Cornelius Hunter seems very confused.
…This brings us back to the UC Berkeley “Understanding Evolution” website. It abuses science in its utterly unfounded claim that “natural selection can produce amazing adaptations.”
In fact natural selection, even at its best, does not “produce” anything. Natural selection does not and cannot influence the construction of any adaptations, amazing or not. If a mutation occurs which improves differential reproduction, then it propagates into future generations. Natural selection is simply the name given to that process. It selects for survival that which already exists. Natural selection has no role in the mutation event. It does not induce mutations, helpful or otherwise, to occur. According to evolutionary theory every single mutation, leading to every single species, is a random event with respect to need.
He has forgotten what “adaptation” means. Of course he is correct that “Natural selection is simply the name given to [differential reproduction]”. And that (as far as we know), “every single mutation …is a random event with respect to need”.
And “adaptation” is the name we give to variants that are preferentially reproduced. So while he would be correct to say that “natural selection” is NOT the name we give to “mutation” (duh); it IS the name we give to the very process that SELECTS those mutations that promote reproduction. i.e. the process that produces adaptation.
Cornelius should spend more time at the Understanding Evolution website.
ETA: CharlieM points out below that…
When CH says that natural selection does not produce adaptations he is talking about individual organisms. He is discussing mutations in individuals and adaptations in individuals. Natural selection has nothing to do with the first appearance of an adaptation in an individual.
And that of course is the confusion – I hadn’t seen just where Cornelius’s confusion lay. Because, of course, the term “adaptation” is a population-level concept. At the level of the individual, the equivalent would be “advantageous mutation”. And that makes the Understanding Evolution website absolutely correct.
Possibly.
Lizzie,
I’ve provided quotes showing that Darwin considered the “struggle for life” to be a prerequisite of natural selection. There is no “struggle for life” in my scenario, because every organism lives and reproduces successfully — including the slower-reproducing ones.
When did population geneticists’ usage of “selection coefficient” become binding on everyone else’s usage of “natural selection” (including retroactively on Darwin’s)?
As I said, I have no problem with their usage of “selection coefficient” because it meshes just fine with Darwin’s definition in real life, where there always is a “struggle for life”.
As for your snide reference to “Darwin Literalism”, I wrote earlier:
That is hardly literalism.
It’s not binding on everyone, of course — only on those who wish to discuss evolutionary biology scientifically. “Natural selection” is a technical scientific term, and if you want to be comprehensible in a scientific discussion, you’d better use it in its technical sense. What Darwin meant by it no longer matters, any more than usage by Newton or Boyle matters. And in contemporary technical usage, your example undoubtedly represented the operation of natural selection.
ETA: An example from particle physics: the term “meson” was invented by Yukawa to mean a particle of intermediate mass (between the electron and the proton) that carried the nuclear force. Today, the term covers particles that are much heavier than the proton, and that have little role in binding nucleons together. You’re free to use “meson” in Yukawa’s sense, but you’re going to have a tough time having a conversation with a physicist if you do. (To add to the confusion, the first meson discovered, the mu meson, turned out not to really be a meson by any definition.)
Steve,
Population geneticists are far from the only evolutionary scientists.
Your analogy would make sense if Darwin’s definition no longer worked, like Yukawa’s, and the cognoscenti had moved on to a new one. But it does still work, and people haven’t moved on.
The conflict between “selection coefficient” and Darwin’s definition of “natural selection” only arises in my artificial scenario because of the fact that all offspring survive and reproduce. That is highly unrealistic, of course, so no one need be troubled by the conflict. In real life the environment imposes strict constraints on the survival and reproduction of organisms — it’s a “struggle for life”, as Darwin put it — and under those circumstances the conflict doesn’t arise.
Those strict real-life constraints are absent from my scenario, which is simply a thought experiment designed to demonstrate that heritable differential reproduction is (in principle) possible in the absence of selection, contrary to Lizzie’s claim.
Lizzie seems to have conceded this earlier in the thread, perhaps unwittingly:
In my scenario, no organisms reproduce less than they otherwise would. Everyone “maxes out”. So you have heritable differential reproduction without anyone reproducing “less than they otherwise would”. Selection isn’t curtailing anyone’s reproduction because there is no selection.
keiths,
That sounds like Creationist talk! Even in Darwin’s terms, a variation ‘beneficial to itself’ would increase in the population, and this he termed Natural Selection. If a variant gets to reproductive maturity faster (and there is no offsetting cost), then that is a variant beneficial to itself increasing in the population.
Essentially, there is always competition in a finite world, even when there is no direct competition for anything other than available slots in the ‘existence matrix’. In a finite population, existence is a resource which a faster breeding lineage will garner more of. Generation time is a parameter that can be tuned by selection like any other.
This has resonance with an interest of mine, the evolution of sex. Some treatments regard asexuality as ‘twice as fit’ as sexuality because they produce grandchildren twice as fast. I think they are wrong, for interesting reasons, but there is not enough space in this margin to go into that! But within the simpler scenario you present, rapid growth is a strategy which is entirely in accord with simple ‘Darwinian’ selection, even before the population geneticists got their hands on it.
Though if you are talking about an infinite world, there is indeed no selection. A set of individuals in the same infinitely-expandable flask, each surrounded by sufficient resource, is essentially the same as putting every new individual in its own flask, infinitely, with no interaction. You just get more of Flask Type A, exponentially.
So sayth WJM, noted evolutionary scholar.
BTW I missed your rendition of what the website should have said instead. So, as usual, something is wrong but you can’t say what the right choice would be.
Never.
And nor is your usage binding on mine.
Exactly. Keiths sounds like a creationist here, insisting on an idiosyncratically literal interpretation of a sacred text.
Darwin’s account of his own proposed mechanism is perfectly consistent with the application of the word “selection” to keiths’ scenario, and there is absolutely no reason to exclude separate out such a scenario from the rest of Darwinian theory.
The reason it’s worth mentioning at all, is that while keiths makes a good point that carving and building are both forms of making, natural selection is not one rather than the other. Its effect is to concentrate traits that tend to promote reproductive success in the same individuals, and by the same token cause individuals with those combined traits to dominate the population.
Reducing the prevalence of one trait is inseparable from increasing the prevalence of its alternative, because prevalence is a proportional term, not an absolute one.
So you get increased prevalence.
So even in the infinite monkeys scenario, in population genetics terms the selection coefficient of the slow breeders is .5 relative to the fast.
But mythical scenarios tell us very little about anything. If Darwin didn’t consider the case of infinite resources, it’s because he had too much sense. Nonetheless, it turns out his theory covers it, if you abstract the concept of “struggle for resources” to “struggle for prevalence”. It really was a brilliant idea (even if Wallace had it too).
Exactly squared!
A thought experiment that rules out competition for resources is a pointless experiment since it makes no difference to the result whether the wild-type and variant are together or in separate identical environments. In reality, resources are not infinite and there will be competition when nutrients get scarce. Lenski’s LTEE is one long round of replenishing the experimental environment by adding samples to fresh nutrient medium.
EL said:
I’ve already agreed that there is no adaptation without NS. BTW, you don’t have to show experimentally what is true by definition. You might as well run an experiment to see if you can generate any married bachelors.
Why did you chop off the rest of my sentence, William?
It was the important part: the “building” part that you objected to – the production of complex functional features that are vanishingly improbable in the absence of NS.
They actually tried it and it never happened.
EL said:
I’m not so sure about that, but I took that question to the other thread.
OK.
Allan and Alan get it.
Allan:
Exactly. There is clearly no selection in my scenario, yet Lizzie insists that there is.
To defend that odd idea she has resorted to more labeling:
Of course I’m doing no such thing. You don’t see me talking about ‘gemmules’, after all, and I would gladly abandon Darwin’s definition of ‘natural selection’ if there were good reason to do so. Lizzie hasn’t offered any.
Meanwhile, Alan also gets it (almost) exactly right:
That’s why I keep stressing that the minor conflict between “selection coefficient” and “natural selection” that arises in my highly unrealistic thought experiment is nothing to be concerned about because it doesn’t affect real-life scenarios.
The only thing Alan gets wrong is that the thought experiment isn’t ‘pointless’. It achieves precisely what it was intended to:
Lizzie,
That’s exactly the point I’ve been making to William. Selection eliminates unfit individuals, and by so doing it opens up more ‘slots’ in the population for good ones. Without selection doing its eliminative work, complex adaptations could not arise, so we rightly credit selection with helping to build or produce those adaptations.
Likewise, the house doesn’t get built without someone cutting boards to fit. He or she is simply removing wood, but that activity is nevertheless part of building a house.
Selection culls unfit individuals. No individuals are culled in my scenario, so there is no selection, as Allan states above.
I didn’t say anything about population geneticists. As far as I know, the meaning of “natural selection” is the same throughout evolutionary biology, and always includes your scenario. If you have evidence to the contrary, please present it. I have no difficulty finding biologists who explicitly state that differential reproduction without resource limitation is natural selection, and I can’t find any who state otherwise. (Whether they count as population geneticists is a little hard to say — anyone who’s writing about that specific question is doing population genetics at that moment, regardless of their specialty.) What was in dispute, at least among population ecologists in the 1960s, was whether the term “competition” should be applied to this case, but all were agreed that “natural selection” applied.
Yukawa’s definition would still work fine. It just turned out to be less useful than a different definition.
No, it’s really not essentially the same, because they’re in the same flask and thus form one population, and yes, there is selection. There are plenty of real-life scenarios which approximate the infinite-flask case: introduction of invasive species, colonization of new habitats, the initial stages of infection of a host by a pathogen. It’s true that all offspring don’t survive in these situations, but in all of them, the presence of one allele has little or no effect on the success of the other. Differential reproductive under these conditions can have large effects on the evolution and composition of the population, and the way evolutionary biologists and population ecologists and virologists describe the process is in terms of “natural selection”. There is no other technical term available.
keiths said:
Without natural selection, the exact same house can and will be built because “limited resources” is an aspect of Natural Selection. If there is no resource limitation,absolutely nothing that NS brings to the table cannot be accomplished via RM alone. NS is essential only in the generation of adaptations, not features, because an adaptation by definition requires NS.
Thus, NS is utterly unnecessary in terms of feature-building. Take away RM, though, and you get squat.
That seems to be saying “without natural selection there will still be natural selection.”
In a world that does not actually exist, you are right. And that’s the best kind of right eh William? Right with no consequences.
If my auntie had bollocks she’d be my uncle. In a non-resource constrained infinite universe anything built by RM+NS can also be built by RM.
And so?
I didn’t say a “no-natural selection” world actually exists. I said that without NS, RM could and would build every feature possible to make a point about what is actually doing the building.
So, you and keiths agree with me. Without NS, RM would build everything that RM & NS could build. Without RM, NS doesn’t do any building.
No, it’s saying that without natural selection, RM can and will build any feature, and probably a ot more features, than RM & NS can build together, and probably build it just as fast if not faster. It sorta puts a spotlight on what is actually doing the building in any meaningful sense of the word.
The unconstrained resources scenario is a red herring.
So is the idea that the only way you can get a “No NS” scenario is by eliminating death. It isn’t.
You can get a No NS scenario in AVIDA (and in RL) by having no correlation between genotype and fitness.
If you do that in AVIDA, the complex functions don’t evolve. If you provide a reproductive advantage for both simpler and complex functions they all do. If you only provide a reproductive advantage for the most complex function, it does.
Therefore RM alone doesn’t result in EQU (the most complex function).
RM+NS does, reliably.
Therefore adding NS to RM makes the difference.
Obviously you can’t have NS without RM (or without a population that lacks any variance e.g. as provided at some point by RM or other variance generating mechanism) because duh.
In AVIDA, there is no variance at the start. It is all introduced by RM.
Clearly RM is necessary – because it’s necessary for NS.
But alone it is not enough. NS is also required if you want to end up with the most complex function.
William: do you disagree with any of this?
Except in the real world it does not and cannot, as has been explained at length. And I explicitly set my example in a universe that does not exist and you simply ignored that.
So no, I don’t agree with you at all.
“probably”, “if not faster”, “sorta”, “probably”
If only there was some sort of way you could find out for sure rather then using words like “sorta”.
Without variation (e.g RM) NS doesn’t exist.
EL said:
No, it’s what I originally meant when I made the case, and that meaning has been referred to me consistently throughout this debate. “Limited resources” which organisms compete over is an aspect of natural selection. Your failure to realize what an actual lack of natural selection would mean doesn’t turn it into a “red herring”. It just means you failed to account for the lack of all aspects of NS, even though I made it clear throughout the debate what I was talking about.
IOW, your conclusion about what RM can produce is only true if you don’t actually remove natural selection, but only the parts of it that would result in a lack of feature generation. IOW, you’re only excluding the aspects of NS that would serve to make your conclusion true.
Let’s revisit some comments I made in this thread:
EL said:
Death is a natural selection process. So is limited resources. I’ve included this in my argument since the beginning, as I’ve shown. Now, suddenly, it seems you realize that you forgot that death and limited resources are aspects of natural selection, so you are attempting to move the goal posts away from that which I have been saying all along – because you realize I’m right. Without NS, RM will build everything (and more) that RM&NS can build, in terms of features.
As far as it being a “red herring” … a red herring leading us away from what? This has been an explicit aspect of my argument the whole time.
Variation is not the same thing as RM. Without variation, NS doesn’t exist. Without, RM, NS can still act on variations.
OMagain said:
No, I didn’t ignore it. I never said that a world without NS actually exists. It could exist, for a while anyway, in a computer program. I never remotely implied that I was asserting that somewhere in the real world NS doesn’t exist; I said that if it did not exist, then logically (as you and keiths agree), what I say RM would do is true.
I realize you don’t want to agree with me, but the fact is that you have agreed with the actual point I’ve been making throughout this debate, as illustrated by the past comments of mine that I’ve collected above which show that I have been talking about exactly what you and keiths agree to above.
RM is a source of genetic variation. Without new variations, variance will rapidly disappear.
Until they do, though NS will act on them, and may well produce new adaptive features in that time. Not that it’s a realistic scenario. RM doesn’t stop.
Not in the sense of “natural selection” as used by Darwin. Death can be discriminate or indiscriminate. If it is indiscriminate it isn’t selective.
But let’s not get back to quibbling over metaphors.
In AVIDA, as in life, you can have selection present (an environment in which some genetic variants have a greater probability of reproducing than others) or you can have no selection (your genotype makes no difference to your probability of reproducing).
In the former case, EQU evolves.
In the latter case it does not.
In both cases the starting population has no variance. All variance is introduced by RM.
Without RM nothing happens.
With RM very little happens.
With RM+NS, EQU evolves.
Do you disagree?
EL:
The fact is that what I have argued in this thread is true. Keiths and Omagain have agreed to it, although they seem loathe to admit it. I have consistently and explicitly shown (as the above quotes demonstrate) that when I talk about “Natural Selection” (as in an imagined world without it), I’m including death and limited resources. Randomized selection via “indiscriminate death” is still an aspect of “natural” selection – like meteor strikes, volcanoes and floods.
I have no idea what to say specifically about the AVIDA experiments because I have no idea what actual parameters have been set up or how the program works. If they include death and limited resources, then they have not removed NS; they have only removed some aspects of it. Under those specific settings and if their results are true, I would say that what natural selection parameters they have left in place kept RM from generating new features.
Now you can answer my question:
In a hypothetical world, EL, where there is no death and resources are not limited, wouldn’t you agree that RM does exactly what I’ve said it does? It will produce any feature RM+NS can produce, plus probably a lot more that RM+NS will likely never produce? And probably just as fast, if not faster?
William,
No, and this comment from the other thread explains why:
William,
I agree that finite resources impose a selective constraint, and that RM would eventually produce every possible adaption in the total absence of selection; but I disagree with what you are claiming in these threads: that selection has no role in building or producing complex adaptations.
Finite resources are a given in nature. Populations can only support a limited number of individuals. If selection were not awarding the limited “slots” to fitter organisms, on average, then complex adaptations would not arise on evolutionary timescales.
Again, this is exactly why Weasel works. Without selection, Weasel would not converge in your lifetime. With selection, it converges in seconds. In fact, I remember calculating somewhere that even under generous assumptions, Weasel without selection would take quintillions of human lifetimes to hit the jackpot.
I am a professional trained theoretical population geneticist (though much of my work has been on phylogenetic inference). I’ve been working for many years on a textbook called Theoretical Evolutionary Genetics. It is available free online and is nearly complete (it will continue to be available online). You can download it here.
If you look at it a few points may jump out at you:
1. When we consider different models, they may or may not have mutation, and they may or may not have natural selection. In our terminology, mutation can be present and natural selection absent.
2. Limited resources are not necessary for there to be natural selection. Yes, it was part of Darwin’s argument. But once the quantitative theory was developed, from 1903 on, it was realized that if one genotype is twice as fecund as another, there will be natural selection even when the population is growing without limit.
3. If there is mutation in an infinite population which has no natural selection, then yes, all possible genotypes will appear. But the number of them is vast. So if there is a stretch of bases 1000 long, there will be
possible haplotypes, and 
possible (unphased) diploid genotypes. Those are numbers way bigger than the number of elementary particles in the visible Universe. By contrast, if natural selection favors one haplotype, its probability of occurrence is high.
4. Oh yes, and saying that natural selection “constructs” an adaptation is not standard terminology in the field. People may or may not use a term like that or say that natural selection “designs” a phenotype, but they are speaking informally, bot technically.
I know many people have said the equivalent in this thread but I just wanted to try to explain what terminology the professionals in the field use.
Christ man, I can admit a logical possibility without being “loathe to admit it” and I’ve clearly said as much. I’ve also then said “If my auntie had bollocks she’d be my uncle” which was my way of qualifying my (and your) statement as essentially pointless. If 1 was 2 then 1 would be 2. How interesting. If you want to propose a thought experiment or similar just say as much.
If we’re in “if” world than anything and everything can be true. That’s not something I’m “loathe” to admit. If you start from a point then explore all points one step away, then explore all points from those new points and so on then yes, you’ll eventually explore every option, no doubt including ones that would be unavailable to RM+NS for whatever reasons.
You win. Congratulations. Now you can make some point about what “build” means and walk away correct.
Keiths said:
I’m haven’t claimed that. In fact, I’ve explicitly stated the opposite. See the other thread where I responded more completely, We are in agreement.
Would you have a problem with saying selection shapes a population, and that complex adaptations are one possible outcome of selected change?
No problem with the one-possible-outcome part. “Shapes the population” is not normal technical terminology.
I haven’t claimed it is the preferred technical description. I just wondered if it isn’t a better metaphor than construct.
How about “bring about”?
Thank goodness you showed up!
Lizzie, to Joe:
Don’t get too excited yet, Lizzie. 🙂
Joe:
That isn’t a realization; it’s a truth by definition. We agree on the mechanics of evolution and the mathematics of population genetics — it’s the definition of “natural selection” that is in dispute.
If you define “natural selelection” as anything that systematically changes allele frequencies in populations, then of course there is natural selection in my scenario. If you define “natural selection” as Allan and I (and our friend Charles D. and a whole bunch of other scientists) do, with the environment favoring some variants over others, then there is no selection in my scenario.
In pop gen, “selection coefficient” is defined as it is as a matter of convenience, and I don’t begrudge you that. What I don’t see is why the pop gen definition should prevail in the current discussion, which is about natural selection’s role in the production of adaptations. As far as I can see, Lizzie has offered no good reason why it should prevail.
I argue that it shouldn’t, and have provided several reasons why.
Here’s my latest. On the AVIDA thread, Lizzie and I are arguing (against William) that there really is no selection in the AVIDA no-selection case. Here’s how I put it to William:
keiths, to William:
William:
No, you have claimed that. You keep saying that selection doesn’t build or produce anything — only mutation does. That depends on a woefully narrow and unreasonable definition of “build” and “produce”.
I’ve offered counterexamples. By your logic, a woodworker turning a decorative table leg on a lathe hasn’t produced anything. A worker carrying bricks isn’t building a house; only the bricklayer is. Ditto for someone cutting boards to fit vs. the person nailing them in place.
What’s ironic is that even mutation doesn’t build adaptations according to your narrow definition. All it does is build sequences. The genes have to be expressed for the adaptation to be built, and mutation doesn’t express genes — it just changes them.
Allan:
Steve:
Steve,
Your position is self-refuting.
You argue that there is selection in the same-flask case but not in the separate-flask case. Yet the results in the two cases are exactly the same at any point in time.
If the results are exactly the same, then selection has no effect. In other words, there is really no selection in either case.
Steve Schaffner,
Well, I realise this is angels-on-a-pinhead stuff, but to me having every individual surrounded by a sphere containing sufficient resource is pretty much the same as wrapping that bubble in glass, or sending each new individual into a diifferent dimension. I was trying to find a scenario in which keiths’ ‘no selection’ might occur.
ISTM that, in this particular version of it, there is no more selection than if the fast-breeders were in a flask in the US and the slower ones in Australia. I think the fact that they form a unified population in one flask only matters if they interact in some way. There has to be some ecological context; something for which the individuals are competing, even if only a place in a finite population. If the population is allowed to expand without limit, then I don’t see how that ecological restriction is implemented.
Consider two completely unrelated bacteria which utilise completely different resources, but happen to be in the same ‘infinite flask’. In what way are they different from a grass species and a whale, in a more broadly circumscribed space? We don’t talk of selection between the latter, and if there was no competition, we wouldn’t between the former either.
Increasing the relatedness of the species increases the relevance of selective language, but only because the more alike they are the more they will compete. (eta: still talking asexuals here; sexuals have other dynamics).
Fecundity of different types can only be relevant when there is some kind of competitive milieu – some kind of ecological competition, even if only for finite ‘slots’. But I disagree with keiths that simple doubling of fecundity in itself evades that constraint. It only evades it when individuals are artificially kept from competing for anything.
I suppose (rereading my thoughts) that what could be argued is that the chance of sampling the more fecund genotype goes up indefinitely in the single flask. It would not take long before it was virtually impossible to even find a cell of the less fecund type, even though there were squillions of them. And of course the ‘force’ that causes this is indeed NS (pop gen version).
Elizabeth,
Heh. Population geneticists are nonetheless fond of the ‘infinite population’. It bugs the heck out of me sometimes, and here I am defending such a scenario!
No, I’ve talked myself out of it! An infinitely expanding space still has finite spaces within it. Within each of those (of any size you like) allele frequency is changing, to the ultimate extinction of the less fecund type in that local area.
NS it is!