Want to see Freeman Dyson and Richard Dawkins butt heads? It begins with a talk Dyson gave to an Edge conference in which Dyson (watch the whole talk) made these comments:
"By Darwinian evolution he [Woese] means evolution as Darwin understood it, based on the competition for survival of noninterbreeding species."
"With rare exceptions, Darwinian evolution requires established species to become extinct so that new species can replace them."
Dawkins rightly objects. Those are weird claims: the first is, I think, a misreading of Darwin. Darwin says something very different.
In the survival of favoured individuals and races, during the
constantly-recurrent Struggle for Existence, we see a powerful and
ever-acting form of Selection. The struggle for existence inevitably
follows from the high geometrical ratio of increase which is common,
to all organic beings. This high rate of increase is proved by
calculation,- by the rapid increase of many animals and plants
during succession of peculiar seasons, and when naturalised in new
countries. More individuals are born than can possibly survive. A
grain in the balance may determine which individuals shall live and
which shall die,- which variety or species shall increase in number,
and which shall decrease, or finally become extinct. As the
individuals of the same species come in all respects into the
closest competition with each other, the struggle will generally be
most severe between them; it will be almost equally severe between the
varieties of the same species, and next in severity between the
species of the same genus. On the other hand the struggle will often
be severe between beings remote in the scale of nature. The
slightest advantage in certain individuals, at any age or during any
season, over those with which they come into competition, or better
adaptation in however slight a degree to the surrounding physical
conditions, will, in the long run, turn the balance.
That's plain: Darwin was saying competition between individuals within a species was most important, but that competition to a lesser degree would occur between higher taxonomic categories. The synthesis narrowed and hardened this position; Dawkins reply represents the view that that there is no species selection (which I would disagree with) and that all the action takes place between individuals. Dawkins is closer to Darwin, and much closer to the modern understanding of Darwinian evolution, than Dyson is. (Whether Darwin or the modern understanding are actually right is a different issue, though.)
The second quote is also a little strange — Dyson has phrased it very poorly, as if Darwin was proposing extinctions first, replacement second. Darwin actually proposed replacement first followed by likely extinction (I have to assume that Dyson merely misspoke, because a literal reading would be far too absurd.) Darwin did place a greater importance on selection than I would, and also phrased extinction as a result of competition between species — something that again, the neo-Darwinian synthesis and Dawkins would not support.
The extinction of old forms
is the almost inevitable consequence of the productions of new
forms. We can understand why, when a species has once disappeared,
it never reappears. Groups of species increase in numbers slowly,
and endure for unequal periods of time; for the process of
modification is necessarily slow, and depends on many complex
contingencies. The dominant species belonging to large and dominant
groups tend to leave many modified descendants, which form new
sub-groups and groups. As these are formed, the species of the less
vigorous groups, from their inferiority inherited from a common
progenitor, tend to become extinct together, and to leave no
modified offspring on the face of the earth. But the utter
extinction of a whole group of species has sometimes been a slow
process, from the survival of a few descendants, lingering in
protected and isolated situations. When a group has once wholly
disappeared, it does not reappear; for the link of generation has
been broken.
Dawkins' argument is neo-Darwinian, that predators don't compete with their prey, for instance, they compete with their fellow predators of the same species to see which is better at catching prey. A more relevant situation, though, is to ask what happens when two predatory species are feeding on the same prey species—we can imagine that whichever one was better at catching the available prey would be better able to survive.
Darwin himself was arguing for one species outcompeting a second species, leading to the extinction of the second, so Dyson was sort of closer to being right…but again, Darwin's vision was a bit more complicated than species A → species B → species C, with extinction separating each succession cleanly; note that he mentions both branching speciation and that extinction can be very slow and prevented by isolation.
I'd say Dawkins better represented the conventional modern view of evolution than Dyson, and as usual, trying to fall back on support from Darwin is a bad idea, except in a historical argument. Darwin was a smart guy, but he's so 19th century — his ideas are interesting but out of date.
Dyson's counter-reply is more effective, even though it really doesn't address his original assertions and makes some new arguable claims. He claims that species once establish evolve very little; instead, I'd say that evolution is still going on at a good pace, but that selection is acting to conserve phenotypic stasis, and that that process is evolution, too. Dawkins and Dyson could probably have a good fight over punctuated equilibrium, and I might well side more with Dyson, but the subject doesn't really address Dawkins' argument in any obvious way.
Dyson says it's absurd to think that group selection is less important than individual selection; I sympathize and agree that evolutionary theory needs to incorporate more higher level mechanisms, but since group selection is categorically rejected by probably the majority of population geneticists, it's going to be a tougher sell than just dismissing them. Since Dawkins conceded that "the extinction of species is extremely important in the history of life, and there may very well be non-random aspects of it (some species are more likely to go extinct than others)", we're really down to arguing relative importance of these events.
There is a real, legitimate debate on this issue, and Dawkins would be an effective advocate for one pole. Dyson is a brilliant fellow, but I get the impression he's wandering a bit far from his domain of expertise here — I think I'd rather see a Dawkins-Moran head-butting session on this one.
- Log in to post comments
i'm always heartened to see Dawkins stand up and object to shit like that. in college i never had the guts to do it, and i still feel like a jerk for doing it now, so it's good to see someone being even more of a jerk even if he could be attacking more worthy targets.
i wonder if he's ever tried to comment at Uncommon Descent.
Lepht
In mostly unrelated but no less interesting news to the regular commenters/readers of this blog: D. James Kennedy has expired. Short retirement, eh? How's the song go? "Ding dong, the jerk is dead....."
A subject fairly close to Dawkins heart, this one :-)
Perhaps he could redress the balance a little and comment on spheres?
Dyson has also made some comments on the subject of climate change which, like his comments on evolution, are rather contrarian. I believe there was some discussion of his views on one of the other blogs but apparently the search feature doesn't include comments. The consensus of the commentators was that Dyson doesn't know what he's talking about.
Based on Prof. Myers' statement of Dr. Dysons' view, it appears that the latter doesn't understand allopatric speciation. In an allopatric speciation event, there is no necessity for the original and founder populations to come into contact again. As an example, there was a substantial overlap of Homo Erectus, Homo Neanderthalis, and Homo Sapians in Europe some 30,000 years ago. It is generally considered that the latter two species evolved from the former.
Why can't the eggheads at The Edge produce a more readable website?
We should include Nei (http://www.pnas.org/cgi/content/abstract/0703349104v1) in the Moran v. Dawkins battle as well.
Dyson has 'earned the right to be wrong'. He's done so much for physics that I can forgive him spending the last few years of his life being a (quite good-natured) contrarian.
Off-topic:
Very funny article about our friends, the orthodox Jews:
http://www.thisisby.us/index.php/content/the_time_i_helped_some_jews
They make Mormons look sensible.
Next time you hear a creationist saying "teach the controversy!" point them to this post. This is what scientific controversy looks like. As Michael Palin once said, it's not just saying "no it isn't."
I hate to say it but it's starting to look like Dyson is allowing his religious bent to cloud his judgment and making him attack Dawkins for reasons other than scientific ones.
Anyone want to start a pool on how soon this will appear out of the quote-mine? Maybe with a side bet on which creationist will grab it first?
Really? I thought it was well accepted that it can happen (The Price was right! Sorry, British joke), but that it's probably uncommon. There are a few good examples, but not many.
During coffee at ESEB, I was chatting to a friend who was saying that the Price equation is universally applicable - you could, if you so desired, use it for comparing species on different planets.
Bob
"As Michael Palin once said, it's not just saying "no it isn't.""
Yes it is.
I've never understood the need for or enjoyment of seeing two men "butt heads" over ideas. This conversation appears to be just that. As well as perhaps a fitting lab demonstration for the evolution of ideas. Is the advantage of cheering them on another demonstration?
It does appear that Dyson mispoke and it is painfully obvious that the two men are of two cultures with differing standards of civility, but that neither will back off except to charge again.
See, there's controversy in science! Obviously scientists can't keep their stories straight and therefore shouldn't be trusted to describe the origins of mankind to impressionable children.
Unlike Christianity, where all members adhere to exactly the same truths that are easily gleaned from the bible, and there are no arguments whatsoever about its content. These are the only people who should be trusted to describe the universe around us.
http://www.newscientist.com/article/dn12591-tasmanian-tigers-weaker-bit…
No it isn't.
He's behind you!
Population geneticists have a habit of getting stuck in the model of panmictic populations with no spatial distribution. When you go to a model with explicit spatial dependence built in from the beginning, interesting things happen: the proof that selfish traitors will always take over a cooperative population no longer goes through, for example. New effects appear, which aren't exactly group selection but are still interesting.
Of course, how much of this applies to actual wet-and-wiggly biology is beyond my ability to say.
"As Michael Palin once said, it's not just saying "no it isn't.""
Yes it is.
No it isn't.
Look, this isn't an argument, it's merely contradiction!
No it isn't.
I was interested to see this disclaimer on The Edge's intro to the exchange between Dawkins and Dyson:
I wonder if the "JB" who posted it subsequently got permission to do so.
Thanks for that link, Blake. Interesting paper --admittedly, I just now skimmed it, and will have to give it a closer read when I have more time, but the authors say:
I'm not sure that last sentence follows. Are not individuals increasing their fitness within a population at risk of extinction, or, put another way, isn't the risk of extinction a selection pressure on individual members of the population? (Following the logic that extinction necessarily entails a period of very low fitness for members of the population undergoing it)
Anyway, like I say, I need time to read the paper carefully.
Excuse my ignorance (haven't had formal biology since 10th grade, but I've read some popsci literature), but it seems to me that species selection certainly exists... obviously different species compete for the same resources and some are better in certain environments than others, and species extinctions do happen. The controversy and argument, to me, appears to be whether or not species selection is its own phenomenon. Dawkins goes further than individual selection, he likes genes themselves as the units of selection. Aren't individual, group, species, and indeed all higher forms of selection merely emergent behaviour of gene selection? We use different equations to talk about orbiting planets and falling apples, but we recognize they're both just special cases of the very same law of gravity. Am I mistaken to think that individual and species selection are special cases of gene selection?
Followed the link to Moran's blog and his comment section is fulled with those folks you seem to have placed in your "dungeon." They seem to come out whenever Dawkins or PZ are mentioned.
It's not surprising that Dyson, a theoretical physicist and mathematician, might not have his facts exactly right on Darwinian evolution.
I'd suspect Dawkins would be susceptible to similar misunderstandings about quantum physics.
Hopefully not too many minds were ruined by Dyson's wrongness on that point.
PZ, you REALLY shouldn't start spouting off tacit support for group selection theories without providing proper evidence for doing so.
there is a reason that 99% of working evolutionary biologists still support the individual selection model.
...and it's not inertia.
I am also not formally educated in biology beyond undergrad. But, as I understand Dawkins, his position is that all selection is best understood at the level of the replicators that do so with the greatest fidelity: the genes. He allows that, in most cases, the level of the individual organism is a convenient shorthand for what is going on at the level of the genome. But, in certain cases, I think his argument would be that the shorthand is not sufficient and to understand what is going on, one simply has to analyze what is happening at the genetic level.
Populations, which do not replicate with much fidelity at all, are too far removed from the action to even provide the integrative understanding that individual-level selection ordinarily does.
So, Jolly, I'd say it's more the reverse in terms of "special cases." In Dawkins's view, you never really even need to appeal to the level of the organism, but we do so in most cases because we can get the same answer in fewer terms. If that makes sense...?
Am I mistaken to think that individual and species selection are special cases of gene selection?
to put it bluntly, yes.
That would be fascinating, no question. Best done in writing, though. Although a taped discussion would be just as interesting...
Jolly B.:
Not in my opinion, FWIW :) But the question is, are group and/or species selection strong enough to overwhelm individual selection, and create traits that would not evolve by individual selection? When you choose the parameters right it seems that they can be. But are these just isolated cases, or a major driving force in evolution?
Blake:
Well, it's trivial to find examples of reproductive restraint that look like altruism, as Wynne-Edwards did. For example, living in very crowded conditions makes mice stressed and inhibits reproduction. But is that really altruism? A criticism of the group selection article points out:
Zahavi also points out, that a lot of microbiologists happily infer group selection while in natural animal populations people seem to have trouble finding it. He thinks that microbial "altruism" could also often involve dominant cells manipulating weaker ones.
justawriter (#11) wrote:
No, the Creationists will probably not quote-mine this. It goes against their assertion that "Darwinists" have shifted to worshipping Darwin, and evolution is nothing more than another religion. Try substituting "Jesus" for "Darwin" and see if it still works as an example of a religious way of looking at the world. Religions only "progress" by claiming to get back to the pure, uncorrupted, original source that started out perfect and good.
Tell a creationist that modern evolutionary biologists might not even bother to read Darwin and they find that as incredible as someone claiming to be a Christian, but not caring what Jesus really said. No, that can't be right. Not if you look at science through the same lenses you look at religion.
Icthyic: The reason that 99% of biologists reject all group selection is that it's damn hard to study, and most people seem to be bad at simply saying "We can't find a way to look at it; file it as unknown."
Even though such things as sexual reproduction clearly hint at mechanisms that occur at higher taxonomic levels.
Another example: the dogma of anti-Lamarckianism, for lack of a better word. Even though there have been hints in recent papers of non-synthesis modes of evolution (RNA writeback in plants; self-modifying DNA in immune system cells; "intentional" searching of DNA space from the ENCODE project), it is still damn hard to study whether or not there exists DNA computations that aren't just simple mutate, live-or-die, propagate, but instead occur by purely local feedback mechanisms. That isn't evidence that it doesn't happen - it just means that we are best of withholding judgement for quite a while.
frog:such things as sexual reproduction clearly hint at mechanisms that occur at higher taxonomic levels.
Would you mind unpacking this a little more? I ask out of genuine interest in what you mean.
The reason that 99% of biologists reject all group selection is that it's damn hard to study, and most people seem to be bad at simply saying "We can't find a way to look at it; file it as unknown."
meh, I disagree.
there have been quite a large number of studies experimentally testing group selection theories in the field, and the results so far have simply not been in the large supportive of most of the extant higher level selection models.
seriously, the level that PZ "hints" at implies a much larger division within the field of evolutionary biology wrt to role of higher level selection models than there is.
It's that that I really object to, the implication without going into detail as to what models have really been proposed by EVOLUTIONARY BIOLOGISTS, and NOT science philosophers, which is where the vast bulk of conceptualizations of higher level selection has come from.
I think you might be confusing "difficult to test" with "not based on evidence to begin with".
science philosophers often come up with interesting ideas, that, because they aren't actually based on extant data, aren't really testable in the field.
bottom line, all the actual experimental evidence to date (and before you start, I HAVE spent years in this field, looking at all the relevant papers) does not suggest an important role for higher level selection models in the field, and to imply otherwise suggests more wishful thinking than reliance on evidence.
windy:
Define "altruism" in a way which does not refer to emotional states or other anthropocentric not-very-tangibles, and we might be able to decide! ;-)
More seriously, I have seen people — in Q&A sessions after talks, and such — get into great spats over what counts as "altruism", which makes me think there's a certain amount of unproductive sparring over definitions. Better than that I can't really say — I'm a physicist!
I recall finding that Zahlavi article you linked some time ago and feeling rather underwhelmed by it. For example, the objection about the hosts not being able to evolve doesn't seem devastating to me when you're considering a situation like a host/pathogen system. Is it worth looking at? Probably. Is it a death knell? I dunno.
Maybe when I've got slightly less to fret about I can get Werfel's code and see what happens when I change this, that and the other.
D'oh! I forgot to say in my penultimate paragraph that the interesting thing is the difference in timescales between host evolution and pathogen evolution: the situation in which the host doesn't evolve is the limiting case of the pathogen reproducing much more rapidly than the host. When you're talking about something like a disease spreading through a human population, that seems like a reasonable first approximation to me.
[waves banner saying, "Remember, I'm a physics guy!"]
Well, pathogens are a bit iffy. Evolution of decreased virulence is one of the best examples of group selection for some (short intro here) But since pathogens within the same host are often related, it's hard to separate from kin selection. Since the group selection model of Werfel & Bar-Yam does not permit "mixing" between consumers, it could perhaps be argued that it's closer to kin selection than group selection in general.
One concept of fitness, inclusive fitness, seems to me to support altruism. Your inclusive fitness is a measure of how many copies of your genes make it into the next generation. Because your relatives share a proportion of your genes, (a smaller portion the more distant the relative, of course) it makes sense to be concerned about the welfare of your relatives. I'm not much of a theorist, but it seems to me that a good theoretical thinker could go from inclusive fitness to some aspects of group selection.
I doubt that the logistic growth curve is the result of altruistic reduction of reproduction as carrying capacity is approached.
The problem with classifying the Werfel & Bar-Yam host/pathogen results under "kin selection" is, as I recall reading somewhere, that the competition among pathogens occurs over multiple generations, via an interaction between pathogens and their host environment. (Feedback, via the environment, which impacts fitness several generations later isn't the same thing as reproductive restraint within the current generation.) This violates a simplifying assumption which I've noticed tends to get invoked tacitly when it really shouldn't, namely that "fitness" or reproductive success is constant across generations for individuals of the same genotype and phenotype. In a spatially extended prey/predator or host/pathogen model, overconsumption by the predators kills off the prey in that region, changing the fitness landscape because a predation style which used to work, works no longer.
What concerns me isn't so much whether a particular effect seen in a model should be labeled "group selection", but whether that effect (a) is robust against perturbations in the model and (b) has any counterpart in the wet-and-wiggly world. My choice of career rather restricts which of these desiderata I'm able to investigate!
Jim, you are right on inclusive fitness.
and
This article amply verifies Blake's observation of confused terminology in these issues. But I think that their case for the 'new' group selection not differing from kin selection is very compelling. Of course, other forms of group selection not covered by Hamilton and Price could still be waiting out there.
Ah, but once you plug variation across generations into plain old individual fitness models, all sorts of interesting things start happening even without group selection!
Herpes viruses hedge their bets
*Maxwell Smart voice* The old maximising geometric mean fitness trick!
Ichthyic wrote ", all the actual experimental evidence to date (and before you start, I HAVE spent years in this field, looking at all the relevant papers) does not suggest an important role for higher level selection models in the field, and to imply otherwise suggests more wishful thinking than reliance on evidence"
Of course, you can't answer the evidential question without solving the conceptual one. Notice that even biologists such as G. C. Williams (he counts, I hope, as an evolutionary biologist) conceptualized the issue so that female biased sex ratios count as evidence for group selection. And they're (female-biased sex ratios) not all that unusual. (See Williams , Adaptation and Natural Selection). Of course, at the time Williams wrote A&NS, we didn't know about many female biased sex ratios in insects. The point is that he thought that if there were, they would count as evidence of group selection.
No, female-biased sex ratios can be explained with individual selection. See this debate (Trivers against David Sloan Wilson and Elliott Sober)
Also recommended for fans of head-butting. Hilarious!
CJO: I haven't seen a convincing explanation as of yet why two-gender sexual reproduction is stable, if only individual selection is occurring. It would always be in the interest of the female's genes, at the individual level, to recoup her total investment by cloning herself. This is different from bacterial sexual reproduction, where just a subset of the genes are traded - in two-sex multicellular animals, the female is sacrificing half of her entire reproduction potential in order to shuffle genes many generations down the road. In the short term, self-cloning should be selected for.
Icthyic: give it a few centuries. What's the damn hurry? Lack of evidence isn't evidence of lack when sufficient data hasn't been collected yet because we don't have the right tools. Group selection hasn't been supported yet, but it hasn't been eliminated either. We'd require huge amounts of data over phylogenetic history, and models spanning 10s of millions of years to detect important group selection.
You're right that any biologist who advances it in a formal setting would be a crackpot. But that doesn't mean that speculation in informal settings is crazy. On the contrary, being aware of our limitations is the only way to leave a door open for future discovery.
Remember, until a year ago self RNA reverse transcription (as opposed to viral reverse transcription) as a method of evolution was unthinkable in most biological corners. How ubiquitous may it be? We don't know. A recent paper proposed that bacterial gene copies into arthropod host gene lines may be common - once again, crazy to have said it a few years ago.
Dyson is a delightful guy, who I and a couple of buddies took out to see The Matrix when it came out, but nonetheless like a lot of engineers, mathematicians, and physicists, he's heard that all biology/climatology=chemistry=physics, and assumed that his work in one field makes him an expert in all others.
I've read some popsci literature), but it seems to me that species selection certainly exists... obviously different species compete for the same resources and some are better in certain environments than others, and species extinctions do happen.
I plead to being unclear on the concept, but species rather than individual selection sounds like you'd have individuals suddenly saying to themselves "hrm, I'm a slightly more fit member of Species X, but in general, my species sucks compared to Species Y; time for me to lay down and die, I guess." I'm not sure how that works.
I don't think Wolbachia genes in arthropods are processed from RNA. Or what example of self reverse transcription do you mean?
Hi PZ,
Thanks for the link to Sandwalk! Great to know that the selectionist/neutralist debate continues. Now my life has meaning again.
Oh what a beautiful (mutational, non-adaptive and structurally-limited) rainbow!
Define "altruism" in a way which does not refer to emotional states or other anthropocentric not-very-tangibles, and we might be able to decide!
"A form of behaviour in which an individual risks lowering its fitness for the benefit of another."
Windy: I was thinking about the plant reverse RNA transcription published, I believe, in Nature about a year ago. They did a knock-out on some gene, then found the gene reappearing a few generations later. They were able to isolate it to RNA that was produced before the knockout, survived several propagations and then was transcribed back into the DNA. Cool stuff - the review process took several years because it so upset previous understanding.
"Whether Darwin or the modern understanding are actually right is a different issue, though."
Why would you say something like that? I give it until tomorrow until the quote miners start announcing to the world that PZ thinks evolution is wrong. You have to be more careful.
I don't think they have been able to pinpoint that RNA is the cause of the hothead reversion, but it is a reasonable inference if they are able to exclude all DNA-based mechanisms (I don't know what the current situation of the competing explanations is).
Here is an example of RNA->DNA where I think the conclusions are more solid:
RNA-templated DNA repair
I never cease to marvel at how Freeman Dyson continues to have this reputation as a brilliant thinker, whereas in fact he talks with the conviction of a genius about things he knows next to nothing about, like climate science, or now, it appears, biology.