The free-rider problem

This is inspired by a discussion I had over on Evolving Thoughts.

Perhaps one way to approach part of Fodor’s argument against natural selection is to focus on the lack of a “free-rider” type problem in other areas of science. I’m curious what others think of this.

So, define a selection free-rider as a phenotypic trait that was not “selected for” in some particular instance, and is correlated with a trait that was selected for. Further, the free-rider could have been selected for (making the correlated trait the free-rider), either actually in some other species or population, or by some possible biological scenario. For instance, the brownness of fur might be a free-rider in one species, because it is correlated with the fur having the insulating properties that it does, which is what was actually selected for. But in another species (or maybe even in another population of the same species), it is the insulating properties that free-ride, because the brownness is selected for (because it provides camouflage, or whatever). If we think of selection as a mechanism or force (which is not uncommon in evolutionary biology), we might say that if selection does not act on a trait T1, but does act on a correlated trait T2, then T1 is a free-rider. If selection had acted on T1, then T2 would be the free-rider.

Now it seems to me that there is nothing like a free-rider problem in sciences such as physics or chemistry. Consider gravity (and I’ll stick to Newtonian mechanics just for simplicity): we know that gravity acts on things with mass, and not on things without mass. So we might think that the massless properties of objects are gravity free-riders, because they ride free on gravity. Maybe color is a gravity free-rider: each of my 2kg red books fall exactly the same when dropped from the same height. But this can’t be right: we already knew that gravity only acts on mass, nothing else. Gravity doesn’t act on mass in some contexts, but not in others.

The problem then is that selection sometimes acts on brown fur and sometimes doesn’t, and this is true for every phenotypic trait. But gravity always acts on mass, magnetic forces always act on iron-containing things, and so on. I take this to be closely related, if not identical to, Fodor’s complaint that there aren’t laws of selection. Because gravity doesn’t have free-riders, I can tell you exactly how it will act on some object if you tell me its mass, which is just applying the Newtonian theory of gravity. But if you tell me all the traits that an organism has, I can’t tell you which one (or ones) selection acted upon. I can’t even generate a list of traits that selection will always act upon. Of course, in one particular case, I may be able to actually figure it out, but that’s not because of the application of theory. The theory simply tells me that some one (or more) of those traits was acted upon by selection.

The above is pretty sketchy, but as I said, I’m curious of what others think. Even if Fodor draws the wrong conclusion, this does seem like an interesting difference between biology and other sciences. Are there examples of free-riders in other sciences that I’m just not thinking of? There are lots of examples where one has to mind the difference between correlation and causation, but the selection free-rider problem seems like a special case: causes and correlations can be switched around depending on context.


  1. gualtiero

    Corey, what you say sounds reasonable to me. I don’t think there is a free rider problem in more basic sciences, and this is one example of the difference between evolutionary biology and more basic sciences. But isn’t it uncontroversial that evolutionary biology is a historical science and thus different from non-historical sciences? Sciences differ in all sorts of ways, don’t they?

  2. tad

    Interesting post. I think basic sciences aren’t the appropriate comparison class though. You need to look into special sciences, like geology. Haven’t been able to come up with an analogous case though. It strikes me that a lot of the complaints Fodor raises about evolutionary biology count against all parts of psychology. So if you think that most special sciences dealing with living things are based on a fundamental conceptual flaw, what can one say? The flaw can’t be that important if so many find this kind of research useful. In any case, it seems that the free rider problem you identify is purely epistemological. But Fodor wants to extract a metaphysical conclusion from this – that there is no fact of the matter regarding what some trait was selected for. Even if it’s really hard to discover, it doesn’t follow that there is no fact of the matter.

  3. Bill

    Tad: good idea. In geology, erosion rates of something such as a mountainside depend on different factors in different environments. So, when we look at an eroded area, we can say that the amount of rock remaining is due to various factors in the historical environment (glaciation, rain, sun, plant life, etc). Some of the rock’s characteristics will be relevant and some not.

    To take the analogy further, can we say something poetic, like “Nature created this beautiful canyon over a million years,” and then watch sparks fly?

  4. We could turn Fodor’s problem into a problem for an undergraduate intro to evolutionary biology class.

    Evolutionary Bio 101:
    1. Fodor (a philosopher) has posed this argument : blah blah blah. Suggest experiments and observations that would help us determine which of the correlated traits was selected in the given environment. Be sure to explain the rationale behind such experiments and observations.

    Does anyone think that is such a hard question? It seems fairly trivial to me. For one, find dissociations between the traits in multiple species, and see if their values correlate with environmental variables in predicted ways. This is basically what Block and Kitcher point out.  If there is variability in the traits, you could directly measure the relative fitness in different environments (e.g., brown versus white fur in snowy versus balmy environs).

    Just to repeat some bits from an earlier comment…

    Does Fodor discuss the evolution of antibiotic-resistant strains of bacteria in any detail? Seems like a good case study to determine how well his argument holds up.

    Fodor brings up data that any Freshman biology major already knows (e.g., traits tend to be correlated), and points out something evolutionary biologists already know (that this makes selectionist explanations tricky). He then proceeds to kill the baby. It seems Block and Kitcher do a pretty good job of dismantling his argument, and expressing the biologist’s perspective.

    I guess when he wrote “If you want to know about the mind, study the mind –not the brain, and certainly not the genes” he pretty much stopped mattering to me as we are living on different planets. Obviously I’m super biased so encourage active neglect of anything he says about biology. 🙂 Why read Fodor when you can read Wimsatt and Bechtel?

  5. Martin Roth

    Among the conclusions F & PP want to draw is that appeals to natural selection do not support certain function attributions, e.g., the claim that the function of the heart is to pump blood (and not any of the possible “free-riders” that may accompany this effect, e.g., making thump-thump noises). Since I think there are good reasons for accepting the conclusion, I find it ironic that so much attention has been paid to their (admittedly) bad arguments for it. Discouraging, too, for I suspect that many will view the faulty inference as somehow giving cover to the view of functions they reject (we all have experience with this, right? Someone gives an obviously terrible argument for a non-obvious conclusion (one that you accept, but for different reasons), with the consequence often being that the non-obvious conclusion is tainted by the bad argument for it). Thanks F & PP!

  6. Gordon

    You’re making a mistake regarding what it is for a trait to be selected. Remember, “Survival of the fittest” is a necessary truth because evolutionary fitness is defined as “that which is passed on to the next generation of organism.” As a result it’s impossible for something to be passed onto the next generation of organism and not be selected for (the free rider trait) being selected for just means that which is passed on.

    You could think about evolution as a sieve of death through which some things pass, the process of passing through the sieve is natural selection, whether or not you can tell a plausible story about the usefulness of that trait.

Comments are closed.

Back to Top
%d bloggers like this: