Thoughts in a Haystack

Fodor may be wrong in whole or in part, but I think it might be premature to lump him in with the "IDiots," and I'm a little disappointed that a philosopher of science (as opposed to someone like Coyne) would treat his argument as though it were so patently limp.

Fodor uses the spandrels argument as an illustration of his larger point, that evolution is historical, not scientific (in the sense of adhering to explicable, generalizable laws). Ruse's suggestion that we can expose co-extensive traits to experiment ("was it it the eyes or the fur"?) begs the central question of Fodor's thesis: can we ever hope to replicate the melange of influences a certain organism with a certain genome experiences? (As a corollary to that question, if it is true, as Dennett likes to say, that "evolution is smarter than you," then how can we ever know we have gotten it right?)

Fodor does not say that evolution did not take place of course, he just says that the metaphor of selection has extremely limited explanatory power. If permits few if any predictions.

Ruse also misses Fodor's point about "constraints," which is not that they leave no room for other factors, but that they make the adaptationist's job close to impossible. Since phenotypic traits are not tabula rasa, they make "reverse engineering" largely a matter of speculation. We can come up with plausible explanations, but this is, again, history, not science, akin to enquiring into why Brutus killed Caesar. Scholarship on this increases our knowledge, but it uncovers no laws.

I don't follow Fodor all the way to the end of the line (especially on Philosophy of Mind), but I do think he is correct to say that we cannot build a science on the principle that "something advantageous happened," since we need the differential survival to show that it was advantageous in the first place. Yes, it's the old tautology argument. But it's important to make the distinction, since without it there is little to bar the door against every panadaptationist explanation that comes walking down the street.
 

... can we ever hope to replicate the melange of influences a certain organism with a certain genome experiences?

Will we ever be able to replicate the history of a certain sample of lead to its formation in a supernova or by decay from uranium? Does that mean we can't know how lead comes to be or even how frequently each process occurs?

... how can we ever know we have gotten it right?

We never will, since science is a fallibilistic program. We'll just go on testing it.

I do think he is correct to say that we cannot build a science on the principle that "something advantageous happened," since we need the differential survival to show that it was advantageous in the first place.

True enough if we are talking about claims that particular traits evolved for particular reasons. But we don't need to do that to be confident that natural selection is a major mechanism driving evolution from the clear evidence we do have of its efficacy and the simple logical outcome we'd expect given all the other things we know about life: the superabundance of offspring; the death before reproduction of most of those offspring and the fact, that no one denies, that organisms are adapted to their environment.

... there is little to bar the door against every panadaptationist explanation that comes walking down the street.

But since Darwin wasn't a panadaptionist, why is the title of the book "What Darwin Got Wrong"? ... other than the sensationalist marketing aspects, that is? ... a reason to distrust the authors in and of itself.
 

Odd that they use Steve Gould to support their attack on natural selection. While Gould certainly argued against natural selection as 'the' or indeed the most important, mechanism for evolution, nevertheless he wholeheartedly supported natural selection as 'a' mechanism for evolution
 

My main point here is that Ruse does not give credit where credit is due. Fodor is correct that natural selection is overly emphasized as a factor in evolution, to the degree that most laypeople and many biologists have a panselectionist conception of evolutionary biology. It seems disingenuous of Ruse to say "of course we know all about laws of form and self-organization." These are nearly absent from popular accounts of evolution. They are certainly not taught in schools. University students studying embryology make pick up a little along the way, but in the popular press evolutionary factors that are not under the purview of population genetics are treated as a sort of embarassment.

As for selection itself, we seem to agree it is not generalizable to a law in the manner of Boyle, Newton or Ohm. No predictions follow from natural selection. We can't say whether a new environmental pressure will bring forth a phenotypic change, an extinction of the line, or no change at all. (No one predicted peppered moths would darken in response to industrialization in England. First the melanic variant was discovered, and then the explanation given, ad hoc. That Kettlewell and others confirmed it does not make it any more predictable from a general principle of natural selection.)

Selection was an important metaphor in Darwin's day, when the alternatives to the watchmaker analogy were so weak. But elevating it to a law today results in the ideology of adaptationism. That Darwin wasn't a pan-selectionist doesn't completely absolve him from encouraging it, though in later editions of the Orgin he did what he could to stem the tide. Either way I agree that it was both unwise and unfair of Fodor to put St. Charles at the center of the debate, where he doesn't belong.

True enough if we are talking about claims that particular traits evolved for particular reasons. But we don't need to do that to be confident that natural selection is a major mechanism driving evolution from the clear evidence we do have of its efficacy and the simple logical outcome we'd expect given all the other things we know about life: the superabundance of offspring; the death before reproduction of most of those offspring and the fact, that no one denies, that organisms are adapted to their environment.

It's not clear to me. Most of life, historical and present day, does not fit this description. Most organisms are asexual, for example. For 3 billion years the Malthusian condition did not exist. Many organisms have different phenotypes as juveniles, seeds, or spores than the adult forms, obviating the role of selection where most wastage takes place. And "adaptation" to an environment is a poor analogy for what really often happens, when an organism chooses, or even transforms its environment.

But I don't really want to squabble about how prominent selection is, when the answer is not really pertinent to what it can explain which is, we agree, much less than most people credit it with.
 

These are nearly absent from popular accounts of evolution. They are certainly not taught in schools. University students studying embryology make pick up a little ...

Education, as Terry Pratchett said, is organized lying to children. Of course, it's doubtful that 8th graders would be able to absorb more than a simple explanation of evolution. And it is even more doubtful that children, who are lucky to get a week or two of instruction in evolution even in high school, would be able to absorb much of anything while being taught at the kind of hyper-speed needed to get through just the selection and/or its alternatives debate in that time. NS is easy to explain and plausible. That's life.

Most organisms are asexual, for example.

I'm sorry ... how does that effect NS? They are still reproducing (and mutating and/or engaging in lateral genetic transfers) and still competing for limited resources.

Many organisms have different phenotypes as juveniles, seeds, or spores than the adult forms, obviating the role of selection where most wastage takes place.

But the different phenotypes are determined by the same set of genes. "Fitter" juveniles, seeds, or spores will produce more of the adult forms that produce those juveniles, seeds, or spores and vice versa. (I'm reminded of Robert Heinlein's line that "zygotes are just gametes' way of producing more gametes.") That linked traits can have both good and bad effects on "fitness" is not, I think, controversial.

"adaptation" to an environment is a poor analogy for what really often happens, when an organism chooses, or even transforms its environment.

But the trait that allows it to move to different environments or gives it the ability to transform them is subject to selection if it is inheritable. That doesn't mean it has to have been the result of selection but it is hard to see how, somewhere along the line, a trait significant to the reproductive success of an organism won't be subject to selection.

what it can explain which is, we agree, much less than most people credit it with.

Yes, but Gould was able to explain that without such silly arguments as Fodor and Piattelli-Palmarini are making, if Ruse is reporting correctly.
 

Yes, but Gould was able to explain that without such silly arguments as Fodor and Piattelli-Palmarini are making, if Ruse is reporting correctly.

I don't think he is, unless the book is wildly deviant from the papers that led up to it.

But the trait that allows it to move to different environments or gives it the ability to transform them is subject to selection if it is inheritable.

And this ("if it is inheritable") is one of those things taken as obvious a priori. Pleiotropy and plasticity pose a logical problem here. Genomes represent potential, not actual, phenotypic form. This doesn't render selection impossible, but it makes it much harder to demonstrate, and to separate from other factors. (Remember that Fodor is not saying NS doesn't exist). This goes to Ruse's remark that "And in any case, we can ferret out which features are most useful and which are just along for the ride." As a matter of science, we cannot.

I'm sorry ... how does that effect NS? They are still reproducing (and mutating and/or engaging in lateral genetic transfers) and still competing for limited resources.

This was in response to your comment about superabundance, which is a recent development in evolutionary time. But yes, some rudimentary Malthusian logic still might apply even to the simplest bacteria, if we presume a closed system (which goes back to the point about colonizing new environments. It's not always easy to trace the causal chains in selectionist arguments.) My main point here was more about the ideology, than the facts, of evolution, which are often confused, even by big-league biologists and philosophers of science). As Mary Midgley wrote: Nature was green long before she was red (in tooth and claw, that is.)

"Fitter" juveniles, seeds, or spores will produce more of the adult forms that produce those juveniles, seeds, or spores and vice versa.

Again, just in response to superabundance. If differential traits don't appear until adulthood (which is often true, and even when unknown is presumed in reverse engineering speculations), then the wastage of multitudes of juvenile forms is not germane.

I'm not denying selection a role, and I don't think Fodor is either. Rather he seems to me to be saying that orthodox neo-Darwinism often tends toward an "environmentalist" (not in the ecology sense) ideology where the forms of life are whatever the forces of nature require them to be, and it's this selectionist tendency his critique is directed at.
 

And this ("if it is inheritable") is one of those things taken as obvious a priori.

Yes, and it is taken to be obvious, and just as "a priori," that gravity is linked in a lawlike manner to mass based on our measurement of only a tiny number of the bodies in the universe. We've measured natural selection in action in a (relatively) small number of cases of life but those cases confirm our logic as to what is going on and why should we deny that logic?

As to it be heritable, surely we can test that. In any event, are we going to take alligators in the Everglades, who dig out "holes" that act as ponds during dry periods and increase their "fitness" to survive such periods (all testable), are, given what we know about their brains and learning abilities (again testable) to be doing the digging because of "cultural transmission" or, worse, divine design? Or do we take it, at the very least, as a good first approximation, that this is a heritable trait? It is certainly hard to see how it would be the result of "constraints" or "evo-devo". Genetic drift could be an explanation of how the trait arose but traits arising from drift are, by definition, heritable.

But yes, some rudimentary Malthusian logic still might apply even to the simplest bacteria, if we presume a closed system (which goes back to the point about colonizing new environments.

Unless the new environment is infinite, there will always be scarce resources in the end. The geometric reproduction of organisms will, eventually, sooner rather than later, guarantee that.

This doesn't render selection impossible, but it makes it much harder to demonstrate, and to separate from other factors.

Herein lies the nub of our disagreement, such as it is. You could say exactly the same thing about the snowfall that my area is experiencing at this moment. We are, I think, reasonably certain that there are a number of factors that go into how much snow I'll be getting. Air pressure, dew point, temperature, wind patterns, etc., etc. We can make reasonable conclusions about which factors are more important and which are less important as to this particular snowfall. After looking at a large number of snowfalls, aren't we able to identify the ones often associated with certain types of large snowfalls and those associated with lesser ones? In short, is the fact that it is harder to demonstrate the same as saying the factors are not demonstrated?
 

The difference, though, is that knowing air pressure, dew point, temperature and wind direction allow us to predict snowfall pretty accurately. We cannot predict phenotypic change except in controlled conditions, just as we cannot predict what will happen when a ruler is assassinated, when philosophical ideas are introduced to new cultures, or when a population is decimated by disease. We have reasonable opinions about these things, enough to, say, oppose military adventures in Iraq. But we could never say that it was a scientific fact that invading Iraq would cause civil war, increase terrorism, destabilize the infrastructure, raise the price of oil, kill millions of civilians, (or, in the converse, spread democracy, increase the production of cheap oil, liberate the Shia) in the same sense that we could say it's a scientific fact that moisture, air pressure and temperature interact in specific ways to produce snow.

On heritability, too, I think we're still miscommunicating--equating phenotypic persistence with genetic heritability is the very assumption I'm trying to question--but rather than hash it out here I'd just say that I continue to think that Fodor's point is much more modest than his critics take it to be (though it doesn't help that he sometimes seems to deprecate the concept of selection entirely, and that he wants to blame Darwin for the whole thing, when what he is really attacking is neo-Darwinism).

What I take away from Fodor is this: putting too much pressure on the metaphor of selection as an explanandum almost invariably ends up promoting an ideology, not a science. A real scientific theory would put due emphasis on the factors that militate against variation (and thus selection), which have been elaborated to various degrees, but don't fit in as well with the drama of the struggle for existence. Other writers have been saying this for decades, including major biologists like Bertalannfy, Waddington, Stuart Kaufmann, and Brian Goodwin who died this summer, and philosophers of science like John Dupre, Susanne Langer, Mary Midgley--or John Wilkins. Fodor is more combative than these, and already has bad blood with Dennett over philosophy of mind. Plus he over-reaches and goes for the cheap shot. But he's not the crackpot that Ruse or Coyne make him out to be on this topic, and we can't hold Fodor's personality against him any more than we mark down Dawkins' selfish gene thesis based on his own famous stridency and over-exuberance. If trying to sell books is the crime, then the only innocents are the ones we've never heard of.