Scientific Naturalism and the Argument from Reason

Print Friendly, PDF & Email

This article outlines the “argument from reason” against scientific naturalism and for theism.

1 What is Scientific Naturalism?

Scientific naturalism, or strict naturalism, is the view that the spatio-temporal universe of entities postulated by our best current (or ideal) sciences, particularly physics, is all there is (Moreland, 2012, p.284) This world-view aims to provide an event-causal, scientific account of how everything has come to be and a general ontology in which the only entities which exist are those which bear a relevant similarity to those which would characterise a completed form of physics (Moreland, 2012, pp.282-285).

If naturalism is true, all of reality can be accounted for by physics, cosmology, and the processes of evolution. The two core theories of scientific naturalism are the atomic theory of matter and evolutionary theory (Moreland, 2012, p. 285)  These scientific theories use combinatorial modes of explanation. Wholes at each level above the ground level of elementary microphysics are explained by the composition of the separable parts at lower levels. So living organisms are composed of cells; cells are composed of molecules, which in turn are composed of atoms, which are composed of subatomic parts, which are composed of microphysical entities.

Every physical event has physical causes which are sufficient to produce it; therefore, scientific naturalism subscribes to “the causal closure of the physical domain” (Goetz, 2011; Reppert,2003, p.52). Whenever we seek to explain an event or entity we need appeal only to physical laws and physical objects. This entails that we need never appeal to subjective consciousness, thoughts, plans, intentions or desires to explain the properties of any organism. The scientific naturalist can either attempt to reduce mental properties, like “intentions” and “beliefs”, to physical properties ( for example, the neuro-phsyiological states of organisms); or, the scientific naturalist might argue that mental properties supervene on, or are emergent from, physical properties. In this case mental properties would be determined by physical properties.

2 How does Naturalism account for Rationality?

Why would a naturalist believe that our cognitive mechanisms reliably produce true beliefs? Darwinian evolution offers an explanation. Organisms with mechanisms which reliably produce true beliefs are better adapted to their environments; therefore natural selection ‘favours’ these organisms. Garvey (2007) suggests that “it is not too controversial to state that it is something about the world that determines which beliefs are true. Different ways of expressing this might include ‘true beliefs describe the world accurately, ‘true beliefs correspond to the world and so on” (Garvey, 2007, p.181).

A trait is adaptive if it helps a creature which has it survive and reproduce given the way its environment is (Garvey, 2007, p.182). If an environment contains a dangerous predator, then traits which keep organisms away from that predator are adaptive. So traits which provide animals with true beliefs will be adaptive; of course, the trait must also give the animal an appropriate propositional attitude towards that belief. So, the belief “there is a dangerous predator in the vicinity” will not be adaptive if the animal also has a desire to play with dangerous predators.

Darwinian evolution also guarantees the reliability of our inductive practices. We regularly make generalisations because we cannot treat every situation as if it is entirely new. To make generalisations we must be able to group things into categories or classes. “Only if we recognise lions as forming a class are we able to have the useful belief ‘lions are not safe to approach’” (Garvey, 2007, p.182). But there are many different ways in which we could classify things. Our inductive practices will only be reliable if we have the ability to pick out the relevant similarities between things.

We could classify tigers with lions, as both are four legged predators; or we could classify tigers with zebras, as both are four legged animals with stripes. The prehistoric man who grouped the tiger with the lion would be more likely to survive than the man who grouped the tiger with the zebra. Darwin’s theory provides some encouragement that our innate subjective spacing of qualities ‘carves reality at the joints’ and makes our inductions come out right. If the spacing of qualities is a gene linked trait, then the spacings which made for the most successful inductions will have tended to predominate through natural selection (Garvey, 2007, p.183).

3.Plantinga’s Evolutionary Argument Against Naturalism

Alvin Plantinga has offered an “Evolutionary Argument Against Naturalism” (EAAN) which purports to show that scientific naturalism is self-defeating. If Plantinga’s argument is successful, it would not provide direct evidence for theism, but it would provide a reason for rejecting one prominent alternative to theism. Plantinga argues that if scientific naturalism is true, it is unlikely that humans have reliable cognitive faculties; this causes grave problems for scientific naturalism. The most recent statement of  EAAN is given in “Where the Conflict Really Lies” (Plantinga, 2011 pp. 307-350); It can be summarised as follows:

1)      It is improbable that human beings will have reliable cognitive faculties if both (a) scientific naturalism is true and (b) our cognitive faculties have been produced by natural selection acting on random genetic mutations scientific.

2)      Anyone who accepts that (a) naturalism is true and (b) that our cognitive faculties have been produced by evolution by natural selection has a “defeater” for the belief that humans have reliable cognitive faculties.

We can call the belief that humans have reliable cognitive faculties “R”. Call the belief that Naturalism is true “N” for short.Call the belief that our cognitive faculties have been produced by evolution by natural selection (“E” for short).

3)      Anyone who has a defeater for R has a defeater for any other belief she thinks she has, including the belief that scientific naturalism is true and the belief that the theory of evolution by natural selection is true.

4)      If one who accepts scientific naturalism and the theory of evolution (N&E) thereby acquires a defeater for those beliefs. One cannot rationally accept that both are true.

Conclusion: one cannot rationally accept that both the theory of evolution and scientific naturalism are true  (Plantinga, 2011, p.344-345).

Plantinga’s argument centres on our cognitive faculties, faculties or powers which produce beliefs. Examples of such faculties include memory, perception, our intuitions of a priori truths. A cognitive faculty is only reliable if the majority of the beliefs which it produces are true. However, Plantinga argues that the probability of our cognitive faculties being reliable, given naturalism and evolution, is low (Plantinga, 2011, p. 311-314).

In general, psychologists agree that we have an innate cognitive apparatus, which includes belief forming mechanisms and innate beliefs. Each cognitive mechanism, or module, has its own specific domain. The module which remembers and distinguishes faces is separate from the part of the mind which detects agency or purposiveness. Each cognitive mechanism was selected by evolution to deal with a specific problem or meet a specific kind of situation. (Garvey, 2007, p.212).

Now, Plantinga points out that if naturalism is true there is no teleological explanation for the development of our faculties; that is, the course of our evolution was not designed. (As we noted in section one, naturalism only allows for event-causal explanations and does not allow for agent-causal or purposive explanations). Additionally, natural selection will only promote adaptive behaviour, not true belief. Evolution can only select for behaviour which enhances chances of an organism’s genes being widely represented in the next and subsequent generations.  (Plantinga, 2004).

Our beliefs are “visible” to natural selection only insofar as they influence, or are influenced by, adaptive behaviour. So, given Darwinian evolution and scientific naturalism, will adaptive behaviour be reliably correlated with true belief? A belief, on scientific naturalism, will have two types of property: it will have neuro-physiological properties and mental content properties. The neuro-phsyiological properties will include a number of neurons, the connections between those neurons, difference potentials, rates of fire, and all the causal inputs and outputs. The content will be the belief that p, for some proposition p (Plantinga, 2011, p. 322).

What is the relationship between neuro-physiological properties and content properties? In section one, we noted that scientific naturalism subscribes to “the causal closure of the physical domain” because every physical event has physical causes which are sufficient to produce it. Given the principle of causal closure scientific naturalists  must either reduce mental properties to physical properties or they must  give a non-reductive analysis. According to the reductive strategy, mental properties are reducible to neuro-physiological properties; as Plantinga puts it “there is only one kind of property in the neighbourhood” (2001, p.323).

According to non-reductive scientific naturalism, content properties are not reducible to neuro-physiological properties. However, given the principle of causal closure, content properties are determined by neuro-physiological properties. Two organisms which are identical in their neuro -physiological properties will be identical in their content properties; furthermore, all of the organism’s content properties are caused by the organism’s physical properties and no neuro-physiological property is caused by the organism’s content properties.

Would true beliefs be correlated with adaptive behaviour given non-reductive scientific naturalism? Plantinga argues that this is improbable. Behaviour is caused by neurological structures in an organism’s nervous system; these neurological structures also determine belief content. But there is no reason to suppose that the belief content must be true. So long as the accompanying behaviour  is fitting for the organism’s environment there is no need for the organism’s beliefs to correspond to its environment. The organism’s mental properties do not cause any physical behaviour; therefore content properties are entirely irrelevant in determining how adapted an animal is to its environment.

Of course, organisms would have certain neurological structures which would be reliably correlated with certain features of the environment. However, Plantinga points out that a physical structure is not equivalent to a belief. Anaerobic marine bacteria contain magnetosomes which indicate magnetic north. This guides the bacteria towards oxygen-free regions of the ocean. However, no one would claim that these bacteria have beliefs with content. An animal might reliably form a particular neurological structure in the presence of a predator, especially if that structure promoted behaviour which helped the animal evade the predator. However, this structure is not a belief. Furthermore, whatever belief content was determined by the structure would be irrelevant so long as behaviour determined by the structure was adaptive.

So only the neuro-physiological properties of a belief determine the resulting behaviour. The neurological properties will be selected for if they are adaptive; it will not matter if the content of the belief is true or false. So what is the probability of the belief being true? Plantinga suggests that  we estimate the probability as 0.5. Given that the probability of any belief being true is 0.5 it is extremely improbable that our faculties will be reliable. Suppose a faculty must produce three times as many true beliefs as false beliefs to be considered reliable. If that faculty produces 1000 independent beliefs there is less than one in 1058 chance of that faculty being reliable (Plantinga, 2011, p.333).

If we come to believe that our cognitive faculties are unreliable we have a defeater for every other belief produced by our cognitive faculties. This would include the beliefs that humans are the result of evolution by natural selection and the belief that scientific naturalism is true.

3. What Exactly  is a Defeater?

What does it mean to have a defeater for another belief? “A defeater for a belief B I hold is another belief B* I come to hold which is such that, given I hold B*, I can no longer rationally hold B” (Plantinga, 2011, p340). There are at least two types of defeater: “rebutting” defeaters and “undercutting defeaters”. A rebutting defeater gives us some reason to think that a particular belief is false. An undercutting defeater is one that calls into question the trustworthiness of a belief’s source (Mirza, 2008, p.128).

To illustrate: suppose I watch Oliver Stone’s JFK, and come to believe that one sniper could not have been responsible for the shots that killed President Kennedy. There are two ways this belief could be defeated. I could study the testimony of numerous forensic scientists who have argued that one “shooter” fired all three shots on Dealey Plaza (Sabato, 2013; Shenon, 2013). This would be a rebutting defeater of the “second shooter” hypothesis: expert testimony gives me some reason to think that Stone’s film was false on this point.

However, suppose I read numerous reviews by historians, journalists and lawyers which claim that Stone was more interested in creating a myth, or paranoid fantasy, than reporting the facts about JFK; that Stone followed the theories of a largely discredited lawyer, Jim Garrison; and that Stone used ‘creative’ editing of the Zapruder film to mislead audiences. In that case, the film JFK can no longer be considered trustworthy. I have good reason to doubt most the claims in Stone’s film. In this case I have an undercutting defeater for my belief in a second shooter, because my belief had an untrustworthy source.

Omar Mirza points out that Plantinga’s EAAN deals better with objections to its second premise if we consider the “defeater” in EAAN to be an undercutting defeater. The evolutionary process responsible for creating human cognitive faculties did not filter out unreliable cognitive faculties: therefore we cannot trust the beliefs produced by our cognitive faculties; and one of the beliefs produced by our cognitive faculties, of course, is the belief that our cognitive faculties are reliable (Mirza, 2008, p.131-146).

So Plantinga’s argument is calling the very source of our confidence in R into question. We have no reason to have any confidence in the truth-tracking capabilities of our cognitive equipment. Our belief that our cognitive systems often tell us the truth is suspect if we gained that our belief through those same cognitive systems and if we have reason to doubt that those systems mostly produce true beliefs. It is does not matter if the sceptic amasses apparent evidence that our cognitive faculties are reliable (this could be “propositional” or “non-propositional” evidence). Given N&E, all our beliefs could be unreliable. This would include all our inductive inferences; we simply could not tell when we had judged the evidence correctly.

Some naturalists have pointed out that we do not gain a defeater for a belief ‘B’ merely by pointing out that the probability of ‘B’ is low given another belief ‘D’. The probability that I drive an Audi given (just) that I drive a car is low. The probability that the function of perspiration  is to cool the body (call this belief ‘C’) given (just) evolution and scientific naturalism is low. That does not mean that I have a defeater for my belief that sweat’s function is to cool the body. So we do not have a defeater for R simply because it is low given N&E.

However, if the defeater of R is an undercutting defeater there is no analogy between “the probability that sweat functions to cool the body given N&E” and “the probability that our cognitive faculties are reliable given N&E”. An undercutting defeater calls a belief’s source into question. The fact that C is low given N&E does not question the reliability of the source of belief C. Nor does the fact that I drive a car call into question the source of a belief that I drive an Audi.  But the low probability of our cognitive faculties being reliable (R) given N&E explicitly calls into question the source of our confidence in R (Mirza, 2008, pp.136-138).

4 Would Evolution by Natural Selection Produce Unreliable Cognitive Faculties?

Paul Draper (2007) has responded to Plantinga by arguing that “… it is very unlikely that belief-producing mechanisms that do not track the truth would systematically promote survival in a very diverse and often rapidly changing environment.” This assumes that belief can cause behaviour – something that scientific naturalism does not allow. However, Sober and Fistleton argue “[a]ssuming that beliefs don’t cause actions is not the same as assuming that they are wholly unrelated”(1998, p.122).  However, even if certain beliefs and certain actions are somehow related Schloss and Murray (2009) argue that we still would not have any reason to think that natural selection would filter out unreliable beliefs.

They consider adaptive misbeliefs, maladaptive misbeliefs and reproductively neutral misbeliefs. An “adaptive misbelief” is a belief which is not true but which is nevertheless adaptive; it is useful for the organism to believe this particular falsehood. Schloss and Murray argue that such beliefs are “extraordinarily plentiful, persistent and influential”. They point to our “promiscuous attribution of agency and teleology”, the placebo effect and unreasonable loyalty “to beloved people, places and nations”. Adaptive misbeliefs can persist even though there is substantial evidence that they are not true:

“Many of these widespread beliefs entail almost delusional denials of repeated experience. Notions that Eros lasts forever, this time it’s real and … ‘when we’re hungry love will keep us alive’ are effective and virtually ubiquitous catalysts for reproductive pair-bonding” (Schloss and Murray, 2009, p.533).

A maladaptive misbelief is a belief which is not true and which makes the organism less adapted to its environment. However, even though these misbeliefs carry a cost for the organism, there is no guarantee that natural selection will weed out every maladaptive misbelief. “In evolution, forgivable malfunctions may be common and achieving proper function may be positively rare” (Schloss and Murray, 2009, p.534).  Even if true beliefs are the best beliefs for organisms, design constraints and historical contingencies may prevent natural selection from finding the best solution; it might simply settle on a solution which enables organisms to survive and reproduce. As Fodor and Piatelli-Palmarini remind us:

Neo-Darwinists are keen to say that natural selection never optimises, it only finds locally satisfactory solutions. From Francois Jacob’s evolutionary ‘tinkering’ to Maynard Smith’s and Dennett’s ‘satisficing’, emphasis is always put on this consideration (Fodor and Palmarini, 2010, p.92)[1].

Reproductively neutral beliefs are beliefs which either do not influence reproductively salient behaviour or which do not influence any behaviour at all. Most theorising is reproductively neutral – for example, beliefs about the true nature of sub-atomic particles or metaphysical beliefs about the nature of causality. These beliefs remain invisible to natural selection because they do not have any adaptive consequences.

Finally, Schloss and Murray note that natural selection can only favour belief forming mechanisms which form representational models that motivate and orient organisms to adaptive behaviour. Are representational models which are true better at causing adaptive behaviour? This doesn’t seem very likely. They ask us to consider “…the task of designing “thinking” robots for a competition in which the winners were duplicated (with minor program variations) for future competitions (p.534).” There are numerous ways to program the robot to “conceptualise” its environment which would allow it navigate through its environment safely “which are representationally biased or even radically false.” Models need only “save the appearances” to be successful; and natural selection is fitness seeking, not truth seeking (Schoss and Murray, 2009, p.534).

5. Less Shocking, More Appealing Versions of EAAN

David Glass in “Atheism’s New Clothes (2012) points out that Plantinga’s conclusion is quite shocking, and that naturalists will strongly resist the conclusion that this version of atheism is self-defeating and will explore every possible avenue before accepting it. However, he points out that there is a less shocking version of Plantinga’s argument, which merely points out that scientific naturalism has difficulty explaining the reliability of our cognitive faculties:

“…suppose, as we do in practice, that our cognitive mechanisms are generally reliable in terms of giving rise to true beliefs. This reliability is what we would expect if there is a God since it is very plausible to believe that he would create us in such a way that our cognitive mechanisms would be reliable. But it is much less to be expected, for the reasons Plantinga proposes, if atheism is true. If so, then the reliability of these mechanisms provides evidence in support of belief in God. This fits in very well with the approach adopted so far in this book. It is not that particular pieces of evidence prove the existence of God or disprove atheism, but that they count in favour of God’s existence. The reliability of our cognitive mechanisms is just one more piece of evidence to support belief in God. Arguably, it is quite a strong piece of evidence since it is much more easily accounted for by theism than atheism (Glass, 2012)

Indeed, even Plantinga notes that there is a slightly different, more appealing, version of his argument which does not argue that all our cognitive faculties are unreliable. We can grant that those faculties which are relevant to survival and reproduction are reliable; however, those faculties which are not relevant to survival should be considered unreliable. Indeed, critics of the EAAN Sober and Fistleton admit that “It is perfectly possible that our mental machinery should work well on simple perceptual tasks, but be much less reliable when applied to theoretical matters.” (1998, p129)

So, instead of R, we could consider those faculties which produce our beliefs about theoretical and metaphysical matters. Theoretical and metaphysical rationality (‘TMR’ for short) would include large scale scientific theories, like quantum mechanics and general relativity. TMR would also include metaphysical schemes, like idealism, theism and naturalism. Our cognitive faculties were selected because they were useful to our distant ancestors, and the problems of string theorists are not the problems of stone aged hunters. Our faculties were not selected to form reliable beliefs about the sorts of unobservable forces, processes and entities that physicists appeal to.

In section two it was noted that our inductive practices will only be reliable if we have the ability to pick out the relevant similarities between things. But perhaps we can only allow that our faculties allow us to make accurate generalisations about the observable world (Garvey, 2007, p. 185-88).  Darwin’s theory provides little encouragement that we will make reliable inductions about subatomic particles or evolutionary processes which are, in principle, unobservable. It is also fairly clear that reliable metaphysical beliefs would have no clear survival value. Beliefs about the true nature of causality or time would have belonged to the reproductively neutral beliefs described in section four.

Now, scientific naturalism is a metaphysical scheme which claims that only the entities postulated by our best current (or ideal) sciences, particularly physics, exist. So Plantinga’s EAAN can be reformulated:

1) The probability that our theoretical and metaphysical beliefs are reliable (TMR) given N&E is low.

2) Anyone who accepts (believes) N&E and sees that the probability of TMR given N&E is low has a defeater for TMR.

3) Anyone who has a defeater for TMR has a defeater for any metaphysical scheme, including Naturalism, and any large scale scientific theory, including the theory of evolution.

4) So anyone who has a defeater for TMR has a defeater for N&E.

5) If one who accepts N&E thereby acquires a defeater for N&E, N&E is self-defeating can’t be rationally accepted.

Conclusion: N&E can’t be rationally accepted.

6. A Deeper Problem for Scientific Naturalism?

We have noted that scientific naturalism assumes that the physical world is causally closed: every physical event has another physical event as its sufficient cause. When we explain physical events (such as events in the brain or the human body) we should only appeal to physical forces, physical objects and the laws that govern them. We would be able to fully explain desires, beliefs and actions of humans merely by appealing to the operation of the brain’s physical parts.

We have also noted that, on scientific naturalism, psychological and conceptual content have no causal power at all; on a materialist ontology, a chain of physical cause and effects governs all our brain states and behaviour. Even if some physical events are somehow identical with mental events (and it is practically impossible to see how this could be so) it is the physical structure of the events in the brain that will determine what we report, how we act and what we believe (Hasker, 2011). This has an unusual consequence.

If the universe is causally closed, then our beliefs are not caused by mental properties or events. It may seem as if human minds consider the content of propositions and their logical connections when forming new beliefs. But, in fact, each belief is caused by an underlying brain state and its connections to other brain states. If brain states fully explain our beliefs and their connections, then there is no need for an accompanying explanation in terms of persons and their reasons. In fact “reasons” have no causal power at all. Human rational reflection is an illusion – merely the afterglow of neural processing.

Yet to be rational one subject must be able to form premises, see logical connections, make inferences and draw conclusions. The subject – or self – needs to be able to remember, consider and decide. A rational inference is not the output of a chain of non-rational events in a physical system. To make a good inference one subjective awareness must grasp the content of numerous beliefs and see the rational connections between them.  A calculator may be reliable, but it is not rational. Furthermore, some sort of responsibility for a proportion of our beliefs seems to be necessary for rationality. A deterministic system doesn’t meet the requirements for rationality. (See the detailed discussion in Reppert,2003).

Rationality demands that we are substantial selves that persist through time: there must be more to each person than their brains and its various outputs and inputs. Yet scientific naturalism cannot make room for a substantial self: for naturalism, scientific explanations act as a “universal acid”, driving other explanations from the field. As Kai Nielsen points out:

“Naturalism denies that there are any spiritual or supernatural realities. There are, that is, no purely mental substances and there are no supernatural realities transcendent to the world or at least we have no good ground for believing that there are such realities…It is the view that anything that exists  is ultimately composed of physical components” (quoted in Goetz and Taliaferro, 2008, p.9).

This being the case, there is no room in naturalism’s ontology for an immaterial self who can consider the content of beliefs, draw inferences and act on them. On scientific naturalism, then, it seems that no-one can be a rational agent.


Draper, P. (2007), “In Defence of Sensible Naturalism”, (retrieved 8th May 2014)

Fistelton, B and Sober, E. (1998) “Plantinga’s Probability Arguments Against Evolutionary Naturalism”, Pacific Philosophical Quarterly, 79: 115-129.

Fodor, J. and Piatelli-Palmarini M. (2010) What Darwin Got Wrong, London:; Profile Books

Garvey, B. (2007) Philosophy of Biology, Stocksfield: Acumen

Glass, D. (2012) Atheism’s New Clothes Leicester: Apollos

Goetz, S. (2011) “Making Things Happen: Souls in Action” in Goetz and Baker ed. The Soul Hypothesis New York: Continuum

Goetz, S. and Taliaferro, C. (2008) Naturalism Grand Rapids: Eerdmans

Hasker, W. (2011) “Souls Beastly and Human” in Goetz and Baker ed. The Soul Hypothesis New York: Continuum

Mirza, O. (2008) “A User’s Guide to the Evolutionary Argument Against Naturalism”, Philosophical Studies, 141: 125-126

Moreland, J.P. (2009) “The Argument from Consciousness”, in Moreland and Craig ed. The Blackwell Companion to Natural Theology, Malden: Wiley-Blackwell

Plantinga, A. (2004) “Naturalism Defeated” (retrieved, 18th May 2014)

Plantinga, A. (2011) Where the Conflict Really Lies, New York: Oxford University Press

Reppert, V. (2003) CS Lewis’s Dangerous Idea, Downers Grove: IVP

Sabato, R. (2013) The Kennedy Half Century New York: Bloomsbury

Schloss, J. and Murray, M. (2009) “You Can’t Always Get What You Want: Evolution and True Beliefs” Behavioural and Brain Sciences, 32:6, 533-534

Shenon, P. A Cruel and Shocking Act: The Secret History of the Kennedy Assassination London: Little, Brown

[1] However,  Fodor and Palmarini’s argument that Darwin’s theories have no explanatory power does not seem very plausible.

This entry was posted in Existence of God, Science and Christianity and tagged , , , , , , , , , , , . Bookmark the permalink.