Laws of Selection?

Jerry Fodor and Massimo Piattelli-Palmarini’s much
anticipated book “What Darwin Got Wrong” is coming out in February.  I am sure many here followed the heated
discussion prompted by Fodor’s LRB article “Why Pigs Don’t Have Wings”
(available here:,
and perhaps many have read Fodor’s (still unpublished?) manuscript “Against
Darwinism” (available on his faculty page:

So what did Darwin get wrong, according to Fodor? One of the
arguments developed in “Against Darwinism” goes like this:

P1: If the theory of natural selection explains the distribution
of phenotypic traits in a biological population, then there are laws about
traits-that-are-selected-for as such.

P2: There are no laws about traits-that-are-selected-for as

C: Hence, the theory of natural selection doesn’t explain
the distribution of phenotypic traits in a biological population.

Regarding P1, Fodor has it that the theory of natural
selection aspires to a kind of generality that requires laws (or, at any rate,
it needs laws if the theory is to do more than provide historical narratives
which reconstruct causal chains leading to particular occurrences).  Regarding P2, Fodor argues that whether a trait
increases fitness is massively context sensitive, and so it is highly unlikely
that there will be any laws concerning the fitness of traits as such (and thus
traits that are selected for as such). 

Now, Fodor has no objections to historical narratives, but
he thinks that if there are no nomologically necessary generalizations about
the mechanisms of adaptation as such, then natural selection reduces to a banal
truth: if a creature flourishes in a certain environment, then there must be
something about that creature or the environment (or both) in virtue of which it
does so.  Well, duh.

To which I am inclined to respond as follows: does anyone
really think that the adequacy of the theory of evolution of by natural
selection depends on there being laws of selection? At its bare-bones, the
theory says that when you have variation plus inheritance plus competition, you
are likely to get evolution.  This isn’t
a “law” of selection (as Fodor understands laws of selection), for it is
completely silent about which, when, and why traits are fitness enhancing.  To answer the latter questions, you actually
have to look at the details of particular cases.  But there is no reason to suspect (at the
outset of investigation) that any nomologically necessary generalizations about
traits will emerge, after having examined the cases.    Should
this generate Sturm and Drang in anyone, though?  Am I missing something?


  1. gualtiero

    Martin, thanks for your nice post. Your response to Fodor sounds right to me.  Now do I get to skip all the writings by Fodor on the subject or am I missing something?

  2. One worry if there are no “laws of selection” might be that it is an entirely contingent matter that things have evolved the way that they evolved. That is to say, it is an entirely contingent matter that there are humans or (for that matter) complex life forms and not simply a bunch of pre-amoebae.

  3. Allen

    So if deterministic physicalism is true, then evolution is just a descriptive narrative that we use as a “shorthand” description for the record of our past observations, right?

    In this case, the initial configuration of matter at the universe’s first instant, plus the causal laws that govern the subsequent behavior of this matter as applied over 13.7 billion years FULLY determines the current state of the universe today.

    There is nothing for evolution to do. It is purely a description of what we observe, not an explanation of it.

    There IS NO “competition”. There is no selection. Instead, events involving fundamental particles unfold as they must…in the only way that they can.

    When you say “competition among creatures”, what you really mean is “it is as though there were competition among creatures”. Because what really exists are fundamental particles (quantum fields, strings, whatever), not “creatures”. It is only in our minds that we take collections of quarks and electrons and form them into creatures.

    Evolution and natural selection have no causal power, we just speak of them as if they did.

    Further, even allowing for some kind of quantum randomness still doesn’t give “evolution” anything to do. Though it does muddy the water a bit.

  4. Taylor

    I think you have it right. An issue of Mind & Language was dedicated to Fodor so you may want to look at that if you are interested. From what I have seen so far, there are good responses to Fodors argument and Fodor has yet to explain why ther do not succeed (see the M&L issue). My guess is that Fodor either does not know what he is saying or he is purposely being obtuse and keeping his arguments a secret (assuming he does have responses to his critics).

  5. Malcolm Dean

    Yes, you are missing something.

    There is strong resistance to publishing scientific alternatives to Darwinism, because no serious alternatives are supposed to exist.

    Many papers which end with de rigeur conclusions allegedly supporting Natural Selection actually deal with alternative models, such as thermodynamics and Information.

    One alternative approach is from Stanley Salthe:

  6. Hm, well, Fodor has a pretty diluted notion of a “law” (special sciences do have laws, right?), so that any regularity, even ceteris paribus would count. And of course there are some regularities in natural selection, that’s why the eye evolved several times etc.
    If he replies that such laws aren’t counterfactual, then well, he’s using a stronger (and confused, IMHO) notion of a law than in his ‘Special Sciences’, so it seems a bit incoherent. (Maybe I’m simplifying this but it seems to me that the notion of a law being defined via counterfactuality is confused insofar as it relies only on our linguistic intuitions that do not fix the truth-value of conditionals at all – it is the regularities in the world that do, and it’s completely independent from the fact how you put it into words; moreover, it must presuppose the notion of projectibility, and you cannot define it in a purely linguistic way, as Zabludowski has shown in his brutal rebuttal of Goodman.)

Comments are closed.

Back to Top