My Chicago colleague James Shapiro appears to have been badly stung by my repeated criticisms of his attempts to forge a new evolutionary view based on the “self-engineering” of organisms and their DNA (see here, here, here, and here, for instance).
This will be the last time I comment on Shapiro’s PuffHo pieces, as I don’t want to give him the attention that, as someone who wants to be seen as the founder of a new evolutionary pardigm, he so desperately wants.
Let me just point out Shapiro’s Big Error (also committed by those who approve of his PuffHo piffle) one more time. It is not rocket science, and even non-scientists should be able to grasp it. I’ve put it in bold so Shapiro and his followers will take note:
Regardless of the source of genetic variation, if new variants are to become “fixed” (i.e. ubiquitous) in natural populations after they arise, and to become part of complex adaptations, there is no credible alternative to natural selection for the process causing that fixation.
Like some molecular biologists, Shapiro repeatedly mistakes the source of genetic variation creating new “mutations” with how that variation comes to predominate in a population. He thinks that the former explains the latter, ergo that natural selection plays a very small role in evolution. And yes, we now know of a world of “mutations” undreamt of by earlier geneticists, including horizontal gene transfer (especially important in bacteria), gene duplication, mobile genetic elements, and the like. But what happens to variation generated by these processes depends on whether they give a reproductive advantage to the organism. If they don’t (i.e., natural selection doesn’t favor the variant), they get eliminated from populations. If they do, they get fixed.
It’s still natural selection, Jim. Why do you fail to understand that?
(Let me add right away that I accept other mechanisms of evolution such as genetic drift, which is certainly of importance in much molecular evolution. What I’m talking about here are complex adaptations, not only of morphology [e.g., the heart], but also complex molecular mechanisms like DNA repair or the immune system.)
Sadly, in his latest PuffHo piece (really, does the science section even look at what’s appearing on their site?),”Inconvenient truths: why are some self-styled defenders of evolution so resistant to lessons from molecular genetics?“, Shapiro is reduced to criticizing two readers who have left comments on my latest critique of Shapiro’s Big Idea. You’ll recognize their names if you’re a regular: Ben Goren and Torbjörn Larsson.
And Shapiro’s criticisms of both men’s comments are of the same ilk: the modern theory of evolution (“neo-Darwinism,” though I am going to stop using that label) is wrong because it doesn’t deal with “mutations” that aren’t simple changes in single nucleotides in the DNA. For a rebuttal of Shapiro’s thesis, just reread what’s in bold above. Horizontally transferred elements, transposable genetic elements, and the like, must be acted on by natural selection if they’re to become part of an adaptation.
I’m not a willy-nilly defender of my commenters, but in this case Goren and Larsson are right and Shapiro is wrong. Take, for instance, what Shapiro says about Torbjörn:
A second commentator, Torbjörn Larsson, argued that horizontal DNA transfer posed no challenge to the neo-Darwinian theory: “In other words the generic gradualism of Darwin mentioned in the article isn’t rejected by the observed degree of horizontal gene transfer.”
But Torbjörn’s claim does not make sense scientifically. Horizontal transfer is not the gradual Darwinian accumulation of changes. Horizontal transfer episodes rapidly incorporate complex evolved DNA structures into new genomes by coordinated molecular events.
There is no way we can reasonably apply the term “random mutation” to a DNA transfer process that utilizes dedicated surface structures for bringing two cells together, assembles a multi-protein DNA transport pore connecting the cells, and initiates DNA transfer replication at a specific site on plasmid DNA. [JAC: yes we can; for the transfer is a genomic change that occurs randomly!]
The well-established molecular details of horizontal transfer in the evolution of bacterial antibiotic resistance are difficult to reconcile with neo-Darwinism.
No they aren’t—not at all.
What Shapiro can’t see here is that it doesn’t matter how new changes in a genome come about: via classical base-pair “mutations” or by the more recently discovered processes of horizontal gene transfer (e.g. the movement of units of DNA from one species of bacteria to another via a virus vector). If those chunks cause antibiotic resistance in the recipient species, and there are antibiotics around, natural selection will go to work on the new “variant.” Indeed, this is how a lot of antibiotic resistance evolves in bacteria. But it doesn’t matter how a variant arises: they are all equivalent to classical mutations and their dynamics can be described by the classical equations of population genetics. After all, antibiotic resistance in bacteria is seen as a classic and easily-understood example of natural selection!
So no, there isn’t an insuperable problem here for “neo-Darwinsim.” We just have to expand our idea of “mutation.”
Shapiro can’t—or won’t—get this, and I give up on him. Time will tell if his “non-Darwinian” ideas gain any traction, but the fact that he propounds them at PuffHo rather than in peer-reviewed scientific journals (how about sending a piece to Evolution, Jim?) is not propitious. As one of my friends, who is both an evolutionist and molecular geneticist, emailed me after reading Shapiro’s piece:
Part of his problem has always been the one that some molecular biologists have: they confuse molecular causation with evolutionary causation. All of the mechanisms mediating horizontal transfer and so forth that he cites are indifferent to their effects on fitness: adaptation still arises as a consequence of the sorting agency of natural selection.
I sometimes wonder if there is a polymorphism in the human population for ability/inability to see the distinction. Jim would then be an example of someone who got a really bad version of the mutant gene for inability.
I’ll let Ben and Torbjörn defend themselves, but for me Shapiro sleeps with the fishes. He will go on crying in the wilderness that is the PuffHo science section until he leaves this earthly vale and his ideas fade into well-deserved obscurity. I’ve made my point, and I’m done with him. I may seem “strident and militant” about this, but one reaches a point when fulminating ignorance, pointed out but not corrected—much less admitted—must be shown for what it is.
Golly, even I of all people, a non-scientist if there ever was one, can see the point. It still takes natural selection to work on the mutation, however acquired, if it is to become dominant (or not) in a population. Shapiro seems to have forgotten that whatever results from horizontal gene transfer has to survive in order to pass on the mutation to the next generation. If not, it does not get passed on, and cannot become widespread in the gene pool.
Same here. In fact I don’t really comprehend why anyone has a hard time accepting natural selection at all. I certainly don’t understand all the details but the basic framework can’t be denied.
Mike.
Perhaps Shapiro has primarily failed to absorb what Ernst Mayr used to call “poulation-thinking” in understanding evolution. And so he discounts the mechanisms (selection, drift) by which genetic variants (whatever their source) rise and fall in frequency in populations. One might begin to wonder about his processing of quantitative phenomena.
I don’t understand – is he really confusing the processes which produce the raw material for evolution with the process which filters that material.
Surely, the distinction is trivial.
If a non-scientist such as myself can understand this, how can he not?
Trivial?
The distinction is not trivial.
I think I understand what you were trying to say; I think you were trying to say something like, “being able to distinguish the two seems like it should be a trivial matter.” (meaning the difference between the two should be very obvious to anyone who’s even vaguely interested in the subject.)
But the difference between what causes a mutation and what causes the mutation to become fixed in a population is in no way trivial. They’re fundamentally different things.
Perhaps “distinguishing between the two should be a trivial matter”. But yes, that is what I meant. One process creates variety, the other filters that variety dependent upon “fitness”.
Congratulations Ben and Torbjörn !!
+1
+ 2
And I’m going to repeat what Dan Dennett has said so well: In the “design space” of living things, the further you “leap” in that design space, the more likely you are (the odds explode very quickly) to cause a detrimental change. Only very small changes can approach a 50:50 chance of causing a beneficial change.*
This, of course, remains true regardless of the mechanism of that change!
This seems to be the point Dr. Shapiro is missing.
Or, even worse, is he positing a directed exchange of genes (for example) or a directed mutation of base pairs, etc.?
(* BTW, this is also true of human-engineered products. We are broadly guilty of “counting the hits and ignoring the misses.” We celebrate the miniscule fraction of inspired inventions and ignore the mountains of failures.)
I agree with your point on what he’s missing, Dr. C.; but further than that, I think he is in fact positing big changes that overturn gradualism.
This can never be (short of intelligent direction, which we all firmly reject, presumably even Dr. Shapiro.) The laws of statistics and large numbers (how many dimensions does this “design space” have? I’m not going to even try to count the ways organisms can vary …)
… the laws of stats … rule that out, full stop.
How important is gradualism though? Does it matter if change happens at a constant pace or in fits and starts?
As I understand it, the core of Darwinian evolution is natural selection, not gradualism (even if Darwin posited that change was gradual and constant) and nothing Dr Shapiro has said alters that.
Well … I’m just reading Dr. C.’s book for a third time (this time with my young son) and he does present gradualism as one of the key concepts of the modern view of EBNS.
The concept of the “fortunate monster” (saltation) has been pretty well shown to be extremely unlikely and extremely rare in fact.
You are correct that nothing in natural selection that rules out a saltation. But consideration of the statistics and evidence show them to be extremely difficult, making them realtively unimportant in EBNS.
All that said, were: Eukaryotic cells, sex, segmentation, cell walls, multicellular life, or similar steps in evolution, which from our limited viewpoint, seem to be really big or revolutionary, caused by some sort of (undirected, random) genetic saltation? Maybe. It could happen. Like the origin of life, which we generally agree was very unlikely, it only had to happen once. And there has been a lot of time (trials).
“Gradual” literally means ‘stepwise’, from Latin gradus, a step or pace.
So every time you hear or use the word, think of a staircase, not a fucking ramp.
Then, a lot depends on timescale. At the basic level, mutation and selection happen in finite increments (usually small, often reversible), not a continuous glide.
At longer timescales, Darwin discussed the possibility of extreme variability in rate of change, a fact that S.J. Gould mostly failed to mention in the context of punc-eq. Dawkins (in TBW) has sweetly parodied Gould’s misbegotten critique of gradualism as constant-speedism.
Indeed – it’s been a while since I read Origin, but I doubt that Darwin really though that change happened at a constant pace – I’d have thought his own experience of speciation on the Galapagos would’ve been evidence enough of rapid change.
Crocodiles and whales. Extemely different rates, no saltations.
Good point. Each step is literally a step function. It’s different before and after, but some distinct increment. It’s summation (Σ) not integration (∫). [Let’s see if my html coding works …]
Those increments are, almost always, very very small.
A universal acid.
That’s a very interesting thought. I bet it has practical considerations to my own field – software development.
Thanks,
Mike.
Of course, to re-phrase one of Dr. Coyne’s points here, “how unlikely an improvement is” also doesn’t matter to “how fast it will come to predominate once arisen”.
Yes.
If this is Dennett, he’s likely got it from Dawkins, who makes this point forcefully, e.g. in chapter 9 of The Blind Watchmaker, attributing it to R.A. Fisher.
Here’s the Fisher quote, which is lovely:
Bravo! Simple language that even a molecular biologist might understand.
(laughing)
neo-pwnd
Well come on Jerry: don´t sugar coat it. Tell us what you REALLY think about Shapiro!
Coyne rejects neutral theory. Larry Moran ken haz a sad.
Jerry Coyne sez: “if new variants are to become “fixed” (i.e. ubiquitous) in natural populations after they arise, and to become part of complex adaptations, there is no credible alternative to natural selection for the process causing that fixation.”
.
Sorry, I missed the cheater phrase the first time around. Neutral traits can be fixed, but are not referred to as adaptations.
Yes, OF COURSE I accept genetic drift; in fact it MUST operate in finite populations and probably accounts for a lot of aspects of molecular evolution,as Mike Lynch and Larry Moran have noted eloquently. But I am talking about adaptations here.
I’ll clarify above.
Michael Lynch (and I) also talk about so-called “adaptations.” The problem with your statements is that there is legitimate disagreement about your premises. You seem to assume that all complex things in a cells are adaptations. You also assume that all adaptations are fully explained by natural selection.
As you know, Lynch (and I, and others) disagree with both premises.
Lynch makes a convincing case that our complex genome is largely a result of nonadapative processes. See “The Passive Emergence of Genome Complexity by Nonadaptive Processes” in his book “The Origins of Genome Architecture.”
He then goes on to ask “Do the Roots of Organismal Complexity Also Reside in Nonadaptive Processes” and he makes a good case that they do.
Lynch concludes, “Unfortunately, the emerging field of evolutionary developmental biology is based almost entirely on a paradigm of natural selection and the near-absence of the concept of nonadaptive processes from the lexicon of those concerned with cellular and developmental evolution does not follow from any formal demonstration of the negligible contribution of such mechanisms, but simply reflects the failure to consider them. There is no fundamental reason why cellular and developmental features should be uniquely immune to nonadaptive evolutionary forces.”
In other words, complex things may not be exclusively “adaptations” and, furthermore, nonadaptive forces (i.e. random genetic drift) make a substantive contribution to adaptation.
As for mutationism—the idea that mutation itself plays an important role in the direction of evolution—your readers might want to read this excellent series of articles by Arlin Stoltfus: The Mutationism Myth.
Please note that I am NOT defending Shapiro.
Erm, if I know Jerry’s position on Evolution at all, I’m pretty sure that he would emphatically disagree that everything complex in a cell is adaptive. Rather, everything in the cell would be subject to the same selective pressures as everything else.
For example, something maladaptive might survive if it’s inextricably linked with something that provides a greater advantage. On a large scale, the size of a human infant’s skull fits the pattern: it’s very maladaptive, but the advantage to having that much more brainpower soon after birth outweighs the harm it causes.
I’m not a cellular biologist so I wouldn’t presume to offer a similar example at that scale, but I’m sure somebody with Shapiro’s knowledge would be able to come up with one were he not blinkered by his active misunderstanding of Evolution.
Since the rest of your post hangs on that mistraken assumption, I don’t think there’s anything else left to discuss.
Cheers,
b&
I am always a bit puzzled by your stance on this issue (LM). Of course drift should be the null hypothesis/null model before one concludes selection, and of course one should avoid to tell just-so stories.
But, sorry to say so, I am literally unable to imagine how the world would have to look like for any trait or coding region not to be under selection: Surely it is clear that many mutations are deleterious and strongly selected against, and thus all biological structures that are of importance (like complexities of the cell, say) are under stabilizing selection at all times?
This seems like something that we would not even have to demonstrate, but rather it follows as logically from some thought as natural selection itself follows from variation + exponential population growth + limited resources.
No one is arguing against purifying selection. Otherwise genomes would degenerate to random nucleotide strings.
The debate is more about “characteristic X exists, is it a selective adaptation, or the result of drift/ non-adaptive complexity?”
There is more debate on this in the molecular side of things. One well argued example is RNA editing, using gRNAs to “correct” a coding frame of a protein mRNA. The appearance of this mechanism lead to a “protects against mutations” adaptive theory, while others point out that the deleterious mutation must have existed first, by necessity, for the editing mechanism to prove beneficial (the wiki entry on RNA editing points to some references under the Evolution Of… part). Now V-ATPases appear to have complex ring structures due to duplication-mutational drift scenerios (doi:10.1038/nature10724).
I fear that a lot of that is semantics. Some people say “survival of the fittest”, which sounds very adaptationist, while it seems more appropriate albeit slightly tautological to say “survival of everything that is not so defect that it dies”.
So when may I call it an adaptation? In a way, everything is adapted to surviving in its current form in its current place even if there could well have been a different form produced by random drift that would have served just as well. The point is, we could easily imagine a thousand forms that are too bad to survive.
As an example, I remember some time ago Larry Moran complained about an adaptationist explanation for the red (?) belly of some weird bird and asked why it could not simply be random. I am not sure what the idea is supposed to be – obviously the null hypothesis is that there would be no special colour there but simply the same as in the rest of the body, no? Now those birds have a red belly and while I have not gone and tested it it is no stretch to presume that the other sex considers this colour to be part of what an attractive, healthy member of their species should look like. Sure the red was obviously the result of random mutation (everything is) and it could just as well have ended up yellow. The point is, it may now just not be a good idea for one of those birds to be born without the red colour. Is that now adaptation or not?
Alex SL:
Another way of making your point is just to say that at the molecular level there is still selection for genes that work together as a complex. So genetic drift can be responsible for all kinds of phenotypic characteristics, but obviously not for unviable genetic combinations. When drift happens, genes still have to adapt to each other!
Wow. I’m flattered!
For the record, here’s the post where I feel reasonably confident I got the right answer in comparing HGT to either a big chunk of mutation all at once or sexual recombination:
http://whyevolutionistrue.wordpress.com/2012/11/28/nicholas-wades-ridiculous-prescription-for-curing-creationism/#comment-328835
My next response is the one where I compared James to Behe…but where I also hypothesized an evolutionary origin for HGT all the way back at the dawn of life on Earth. I find it quite telling that James never did address my explanation and instead latched on the Behe comparison.
It’s also in that exchange where Torbjörn jumped in, echoed my Behe comparison, and added a great deal of detail to the possible origins of HGT based on his RNA world hypothesis.
I’m sure James would mop the floor with me on the details of molecular biology, but I find it quite saddening that I actually understand the foundational underpinnings of biology better than he.
b&
HGT and RNA World will show up on my MolEv final exam this afternoon. *Off to read said links…!!*
(Apologies for the substance-free comment.)
So what complex adaptations exist that were not shaped by natural selection?
We’re talking adaptations right? Let’s have a look at them.
…the bacterial flagellum?
Ouch. Unless Shapiro wants to be tarred with the Behe brush, one thinks he might want to rethink his position.
In Coyne’s blog post, what he *says* is “become part of complex adaptations”.
In my view it is perfectly normal for neutral variants to become fixed and then subsequently become incorporated into complex adaptive structures.
The idea itself is — or should be — uncontroversial. This is one of the ways “irreducible complexity” arises — though I much prefer the term “interlocking complexity”, which is precisely the same concept but described AND PREDICTED by Herman Muller back in 1918. (Yes, Behe is that out of date.)
See: “Mullerian Two Step” (part of “evidences for evolution” by Doug Theobald at talkorigins)
http://www.talkorigins.org/faqs/comdesc/ICsilly.html
(I also wrote an earlier article on this subject and got a POTM at talk.origins for it in Sept 2006.)
The Mullerian two step is when complex adaptation arises in these two stages.
(1) Fixation of a part (possibly neutral fixation, no adaptive impact)
(2) Incorporation of that part into the complex structure. (By introduction of other parts, that are adaptive, but which depend for their operation on the first part).
Unfortunately, the examples in the FAQ are for cases where the fixed part is adaptive at the time of fixation, and only subsequently becoming a critically important part as other changes occur that are dependent on it.
But IMO there’s nothing exceptional about a completely neutral part arising, becoming fixed, and only subsequently becoming at all adaptive, in the context of other changes in the whole structure.
Hence I blanch somewhat at Jerry’s phrasing:
>>>
Regardless of the source of genetic variation, if new variants are to become “fixed” (i.e. ubiquitous) in natural populations after they arise, and to become part of complex adaptations, there is no credible alternative to natural selection for the process causing that fixation.
<<<
Here's a credible alternative.
(1) Duplication of some gene (neutral, fixed by drift) (part "A")
(2) Variation in one of the duplicates (neutral, fixed by drift) (part "A" fixed)
(3) Adaptive mutation (fixed by selection) which makes use of the variant earlier fixed by drift. (part B, adaptive but only in the context of part A)
The difference is subtle but significant. The Part A has become adaptive. But it was not fixed by selection. It was fixed by drift, and selection only applied later to a DIFFERENT part, which took advantage of the existence of the fixed part A for its function.
This kind of process is likely involved in the origins of complex cascades like the famous blood clotting cascade.
oops.
you just unwittingly supported Jerry’s argument.
by adding a source of variation that selection can then act on in your part (3), all you’ve done is EXACTLY what Shapiro did.
I think you are mis-interpreting this. Please look at this paper
et al. & JW Thorton. Nature (2012) doi:10.1038/nature10724
http://pages.uoregon.edu/joet/PDF/finnigan-thornton_nature2012.pdf
I did. all they are arguing is for a PIECE of the puzzle to have become fixed in a specific fashion.
selection still operates on the result.
this is NOT a case where selection did not cause fixation of an ADAPTIVE trait.
…on the result when combined with another trait, I mean.
obviously selection is no longer acting on a fixed trait.
point is, you have to be careful to delineate neutral traits (arising for any reason) from ones that any given selective pressure can act on, or even old traits that are fixed that have new sources of variation added, as is the case with your example.
This is becoming a semantics argument, so I’m stopping here for the sake of the comments field’s length.
You view the second mutation as an adaptation to the first. I view the second mutation as a compensation to return fitness to the point prior to the first mutation. In the end the change in this enzyme shows no evidence of increasing fitness for the organism, so I consider this non-adaptive evolution (see doi:10.1038/nrg3015).
I’ll have to track you down at a conference some time to continue this (with beers, or course ;-))
again, stressing this part of your idea:
The Part A has become adaptive.
no, that is NOT what happened. otherwise, it would not have been defined as neutral when it became fixed to begin with.
“A” BECAME adaptive when combined with the second mutation, which added variability that selection could then act on.
If my reading of Sean B. Carroll’s work is correct, a trait can not be fixed in the absence of selection. Being “fixed” means that some set of genes appears in all individuals in the population. And since all genes are equally subject to the decay of mutation over time, to remain fixed selection must remove these mutations when they occur.
Assuming that it takes some time to “fix” a gene in a population, I don’t see how a gene could become ubiquitous without selection operating on it. I suppose it could happen in a trivially small population but out in the real world of gene pools I am doubtful.
Neutral mutations fix at about the same rate as mutation.
http://en.m.wikipedia.org/wiki/Fixation_(population_genetics)#_
See also ‘Mitochondrial Eve.’
I don’t think that can be correct. The rate of mutation would be driven by (for example) exposure to radiation that causes random bases to change. Whatever this rate, it would have zero relationship to how quickly a gene could spread in a population in the absence of selection. And in the absence of selection the gene under consideration would be subject to mutation which would work in the anti-fixing direction.
And… what about MitoEve are you suggesting is relevant to the question?
In Coyne’s blog post, what he *says* is “become part of complex adaptations”.
In my view it is perfectly normal for neutral variants to become fixed and then subsequently become incorporated into complex adaptive structures.
The idea itself is — or should be — uncontroversial. This is one of the ways “irreducible complexity” arises — though I much prefer the term “interlocking complexity”, which is precisely the same concept but described AND PREDICTED by Herman Muller back in 1918. (Yes, Behe is that out of date.)
See: “Mullerian Two Step” (part of “evidences for evolution” by Doug Theobald at talkorigins)
http://www.talkorigins.org/faqs/comdesc/ICsilly.html
(I also wrote an earlier article on this subject and got a POTM at talk.origins for it in Sept 2006.)
The Mullerian two step is when complex adaptation arises in these two stages.
(1) Fixation of a part (possibly neutral fixation, no adaptive impact)
(2) Incorporation of that part into the complex structure. (By introduction of other parts, that are adaptive, but which depend for their operation on the first part).
Unfortunately, the examples in the FAQ are for cases where the fixed part is adaptive at the time of fixation, and only subsequently becoming a critically important part as other changes occur that are dependent on it.
But IMO there’s nothing exceptional about a completely neutral part arising, becoming fixed, and only subsequently becoming at all adaptive, in the context of other changes in the whole structure.
Hence I blanch somewhat at Jerry’s phrasing:
>>>
Regardless of the source of genetic variation, if new variants are to become “fixed” (i.e. ubiquitous) in natural populations after they arise, and to become part of complex adaptations, there is no credible alternative to natural selection for the process causing that fixation.
<<<
Here's a credible alternative.
(1) Duplication of some gene (neutral, fixed by drift) (part "A")
(2) Variation in one of the duplicates (neutral, fixed by drift) (part "A" fixed)
(3) Adaptive mutation (fixed by selection) which makes use of the variant earlier fixed by drift. (part B, adaptive but only in the context of part A)
The difference is subtle but significant. The Part A has become adaptive. But it was not fixed by selection. It was fixed by drift, and selection only applied later when a DIFFERENT part came up which took advantage of the existence of the fixed part A for its function.
This kind of process is likely involved in the origins of complex cascades like the famous blood clotting cascade.
I admit I haven’t been following this very closely (though now I do want to go back and hit all the links mentioned above). But it seems quite straightforward and simple. This semester I have been taking a molecular evolution class (and lucky to have some very bright faculty in the department), and for every genetic change we talked about — whether single nucleotide substitutions, indels, gene conversion, HGT, etc. — its probability of persistence in the population always depended on its effect(s) on fitness. I feel like I should be adding an “obviously” at the end of the previous sentence, but apparently that concept isn’t as obvious as I’d thought. I like the email you excerpted above, suggesting presence of a polymorphism for ability/inability to make the distinction, because it really does seem like a fairly black and white issue — either you get it and it clicks, or you just can’t seem to wrap your head around it, and to us normal folks trying to explain it, it feels like we’re just hitting a brick wall.
Perhaps good practice for my General Exam…
It may be that people are confusing the probability for a variant to arise, and the probabilistic character of benefit that causes the trait to propagate once arisen.
I suspect it’s tied to some P(A|B) Bayesian reasoning analog of the material conditional “if” tested by the Wason selection task. Might be interesting to try to find an experimental analog to that task, but I expect people would tend to do depressingly poorly.
I love the part where Shapiro is very impressed by the complex molecular machinery involved in horizontal gene transfer. He should take a look at ordinary sexual reproduction some time. Motile sperm, ovaries, placentas, and in humans even angst and weltschmerz!
Hey, this is a family blog (edit: website)! Keep your posts G rated, please.
If that isn’t G rated the problem isn’t with the comment but with the construction of how the rating is determined.
“Matt G” is probably biased in insisting on G.
No – it’s just that “motile sperm” made me blush.
I blushed at “weltschmerz,” myself. But then, I always do.
Is Shapiro really stooping to playing a game of “gotcha!”? Wow! Here is a guy who is desperate to not understand evolution. Horizontal gene transfer makes things a little messier, but does not undermine the fundamentals of the Modern Synthesis. Biology is complicated – get used to it. Sarah Tishkoff just wrote in a Nature News story that many genetic diseases do not have the same genetic basis – contrary to expectations ( http://www.nature.com/news/past-5-000-years-prolific-for-changes-to-human-genome-1.11912 ).
Mat says: “Horizontal gene transfer makes things a little messier, but does not undermine the fundamentals of the Modern Synthesis.”
I think if you go back a read about how the “Modern Synthesis” was defined in the early 1960’s you would multiple contractions with what we now accept and has been suggested about horizontal gene transfer.
The ideas of natural selection, descent with modification and common ancestry as an explanation for homology are quite compatible. But what is going to be lost is the notion that all life emerged from any single ancestor.
By definition, there’s a LUCA unless you postulate that DNA and the genetic code were invented at least twice.
The chance that the last was also the first common ancestor is negligible, which means that there was a whole extended lineage of universal common ancestors of all extant life. I don’t think that idea’s going to be lost from evolutionary theory, ever.
HGT just makes it harder to draw correct trees.
Yes and no.
If you compare HGT with sexual recombination, it becomes clear that all it really represents is the same sort of messiness we already see in certain stereotypical backwater parts of the world where close cousins and even siblings intermarry.
We already know that the Tree of Life is more of a thickly-matted briar patch. All HGT does is demonstrate that the roots extending outward from the base are as tangled as the branches reaching upwards.
Cheers,
b&
I just tried to leave this as a comment at HuffPo. We’ll see if I jumped through all the right hoops.
Well said, Ben.
Thanks.
Still no sign of my comment, but it says there’re a half-dozen comments awaiting review. We’ll see how much he and / or PuffHo cares about open debate….
b&
+ 1 for that analogy.
😀
Thanks.
The comment finally got posted, and somebody there in response wanted to know if GMO crops fit into the modern understanding of evolutionary theory. I’m so proud that I explained why they do without questioning said person’s own genetic fitness….
b&
>>You’re like a percussionist in a symphony orchestra insisting that there isn’t any melodic structure to the piece….
A very good clarifying simile!
Naturally Jerry is right. But I take exception to his notion that “we need to expand our idea of “mutation””. When the original population genetics was written (1920s-1940s) the molecular basis of mutation was unknown. It was later discovered that many mutations happen to be point mutations, but the theory never presumed this, does not require this and had, by necessity, to be agnostic as to the mechanistic basis of mutation. I’m sure Fisher, Haldane et al would be most interested to hear the various means by which new variants arrive in populations, but would see no reason to rewrite any of their models (which is Jerry’s point). If any expansion is needed it follows only because of a relatively recent and unnecessary contraction that stems from advances not in evolutionary theory but in molecular biology.
Yes. I think this is also a very, very common tactic among the creationist/ID crowd.
The deliberate mischaracterization of what a mutation is is practically definitive proof that the speaker is a creationist.
Oh, I’d say anyone who hangs out in the PuffHo Science Section has already told us exactly who he is. Wouldn’t you?
sub
This remark from Dr. Coyne’s scientist friend illustrates the heart of the matter: Shapiro grossly misuses the term “random”. In evolutionary biology, “random mutation” means “one that happens regardless of its potential utility”. Shapiro tends to think that complex genetic changes that happen through a complex mechanism cannot be “random”.
Possibly not understanding that “random” does not require being “uniformly random”. Bertrand’s Paradox is a nice bit of geometric probability that helped get that notion across to me; maybe including it at some point in K-12 education would be of benefit.
That was exactly my thought, except I would upgrade “possibly” to “probably”.
For most people the word “random” has connotations of equiprobable. Most people would probably not consider a biased die to be “random”. (I mean a die which is biased towards one number, but doesn’t always come up with that number.)
It is as if a description of the detailed mechanism of a slot machine must necessarily render a “jackpot” as a nonrandom event.
Shapiro is using that definition of random. He theorizes that cells can co-ordinate mutations solve a survival problem.
Here is David Deutsch’s take on the matter:
“Neo-Darwinism does not refer, at its fundamental level, to anything biological. It is based on the idea of a replicator (anything that contributes causally to its own copying).”
[…]
“The most general way of stating the central assertion of the neo-Darwinian theory of evolution is that a population of replicators subject to variation (for instance by imperfect copying) will be taken over by those variants that are better than their rivals at causing themselves to be replicated. This is a surprisingly deep truth which is commonly criticized either for being too obvious to be worth stating or for being false.”
[…]
“So what would refute the Darwinian theory of evolution? Evidence which, in the light of the best available explanation, implies that knowledge came into existence in a different way. For instance, if an organism was observed to undergo only (or mainly) favourable mutations, as predicted by Lamarckism or spontaneous generation, then Darwinism’s ‘random variation’ postulate would be refuted. If organisms were observed to be born with new, complex adaptations — for anything — of which there were no precursors in their parents, then the gradual-change prediction would be refuted and so would Darwinism’s mechanism of knowledge creation. If an organism was born with a complex adaptation that has survival value today, yet was not favoured by selection pressure in its ancestry (say, an ability to detect and use internet weather forecasts to decide when to hibernate), then Darwinism would again be refuted. A fundamentally new explanation would be needed.”
David Deutsch is quoted as saying,
“The most general way of stating the central assertion of the neo-Darwinian theory of evolution is that a population of replicators subject to variation (for instance by imperfect copying) will be taken over by those variants that are better than their rivals at causing themselves to be replicated. This is a surprisingly deep truth which is commonly criticized either for being too obvious to be worth stating or for being false.”
Most evolutionary biology textbooks discuss population genetics. They conclude that some populations can be “taken over” by newly arising alleles that are deleterious (not beneficial). They also conclude that, in most cases, beneficial alleles will be eliminated before they can “take over” a population.
Both of these facts seem to contradict what David Deutsch says.
“They conclude that some populations can be “taken over” by newly arising alleles that are deleterious (not beneficial). They also conclude that, in most cases, beneficial alleles will be eliminated before they can “take over” a population.
Both of these facts seem to contradict what David Deutsch says.”
Only the first statement does; the second does not. I’d like to see a citation for the first statement.
I think the key caveats there are, “some”, “can be”, and “in most cases.” These could still be true and conform with Deutsch’s statement. A matter of statistics and time I think. In the end, the beneficial wins out. Sometimes not: Species and clades go extinct.
Hello Mr. Moran! It’s a distinct pleasure to interact with you, and I’ll confess that I’m well out of my comfort zone. But your statement…
“Most evolutionary biology textbooks discuss population genetics. They conclude that some populations can be “taken over” by newly arising alleles that are deleterious (not beneficial).”
I believe David Deutsch discusses one example of this in his book from which I quoted. He talks about a population of nesting birds, that originally nests in April (which is the optimum time of year for various reasons like availability of food, weather, etc.), which is then “taken over” by an allele for nesting one month earlier in the year in March. Even though March isn’t as optimal as April, it does fall one month earlier in the year, therefore the allele birds can take their pick of the best nesting sites without much competition from the dominant allele.
If this new allele takes over, as Deutsch proposes, it is actually a deleterious effect for the bird population, because now the majority of birds are not nesting at the optimal time of year.
I don’t know if this answers your criticism of Deutsch, but I hope I’ve at least explained this example clearly.
In other words, Deutsch is saying it is the fact that variants copy themselves better that is important in Darwinism, regardless of whether the phenotypical effects are good or bad.
I think David recognizes that a variant that is superior at copying itself (and therefore spreading through the population) could definitely drive the species to extinction.
Your description of Deutsch’s explanation seems contradictory. Since there are many variables in nesting patterns then there is a trade-off between location, food, weather, etc. I haven’t read Deutsch, but from what you said, it doesn’t make sense then to say the allele, which seems act on a single variable, is deleterious for the population. That would have to be tested by measuring its effect on the population.
Deutsch continues the argument, and in my opinion reinforces it, by saying that further variants may arise that continue to push back the nesting period, i.e. from March to February, then to January, because first-come-first-serve applies to the prime nesting sites.
But because the nesting period has now been shifted well away from the optimal month of April (into the harsh winter months?), this can only be a deleterious effect for the bird population as a whole, which will result in fewer offspring for the species, and possibly drive it into extinction.
His argument, not mine!
Eventually they will be nesting in April •of the year before•.
LOL
Damn the quantum physicists with their sine waves and cyclical universes.
While HGT may have a large and sudden impact on the fitness of the recipient species/lineage, the origins of any adaptive cassette of encoded information in the donor lineage surely cannot have been by a means other than gradual and random mutations.
This isn’t the case. See http://huff.to/TAWx5G
Shapiro is a case example of why good universities typically include both molecular AND population genetics as required courses for biology oriented majors.
It appears he missed the entire population genetics part.
I thought Shapiro’s view was that mutation rate, or some form of sequence variation production, is THE dominant and rate limiting component of evolution, instead of it being natural selection. Not that he doesn’t understand or appreciate natural selection but that he thinks it plays a less important role than sequence variation mechanisms.
Anyway, I don’t understand that point of view, maybe because I’ve got it wrong.
Shapiro seems to think that organisms can direct their own mutations and that, therefore, natural selection is not needed.
But who knows.
He writings are a mish mash and he never seems to quite say why the modern synthesis is dead and what exactly he is going to replace it with.
I think it’s also important to make clear that by fixing a mutation (or making it very common), you are also increasing the probability of a mutation that can build on the first and create a more complex system.
Orthogenesis?
Dr. Shapiro, as I read his posts, is not denying that Natural Selection is the primary means by which variations spread through a population. He is only claiming that means by which variations arise has to be reevaluated.
I don’t understand why Jerry thinks that this is insignificant to modern evolutionary theory. HGT sounds very Lamarckian to me and if it, or similar faster acting mechanisms for variation, are common, then we need to revisit the gradualism that is taken for granted.
” He is only claiming that means by which variations arise has to be reevaluated. ”
This would have been a good point a few decades ago before it was widely accepted how important horizontal gene transfer is.
True. Shapiro is, however, taking this to whole new degree. He is claiming that cells have a mechanism that, when faced with selection pressure (e.g. starvation), will produce co-ordinated mutations in offspring to survive.
Shapiro, as I understand him, is arguing that the cell literally is attempting to engineer changes to increase chances of survival.
Well that’s not horizontal gene transfer. If it exists, it’s something else. But even if it exists, I don’t see how it conflicts with the modern synthesis.
“Conflicts” is a strong word. Technically they aren’t in conflict. However the Modern Synthesis had claimed that biological diversity could be be explained, for the most part, by random mutation. Shapiro, if I understand him correctly, is claiming that the cell is actually trying to mutate to survive and therefore under strong selection pressure will produce multiple mutations in one organism.
And those mutations are random.
I am not sure if he would say that. I haven’t read his book yet, but it could be that he would say that they are directed. So, for example, if the trouble facing the organism was in food intake, we would see multiple mutations in related areas but next to none in non-related areas.
Even if his view is just that under pressure, some sort of mechanism goes off and it starts mutating like crazy, it is still a step beyond the Modern Synthesis (at least as it is usually taught) which claims not to require any special hyper mutation mechanism.
Ben, those mutations are still RANDOM. Get it? RANDOM.
Random means not favouring a particular outcome.
You do know how the immune system works don’t you? Do you think that the immune system is a threat to the modern synthesis also?
Shapiro claims these mutations are directed and he is simply wrong.
You appear to be talking about epigenetics, not horizontal gene transfer.
The background problem is someting that I find increasingly worrying: many molecular biologist get no training in evolution and, as a consequence, even those working on molecular evolution do not understand the main tenants of evolutionary theory – in particular, the key role of natural selection.
Two weeks ago, in a meeting on Evolution and ecosystem management, I witnessed with incredulity how a colleague defended that plasticity and epigenetics represent a challenge to Darwinism and a revival of what he termed “neo-Lamarckism”. He seemed totally ignorant of the genetic basic and regulation of plastic responses, and the fact that their persistence in population is mediated by natural selection.
Things are made worse by the apparent complexity of the molecular world. People is analysing now the “molecular phenotype” of individuals, all under fency names (genomics, proteomics, metabolomics…), but making easier to loose sight of the connection to the phenotype to the process of natural selection.
Of course, an additional reason is that fency names make for sexier conferences and publications – even more in the media.
This seems to be a misunderstanding. While plastic responses can indeed come under genetic influence, morphological and other phenotypic plasticity are general properties of uncanalized systems and don’t need to have been specifically evolved. There are numerous cases of environmentally induced ecophenotypes becoming genetically fixed after the fact. Segment number in centipedes, for example, tracks temperature clines. Any why shouldn’t it? Segmentation is determined by biochemical oscillators which have temperature-sensitive dynamics.
Whoa, next he’ll be reinventing the Baldwin effect.
Has the Baldwin effect been discredited? Or just marginalized.
I am pleased to see this dispute over Shapiro’s ideas have devolved into a discussion of horizontal gene transfer with both sides accepting that HGT might be an important factor in evolution.
I want to take issue with those who are maintaining that the influence of HGT can be easily incorporated into our thinking about evolution especially as has been called “modern synthesis”, or “new synthesis” or “neo-Darwinism”. I can say from personal experience in efforts to get my work published over the past 30 years that this is not true. (See here http://www.vme.net/hgt/ for list of papers available as pdfs). Paper #1 at the site was submitted to 7 different journals before being accepted. Many of the referees were definitely disciples of one or more of the isms and felt the notion of HGT was unacceptable for ideological reasons. Obviously the climate has improved considerably. Perhaps one of those isms can be sufficiently modified to accept HGT (maybe neo-Darwinism since it seems to be a little more loosely defined). But it will be a new theoretical construct and will have to accommodate many new factors.
If we keep the ideas very general there will be little problem with existing theory. The role of natural selection, descent with modification and common ancestry as an explanation for homology are quite compatible. But here is a list of things that are more problematical.
Transposable elements are clearly involved in HGT; this has been documented in not just prokaryotes but also in plants, animals, fungi and other eukaryotes. In efforts to understand why they are so widespread and why they persist in nature it is often much easier to consider group selection rather than individual selection explanations as I argued in papers #2 and #15 in the above link. When a TE enters a new population, its rapid spread is easier to explain as an infectious process. This has been observed with I and P factors in contemporary populations of Drosophila. In historic times we have witnessed what is equivalent to a sympatric speciation event. Now this is something a Modern Synthesizer would have difficulty accepting in 1959. (Paper #2 above link discusses this).
Common ancestry remains the best explanation for homologous genes, but there is no longer any need to cite the need for common ancestors in which those genes originally evolved. That is to say, what has been called the Last Universal Common Ancestor is probably something that never existed. (Papers #9 and #15 at above site).
Also the phylogenetic tree, which is widely sought by many, is probably another ideological construct with little support in reality. I am not sure if the loss of trees (to be replaced with nets) necessarily violates any of the above isms, but I do know from my own experience that many very knowledgeable biologists with sound backgrounds in evolutionary topics (especially systematics) have very strong opinions on this topic.
It has always been difficult to find a good definition of “species”. Some of the above problems may be overcome by changing definitions but if it is broadened too far the whole definition of speciation itself will have to reconsidered.
Just one final note: HGT can roughly look like Lamarkian inheritance, but that is just a superficial similarity. Darwinism provides a better explanatory framework.
I’ve seen your work.
Have you managed to show that HGT has significance above the level of microbial lineages yet?
seriously, the idea that somehow genes are being transplanted in metazoa, from one species to another, and that has evolutionary significance… puts you right on the edge of crankdom.
What are you, a member of Synthesis Stasi? What about this? http://www.ncbi.nlm.nih.gov/pubmed/21838889
Or this?
http://www.ncbi.nlm.nih.gov/pubmed/22997182
stop trying to play gotchhya and actually READ the papers.
“However, some of these genes are also found in animal genomes”.
hmm… this is incredible, I had this same discussion with someone who know nothing about biology and I though “Ok… this can be a common misconception if you don’t know enough about the subject”, but a (molecular) geneticist saying that thing is just absurd.
Why can this happen?
We’re supposed to direct the “why” questions to the theologians. Or so they tell us.
Maybe, just maybe, he isn’t the one operating under a misconception.
And what that misconception would be?
Is it a misconception to say that natural selection acts after the genetic variants already exist in a population? The origin/explanation/reason of the variant being there (HGT, migration and mating, random mutation, etc) would not stop natural selection from happening and will not diminish its importance.
You are misconceiving Shapiro’s point. Shapiro and those in his camp don’t think that natural selection is unimportant or unnecessary. Rather they don’t think it is sufficient if it is only acting on random single place mutation as was thought when the Modern Synthesis was formed.
Shapiro posits the existence of mechanism within cells that creates mutations to handle stresses from the environment.
Behe says that there had to have been an outside designer who engineered favorable, but otherwise ridiculously improbable, mutations for natural selection to act upon. Shapiro says that there is no need for an outside designer. Rather the cells themselves contain a mechanism that engineers favorable mutations for natural selection to act upon.
Neither of them are denying the importance of Natural Selection. They are saying that random mutation as conceived within the Modern Synthesis is insufficient.
Ben, I have to disagree.
Saphiro states that: “The well-established molecular details of horizontal transfer in the evolution of bacterial antibiotic resistance are difficult to reconcile with neo-Darwinism.”
Which is utterly wrong. Or maybe I just don’t know what neo-Darwinism is, because the theory I am familiar with, do not have the problem stated by Saphiro.
Neo-darwin evolution is not simply the claim that natural selection operates, it is the claimed that natural selection is THE process which gives evolution its adaptive bias. Natural selection is sufficient to ensure adaptation, but HYPOTHETICALLY it is not necessary. If mutations are fundamentally biased in adaptive directions, then natural selection is not creating the adaptive bias of evolution*. In this case you could argue that natural selection is reduced to the secondary roll of simply removing what is bad, instead of being fundamentally involved in accreting what is good. In this regard Shapiro’s view (if I understand it) is radically different from neo-darwinian evolution. I think we can at least agree that a view of evolution where God herself introduces all beneficial mutations is incompatible with a modern understanding of evolution. In the same way Shapiro’s views are incompatible with a modern understanding of evolution. Luckily no evidence supports Shapiro’s notions.
*This is basically just lifted strait from Gould.
matthewackermann:
I’m not sure I understand what you tried to say here. Accreting what’s good and suppressing what’s bad are, to my mind, exactly the same thing: differential reproductive success increasing allele frequencies in population gene pools. Could you elaborate?
Wow, Jerry, James Shapiro really doesn’t like you! I just spent an interesting few days trading comments with him on that HuffPo blog (and learnt a lot about his position, I must say) but as soon as he found out that I was an atheist and liked your book, he went totally crazy and started accusing me of all sorts of stuff. Quite sad, really, especially as I had already identified myself as a commenter on your website and made it clear (I thought) that I was there purely of my own accord and expressing my own opinion.
I did wonder – and posed the question to him (no answer) – why it does not seem to bother him that so many evolutionary biologists think he is wrong. (It would bother me!) Now, I am beginning to wonder if he’s one of these strange conspiracy theorists who thinks that we all sign up to some secret creed in the church of Dawkins and Coyne.