Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.

x

The Evolutionary Argument Against Naturalism
Omar Mirza*
St Cloud State University

Abstract

Metaphysical naturalism can be taken, roughly, to be the view that there is no God, and nothing beyond nature. Alvin Plantinga has argued that naturalism, in this sense, is self-defeating. More specifically, he argues that an evolutionary account of human origins gives the naturalist (but not the theist) compelling reasons for doubting the reliability of human cognitive faculties, and thus compelling reasons for doubting the truth of any of his beliefs, including naturalism itself. This argument, which has come to be known as the ‘evolutionary argument against naturalism’, has generated a great deal of controversy, and a substantial literature concerning it has grown up as a result. In this paper, I will introduce readers to this literature. I begin by explaining the argument itself, and making clear its intuitive force. I then survey the main objections to it, such as the Perspiration Objection, the ‘Can’t the Naturalist Just Add a Little Something?’ Objection, and the Tu Quoque Objection: in the course of this survey, I pay particular attention to the most interesting of these, a version of the Tu Quoque Objection according to which the problem of evil results in a form of epistemic self-defeat for the theist that is exactly analogous to the self-defeat with which the naturalist is allegedly faced in the evolutionary argument. I go on to suggest that, despite the wide range of objections in the literature, the challenge of the evolutionary argument against naturalism is still very much with us, and I conclude by describing some promising directions for future research.

Much discussion in the philosophy of religion has been about whether or not theists are rational in believing as they do. It has frequently been argued that theistic belief is in some way epistemically defective, deplorable, or outside the pale of legitimate intellectual endeavor. These arguments have generally been put forward by adherents of metaphysical naturalism, the view that there is no God and nothing like God (henceforth, naturalism). It is in this context that Alvin Plantinga has advanced the evolutionary argument against naturalism (henceforth, EAAN): the aim of this argument is to show that, under certain easily realized conditions, it is irrational to hold onto belief in naturalism (Plantinga, ‘Introduction’,12). The charge of irrationality that has been repeatedly leveled at theists is now directed at naturalists. Naturalism, says Plantinga, is self-defeating (‘Introduction’, 2). This self-defeat is meant to follow from the naturalist’s commitment to evolutionary theory, and her acceptance of a certain probability statement.1 Plantinga also argues that the same problem (of self-defeat) is not faced by theism (Warrant and Proper Function, 236).2 What exactly is meant by ‘self-defeat’ here? To understand this, we need to consider the idea of a defeater. To a first approximation, a defeater for a belief is a compelling reason to give up or doubt that belief.3 More precisely, to say that D is a defeater for B, relative to a person S at time t is to say the following: given that S believes D at time t, it is irrational for S to also believe B (Plantinga, ‘Naturalism Defeated’). In general, defeaters can themselves be defeated when a person acquires new beliefs. These defeaters, which defeat earlier defeaters, will be called defeater-defeaters. If D is a
ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd

Evolutionary Argument Against Naturalism

79

defeater for B, relative to S, and S has no defeaters for D, D is said to be an undefeated defeater for B (Plantinga, ‘Naturalism Defeated’). We can now explain the conclusion that naturalism is self-defeating as follows: under certain easily realized conditions, a naturalist will acquire an undefeated defeater for his belief in naturalism.4 Let’s look in a bit more detail at how the argument for this is supposed to work. To begin with, let N stand for naturalism; let E stand for the view that human cognitive faculties have evolved by way of the mechanisms that are studied by contemporary evolutionary theory; and let R stand for the claim that the beliefs produced by those cognitive faculties are for the most part true.5 EAAN has three stages, each of which involves defending a certain premise: 1. The objective conditional probability P(R ⁄ N&E) is low or inscrutable (meaning that we cannot determine whether it is low or high): call this the Probability Thesis. 2. Anyone who accepts N and E and the Probability Thesis thereby has a defeater for R: this is the Defeater Thesis.6 3. Anyone who has a defeater for R has an undefeated defeater for each of his beliefs. From the second and third premises, it follows that anyone who accepts N and E and the Probability Thesis has an undefeated defeater for each of his beliefs; this means that one who accepts N and E and the Probability Thesis thereby has reason to doubt all his beliefs, and hence is committed to a kind of global skepticism about all his beliefs.7 But one of those beliefs is, of course, naturalism itself; hence, anyone who accepts N and E and the Probability Thesis has an undefeated defeater for his belief in naturalism. Now, a rational and well-informed naturalist must accept E: it is, thinks Plantinga, the only option for the naturalist when it comes to explaining the origin of our faculties, given the current state of our evidence (‘Introduction’, 12). Moreover, a rational naturalist who grasps the argument for the Probability Thesis will accept it too (granting, for the moment, that this argument is sound). So rational, well-informed naturalists who grasp the argument for the Probability Thesis will accept E and the Probability Thesis, and thereby acquire an undefeated defeater for their belief in naturalism. Hence, naturalism is self-defeating. Let us see how these premises are defended. 1. Defense of the Probability Thesis The Probability Thesis is a claim about the value of a particular objective conditional probability, namely P(R ⁄ N&E).8 What, intuitively, is the significance of this probability? To see this, consider the case of a hypothetical population of creatures on a planet a lot like earth, formed by blind, undirected evolution, and suppose that naturalism is true (Plantinga, ‘Introduction’, 5). Assuming further that they have and change beliefs, but that we know nothing else relevant about them, how likely is it that their cognitive faculties are reliable? The answer to this question is the value of P(R ⁄ N&E). Now there are two broad classes of view regarding the likely effect of evolutionary origins on cognitive reliability (Warrant and Proper Function, 218–19). On the one hand, there are those who think that the fact that we have evolved and survived makes it likely that our cognitive faculties are indeed reliable9: on this kind of view, P(R ⁄ N&E) will be high. On the other hand, there are those who think that our evolutionary history gives us grounds for doubting our cognitive reliability, or at least fails to give us any grounds

ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd

Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x

80 Evolutionary Argument Against Naturalism

for affirming it10: on this kind of view, P(R ⁄ N&E) will be low, or at any rate, can’t reasonably be affirmed to be high. So, who is right here? What is P(R ⁄ N&E)? Let’s start by thinking about those hypothetical creatures: this will help us avoid any undue bias that could result from our prior conviction that our own cognitive faculties are reliable (Plantinga, ‘Introduction’, 5). We are assuming that naturalism is true, and most contemporary naturalists will be materialists with respect to humans and other conscious beings, so we will assume that these creatures are material objects, and that their beliefs are neural states or processes.11 Note also that, so far, we have not been given that the beliefs of these hypothetical creatures are connected to their actions in anything like the way we take for granted. For all we know their beliefs and their actions might have a common cause without in any way causing each other, or their beliefs might cause actions by virtue of the neurophysiological properties of those beliefs, but not by virtue of their content.12 A number of such possibilities are considered by Plantinga in different presentations of the argument, but these can be reduced to two (‘Introduction’, 10): C The contents of the beliefs of these creatures enter into the causal chains leading to actions. C The contents of the beliefs of these creatures do not enter into the causal chains leadings to actions. The probability calculus now gives us the following formula: P(R/N&E) ¼ P(R/N&E&C) P(C/N&E) þ P(R/N&E&~C)P(~C/N&E) P(R ⁄ N&E) is then evaluated by estimating each of the four terms that appear in the formula (Plantinga, ‘Introduction’, 10). P(R ⁄ N&E&C) is estimated as low, because in this case the contents of beliefs will be invisible to natural selection and so there will be no selection pressure towards those contents being mostly true (Plantinga, ‘Reply’, 253). It seems, initially, as though P(R/N&E&C) is going to be very high, but Plantinga contests this estimate by arguing, on the basis of examples, that there are indefinitely many possible belief systems with mostly false beliefs, but which nevertheless lead to felicitous action (Plantinga, ‘Reply’, 253). Plantinga concludes that P(R ⁄ N&E&C) will be at best moderately high, not very much more than ½. How do we estimate the probabilities P(C ⁄ N&E) and P(C ⁄ N&E)? Note that C corresponds to the view labeled ‘semantic epiphenomenalism’ in the philosophy of mind. Plantinga thinks that, because of the enormous difficulties that naturalists face in avoiding semantic epiphenomenalism, P(C ⁄ N&E) should be estimated as very high, which in turn means that P(C ⁄ N&E) is very low (‘Introduction’, 10). Plantinga claims that a reasonable evaluation of the probabilities leads to an estimate of P(R ⁄ N&E) as being somewhat less than ½ (‘Introduction’, 10). We can reconstruct the reasoning behind this evaluation as follows: the formula above is a weighted average of P(R ⁄ N&E&C) and P(R ⁄ N&E&C), and since P(C ⁄ N&E) is very high, P(R ⁄ N&E&C), which is very low, is weighted much more heavily than P(R ⁄ N&E&C); hence, it is reasonable to think that P(R ⁄ N&E) is low.13 A natural objection to this argument is that we should be agnostic about these probabilities, and hence we should think that P(R ⁄ N&E) is inscrutable to us (Plantinga, ‘Introduction’, 10). Plantinga concedes that this is a reasonable stance, and says that P(R ⁄ N&E) is either low, or inscrutable, specified to the hypothetical creatures. But, he thinks, the
ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x

Evolutionary Argument Against Naturalism

81

value of this probability will be the same for us as it is for those creatures (‘Introduction’,10). And this concludes the defense of the Probability Thesis.14 2. Defense of the Defeater Thesis The Defeater Thesis is defended by appealing to epistemic intuitions about hypothetical cases that, it is claimed, are clearly analogous to the case of the naturalist in EAAN: since, in these cases, the subject has a defeater for R, the same is claimed to be true of the naturalist who accepts the Probability Thesis. One hypothetical case that has been central in discussions of EAAN is the XX Case (Plantinga, ‘Reply’, 208). XX Case. Suppose a man believes that he has taken an unreliability-inducing drug called XX. He also believes that the drug is capable of causing massive unreliability in those who take it, so that even the most elementary and intuitively obvious cognitive operations are affected; he believes this unreliability to be of such an extreme kind that its victims are often unable to detect their own unreliability (that is, they continue to believe they are reliable when in fact they are not). He comes to believe that the probability that a person has remained cognitively reliable given that they have taken this drug is low or inscrutable. That is, if we let XX also stand for the statement that the man has taken the drug, then P(R ⁄ XX) is low or inscrutable. Then the man has a defeater for R, specified to himself. The XX case is the most widely cited analogy used to support the Defeater Thesis. Other cases are also used: what is important to note here is that the Defeater Thesis is defended, not by the use of any principle, but rather on the basis of analogies (Plantinga, ‘Reply’, 240). 3. Defense of the Third Premise Plantinga defends the third premise by arguing that, if the naturalist has a defeater for R, this generates a defeater for the rest of his beliefs as well (‘Introduction’,12). The reason is that all of the naturalist’s beliefs are products of his cognitive faculties, which constitute their source. Once the reliability of that source comes into question, so do the beliefs generated by the source. Moreover, the defeater for R that the naturalist acquires cannot itself be defeated, since everything that could be a defeater-defeater is itself subject to defeat. 4. Objections to EAAN Let us now consider some objections to EAAN. My goal here is not to settle anything, but just to give readers a taste of the discussions that EAAN has provoked. Obviously, I will only be able to cover a small sample of the objections that have been made. And for those objections I do describe, I won’t be able to do justice to the detail and sophistication with which they have been proposed. Instead, I will content myself with introducing two important categories of objection: objections to the Defeater Thesis, and Tu Quoque objections. 5. Objections to the Defeater Thesis Most of the controversy regarding the argument has focused on the Defeater Thesis. There has been one main worry that critics have had about this claim. The objections to it that we will describe are manifestations of this worry, which can be expressed as
ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x

82 Evolutionary Argument Against Naturalism

follows: what exactly is the connection between the naturalist’s acceptance of the Probability Thesis on the one hand, and her acquisition of a defeater for R on the other? One natural expression of this worry is the Perspiration Objection (Plantinga, ‘Naturalism Defeated’). The Perspiration Objection. ‘Let F be the claim that the function of perspiration is to cool the body. Now, the probability of F given (just) N&E is also low. But surely it would be absurd to claim that this gives the naturalist a defeater for the belief that F is true. Thus, it is also absurd to claim that the naturalist has a defeater for R in virtue of accepting the Probability Thesis.’ We can think of the Perspiration Objection as putting forward a certain analogy; only this time, unlike the case of EAAN, the analogy is put forward, not to defend the Defeater Thesis, but rather to attack it. In other words, this objection can be seen as an argument by analogy for the falsity of the Defeater Thesis. Does the Perspiration Objection succeed? I think not. More specifically, it will succeed only if the perspiration case is sufficiently similar, in epistemically relevant respects, to the EAAN case (perhaps by being at least as similar to the EAAN case as the latter is to the XX case). I don’t have the space to go into this in detail, but it seems clear to me that this is false, and hence that the Perspiration Objection fails15; nevertheless, it does suggest a helpful approach to evaluating the Defeater Thesis: look for hypothetical cases in which seeing that a probability is low or inscrutable does not yield a defeater, and which are at least as strongly analogous to EAAN as is the XX case (Talbott and Wielenberg adopt this strategy). If there are such cases, they would seriously undermine the support on which the Defeater Thesis rests. A plausible suggestion is that there is no defeater in the perspiration case because the naturalist has other beliefs relevant to the function of perspiration, beyond just N&E, and the probability that the function of perspiration is to cool the body relative to these other beliefs is high. So could not the naturalist appeal to other beliefs to raise the probability of R? This thought leads naturally to the following objection (Plantinga, ‘Naturalism Defeated’). The ‘Can’t the Naturalist Just Add a Little Something?’ Objection. ‘The naturalist believes many things besides N&E, such as R itself, or the proposition L that ‘‘we have won the evolutionary lottery’’. Relative to N&E conjoined with these other beliefs, the probability of R is very high, and thus the naturalist need not have a defeater in virtue of accepting the Probability Thesis.’ Plantinga has argued that this objection is very implausible, for one cannot escape having a defeater for a belief just in virtue of the fact that one already has that belief (‘Reply’, 222). Otherwise, it would be impossible for anyone to acquire a defeater for any belief at all. Plantinga also notes that appealing to L is dubious on the face of it: he suggests that it would be like the response of a theist who hears an argument against the existence of God on the basis of the problem of evil, and then says that although he can see that the existence of God is unlikely given the vast quantity of evil in the world, he is nevertheless prepared to continue believing in God because he thinks we have won the ‘Divinity lottery’, and there really is such a person as God after all (‘Reply’, 223). The objection above does not seem very compelling, but it does suggest a strategy for responding to EAAN. That strategy is to come up with a proposition, X, which the naturalist believes, and which is such that P(R ⁄ N&E&X) is high. This is not enough, however: we also want X to be such that, if P(R ⁄ N&E&X) is high, the naturalist who sees this does not have a defeater for R in virtue of seeing that P(R ⁄ N&E) is low or inscrutable. Following Plantinga, we will say that a proposition X is a defeater-deflector of
ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x

Evolutionary Argument Against Naturalism

83

EAAN just in case it satisfies both of the above conditions (‘Reply’, 224). An important question can now be expressed as follows: are there any defeater-deflectors of EAAN? If so, which ones? This is known as the Conditionalization Problem for EAAN (Plantinga, ‘Reply’, 220–25). A defeater-deflector of EAAN is a proposition that is naturally thought of as providing evidence for R. However, Plantinga and others hold that, in addition to propositional evidence, beliefs can also be warranted in virtue of nonpropositional evidence (Bergmann, ‘Commonsense Naturalism’,65). Such beliefs are typically held in the basic way, in that they are not based on any other beliefs. On this sort of view, for example, the experience I have when it appears to me that there is a computer in front of me is nonpropositional evidence, for me, that there is a computer in front of me. That may be so, you think: but what does this have to do with EAAN? Well, one belief that appears to be held in the basic way is R itself. It is an example of what Thomas Reid called first principles (Bergmann, ‘Commonsense’, 66). Now, denying such principles, thought Reid, elicits in us the emotion of ridicule, which serves the function of helping us to avoid doubt in these principles. Perhaps, then, the emotion of ridicule can, at least in some circumstances in which it is experienced, be seen as nonpropositional evidence for belief in R (Bergmann, ‘Commonsense’, 68).16 This leads to the following objection: The Nonpropositional Evidence Objection. ‘Even if R has low probability on all the available propositional evidence, the naturalist could still have nonpropositional evidence for R which makes it rational to continue to hold onto R. Hence, the naturalist need not have a defeater for R merely in virtue of accepting the Probability Thesis.’ One challenge to this objection comes from considering the XX case (Plantinga, ‘Reply’, 232). Recall that this is a widely cited analogy used to support the Defeater Thesis, and EAAN proponents typically assume that it is indeed analogous, in all relevant epistemic respects, to the case of the naturalist who accepts the Probability Thesis. In what follows, let us grant this assumption for the sake of argument.17 Now, suppose I think I have taken XX, and I have nonpropositional evidence for R. Am I rational in continuing to hold onto belief in R on the basis of this evidence? My own intuition says ‘No’. For I believe that, if the drug were to render R false in my case, I would still quite likely have exactly the same nonpropositional evidence for R (except that this evidence would then be misleading); if I reflect on this, I will surely have serious doubts about whether this nonpropositional evidence is trustworthy, and so I will not be able to escape defeat. By analogy, the EAAN proponent will argue, the naturalist does not escape defeat by virtue of nonpropositional evidence for R. 6. Tu Quoque Objections EAAN was put forward to show that theism has a certain advantage over naturalism. Some have challenged this claim, by arguing that, if EAAN is sound, then theism faces the same kind of self-defeat as that with which naturalism is charged in EAAN. Objections of this type generally go by the label ‘Tu Quoque Objection’. An early version of this is the following (Ginet, 407): The Austere Theism Objection. ‘The theist believes that there is a God: an omniscient, omnipotent and supremely benevolent being. This means that, unless his powers of inference are severely limited, the theist also believes what is entailed by theism, namely that there is a very powerful being. Call this latter belief austere theism, (henceforth, A). Now, the conditional probability P(R ⁄ A) is very hard to determine, and so the most plausible
ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x

84 Evolutionary Argument Against Naturalism

opinion about its value is that it is inscrutable, and hence low or inscrutable. Since every theist is an austere theist, it follows that any theist who sees that this probability is low or inscrutable has a defeater for R.’18 An obvious rebuttal to the Austere Theism Objection is that the theist surely believes much more than just austere theism (Ginet, 407). If we let T stand for traditional Christian theism, then T will include a range of beliefs besides A. Perhaps, then the theist is saved from self-defeat in virtue of the fact that P(R ⁄ T) is high, even if P(R ⁄ A) is inscrutable. This response sounds promising, but there are two problems with it. To begin with, the naturalist can appeal to similar considerations, by claiming that he too believes more than N&E, that relative to his expanded set of beliefs the probability of R is very high, and that this saves him from the defeat with which he is threatened (Ginet, 407). Unless there is a reason to think that this kind of move is acceptable for the theist but not the naturalist, EAAN will fail to secure a distinctive advantage for theism over naturalism. The second problem is that of coming up with a good reason for thinking that P(R ⁄ T) is high. One strategy is that of appealing to the perfection and goodness of God: it might be inferred, from the premise that God exists and has these attributes, that he has, most likely, given us faculties which yield mostly true beliefs. This is certainly a natural thought; it is reminiscent of Descartes’ claim that a perfect God is not a deceiver (Roeber).19 One problem with this inference is that it relies on the presupposition that God does not have a morally sufficient reason for creating us with unreliable faculties. This presupposition is initially plausible, because it is hard to think of such a reason. However, it can come to seem doubtful in the light of responses to the problem of evil that are widely endorsed by theists, and which are initially plausible in their own right. Let’s examine this in some more detail. Theists have rejected some forms of the evidential argument from evil on the grounds that, for all we know, God has morally sufficient reasons for permitting horrible evils, even if we can’t tell what these are. This approach to the argument from evil goes by the label ‘skeptical theism’ (Bergmann, ‘Skeptical Theism and Rowe’s New Evidential Argument from Evil’, 278). We can characterize this as the conjunction of theism with the claim that we have no good reason to suppose that the goods, evils, and connections between goods and evils of which we are aware, are representative of the goods, evils, and connections between goods and evils that there actually are (Bergmann, ‘Skeptical Theism’, 279). As Bergmann points out, this second conjunct has considerable initial plausibility, even independently of theism, since ‘it wouldn’t be the least bit surprising if [axiological] reality far outstripped our understanding of it’ (‘Skeptical Theism’, 284).20 Return to the claim that there is no morally sufficient reason for God to create us with unreliable faculties. How could we be justified in accepting this claim? If skeptical theism is correct, then there could, for all we know, be some great good, of which we are unaware, but which is inextricably connected with our being massively unreliable, and which gives God a morally sufficient reason to make us that way (Roeber, 317–19). To the extent that skeptical theism seems reasonable, the plausibility of this claim is undermined, and hence so is the inference from God’s goodness to our reliability. Plantinga’s strategy for arguing that P(R ⁄ T) is high is different from the one we have just been considering, in that he appeals to the religious tenet that ‘God is the premier knower, and has created us human beings in his image, an important part of which involves his endowing them with a reflection of his powers as a knower’ (Warrant and Proper Function, 236); thus, the Christian has reasons to think that humans have knowledge. But a belief can only count as knowledge if it is produced by a reliable
ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x

Evolutionary Argument Against Naturalism

85

belief-forming faculty.21 And thus the theist has no reason to doubt R that corresponds to the EAAN defeater (Warrant and Proper Function, 236). One challenge to this argument stems from the fact that it is a part of theism that human nature can be corrupted; even if humans were originally made in the image of God, all kinds of things could happen to them which lead to their cognitive malfunctioning. For example, such malfunctioning could be caused by the free actions of created agents, whether these are humans or supernatural agents like Satan (Lehrer, 29); and God might permit this malfunctioning to occur either because of the intrinsic value of creaturely free will or, if skeptical theism is plausible, because of morally sufficient reasons that are beyond our ken (Roeber, 318). At any rate, this second strategy for arguing that P(R ⁄ T) is high will be convincing only if such possibilities can be ruled out. The reflections above suggest something like the following argument. The Satanic Deception Objection.22 ‘What is P(R ⁄ T)? To evaluate this, let us consider a hypothetical population on a planet a lot like earth: this will help us to avoid any undue bias arising from our prior conviction that our own cognitive faculties are reliable. Assuming that these creatures have and change beliefs, that God exists, and that something roughly similar to the Christian story is true of them, what is P(R ⁄ T) specified to them? ‘Let S be the following statement (S) On the basis of morally sufficient reasons which are beyond our ken, God allows Satan to create incredible amounts of deception among these creatures, for a significant period of time, leading to the massive malfunctioning of their cognitive faculties during that time. Then the probability calculus yields the following formula: PðR/TÞ ¼ PðR/SÞPðS/TÞ þ PðR=SÞPð S/TÞ: Now, P(R ⁄ S) is surely very low. And, since theists agree that God has allowed Satan and humans to create all kinds of evil on the earth, it is hard to see how S can be ruled out, given theism. In the light of our cognitive limitations, and the infinite gap between our knowledge and that of God, we do not know how likely it is that God has morally sufficient reasons to allow Satan to deceive these creatures. Thus, P(S ⁄ T) is inscrutable, and so P(R ⁄ T) is inscrutable as well. Since this is true with respect to the hypothetical population, it must be true with respect to us as well. ‘Note that if the theist sees this, she will have a defeater for R, on the basis of exactly the same reasons that lead the EAAN proponent to conclude that the naturalist has a defeater for R.’ This argument, if successful, shows that theism can lead to the same kind of universal skepticism as that with which the naturalist is threatened in EAAN, assuming the latter argument is sound. Does the EAAN proponent have a good reply? One response might be to develop a version of EAAN in which P(R ⁄ N&E) is shown to be low, rather than just low or inscrutable.23 If there is no corresponding reason to think that P(R ⁄ T) is low, then theism appears to escape facing a problem of self-defeat in exactly the same way that naturalism does. Some sort of Tu Quoque response might still be made, but it would not be as natural or straightforward as the one we have considered above. Can there be an argument to the effect that P(R ⁄ N&E) is low, rather than low or inscrutable? In more recent work, Plantinga has already developed such an argument (Plantinga, ‘Self-Profile’).24 Is there a natural Tu Quoque Objection to this new argument? That seems to me an important question.
ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x

86 Evolutionary Argument Against Naturalism

Another response would be to appeal to further beliefs endorsed by Christian theology to show that the Christian theist need not have a defeater for R. For example, Plantinga argues that the Christian will believe that, by virtue of her accepting Christian faith, the action of the Holy Spirit within her is progressively ameliorating the effects of sin on her human nature, and is restoring her to the image of God in which she was created, a part of which involves her reflecting God’s powers as a knower and having reliable cognitive faculties (‘Respondeo’,337). This kind of response concedes that God could allow vast amounts of deception to occur among humans, while nevertheless emphasizing that those who accept Christianity are saved from this fate. 7. Further Directions for Research Much discussion of EAAN is polemical, and focused on proposing rebuttals to the argument, or proposing counter-rebuttals to those rebuttals. Little consensus seems to be emerging in this discussion, and so, I think, the challenge of EAAN is still with us (Mirza, 126). Where do we go from here? I suggest that, rather than merely attacking or defending EAAN, philosophers spend more time trying to carefully and systematically understand the intuitions that make the argument so compelling to its proponents. This approach is familiar from other topics in philosophy. For example, most epistemologists want to refute skepticism, but they are also interested in explaining why it seems compelling to so many. In fact, very often a refutation of skepticism will be seen as unsatisfactory if it seems too easy, that is, if the refutation makes it seem a mystery that anyone would be worried about skepticism in the first place (Reed, ‘A New Argument for Skepticism’). Ideally, a refutation of EAAN will meet a similar standard: such a refutation should also yield some insight into why the argument, when properly understood, is intuitively compelling to at least some philosophers. The Defeater Thesis is at the heart of EAAN, and the most widely cited defense of this thesis involves an analogy with the XX case. But few philosophers have tried to challenge or investigate the intuition that is meant to be elicited by the XX case, namely the judgment that the subject in that case has a defeater for R. It is generally just granted, even if only for the sake of argument, that this intuition is sound. I conjecture that a deeper understanding of this case will lead to one of two results: either the intuition will be rejected, in which case the Defeater Thesis will be undermined; or else we will be able to determine which epistemic features of the XX case best explain the intuition. In the latter case we can investigate whether or not the EAAN case has the very same features, and hence whether or not it is genuinely analogous to the XX case. Acknowledgement Thanks to Kevin Sharpe and an anonymous reviewer for helpful comments. Short Biography Omar Mirza is Associate Professor of Philosophy at St Cloud State University in St Cloud, Minnesota. He has been a Visiting Fellow at the Center for the Philosophy of Religion at the University of Notre Dame, and has previously held posts at George Washington University, Washington D.C. and Luther College in Decorah, Iowa. Omar holds a doctorate in Logic and the Methodology of Science from U.C. Berkeley, and his interests lie in epistemology, logic, and the philosophy of religion; his work
ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x

Evolutionary Argument Against Naturalism

87

on the evolutionary argument against naturalism has recently been published in Philosophical Studies, and in the latest edition of A Companion to Epistemology (ed. Dancy, Sosa, Steup). Notes
* Correspondence: Department of Philosophy, St Cloud State University, Centennial Hall 365, 720 4th Avenue South, St Cloud, MN, 56301, USA. Emails: oamirza@stcloudstate.edu and omarazizmirza@gmail.com. This is not to say that Plantinga wishes to criticize the theory of evolution itself. That theory, he says, is perfectly compatible with theism. Rather, it is the conjunction of naturalism and evolutionary theory that Plantinga finds to be self-defeating, under certain conditions (‘Introduction’, 1). 2 This means that there is no self-defeat problem that arises for theism in the same sort of way as it does for naturalism in EAAN; Plantinga’s arguments on this point leave open the possibility that there is some other, very different, way in which theism is self-defeating. 3 Note that it is not a part of the notion of defeater that a defeater itself has to be rationally held. A defeater can be as irrational a belief as one pleases. For example, suppose that I acquire the belief that there exists a Cartesian evil demon, and I acquire it by way of misunderstanding Descartes: then my belief ‘There exists a Cartesian evil demon’ is a defeater, for me, of my belief that I am typing this at my computer right now, even though it is wholly irrational; the reason is that given that I believe in a Cartesian evil demon, it is irrational for me also to believe that I am typing. Alston offers criticisms of Plantingas suggestion that an irrational belief can function as a defeater (‘Plantinga, Naturalism, and Defeat’, 186–95), and Plantinga replies in the same volume (‘Reply’, 274). 4 As will become clear, these ‘easily realized conditions’ are that the naturalist should accept evolutionary theory, and the Probability Thesis (to be described below). 5 Strictly speaking, this should read ‘The beliefs produced by human cognitive faculties are for the most part true over some significant range of their operation’. This qualification is necessary, because there will be various domains relative to which human cognitive faculties are unreliable (Plantinga, Warrant and Proper Function, 216–17). 6 What exactly is this defeater? It is not N&E, but rather the conjunction of N&E with the Probability Thesis (Plantinga, ‘Reply’, 206). 7 This is not intended to be a prescription for how the naturalist who accepts the Probability Thesis ought to change his or her beliefs. It merely indicates the belief changes to which the naturalist is rationally committed by virtue of belief in naturalism; if these belief changes seem absurd, note that EAAN is akin to a reduction ad absurdum of naturalism. If there is a prescription involved in EAAN, it is that one should not be a naturalist. 8 Objective probability is analyzed by Plantinga in terms of logical probability (Warrant and Proper Function, 162). 9 This is one way to understand Quine’s statement that ‘creatures inveterately wrong in their inductions have a pathetic but praiseworthy tendency to die before reproducing their kind.’ (126) 10 Stephen Stich, for example, has challenged the view that our reliability is in any way guaranteed by our having survived and evolved (56). 11 Clearly, Plantinga thinks that materialists should hold that beliefs are neural states or processes. In private communication, he has mentioned that some version of EAAN will still be cogent even on other materialist accounts of what beliefs are. 12 This last possibility is normally discussed by philosophers of mind under the heading ‘semantic epiphenomenalism’. 13 The appropriateness of this estimate can be illustrated by putting some actual numbers into the formula and computing the result for P(R ⁄ N&E). Of course, trying to come up with accurate numerical values for these probabilities is absurd, but the fiction that we have such values can help in understanding why the Probability Thesis is defensible. So let us say that P(C ⁄ N&E) = 0.7, and therefore that P(C ⁄ N&E) is 0.3; let P(R ⁄ N&E&C) = 0.2. Then, even if we set P(R ⁄ N&E&C) to the value of 1 (a concession, since Plantinga has argued this is at best moderately high), P(R ⁄ N&E) comes to 0.44, which is less than ½, and is hence low. 14 In more recent work, Plantinga has given a new argument for the Probability Thesis which supports the stronger conclusion that P(R ⁄ N&E) is low, rather than the conclusion that it is low or inscrutable; he writes that ‘I’d like to give an argument for [the Probability Thesis], and argument somewhat different from (and, I hope, superior to) the arguments I’ve given for it elsewhere.’ (‘Self-Profile’, 175). However, most presentations of EAAN defend the Probability Thesis along the lines described in this paper, and hence that has been our focus. 15 One way to see this is to note that the defeater for R that the naturalist is said to acquire in EAAN is plausibly supposed to be an undercutting defeater (which works by casting doubt on the trustworthiness on the source of the belief R), while there is no plausible way to construe the naturalist as having an undercutting defeater for F in virtue of seeing that P(F ⁄ N&E) is low or inscrutable. This response to the Perspiration Objection is developed in detail by Omar Mirza (‘A User’s Guide to the Evolutionary Argument Against Naturalism’).
ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x
1

88 Evolutionary Argument Against Naturalism
16

This idea will initially seem counterintuitive to most people. Baron Reed objects to it (‘Epistemic Circularity Squared? Skepticism About Common Sense’), and Michael Bergmann replies to the objection in the same journal (‘Epistemic Circularity and Common Sense: A Reply to Reed’). 17 This assumption is typically ignored or conceded in discussions of EAAN, but it might be viewed as controversial, as a reviewer for Philosophy Compass has rightly pointed out. In the conclusion of this paper, I note the importance of investigating the intuitions elicited by the XX case, as a preliminary to determining whether or not this case is genuinely analogous to the case of the naturalist in EAAN. 18 The Tu Quoque Objector is, of course, not committed to this claim. He is trying to show that, if EAAN is sound, the EAAN proponent is committed to this conclusion, and hence is not in a position to claim that theism has a serious advantage over naturalism. 19 To the best of my knowledge, Plantinga does not appeal to this inference anywhere. However, not only is this a natural first thought about how the theist could argue that P(R ⁄ T) is high, but it is also a good way to introduce themes that will be used in stating the next version of the Tu Quoque Objection, and hence I begin with it. 20 The viability of this kind of approach is challenged by Richard Gale (208). 21 This follows from Plantinga’s own analysis of knowledge, and also from other externalist accounts of knowledge (Warrant and Proper Function, 17). 22 I am not aware of anyone presenting the Tu Quoque Objection in exactly this form before, but I have put it together by combining ideas from a number of commentators, especially Fales, Lehrer, and Roeber. 23 All versions of the Tu Quoque Objection so far have argued that the probability of R on the relevant beliefs of the theist is inscrutable, rather than low. Roeber, for example, calls his argument the ‘Inscrutability Argument Against Theism’. 24 This is beyond the scope of this paper.

Works Cited
Alston, William. ‘Plantinga, Naturalism, and Defeat.’ Naturalism Defeated? Essays on Plantinga’s Evolutionary Argument Against Naturalism. Ed. James Beilby. Ithaca, London: Cornell University Press, 2002. 176–203. Bergmann, Michael. ‘Skeptical Theism and Rowe’s New Evidential Argument from Evil.’ Nous 35. (2001): 278– 96. ——. ‘Commonsense Naturalism.’ Naturalism Defeated? Essays on Plantinga’s Evolutionary Argument Against Naturalism. Ed. James Beilby. Ithaca, London: Cornell University Press, 2002. 61–90. ——. ‘Epistemic Circularity and Common Sense: A Reply to Reed.’ Philosophy and Phenomenological Research 73.1 (2006): 198–207. Dancy, Jonathan, Ernest Sosa, and Matthias Steup, eds. A Companion to Epistemology. Malden, MA: Wiley-Blackwell, 2010. Fales, Evan. ‘Darwin’s Doubt, Calvin’s Calvary.’ Naturalism Defeated? Essays on Plantinga’s Evolutionary Argument Against Naturalism. Ed. James Beilby. Ithaca, London: Cornell University Press, 2002. 43–58. Gale, Richard. ‘Some Difficulties in Theistic Treatments of Evil.’ The Evidential Argument from Evil. Ed. Daniel Howard-Snyder. Bloomington, Indianapolis: Indiana University Press, 1996. 206–18. Ginet, Carl. ‘Comments on Plantinga’s Two-Volume Work on Warrant.’ Philosophy and Phenomenological Research 55.2 (1995): 403–8. Lehrer, Keith. ‘Proper Function vs. Systematic Coherence.’ Warrant in Contemporary Epistemology: Essays in Honor of Alvin Plantinga’s Theory of Knowledge. Ed. Jonathan L Kvanvig. Lanham: Rowman and Littlefield, 1996. 25–46. Mirza, Omar. ‘A User’s Guide to the Evolutionary Argument Against Naturalism.’ Philosophical Studies 141 (2008): 125–46. Plantinga, Alvin. Warrant and Proper Function. New York, Oxford: Oxford University Press, 1993. ——. ‘Naturalism Defeated.’ 1994. Michael Sudduth’s Analytic Philosophy of Religion Website. <http://philofreligion. homestead.com/files/alspaper.htm>. ——. ‘Respondeo.’ Warrant in Contemporary Epistemology: Essays in Honor of Plantinga’s Theory of Knowledge. ed. Jonathan L. Kvanvig. Lanham: Rowman and Littlefield, 1996. 307–78. ——. Warranted Christian Belief. New York, Oxford: Oxford University Press, 2000. ——. ‘Introduction.’ Naturalism Defeated? Essays on Plantinga’s Evolutionary Argument Against Naturalism. Ed. James Beilby. Ithaca, London: Cornell University Press, 2002a. 1–12. ——. ‘Reply to Beilby’s Cohorts.’ Naturalism Defeated? Essays on Plantinga’s Evolutionary Argument Against Naturalism. ed. James Beilby. Ithaca, London: Cornell University Press, 2002b. 204–75. ——. ‘Self-Profile.’ A Companion to Epistemology. Ed. Jonathan Dancy, Ernest Sosa, and Matthias Steup. Malden, MA: Wiley-Blackwell, 2010. 173–8. Quine, Willard V. O. ‘Natural Kinds.’ Ontological Relativity and Other Essays. Ed. Willard V. O. Quine New York, London: Columbia University Press, 1969. 114–38. Reed, Baron. ‘Epistemic Circularity Squared? Skepticism About Common Sense.’ Philosophy and Phenomenological Research 73.1 (2006): 186–97.
ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x

Evolutionary Argument Against Naturalism

89

——. ‘A New Argument for Skepticism.’ Philosophical Studies 142 (2009): 91–104. Roeber, D. Blake. ‘Does the Theist Have an Epistemic Advantage over the Atheist? Plantinga and Descartes on Theism, Atheism, and Skepticism.’ Journal of Philosophical Research 34 (2009): 305–28. Stich, Stephen. The Fragmentation of Reason. Cambridge: MIT Press, 1990. Talbott, W. J. ‘The Illusion of Defeat.’ Naturalism Defeated? Essays on Plantinga’s Evolutionary Argument Against Naturalism. Ed. James Beilby. Ithaca, London: Cornell University Press, 2002. 153–64. Wielenberg, Erik. ‘How to be an Alethically Rational Naturalist.’ Synthese 131 (2003): 81–98.

ª 2011 The Author Philosophy Compass ª 2011 Blackwell Publishing Ltd

Philosophy Compass 6/1 (2011): 78–89, 10.1111/j.1747-9991.2010.00372.x

Sign up to vote on this title
UsefulNot useful