Insofar as he made no logical mistakes in his reasoning, we need a different strategy to defend the theory of evolution. In this post I will argue that his argument is an instance of gerneral underdetermination, and hence not a problem of evolution but of philosophy of science.
Underdetermination means that we can’t specifically identify the exact cause of scientific phenomena. For example, given some phenomenon, say darkness during the day, there can be many possible explanations: an eclipse, an exploding volcano shooting ash into the air, the sun has gone out, the electric company has blocked the sun to make more money, it was the work of Claw Vipers, etc. The exact cause of the darkness is underdetermined; sure we can research the problem and eliminate some of the possible explanations, but because of our limitations we will never be able to check everything. So the cause of the darkness can be said to be underdetermined, i.e. there is just not enough determining evidence.
Fodor argues that the theory of evolution is vacuous becuase given any trait we identify as benficial to the fitness of the organism is arbitrarily selected. Since there are too many factors to identify within an ecosystem or organism acting within that ecosystem, any hypothesis we propose about the fitness of that organism in that ecosystem will be trivially compatible with evolution.
For example assume there is an argument that having a certain trait, say longer legs, increases a zebra’s fitness. We can recognize that this argument could be unfounded because it might not be the longer legs but something else that increases the zebra’s fitness. It just happened that increased leg length was a harmless side affect of this truly beneficial trait. Either way, if it is the longer legs or some other unidentified trait, evolution is always compatible with our theories, and so it is trivially vacuously true.
In short I would say that he is arguing the cause of natural selection is underdetermined. The task is to identify whether this is a unique case of underdetermination or an instance of general underdetermination. I will now show that this sort of underdetermination can exist in physics*:
Imagine we are doing physics and we want to know which of two metal ingots is the more massive. We pull out our scale, place each object on one of the trays and wait for the scale to indicate which is the more massive.
Why does the scale tip in the direction of object A? We could argue that object A has a trait, it is composed of iron, and that trait makes it more massive than some other object. However, maybe object B is connected to a helium balloon. Maybe there is a gravitational anomaly in the location where we are doing our experiment. Maybe the iron is magnetized and there is another ingot with the opposite polarity under the table. Maybe a God is tampering with our experiment with a noodly appendage. Feel free to make up as many of these as you want. There any number of reasons why one object could tip the scale in its favor, and being more massive is among them, though selecting this as the reason is arbitrary.
(One of the things that is wrong here is that we don’t expect General Relativity to predict which objects are more massive. The mass of an object is the result of the history of its creation and ‘life’ up till the point we measure it. We do expect Relativity to suggest methods for testing such claims, which it does. Likewise Evolution should not be expected to predict which organism is fitter, but to suggest methods for testing fitness.)
If I now recast Fodor’s criticism into physical terms, in reference to the above thought experiment, this is the result: The theory of General Relativity (gravity) is vacuous because any given trait we identify as increasing the mass of an object is arbitrarily selected. Since there are too many factors to identify within a physical system, any hypothesis we propose about mass of the object in that physical system will be trivially compatible with General Relativity.
Therefore physics suffers from the same kind of underdetermination that Fodor accused of evolution. Anyone who persists in disbelieving evolution on these grounds should also deny General Relativity. Of course this is excessive: since the underdetermination criticism goes to the heart of our scientific theories in general, it is a problem of philosophy of science and not a problem of biology or physics specifically. Insofar as underdetermination remains an issue within the philosophy of science we still have to take it into consideration, but this should not be seen as a reason to think our current scientific theories are wrong.
[EDIT: I’ve put up a new analysis (24 March 2010) of Fodor’s argument here: Hypotheses Natura Non Fingo]
* This is the argument I presented Fodor with during our brief conversation after his talk at CUNY. He tried to block it by saying that Natural Selection is statistical, whereas General Relativity is not. In my previous post, What Fodor Got Wrong, I argued that this position begs the question or is just wrong.