55 of 68 people found the following review helpful
This review is from: What Darwin Got Wrong (Hardcover)
Ever since Stephen Jay Gould and Richard Lewontin launched their attack on the "Panglossian paradigm" of adaptationism, biologists have been cautious about claiming some well-adapted trait was shaped by selection for its current function. An adaptive trait, Gould and Lewontin argued, could simply be a lucky by-product of selection for some other trait.
They drew an analogy with the spandrels of San Marco: at first glance, these features linking the dome and arches look to have been designed for the sake of the beautiful images that adorn them. But further reflection reveals otherwise: they were actually a by-product of resting a dome on arches! The moral for biologists: take care to distinguish the real products of selection from the "free-riders".
In their provocative new book, Jerry Fodor and Massimo Piatelli-Palmarini want to draw a different moral from this story. What it really shows, they argue, is that the idea of a trait being "selected for" is incoherent. To say the spandrels were put there to hold up the dome is, after all, to make a claim about what the architect had in mind. Since, by contrast, there is no mind in charge of natural selection, it makes no sense to say that some trait was "selected for" while another was a "free rider". Though they add a lot of complicated extras, this is the core of their master argument against Darwinism, as set out in Chapter 6.
So here's the obvious reply: the difference between selected-for traits and their free riders is a causal difference. Selected-for traits causally contribute to the reproductive success of organisms, whereas free riders don't. To say some trait is a "free rider" is to say that, regardless of its current function, it evolved without contributing to the success of its bearers. This is going to be hard to find out, of course, but the conceptual distinction is clear enough. No minds are needed.
This is what Fodor's critics have been saying for years. Yet it's not an objection explicitly addressed in the book, despite being, as far as I can tell, a perfectly good one.
It is important to realize the full scope (and extraordinary arrogance) of what Fodor and Piatelli-Palmarini want to achieve here. Their ambition is not merely to downplay the significance of selection in evolution, as defenders of "evo-devo" often seek to do. Rather, they intend to annihilate the entire theory of evolution by natural selection a priori, from the recesses of their armchairs, with a knockdown objection at the conceptual level.
If they were right, biologists would certainly finish the book with egg on their faces. How stupid! To spend 150 years thinking some incoherent nonsense was the best idea ever! Unsurprisingly, all the egg goes the other way. In Daniel Dennett's words, "What could drive Fodor to hallucinate the pending demise of the theory of evolution by natural selection?"
Sort: Oldest first | Newest first
Showing 1-5 of 5 posts in this discussion
Initial post: 2 Mar 2010 20:44:42 GMT
Last edited by the author on 2 Mar 2010 20:47:32 GMT
Mr. Allan Crossman says:
"An adaptive trait, Gould and Lewontin argued, could simply be a lucky by-product of selection for some other trait."
That's not strictly correct. An adaptive trait will either have been selected for directly, or will be a modification of a trait that first arose for some other reason, but is now selected for its current purpose (this is called exaptation).
What Gould and Lewontin argued was that some traits are not adaptive at all. The "Panglossian" program they criticised was the attempt to explain absolutely everything in adaptationist terms.
Posted on 4 Apr 2010 21:03:15 BDT
Last edited by the author on 4 Apr 2010 21:18:44 BDT
"So here's the obvious reply: the difference between selected-for traits and their free riders is a causal difference. Selected-for traits causally contribute to the reproductive success of organisms, whereas free riders don't. To say some trait is a "free rider" is to say that, regardless of its current function, it evolved without contributing to the success of its bearers. This is going to be hard to find out, of course, but the conceptual distinction is clear enough. No minds are needed."
What you are missing here JB is one of Fodor and Piattelli most significant points. It follows from their argument that any attempted refutation of this kind commits you to a wholly circular form of reasoning. You only know which traits have been selected for and which are free riders if you assume natural selection is the mechanism of evolution, and you can only evidence natural selection as the mechanism of evolution if you independently know which traits have been selected for and which are free-riders.
The logic behind their argument needs to be carefully interrogated and elaborated to make sure it is completely sound, but it is going to take a more subtle argument than this to squash it. As a committed Darwinian, I think their analysis is pretty scary.
I'm not going to elaborate here because I've detailed this argument in my own review and in a couple of comments to it.
True science thrives on challenges of this kind. They are its life's blood. Treating theories like holy writ is exactly what we need to avoid. I think the hubris belongs very much to Dennett ("How dare they challenge the theory I have spent my life defending") and not with Fodor and Piattelli who, as a matter of fact, write very modestly and unassumingly.
In reply to an earlier post on 16 Apr 2010 10:47:56 BDT
Last edited by the author on 16 Apr 2010 11:29:47 BDT
Jonathan Birch says:
It's important to realize that Fodor is not doing epistemology. Naturally, *finding out* whether a trait has been selected for is not an easy business. But that is not the issue. Fodor's claim is that the very notion of "selection for" is incoherent.
At bottom, Fodor's argument appears to rest on a fundamental misunderstanding of what biologists mean by "natural selection". Fodor envisages a process in which a black-boxed mechanism (called "natural selection") scrutinizes traits for their contribution to fitness and "selects for" the traits which make the biggest contribution, thereby ensuring that these traits increase in frequency.
Fodor (rightly) demolishes any such notion. As he points out, this sort of scrutinizing process would have to be intensional rather than extensional, in so far as it would need to distinguish between the causal contributions to fitness made by coextensive traits. Yet only a mind, or perhaps a process governed by its own peculiar (intensional) laws, could be intensional in the required sense. An ordinary causal process cannot do that job.
It is a subtle argument, but Fodor has demolished a straw man. In reality, biologists simply do not posit a scrutinizing mechanism called "natural selection". The idea of natural selection as an agent, "daily and hourly scrutinizing", is metaphor. Darwin's insight was that heredity plus differential reproductive success does the job all by itself: there is simply no need for a further mechanism to "do the selecting".
In this picture, talk of "selection for" some trait does not require the existence of some black-boxed selective mechanism. It is a term of art, identical to saying that the trait causally contributes to fitness. Construed in this way, it is obviously coherent.
You may think I am being harsh here. But I need to be, for Fodor's conclusions are deeply and dangerously wrong. If you want to read more about this, see the responses by Elliott Sober, Samir Okasha or Philip Kitcher & Ned Block: all excellent philosophers, all utterly damning in their assessment of this dreadful book.
In reply to an earlier post on 16 Apr 2010 11:06:32 BDT
Last edited by the author on 16 Apr 2010 11:30:29 BDT
Jonathan Birch says:
I'm not sure there is any genuine disagreement between us here.
Gould and Lewontin criticize the adaptationist tendency to see natural selection as an "optimizing agent"; they also criticize a related tendency to draw no distinction between a trait's "current utility" and "reasons for origin". Their problem with the latter tendency is that a useful trait need not have originated through selection for its current function. It may originally have been shaped by selection for a different function, or it may have originated as a developmental by-product of something else.
Posted on 16 Jul 2012 15:01:02 BDT
Mr. Daniel J. Merceron says:
"Hubris" - that's an interesting term. Do I detect a little concern.
‹ Previous 1 Next ›