*Forthcoming in The International Journal for Philosophy of Religion, 1999

MICHAEL CZAPKAY SUDDUTH

In his forthcoming Warranted Christian Belief (hereafter, WCB), Alvin Plantinga develops a case for the positive epistemic status of theistic and Christian belief. One of Plantinga’s primary claims in WCB is that theistic belief can be warranted, and if true constitute knowledge, even in the absence of propositional evidence. This model of the proper basicality of theistic belief is an externalist one, employing the concepts of warrant and proper function that were the focus of his previous two volumes in the trilogy. Like his earlier defense of the deontological rationality of theistic belief, the externalist model appears to stand in sharp contrast to the classical evidentialist demand for evidence, traditionally based on internalist epistemological assumptions. But does it? Plantinga’s introduction of what he calls a "defeater system" as an essential part of the human cognitive design plan suggests two important internalist constraints within Plantinga’s otherwise externalist epistemology of religious belief. In this paper I unpack these constraints and investigate their evidentialist implications. If my argument is sound, then Plantinga’s epistemology entails two important internalist conditions for warrant, as well as circumstances in which evidence is necessary for warranted theistic belief.

In Warrant and Proper Function Plantinga argued that a belief has warrant, roughly, just if it is produced by cognitive faculties functioning properly in a congenial environment according to a design plan successfully aimed at truth. Fundamental to proper function is the idea of a cognitive design plan - a set of blue prints or specifications for a well-formed, properly functioning human cognitive system. Since the kind of specifications relevant for warrant are truth-oriented, they are specifications for that segment of the cognitive design plan that has as its purpose the production of true beliefs (as opposed to non-alethic purposes, such as survival or relief from suffering). The design plan specifies what the appropriate doxastic response of our cognitive faculties should be in a wide range of circumstances to achieve this alethic goal in a reliable manner.

The position here is externalist since the proper functioning of one’s cognitive system is not the sort of thing to which a person has introspective access or could come to know just by reflecting on one’s mental states. Viewed this way warrant entails a kind of externalist rationality, which should be distinguished from Plantinga’s earlier notion of deontological justification or the rationality associated with being within one’s intellectual rights. A belief is externally rational if it is a deliverance of one’s rational faculties, produced by properly functioning (and unimpeded) cognitive faculties successfully aimed at truth. I will refer to this as proper function rationality (hereafter PF-rationality).¹ A belief will lack warrant if it is PF-irrational, and a belief is PF-irrational if it is produced by (i) cognitive malfunction, (ii) properly functioning cognitive faculties overridden by certain emotions, or (iii) properly functioning cognitive faculties not aimed at the production of true beliefs. Since a person could satisfy any of these conditions without violating any epistemic duties, it is possible to be deontologically rational without being PF-rational.

Plantinga argues in WCB that if theism is true, then it is likely that our cognitive design plan includes specifications for holding theistic belief(s). In this way, Plantinga unpacks the epistemological implications of a theistic metaphysics. More specifically, he argues that if theism is true, then it is likely that:

By "circumstances" I will understand any of the widely realized experiential conditions that Plantinga points out in chapter 7 of WCB, and which have been the stock and trade of his proper basicality thesis since the early 1980s. These include things like the starry night sky, the crashing waves of the ocean, the majestic grandeur of the Vermont Mountains, and the lovely melody of a Bach Concerto. Plantinga maintains that these sorts of circumstances trigger the formation of various kinds of theistic beliefs: God is present, is powerful, is forgiving, etc. (all of which self-evidently entail God exists). By "relevant" cognitive faculties I refer to Calvin’s sensus divinitatis, which Plantinga takes to be a natural faculty or mechanism that is responsible for producing various theistic beliefs. More specifically, "the sensus divinitatis is a disposition or set of dispositions to form theistic belief in various circumstances, in response to the sorts of conditions or stimuli that trigger the working of this sense of divinity."² The experiential circumstances, of course, are not evidences that are taken as premises from which theistic beliefs are derived by a process of inference. The circumstances are simply occasions that trigger the formation of theistic belief, so theistic belief is basic (i.e., not held on the evidential basis of other beliefs). Plantinga further thinks that when the sensus divinitatis is functioning properly we will hold a firm theistic belief.

More precisely, then, if a person’s (relevant) cognitive faculties are functioning properly and she is in any of the widely realized experiential circumstances (specified by the design plan), she will firmly hold some theistic belief. Whereas in his earlier epistemology Plantinga had maintained that in these circumstances a person’s theistic belief is typically justified or rational (in the deontological sense), the claim here is that such beliefs are PF-rational and warranted, and if true, also constitute knowledge.

The design plan for humans also includes what Plantinga calls a defeater system, a cognitive subsystem that is designed to regulate modifications in a person’s noetic structure given new experiences and the acquisition of new beliefs which come with social exposure, mental maturation, and education. In short, these are specifications as to the correct or proper ways of changing beliefs in response to experience (doxastic and otherwise).³ Roughly, if a person acquires a defeater for a belief B, then there ought to be a certain kind of revision in the person’s noetic structure. (The "ought" here is the "ought" of proper function, not obligation). In some cases the person ought not to hold B with the same degree of firmness (partial defeat). In other cases the proper response is to withhold B (complete defeat), that is either hold the denial of B or hold neither B nor its denial. A belief D is a defeater for a belief B of some person S just if, given S's noetic structure, S cannot rationally hold B (at least not as firmly) given that S also holds D.⁴

Following a distinction introduced by John Pollock, Plantinga distinguishes between rebutting and undercutting defeaters.⁵ In the first case, a person acquires reasons for holding a belief incompatible with some belief B. I see in the distance what appears to be a sheep in the field and form the belief that there is a sheep in the field. The owner of the field then comes along and tells me that there are no sheep in the field, but he adds that he does have a dog that looks like a sheep from a distance. Here I have acquired, via testimony, a rebutting defeater for the belief that there is a sheep in the field. I have acquired an overriding reason for supposing that there is no sheep in the field. Although in other circumstances it would be PF-rational for me to hold my belief that there is a sheep in the field, in this circumstance it is no longer PF-rational for me to hold that belief (at least not with the same degree of firmness). In the case of an undercutting defeater, a person merely loses his reason for holding B. A person enters a factory and sees an assembly line on which there are a number of widgets that appear red. Being appeared to red-widgetly, the person forms the belief that there are red widgets on the assembly line. The shop superintendent then informs the person that the widgets are being irradiated by an infra red light, thereby enabling the detection of otherwise undetectable hairline cracks. Here the person merely loses his reason for supposing that the widgets are red, as opposed to acquiring a reason for supposing that they are not red. Both of these cases present us with a person (i) who holds some belief B at time t1, (ii) whose noetic structure undergoes modification with respect to B at time t2 after it comes to include the additional belief D, and (iii) whose relevant cognitive faculties (we correctly judge) are functioning properly at t1 through t2.⁶

The defeater system is aimed at the production and sustenance of true beliefs and the avoidance of false beliefs. Given the reliabilist constraint in Plantinga’s epistemology, and its connection with objective probability, we might say that the objective (statistical) probability of there being a red object in front of a person is high when being appeared to redly. But if one is both appeared to redly and informed by an authority that the object in question is being irradiated by a red light, the objective (statistical) probability of the object being red is not high.⁷ So the proper function requirement for warrant extends to the proper functioning of one’s defeater system, what we might call defeater-system PF-rationality. A person S’s belief B is warranted only if S is PF-rational in holding B (to the degree that S does), and S is PF-rational in holding B (to the degree that S does) only if the relevant portions of S’s defeater system are functioning properly. In which case the appropriate revisions will take place with the acquisition of any defeaters against B. If there was no noetic modification in the above cases, then the person’s cognitive state would be PF-irrational (to some degree). The failure to revise B given the acquisition of a defeater indicates a malfunction in one’s defeater system. Consequently, B would lack warrant (at least to some degree).

In Plantinga’s account of rationality defeaters, the conditions that serve to defeat a belief are typically other beliefs of the person, though Plantinga also allows experiences to count as defeaters.⁸ Since a person’s own beliefs and experiences are the sort of thing to which she has cognitive or introspective access upon reflection, rationality defeaters are internalist in nature. In the case of "beliefs," the defeating conditions would be cognitive items that fall within the believer’s perspective of the world.⁹ It is important to distinguish these internalist defeaters from defeaters construed merely as some external fact about the subject’s environment or cognitive situation that negatively affects the warranted status of a person’s belief or which prevents an overall justified true belief from counting as knowledge.¹⁰ Consider the situation in which mischievous teenagers place several fake sheep (made out of plastic and fur) on a hill. Driving by I form the belief that there is a sheep on the hill. It so happens that I form this belief while looking at what is in fact a real sheep that has wandered into the company of the fake ones in the effort to find a mate. Although the mere existence of fake sheep does not make it PF-irrational for me to believe that there is a sheep on the hill, it does prevent the belief from having much by way of warrant for me. So I do not know that there is a sheep on the hill. If I believed that local teenagers had contrived a fake-sheep-on-the-hill plot, then I would have a rationality defeater.

Plantinga recognizes both rationality defeaters and warrant defeaters, and though they are distinct they have an important relation. A person’s belief will fail to be warranted if produced by cognitive faculties that are not functioning properly, not aimed at truth, or if there is something awry in one’s epistemic environment. Since these are all design plan specifications to which a person does not have introspective access, they are externalist defeaters. What gets defeated is the warrant a belief has (and so the belief is not an item of knowledge). But if I acquire a rationality defeater for my belief B and the design plan specifies that I hold B less firmly, but I continue to hold it with the same degree of firmness, my belief is not externally or PF-rational. So my belief will lack warrant for externalist reasons that involve the failure of my cognitive system to respond properly to certain internalist conditions, namely some new experience or the acquisition of some new belief(s). These internalist rationality defeaters, then, (because they defeat what can rationally be believed in the sense of proper function) can also be externalist warrant defeaters (though not all warrant defeaters are rationality defeaters).¹¹ This highlights a more general point about internalist rationality in relation to warrant. A belief is internally rational just if it is the appropriate doxastic response to what is given to a person by way of her previous beliefs and current experience. Warrant (at least to a degree sufficient, along with true belief, for knowledge) requires both internal and external rationality.

So Plantinga holds to a negative internalist condition that is necessary for proper function and warrant. Call it the no-defeater condition:¹²

It follows from [ND] that if a person acquired an undefeated defeater for theistic belief, then holding theistic belief (or doing so firmly) would be PF-irrational and the belief would fail to be warranted (and hence would also fail to be knowledge).

Consider a case not too far removed from one that Plantinga himself considers in a few places.¹⁴ Lisa has been raised in a Christian family. During her youth she holds her theistic belief in a basic way, but in her later teenage years her theistic belief isn’t as strong. The cares of college life, sexual indulgence, and late-night parties slowly erode her thoughts of God. While a senior in college, she is exposed to Sigmund Freud’s idea of wish fulfillment. She becomes convinced that the belief she had in an invisible friend called Merlin while a young girl was one such belief, a convenient defense mechanism against the hostile forces of her childhood environment. Upon further reflection, though, she concludes that her belief in God is significantly analogous to the belief she once had in Merlin. So she comes to believe that (p) her belief in God is really the product of wish fulfillment. Her readings in Freud confirm this. Moreover, she believes that (q) the objective probability of a belief being true given that it is produced by wish fulfillment is either low or inscrutable. (I say "or inscrutable" here because perhaps she is simply agnostic about the probability of a belief being true given that it was produced by wish fulfillment, rather than estimating that probability to be low). She then believes that (r) the objective probability of her theistic belief being true is either low or inscrutable. Lisa has acquired an undercutting defeater for her theistic belief. If her defeater is partial and itself undefeated (as might be the case if she didn’t hold either p or q very firmly), then the rational thing to do would be to hold her theistic belief less firmly than she did before acquiring this defeater. Perhaps her defeater is complete and itself undefeated (e.g., she has great enthusiasm for projective theories of religious belief or alternatively her theistic belief is very weak before believing p and q), then if she is PF-rational she will withhold her theistic belief.¹⁵

It seems, then, that Plantinga should maintain:

It is important to point out that [P2] is logically consistent with [P1]. What follows from [P1] is that if a person is in the relevant circumstances and holds a firm theistic belief or fails to withhold theistic belief, he suffers from some cognitive malfunction or impedance, or perhaps his cognitive state has been produced by something other than truth-aimed cognitive faculties. But in addition to circumstances C (that require firm theistic belief), there are plausibly circumstances C* that include having an undefeated defeater for theistic belief. Here, depending on the actual defeater, the appropriate doxastic response for a reasonable person will be holding a less than firm theistic belief T, holding the denial of T, or holding neither T nor its denial. The relevant analogy here can be taken from the conditions that govern the PF-rationality of sensory perceptual, testimonial, and memorial beliefs. If a circumstance includes being appeared to rainly, then I am PF-rational in holding the belief that it is raining outside, unless of course the relevant circumstance includes my having rationality defeaters for such beliefs. So although there are conditions in which holding theistic belief is PF-rational, there are also conditions in which holding theistic belief would not be PF-rational.¹⁷

In the case above Lisa holds a theistic belief and then acquires an undefeated defeater (the conjunction of p, q, and r) and PF-rationality requires that Lisa no longer hold the belief T or that she hold it with a less degree of firmness. But the question that immediately arises is this: given that a person acquires an undefeated defeater for her theistic belief, what is required for her to be PF-rational and warranted in holding theistic belief at some later time, after the acquisition of the defeater? I think the natural response here is to say that what is needed is a cognitive state in which the defeater no longer carries defeating force, either because other conditions now neutralize its defeating force, undermine it, or eliminate it altogether. In other words, the defeater must be defeated. What is needed is a defeater-defeater (to coin Pollockian/Plantingian terminology). Plantinga himself suggests this. Writing with reference to a theist who finds herself with an undercutting defeater due to reading too much Freud, he says: "if that defeater remains itself undefeated and if she has no other source of evidence, then the rational course would be to reject belief in God."¹⁸

Consider the range of possible defeater-defeaters Lisa could acquire.

One example of a defeater-defeater in Lisa’s case would be an undercutter that attacks the connection between [(p) and (q)] and (r). Suppose she comes to believe that (s) wish fulfillment is a natural mechanism that God has implanted in humans to act as a secondary cause in the production of theistic belief.¹⁹ In that case, even if (q) is true for most beliefs, the conjunction of (p), (s), and (q) would fail to be indicative of the truth of (r). So Lisa could retain the original defeating reasons but their defeating force would be neutralized by a relevant addition to her noetic structure. A second option would be having reasons for supposing that either (p) or (q) is false; that is, either it is not the case that theistic belief is produced by wish fulfillment or it is not the case that beliefs produced in that way have an objective probability that is low or inscrutable. For instance, perhaps upon reflection she discovers important disanalogies between her Merlin belief and theistic belief. Here she acquires a rebutting defeater for (p). A premise of the original defeater is attacked. Call this a reason-defeating defeater.²⁰ Alternatively, Lisa might have a rebutter for (r), a reason for supposing that the objective probability of theistic belief being true is neither low nor inscrutable. Perhaps Lisa acquires reasons for supposing that she has been created with a reliable mechanism for forming beliefs in God, that there is a mechanism that produces such a belief and that beliefs produced by it have a high objective probability of being true.

At this point, natural theology enters the picture. First, where the defeater against theistic belief is a reason for supposing that theism is false (say by way of an evidential argument from evil), a rebutter against this belief will be a reason for supposing that theism is true. For instance, a person who is agnostic about the existence of God at t1 because of an argument from evil may find at t2 that theism carries significant explanatory power for the existence of the Universe, its spatial and temporal regularities, and the degree of fine-tuning it exhibits. Secondly, notice the connection between this and the kinds of defeater-defeaters discussed in the previous paragraph. Recognizing the explanatory power of theism might provide Lisa with a reason for supposing that theistic belief is relevantly disanalogous to her Merlin belief (if she sees that the latter does not have explanatory power). Moreover, if a person acquired reasons for supposing that there is a God, she could thereby have grounds for supposing that she has been created with a cognitive design plan that includes the production of theistic beliefs. She might then have reasons for supposing that she has been created with a faculty or mechanism for producing true beliefs about her creator, and perhaps that when her faculties are functioning properly they reliably produce such beliefs. So natural theology would also be a way to acquire a rebutter to the conclusion of Lisa’s original undercutting defeater.

So whereas the classical evidentialist requirement for theistic belief (stated in terms of warrant) affirms

the defeater-defeater requirement (applied to theistic belief) affirms:

[ND] and [DD] together constitute two important internalist requirements in Plantinga’s otherwise externalist epistemology. Or, to be more technically accurate, it seems to be an entailment of Plantinga’s epistemology that any person who fails to satisfy the negative internalist requirement [ND] with respect to some belief B is under a second positive internalist requirement, the defeater-defeater requirement, with respect to B.

Several philosophers have drawn attention to the negative evidential significance of defeaters: they (i) remove (deontological) rationality (Plantinga, Wolterstorff),²² (ii) override prima facie (truth-conducive) justification (Alston),²³ and (iii) undermine knowledge (Goldman, Nozick).²⁴ The idea of a defeater-defeater requirement has also been developed by Philip Quinn and John Greco, though neither is concerned with that requirement as a condition for cognitive proper function.²⁵ More importantly, there remains the task of spelling out the positive evidential significance of defeater-defeaters.

In cases where a defeater-defeater is a rebutting defeater against some original rebutting defeater against theistic belief, the defeater-defeater will be evidence for the truth of theism. I have already noted this with reference to the problem of evil serving as a reason for supposing that theism is false. A defeater-defeater that rebuts this defeater will be a reason for supposing that theism is true (since it aims to rebut the claim that theism is false). Natural theology would be essential to this task. But not all defeater-defeaters will amount to evidence for the truth of theism, at least not directly. For instance, in the case of an undercutting defeater-defeater I will have overriding reasons for supposing that the premises of the argument from evil fail to be indicative of the truth of the conclusion "there is no God." Alternatively, in what I described earlier as a reason-defeating defeater-defeater, one may have overriding reasons for supposing that one of the premises of the argument is false. In each of these cases I lose my reason for supposing that theism is false, but I do not acquire a reason for supposing that theism is true. My defeater-defeater does not give me evidence for the truth of theism. But that is exactly what classical evidentialism demands by way of evidence.

I think that the resolution to this difficulty lies in drawing a distinction, two distinctions to be exact. First, there is the distinction between first-order beliefs B and second-order beliefs B* of the form <S’s belief B is Q>, where Q = some epistemic property (e.g., warrant, rationality, or justification). Secondly, there is the distinction between the kinds of evidence that support the truth of beliefs on each level. Defeater-defeaters (including undercutters), I want to argue, provide evidence for the truth of certain higher-level beliefs, and such evidence is epistemically significant for rationality and warrant at the lower level.

Rationality defeaters are closely related to internal rationality. Internal rationality, loosely speaking, involves things being epistemically right from the believer’s perspective as a knowing subject. A person who gets a rationality defeater comes to see (in a way appropriate to her level of conceptual development) that something significantly counts against the truth of B (or outweighs considerations in favor of its truth) to such a degree that holding B is no longer appropriate given the truth goal of believing. At any rate, she certainly has reasons for supposing this by virtue of having overriding reasons for supposing that a belief is false or inadequately grounded. But then it seems that a person who acquires a defeater acquires an overriding reason for supposing that holding B is epistemically irrational, irrational from the epistemic point of view.²⁶ If the subject acquires a defeater then she has reasons for supposing that things are not epistemically right, that continuing to hold B would run counter or against the epistemic goal of believing. In other words, when a person acquires a defeater, she acquires reasons that support a certain negative epistemic evaluation of her belief B.

What happens when the defeater gets defeated? On the internal view of rationality under consideration here, a defeater-defeater provides a person with reasons for supposing that holding B would be epistemically right or appropriate again. Since the defeater-defeater removes the grounds for S’s taking B to be epistemically irrational, it would seem that the defeater-defeater provides S with reasons for supposing that B is epistemically rational. This of course depends on S not having additional (defeating) reasons for supposing that B is epistemically irrational at that time. Equally, it would seem that the strength of the original defeater as well as the defeater-defeater is also important here. S might originally acquire fairly strong reasons for supposing that theism is false, such that S takes it on these grounds that holding theism and withholding theism are equally epistemically irrational. In that case, some defeater-defeaters (even rebutters) might only provide S with reasons for merely withholding theism (rather than holding theism or its denial). Stronger defeater-defeaters would provide S with reasons for supposing that holding theistic belief is epistemically rational only if S does not hold that belief very firmly. Moreover, we can add to the informal equation here that the ability of a belief to survive defeat can reasonably be thought to strengthen reasons for supposing that B is epistemically rational. So what we should say then is that defeater-defeaters provide a person with reasons for supposing that holding theistic belief (to some degree) is epistemically rational, provided that the defeater-defeater is strong enough relative to the original defeater and other epistemically relevant items at the time.²⁷ So, given these qualifications, when a defeater gets defeated, though one doesn’t necessarily have evidence for the truth of B, one does have evidence (in varying degrees) for the truth of a higher-level claim about the (restored) epistemic rationality of holding the original belief. As the original defeater provides one with reasons for supposing that there is a true negative higher-level epistemic proposition, acquiring a defeater-defeater for that defeater provides one with reasons for supposing there is some true positive epistemic proposition of the form <S is (now) epistemically rational in holding B>.

What we have is a kind of diachronic epistemic level ascent, which can be applied to theistic belief and represented as follows:

At t₁ {S is PF-rational (and warranted) in holding some theistic belief T}

* The italicized statement at the 1^st level indicates a sufficient condition for the obtaining of the statement in bold at the 1^st level. The italicized statements at the 2^nd level indicate necessary conditions for the obtaining of the statement in bold at the 2^nd level.

In terms of PF-rationality, according to [DD] defeating circumstances require a defeater-defeater for S to remain PF-rational in holding a theistic belief (subject to the qualifications discussed in note no. 21). Since warrant requires proper function, the defeater-defeater is also required for warrant. More technically (and following the diagram above), given the ascent to the 1^st level at t2, a move to the 2^nd level is necessary to be warranted in holding theistic belief at tn+2 (at least holding it with the same degree of firmness). Since the defeater-defeaters at the 2^nd level include reasons for supposing that certain (positive) higher-level epistemic propositions are true, reasons for the truth of theism are not the only relevant evidential factors in determining whether theistic belief is warranted or PF-rational. Evidence for the truth of certain higher-level beliefs is also relevant.

I think this captures an important internalist intuition at the heart of the evidentialist tradition. The intuition springs in part from the Enlightenment understanding of the human person as naturally reflective and the opposition to religious beliefs that are either not open to critical inquiry or are immune from rational argumentation or evaluation. But this internalism really suggests a more general epistemic desideratum, something like reflective rationality – the rationality associated with judgments about the epistemic status of one’s beliefs, where such judgments are the product of reflection on the adequacy of the grounds of one’s belief. Evidentialist requirements seem appropriate here, in relation to reflective rationality and beliefs that encapsulate epistemic evaluations of lower-level beliefs.²⁸ But this is higher-level activity. One of the mistakes of classical evidentialism was to apply such requirements without restriction at the lower level, rather than specify a limited range of circumstances that would require reflective rationality for warrant at the lower level. The internalist no-defeater condition provides an important criterion for this. Given the state of cognitive internal defeat, reflective rationality (to some degree) is necessary. The evidence, however, that is required for reflective rationality is not restricted to evidence for the truth of the lower-level belief but includes evidence for the truth of higher-level beliefs about the epistemic rationality of lower-level beliefs. So defeater-defeaters, whether they be rebutters or undercutters, are evidentially and epistemically significant.

According to [P2] theistic belief is not PF-rational or warranted in certain circumstances. But this conclusion is actually denied by Plantinga in WCB (see especially chapter 13). Although Plantinga would agree that something like [ND] is true, it would not follow that there are conditions in which a person can acquire a defeater for theistic belief. Acquiring a defeater for any belief, so his argument goes, really depends on what is specified by the design plan. In the case of theistic belief, it is reasonable to assume that no such provision is made. What about the existence of alleged defeaters? Plantinga’s response here is that alleged defeaters arise because of cognitive malfunction elsewhere in a person’s cognitive system, chiefly in the sensus divinitatis. Although there are always unintended consequences of how a thing will function if it is damaged, it doesn’t follow that this is part of its design plan. So also with theistic belief. Sin has in all likelihood damaged the sensus divinitatis, and one of the noetic effects of sin could be the generation of alleged defeaters for theistic belief. With reference to the problem of evil, Plantinga writes that a person "has a defeater only if it is part of our cognitive design plan to give up theistic belief in those circumstances; and we have no reason to think that it is. The design plan includes the proper function of the sensus divinitatis; how things actually go when that process does not function properly could be part of the design plan; more likely, though, it is an unintended byproduct rather than itself part of the design plan."²⁹

The claim that every instance of religious unbelief or defeaters for theistic belief is the result of cognitive malfunction is problematic, as is the related claim that the design plan likely makes no provision for defeaters for theistic belief. The main difficulty with these claims is that they entail a substantial difference between properly basic theistic belief and other paradigmatic properly basic beliefs, but such a difference is what Plantinga elsewhere seems to deny. In his earlier papers and in WCB (chapter 11) Plantinga argues that there are many properly basic beliefs that are susceptible to defeat, e.g., sensory perceptual, memory, and testimonial beliefs. What does it mean to say this except that the design plan specifies circumstances in which holding such beliefs would be PF-irrational? For instance, the design plan presumably specifies that in the simple circumstance of <being appeared to redly>, I form the belief that there is a red object present. But if this circumstance is complex and also includes <I have an overriding reason for supposing that my relevant sensory perceptual faculty is not reliable>, then PF-rationality requires that I withhold the belief. Similarly, I am PF-rational in believing that there are sheep on the hill if the owner of the land tells me so. But if I have good reason to believe that he is lying or have overriding reasons for supposing that in fact there are no sheep on the hill, then I will be PF-irrational in holding the belief he affirms to me.

Plantinga writes:

But how can theistic belief not be immune to defeat and yet the design plan not include specifications for withholding theistic belief? How can basic theistic belief have defeasibility parity with other paradigmatic properly basic beliefs and yet the design plan make no provision for withholding theistic belief? In fact, Plantinga has stated that "nearly any belief is possibly subject to defeat." The design plan probably excludes rationality defeaters for some beliefs, beliefs that have certain epistemic immunities (e.g., beliefs about one’s own mental states), be it immunity to doubt, error, or revision.³¹ But theistic belief doesn’t quite meet these criteria. Moreover, given that Plantinga takes internalist defeasibility as essential to rebutting the charge of fideism, if the no provision claim is true, then it seems that the charge of fideism will be correct; for the rebuttal to that charge rests on a parity argument that shows that theistic belief and other paradigmatic basic beliefs share the property of internalist defeasibility.

Moreover, even if we assume that a person has a defeater for theistic belief because of cognitive malfunction, this does not necessarily undercut the claim that withholding theistic belief can be PF-rational and a specification of the human cognitive design plan. In "Naturalism Defeated" (unpublished) Plantinga himself claims that an irrational belief can be a defeater. He cites an example of someone who believes, due to some canine-oriented cognitive malfunction, that his neighbor’s dog is intentionally trying to annoy him. Such a belief, he says, "can nevertheless function as a defeater for a previously held belief that dogs never intentionally try to annoy people."³² Now, an irrational belief is by definition one that is not PF-rational. Plantinga’s point, then, seems to be an admission that in some cases the specifications of the design plan include withholding belief for reasons that are themselves produced by cognitive malfunction (non truth-directed cognitive faculties, or faculties overridden by emotions of some sort). So the mere fact that a defeater for theistic belief arises because of irrationality elsewhere in one’s cognitive system would not be a sufficient basis for asserting that the design plan makes no provision for withholding theistic belief in some circumstances.³³

(b) Plantinga’s Intrinsic Defeater-Defeater Argument 

But Plantinga has a second move.

Supposing that we concede that it is possible for a person to acquire a defeater for theistic belief, couldn’t basic theistic belief, by virtue of its own degree of non-propositional warrant, serve as an adequate defeater-defeater against any alleged defeater against theism? In which case there would be no need for the kind of defeater-defeaters discussed earlier. The argument here builds on Plantinga’s distinction between intrinsic and extrinsic defeaters, originally introduced in an interesting dialogue between Plantinga and Philip Quinn.³⁴ In the former case, the belief that is the target of a potential defeater, by virtue of its degree of warrant, acts as a defeater-defeater against the potential defeater. In the latter case, the defeater-defeater would be reason to think that the potential defeater is false (or its ground is inadequate), but where this reason is independent of the belief that has been targeted for defeat by the potential defeater.

In "The Foundations of Theism: A Reply," Plantinga supports the existence of intrinsic defeater-defeaters with "the disappearing letter" scenario. I write a letter to the chair of my department trying to bribe him to write a highly exaggerated letter on my behalf for an NEH fellowship. The letter mysteriously disappears from the chairperson’s office. I have a motive to steal it, the opportunity to do so, and I have been known to do such things in the past. Moreover, a reliable member of the department claims to have seen me hanging around the chairperson’s office about the time the letter must have been stolen. Given the evidence, my colleagues believe that I stole the letter. However, I believe that I spent the day in the woods and so could not have stolen the letter. My memory belief has a great deal of nonpropositional warrant for me, and so, despite the counter-evidence, I am rational to believe that I did not steal the letter. Here it seems that the potential defeater is defeated by the (targeted) defeatee. By analogy, Plantinga claims that basic theistic belief can, by virtue of its own degree of nonpropositional warrant, function as an intrinsic defeater-defeater. It would seem, to use Plantinga’s example, that Moses’ belief that God was speaking to him from the Burning Bush would have more by way of warrant than a defeater in the form of either an atheological argument from evil or the projective explanation of theistic belief advocated by Feuerbach or the Freudian wish fulfillment equivalent. So there is no need to fall back on any defeater-defeater that is independent of the original basic theistic belief, contrary to what is suggested by [DD].

The idea of beliefs, theistic and otherwise, being intrinsic defeater-defeaters is an interesting one, but I think ultimately insignificant. The force of the disappearing letter case depends crucially on the strength of my belief that I spent the day in the woods. And this in turn could be adversely affected by my possessing overriding reasons for supposing that I suffer from memorial hallucinations, multiple personality disorder, or some other disorder that undercuts my belief in the reliability of my memory. If I did have such reasons I would have a defeater that could not be defeated by the defeatee. So how we describe the situation will determine whether we have an intrinsic defeater-defeater.³⁵ More importantly, as Philip Quinn pointed out in response to Plantinga, Moses’ situation is clearly extraordinary. Few of us would seem to have theistic beliefs that enjoy that degree of warrant. The range of degree of warrant that may plausibly be attributed to many (or most) cases of properly basic theistic belief (even if sufficient for knowledge) will preclude the possibility of that belief being an intrinsic defeater-defeater for most people under normal circumstances. This is especially true if we use degree of belief as a partial determinant of the degree of warrant. The fact that the firmness with which some people hold theistic beliefs is often (significantly) reduced by putative defeaters shows that on these occasions even if basic theistic belief is an intrinsic defeater-defeater it is so only for a limited range of potential defeaters. In fact, it would appear that the notion of intrinsic defeater-defeaters is not even confined to theistic belief, but every belief can serve as an intrinsic defeater-defeater against some (weak enough) defeater. The notion that theistic belief can be an intrinsic defeater-defeater is trivially true.

But perhaps the whole idea of intrinsic defeater-defeaters is really misleading. It seems that an intrinsic defeater-defeater is really not an instance of an acquired defeater being defeated at all. If a person holds a theistic belief, the degree of belief index provides an initial degree of insulation from some range of possible defeaters. But this is not to say that the basic theistic belief held at t1 is a defeater-defeater against a defeater acquired at tn+1. It is only to say that theistic belief (or any belief for that matter) carries with it implications for what can or will in the future count as a rationality defeater for the belief. Only beliefs with certain epistemic immunities (e.g., beliefs about one’s mental states) will exclude all possibility of defeat. They will have maximal insulation from defeat. Theistic belief, though it will set certain limitations on the conditions under which a defeater can be acquired, does not rule out a person’s acquiring a defeater for the belief. And if a person acquires a defeater, then the defeater must be defeated by some other belief (or experience), independent of the original belief (and its grounds). I think it must be said that no belief can act as an intrinsic defeater-defeater to an acquired defeater; rather, every belief sets constraints on the range of possible acquired defeaters. There will always be potential defeaters that are not actual ones. Perhaps this was Plantinga’s only point. But then the issue is whether a person can acquire a defeater for basic theistic belief, and it certainly seems an important part of Plantinga’s epistemology to allow that possibility.

I have argued that Plantinga's largely externalist epistemology of warrant and proper function (implicitly) contains at least two important internalist features generated by his notion of a "defeater system" as essential to proper cognitive function. This component in turn entails that, given defeating circumstances, the PF-rationality of theistic belief depends on a person’s having a defeater-defeater for the original defeater against theistic belief. The defeater-defeater requirement suggests that the rationality of theistic belief, even construed in an externalist manner, sometimes demands a degree of internalist reflective rationality whereby we acquire reasons for supposing that theistic belief has some epistemic excellence (e.g., is epistemically rational). To the extent to which reflective rationality contributes to the PF-rationality of theistic belief it also contributes to the warrant of theistic belief. Here the internalist intuitions of classical evidentialism are retained within an otherwise externalist epistemological theory. Given the acquisition of defeaters, whether one knows that God exists sometimes requires reflective rationality. Plantinga has provided not merely an externalist model of proper basicality but an epistemological theory that can accommodate the more reasonable intuitions of the classical evidentialist tradition.

Acknowledgments

Many thanks to John Zeis, William Alston, and Alvin Plantinga for either commenting on earlier drafts of this paper or discussing with me the arguments contained therein.