Thursday, October 3, 2013

The Evolutionary Argument Against Naturalism: A Response to Stephen Law

In my recent review of Alvin Plantinga's book Where the Conflict Really Lies, I sketched his Evolutionary Argument Against Naturalism (EAAN) and briefly considered a reply to the argument by Stephen Law.  Stephen Law published his reply to the EAAN called "Naturalism, Evolution, and True Belief" in Analysis in January 2012.  It is available on his blog here.  In this essay, I would like to examine Law's response in more detail and show why I think it fails to refute the EAAN.

The Evolutionary Argument Against Naturalism (EAAN)

I will begin by summarizing the EAAN (those already familiar with the argument may skip to the next section).  The EAAN attempts to establish an internal inconsistency between belief in naturalism on the one hand, and belief in the reliability of human cognitive (or belief-forming) faculties on the other.  Specifically, the argument suggests that naturalism--which supposes humans are thoroughly material beings who came to be by a purposeless, unguided process of Darwinian evolution--provides no mechanism by which human belief-forming faculties could be calibrated to track "true" propositions.  This is so for the following two reasons.

First, under naturalism, the only causally efficacious entities in existence are material forces and the material objects that absorb and transmit those forces.  Since the contents of beliefs (e.g. propositions) are not material forces nor material objects, the contents of beliefs do not cause physical effects.  If Smith's hand, for example, scribbles the words "seven is a prime number" on a piece of paper, it is nowhere permitted in the naturalist's account of Smith's behavior to add "...and then the proposition that seven is a prime number caused Smith's hand to write 'seven is a prime number' on the paper."  Under naturalism, to the extent that propositions exist, they are not the sorts of entities that cause physical objects to change their positions, momentums, and the like.  Thus, under naturalism, propositions cannot be part of any physical account of cause and effect.

Second, naturalism prohibits the existence of any immaterial substance that could causally interact with the physical world.  Specifically, naturalism rules out versions of substance dualism--the position that immaterial minds exist independently of, but causally interact with, the physical world.  Therefore, under naturalism, Smith's hand does not write "seven is a prime number" because Smith's immaterial mind apprehends the proposition "seven is a prime number," believes this proposition to be true, and then causally affects Smith's body to write the proposition on paper.  To the extent naturalism permits something called a "mind" to exist, it is either simply another name for a regular material object (like the brain), or it is a causally impotent abstraction--a word with no concrete referent.  If the mind is just the same thing as the brain, then the mind can only do what the brain does.  And, since the brain is a physical object, it can only do what physical objects do.  Since propositions are not physical objects or forces, propositions do not interact with the physical states of the brain.  Thus, the proposition "seven is a prime number" does not cause Smith's brain to do anything, and therefore plays no role in explaining why he wrote the proposition on paper.  If the mind is a causally impotent abstraction, then obviously it still plays no role in explaining Smith's behavior.  Thus, under naturalism, propositions cannot enter the causal story of Smith's behavior by virtue of some special properties of Smith's mind, like the special powers of mental causation that are available to substance dualists.

So, under naturalism, neither the proposition "seven is a prime number" itself caused Smith's hand to scribble the proposition on paper, nor did Smith's immaterial, causally efficacious mind (because such a thing does not exist) cause his hand to scribble the proposition on paper.  Given naturalism's commitments to materialist explanations of events, it seems naturalism does not have any other resources available to explain how the proposition "seven is a prime number" could be linked to Smith's hand's writing "seven is a prime number" on paper.  Furthermore, the same could be said of any proposition Smith or anyone else might believe and the actions any of them might undertake.

Now, here's the thrust of the EAAN with respect to this point: If naturalism is true, and humans came to be the way they are by a Darwinian version of evolution, then there is no physically efficacious link between one's beliefs and one's behaviors.  Evolution sifts between adaptive behaviors and maladaptive behaviors only; evolution has no impact on the realm of abstract concepts, ideas, and propositions.  Thus, evolution's circumscribed access to physical behaviors alone renders it incapable of calibrating human belief-forming capacities toward true propositions.  But if evolution did not calibrate human belief-forming capacities toward true propositions, if naturalism prohibits anything like a God or designer from calibrating human belief-forming capacities toward true propositions, and if naturalism denies the existence of immaterial rational minds that can apprehend true propositions and cause bodies to behave by virtue of those propositions, then what grounds are left for naturalism to assert that human belief-forming capacities reliably detect true propositions?  The EAAN claims there is nothing left under naturalism that could plausibly calibrate human belief-forming capacities toward tracking true propositions, and thus the EAAN concludes that, under naturalism, human belief-forming capacities unreliably produce true beliefs.

Finally, having attempted to establish that naturalism entails human belief-forming capacities unreliably produce true beliefs, the EAAN argues that anyone who believes naturalism is true simultaneously accepts that his own belief-forming capacities are unreliable.  This of course then undermines or defeats his original belief that naturalism is true, since that belief was generated by the very belief-forming faculties that naturalism makes unreliable.  In this way, the EAAN argues that beliefs in both naturalism and the veracity of one's own belief-forming faculties are rationally incompatible--one cannot sensibly believe them both.

Stephen Law's Rebuttal

Stephen Law's rebuttal challenges the EAAN's claim that behaviors are entirely independent from the propositional content of beliefs.  Law summarizes his position this way:
"I argue that, even in its most recent incarnation, the EAAN fails. In particular, Plantinga overlooks the fact that adherents of naturalism may hold, seemingly quite plausibly, that there exist certain conceptual links between belief content and behaviour. Given conceptual links of the sort I envisage, natural selection will indeed favour true belief."
Law postulates that belief contents (e.g. propositions) and behaviors are conceptually linked in some way.  What he means by this is not altogether clear to me, but my best guess is the following.  Law appears to suggest that beliefs and behaviors are both caused by the same neurological substructures (e.g. brain, spinal cord, efferent nerves, and so forth--the anatomical details do not matter here).  This is to say, whatever neurological structures are responsible for causing Smith's hand to write "seven is a prime number" on the paper, the congeries of physical properties associated with those neurological structures also produces a unique kind of belief.  Law clearly asserts, "one cannot plug any old belief content into any old neural structure"; rather, when a certain neural structure is fixed, there are conceptual limits or constraints on what sorts of beliefs that neural structure will produce.  On this point, Plantinga agrees.  Plantinga also accepts that unique neural structures will produce unique beliefs, given the prior commitments of naturalism.  Where Plantinga and Law disagree, though, is whether the unique belief determined by the neurological structure has anything conceptually to do with the nature of the behavior, let alone whether it has anything conceptually to do with true propositions.  Law seems to argue it is more probable than not that a belief generated by a neurological structure is conceptually about the same subject that the behavior is conceptually about.  For example, since Smith's hand wrote a proposition about prime numbers, Law would suggest that whatever belief the neurological structure directing the motions of Smith's hand produces, it will more likely be about prime numbers than anything else.  So, my best assessment of Law's argument is Law suggests that the neurological structures that cause both behaviors and beliefs are linked by virtue of the "conceptual aboutness" associated with the behavior produced by the neural structure.  Law offers the following analogy to illustrate:
"Suppose that, solely in combination with a very strong desire for water, a certain belief/neural structure typically results in a subject walking five miles to the south....We can know a priori, solely on the basis of conceptual reflection, that, ceteris paribus [everything else equal], the fact that a belief/neural structure causes that behaviour in that situation significantly raises the probability that it has the content there’s water five miles south. Among the various candidates for being the semantic content of the belief/neural structure in question, the content that there’s water five miles south will rank fairly high on the list."
So Law suggests that, supposing one's neurology is such that the neurological structures that cause thirst for water also cause movement in a southerly direction, the same neurological structure causing southward movement also more likely than not produces the belief that "there's water south of me," or "there's water thataway," or "down there water me drink."  In other words, since the behavior is conceptually about movement toward water, Law argues that whatever belief is also caused by the same water-seeking neurological structure should more likely than not also be conceptually about movement toward water.

It might go without mention that if the sorts of conceptual constraints between beliefs and behaviors Law highlights exist, then natural selection certainly could select for true beliefs.  Following his example, out of those who were evolutionarily situated north of fresh water, those whose neural structures produced southerly movement would have survived while those whose produced movement elsewhere or no movement at all would have died.  Since Law posits that beliefs are linked to the conceptual aboutness of behavior, those who moved south and survived would also have believed that water was situated south of them--a true belief, as it were. 

My Reply to Law

Those who read Law's full rebuttal to Plantinga's version of the EAAN might recognize, as I did, that Law falls short of flushing out the biological nature of the conceptual links he suggests exist between beliefs and behaviors.  He rather simply posits their existence.  He does not tell us much about the way neurological structures produce beliefs or even why neurological structures produce beliefs at all.  So it certainly is not obvious why a reader should simply accept that, under naturalism, beliefs are conceptually linked to behaviors. Interestingly, Law softens the potency of his argument in his conclusion.  He writes:
"Of course, I am merely making a suggestion. Perhaps there exist no such conceptual constraints on belief content of the sort I envisage" (my emphasis).
I wish to note that "merely" conjuring a suggestion in response to Plantinga's argument is by no means a "refutation" of that argument.  A refutation involves demonstrating that an argument is false.  A mere suggestion is not a demonstration.  Law, nevertheless, concludes his piece as though he has offered more than a mere suggestion: " things stand, it is not naturalism that is defeated, but the EAAN."

As I see it, Law could only defeat the EAAN with his strategy if he could demonstrate how and why these conceptual links between beliefs and behaviors would exist in a world where naturalism were true.  This task is by no means simplistic; it appears to involve taking a mechanism that is calibrated for one function and superimposing onto that mechanism a secondary function qualitatively distinct from the original function.

Specifically, following Law's thirsty man, it involves constructing a neurological-muscular mechanism that can generate the function of mobility--a mechanism that can walk to the south.  Now, in addition to walking to the south, Law would have to show how and why that same mechanism would also perform the secondary function of sustaining not only a belief, but a belief with content to the effect "there's water south of me."  This seems analogous to showing that a thermometer not only performs the primary function of measuring temperature but also--by the same mechanism--performs the secondary function of measuring, for instance, wind speed.  It would indeed be surprising, given naturalism, to find that the mechanism enabling thermometers to measure temperature in degrees Centigrade also enables them to measure the distinct phenomena of wind speed in kilometers per hour.  Such a state of affairs would cry out for an explanation.  One would not just accept it because, let's suppose, it aids in the success of naturalist metaphysics.  If my analogy is remotely successful, it is plausible to suppose that human behaviors and beliefs are as distinct from each other, given naturalism, as temperature and wind speed are from each other.  Again, why should a mechanism that causes movement to the south also cause a subjective experience with the content "there's water south of me"?  Physical movement and subjective experience are arguably more radically unlike each other than temperature and wind speed.  It should follow, then, that if one is not prepared to accept without demonstration that, in a naturalist world, thermometers perform both the function of measuring temperature and the function of measuring wind speed by the same mechanism, then one ought not to blithely accept without demonstration that neurological structures, in a naturalist world, perform both the function of causing behavior and the function of causing true beliefs by the same mechanism.  It certainly seems more parsimonious, given naturalist metaphysics, to simply assign one function to one mechanism and leave it there.  Certain neurological structures might cause movement and that's it; other neurological structures might cause beliefs and that's it.  The two structures may be housed within one brain, the way multiple functions can be housed within one electronic device, but the mechanisms that produce each unique function are still distinct from each other.1

The only defense Law provides of his claim that beliefs and behaviors are conceptually linked arises in his conclusion.  He writes:
"Still, the view that there are such constraints on content is widespread (it is by no means restricted to those wedded to some form of logical behaviourism or functionalism, for example). It seems intuitively obvious to many of us that belief content is not entirely conceptually independent of behavioural output: that one cannot plug any old belief content into any old neural structure (or soul-stuff structure, or whatever) entirely independently of its behavioural output. That intuition would appear to be, philosophically speaking, largely pre-theoretical. It cannot easily be dismissed by Plantinga as a product of some prior theoretical bias towards naturalism and/or materialism" (my emphasis).
Law's primary support for his claim that beliefs and behaviors are conceptually linked turns out to not rely on any physics, chemistry, or biology, but rather on an obvious intuition.  Granted, I agree with Law that the intuition is manifestly obvious--I, for one, will accept that claim without argument.  However, recording the observation that beliefs and behaviors are obviously connected in some way does nothing to defend naturalism against the EAAN.  Ironically, this intuition serves to strengthen an argument against naturalism!

Why do I say this?  Let me put the matter a bit more formally to illustrate.  First, recall that for any proposition P and any proposition Q, the following versions of conditional reasoning hold:

Modus Ponens
1. If P is true, then Q is true
2. P is true
3. Therefore Q is true

Modus Tollens
1. If P is true, then Q is true
2. Q is false
3. Therefore, P is false

Two examples, one of each:

Modus Ponens
1. If it is raining outside, then there are clouds in the sky
2. It is raining outside
3. Therefore, there are clouds in the sky


Modus Tollens
1. If 10 is a prime number, then 10 is only divisible by itself and 1
2. 10 is not only divisible by itself and 1
3. Therefore, 10 is not a prime number

Now, I will construct two arguments involving naturalism's relationship to human beliefs and behaviors, each sharing an identical premise (1), but one claiming that naturalism is false, while the other claims naturalism is true.

Argument A: Naturalism is False
1. If naturalism is true, then human beliefs are generated conceptually independently of human behaviors
2a. Human beliefs are not generated conceptually independently of human behaviors
3a. Therefore, naturalism is false

Argument B: Naturalism is True
1. If naturalism is true, then human beliefs are generated conceptually independently of human behaviors
2b. Naturalism is true
3b. Therefore, human beliefs are generated conceptually independently of human behaviors

Argument (A) is a version of modus tollens, while argument (B) is a version of modus ponens; both arguments, therefore, are logically valid. Within both arguments, you'll see that premise (1) is identical: This is a premise subsumed within Plantinga's EAAN argument.  My intention is to illustrate what happens when one shifts the conclusion of an argument to a premise, and vice versa.  I understand Plantinga's EAAN argument to implicitly adopt the form of argument (B): Plantinga intends to grant, for the sake of argument, that naturalism is true in order to demonstrate how it leads to self-defeat.  As such, the truth of naturalism is a premise in that argument rather than the conclusion.

Now, notice that premise (2a) is precisely Law's intuition: human beliefs are not generated conceptually independently of human behaviors.  If we insert that intuition into an argument containing the same premise as the one implicit in Plantinga's EAAN argument, we should end up with argument (A) above--an argument affirming the falsity of naturalism!  Surely, I do not think Law intends to do that.

But, what place does Law's intuition have within argument (B)?  All it seems to do is deny the conclusion.  Denying the conclusion, though (aside from begging the question), says nothing about the first two premises.  Perhaps it implies that at least one of them is false, but which one and why?  We'll grant that Law will not reject premise (2b), even though he does not self-identify as a naturalist.  That then only leaves premise (1).  But, to reject premise (1) would require some sort of argument--specifically, some sort of account of how naturalism is compatible with the conceptual linkage between human beliefs and behaviors.  The mere observation--however correct it may be--that, in our world, behaviors are obviously tied to beliefs fails to place that observation within, and because of, a naturalist world.  Without specifically placing the observation in a naturalist world, the observation is simply compatible with any kind of world--including Plantinga's theistic world.  Such a maneuver would basically turn the connection between behaviors and beliefs into a necessary truth in all possible worlds.  In other words, it would make the conceptual linkage between beliefs and behaviors an inexplicable "brute fact."  If this is the case, then it is difficult to see how this is good for naturalism.  Naturalists appear to take pride in their ability to explain the natural world by way of a relatively small set of forces and particles.  The greater the size of the naturalist's "bag of brute facts," the less he or she can explain by way of forces and particles.  So, as best as I can make out, short of situating the conceptual linkage between beliefs and behaviors within, and because of, a naturalist world, naturalism is weakened by the obvious existence of such a linkage: if the linkage is, as Law contends, "pre-theoretical," then it is an inexplicable brute fact; if the linkage is not a brute fact but rather a product of a naturalist world, then it has yet to be explained.  Either way, naturalism is uncomfortably short on explanations--too short to refute the EAAN.


In sum, I find Law's reply to the EAAN incomplete at best and fatal to naturalism at worst.  He attempts to sidestep the EAAN's gambit by inserting special conceptual links between beliefs and behaviors into the naturalist world without regard for how the naturalist world could produce them.  He rightly observes that in our world these conceptual links are obvious; but, rather than affirm that our world is a naturalist world, such an observation could a) equally (if not more so) affirm that our world is not a naturalist world, or b) put the observation beyond the purview of naturalist explanation in principle.  These options neither decisively refute the EAAN nor make naturalism a more adequate metaphysical description of our world than it was before.  As it stands, then, naturalists only stand to lose some ground while in the thrall of the EAAN.

1. Just in case some readers find the analogy with temperature and wind speed wide of the mark, they can replace "wind speed" with "belief in the temperature reading."  The analogy would then suggest that the mechanism that permits a thermometer to measure temperature also--again by the same mechanism--permits the thermometer to believe that the temperature is whatever the mechanism says it is.  Thus, in addition to indicating "20 degrees Centigrade," the exact same mechanism would also produce the belief "the temperature is 20 degrees Centigrade."  I maintain that this state of affairs is at least as absurd as the thermometer measuring both temperature and wind speed--in fact, I think it is more absurd because now thermometers have subjective states of consciousness.


  1. As someone who would certainly be labeled a “naturalist” by Ashton and Plantinga, I suppose I’d better begin this response with a quick indication of my ontological position. I reject Cartesian dualism of the sort promoted by the aforementioned thinkers; setting aside its religious associations, I find it metaphysically superfluous, postulating as it does a sort of parallel universe whose principal occupants, minds, interact with physical reality (brains particularly) in a systematically obscure way. As scientists come to understand the brain’s operation in ever greater detail and begin to emulate its behavior with computers, the unilluminating vagueness and inescapable mysteriousness of dualism become ever less compelling or defensible. That said, I will nevertheless add that I am willing to countenance an ontology that includes certain types of abstract, non-physical entities: to logical constructs like numbers I would certainly concede a sort of “existence,” and likewise for properties of physical objects, which come in a variety of types with varying complexity and subjectivity (“color” being a relatively simple example, and “honesty” a much thornier one). “Propositions” are another sort of abstraction that indubitably perform valuable service when we describe and contemplate states and events in the physical universe, in particular the behaviors and decisions of intelligent organisms. In sum, Platonism in some form or other seems much less disreputable to me, much more useful in coming rigorously to grips with reality, than dualism.
    A “belief” can be viewed as either a Cartesian entity, another of the mysterious inhabitants of the domain where minds live, or as a sort of property, a way of describing the dispositions and (perhaps hypothetical) actions of agents whose lives we are observing. If we employ the latter approach — the only scientifically respectable option to my mind — then a statement of the form “Jones believes that there is beer in the fridge” has to be analyzed in terms of Jones’s likely behaviors, both non-verbal and verbal. The links between the statement and the possibly relevant behaviors are certainly complex, but humans in everyday life are undeniably adept at discerning them, at adducing evidence, and at coming to near-unanimous consensuses on whether, say, Jones really believes the proposition regarding the beer. If we wish to move to the next stage and figure out the physical underpinnings of Jones’s beer-related actions and tendencies, we will be obliged to enter the labyrinth of his brain, unquestionably a fearsome undertaking, dwarfing in difficulty the debugging of the most intricate computer program in use today. To explain fully Jones’s particular belief, we might well have to describe in detail the configurations of thousands or millions of neurons, along with their ability (in conjunction with other beliefs and desires) to influence his behavior in conjectured situations; but at this point there is no compelling reason to consider such a task impossible in principle. I find the complexity of the underlying science completely unsurprising and cannot see anything in it that would discredit the approach in any way. On the contrary, what I would consider discreditable would be someone who hoped to bypass the science by invoking vague and ill-specified interactions between Jones and some shadowy Cartesian-style beliefs lurking off in his incorporeal mind.


  2. Hi CPT,

    Thanks for your thoughtful replies both here and on my Amazon review of Where the Conflict Really Lies (for those interested, you can find that review here:

    Your major point here seems to be that there is no such thing as a "belief" apart from "behavior." You write, "For the naturalist, the link between belief and behavior exists by definition; ascribing a belief to Jones is nothing more than describing, in an abstract but useful way, certain types of behavioral dispositions he possesses."

    You then use this definition of belief to challenge this conditional from the EAAN: (1) If naturalism is true, then human beliefs are generated conceptually independently of human behaviors. Your challenge is that it is impossible for beliefs to be conceptually independent from behaviors because beliefs just are behaviors. So, on your view, if naturalism is true, beliefs are identical to behaviors, and premise (1) is then false.

    My first reply is to point out, as you did, that Stephen Law himself does not share your definition of belief (and, obviously, neither do I nor Plantinga). Insofar as my essay is in reply to Law in the context of Plantinga's EAAN, it seems appropriate to take their understandings of the term belief in the discussion.

    My second reply is that, insofar as your revised definition of belief nullifies the force of the EAAN, it equally nullifies the force of any propositional claim about what's true or false, including propositions like "dualism is false," "naturalism is true," or "beliefs are behaviors." I say this is so because behaviors do not come about by virtue of true propositions. Put another way, behaviors do not exist in a direct cause-effect relationship with true propositions. It is not the case that true propositions are causes and behaviors are effects. For example, the proposition "there's beer in the fridge" (supposing it is true) does not in any way cause Jones to both announce "I'm getting a beer" and walk to the fridge. The proposition "there's beer in the fridge" is neither a physical force nor a physical object; so, under physicalism, that proposition cannot cause anything. This means that whatever behavior Jones exhibits, it does not bear any relation to the truth of the proposition "there's beer in the fridge." If this is the case, then it would also be the case that no propositional truth whatever has any effect on behavior, which means no behavior has any connection to the truth of propositions. Since, on your account, beliefs and behaviors are identical, it follows that no belief has any connection to the truth of propositions either. As a result, propositions just have nothing to do with anything in human life. This includes propositions like "dualism is false," "naturalism is true," and "the EAAN is a bad argument."


  3. Continued....

    For instance, let's say you behave as if naturalism is true and I behave as if naturalism is false. Since the truth of naturalism does not cause anything, including both of our behaviors, then neither of our behaviors reliably say anything about whether naturalism is true. Naturalism might be true and it might be false, but its status is in no way indicated or reliably captured by our behaviors. It's like we're both blindfolded, thrown into a helicopter, randomly dropped in the middle of the ocean, then told to "swim north." You go off in one direction, I go in another, and just given that information, we try to sort out which one of us is actually going north. Maybe one of us is going north and maybe neither of us is going north, but we won't learn anything about which is which merely by our behaviors alone.

    Now, this approach of yours does accomplish the goal of nullifying the force of the EAAN. However, in so doing, your approach also undermines the force of other claims you wish to make, like the claim that "dualism is false." The situation brings to mind the following analogy.

    Suppose Bob is unfortunately afflicted by a virus that paralyzes him from the waste down, impairing the use of his legs. There is no known medical cure for this. Bob, of course, would like to regain the use of his legs. I'll call regaining the use of his legs his "ultimate" goal. Bob also realizes that, to achieve his ultimate goal, he'll have to get rid of the virus. I'll call getting rid of the virus his "intermediate" goal.

    Now, suppose Bob's doctor says to him, "I know how to get rid of the virus! I'll just amputate both of your legs. Problem solved." If Bob is shortsighted, he might like this idea because it satisfies his intermediate goal of getting rid of the virus. But, if Bob thinks a bit more long-term, he realizes that by satisfying his intermediate goal in this way, he permanently frustrates the satisfaction of his ultimate goal of regaining the use of his legs.

    I claim that you stand in a similar situation to Bob. Your intermediate goal is nullifying the EAAN. But your ultimate goal is validating naturalism or refuting dualism. Your strategy of equating beliefs with behaviors satisfies the intermediate goal at the expense of the ultimate goal. I've met many naturalists who accept this gambit and pronounce that "no one knows anything about anything," including themselves. Whether you are one of those naturalists is up to you, but I don't see how naturalism is worth that price.