Skip to main content

John M. DePoe points out that Alvin Plantinga’s evolutionary argument against naturalism is often considered to follow similarly to an argument given by C. S. Lewis. However, in this essay he suggests there are significant differences in their arguments against naturalism, which his analysis of their arguments emphasizes. The most significant contrast is the standard for rational thought that is used in each argument. By examining these differences DePoe raises some criticisms of Plantinga’s argument and suggest some reasons to prefer Lewis’s version. In particular he maintains that the account of rationality in Plantinga’s argument makes Lewis’s argument the superior way to argue against naturalism. Mr. DePoe is Assistant Professor of Philosophy at Marywood University.

One of Alvin Plantinga’s many important contributions to philosophy is his evolutionary argument against naturalism (EAAN).1 It has been suggested that Plantinga’s argument shares a kinship with C. S. Lewis’s argument against naturalism contained in chapter three of Miracles that is now commonly referred to as the argument from reason (AFR).2 Victor Reppert, a philosopher with expertise on C. S. Lewis’s AFR, has noted that Plantinga’s EAAN bears a striking similarity to Lewis’s case against naturalism.3 James Beilby claims that there is an affinity between the arguments given by Plantinga and Lewis.4 Plantinga himself declares that Lewis’s argument is an ancestor to his own.5 While there are some commonalities between Plantinga’s EAAN and Lewis’s AFR, my contention is that there are important, fundamental differences in these two arguments. Moreover, by identifying these noteworthy differences between the arguments given by Plantinga and Lewis, I believe that some significant problems with Plantinga’s EAAN will be evident.


Let us begin where Plantinga’s EAAN and Lewis’s AFR concur, namely on what they intend to refute, naturalism. Lewis defines naturalism as the belief that the world is nothing more than the events in space and time that operate on their own accord.6 Naturalism is the belief that the natural world is the “total picture,” there is nothing outside, above, or beyond the events in nature. Lewis writes that naturalists believe nature is the “whole show”; there is nothing that is independent of nature.7 Although Plantinga is somewhat coy in his attempts to define naturalism, he regularly cites Richard Dawkins and Daniel Dennett as paradigmatic cases of naturalism.8 Dawkins believes that nature through the purposeless forces of evolution is the “blind watchmaker,”9 and Dennett maintains that Darwin’s “dangerous idea” is that undirected evolutionary processes can produce order from chaos without the planning or providence of a Designer or Mind.10 The consistent image of naturalism that I take from these examples (and other things Plantinga writes about naturalism) is the understanding that naturalism is the metaphysical position that reality’s fundamental constituent parts operate according to non-purposive, undesigned causes with no intended goal or direction, and that nothing exists independently of these basic parts.11 While a keener eye may spot some trifling differences I may have overlooked, there is overall agreement between Plantinga and Lewis on the target their arguments intend to refute; naturalism is the belief that nothing exists independently of purposeless natural laws, processes, and things.

It is worth noting that for both Lewis and Plantinga, naturalism is not the same as atheism, agnosticism, or non-theism. It is possible for someone to reject both theism and naturalism. For example, one could be an atheist who believes that some things exist beyond the interlocking system of natural laws, processes, and things, such as an atheist who is a substance dualist or a moral realist. Lewis even considers the possibility that one could adhere to naturalism and believe that there is a “God” (just so long as “God” is subject to the system of nature; “God” would have to be inside the natural system, not beyond it).12 The target of their arguments, philosophical naturalism, should not be confused with mere atheism. Furthermore, naturalism is broader than an austere materialism; materialism is one kind of naturalism, but one could be a naturalist without being a materialist. Lewis and Plantinga agree that an essential component to naturalism is the commitment to a total, systematic conception of the world being driven most fundamentally by blind, undirected, impersonal causes.

Before moving on to the substance of their arguments, there is a first difference we can note between Lewis and Plantinga. Plantinga’s EAAN is not aimed at naturalism simpliciter. The EAAN is an argument against the conjunction of both naturalism and evolution (understood at least in part as the theory of common descent or common ancestry).13 Plantinga’s argument, thus, is not a general argument against naturalism. Rather, Plantinga’s argument is directed against a naturalist of a specific stripe, namely the kind of naturalist that believes in evolution. On the other hand, Lewis’s AFR is an argument against any kind of naturalism.

Two anonymous referees for this journal have urged me to consider whether it is worthwhile for me to make so much fuss over the fact that Plantinga’s argument is not aimed at a pure naturalism but rather at the conjunction of naturalism and evolution. After all, naturalism has been on the rise almost entirely as a result of the advancement of Darwinism. As a matter of fact, most naturalists today cannot plausibly deny evolution and keep their naturalism. It may seem like I am raising a trivial and insignificant weakness against Plantinga’s argument. In response to these concerns I have two brief rejoinders. In the first place, there have been and currently are some naturalists who are not strongly committed to Darwinian evolution as a basis for their naturalism.14 Historically, naturalism has been advanced without Darwinism. It presently has some supporters who are skeptical of current evolutionary theory, and I suspect it will continue to have supporters who do not rely on Darwinism. Plantinga’s argument will have no impact on these sorts of naturalists. Second, if the mainstream scientific community eventually replaces Darwinian evolution with some other account of the diversity of biological life, then Plantinga’s argument will likely cease to be a serious challenge to naturalism. While it may seem like the bond between naturalism and Darwinism is inseparable for the moment, it is by no means an essential union. For these reasons, I believe it is important to highlight that Plantinga’s argument is not against a pure form of naturalism.

Plantinga’s EAAN

Plantinga’s most recent formulation of the EAAN is given in the last chapter of his book, Where the Conflict Really Lies. To get a taste of the kind of argument Plantinga is proposing, I will quote him at length where he states the basic gist of the argument:

First, the probability of our cognitive faculties being reliable, given naturalism and evolution, is low. (To put it a bit inaccurately but suggestively, if naturalism and evolution were both true, our cognitive faculties would very likely not be reliable.) But then according to the second premise of my argument, if I believe both naturalism and evolution, I have a defeater for my intuitive assumption that my cognitive faculties are reliable. If I have a defeater for that belief, however, then I have a defeater for any belief I take to be produced by my cognitive faculties. That means that I have a defeater for my belief that naturalism and evolution are true. So my belief that naturalism and evolution are true gives me a defeater for that very belief; that belief shoots itself in the foot and is self-referentially incoherent; therefore I cannot rationally accept it.15

He then proceeds to give a more formal and precise statement of the argument. Where “R” stands for the proposition that our cognitive faculties are reliable, “N” for the proposition that naturalism is true, and “E” for the proposition that our cognitive faculties have come to be in the way proposed by contemporary scientific theory of evolution, it has the following structure:

(1) P(R|N&E) is low.
(2) Anyone who accepts (believes) N&E and sees that P(R|N&E) is low has a defeater for R.
(3) Anyone who has a defeater for R has a defeater for any other belief she thinks she has, including N&E itself.
(4) If one who accepts N&E thereby acquires a defeater for N&E, N&E is self-defeating and cannot rationally be accepted.
(5) N&E cannot rationally be accepted.
The first premise is often called “Darwin’s Doubt” because Plantinga introduces it through a provocative confession found in one of Charles Darwin’s letters:

With me the horrid doubt always arises whether the convictions of man’s mind, which has been developed from the mind of the lower animals, are of any value or at all trustworthy. Would any one trust in the convictions of a monkey’s mind, if there are any convictions in such a mind?16

The concern voiced by Darwin stems from realizing that evolution “selects” traits that promote behavior suited for survival or reproductive success, not necessarily for forming true beliefs reliably. If the undirected hand of evolution is responsible for the cognitive faculties we possess, then it is not likely that we are going to have cognitive faculties that produce true beliefs reliably. Plantinga supports this idea with a quote from Patricia Churchland, a prominent naturalist philosopher:

Boiled down to essentials, a nervous system enables the organism to succeed in the four F’s: feeding, fleeing, fighting and reproducing. The principle chore of nervous systems is to get the body parts where they should be in order that the organism may survive. … Improvements in sensorimotor control confer an evolutionary advantage: a fancier style of representing is advantageous so long as it is geared to the organism’s way of life and enhances the organism’s chances of survival. Truth, whatever that is, definitely takes the hindmost.17

Plantinga develops his first premise by considering how the content of one’s cognitive faculties would be related to survival on reductive and nonreductive materialism.18 For my purposes, I can skip over these details and take for granted Plantinga’s underlying point, namely that given naturalism and evolution, the probability that our cognitive faculties produce true beliefs reliably is very low on either reductive or nonreductive materialism. The salient idea is that the problem with one’s cognitive faculties given naturalism and evolution, according to Plantinga, is that they are very likely to be unreliable—they are probably going to produce more false beliefs than true ones.

Many naturalists oppose Plantinga’s first premise and argue that Darwinian evolution includes processes that are inclined to “select” traits conducive to reliable cognitive faculties.19 Contrary to Patricia Churchland, these naturalists believe that it is an evolutionary advantage for creatures to possess cognitive faculties that produce true beliefs reliably. Plantinga responds to this point by arguing that possessing true beliefs might possibly stand as an evolutionary advantage if survival and gene-spreading activities could be influenced causally by virtue of the content of one’s beliefs given naturalism and evolution.20 However, the conjunction of evolution and naturalism cannot allow actions to be caused in virtue of their belief content. Thus, even if it produced creatures with reliable cognitive faculties, Plantinga maintains the reliability would be a felicitous coincidence of the evolutionary process rather than a trait elected by natural selection.

The second premise states that if one accepts naturalism and evolution and realizes that naturalism and evolution render probable that one’s cognitive faculties are unreliable, then one has good reason to believe that one’s own cognitive faculties are unreliable. However, since one’s own cognitive faculties are the fundamental source of all of one’s beliefs, the third premise states that once one admits that one’s cognitive faculties are unreliable all of one’s beliefs are epistemically undermined, including one’s beliefs in naturalism and evolution. Finally, since accepting naturalism and evolution can provide one with a reason to disbelieve naturalism and evolution, it follows that the conjunction of naturalism and evolution is self-defeating and cannot be accepted rationally. And on the basis of these reasons, Plantinga concludes that naturalism and evolution cannot be accepted rationally.

For the purpose of contrasting Plantinga’s EAAN with Lewis’s AFR I will highlight a few important features of Plantinga’s argument. First, as I have already noted, the argument is not strictly against naturalism, but rather the conjunction of naturalism and evolution. The conjunctive position appears to be essential to Plantinga’s argument since Darwin’s doubt is generated by considering the reliability of one’s cognitive faculties as a result of naturalistic evolution and not by naturalism itself. Presumably, a naturalist who did not accept evolution would not be saddled with the defeater presented by Darwin’s doubt.

Second, implicit in Plantinga’s discussion is that “R,” the reliability of one’s cognitive faculties, is a necessary condition for being rational.21 The conundrum with naturalism and evolution, on Plantinga’s analysis, is that their conjunction leads one to believe that one’s cognitive faculties probably do not cause one to have beliefs that are true more often than not. Thus, the problem with naturalism and evolution is not to be found in the kind of way that beliefs are produced (more on this when I discuss Lewis). Rather, it is that the series of causes and effects that produce beliefs are not likely to produce true beliefs reliably. Keep in mind that Plantinga’s alternative for justified belief is his account of warrant through proper function.22 On his alternative proposal, warranted beliefs are (roughly) those that are produced by a cognitive faculty aimed successfully at producing true beliefs, operating in the environment for which it is designed. Thus, when a belief satisfies the conditions of being produced in a properly functioning way, it is warranted (unless there is an undefeated defeater for holding the belief). What Plantinga claims proper functionalism has going for it—and that naturalistic evolution does not—is that proper functionalism promises a system of causes and effects that will deliver reliable cognitive faculties whereas naturalistic evolution cannot. So by Plantinga’s lights, the difference between rational and irrational belief production is not a matter of kind, it is a matter of the reliability of the causes that produce beliefs. This is an important contrast to Lewis’s position.

Lewis’s AFR

In chapter three of Miracles, Lewis presents his “Cardinal Difficulty with Naturalism,” which is the primary source for the AFR.23 Lewis’s argument stems from the observation that naturalism is a system that precludes the kind of relation required for rational belief. Rather than defining rational belief in terms of the causal reliability of one’s cognitive faculties, Lewis sees rationality as being the result of the sui generis ability of the mind to “see” or “grasp” that one thing can be the grounds for inferring something else. The problem with naturalism, according to Lewis, is that naturalism cannot accommodate this condition required for reasoning. Thus, naturalism “discredits our processes of reasoning or at least reduces their credit to such a humble level that it can no longer support Naturalism itself.”24

To understand Lewis’s argument, we must first make a distinction. Lewis distinguishes mental states produced by cause-and-effect from those brought about by ground-and-consequent.25 Lewis illustrates the difference in the two with a pair of examples where the word “because” indicates the two concepts: (6) Grandpa is ill today because he ate lobster yesterday.

(7) Grandpa must be ill today because he hasn’t gotten up yet.
In (6), the word “because” is used to illustrate cause-and-effect. Grandpa’s illness is caused by his lobster consumption. In (7), however, the word “because” indicates something different. Clearly, it is not supposed to communicate that Grandpa’s remaining in bed is the cause of his illness. Rather, Grandpa’s remaining in bed is the grounds for inferring that he is ill (after all, Grandpa normally gets up early when he is feeling fine). Notice the difference in another pair of examples, between claiming “He cried out because it hurt him” (cause-and-effect) and “It must have hurt him because he cried out” (ground-and- consequent). “The one,” Lewis writes, “indicates a dynamic connection between events or ‘states of affairs’; the other, a logical relation between beliefs or assertions.”26 Hence, a mental state that occurs by cause-and-effect is not the product of reasoning but the effect produced by a non-rational cause. Beliefs that are produced as a result of the ground-and-consequent relation are essential for rationality. Lewis supports this notion by noting that we dismiss those beliefs that are the result of wishful thinking, prejudice, and insanity precisely because they have been produced entirely by processes involving cause-and-effect and not ground-and-consequent.27

According to the AFR, the problem with naturalism is that it is committed to the view of reality where purposeless cause-and-effect relations without exception bring about every event, including mental events, like thinking or reasoning. Consequently, given naturalism, what is believed to be an act of reason is actually brought about by the natural world’s closed series of non-rational causes-and-effects. The upshot is that the naturalist would have to conclude that his belief in naturalism is the byproduct of a series of non-rational causes-and-effects. Therefore, if naturalism is true, then it implies that no one can have rational grounds for believing that naturalism is true. Lewis succinctly states the self-refuting problem with naturalism raised in his AFR:

[I]f naturalism were true then all thoughts whatever would be wholly the result of irrational causes. Therefore, all thoughts would be equally worthless. Therefore, naturalism is worthless. If it is true, then we can know no truths. It cuts its own throat.28

While Lewis does not provide a formal statement of the argument, I think it can be expressed faithfully with the following propositions:
(8) Naturalism is the philosophical position that essentially excludes the possibility for beliefs to stand in ground-and-consequent relations.
(9) A necessary condition for rational thought is for beliefs to stand in ground-and-consequence relations.
(10) Naturalism cannot satisfy a necessary condition for rational thought to exist. (from 8 & 9)
(11) Any philosophical position that excludes rational thought is self-defeating. Therefore,
(12) Naturalism is self-defeating. (from 10 & 11)
In sum, Lewis’s argument proceeds by arguing that naturalism is a philosophy that only admits cause-and-effect relations and leaves no room for beliefs to be produced by ground-and-consequent. Since ground-and-consequent relations are necessary for rationality, anyone who arrives at a philosophical position that does not include them must admit that nothing, including his position, can be supported rationally. Finally, a philosophical position that entails nothing can be supported rationally must itself not be supported rationally, and so there are no grounds for believing it is true. Therefore, someone who became convinced that naturalism is true would also be in a position to believe that there is no rational support or grounds for believing it is true. Hence, any proof for naturalism would yield the self-refuting position of being a “proof that there are no such things as proofs.”29

Comparing EAAN and AFR

We are now in a position to compare Plantinga’s EAAN and Lewis’s AFR. While I do not want to give the impression that similarities between the two do not exist, my intention is to highlight some significant differences.

The first and most important contrast in the two arguments is in the way that they characterize the concept of rationality. In Plantinga’s EAAN, rationality is a matter of having cognitive faculties that produce true beliefs reliably—at least the reliability of one’s cognitive faculties is a necessary condition for rationality. His argument tries to show an inconsistency in the naturalist’s believing that naturalism and evolution are likely to be true when the naturalist draws the conclusion that his cognitive faculties are not likely to be reliable causes for true beliefs. The EAAN does not indict naturalism on the grounds that rationality is a different kind of relation, one that cannot consist of non-rational cause-and-effect relations. Rather, the point of the EAAN is that the conjunction of naturalism and evolution is not likely to provide a system of causes and effects where one’s cognitive faculties are likely to cause one to have true beliefs reliably. Lewis’s AFR, however, faults naturalism because it precludes the possibility of ground-and-consequent relations existing between mental states. In other words, Lewis finds naturalism to be self-defeating because naturalism has no place for the kind of relation that is necessary for rational thought to exist. Rationality, according to Lewis, is a different kind of relation than cause-and-effect. They are so different that discovering one’s thought has been produced by cause-and-effect would undermine its claim to being rational.

The difference on the nature of rationality leads to the second difference in the two arguments. The EAAN does not preclude the possibility that one may possess cognitive faculties that are reliable given evolution and naturalism. Plantinga states explicitly that the EAAN is not “an argument for the conclusion that unguided evolution could not produce creatures with reliable belief- producing faculties; … that it couldn’t is neither a premise nor the conclusion of my argument.”30 So, the EAAN does not attempt to argue that the conjunction of evolution and naturalism cannot produce creatures that have reliable, truth-conducive cognitive faculties. In other words, the requirement for rationality stipulated in the EAAN could possibly be met by naturalism and evolution. Rather, the problem is that the conjunction of evolution and naturalism is not likely to satisfy this condition for rationality and upon reflection should lead one to doubt the conjunction of evolution and naturalism.

Due to its conception of rationality, the AFR raises a different problem for naturalism than the EAAN. Lewis argues that naturalism cannot possibly accommodate an essential condition for rational thought because naturalism is the belief that the world consists of nothing more than the series of events that are interlocked by non-rational causes and effects. Since a requirement for rationality cannot possibly be met on Lewis’s conception of rationality, the AFR intends to demonstrate the impossibility of rationally accepting naturalism.

The third difference is that Lewis’s AFR is focused on an essential problem with naturalism, while Plantinga’s EAAN strikes only at a tangential feature of naturalism. As I have already noted, Plantinga’s EAAN is not an argument against naturalism, but the conjunction of naturalism and evolution. While it is no trifling conclusion to show that naturalism and evolution entail self-defeat, the consistent naturalist may keep his naturalism and reject evolution. Naturalism in its essentials remains unscathed by the EAAN. The AFR, however, challenges the essence of naturalism. No consistent naturalist could accept Lewis’s argument and maintain his naturalism.

Which Conception of Rationality is Right?

The crux of the differences between the arguments given by Plantinga and Lewis rests on the requirements for rational belief. Plantinga’s EAAN operates under the condition that it would be irrational to believe something that is produced by an unreliable faculty. Lewis, on the other hand, takes rationality to be a unique kind of mental activity that is inconsistent with being the product of a cause-and-effect process. Which standard is the right one for evaluating naturalism? On this issue, I believe Lewis holds the high ground, and I will present some initial reasons for thinking that Lewis is right. First, I will argue that the reliability of one’s cognitive faculties is not necessary for rational belief. Second, I will argue that the reliability of one’s cognitive faculties is not sufficient for rational belief. Third, I will present some reasons for thinking that Lewis’s view that rational beliefs occur by standing in the relation of ground-and-consequent is evidenced by first-hand experience. As I present these arguments, I will stipulate that Plantinga’s concept of causally reliable cognitive faculties and Lewis’s notion of a rational belief that requires the relation of ground-and- consequent are mutually exclusive.31 In other words, the concept of reliable cognitive faculties will be understood to denote when a person’s belief has been brought about by a cognitive mechanism that is part of the series of causes and effects where the belief-forming mechanism produces beliefs that are true more often than not. To emphasize this point, I will use the locution, “cognitive mechanisms,” to refer to cognitive faculties that function by cause-and-effect.

Reliable Cognitive Mechanisms not Necessary for Rationality

The first point in favor of Lewis’s conception of rationality is that the reliability of one’s cognitive mechanisms is not necessary for rational belief. Imagine that someone has cognitive mechanisms that cause the person to have beliefs that are more often false than true. This can be done by considering classic skeptical cases, like being a brain-in-a-vat or the Matrix.32 The person who is fictionalized in these epistemically unfortunate scenarios has cognitive faculties that produce beliefs that are false more often than true. In these scenarios, the person’s brain is stimulated in just the right way to cause the person to believe, for example, that there is a tree in front of him when, in fact, there is no tree in front of him. For each of this person’s cognitive mechanisms, let us assume that it is unreliable.33 Does it follow that the person’s failure to possess reliable cognitive mechanisms prevents him from being rational?

The answer is clearly no. Even though a person may have completely unreliable cognitive mechanisms, the person could still make rational inferences. For example, the person in the Matrix who has abysmally unreliable cognitive mechanisms (not a single belief produced by the cerebral cortex is true!) can still make reasonable inferences based on the beliefs produced by his cognitive faculties. If the Matrix causes him to believe that there are five apple trees in front of him, and he reasons, since there are five trees in front of me, then there must be an odd number of trees in front of me, then I see no reason to think this is an irrational belief. After all, the person is not aware that his cognitive mechanisms are not reliable. His reasoning takes place exactly like you or I might reason if we were standing in front of five trees. Of course, the person does not know that there are five trees in front of him (because his belief is not true and knowledge is factive). But lacking reliable cognitive causal mechanisms does not seem to prevent him from being rational in holding his belief. In other words, the unreliability of cognitive mechanisms does not keep the person from inferring a belief (there must be an odd number of trees in front of me) on the grounds of another belief (there are five trees in front of me). Since possessing unreliable cognitive mechanisms is no barrier to forming beliefs by “seeing” the logical relation that holds between two beliefs, it follows that possessing reliable cognitive mechanisms is not necessary for rational belief.

Reliable Cognitive Mechanisms not Sufficient for Rationality

The reliability of one’s cognitive mechanisms is also not sufficient for rationality. While Plantinga’s EAAN does not depend on the sufficiency claim, I will cover this possibility for the sake of thoroughness. That the reliability of one’s cognitive mechanisms is not sufficient for rational belief has been illustrated by Laurence BonJour’s thought experiment involving Norman the clairvoyant:

Norman, under certain conditions which usually obtain, is a completely reliable clairvoyant with respect to certain kinds of subject matter. He possesses no evidence or reasons of any kind for or against the general possibility of such a cognitive power or for or against the thesis that he possesses it. One day Norman comes to believe that the President is in New York City, though he has no evidence either for or against this belief. In fact, the belief is true and results from his clairvoyant power under circumstances in which it is completely reliable.34

Keith Lehrer proposes a similar thought experiment using the example of Mr. Truetemp.35 Lehrer imagines that a small device, a “tempucomp,” could be inserted in someone’s head such that immediately it causes that person to believe truly the current temperature of his environment. Then, Lehrer asks us to imagine a person, Mr. Truetemp, who unbeknownst to him has had the tempucomp implanted. Even though Mr. Truetemp is reliably forming true beliefs about the temperature, virtually no one would deem his beliefs about the temperature to count as knowledge, because Mr. Truetemp is clueless as to how he is acquiring his beliefs about the temperature. Although Lehrer couches his example in terms of knowledge, his point applies equally to the concept of rational belief.

The underlying significance behind these thought experiments is that the possession of reliable cognitive mechanisms is not sufficient for rational belief. Just because Norman and Mr. Truetemp have beliefs that are caused by mechanisms that are impeccably accurate, this is not enough to count their beliefs as rational.36

First-hand Experience of Ground-and-Consequent Relations

I have argued that Plantinga’s condition is neither necessary nor sufficient for rational belief. The next point I want to make in favor of Lewis’s conception of rationality is the positive case for accepting ground-and-consequent relations as the basis for rationality. This case is based on a person’s immediate and direct awareness of ground-and-consequent relations. Anyone who has taught logic courses is familiar with attempts to get students to “see” the logical relations between propositions by asking students to think carefully about them. For example, I may write on the board, “All men are mortal and Socrates is a man,” and then ask students to draw the conclusion that logically follows from what I have written. Students typically are able to “see” the logical implication and answer, “Socrates is mortal.” This exercise works because I trust that students are capable of “seeing” the ground-and-consequent relation between “All men are mortal and Socrates is a man” and “Socrates is mortal.”

Likewise, Lewis believes that we have first-hand evidence that indicates that we do, in fact, arrive at some beliefs on the basis of “seeing” the ground-and-consequent relation. Lewis appeals to examples where we use “because” in the two different senses to call upon our own introspective awareness of a difference each of us has experienced first-hand.37 The problem with naturalism is that it must cast doubt upon this very experience of drawing inferences on the basis of ground-and-consequent. If our experience of ground-and-consequent is ultimately illusory, then it implies that one’s reliance on the experience of inferring naturalism on the basis of evidence is an illusory experience as well. Thus, the problem with naturalism is that it undermines itself, and it flies contrary to our own experience.38

Speaking of reasoning as an experience is somewhat misleading and Lewis would disagree with the implication that it is experienced passively rather than performed actively. He describes reasoning in the following way: “It is not an object that knocks against us, nor even a sensation which we feel. Reasoning doesn’t ‘happen to’ us: we do it.”39 Thus, when I speak of experiencing reasoning on the basis of ground-and-consequent, the experience should be understood as an active experience, the experience one has of engaging and enacting one’s mental life actively, perhaps analogous to the experience of exercising free will. Lewis’s contention that rationality essentially involves the activity of inferring beliefs on the basis of ground-and-consequent is supported by our own first-hand experience. The truly rational act of the mind is its “seeing” one idea “follows” from another; everything else is non-rational.

Objections and Replies

In this final section, I will consider two objections to my position. First, I will consider a response on behalf of the naturalist who might claim that cause-and-effect and ground-and-consequent relations are not mutually exclusive. The second objection that I will consider is a Plantingan rejoinder to the effect that ground-and-consequent belief formation is just another cognitive faculty that one should include in the class of cognitive faculties.

On Behalf of the Naturalist

The naturalist may try to salvage his position by maintaining that ground-and-consequent relations and cause-and-effect relations could both be responsible for producing the correct resultant belief in acts of reasoning. For example, one kind of naturalist may claim that believing “All men are mortal and Socrates is a man” has both mental properties (those associated with the content of the belief) and natural properties (those associated with the physical features of the corresponding brain state). Thus, the naturalist may allege that the resultant belief that “Socrates is mortal” is the product of both cause-and-effect due to the cause-and-effect relationship between the natural properties of the two brain states and the ground-and-consequent relation that holds for the content of the beliefs.

There are at least two irreparable problems with this objection. First, the ground-and-consequent relation does not always coincide with the cause-and-effect relation because there are innumerable logical consequences that can be inferred from any logical grounds, whereas there is only one effect that can follow from a given cause.40 Having a particular belief (which the naturalist claims must be associated with a particular brain state) is not sufficient to “see” all the logical implications that follow from it. Lewis makes this point when he writes:

We know by experience that a thought does not necessarily cause all, or even any, of the thoughts which logically stand to it as Consequent to Ground. We should be in a pretty pickle if we could never think `This is glass’ without drawing all the inferences which could be drawn. It is impossible to draw them all; quite often we draw none.41

In other words, if being in a particular brain state was sufficient to bring about the same resultant beliefs through cause-and-effect as “seeing” the logical grounds for drawing an inference in an act of reason, then being in a particular brain state would always lead one to “see” all the beliefs logically related to the original belief. But clearly (and thankfully!) we do not (and cannot) be caused to draw all of the logical implications that follow from the truth of a given belief. In fact, the natural properties of a given brain state, taken in a closed naturalistic system, yield one result from the constitution of the initial conditions and causal laws of the system.42

The second problem with the naturalist’s response is known as the problem of mental causation or the causal exclusion argument.43 The problem is that both the natural properties of a particular brain state and the mental properties of a particular belief’s contents cannot both bring about the effect causally.44 Either the result is brought about by the causal efficacy of the physical properties or by the mental content of the belief. However, given the closed system of naturalism, clearly only the natural properties can cause the effect. Indeed, given naturalism, the mental content of one’s beliefs must be relegated as being merely “epiphenomenal,” that is, causally impotent. Since the most fundamental forces in naturalism are the blind, purposeless causes of nature, the naturalist must admit that even if there is a nice correlation between beliefs existing that have ground-and-consequent relations, the existence of these beliefs would be brought about entirely by cause-and-effect, and not by virtue of the content of the belief states.

On Behalf of the Plantingan

A defender of Plantinga’s general approach to rationality might object to my bifurcating ground-and-consequent belief production from the set of cognitive faculties that produce beliefs reliably. After all, can we not speak of the reliability of one’s mental faculties, that is, the cognitive faculties responsible for producing one’s beliefs arrived at through inferences? Are not some patterns of inference more reliable than others? In my critique of Plantinga’s account of rationality, it may appear I have overlooked the reliability of those very cognitive faculties responsible for forming beliefs based on ground-and-consequent relations. Therefore, the Plantingan can argue that the reliability of one’s “rational faculties” (that is, those that produce beliefs by ground-and-consequent) is intended to be included in Plantinga’s standard for rationality. In other words, they allege that my attempt to separate ground-and-consequent belief production from the reliability of one’s cognitive faculties does not reflect Plantinga’s point.

In the first place, it would be wrong to assess the epistemic value of ground-and-consequent reasoning from the perspective of its reliability to produce true beliefs. The underlying rational insight of ground-and-consequent reasoning is the basis from which we judge the rationality and reliability of other epistemic practices. In addition to being wrong, it would also be bizarre to try to defend the rationality of ground-and-consequent reasoning on the basis of its reliability. William Hasker, a contemporary defender of the AFR, explains the wrongheaded nature of such a move in his book, The Emergent Self:

How are we to satisfy ourselves as to which inferential practices are reliable? By hypothesis, we are precluded from appealing to rational insight to validate our conclusions about this. One might say that we have learned to distinguish good reasoning from bad reasoning, by noticing that good inference-patterns generally give rise to true conclusions, while bad inference-patterns often give rise to falsehood. (This of course assumes that our judgments about particular facts, especially facts revealed through sense perception, are not in question here—an assumption I will grant for the present.) But this sort of “logical empiricism” is at best a very crude method for assessing the goodness of arguments. There are plenty of invalid arguments with true conclusions, and plenty of valid arguments with false conclusions. There are even good inductive arguments with all true premises in which the conclusions are false. There are just the distinctions which the science of logic exists to help us with; basing the science on the kind of ham-fisted empiricism described above is a hopeless enterprise.45

Next, we encounter a formal difficulty in Plantinga’s EAAN. If we are supposed to assess the probability of possessing reliable cognitive faculties given naturalism and evolution, and included amongst one’s reliable cognitive faculties are the very thought processes being used to think about Plantinga’s EAAN, the conditionalization will not work.46 Since one must work on the assumption that one’s own thinking is sound in order to think about Plantinga’s argument, it is not possible to consider one’s ability to reason from ground-and-consequent independently from the reliability of one’s cognitive faculties. While it may be possible to wonder whether the information caused by one’s sensory organs is reliable, one cannot question in the same way whether one’s ability to reason based on ground-and-consequent is reliable. After all, one’s reasoning itself must be relied upon to draw any inferences about its reliability. Since we all take some of our rational faculties to be reliable, then when we assess the reliability of all of our cognitive faculties, we do so given naturalism, evolution, and the reliability of some cognitive faculties. Since the reliability of some cognitive faculties is part of what is given, the probability that one has reliable cognitive faculties turns out to be almost certain when one includes one’s ability to reason from ground-and-consequent into the reliability of one’s cognitive faculties in Plantinga’s EAAN. In other words, one cannot consider the reliability of thinking based on ground- and-consequent without relying on ground-and-consequent. One might as well try to use one’s conscious experience to prove that consciousness does not exist.

Lewis makes a related point when he writes:

If the value of our reasoning is in doubt, you cannot try to establish it by reasoning. If, as I said above, a proof that there are not proofs is nonsensical, so is a proof that there are proofs. Reason is our starting point. There can be no question either of attacking or defending it. If by treating it as a mere phenomenon you put yourself outside it, there is then no way, except by begging the question, of getting inside again.47

While the reliability of one’s cognitive mechanisms can be called into question without falling into inconsistency, the same cannot apply to reasoning based on ground-and-consequent. Since the soundness of beliefs arrived at by ground-and-consequent is the starting point for all thought based on inference, we must take its truth-conduciveness as axiomatic and underived. Any “proof” that reasoning by ground-and-consequent is unreliable must indicate the “proof” has a serious flaw rather than leading one to draw the inference (by ground-and-consequent) that reasoning based on ground-and-consequent is unreliable. For this reason, one cannot conditionalize the probability of possessing truth-conducive rational faculties apart from the exercise and acceptance of them. Thus, the Plantingan attempt to incorporate one’s rational faculties as part of one’s cognitive mechanisms is not a feasible way to work the EAAN.48


While Plantinga’s EAAN and Lewis’s AFR have some common features, there are significant differences between the two. I have highlighted that the EAAN argues against the conjunction of naturalism and evolution, whereas the AFR makes a case against naturalism simpliciter. Furthermore, I have shown that there are significant differences between the concepts of rationality employed in the two arguments. These two different conceptions of rationality are important. Ultimately, it is because of the incorrect standard for rationality employed in the EAAN that it fails. The better hope for a successful argument against naturalism rests with Lewis’s AFR.49

Cite this article
John M. DePoe, “The Self-Defeat of Naturalism: A Critical Comparison of Alvin Plantinga and C. S. Lewis”, Christian Scholar’s Review, 44:1 , 9-26


  1. See Alvin Plantinga, Warrant and Proper Function (Oxford: Oxford University Press, 1993), ch. 12; Plantinga, Warranted Christian Belief (Oxford: Oxford University Press, 2000), 227-240; Plantinga, “Introduction” in Naturalism Defeated? ed. James Beilby (Ithaca, NY: Cornell University Press, 2002), 1-12; Plantinga, Where the Conflict Really Lies (Oxford: Oxford University Press, 2011), ch. 10.
  2. C. S. Lewis, Miracles, rev. ed. (New York: Touchstone, 1996 [orig. 1960]). This argument also is given in chapter 13 of Miracles and Lewis’s essay, “Religion without Dogma?” in God in the Dock, ed. Walter Hooper (Grand Rapids, MI: Eerdmans, 1970), 126-146.
  3. Alvin Plantinga, Warrant and Proper Function, 237 n. 28. See also Victor Reppert, “The Argument from Reason,” in The Blackwell Companion to Natural Theology, eds. W. L. Craig and J. P. Moreland (Malden, MA: Blackwell, 2009), 362.
  4. James Beilby, “Preface,” in Naturalism Defeated? ix.
  5. Alvin Plantinga, Where the Conflict Really Lies, 310 n. 4.
  6. C. S. Lewis, Miracles, 13-17. See also Lewis, “Religion without Dogma,” 133. William Hasker offers a strikingly similar account of “physicalism,” which he defines as the position that accepts all reality is constrained by causes and explanations that are mechanistic and all mechanistic causes and explanations are non-teleological. See William Hasker, The Emergent Self (Ithaca, NY: Cornell University Press, 1999), 62-64.
  7. Miracles, 14.
  8. See, for instance, Warranted Christian Belief, 229 n. 44.
  9. Richard Dawkins, The Blind Watchmaker (New York: Norton, 1986), 5.
  10. Daniel Dennett, Darwin’s Dangerous Idea (New York: Simon & Shuster, 1995), 50ff.
  11. For example, see Warrant and Proper Function, 217-218; and Warranted Christian Belief, 225-227.
  12. Miracles, 16-17, 42-43.
  13. See the references in footnote 1, and notice that Plantinga’s argument focuses on the P(R|N&E), rather than P(R|N). See, however, the suggestion that the argument is directed “perhaps just [at] naturalism” in Where the Conflict Really Lies, 314. For reasons I will briefly state below, I do not think Plantinga’s EAAN can be made against a non-conjunctive naturalist position.
  14. Of course, naturalists of the past include famously Lucretius and Paul-Henri Thiry d’Holbach (among many others) that preceded Darwin’s theory. Today, Thomas Nagle stands as an example of a naturalist who is skeptical of Darwinian evolution. See Nagle, Mind and Cosmos: Why the Materialist Neo-Darwinian Conception of Nature is Almost Certainly False (Oxford: Oxford University Press, 2012).
  15. Where the Conflict Really Lies, 314; italics in original.
  16. Charles Darwin, Letter to William Graham, July 3, 1881, quoted in Where the Conflict Really Lies, 316.
  17. Patricia Churchland, “Epistemology in the Age of Neuroscience,” The Journal of Philosophy 74 (October 1987): 548-549; emphasis in original.
  18. Where the Conflict Really Lies, 326-335.
  19. See, for example, Evan Fales, “Darwin’s Doubt, Calvin’s Calvary” in Naturalism Defeated?, 43-60. Even Richard Swinburne, a prominent theistic philosopher, believes that possessing true beliefs provides some survival advantage in the evolutionary process. See Richard Swinburne, The Existence of God, 2nd ed. (Oxford: Oxford University Press, 2004), 350-354.
  20. Where the Conflict Really Lies, 335-337.
  21. In this discussion, I am focusing on the requirements for rationality as required for Plantinga’s EAAN. Importantly, this discussion is not going to cover Plantinga’s account of rationality as discussed in his many writings on epistemology.
  22. See Warrant and Proper Function and Warranted Christian Belief for hallmark statements of his position.
  23. See also C. S. Lewis, Miracles, ch. 13; and Lewis, “Religion without Dogma.”
  24. Miracles, 24.
  25. Ibid., 24-26. Incidentally, this distinction is not original to Lewis but was a common and accepted distinction going back at least to the early parts of twentieth-century epistemology. See, for example, the American philosopher, C. I. Lewis, An Analysis of Knowledge and Valuation (La Salle, IL: Open Court, 1946), 11-16.
  26. Miracles, 25.
  27. Ibid., 26.
  28. Lewis, “Religion without Dogma,” 137.
  29. Miracles, 24.
  30. Where the Conflict Really Lies, 310; italics in original.
  31. This is an important qualification since Plantinga may want to hold that sometimes the two coincide, especially when it comes to his account of warrant or knowledge. In order to focus on the essential differences in the EAAN and AFR, I will need to distinguish pure cases of beliefs produced by cause-and-effect compared to those based on ground-and-consequent.
  32. For background on these thought experiments, see Tony Brueckner, “Skepticism and Content Externalism,” in The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta (Spring 2012), <>.
  33. If the concern is that the cognitive faculties of such a person would be reliable if they were to operate in the environment for which they are designed, we can modify the thought experiment so that the person’s cognitive faculties produce unreliable beliefs in most environments, including the one in which they operate normally.
  34. Laurence BonJour, The Structure of Empirical Knowledge (Cambridge, MA: Harvard University Press, 1985), 41.
  35. Keith Lehrer, Theory of Knowledge, 2nd ed. (Boulder, CO: Westview Press, 2000), 187-188.
  36. There is much discussion about the role of the subject’s awareness as a general requirement for justification. Plantinga’s accomplished protégé, Michael Bergmann, has argued forcefully that such a requirement is hopeless in Justification without Awareness (Oxford: Oxford University Press, 2006). For a response to this charge, see John M. DePoe, “Bergmann’s Dilemma and Internalism’s Escape,” Acta Analytica 27.4 (2012): 409-423.
  37. See examples of (6) and (7) in this paper under the section heading “Lewis’s AFR.”
  38. Lewis makes this point about the naturalist’s inconsistency in affirming reason especially in Miracles, 31-32; and “Religion without Dogma,” 136-137.
  39. C. S. Lewis, Miracles, 41; emphasis in original.
  40. The statement “there is only one effect that can follow from a given cause” is a bit of an oversimplification. Given some indeterministic interpretations of quantum physics, there may be some cases where many effects could follow from one cause. However, this wrinkle is of no consequence to my point. The naturalist must be committed to the idea that brain states (not quantum events) can produce precisely as many effects as there are possible logical inferences to be drawn from one logical “ground.” To my knowledge, this is not a plausible position to take, even granting some of the wildest interpretations of quantum physics.
  41. Miracles, 26-27.
  42. Along the lines of footnote 40, I may amend this more accurately to state that if naturalism is correct, then two people in identical brain states would necessarily be caused to have a brain state consistent with a small subset of brain states that are the product of the initial conditions and causal laws. This amendment is not sufficient for a naturalist to overcome the material point, namely that the natural properties of a brain state do not accommodate sufficiently the same openness to the resultant belief states afforded by the mental properties (for example, belief content) involved in ground-and-consequence relations.
  43. Most famously raised by Jaegwon Kim. For a recent version of his argument, see Jaegwon Kim, Physicalism or Something Near Enough (Princeton, NJ: Princeton University Press, 2005), ch. 1. For an overview with a bibliography, see David Robb and John Heil, “Mental Causation,” in The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta (Spring 2013 Edition), <>.
  44. Importantly, the causal exclusion argument assumes a non-reductive account of causation, which I will take for granted in this brief discussion.
  45. Hasker, Emergent Self, 74-75.
  46. Related problems have been raised by Michael Bergmann, “Commonsense Naturalism,” in Naturalism Defeated? 61-90; Richard Otte, “Conditional Probabilities in Plantinga’s Argument,” in Naturalism Defeated? 135-149; Trenton Merricks, “Conditional Probability and Defeat,” in Naturalism Defeated? 165-175; Yingjin Xu, “The Troublesome Explanandum in Plantinga’s Argument against Naturalism,” International Journal for the Philosophy of Religion 69 (2011): 1-15.
  47. Miracles, 32-33.
  48. In Plantinga, Where the Conflict Really Lies, 348-349, Plantinga suggests in response to those who press the kind of objection I am raising that he may have to offer a weaker argument against naturalism. Instead of arguing that the reliability of all of one’s cognitive faculties are inconsistent with naturalism and evolution, perhaps he only needs to argue that the probability is low for the reliability of only those cognitive faculties responsible for producing metaphysical beliefs given naturalism and evolution. I confess I have no clue what kinds of cognitive faculties one may have besides one’s rational faculties that are used to produce metaphysical beliefs. In my own experience when I consider metaphysical beliefs like idealism, mind-body dualism, or the reality of universals, I am aware of no special mechanism that is responsible for producing these beliefs. Introspectively, I form my beliefs about these matters through reasoning based on propositions that I take to be true.
  49. I would like to thank Dax Bennington and David Naugle who invited me to read an earlier draft of this paper at a faculty colloquium at Dallas Baptist University. I also wish to thank two anonymous referees for this journal, John Churchill and Tyler McNabb, who read earlier versions of this work and offered insightful critical feedback.

John M. DePoe

Marywood University
John M. DePoe is Academic Dean of the Schools of Logic and Rhetoric at Kingdom Preparatory Academy in Lubbock, Texas and former Professor of Philosophy at Marywood University.