In The Hopeless Monster? Not so fast! Bora says:
In a back-and-forth with a commenter, Coyne defends himself that he is talking about the changes in genes, not evolution. This just shows his bias - he truly believes that evolution - all of it - can be explained entirely by genetics, particularly population genetics. His preferred definition of evolution is probably the genocentric nonsense like "evolution is a change of gene frequencies in a population over time". I prefer to think of it as "evolution is change in development due to ecology" (a softening of Van Valen's overly-strong definition "evolution is control of development by ecology"). Population genetics is based on the Hardy-Weinberg equilibrium - pretty much all of it is a build-on and embellishment of it. Population geneticists tend to forget, once they get into complex derivations of HW, that HW has about a dozen completely unrealistic assumptions underlying it. Now, in a case-to-case basis, some of those assumptions can be safely ignored, some can be mathematically taken care of, but some are outside of the scope of mathematics (or at least the kind of math that can be integrated into the development of HW). Those are ignored or dismissed and, if this is pointed out by those working on evolution from a Bigger Picture perspective, met with anger.
First I would like to observe that the HW is the jumping off point for many models, but obviously population genetics' bread & butter explores the large space of deviations from that ideal (i.e., a population at HW equilibrium is not subject to selection, mutation, migration, etc.). I think Bora's exposition here would definitely mislead; don't confuse the brick for the house. But is this "nonsense"? I'm not very "religious" about the whole "evolution is change in allele frequencies" or it isn't debate; other definitions just don't seem as clear and useful, but I'm willing to entertain them. Also, Bora's fixation on "genocentrism" is often hard to figure out, but I really stopped paying attention to that when he started posting strange things like Genocentrism aids Anti-Abortion Arguments. I am asking for an intervention if I ever post something titled "Evo-devo supports Fabian socialism," just so you know.
Since a fair number of evolutionary biologists read this bog, I'd like some input on the whole "evolution is allele frequencies" debate. Because I tend to think from a genetic angle in terms of models and formalisms operating upon them it makes sense that this is a clear way for me conceptualize the issues (I mean, to a great extent population genetics has always been rooted in the allele frequency line of thinking). But I am interested in what others have to say, as my interest in paleontology of late should make clear. The standard population genetic definition captures the essential point about heritable transmission; which seems critical to highlight in any biological evolutionary context. It's a clear & distinct idea. But that doesn't mean that other ideas don't offer value-add in terms of insight or utility in their own contexts of course, I just don't know as much about alternatives.
Note: I am aware of those who argue that gene-gene interactions & networks really modify the "change in allele frequencies" mantra, I definitely think there is something to that (I have a strong interest in statistical epistasis). But that seems to just a variation of the "genocentric" outlook. We're starting with the same brick and just assembling a bit differently. Sewall Wright was one of the founders of population genetics and for most of his career gene-gene interactions were central elements within his substructure based models.
Update: Greg Laden weighs in:
In the end, however, there is a larger question: What the hell are you'all talking about anyway? I find that the discussion of "hopeful monsters" and saltational evolution has not addressed the essential, fundamental question of adaptation. This may be because most of the people who are talking about it are not adaptationists, and the current trend in the blogosphere is to be anti-adaptationist (it seems to me). But this is a conversation about adaptations and how they arise, so this is something we should talk about.
Right. Bora has a tendency to talk about "nonsense" or how "old-fashioned" "genocentrism" is, but I really don't get a clear sense of what he's positing to replace it. To say, for example, that multi-level selection is the future is fine, but even those who prioritize individual selection agree that selection occurs on multiple levels. To really proceed with any discussion you need to clarify how exactly the selection occurs and what the relation between the levels are in the real world. Too often critics of "orthodox" population genetics play the game of critique without generating much in terms of a counter-system. The beauty of population genetics is that its relatively simple formalism does make it easy to critique; some of the algebra and other mathematical methods can be a bit hard to follow, but ultimately all that's needed is time & effort. No hermeneutical analysis of cryptic texts is needed. I don't think one can say the same for the arch-anti-adaptationist, Stephen Jay Gould, who liked to deal in words. One might wonder as to the clarity of someone so often misunderstood.
Finally, in the comments Bora states:
Coyne is a knee-jerk anti-Gouldian and he will use any opportunity to slander Gould, appropriate or not. And I am not an adaptationist myself, but the questions of the origin of diversity and the origin of adaptation are central questions of Biology which can partly, but only partly, be explained at the level of the genes.
First, I doubt as a geneticist such as Jerry Coyne spends much time thinking about Steve Gould's ideas much. Second, I just checked Speciation (use search inside on Amazon) and I don't see a strong anti-Gouldian ax grinding. There's some skepticism, but we're not talking about Darwin's Dangerous Idea here. Finally, I think it might be relevant to point out that from what I recall Jerry Coyne comes out of Dick Lewontin's lab at Harvard, and Lewontin was long a collaborator with Gould (e.g., spandrels). So I am wondering why Bora claims Coyne is an "anti-Gouldian"?
In this thread and over at Greg's Bora has mentioned that this sort of controversy is going to be great for traffic. All for the good, but controversy should be grounded in accurate representations of the arguments. Jerry Coyne an "anti-Gouldian"? Population genetics just an embellishment of the Hardy-Weinberg equilibrium? Come on. Many of the readers of ScieneBlogs don't know much about population genetics or Jerry Coyne, and comments by ScienceBloggers carry some weight and authority. Can we be a little cautious and sacrifice rhetorical positioning for the sake of some fidelity to reality? I do believe that in the end Truth will win out, but cutting out the noise can help in reducing the time until resolution.
- Log in to post comments
I love doing this back-and-forth with you! Good for the discussion, good for traffic, good for educating the readers. And over the years we became a little bit more civil with each other as well which is a Good Thing TM ;-)
Also, I never link back to that anti-abortion post of mine since it is 3.5 years old and I have refined my stances since. It was originally designed to be provocative more than informative anyway. I linked to other stuff in my post instead.
first hit on google for "genocentrism".
Holly shit! When you are logged in or logged out of Google?
both dude. i double-checked.
If we stipulate (as I think we must) that a) evolution is an emergent property of populations, not individuals, and that b) heritable variation in development reduces to variation in developmentally expressed alleles, then I think we can conclude that Quail is full of crap.
What definition of evolution is there that does not involve changes in allele frequencies? It seems to me that this process necessarily underlies everything that occurs in evolution. Sometimes, it may not make sense to break things down to this fundamental level of explanation, like when you are trying to explain how tetrapods evolved from fish-like ancestors. But even then, under all of the layers of complexity, involving evo-devo, the fossil record, etc., about fins becoming feet, this evolutionary process (and all the others) ultimately was about an increase in frequencies of alleles for feet-like protuberances than fin-like protuberances.
Is there a non-allele frequency alternative?
I'm unclear how 'evolution is change in development due to ecology' ties in to inheritence without necessarily passing through the nexus of allele frequencies. And while this might betray the narrowness of my exposure to evolutionary theory, I am quite comfortable with Richard Dawkins' suggestion that an organism's genome be treated as part of the environment in which an allele must survive and propagate, just as competing species and atmosphere compose parts of the environment. Further, isn't limiting that subject to selection to development prone to being countered by exceptions whereby differences in 'fitness' manifest later in life?
Population genetics is based on the Hardy-Weinberg equilibrium - pretty much all of it is a build-on and embellishment of it.
wtf? not even wrong.
Speciation doesn't even mention punc-eq, from what i recall. but coyne and orr set out to discuss the evolution of reproductive isolation from the start, so it's hard to see how punc-eq would fit in.
And, yes, Coyne did get his phd with lewontin. He may have even overlapped with Rosie Redfield, who may have something interesting to blog about this topic.
yeah, i didn't recall mentioning PEQ, but it is referenced. though mostly to reject its non-scale independence.
Evolution: Heritable change in a biological entity. Inheritance is not just genes, but also epigenes, memes, constructed niche. To be sure that's pretty broad. Epigenetic change in a cell lineage leading to cancer is evolution. A population of farmers depleting their soil and becoming smaller over the generations due to malnutrition is evolution.
My perspective: Synthesis of standard evolutionary ecology, Baldwinism, dual inheritance, hierarchical selection. I used to be more of an orthodox sociobiologist. (I used to be a lot more left wing too. Is there a causal relationship between the changes in these two aspects of my worldview? No, not really.)
But perhaps the most important change in my views is that I have concluded that a lot of biological controversies that some believe have implications for human affairs really do not, or at least should not. Many of these debates are just esoterica of interest to scientists and popular science fans alone, and their outcomes are of little or no practical relevance to societal issues or even meaningful self-knowledge.
Razib: I hate to be the local smart ass but I think your deffinition should read "evolution is change in development due to a change in ecosystem over time" , remember, ecology is the study of ecosystems not the ecosystems themselves.
Mason
Changing allele frequency is one thing, but there is a lot more to genetics than the genes. Small changes in the expression of particular genes can have big implications on not only phenotype but behaviour as well, see: http://eprints.iisc.ernet.in/archive/00002457/
Its not so much the genes in the population that counts, its what they are doing. Having totally different alleles for a gene can have absolutely no effect if the gene is not expressed. And likewise 2 individuals with exactly the same alleles can be very different creatures based entirely on genetic changes to regions outside their genes, (promotors etc) So before you even begin to close things down to evolution relating specifically to allele frequency, you have to look at the broader picture. You know, the stuff that isn't a gene.
I think if you are going to use it as a definition of evolution then it is a very narrow definition that misses a lot of what actually goes on in nature. Sure one of the features of evolution is that it often involves a change in allele frequency in a population, but its hardly the only feature.
The excitement over the traffic this will bring, and the comment about the anti-abortion post being purposefully more provocative than informative, remind me of the Propaganda of the Deed type of anarchists. Throw a bomb onto Wall Street, and create a controversy to get people all riled up about whatever statement you're trying to make. In their eyes, that's more revolutionary than organizing a union or knocking on people's doors to talk to them about an important vote coming up.
During the countercultural chaos of the late '60s and early '70s, some young guys thought it would be revolutionary to just tear shit up all throughout Seattle (iirc), using the same logic. I believe it was called Days of Rage, but could be wrong. The women's movement organizers drove them out of town for acting like 13 year-old street punks.
Having Molotov cocktails lobbed into the blogosphere gets awfully tiresome.
Sure one of the features of evolution is that it often involves a change in allele frequency in a population, but its hardly the only feature.
Joss,
Can you please supply an example of evolution where a change in allele frequency is not involved?
By the way, promoters are parts of genes, and if two genes differ by a nucleotide or two in their promoters, then they are allelic to each other.
There may be more to biological evolution than changes in heritable characteristics of a population over time but that's the very best minimal definition that I know of [What IS Evolution?]. Nobody has come up with a better one for telling when something is biological evolution and when it isn't.
I don't agree with Bora about how to define evolution but I do agree with his opinion about Coyne. Coyne voices one of the standard misrepresentations of Gould when he confuses saltation with punctuated equilibria. I've tried to explain this at [Macromutations and Punctuated Equilibria]. It seems as though it's not only creationists who have trouble understanding punctuated equilibria.
when population geneticists speak of changes in gene frequency they do not mean changes purely within a gene. This choice of phrase is purely a historical artifact. An allele affecting protein function or gene expression (or any other type of functional change) fall well within the realm of population genetics.
Evolution: Heritable change in a biological entity. Inheritance is not just genes, but also epigenes, memes, constructed niche. To be sure that's pretty broad. Epigenetic change in a cell lineage leading to cancer is evolution. A population of farmers depleting their soil and becoming smaller over the generations due to malnutrition is evolution.
that sounds fine. i wonder how much of a long-term impact stuff like heritable epigenetics has...but i guess it only has to have a short term impact if there's a lot of stochasticity in the system.
joss, yeah, i don't think that that gene expression variation is that diff. as long as it is heritable. we might change the terminology (wording) some, but the models don't change that much. i'm not religious in getting huffy about sequence variation vs. cis-regulatory elements and what.
Razib: I hate to be the local smart ass but I think your deffinition should read "evolution is change in development due to a change in ecosystem over time" , remember, ecology is the study of ecosystems not the ecosystems themselves.
;-) is that a joke? that's not my definition but someone i was quoting. a smartass should be a closer reader!
The excitement over the traffic this will bring, and the comment about the anti-abortion post being purposefully more provocative than informative, remind me of the
art should provoke, science should inform (try).
Evolution is change in the frequencies of heritable innate characteristics within a population, from one generation to the next.
Most of the time, the support of heredity is genes, and so evolution can be reduced to genetic evolution. However, there are rare, but important evolutionary events where the heritable characteristic of note is not genetic in nature: the most commonly cited example is endosymbiosis. When a cell decides to "engulf" another, that's not a genetic change (although the change in question may be made easier or more efficient by genetic changes, those are peripheral to the "main action", which is a physical, non-genetic event, and could in theory occur without any gene change at all).
Now most people would accept that endosymbiosis can be a massively important evolutionary event (dissenters are free to revert to an anaerobic lifestyle at their leisure...)
On a more important note, I had to read the first quote from Bora's post about 4 times looking for the joke before I realized that it said Van Valen, not Van Halen. Way funnier the way I read it...
[i]By the way, promoters are parts of genes, and if two genes differ by a nucleotide or two in their promoters, then they are allelic to each other.[/i]
yeah my bad.. .got to check my definitions better in the future before weighing in :)
The "evolution is a change in allele frequency" idea has to be one of the most dishonest definitions in the history of science. At one time you have an organism representative of a species and having some specific set of genes. At a later time you have a descendant organism representative of a different species. There are, as in any database, adds, changes and deletions in the quota of genes. Possibly deletions could be classed as change in frequency, but not additions. This is important because the "random" genetic change postulated for neo-Darwinian evolution comes with an important statistical prediction - the rate at which novel variations appear in a population of a species is directly proportional to the number of members of the species. The subsequent filtering can partially mask this statistic but not eliminate it. While this statistic is true for allele variation it is not true, spectacularly not true, for novel genes for which the appearance rate would appear to be essentially independent of population size. The neo-Darwinian model of "random" change does not remotely match the data for gene addition, still less so for addition of groups of genes. This suspect definition appears to be an attempt to conceal the absolute necessity for some additional non-neo-Darwinian evolutionary mechanism to account for the areas in evolution where neo-Darwinism is notoriously unhelpful.
I have an example of a non-neo-Darwinian mechanism at nsof.co.nz.