Gene.
UPDATE: Coyne has now posted a new article noting that he "... forgot to mention one important supplement to evolution’s 'Modern Synthesis.'" [The Neutral Theory: a genuinely important supplement to evolution’s “Modern Synthesis”]
Futuyma, D.J. (2023) The Evolving Evolutionary Synthesis. Evolutionary Biology: Contemporary and Historical Reflections Upon Core Theory. T. E. Dickens and B. J. A. Dickens, Springer: 517-541.
I read Comfort's article when it came out, and I agree it's a joke (especially referencing Ball). I would argue for The Selfish Gene as a still-engaging pop-sci book when it comes to behavioral ecology. However, I would not argue for or against it as if it was/is a textbook on everything evolutionary biology.
ReplyDeleteAnd yes, while neutral theory made its debut in the late 1960s as you've noted, it was still missing crucial data: "Enthused by their discovery, they slightly overstated the prevalence of neutrality. With the gradual accumulation of DNA sequence data, subsequent research converged toward a more balanced view of molecular evolution" (Galtier, 2024, DOI:10.1093/gbe/evae003).
Also see how "Kimura's paper had at least two deficiencies" in Nei (2005, DOI:10.1093/molbev/msi242).
I'm only humbly pointing out that by the late 60s to mid-70s it wasn't a slam dunk.
Rowan Hooper's article on The Selfish Gene's 50th was much more balanced, and crucially, tempers the overhyped epigenetics and company.
(I still can't seem to be able to create an account, so apologies for the yet again nameless account)
A quick addendum:
DeleteRevisiting the book, e.g. chapters 3 and 8 (rowing crew) provide a omnigenic view:
"No one factor, genetic or environmental, can be considered as the single ‘cause’ of any part of a baby. All parts of a baby have a near infinite number of antecedent causes" (ch. 3).
Also the gene is defined in its evolutionary, not biochemical, sense.
And revisiting Kimura's 1988 book (English translation: DOI:10.1007/978-981-15-6165-8):
"When we consider evolution at the phenotypic level, what is indisputably interesting is macro-evolution and the associated question of evolution at the phenotypic level. In this case, Darwinian natural selection undoubtedly plays the major role, but the simple panselectionism that was entertained during the golden age of the synthetic theory of evolution needs to be revised."
@nonymous: It's important to understand that many biologists are only interested in the evolution of genes, which they often assume are synonymous with coding sequences. For them, the entire neutralist-selectionist debate is about substitutions in coding regions and opponents of neutral theory are more than happy to focus on slight fitness effects of synonymous codon changes.
ReplyDeleteThis is why they object to the idea that most changes are due to drift.
That's a very narrow view of evolution at the molecular level. It ignores all the changes that are going on outside of coding regions.
But that's not the main point I was making. When Futuyma said that evolutionary theory was transformed in the late 1960s he wasn't referring to the idea that every proposal by Kimura et al. had to be correct. He was referring to the fact that the old ideas about the dominance of natural selection were correctly challenged.
Anonymous says, "Also the gene is defined in its evolutionary, not biochemical, sense."
ReplyDeleteYes. I make that distinction very clear in my book and in the Wikipedia on 'gene.' Dawkins was referring to any bit of DNA where changes in the genome occur - that's the Mendelian gene not the molecular gene.
However, Dawkins missed the opportunity to distinguish between genes and alleles. That's a shame because it would be very helpful if the general public understood alleles.
He also missed the opportunity to clearly distinguish between his use of 'gene' and the molecular gene used in biochemistry and molecular biology. Thus, he conveys the impression that all of evolution is about changes in selfish genes where genes are bits of DNA that encode proteins.
Changes in a base pair in junk DNA count as a gene in Dawkins' terminology.
In your book (which is my go-to reference, by the way!) - re the changes in a bair pair in junk DNA - you make the excellent case for maintenance and origin functions. That's the very "longevity, fecundity and fidelity" in Dawkins's "gene", i.e. it readily accommodates that our genome is 90% junk. If generation to generation x segment is not under selection, then it isn't a gene in the sense he explained in the early chapters of the book.
DeleteAnd he did explain and stress the alleles, e.g. out of 38 hits:
"As far as a gene [in a body] is concerned, its alleles are its deadly rivals, but other genes [in the same body] are just a part of its environment, comparable to temperature, food, predators, or companions. The effect of the gene depends on its environment, and this includes other genes" (that's the poly- or omnigenic model again, with the alleles in the gene pool being the "rival").
A similar argument can be found in his The Extended Phenotype regarding the "spacer DNA" argument of the 1970s, which he dismisses since such sequences are not allelic in the evolutionary sense.
I don't get why the book generates so much controversy - in my earlier reply I quoted a part (out of many) that stresses that a gene on its own (more so without the environment) doesn't dictate everything. And the statistical (contingent) nature of evolution is also addressed.
A possible answer I came across was Haig's in From Darwin to Derrida; his point on the type/token distinction.
Funnily enough that's exactly the point I make in my own retrospective on TSG in Scientific American: https://www.scientificamerican.com/article/the-selfish-gene-at-50/
Delete@Philip Ball: In that article you say, "The evolutionary biologist doesn’t much care what a gene does, so long as it somehow affects the traits and thus the fitness of the organism."
ReplyDeleteReally? Are you saying that evolutionary biologists are only interested in the evolution of genes and not in the evolution of regulatory sequences or centromeres? Are you saying that evolutionary biologists aren't interested in neutral evolution that has nothing to do with traits or fitness?
It sounds like you are assuming that all evolutionary biologists are like Richard Dawkins.
You also say, "The developmental or molecular biologist doesn’t much care how a gene got to be the way it is but wants to know how it influences growth, cell function and health."
That may be true of some developmental and molecular biologists but there are a great many developmental biologists who are interested in evolution - there's a whole field devoted to evo-devo, perhaps you've heard of it? There are also many biochemists and molecular biologists who are passionately interested in evolution - some have even written books about it.
@Philip Ball: The way we define a gene is incredibly important. Let's take the molecular gene as an example. In your book, "How Life Works," you devote a long chapter to genes but you never get around to the definition that's been common in many molecular biology textbooks for the past 50 years.
ReplyDeleteInstead, you bring up an old-fashioned definition that defines a gene as a stretch of DNA that encodes a protein. On page 82 you attribute this definition to Crick when you say, "Crick's definition of a gene as part of DNA that encodes a protein is no longer adequate, for reason that will become more clear in the next two chapters." You then go on say that "we now know" about non-coding genes that don't encode proteins at all but instead specify functional RNAs.
Then you say, "There is no consensus about whether or when such non-coding DNA sections warrant being called (noncoding) genes - simply because there is no longer any agreement about what, at the molecular level, 'gene' should designate.
But it's been very common to define a gene as a DNA sequence that's transcribed to produce a functional product (RNA or protein) and that definition has been in use for more than 50 years. Why don't you mention that?
This is such an easy game to play Larry - I could find nits to pick in just about every blog piece you write, but I won't, not just because it's not worth my time but also because it would be an ungenerous denial of what writing for a general audience is all about. (No I don't think all evolutionary biologists think X and so on - that's just a silly interpretation to put on those words, and we both know it.) All the same, my book does include essentially that definition of a gene (worded more precisely), so there isn't even here a nit to pick, just yet more misrepresentation. Plus ça change. (And there I was thinking: Why not point out something on which we seem to agree? But heaven forbid that you should accept that!)
DeleteTwo eminent historians of science (William B. Provine and Mark B. Adams) came to the conclusion around 1990 that they had been misled in the 1960s by early mentors. By the 1990s bioinformatic analyses of freshly available DNA sequences had led a scientist historian to the conclusion that there were at least two levels of information in DNA, one genic and localized to genes and one global (genomic), which could be looked upon as the accent of DNA. Like the accent of Eliza Doolittle in Pygmalion, this could have reproductive consequences.
ReplyDeleteOver the 1990s decade the paths of these three historians converged. Of their many subsequent papers and books, summaries of key points (2024) may be found in the Journal of Theoretical Biology (2024) 595, 111943 and in Theory in Biosciences 143, 1-26. The latter is entitled "Speciation, Natural Selection, and Networks: Three Historians Versus Theoretical Population Geneticists."
They are difficult reading for non-specialists. However, until influential public commentators such as Nathaniel Comfort, Richard Dawkins and members of the EES movement, reveal a minimal awareness of this viewpoint, it might be best to give their commentaries less weight.
@Philip Ball: As it turns out, I also wrote a book for a general audience and I also wrote up a definition of a gene. I'll post it here and let readers decide for themselves which version is the most historically and scientifically accurate.
ReplyDelete(From page 137 of "What's in Your Genome")
"In Chapter 1 I defined a gene as a DNA sequence that is transcribed to produce a functional product and this is the way genes have been described since the 1960s by many scientists who have thought seriously about the subject. There are lots of textbooks that use a similar definition. Kostas Kampourakis in Making Sense of Genes concludes that the most acceptable definition of a gene is the one that includes production of both proteins and noncoding RNAs (see the quotation at the beginning of this chapter).
The idea here is that the primary gene product is RNA and any segment of DNA that’s copied into RNA (transcription) qualifies as a gene as long as the RNA has a biological function. That last bit is important because there are sequences of DNA that are transcribed but the RNA doesn’t do anything. Transcribed pseudogenes are an example but so are many spurious or accidental transcripts that happen quite frequently when the transcription machinery makes a mistake. Those bits of DNA are not genes even though they are transcribed.
There are very few definitions in biology that are absolute. In fact, one of the main reasons why biology is so difficult is because it is full of exceptions to the rule. In the case of genes, the exceptions are well-known; they include the fact that some genes are made of RNA not DNA—the genes of coronavirus are an example. There are also more complicated exceptions such as cases where several different protein-coding regions are transcribed together to produce a long RNA where several different regions can be translated simultaneously. These complexes are called operons and they are common in bacteria. The problem here is that we will often want to refer to each protein-coding region as a separate gene. There are also rare examples of two separate DNA sequences that are transcribed to produce RNAs that are then joined together to make a functional product; strictly speaking, neither of the separated DNA sequences qualifies as a gene.
What this means is that the best definition of a gene is only a “working definition” because there is no definition that covers all cases. This drives philosophers crazy but biologists are used to such ambiguity."
My discussion about defining a gene differs in three important ways from Philip Ball's.
ReplyDelete1. I make it very clear that knowledgeable scientists have a good definition of 'gene' - they are not confused.
2. I don't waste any time on stupid incorrect definitions that were used by unthinking scientists, especially the idea that genes only encoded proteins. I don't waste any time pretending that the experts were wrong and recent advances have discovered non-coding genes. (People who do this have an agenda.)
3. I make it very clear that the most important feature is whether the gene product is functional. This sets us up for a serious discussion about the number of possible non-coding genes in the human genome and the frequent misuse of the word "gene" in that context.
1. I make it clear that there is still valid discussion among knowledgable scientists about how best to define a gene and that there are subdisciplinary differences in how genes are conceptualized, rather than pretending that there's a single best definition and others are just ignorant.
Delete2. I make it clear that there were historically very reasonable grounds once for supposing that genes encoded proteins and that the scientists who thought this weren't stupid. (For more info on that, see e.g. Lily Kay's Why Wrote The Book of Life?)
3. I also make clear that the molecular definition of a gene hinges on whether the product is functional. I add that there are now good reasons to suppose that functionality is not always easy to give a clear yes/no answer.
So I too figure that readers can judge for themselves.
Hmmm: "any segment of DNA that’s copied into RNA (transcription) qualifies as a gene as long as the RNA has a biological function." From the moment that a section of DNA first becomes transcribed, it becomes a candidate for function, which natural selection might recognize. This might not occur soon enough, so it may disappear.
ReplyDeleteHowever, what if in the following, say, 100 generations, a viral RNA appears in the cytoplasm with a complementary segment? As shown in the early 1970s, dsRNA will form and activate host alarm mechanisms (PKR). This will endow survival value on the host segment. Its transcript has a function. Eureka, it is a "gene"!
And as for "biology is so difficult is because it is full of exceptions," you are in good company, Larry. Perhaps take a peek at Treasure Your Exceptions. The Science and Life of William Bateson (2nd edition, 2022). You are in agreement with one who fought ardently (1900-1906) against a powerful research establishment in order to get Mendel's work recognized. Maybe we will come to view similarly the battles that are being fought today concerning segments that some regard as "junk"?