Jerry Coyne has done a joint review of Dawkins’ “The Greatest Show on Earth” and Fodor and Piattelli-Palmarini’s “What Darwin Got Wrong”. I’m not all that concerned with Dawkins’ book, and think that it might go too deep for my limited interests. However, after reading Coyne’s review, I am interested in “What Darwin Got Wrong”, for two reasons:
1) Because after reading the review I think they might have a point.
2) Because after reading the review I wonder if they’ve also gone insane.
It’s never good to judge a book by the writings of someone who has a stake in them being wrong, since all sorts of misinterpretations can occur. So, to tell if 2) is right, I have to read it myself to make sure that the insane claims really are insane. But I can deal a bit with 1), by looking at what Coyne says about it and seeing if his replies might make sense against insane points but not against sane points (we might end up agreeing on the same points, oddly enough).
Coyne’s “ad” for his review is here: http://whyevolutionistrue.wordpress.com/2010/04/22/i-review-what-darwin-got-wrong-and-the-greatest-show-on-earth-in-the-nation/
The review — and it was most gracious of the magazine to allow it to be put up for free on-line, at Coyne’s insistence, so kudos to both — is here: http://www.thenation.com/doc/20100510/coyne/3
(We start on page 3 because that’s where I’m going to start; the previous two pages talk more about Dawkins’ book which I don’t think I’d object much to, so …)
So, let’s start with Coyne’s criticisms:
“… let’s examine F&P’s claims. These fall into two groups. The first is that scientists have recently discovered a lot of things about genetics and development that make natural selection look ineffective:
“Contrary to traditional opinion, it needs to be emphasized that natural selection among traits generated at random cannot by itself be the basic principle of evolution. Rather, there must be strong, often decisive, endogenous constraints and hosts of regulations on the phenotypic options that exogenous selection operates on.”[reformatting of the quote is my screw-up; this section is a quote from “What Darwin Got Wrong”]
In other words, Darwin’s assertion that species are “quite plastic” is wrong: organisms are so constrained by their biological nature that they’re not free to change, even if it would be good for them to do so.”
Whoa. Wait. That’s not an “in other words”. That doesn’t seem to be what F &P-P (Fodor and Piattelli-Palmarini) are actually saying. In fact, just from that quote they seem to be saying the precise opposite: that in order for any sort of external selection to work, there has to be constraints on the organism, and so you can’t just get random changes selected for. This seems pretty reasonable, as if it was completely random changes you could get an awful lot of changes that either didn’t relate to the previous ones that were successful or that were, in fact, fatal, and even fatal to a more developed organism. And it seems to me that this notion of constraints is, in fact, precisely what Dawkins uses to deny that natural selection is actually “random”; you can’t get any change, but can only get changes that are allowed by the DNA expressions and structures of the organism. You don’t just get wings, but you first get small changes, selected for, that can develop over time into radical changes, but it doesn’t just happen at random.
Now, how much this impacts the ideas on either side is an open question, but I don’t get a good feeling when even I — and, I admit, I’m not an expert in biology and, quite frankly, hate doing it, personally — can look at a quote and think that the interpretation of it is somewhat lacking.
Let’s look at one of these: phenotypic plasticity. This refers to the ability of a phenotype–an observable trait or characteristic of an organism–to change within a single generation in response to environmental fluctuations. This is what happens, for instance, when you get a tan. If you have an outdoor cat, its fur gets thicker in winter. The plumage of Arctic animals like the ptarmigan, ermine and Arctic hare changes color from brown to white as winter comes on. Even the lowly brussels sprout has sophisticated plasticity: when it detects that a sprout-eating butterfly has laid eggs on the plant, it changes its leaf chemistry to attract parasitic wasps that destroy those eggs.
F&P imply that somehow–they’re not clear about how–this ability to undergo adaptive developmental change within a generation prevents natural selection from causing genetic change between generations. But that isn’t the case. In fact, far from being an impediment to natural selection, the ability of an individual to adapt to a changing environment is a product of natural selection! Individuals who can tan in the sun (and thus prevent melanomas) have an advantage over those whose pigmentation is fixed. Cats are better off if the length of their fur suits them to the seasons. Genes that are able to respond to predictable variation in the environment will always outcompete those that produce only a fixed (and hence episodically maladaptive) trait.”
This might be a fair interpretation of their claim, and if this is their claim then, yes, they’re nuts. But let me posit this: Imagine that we have a beneficial trait — say an increase in brain size — that mutates into an organism. But, just by sheer “luck”, it happens to only propogate amongst cat-like creatures that don’t have the ability for their fur to grow thicker by the seasons. Now, from what Coyne has just said above, we know that growing thicker fur confers an advantage. Does it confer a greater advantage at that moment than increases in brain size? If it does, then selection will weed it out, not because it isn’t advantageous but because it is correlated with another trait that isn’t advantageous. Now, let’s imagine another trait, like having lighter coloured fur (so it stands out more in the environment). But let’s say that it happens to be the case that it mainfests in the animals whose fur can thicken. It, then, should survive as long as it is not so deterimental to the organism that it drops its survival below that of the animals that do not have the “thicken fur” trait but do have the “brain size increase” trait.
Now, it would be a fair argument to say that this case is fairly contrived, and I concede that. My concern is that it seems to me that, in principle — at least at this time — you couldn’t tell the difference. Why is this a problem? Well, it means, to me, that attempts — particularly problematic in evolutionary psychology (which is where I encountered and hated it), but which seem to be common in everything dealing with evolution — to explain a trait by assuming it is beneficial and then finding an explanation for how it could have been beneficial seem pretty dangerous, and that if there is disagreement it is really hard — if not impossible — to tell who’s right. If someone claims that a trait is a free-rider, and other people claim that it has benefits (but don’t agree on the benefits) at least for traits developed over the long term we really don’t have a way to settle that. That’s something that should give biologists some pause, at least. (I’ll address the examples where we can, later. I tend to like to go in order, which can be confusing but makes my life easier [grin]).
Indeed, virtually none of the biologists who study the “constraints” described by F&P share their dim view of natural selection. That’s because, over and over again, selection has wrought the most improbable and unpredictable changes in animals and plants. F&P claim, for example, that selection could never produce winged pigs because of developmental constraints: “Pigs don’t have wings because there is no place on pigs to put them. There are all sorts of ways you’d have to change a pig if you wanted to add wings. You’d have to do something to its weight, and its shape, and its musculature, and its nervous system, and its bones; to say nothing of retrofitting feathers.”
Haven’t F&P heard of bats? Bats evolved from small four-legged mammals, probably resembling shrews. You could say the same thing about shrewlike beasts that F&P did about pigs: how could they possibly evolve wings? And yet they did: selection simply retooled the forelegs into wings, along with modifying the animal’s weight, shape, musculature, nervous system and bones for flying (no feathers needed). One of the great joys of being a biologist is learning about the many species in nature whose evolution would appear, a priori, impossible.”
Okay, see, I’ve read about the “pig” example, and every time I’ve read it I’ve ended up interpreting it as really this argument: “It’s not the case that pigs don’t have wings because wings aren’t beneficial to pigs, or that there was some ‘have wings’ trait that was selected out by natural selection. Pigs don’t have wings because they’ve never, ever, exhibited a trait that could lead them to wings. Pigs just don’t have any ‘wing-creating’ trait, and never exhibited one. Selection only works on what can be selected for, and if a trait — even in its basic stages — doesn’t manifest it will never be selected for, no matter how beneficial it would be for the pig to fly”. And this seems pretty reasonable to me. Any attempt to say “pigs don’t have wings because there is/was no benefit to them developing wings” is, in fact, neither empirically nor philosophically sound.
But this in no way discredits the theory about bats. We can in fact presume that some precursor of wings did develop in bats, and in some way that filtered through a number of adaptations until they had full wings. Whether or not all of those — or even most of those — adaptations were selected for as opposed to being free riders seems, to me, to be an open question, but that bats did develop wings and did so gradually from shrew-like creatures is, to me, undeniable.
So, if F & P-P really do say what Coyne interprets them as saying, they’re insane … but that’s not how I’ve taken that every time I’ve read it. I’ll have to read the book to say for certain.
“But wait a minute. If you translate that last sentence into layman’s English, here’s what it says: “Since it’s impossible to figure out exactly which changes in organisms occur via direct selection and which are byproducts, natural selection can’t operate.” Clearly, F&P are confusing our ability to understand how a process operates with whether it operates. It’s like saying that because we don’t understand how gravity works, things don’t fall.”
If F & P-P are really saying that nothing ever gets selected for on the basis of benefit, then they’re insane. But I think the biggest issue is this: how can you say that natural selection is responsible for our traits when you don’t know how many of our traits were propogated through the species because they were beneficial or because they were free riders on other things? If, say, 90% of our traits were either free riders or free riders until an environmental change made them beneficial, you can’t say that things like: brain size, consciousness, eye sight, bipedalism, opposable thumbs, etc , etc are the way they are because of natural selection, since any and all of them might have developed as free riders. And if it could be the case that the vast majority of our traits could owe most or all of their presence to free riding, doesn’t that make natural selection less important, and other factors more so? And so shouldn’t we stop just looking for benefits to explain traits?
Now, Coyne does have a reply to this. Before I discuss it, I want to make it clear that I don’t deny that traits can persist and propogate because they are selected for on the basis of benefit to reproduction. There are clearly some cases where this occurs, and Coyne’s examples may well be of such cases. The issue, for me, really is how we determine this generally, not in specific cases. If Coyne can prove that a specific case is certainly chosen by natural selection, then I’m more than happy to grant that case (and hope that F & P-P would as well).
So, the peppered moth:
Here’s a more realistic example. Perhaps the most famous case of natural selection in action is the color change that occurred in Britain’s “peppered moth” over the past 150 years. Before the Industrial Revolution, these moths had white wings speckled lightly with black, although avid collectors found a few all-black mutants. As pollution from manufacturing increased the concentration of suspended particles in the air, black moths became more numerous, and eventually predominated in many places. When clean air laws reduced Britain’s pollution in the 1950s, the evolution of wing color reversed, and in most places the white color once again became common. The difference between white and black moths was shown to reside at a single gene.
What caused these evolutionary changes? There were several theories. One was that the target of selection wasn’t the moth’s color but the survival of caterpillars that, while not showing the color differences of adults, happened to be affected by the same gene. Another suggestion was that natural selection acted on color: perhaps sharp-sighted birds picked off moths whose color contrasted with that of the trees on which they rested. In unpolluted woods, lichen-covered trees are light-colored but turned black as pollution increased. This would give a selective advantage first to the dark-colored moths and then, as pollution abated, to light-colored moths.
F&P would presumably counsel us to give up at this point, since we can’t, they say, distinguish between the counterfactuals of selection “for” larval survival and “for” adult color. But we can! Breeding experiments in the laboratory showed that the survival of caterpillars couldn’t explain the increase and subsequent decline of the black form. In contrast, field experiments that involved observing predation on dead moths of different colors fastened to trees of different colors, and on live moths of different colors released in unpolluted woods, showed that selection on color was strong, easily able to explain the evolutionary changes observed in nature.”
I think this is a pretty strong case, so let me nitpick that last statement a bit: his claim is that the observations showed that colour is able to explain it. Does that mean that that is what happened? What if there was an unobserved gene that simply allowed black moths to better survive in polluted areas than the whiter ones? You might still see selection by predation, but that wouldn’t be what was doing most of the work, even if the claim was that it was strong enough to do so. How would we test this claim?
And this is in a case where we have lots of readily available data, and one could still raise — pointless, I admit — doubts. Now, let’s take the list of traits that humans or even pigs have. For how many of them could we do that sort of analysis? How many of them can we prove exist only because of a benefit? The stegosaurus example is not the only case where we really have no way of proving it, one way or the other.
See, there’s an additional problem here. To be selected for, a trait has to develop. And so someone might be able to find a case where that trait did develop because of a benefit. But it has to be in the gene pool first, which means it has to a) manifest and b) survive. Take our black-coloured moths. There were black-coloured moths — as far as I know — before the industrial evolution. It wasn’t the most common trait, but it was there. So why was it there? Because it benefitted? Unlikely, and absolutely false if Coyne’s explanation is the right one. So black colouring, then, was a free rider that simply happened to end up beneficial when the environment changed. So, then, it is reasonable to ask what the peppered moth is an example of? The trait is not explained by natural selection, but the numbers of individuals with that trait is. I suspect that F & P-P want explanations for why they have that trait (black-colouring) and don’t really care about why most moths don’t exhibit that one. And the explanation given, it seems to me, doesn’t explain how they got that trait in the first place (and, interestingly, seems to work against it).
So, if we have a trait that is bad enough to limit the number of organisms but survives anyway, then it seems that “selection” didn’t operate on that trait. How many of these traits do we have? I don’t know, but I’m not convinced that anyone else does either.
A quick final comment about artificial versus natural selection. Artificial selection dodges all of these issues because it selects for specific traits: the breeding and culling are aimed precisely at getting that specific trait to manifest. Natural selection, as F & P-P point out, does not. It aims at getting an organism that reproduces. Because of this, artifical selection as an explanation for why a specific trait is as it is is a good and reasonable and testable and settleable explanation. But with natural selection, all we have is “It lived and reproduced”. We don’t know why. We don’t know if it was the trait that we are looking at that made the difference or if it played any real role at all in the survival and reproduction. There is an intensely complicated relationship between all the traits in an organism and its environment that confuses things.
In essence, you can’t apply natural selection explanations to one trait alone. You always have to consider all the traits and the envionment the organism it was in to get an explanation for any particular trait. If F & P-P want to simply draw attention to that, I’m on their side. If they want to claim that therefore natural selection never happens, they’re insane. And without invoking the Fallacy of the Golden Mean, from both my experience and from my reading I strongly suspect that the real point is somewhere in between. Heuristically speaking, it usually is.