More Recent Comments

Tuesday, May 06, 2014

Answering creationist questions about Neutral Theory

Many of the creationists are just learning about Neutral Theory for the first time in their lives. (The basics were published in the late 1960s—over 45 years ago.)

Vincent Torley (vjtorley), a philosopher from Australia, has struggled with the idea for several weeks and now he thinks he has some challenging questions for evolutionary biologists. Those creationists are really fast learners. It took me several years of study before I really grasped the basic concepts and the theory behind population genetics. Torley's questions are at: Will the real Neutral Theory please stand up?. The obligatory piling on by "News" is at: Is there a real neutral theory of evolution?.

Torley begins with ...
So I’d like to ask Professor Moran a few questions:

1. Do you agree or disagree with the view expressed by Motoo Kimura that natural selection is necessary to explain evolution occurring at the morphological level?
Some evolution at the morphological level can be attributed to natural selection and some is due to random genetic drift. The latter category includes neutral morphological changes and a small percentage of detrimental morphological changes.

I was influenced in this view by Masatoshi Nei's book Molecular Evolutionary Genetics (1987). He wrote the following on page 422 ...
In the strict form of neo-Darwinism, very little room is given to neutral variation, and virtually all morphological and physiological characters are thought to be products of natural selection (Mayr, 1963; Ford, 1964). Certainly, most morphological and physiological characters are well-adapted to the environment in which the organism lives, and there is no question about the importance of natural selection in the formation of intricate morphological characters. However, are all individual differences in morphological and physiological characters adaptive as claimed by extreme neo-Darwinians? More than 4 billion people live on this planet, and all of them except identical twins are different with respect to various morphological and physiological characters. Are all these differences adaptive? Is random genetic drift unimportant for generating morphological and physiological diversity among organisms? I doubt it. It seems to me that in some morphological characters a substantial part of genetic variation in nonadaptive.
So, the answer to your question is "yes;" natural selection and random genetic drift are both necessary to explain evolution at the morphological level.
2. How do you respond to Dr. Gert Korthof's Kothof’s claim that the neutral theory "is not a theory of evolution," because it "is not sufficient to explain complex life and adaptations"?1 If not, why not?
I respond by saying that Gert Korthof Kothof—whoever that is—doesn't understand the definition of evolution [What Is Evolution?]. Neutral Theory and random genetic drift are integral parts of evolutionary theory. They are not very good at explaining most adaptations but there's a lot more to evolution than adaptations.
3. Can you point to any complex structures, functions or behaviors which you believe could not have arisen in the absence of natural selection? (You’ve already nominated the change occurring in the human brain over the past few million years as an event in which natural selection played an indispensable role; what else would you put on your list?)
The vast majority of complex structures seem to be adaptations of one sort of another. I suspect there are many "functions" and "behaviors" that are neutral, or even detrimental, but it's difficult to rule out any adaptive component.
4. In which of the following events do you see natural selection as having played a decisive role: the origin of eukaryotes, the origin of multicellularity, the 20-million-year Cambrian explosion, the origin of land animals, the origin of the amniote egg, the origin of angiosperms, and the radiation of mammals immediately after the extinction of the dinosaurs?
I think that natural selection played an important role in all of those events.
5. Or is it simply your contention that natural selection, while not playing an important role in the origin of complex structures and novel morphological features, exerts a refining and purifying effect subsequent to their appearance, weeding out non-viable life-forms?
No. I have always contended that natural selection plays an important role in the origin of most complex structures and novel adaptive morphological features. There are likely to many "novel morphological features" that are non-adaptive.

It's also true that negative natural selection acts as a break on evolution by preventing detrimental changes and "weeding out non-vaible life forms."

HTH HAND


1. Gert Korthof himself supplies the complete quotation in the comments below. He says, "Corneel, thanks for defending me! Small correction: the blog is mainly in Dutch, but the website you linked to is in English. Obviously: the misquote IS in English. And now the full quote: 'Please note that 'the neutral theory of evolution' is not sufficient to explain complex life and adaptations. In that sense it is not a theory of evolution. However it is accepted that the neutral theory explains a lot of differences in DNA. Kimura: 'Of course, Darwinian change is necessary to explain change at the phenotypic level -fish becoming man- but in terms of molecules, the vast majority of them are not like that. (7)'" Looks like Vincent Torley might have quote mined a scientist. Isn't that amazing?

75 comments :

NickM said...

Creationists seem unable to hold in their heads the idea that multiple natural processes can be in operation, and that some of those processes can explain one class of observations, and other processes can explain other classes of observations.

Earth to creationists/Sal/vjtorley: natural selection is the main explanation of complex adaptations (e.g., eyes). Neutral processes are the main explanation of non-adaptive changes (e.g., sequence change in junk DNA). The fact that most molecular evolution is neutral makes sense because most of the genome of humans (and other large-genomed organisms) is junk. This statement does NOT mean ALL genomic change is neutral, and basically any evolutionist would agree that selection plays an important role when it comes to adaptive changes in functional DNA such as genes. Also, most molecular evolution being neutral says nothing in particular about the evolution of morphology, which is controlled by non-junk DNA.

There is more that could be said (as always; see in particular Michael Lynch), but that's a good first approximation. This should have been obvious to you guys from the beginning, if you had bothered to think and read about it for 5 minutes, rather than trumpeting your ignorance in blogposts. Don't you ever get embarrassed about getting such fundamental basics wrong? Don't you see why this level of ignorance, when coupled with accusations that it is the evolutionists who are wrong / evil / misleading the world, is totally infuriating to professional biologists and guarantees that you will be seen as nothing but malicious, intellectually lazy cranks?

(end rant)

John Harshman said...

This is Gert Korthof (note spelling):

http://wasdarwinwrong.com

I suspect the creationists have quote-mined him on this subject; and I bet he meant that the neutral theory isn't a *complete* theory of evolution. Of course nobody ever said it was.

Unknown said...

I don't agree that selection and drift are "multiple natural processes". There is population resampling. Darwin called this natural selection. The first generation of the synthesis developed models of this resampling process. They also made law of large number approximations. These approximations are what are now referred to as selection and the difference between them and the full models is referred to as drift. But that is reification at it's worst - you have one process that accurately describes what's going on and then you split it up into two "processes" that aren't really happening.

I've asked this before: Has anybody found it useful to look at the drift term in isolation in cases apart from neutrality? Can anybody give any good reason we can not ditch selection and drift in the modern conception in favor of population resampling? The drift part becomes really complicated as soon as you have |s|>0 and for that reason alone is rarely treated in depth, which in turn gets people confused because they simply extrapolate from the neutral case, which then leads to misconceptions about drift when there's weak selection for instance.

Can we please put this particular humpty dumpty back together again?

(end rant)

Joe Felsenstein said...

@Simon Gunkel: So you just want to rename selection and drift "population resampling"?

By the way, we don't just throw up our hands and refuse to deal with genetic drift if selection is noticeable. There are nice (and accurate) formulas for fixation probabilities when there is both selection and drift, such as Kimura's fixation probability formula of 1962.

Joe Felsenstein said...

... and geologists lie when they say there are landslides. There aren't. They are nothing but rocks moving under the influence of gravity. And chemists and physicists lie when they talk about Brownian Motion. There is no such thing -- it is just molecules banging together. Nor are there any such things as rocks -- they are nothing but molecules bonded to each other. Nor are there any such things as molecules ...

:-)

Rolf Aalberg said...

For what it may be worth, I knew the answer even before I'd read Prof. Moran's reply I think that natural selection played an important role in all of those events.
I strongly suspect that vjtorley and others of his kind never seriously have applied brainpower to understand evolution form within, so to speak. They are always looking for arguments they think may cast serious doubts on the entire framework of evolutionary thought and theory.

I see it as a fundamental fact that noone yet has been able to design a working workaround to the theory of evolution. That represents a strong indication that the ToE with or without added qualifiers like “Neutral Theory” et ceterea, per of today is the one and only viable solution to what to some appears like a mystery, often touted as a reason to invoke the magic of the religious myths, supernature (and for some, “the little greeen” men theory) as better alternatives.

Until anyone presents a radically different solution, I suggest he/they do some serious research into the alternatives from a scientific viewpoint and put aside their more or less pronounced religious motives. Religion may be good for some people and there’s nothing wrong with that. It may be good for the spirit, but just doesn’t answer questions about the real (non-spiritual) world.

The ToE became both a theory and a fact over 160 years ago.

Creationists are desperately searching for the small tuft that will overturn a big load.

Corneel said...

Gert Korthof blogs at the Evolutie blog (in Dutch, sorry guys). The review where the quotes derive from is here. Gert strongly opposes creationists on his blog, so it is funny to see Vincent Torley quote him in defense. I cannot speak for Gert here, but it seems to me that he was thinking of morphological evolution in his quote. In fact, the entire quote is:
Please note that 'the neutral theory of evolution' is not sufficient to explain complex life and adaptations. In that sense it is not a theory of evolution. However it is accepted that the neutral theory explains a lot of differences in DNA.
...which, as you will recall, is what got this entire discussion going: the number of polymorphisms between the human and chimp genome.

heleen said...


you have one process that accurately describes what's going on and then you split it up into two "processes" that aren't really happening

So, if I cross vestigial F1 in Drosophila : vg/+ x vg/+, and find 1/5 of the F2 offspring are vg/vg, I'm not allowed to split this into 2 processes, Mendelian inheritance and differential survival of genotypes?

heleen said...

In other words, VJTorley quotemined Gert Korthof.

AllanMiller said...

Although Korthof does commit the 'sin' of conflating evolution with selection.

Rolf Aalberg said...

As a Norwegian I can read Dutch but to me it reads like bad "Pidgin-English";)

SPARC said...

He's the same Gert Korthof who contributed to Why intelligent design fails.

Corneel said...

Hey!...and what was that again about Norwegian and Danish? ;-)

Unknown said...

@Joe: If I had a choice I would call the process of population resampling "selection", because that's how Darwin introduced it. I'm well aware of Kimuras 1962 formula - my population genetics may be rusty, but not so rusty that I couldn't explain what he did there.
The key thing to me is that there are nice formulas for selection in isolation (law of large numbers approximations to the process) and there are nice formulas for selection and drift combined (no matter whether thats Fisher-Wright, Moran or the diffusion approximation), but the formulas for drift on it's own get ugly in all but the neutral case. And when you can't neglect drift, you use complete models anyway - you never derive the ugly drift term and combine it with the selection term, you just use an adequate model for the combination.

The problem I have is not with the maths. The problem is with the reification of the maths into two processes. Any random variable X that has an expected value E(X) can be rewritten as E(X)+X-E(X), i.e. written as the sum of the expectation and a centered random variable. That's useful if you want to use the central limit theorem. But it's generally not the case that this partitioning reflects two separate processes.

If you apply the partitioning to individuals, it becomes fairly obvious that there's little sense in treating drift and selection as such:
The number of offspring an individual has is a random variable. For instance an individual might have 0,1,2,3 or 4 offspring with probabilities 0.2, 0.15, 0.3, 0.2 and 0.15. The expected number of offspring is thus 1.95. The individual has 2 offspring. The partitioning now tells us it has 1.95 offspring through selection and 0.05 offspring through drift. If the individual had 1 offspring , it'd still have 1.95 through selection and -0.95 through drift.

Can we partition this way mathematically? Sure and it is useful to do so. But to think of these partitions as actual processes gets ridiculous.

As to the second post. I did not claim people were lying. I claimed that they reified a mathematical partition of a process into two different processes. The landslide example might be illustrative, though. Sedimentologists look at the distribution of grain size in clastic sediments (usually on a log scale). Landslide deposits are badly sorted (high variance of grain size) and very coarse (mean grain size usually gravel). Now, what they don't do is claim that there are two processes involved in landslides, one of which is responsible for the mean size and one for the variance...

@heleen: One can at least make a decent case for that partition to reflect two processes - sampling of gametes in reproducing organisms and sampling of organisms.

Gerdien de Jong said...

Pidgin-Dutch: especially Danish: it sounds like extreme Grunnings.
Norwegian is somewhat redeemed by being pronounced clearly.

B Moritz said...

Dutch is in fact a little bit understandable for me as German native speaker, but it is better in spoken and written version together...
And I have read a couple of post from Gert Korthof (mostly what I could find in English)

VJTorley has the tendency to quotemine everybody in very interesting ways: he has written a very long article on a neoscholastic philosopher criticizing the ID movement and he quoted and misunderstood Thomas Aquinas in a very interesting way - I learned a lot by analyzing where Torley went "off the road".

Claudiu Bandea said...

About a year or so ago, our host Lary Moran wrote a post entitled “Mutation-Driven Evolution” ( http://sandwalk.blogspot.com/2013/06/mutation-driven-evolution.html). The post was a preview of a new book by Masatoshi Nei entitled “Mutation-Driven Evolution. Larry ended his post saying: “I can't wait to get my hands on a copy of this book. Look for a review in a few months.”

I don’t know if Lary kept his promise or not, but in the outline of his book Masatoshi Nei writes that, unlike Motoo Kimura and Jack King, who believed that phenotypic evolution (in contrast to molecular evolution) is caused primarily by natural selection, he believes that both molecular and phenotypic evolution are primarily caused by mutation.

John Harshman’s reaction was: “Larry, this seems to conflict seriously with your (and my) preferred definition of evolution”.

For whatever reason, Larry chose to be silent on this difficult and inconvenient issue about the Neutral Theory because, I presume, he wanted to read the book first. It would be enlightening, if Larry and John would be willing to discuss the significance of Neutral Theory and the Mutation Theory in explaining evolution.

Corneel said...

@Simon Gunkel
[...] the formulas for drift on it's own get ugly in all but the neutral case
...because when it's not the neutral case, it is NOT drift on its own.

The partitioning now tells us it has 1.95 offspring through selection and 0.05 offspring through drift.
No, it doesn't. Selection and drift both operate on populations, not individuals.

Honestly, can you really not see that selection and drift are completely different processes? Suppose that I start walking to the library, but get a ride halfway. Should I then abolish the use of the words walking and driving , simply because I did not complete the journey by one of these distinct means of transportation alone? Should I now start using the verb "move", because walking and driving weren't really happening?

Joe Felsenstein said...

My rather strange comment above about geologists lying was a parody of the comment before it, a typical stupid taunt by JoeG. So I parodied it. JoeG's comment seems now to have been deleted by Larry (understandably).

Joe Felsenstein said...

@Simon Gunkel: I think we have a semantic disagreement here. You want to call genetic drift and natural selercton both "selection". I don't see that this is a useful thing to do, and certainly almost everybody in population genetics does not use the term "selection" in that way. Feel free to use the term the way you want to, but don't be surprised if people misunderstand how you are using it.

Unknown said...

But if these are two different processes, then I would expect somebody to actually care about what either of them does on it's own in all cases. Let's use a very simple model: A Moran model for haploids, ignoring all birth-death pairs in which the allele frequency of the allele under consideration does not change. In this case the complete model looks like this:
For frequencies of 1 and 0, the frequency stays constant.
For other frequencies the change at any step is
1/N with a probability of 1/(1+e^(-s))
-1/N with a probability of 1/(1+e^s)
Now let's partition that increment. The selection part is a constant increment of
(1/(1+e^(-s))-1/(1+e^s))/N=(e^s-1)/((e^s+1)N)
The drift part is a random variable with increments of
(1+(e^s-1)/(e^s+1))/N with a probability of 1/(1+e^(-s))
(1-(e^s-1)/(e^s+1))/N with a probability of 1/(1+e^s)

If you are telling me the latter version is better, I disagree strongly.

Honestly, can you really not see that selection and drift are completely different processes?
Honestly, I really can't see it. What I see is that I start out with a nice stochastic process, like a biased random walk or polynomial resampling, or a Wiener process and then I partition that into an expectation and a centered component. And I find that partitioning useful sometimes, but I disagree with the notion that the partition is something to be reified. In physics it's sometimes useful to break up a force vector in components in the directions of your coordinate system. But if you stated that it made sense to stop talking about gravity as a fundamental force and rather have three fundamental forces x-gravity, y-gravity and z-gravity, I would argue that you've taken something nice and made it less intuitive and ugly.

No, it doesn't. Selection and drift both operate on populations, not individuals.
You can easy get from there to populations, simply by taking means. In a finite population you end up with a number of alleles through selection that is not an integer through selection and often enough with a negative non-integer number through drift. It still leads to the same weirdness.

NickM said...

I was going to reply and try and untangle the layered confusions here, but then, what's the point.

Corneel said...

@Simon Gunkel
Very well. So how would you treat mutation? Suppose we have irreversible mutation and drift. This would also result in a biased random walk, very similar to selection and drift acting together. Would it be silly to reify the process of mutation, even though we can identify the biological cause for it ? If not, why not? Because it is not a sampling process?

Unknown said...

@Joe: I'm using population resampling, precisely because selection has come to mean something different from that. I'd argue that shift happened when the synthesis started to include people with little mathematical training. It's something I lament, but there is no going back now. But I do think that treating selection and drift as two processes, rather than as mathematical abstractions that partition one process is something that doesn't help.

I mean, in physics it took a while to understand that electrostatic and magnetic phenomena were two sides of the same coin and then some more to get the Maxwell equations that describe the electro-magnetic force. And population genetics pretty much started with a unified theory of selection and drift - it even started without separate terms for them. And now we're at a point of "completely different processes". I think this goes beyond semantics at some point.

Unknown said...

@Corneel. It wouldn't be silly to treat these as different processes, because they are. But I don't think partitioning resampling into selection and drift is meaningful. There's nothing that would allow us to distinguish them and both are defined by an abstract demand. We'd like to have a centered component? There we go.

gert korthof said...

Corneel, thanks for defending me! Small correction: the blog is mainly in Dutch, but the website you linked to is in English. Obviously: the misquote IS in English. And now the full quote:
" Please note that 'the neutral theory of evolution' is not sufficient to explain complex life and adaptations. In that sense it is not a theory of evolution. However it is accepted that the neutral theory explains a lot of differences in DNA. Kimura:
"Of course, Darwinian change is necessary to explain change at the phenotypic level -fish becoming man- but in terms of molecules, the vast majority of them are not like that." (7)"

Faizal Ali said...

gert korthof,

Would you consider contacting Torley directly and informing him of his "error"?

gert korthof said...

Larry said "I respond by saying that Gert Kothof—whoever that is—"

You are right, that person does not exist (according to google at least)!!!
You assumed creationists always quote correctly, but the quote was wrong and out of context. The context was (1999) Kimura saying :

"Of course, Darwinian change is necessary to explain change at the phenotypic level -fish becoming man- but in terms of molecules, the vast majority of them are not like that." (7)
'Fish becoming man' was what I meant with evolution.

Chantal said...
This comment has been removed by the author.
Corneel said...

@Simon Gunkel
Of course we can distinguish those. As is the case for mutation, we know the biological cause for selection: differential reproduction between genotypes. In many cases, we can even identify the variation in morphology/ behaviour / physiology that is associated with it, and we can experimentally demonstrate that removing that variation will also remove the constant increment, while the random variable remains. Does that not show that selection is a very real phenomenon, distinct from mere sampling error?

Unknown said...

@ Corneel: The error here is the "the random variable remains" part. The random variable changes as s changes. I made this explicit in the model I gave above and you can make this explicit for other models as well. Differential reproduction between genotypes affects the probability distribution for the change in allele frequency per time step. Selection in the modern use is restricted to the change in the expected change, while the change in the variance, skewedness, etc. all sits in the drift part.

Larry Moran said...

@Claudiu Bandea,

Sorry to disappoint you but I'm not going to write a review of Nei's latest book.

You're going to have to read it yourself.

Claudiu Bandea said...

The relevant issue here is science, not the book: do you agree with Nei’s proposition that Kimura and King’s perspective was wrong and that the molecular and phenotypic evolution are primarily caused by mutation?

Corneel said...

@Simon Gunkel
Please correct me if I am wrong, but doesn't mutation change the probability distribution too?

Unknown said...

@Corneel: Sure. But to look at this yet another way. Let's say I describe the population through a phase space. I take any gene, and give the number of copies for all alleles. That gives you a countable number of states, hence it is possible to write down a transition matrix. Now, I can write a transition matrix for mutation. I can also write a transition matrix for population resampling. I can get the combined effect by multiplying these matrices with each other. I can do the same thing for migration. I can not generally do this for selection. I can not do this for drift. Because both "processes" map existing states to states that are not in the phase space (because they use non-interger numbers of copies for the alleles). It's only when I put them back together that I don't run into problems.

There's just no good reason to treat them a separate processes. In the case of mutation, there is a good reason in that they are different physical processes. But both selection and drift are the same process in which individuals produce a number of offspring before they die.

Unknown said...

I apologize if I seem daft. It has been almost 40 years since I last did any serious reading on genetics, drift and selection. But, don't both drift and selection rely on differential rates of reproduction? If that is the case, is drift really separate from selection, other than the fact that there is no known adaptive benefit in drift?

Anonymous said...

NickM ,

It seems that it has become a habit of yours not to comment on "difficult issues"... I'm still waiting for a comment from you regarding the origins of life.... :) Remember…? Fortunately, someone already explained it to me why you don't even dare to comment... because if you did, you would make such a fool of yourself and that would be stuck to every job application you would ever make... So... You are not taking any chances.... So much for scientific integrity…whatever that means…

Anonymous said...

I don't expect Larry do review very book on evolution there is... Nobody does... However, if Maserati Teppaniaky were leaning, at least a bit towards ID, I'm pretty sure Big Larry would find time in his busy schedule to review his book... Am I bluffing....

Anonymous said...

I've got to give credit to creationists... They are very diligent in their research and they never offend a soul even though Larry fires at them with a lot of s..t... If there is a God, he certainly had something to do with neutral theory of evolution lacking evidence issue...
Can anyone see it but me...?

Unknown said...

So Quest, what does this have to do with origin of life? Evolution, natural selection, drift, etc has to do with the changes in life after it has been formed. And it does it quite well, I might add.

But if you want to talk about the origin of life, I'm sure that people here will be willing to talk about that as well. But please remember (put your big boy diapers on) they are different issues.

Anonymous said...

Really..? I can bet you will disappear from this blog or you will ignore me like NickM the moment we will talk the origins... do you feel lucky.. or...confident enough...?

Unknown said...

Quest, not a problem. I will debate possible, and probable, origin scenarios with you any day. But are you? More importantly, are you willing to admit, as all scientists are, that the origin question is still undecided? Unless you are willing to admit that, there is no point talking.

Unknown said...

Quest, do you also see fairies and leprechauns?

Anonymous said...

Acadia Tonga,

The question is; will you admit to the obvious and logical inference...?
You see... we haven't even started the discussion and you are back-pedaling already... ;)... Can you see my problem...?

Joe Felsenstein said...

Quest,

... backpedalling like you do when we ask you what evidence you would accept for common descent, or whether you even understand the methods used to accumulate such evidence? You raised the issue of common descent, then when we asked you this question, Brace Sir Robin ran away. Only to appear hear and complain about other people avoiding his questions. Hilarious!

Joe Felsenstein said...

Typo: Brave Sir Robin ...

Corneel said...

@Simon Gunkel
Let's go back a little. You just conceded that mutation would affect the probability distribution, just like you argued for selection, yet you seem to feel no urge to lump it in with genetic drift. Why is that? That is because mutation is clearly linked to a distinct biological phenomenon. Look, I appreciate that selection and drift have a lot of similar properties, because they both are samping processes. Also, I can see your point and I freely admit that I am sensitive to your appeal to mathematical parsimony. However, in this matter biological interpretation trumps mathematical aesthetics. Selection has a clear biological interpretation as the only evolutionary mechanism that will result in adaptation. We can use it in empirical research by identifying the heritable variation in fitness that drives this process "Resampling"just will not suffice in this respect.

gert korthof said...

lutesuite, not at this moment. I am more disturbed by Larry's rude behaviour. He managed to introduce a second misspelling of my name in a footnote he added to his blog. There are 6 persons on this blog who spell my name correctly, Larry is not able to do so. I wonder whether he does this on purpose... Anyway, it does not show attention to detail and fact checking habit of a scientist.

AllanMiller said...

Perhaps Quitton could indicate what evidence they would accept for a plausible non-interventionist origin of life? Although perhaps, for tidiness, in a thread that is actually about the OoL.

AllanMiller said...

I have debated Simon on this in the past, and have been forced to concede that allele frequency change is governed by generational birth/death sampling, and that it's just one fundamental process. Mutational pressures, gene conversion etc can supply additional alleles or drive, but sampling is always in train. There is a continuum of s values, and practical s values themselves can vary stochastically, and have differential impacts dependent upon population size etc. s=0, or s<1/2Ne, are the 'null' points in that continuum, where selection is 'off'. But turning up the selective 'heat' in tiny increments does not cross a boundary. I do not have a proof, but I think it ought to be possible for a population to adapt even if s values never exceed 1/2Ne, since successive fixations move closer to the LLN expectation. Fixed alleles should still be biased towards those with positive s, if only marginally.

For practical purposes, it's useful to distinguish the definitively adaptive and the effectively neutral, but the biased and unbiased components intermingle, and much can go on below the radar.

Unknown said...

See, I would agree with this assessment, if selection had retained its original meaning. As I noted before the concept of Darwin is identical to what now is selection and drift combined. Darwin did describe selection as population resampling in the sense I'm using it now, but that process got split into drift and selection.

Corneel said...

@Allan Miller
Fair enough, I agree with what you are saying. But I still resist the idea that selection cannot exist as a distinct concept, but should be subsumed under "population resampling". That would divorce the model from what we are trying to explain in biological systems, namely adaptation. It is the biased component that explains adaptive evolution, so why not put a name to it?

Unknown said...

Well, we still have selection coefficients. My qualm is only with the idea that drift and selection are different processes and the biggest issue with that is that selection coefficients are still relevant parameters in the drift term. You can't really isolate the effects of differential fitness from the process.

If I flip a biased coin, there's one process: flipping a biased coin. I still need to talk about the bias when I describe it. In the same way, I can talk about selection coefficients and don't ever need to contrast selection and drift.

Instead of asking: "Is selection more important than drift here?", I can ask "How large are my selection coefficients? Can I distinguish them from the null model of neutrality?" Not only have I avoided treating drift and selection as different processes, I`ve gone from a yes/no question, to a quantitative one. If I answer the second question, I can give you a value for s, with error bars.

In short: I don't want to name a biased component, I'd prefer to name the bias.

Faizal Ali said...

I am quite confident that Larry's errors were inadvertent, and will be corrected at the earliest opportunity, with an apology included.

I'm not so confident about the creationists correcting their misrepresentation of your ideas.

Corneel said...

@Simon Gunkel
Mmm, OK, I can see that. One last question though: when I teach Darwinian selection, I use infinitely large populations as an example, because that removes the effect of drift. Is N = infinite simply a special case of population resampling, just like the s = 0 scenario, or did we isolate the effect of differential fitness (selection) here?

Larry Moran said...

Dear Gert Korthof,

I do sincerely apologize for not spelling your name correctly. I have corrected it in my post. You are correct when you note that there are many people who are much better at spelling than I am. It was not intentional. I try to catch all spelling mistakes but it's not as easy for me as it seems to be for other people.

Rolf Aalberg said...

Quest, will you please reserve discussion about the origins of life to places where that is the subject and show that you are capable of understanding that evolution is just a subset of the grand set of LIFE? You have been told that several times before, what is that you do not understand?

Unknown said...

Well, infinitely large populations aren't really an example. They are an approximation. For an infinite population the drift term goes to 0 (in fact drift is the difference between what happens in a real population, compared to the approximation). So yes, the infinite population model is a case where you can isolate the effect of differential fitness into the selection term, by virtue of the drift term going to 0.

One of the problems is that you have these two special cases and they might actually hinder you in understanding what's going on in between, because you go "OK, I`ve got the infinite population model, so I know what selection does and I've got the neutral model, so I know what drift does. And now I've got an intuitive idea of what happens when both are in play - it's simply a bit of this and a bit of that."

I referenced weak selection in one of the earlier posts because that's the range where neither neutral models nor infinite population models are good approximations. And it's that range where thinking in terms of drift and selection introduces errors that go beyond semantics.

That being said, I think both infinite population models and neutral ones are indispensable for teaching - they just need to be treated as approximations and in the best case a warning on avoiding the error of trying to get an idea of weak selection by mixing them is added.

Corneel said...

Agreed. Thank you.

William Spearshake said...

@Quest: "...will you admit to the obvious and logical inference...?"

Absolutely. But what I won't do is take the intellectually lazy way out and defer to a god for anything that I don't yet understand. Throughout the last few hundred years, one by one, religious explanations of the natural world have fallen to science and reasoning. The logical inference is that the origin of life will as well.

Claudiu Bandea said...

@NickM,

I agree that there are many layers of confusion here, it’s almost like the confusion is maintained in order keep the discussion alive for ever…which is ideal for a blog.

Uncivilized Elk said...

He doesn't response because you never say ANYTHING. You're inanity personified.

NickM said...

Just posted this at UD, in response to vjtorley's latest silly post: The much-disputed neutral theory of evolution and the book that Professor Moran refuses to review: Larry Moran responds to my questions

http://www.uncommondescent.com/intelligent-design/the-much-disputed-neutral-theory-of-evolution-and-the-book-that-professor-moran-refuses-to-review-larry-moran-responds-to-my-questions/#comment-499784

Anonymous said...

Well, NickM the Shameless,

When am l goring to get my answer... even the shameless one...?

Anonymous said...

Mitsubishi Neil:

"....if you say evolution occurs by natural selection, it looks scientific compared with saying God created everything. Now they say natural selection created everything, but they don’t explain how. If it’s science, you have to explain every step. That’s why I was unhappy. Just a replacement of God with natural selection doesn’t change very much. You have to explain how.…

Mutation means a change in DNA through, for example, substitution or insertion [of nucleotides]. First you have to have change, and then natural selection may operate or may not operate. I say mutation is the most important, driving force of evolution. Natural selection occurs sometimes, of course, because some types of variations are better than others, but mutation created the different types. Natural selection is secondary…

Kimura believed morphology [appearance] evolves through natural selection. He applied neutral theory only on a molecular level. I say it can determine morphological characteristics as well because DNA determines everything, but to prove this has not been so easy. [Laughs.] Forty or 50 years later, I am still trying to prove it…

… Darwin is a god in evolution, so you can’t criticize Darwin. If you do, you’re branded as arrogant.

I particularly like his statement: "say it can determine morphological characteristics as well because DNA determines everything, but to prove this has not been so easy. [Laughs.] Forty or 50 years later, I am still trying to prove it…

Sorry professor Mitsubishi but our guys have proof; it called "I have to be right"... Joe Frankenstein, Larry, Johnny Harsh, Dan Granular and Jerry Coin as weld as Nick The Shameless have proof... is called "My theory is first.... right or wrong".... Congratulation!!!

The whole truth said...

At the risk of stirring up a hornet's nest (I like to take some risks) I have a question.

First, some foundation: If you're not familiar with Great Blue Herons (Ardea herodias), Great Egrets (Ardea alba), and American Bitterns (Botaurus lentiginosus), please look them up. Appearance wise, they are very similar birds except for their coloring, and the bitterns are smaller. The all white feathers of a Great Egret would seem to be a disadvantage since they do not live in a white environment. Great Blue Herons are pretty easy to see too but not as easy as Great Egrets. Bitterns can be very hard to see. Bitterns are the only one of the three that really blends into its environment, although sometimes they are right out in the open or are standing in or next to green grass that doesn't hide them well at all. All three species are successful, in the sense that they are widespread and common. All three species can and often do live in the same environment, hunt for food in the same way, and eat the same things. They can also be prey for the same predators, such as eagles and coyotes.

My question: How does evolutionary theory (mutation, natural selection, adaptation, drift, etc.) apply to and/or explain the big differences in the coloring of those three species? And I'm especially interested in what you all think about the white feathers of Great Egrets (and other white egrets).

John Harshman said...

Evolutionary theory has several possible explanations, and it would take actual study of empirical data to narrow that down, which I expect nobody has done for these species. One can only speculate.

Bitterns can be common, but you very seldom see them. And that's because they're usually hanging out in places where their camouflage works. Egrets and great blue herons, on the other hand, usually feed out in the open, in places where they would be visible even given camouflage. As for color differences among uncamouflaged species, I'll just mumble "sexual selection".

How was that?

Larry Moran said...

Your question contains a hidden assumption that we refer to as the "adaptationist" perspective. You assume that there must be an evolutionary "explanation" of the differences. The null hypothesis is that the differences are unimportant to the survival of individuals within each species. They are due to neutral alleles that were accidently fixed in each population by random genetic drift.

Thus, the correct question is, "Is there any evidence that the differences between these species are due to adaptive change." It may not seem like a different question than the one you asked but it is.

John Harshman said...

You should note that the original question included drift as a possible mechanism. Your question is just a bit more operational, but at the cost of narrowing the focus. One could add, for example, "...and if so, what sort of adaptation?"

AllanMiller said...

My own opinion is that the coloration of the bittern is likely to be adaptive, the other two probably not. Obviously, I'm just an ornery adaptationist and I cannot prove this hypothesis without putting the birds in a 'realistic' predatory scenario with, say, brightly painted reeds or brightly painted birds to break up the pattern. But bitterns stick close to their reedy habitat - they may stand out against other backgrounds, but it's the net losses from time spent against all backgrounds that counts. It's also the smallest of the 3, so predation may be more an issue. The egret, it possibly doesn't matter due to fishing habit and better defence, so an albino gene could readily have spread through drift. It seems unlikely that white is an advantageous colour for them (although a factor may be how it looks to a fish). Blue herons, they fish in the open, and again there is no consistent background selecting for camouflage.

Interesting opening shot here allows comparison of bittern and egret 'camouflage'. https://www.youtube.com/watch?v=LbBeUhRV6jU

The whole truth said...

Thank you for your responses.

John, regarding uncamouflaged species (or the uncamouflaged gender of some species) sexual selection seems like a reasonable consideration to me. What's interesting though, and I'm sure you're aware, is that in many species the males and females look alike, or so close that it's very hard to tell them apart (hard for us humans). And then there are birds like phalaropes where the females are the 'prettier' ones and the duller (more camouflaged) males do the incubating of the eggs and the care of the young. And I wonder, do you guys/gals think that some animals may sexually select their mate(s) because (or partly because) that mate has 'better' camouflage than the other individuals it encounters?

Could it be that instead of an animal adapting to its environment/surroundings so that it blends in (is what we call camouflaged), that sexual selection of camouflaged individuals is the cause/mechanism (or whichever other term you prefer) by which camouflage becomes 'the norm' for some species or the camouflaged gender of some species?

I'm not sure that I'm making my point clearly so let me try this: A lot is said about some animals selecting mates because the chosen mate is flashy, colorful, beautiful, showy, or some other term to describe how much its looks stand out (picture peacocks). I don't recall ever seeing or hearing someone say that some animals select or may select their mate(s) because the chosen mate is well camouflaged (more so than other potential mates). To us, the difference could be unnoticeable but to the animal looking for a mate it may be important.

Here's some food for thought: Which came first, the camouflage adaptation to the environment/surroundings, or the selection of mates for some other reason but that just happen to blend into the environment/surroundings, or the selection of mates that the selector is attracted to because it is 'better' camouflaged, or what? Yes, I realize that 'Which came first?' questions may be dumb or impossible to answer and that my question may be confusing (I'm very tired right now), but I figure that you guys and gals are smart and that you can get the general gist of what I'm wondering about, and that you may have some answers/opinions that are interesting.

I have many years of experience observing nature (including a lot of herons, egrets, and bitterns) but I'm nowhere near as knowledgeable as most of you are when it comes to bio-chemistry, molecular biology, genetics, and the like, and I often wonder how each of you would connect the dots between what you study and some of the things that I observe in nature. I know that there are lots of papers, books, articles, TV shows, etc., that talk about this stuff (and I have read/watched many) but I think that there are some very sharp people here and I like to see what you have to say too.

See part two.

The whole truth said...

Part two.

Larry, I didn't mean to make it sound as though I assume that coloring or other morphological features always have to be adaptations, but I think that it's reasonable to consider whether they may be. In the case of white egrets it would seem that the idea of their coloring as an adaptation could be quickly discarded. Polar Bears would be a different story. Regarding an "explanation" I think that it's okay to say that evolutionary theory explains the differences in the way that you said: "the differences are unimportant to the survival of individuals within each species" or at least to the survival of the species overall, and that the differences in coloring "due to neutral alleles that were accidently fixed in each population by random genetic drift" is or could be a more precise, credible explanation. I only added "could be" because there is some disagreement about how or which evolutionary processes/events affect morphological features and because I don't know enough to be sure that neutral alleles that were accidentally fixed in each population by random genetic drift is the correct explanation. I'm not saying that you're wrong, but I don't know if you're right.

Regarding explanations by evolutionary theory, let me say a bit about how I perceive evolution. To me, there is room in evolution for some slop. In other words, there can be and often is variation between species or the individuals of species that doesn't affect their survival/viability and that doesn't seem to have an 'explanation', at least from an adaptationist perspective. Many species or individuals within a species have morphological features and variations of those features that don't seem to benefit them and also don't seem to harm/hinder them, and the explanation appears to be what you said: "the differences are unimportant to the survival of individuals within each species" or to the survival of the species overall. In some cases the explanation doesn't have to be complicated, and in some cases adaptation doesn't appear to be part of the explanation. Human eye colors are a feature that apparently doesn't matter to the survival/viability of our species or its individuals. Whether eyes are blue, brown, green, or whatever, individual humans or our species overall appears to not be harmed/hindered by different eye colors and various eye colors don't seem to benefit humans either. Some people may prefer a certain eye color in their mate, or may specifically select a mate due to eye color alone but I would think that that is extremely rare or never happens. Maybe 'the designer' just thought that we humans would find each other more interesting if we have a variety of eye colors. LOL.

Sheesh, I had to divide what I've typed into three parts because of the character limit :) so see part three.

The whole truth said...

Part three.

Allan, adaptive does seem to be a good term to describe a bittern's color pattern. It's color pattern certainly helps it blend in much of the time. I wonder if the coloration of a GB Heron could be called partly adaptive since their frontal (throat/breast) feathers have a vertically 'streaky' appearance, kinda-sorta of like a bittern. I have a feeling though that it wouldn't matter if a GB Heron has a frontal streaky appearance or is bright pink all over, at least when it comes to catching prey and escaping predators. If only I could find a bright pink one and find out. :)

I'm sure you noticed the wiggling that the bittern did in that video. Herons and egrets do that too but usually not as much as bitterns do. Sometimes they (herons. egrets, bitterns) wiggle when the wind blows the nearby grass around as though they're trying to match the movement of the grass, sometimes they wiggle just before striking at their prey, and sometimes they wiggle for no apparent reason. I could only watch a couple of minutes of the video because of my slow internet connection and I don't know if the video showed a male bittern trying to attract a female so if it didn't and if you and others want to see and hear something fascinating you should look for a video of a 'pumping' bittern. Bitterns are one of my favorite birds and I often watch them for hours at a time.

nmanning said...

Quest must know a bunch of creationists that I have never encountered, for in my 10+ years of experience in reading the writings of and dealing with creationists, nearly all (to include the 'biggies', such as Wells and Sarfati) are anything but diligent in their research, and all of them are very condescending and insulting to those not of their cult.

nmanning said...

Mitsubishi Neil...Joe Frankenstein, Larry, Johnny Harsh, Dan Granular and Jerry "Coin"...

"I've got to give credit to creationists... They are very diligent in their research and they never offend a soul ..."

Creationists are not known for being consistent, clever, or honest. As Quest demonstrates nicely for us.