Dr. Michael Czapkay Sudduth

June 27, 2001 Draft

ABSTRACT: This chapter is part of a more detailed examination of the idea of immediate and non-inferential knowledge of God in the Reformed theological tradition. In the previous chapter I examined several prominent Reformed theological attempts to explain and defend the idea that the natural knowledge of God is immediate and not based on inference or argumentation. I concluded that such arguments at best establish a purely psychological thesis about the causal origin of belief in God which leaves open a variety of epistemic functions of inferential reasoning about God. In this chapter I consider some contemporary philosophical attempts to develop and defend the idea of immediate knowledge of God.

Contemporary Reformed Epistemology:

The Epistemic Immediacy of Theistic Belief

Dr. Michael Sudduth

The Reformed theologians examined in the previous chapter not only maintained that the source of belief in God is immediate, but that this psychologically basic theistic belief constitutes knowledge. The Reformed view of the natural knowledge of God is thus not merely a claim about psychological immediacy, but epistemic immediacy.

Immediate knowledge of God has been denied by various epistemological traditions in Western philosophy, from medieval Aristotelianism to the janus-faced descendents of Enlightenment empiricism. I will refer to these epistemological traditions under the general rubric of classical evidentialism.[1] In contrast to the Reformed position, classical evidentialism in its various has maintained that theistic belief is rational or constitutes knowledge only if a person bases his belief on evidence, has evidence for this belief, or at least that there is evidence available for it. But the tradition of evidentialism has come into sharp critical focus during the last 30 years or so, both in general epistemology and the philosophy of religion. Several prominent epistemologists have presented interesting and substantial challenges to evidentialism. In religious epistemology these challenges have come largely from Reformed epistemology,[2] the main representatives of which are William Alston, Nicholas Wolterstorff, and Alvin Plantinga. Plantinga and Wolterstorff are both in the Reformed theological tradition, so their work establishes an important connection between contemporary epistemology and the Reformed theological tradition. In fact, both Plantinga and Wolterstorff have regarded their epistemological work as in some respects an attempt to develop with greater philosophical precision the insights of the Reformed thinkers we have examined in the previous chapter. Their critique of evidentialism has been closely tied to the defense of the proper basicality of belief in God, roughly, the idea that psychologically immediate theistic belief can be rational, justified, and even constitute knowledge, independent of whether there is any evidential case available for theism.

As a first step toward a more thorough critical analysis of the epistemic immediacy thesis and its relationship to the role of natural theology, in this chapter I will develop the contrasting epistemologies of religious belief advocated by classical evidentialism on the one hand and Reformed epistemology on the other. More precisely, since Reformed epistemology developed largely as a critique of evidentialism, I will present the Reformed epistemology case against evidentialism. In the first two sections I’ll consider this in terms of the epistemic category of rationality or justification. In the remaining three sections, I’ll look at the idea of immediate knowledge of God. Although I will provide some periodic critical observations, I will reserve the following chapter for a thorough critical engagement of Reformed epistemology and evidentialism.

I. Reformed Epistemology and the Critique of Classical Evidentialism

Reformed epistemology initially developed as a critique of classical evidentialism. Roughly stated, evidentialism requires the possession or availability of adequate evidence for theistic belief to be rational, justified, or to constitute knowledge.[3] So, for instance, in one formulation the evidentialist position maintains that theistic belief is not rational unless it is based on adequate evidence. "Evidence" in this context is propositional evidence, reasons in the form of other (rational) beliefs or propositions that provide the right degree of support for the truth of theistic belief. The evidence must, of course, be theologically neutral (not presuppose theism), and it will be adequate evidence just if it either entails theism or makes theism probable. Given evidentialism, of course, natural theology has an important, indeed necessary, epistemic function. Reformed epistemology, however, attempts to show that evidentialism is crucially flawed and that theistic belief could possess certain positive epistemic statuses in the absence of evidence and argument, especially of the sort provided by natural theology.

If the Reformed epistemology critique of evidentialism is correct then, it would seem that natural theology is unnecessary for theistic belief to be rational or constitute knowledge.[4] Although this objection does not entail that the theistic arguments are in themselves bad or inappropriate, it does imply that a primary motive for engaging in the project of natural theology is unsound, and that the role of natural theology must be substantially rethought. Not surprisingly, many philosophers, including theistic ones, are uncomfortable with this move. It appears to generate an important weakening of the grounds for natural theology. Put more generally, it might be argued that if theistic belief can constitute immediate knowledge, there is little place left for inference or argumentation; at least as far as securing the positive epistemic status of theistic belief is concerned.

A. Strong Evidentialism

But what exactly is allegedly wrong with the evidentialist position according to Reformed epistemology? The problems associated with the evidentialist requirement for theistic belief are closely connected to the precise formulation of evidentialism and the epistemological assumptions on which evidentialism is based. Specific criticisms will vary therefore depending on the version of evidentialism in view.

As a starting point, though, we can begin by considering one of the standard formulations of the evidentialist requirement alluded to above.

[JE] A person S’s theistic belief T is justified only if it is based on propositional evidence that provides adequate support for T.[5]

Why suppose that [JE] is true? One epistemological premise from which [E] can easily be derived would what we might call the strong evidentialist principle:

[SE] Given any belief that p for any person S, S’s belief that p is justified only if it is based on propositional evidence that provides adequate support for p.

[SE] denies the idea that any belief is immediately justified. What is the argument for [SE]? I think there are primarily three.[6]

First, it might be argued that the very concept of justification implies the activity of justifying one’s belief. This activity obviously requires the giving or citing of reasons in the form of propositional evidence. And clearly one can’t do that unless such reasons are cognitively available. Therefore, justification requires reasons. While it is no doubt true that the activity of justifying one’s belief requires reasons, it does not follow that being justified requires reasons. There is an important distinction between the state of being justified and any activity by which one attempts to show that one is justified. Although I can’t show that I am justified without marshalling evidence in the form of argument, it is implausible to suppose that I must do all this to be justified. By parity of reasoning, I can be in the state of having received a letter from Richard Swinburne without showing that I am in this state. And thus what is required to show it need not be required for the state. Indeed, when I show that I am in some state X, my being in X is presupposed. The latter is the more basic concept. Similarly, then, I can be in the state of being justified without showing that I am in this state.

Secondly, it might be argued that all our beliefs are interdependent. No belief stands alone in our noetic structure. Hence, to be justified in believing that there is a brown table in front of me, I need to be justified in other beliefs, e.g., about my physical environment, about how tables look, about colors. It is plausible I think to suppose that a precondition for our holding any belief is that we hold other beliefs. Beliefs are certainly not isolated. We need at least to hold the requisite concepts involved in a belief. The mistake, though, is to suppose that the negation of [SE] would require maintaining that the justification of any belief is in no way dependent on other beliefs. But this is not correct. Rejecting [SE] only requires rejecting the more narrow idea that the justification of some beliefs is justificationally dependent on other beliefs. That is, some beliefs receive their status as justified in a way that does not require their being based on other beliefs. This is quite consistent with maintaining that the existence of any belief depends on the existence of other beliefs in the person’s noetic structure. [7]

Third, it might be argued that if any belief is justified in the absence of evidence or reasons, such a belief would be beyond critical analysis and we would thereby open the door to dogmatism and groundless assertions. Of course, conceding that some beliefs are justified by virtue of something other than their evidential grounding in other beliefs has no such implication. One can always raise the question as to whether such beliefs are true, and one can in principle find reasons for supposing that a belief is true even if its justification does not depend on such reasons. Moreover, for any belief that is immediately justified, one can inquire as to what grounds there are for supposing that it is immediately justified. The criticism here overlaps with the first one above. I can always engage in the activity of justifying a belief that is immediately justified.

So these three arguments for [SE] are really inadequate to support [SE]. Moreover, the Reformed epistemologist will also argue that [SE] is obviously false.

First, [SE] would be logically inconsistent with many paradigm cases of justified belief where a person’s belief that p is not based on propositional evidence. As the result of when being appeared to redly, I believe that there is something red in front of me. Another person takes it as quite obvious that he exists. A young child accepts his teacher's testimony that Sacramento is the capitol of California. Upon feeling tired, I believe that I am tired. Upon having a memorial experience of last summer's hike through the Pinnacles, I find myself believing that I hiked the High Peaks Trail. As I reflect on the argument form modus ponens, the corresponding conditional strikes me as obviously true. In these cases, the target beliefs are not held on the evidential basis of other beliefs. The target beliefs are formed directly, either by way of extrospective experience, testimony, introspection, memory, or grasped intuitively as an a priori truth. Given [SE], belief in such propositions in the described circumstances is not justified. But we are strongly inclined to accept that we are justified in believing these propositions in the circumstances described. To accept [SE] we would have to revise a powerful and widely held intuition about conditions in which certain kinds of beliefs are justified.

A second and more fatal problem, though, is that [SE] threatens to generate an infinite regress that will entail that no human person is justified in believing any proposition.[8] Suppose that some belief B1 is justified only if it is based on "adequate evidence." But justification cannot be transferred from beliefs that lack justification. So if evidence is construed propositionally, the "adequate evidence" that forms the basis of my belief must itself consist of further justified beliefs. But then B1 will be justified only if a person has at least one other justified belief B2 that provides adequate evidence for B1. But then given [SE], there must be another belief B3 that supports B2. But since B3 must also be justified, there must be another belief B4 that provides adequate evidential support for B3. Now there are only three possible outcomes of this chain of justification transfer. It can continue ad infinitum, return back to some belief that has appeared previously in the series, or it can terminate in a belief that is justified but not by virtue of being based on some other (justified) belief(s).

Given the first scenario, no belief can be justified for any person unless the person possessed an infinite number of justified beliefs. But it is not psychologically possible for human persons to hold an infinite number of beliefs, much less an infinite number of justified beliefs. Not only is there not enough time for a human person to acquire an infinite set of beliefs, but any infinite set of beliefs would eventually involve formulations too lengthy for the human mind to understand. And if a propositional formulation can't be understood, it can't be believed. True, we may understand that a particular formulation expresses some truth, but we could believe the truth it actually expresses.[9] The conjunction of these psychological facts and [E*] entails that no human person has any warranted beliefs, and thus knowledge is not possible. Thus [SE] generates an infinite regress that in fact undermines knowledge.

On the second scenario, the evidentialist can avoid the infinite regress problem, but he ends up with the problem of circular justification. Suppose the series of beliefs, each of which is held on the evidential basis of the other, loops back or returns to a belief that has occurred previously in the series. Here we have a circular chain of beliefs. S's belief B1 is based on B2, B2 is based on B3, and eventually we arrive at some belief B_n that is based on B1. There are two considerations here, one causal and the other evidential. Both are involved in the idea of holding some belief on the evidential basis of some other belief. The relation of evidential dependence seems to be transitive. So if B1 is evidentially dependent on B2, B2 evidentially dependent on B3, and B3 eventually dependent on B1, then B1 will be evidentially dependent on itself. Now there is nothing intrinsically implausible about a belief’s providing evidence for itself. However, this really breaks away from the spirit, if not letter of [SE], where justification requires that one belief be evidentially supported by some other, different belief. A more serious problem emerges, though, with reference to the idea of causal dependence that is involved in the "based-upon" relation. If transitivity holds between causal relations, as seems to be the case, then circular justification transfer will entail circular doxastic causality. There will be some belief that is, not just evidence for itself, but its own causal ground. It will be causally dependent on itself. This consequence is not highly attractive.

The conclusion, then, is that an epistemological principle like [SE] can’t be the basis of [E]. More precisely, it appears that mediate of inferential justification implies that some beliefs must be justified by virtue of something other than their being based on other justified beliefs. There must be something like immediately justified beliefs. But in that case, the natural question to ask is: why can’t theistic belief be an immediately justified belief?

B. Plantinga, Evidentialism, and Classical Foundationalism

Alvin Plantinga has devoted considerable attention to this question by examining both the concept of rationality/justification and the epistemological assumptions that underlie the evidentialist position.

The evidentialist takes “evidence” to be necessary for rational theistic belief, but how shall we understand the nature of rationality (or justification) itself? According to Alvin Plantinga, there are two primary ways of understanding the concept of rationality envisioned by the evidentialist.[10] There is first a kind of deontological view of rationality according to which a rational belief is one that we are permitted or entitled to hold. The close analogy here is with action. A person S is permitted or entitled to perform some action A just if S’s doing A does not contravene any of the relevant (moral, legal, etc) rules or obligations S. Similarly, the evidentialist view of justification may be spelled out in terms of a person’s being within his or her epistemic rights in believing as she does. To be rational in this sense is just not to flout one’s intellectual or epistemic duties. Since we commonly speak of a person being justified in her actions, the term justification is an apt term here. Alternatively, one can think of rationality in the sense of non-defectiveness or epistemic health, what Plantinga calls axiological rationality. A person who lacks rationality in this sense is a lot like a person who suffers from a spastic colon or heart murmur; there is some flaw or defect in the person biological system. Thus clarified, the evidentialist requirement may be construed as a requirement for being within one’s intellectual rights or for cognitive non-defectiveness. In either case, Plantinga finds the evidentialist position implausible.

Plantinga’s case against evidentialism rests heavily on its association with classical foundationalism (hereafter CF).[11] According to foundationalism simpliciter our rational beliefs form a structure in which some beliefs are accepted on the basis of other beliefs (nonbasic beliefs) and other beliefs are not accepted on the basis of other beliefs (basic). If a belief is basic and rational, it is properly basic. Properly basic beliefs serve as the ultimate foundations for all inferential reasoning or properly non-basic beliefs, thereby avoiding the infinite regress problem noted above. CF is a particular version of foundationalism that restricts properly basic beliefs to self-evident beliefs (e.g., 2+2 = 4), beliefs that are evident to the senses (e.g., there is a tree outside), and beliefs about one’s current states of consciousness (e.g., I am tired). Plantinga sees the evidentialist requirement as emerging from the framework of CF. If one sets the criteria for properly basic beliefs too narrowly, theistic belief will be excluded from the class of properly basic beliefs. Take the modern version of CF beginning with Descartes. It restricts properly basic beliefs to beliefs that are either self-evident or about one’s immediate states of consciousness. Belief in God, so the argument goes, is not plausibly construed as basic in either of these senses. Hence, it is rational only if it accepted on the basis of propositions that are properly basic in this narrow sense or are ultimately based on beliefs that are properly basic in this narrow sense. On the other hand, the other version of CF, ancient foundationalism (represented, for instance, by Aristotle), maintains that properly basic beliefs are either self-evident or evident to the senses. But the truth of theistic belief is not something evident to the senses since God is not an object of sensory perception. CF, then, seems to have the consequence that if theistic belief is rational, it must be based on evidence, where this evidence consists of truths that are either self-evident, evident to the senses, or about one’s current states of consciousness, or based on beliefs that are ultimately based on such narrowly circumscribed properly basic beliefs.

Plantinga’s case against the evidentialist requirement for rational belief is essentially to argue that evidentialism is rooted in classical foundationalism and classical foundationalism is epistemologically bankrupt. The latter argument comes down to two theses. If we accept classical foundationalism, then (i) we are not rational to accept most of our ordinary everyday beliefs (though we consider them paradigm cases of rational belief) and (ii) accepting classical foundationalism itself cannot be rational.

First, Plantinga argues that classical foundationalism (CF) falls afoul upon reflection on most of our everyday, ordinary beliefs. If we suppose that CF is true, then very few of our commonsense, every day beliefs are rational. Most people believe in the existence of other minds, the reality of an external world, and the occurrence of the past. But these are neither self-evident nor about one's own immediate introspective experience. According to modern classical foundationalism, they are rational only if they are evidentially supported by other rational beliefs. But are they? What beliefs entail the existence of other minds? Do any of our self-evident beliefs make probable the existence of the external world? It seems not, Plantinga thinks. But we are nonetheless rational in holding these beliefs. And here we have a double claim. A person who believed in the existence of the external world or other minds, though he doesn't have evidence for them, would not be violating any intellectual duties, nor would such a person be exhibiting some kind of cognitive malfunction or defectiveness. So something must be wrong with CF, if it logically implies that many of out beliefs are not rational given that it seems very clear that they are rational. We are forced to choose between out intuitions of rational beliefs and CF. Plantinga thinks the correct choice is obvious. It is not CF. We may put Plantinga’s objection here in a more general form. The problem with CF is that it is implausible to suppose that from the rather narrow set of properly basic beliefs postulated by CF we can generate a sufficiently broad structure of rational or justified beliefs that would cover the sort of beliefs we think we are rational or justified in holding.

Secondly, a potentially more devastating objection is Plantinga's claim that CF is self-referentially incoherent. It does not satisfy the conditions of proper basicality that it lays down, and it is difficult to see how it could be adequately supported by beliefs that do meet those narrow criteria. According to modern classical foundationalism, a person would be rational in accepting CF only if CF is properly basic or properly nonbasic. But CF is neither self-evident nor about one's immediate introspective experience, so it is not properly basic. So CF is rational only if it is entailed or made probable by some other rational beliefs. Although there might be such an argument lurking out there somewhere, no classical foundationalist has yet provided such an inference. Therefore, belief in the proposition that asserts the modern classical foundationalist criteria for proper basicality is not a rational belief. Anyone who accepted this principle, in the absence of an argument for it, would be flouting his intellectual duties or exhibiting some kind of cognitive malfunction or defect. As Plantinga writes: "It is evident that classical foundationalism is bankrupt, and in so far as the evidential objection is rooted in classical foundationalism, it is poorly rooted indeed."[12]

If the basis for evidentialism is CF, and CF is implausible, in essence we have lost the reason to endorse evidentialism. Hence, Plantinga’s critique aims to undercut an objection to the proper basicality of theistic belief, and ultimately the claim that theistic belief is not rationally justified. There is no good reason to deny that theistic belief is properly basic given CF.

C. Evidentialism and Coherentism

Evidentialism, of course, need not be based on classical foundationalism.[13] It can also be based on epistemological coherentism.[14] Coherence can be spelled out in different ways. There is first the idea of logical (and probabilistic) consistency between beliefs. There is also the idea that coherence is a function of the number of inferential connections between beliefs in a system. Many non-inferentially connected beliefs can be logically consistent. Perhaps most importantly, though, coherence involves important explanatory relations between beliefs. Hence, coherence is reduced by the presence of unexplained anomalies in a person’s noetic structure.[15]

Plantinga proposes two ways of developing an evidentialist case based on coherentism. Both are developed in terms of what Plantinga considers the best way to take the evidentialist view of rationality, as axiological rationality or warrant non-defectiveness, the absence of noetic defects or cognitive disorder.[16] First, the coherence evidentialist may think of coherence as necessary and sufficient for any belief to be rational, what we might call pure coherence theory. Hence, theistic belief will be rational in this case just if it coheres with the rest of a person’s noetic structure. As Plantinga notes, on this sort of coherentism, theistic arguments become less important since the rationality of theistic belief is determined by coherence relations between theistic belief and the rest of a person’s noetic structure, as opposed to narrowly defined deductive or inductive inferential relations between theistic belief and some limited set of other beliefs. Alternatively, and perhaps more plausibly, the coherence evidentialist may be construed as simply using coherence as a criterion for what counts as a properly basic belief. While some beliefs have their warrant transferred to them by way of other beliefs (through various deductive or inductive inferential relations), some beliefs are warranted just by virtue of their coherence with the rest of one's beliefs. Common sense beliefs, such as the one's introduced above, may be warranted because they do not conflict with the (majority of) rest of one's beliefs or because they are supported in some way by the totality of the rest of one's beliefs. The first represents a kind of negative coherence, whereas the latter is positive coherence. These coherence relations involve a kind of evidential support relation between the target belief and the rest of the person’s noetic structure. The coherence relation here is holistic. While allowing some beliefs the privileged status of being psychologically basic, warrant is nonetheless conferred by evidence in the form of global noetic coherence.[17] On this way of thinking about coherence, coherence evidentialism is in fact a species of foundationalism – a blend of local foundationalism (due to warrant transfer terminating in basic beliefs) and global coherence (warrant generated by coherence vis-à-vis basic beliefs). Hence, this version represents a kind of impure coherentism.

What are we to make of this evidentialist position? Plantinga’s basic position is to argue that coherence is neither necessary nor sufficient for rationality construed as warrant non-defectiveness, and thus both forms of coherentism are inadequate as a basis for evidentialism. Standing outside Balliol College in Oxford, I form a variety of beliefs about my surroundings, e.g., gowned undergraduates on their way to the Examination schools, a red double-decker bus is to my right, Blackwells bookstore is just around the corner. These beliefs all cohere with each other and the rest of my beliefs. Later that day my experience changes while riding the train to London, but I continue to hold the same beliefs as I did while I stood outside of Balliol, e.g., a double-decker red bus is to my right, Blackwells is just around the corner. My beliefs are no less coherent at the later time than there were earlier in the day. If they were rational earlier, they are rational at the later time – if coherence were sufficient for rationality. But clearly my beliefs are not rational at the later time, and thus coherence is not sufficient for rationality in the sense of warrant non-defectiveness. The failure of my cognitive system to respond to changes that occur in my experience is, of course, the source of the warrant defect, but it is logically consistent with coherence. If coherence is not sufficient for the rationality of beliefs in general, it is hard to see how it could be sufficient for the rationality of basic beliefs in particular. After all, most of the beliefs in the example above are basic beliefs. Despite their coherence, they are not rational.

Similarly, Plantinga argues that coherence is not necessary for warrant non-defectiveness. Consider a modification of the above example. At one point along the train ride, suppose that my visual experience registers momentarily and I see that I am on a train. I automatically form the belief that I am on a train. The belief will lack coherence with the rest of my beliefs, since the experiences of being on a train and standing outside Balliol College are quite incompatible. But does my belief that I am on a train at that point fail to rational because it fails to cohere with the rest of my beliefs? There is nothing defective at all about the formation of the belief that I am on the train; defects lie elsewhere. But I am rational in the sense of warrant non-defectiveness to believe that I am on the train.

Plantinga does not consider deontological rationality in his discussion of coherentism, but the implications of the account are easily drawn out. First, it is certainly possible for a person’s beliefs to be coherent and the person free from epistemic blame. But it hard to see that freedom from epistemic blame or being within one’s intellectual rights follows as a consequence of having coherent beliefs. After all, it surely seems plausible to suppose that one could acquire a belief that coheres with the rest of one’s beliefs as the result of being epistemically irresponsible. Suppose that while the prosecution is presenting valuable bits of evidence, Mack is deep in thought reminiscing over his last visit with his mistress and plotting the next. As a result he does not hear the evidence presented. Moreover, while the jury is deliberating, he is staring off into space. Missing the relevant evidence, he believes that the defendant is not guilty. This belief coheres with the rest of his beliefs, perhaps quite well. The defendant is a white-male businessman of age 45 with a family. Mack believes that only blacks, young people, and single people have a propensity toward crime. We might reasonably suppose that a juror has an obligation to consider the evidence. Mack has not done so. Despite the fact that his belief coheres with the rest of his beliefs, surely his belief was not formed in an epistemically responsible manner. He is thus not deontological rational to believe that the defendant is not guilty.[18] So coherence is not sufficient for epistemic dutifulness. But it is not necessary either. A person could very well be deontologically rational in coming to hold a belief that fails to cohere with the rest of the person’s noetic structure. Take the case of Mack. Suppose that, quite opposite to the description give, Mack hears the evidence and is powerfully persuaded by a consideration of it. This belief will not cohere with his beliefs about the propensity toward crime of the defendant. He has been epistemically responsible and thus a belief has been formed that does not cohere with significant sections of his noetic structure. Or suppose take the second scenario on the train ride to London. When I spontaneously form the belief that I am on the train, how I am epistemically to blame for holding this belief, though it fails to cohere with the rest of my beliefs? It is hard to see how. Suppose that I am struck by some bizarre form of cosmic radiation that causes me to believe, several times in the course of the day, that a church bell is ringing, though this fails to cohere with the rest of my beliefs. What epistemic duty have I violated in forming this belief. I haven’t as much as a clue that anything is wrong. I am not epistemically blameworthy.

Since coherentism is an implausible epistemological theory, it is a poor basis on which to resurrect an evidentialist requirement for theistic belief. More precisely, since there are many cases of properly basic beliefs where these beliefs are rational but fail to cohere with the rest of a person’s noetic structure, it cannot be argued that coherence is a necessary condition for proper basicality. Hence, theistic belief can’t be excluded from being properly basic on the grounds of either classical foundationalism or coherentism.[19]

II. The Positive Case for Proper Basicality

But does any of this show that theistic belief is properly basic? A common complaint by philosophers in response to Plantingia’s critique of evidentialism is that he does not show that theistic belief is rational or justified for anyone, only that the examined objections to it are implausible.[20] While Plantinga’s earlier papers were attempts at defending the proper basicality thesis, in subsequent works the positive case for properly basic theistic belief has come through more explicitly.

In Warranted Christian Belief, Plantinga makes it clear that the both the theist and the Christian theist in particular theist are typically justified in their theistic and Christian beliefs, even when these are held in a psychologically basic way.[21] More precisely, it is pretty “obvious” that they are typically deontologically rational in holding their theistic beliefs. Since deontological rationality is simply a matter of being within one’s intellectual rights, it need only be shown that in many circumstances in which people hold theistic belief, they do so without violating any epistemic duties or being epistemically blameworthy.

It is a matter of controversy as how exactly we should understand epistemic duties.[22] One plausible way of understanding the deontological requirement, though, is as a requirement to do all that can be reasonably expected of us to see to it that our beliefs are true. The question is whether the person who holds a basic theistic belief fails to satisfy this sort of requirement, or any more specific duties that follow from it. The answer seems plainly not, not usually. Many believers find themselves with a very strong sense of God’s presence in their lives. Some of these have been raised in religious families and others not. Most of them are also aware that not everyone shares their belief in God, and that there are also various objections to belief in God. Some of these believers are even acquainted with the objections to theistic belief by thinkers like Nietzsche, Freud, Russell, and Flew. Some believers are persuaded by such arguments and find themselves doubting their belief in God. Hence, while they believe in God in a basic way, this belief is subject to revision and doubt in the fact of sufficient counter evidence, or at least what they take to be sufficient counter evidence. Surely this sort of thing is epistemically responsible. But other believers read the arguments, understand them, but ultimately are not persuaded by them. Perhaps they find various flaws in the arguments. Maybe they are aware that, like most philosophical arguments, such arguments are far from compelling. Moreover, these believers are still struck with a sense of God’s presence in their lives. This belief is confirmed in a variety ways, such as their life taking a course that predicted by their religious beliefs (e.g., growing in certain virtues). Should we regard these individuals as epistemically irresponsible? Hasn’t this sort of believer done all that can be reasonably expected of her? She’s considered possible counter evidence, examined her life and grounds for belief, and yet she still believes that God exists. Now perhaps there is some sort of cognitive defect in such a person. Maybe this is a case of Freud’s wish fulfillment hypothesis, or to use the more culturally popular description, a kind of “psychological crutch.” Perhaps; but such a person would not on that account be epistemically blameworthy. One might suppose rather that theist should be pitied rather than blamed for her beliefs. So, given that we have reason to suppose that many believers conduct their intellectual lives in a way that does not involve a violation of any epistemic duties, according to Plantinga, there is a positive case for supposing that these sorts of believers are deontological rational or justified in their basic belief in God.[23]

Nicholas Wolterstorff has also developed a similar case for the rationality of basic belief in God on the basis of the principle of credulity.[24] Like Plantinga, Wolterstorff thinks of the evidentiaist challenge in terms of epistemic duties, obligations, and responsibilities, all rooted in the Lockean conception of rationality.[25] Wolterstorff is critical of the imperialism of "reason as inference or argument" that is suggested by the evidentialist demand for evidence. He sees a more reasonable alternative in the work of the 18^th century philosopher Thomas Reid, specifically Reid's idea that human persons have, among their belief forming and sustaining dispositions, a native disposition of credulity or believability. According to Reid, this disposition to accept what we receive by way of testimony is unlimited and without restriction in childhood, but it is refined and restrained with the experiences of falsehood and deceit as a child matures. Wolterstorff derives from Reid’s discussion a more general principle of rationality: we ought to consider beliefs innocent until proven guilty, not guilty until proven innocent.[26] He formulates a criterion of rationally justified belief on that basis. The criterion, roughly stated, permits us to hold a belief unless we have adequate reasons to cease holding the belief. More specifically, with reference to beliefs that we could refrain from holding ("eluctable" beliefs as he calls them), we are rationally justified in holding them unless we either have or ought to have adequate reasons for supposing that they are false or formed in a way that is not reliable.[27] Reasons play a rationality-removing role. Unlike Plantinga's account, Wolterstorff includes reasons a person doesn't have but ought to in the category of defeaters to rational belief. Wolterstorff sees this criterion as appropriately capturing the normative requirements for deontologically rational belief. Since it is reasonable to suppose that there are cases where people hold theistic belief in a basic way but do not have, nor ought to have, adequate reasons for surrendering that belief, theistic belief can be rational without evidence.

The rather common criticism against Reformed epistemology, that it has done little or nothing to show that theistic belief can be properly basic, seems plainly mistaken. Reformed epistemologists such as Plantinga and Wolterstorff have actually done what evidentialists themselves have not done. They have engaged the conceptual foundations of the evidentialist position, specifically with respect to the concept of rationality itself. They have presented a prima facie compelling case that the dominant, though frequently implicit, deontological concept of rationality in the tradition is unable to sustain an adequate objection to the rationality of basic belief in God. Moreover, assuming that deontological concept of rationality, they have argued that many theists are deontologically rational in holding their theistic beliefs in a basic way.

III. Epistemic Evidentialism Examined

Up to this point we have examined the rationality or justification of basic belief in God. Knowledge, of course, is another matter. Can theistic belief constitute immediate knowledge? While evidentialism is historically developed in terms of rationality and justification (often construed deontologically), it can be developed with respect to knowledge. While the historical preoccupation with rationality and justification largely explains the past emphasis of Reformed epistemologists, the knowledge of God has emerged as point of increasingly emphasis since the early 1990s.[28]

We should accordingly revise the form of evidentialism under discussion. Knowledge (or epistemic) evidentialism maintains:

[KE] A person S’s theistic belief T constitutes knowledge only if it is based on propositional evidence that provides adequate support for T.

As it turns out, many of the observations already made have direct application to evidentialism with respect to knowledge. To see this, we should note a few things about the nature of knowledge, and its relation to justification.

The ascription of “knowledge” to our beliefs belongs to a particular mode of evaluation or appraisal of beliefs. We may, in fact typical do, evaluate beliefs as true or false. But we may also evaluate beliefs in relation to truth as a goal of our cognitive life. For instance, we often speak – as we did above - of beliefs being rational or justified. In doing so, we are appraising or evaluating beliefs, not necessarily as true, but as permissible or commendable in some sense toward the goal of believing what is true. A rational or justified belief, for instance, typically indicates that someone’s belief is positively related to truth as a cognitive desideratum, e.g., the person has good evidence, grounds, indications, or reasons for supposing that his belief is true.[29] This category of belief evaluation is what we can refer to as positive epistemic status. Like rationality and justification, knowledge belongs to the category of positive epistemic status.

Unlike the former terms, though, knowledge involves the actual possession of truth, whereas rational or justified beliefs may be false. So knowledge can’t be the same as a justified belief. Nor is knowledge merely the possession of a true belief. One may possess a true belief quite by accident or believe what is true without having anything that is indicative of the truth of the belief. Knowledge cannot be accidentally true belief. Put otherwise, there is a kind of surplus value in knowledge over true belief, a value found in the possession of a particular excellence of how the believer is related to the truth-goal of believing. Since this excellence is precisely what is traditionally indicated by terms like justification, the traditional view of knowledge is, roughly at any rate, justified true belief. (I will henceforth refer to this as the JTB theory).

A. Evidentialist Justification and Knowledge

We can now see a connection to the previous sections. If knowledge is justified true belief, there is only one plausible way to argue against immediate theistic knowledge. One may, of course, argue that theism is false. If so, God cannot be known, immediately or otherwise. But this won’t be a happy strategy since many evidentialists are themselves theists. So the plausible approach must be a case against immediate knowledge that is logically consistent with the truth of theism. In that case, the only plausible approach is to argue that justification requires evidence. More precisely, given the JTB theory, the truth of [KE] entails the truth of [JE]. Hence, the plausibility of evidentialist views of justification considered in the previous sections is directly relevant to a case against immediate knowledge of God.[30] But the implausibility of [JE] clearly undermines [KE]. And the implausibility of [JE] does seem formidable.

To briefly summarize the results:

(1) [JE] cannot be asserted on the basis of strong evidentialism, the view that all beliefs must be based on other beliefs to be justified. Such a view is logically inconsistent with many paradigm cases of justified belief, and it also faces the dilemma of infinite regress or circularity.

(2) [JE] cannot be asserted on the basis of classical foundationalism, since (a) there are many paradigm cases of immediate justified beliefs that fail to satisfy classical foundationalist criteria for proper basicality and (b) classical foundationalism itself does not satisfy its own criteria for justified belief.

(3) [JE] cannot be asserted on the basis of pure or mixed coherence theories since coherence is neither necessary nor sufficient for justification, understood deontologically or as freedom from cognitive defect.

(4) [JE] cannot be asserted on the basis of various modifications of evidentialism as long as justification is construed deontologically since there are various grounds on which to affirm that many instances of basic theistic belief satisfy the conditions of deontological justification.

Hence, I think we must conclude that the prospects for [KE] do not look very good from the vantage point of [JE] and the JTB theory.

But perhaps this is no significant loss. After all, the above argument is based on the truth of JTB account of knowledge, but most contemporary theories of knowledge recognize a significant deficiency in the JTB theory of knowledge, a philosophical disquietude that descended on epistemology by Edmund Gettier’s 1963 paper, “Is Justified True Belief Knowledge?” [31] Gettier argued that a true belief can be justified, evidentially and deontologically, but fail to constitute knowledge. The post-Gettier settlement has been either to modify the traditional understanding of justification or maintain that justified true belief is not sufficient for knowledge and some fourth condition must be added. In the light of the post-Gettier approach to epistemology, it will be helpful I think to simply use the word “warrant” to indicate whatever conditions are needed to transform true belief into knowledge. Knowledge, then, will be sufficiently warranted true belief. The evidentialist case for [KE] does not depend on [JE], but rather on the following:

[WE] A person S’s theistic belief T is sufficiently warranted only if it is based on propositional evidence that provides adequate support for T.

The evidentialist can then argue as follows. He can concede that theistic beliefs can be immediately justified. That is to say, the theistic can possess deontological rationality, or any sort of subjective rationality associated with how things seem to him and the absence of overriding reasons to the contrary. But he can argue that since justification is at best only part of warrant, the immediate justification of theistic belief would not (in conjunction with truth) entail immediate knowledge of God. And so no defense of the immediate justification of theistic belief amounts to a defense of the immediate knowledge of God. [KE] actually entails [WE], and this is entirely consistent with the falsity of [JE]. So the evidentialist need only show that warrant requires evidence to support [KE]. But does warrant require evidence?

Clearly [WE] cannot be asserted on the basis of any general principle according to which all beliefs must be based on adequate propositional evidence (i.e., other warranted beliefs). Not only are there many paradigm cases of immediate warrant, but such a view would be infected with the same dilemma of infinite regress and circularity encountered earlier with strong justification evidentialism. Nor can [WE] plausibly be asserted on the basis of classical foundationalism, for many paradigm cases of immediate warrant fail to satisfy the narrow classical foundationalist criteria of proper basicality, nor could much of what we take ourselves to believe with warrant be generated from such narrow foundations.[32] Nor can [WE] be asserted on the basis of a coherence criterion of proper basicality. Minor adjustments to the examples looked at earlier, so that they are framed in terms of warrant, would yield similar results. While on the train to London, would I not be warranted in believing that I am seated on a train when my beliefs become responsive to my experience? It seems so. After all, couldn’t I not know that I am seated on the train, even if fails to cohere with much of the rest of my beliefs. Moreover, it is also worth noting here that even if the coherentist criterion of properly basic beliefs is correct, it would not exclude theistic belief from the status of proper basicality on that account. There is little promise in arguing that no person’s theistic belief coheres with the rest of her noetic structure. So theistic belief could rightly be regarded as a properly basic belief. Of course, it does receive its warrant by way of evidence, but only in the sense that it coheres with the rest of what the theist believes. This is not necessarily a matter of actually inferring the belief from other beliefs, or citing evidence in support one’s belief. And it isn’t psychologically based on evidence at all.

The basic problem for [WE], though, is that when we examine the sort of conditions that are typically added to justification to account for Gettier type cases they are typically not matters of the person having evidence for his belief. In fact, this is typically the case just because Gettier cases are a response to a certain inadequacy in evidential conceptions of justification. Drawing evidence out of warrant is simply to acquire a cognitive stone instead of epistemic bread. What sorts of conditions are usually added to justified true belief to yield warrant? One view maintains that a belief will be warranted only if it is not explicitly or implicitly inferred from a false (justified) belief. But this condition only specifies something that is necessary in cases of inferential knowledge. It doesn’t provide any basis for placing an evidential requirement on theistic belief. Another view is simply to maintain that warrant requires some causal relation between the cognizer’s belief state and some feature of the world. While this is applicable to cases of immediate knowledge, [WE] is hardly deducible from it. All that would follow is that one’s belief theistic belief cannot constitute knowledge unless God is the cause of this belief. Surely this condition could be satisfied without [WE] being true. Lastly, and most commonly, there is a broad range of adjustments to warrant that fall under the category of defeasibility theories. Roughly, these theories maintain that S’s belief B is warranted only if there is no relevant unpossessed evidence, e, such that if the cognizer were to believe e, she would no longer be (evidentially or deontologically) justified in holding B. Of course, defeasibility theories differ as to what counts as relevant unpossessed evidence.[33] But in any event, such theories provide no support for [WE] since they are concerned with evidence not possessed by the cognizer, much less evidence that forms the basis of one’s belief.

So if evidence isn’t introduced in the justification component, it is hard to how it can get smuggled in under the rubric of warrant, for what it takes to address Gettier cases is not more evidence.

B. William Alston’s Internalist Directed Critique of Epistemic Evidentialism

Another way of critically examining evidentialism has been proposed by William Alston. The evidentialist positions articulated above all share a fundamental feature in common. They are all internalist epistemologies. Internalist theories of knowledge represent one general approach to warrant, according to which what confers warrant on a belief is internal to the cognizer. Of course, the internalist is thinking of "internal" in a special sense. What is relevantly "internal" in this context is something that is necessarily mentally or introspectively accessible to the cognizer. These include a priori truths (e.g., logical truths), as well as the entire range of the subject's own beliefs and experiences. What is internal to the cognizer thus includes, though is not limited to, matters of privileged epistemic access, truths that a particular cognizer has a way of knowing that no one else does, by introspection (e.g., I feel tired). According to the core idea of epistemic internalism, a warranted belief is one that is based on an internally accessible, adequate ground, and – on some accounts - where the adequacy of the ground is itself something to which the cognizer has access upon reflection.[34] Although some internalists confine these grounds to other beliefs, other internalists include experiential grounds (e.g., being appeared to a certain way). Both sorts of grounds are, of course, internally accessible. Internalism, though, naturally leads to evidentialism since "evidence" is construed as something that the believer recognizes as support for his belief. It is thus not surprisingly that evidence is a basic motif of the internalist tradition in Western philosophy.

William Alston has proposed that evidentialism is best critically approached in the context of its internalist assumptions.[35] Given internalism, the prospects of immediate knowledge of God are quite bleak, and – correspondingly – the evidentialist case very plausible. The crucial issue here is how internalist requirements influence how we think about immediate knowledge, its possibility and range.

In its most radical forms internalism would rule out immediate knowledge altogether. For instance, there are forms of internalism that restrict the range of justifiers to other justified beliefs. This is naturally developed in such a way that it yields the [SE] position critically examined earlier. Of course, if grounds are cognitively accessible items, then we should be able to cite them when our beliefs are challenged. But what we cite are other beliefs. Hence, the internalist framework lends itself rather easily to conflating being justified with being able to justify one’s belief. But since the latter requires reasons, reasons are then viewed as requirements of the former. From the same root grows the pervasive internalist requirement that the cognizer justifiably believe that the grounds or sources of his belief are adequate, or at least that he is able to see this fairly readily upon reflection. On such a view, I am justified in believing that I see a tree only if I am, or can be, justified in believing that my ground for this belief, perhaps in conjunction with all my relevant beliefs, provides adequate support it. This springs naturally from the idea that, in addition to the grounds, the adequacy of grounds is something that is cognitive accessible. Hence access to the justifier is often times also a matter of access to the properties whereby it can justify. We can see this idea as implicit in the concept of an evidentialist requirement for belief. If my beliefs are based on evidence, then it seems reasonable to suppose that I must take my grounds to be evidence for the belief. Why else hold my belief on such grounds? But it would appear that I can only take grounds to be evidence if I take it that my grounds are epistemically adequate or efficacious in their support of my belief (to whatever degrees). [36]

The basic idea of internalism, i.e., justified beliefs are based on internally accessible grounds, is by no means decisive against the prospects for knowledge of God, either immediate or mediate. Some internalists do not restrict grounds to other warranted beliefs of the cognizer, but include a broad range of experiential factors as grounds (e.g., sensory and memorial experience). In that case, the demand for grounds is not logically inconsistent with immediate knowledge.[37] Secondly, most religious people have grounds for their religious beliefs and most can fairly readily identify a good number of them for their beliefs. They would appeal to Scripture, the creeds, religious experience, the views of other religious people, and philosophical considerations. And as we noted in the previous chapter, several prominent Reformed theologians have maintained that God’s existence is in some way a self-evident or intuitive truth grasped by our rational cognitive faculties. Some of these sources of belief would appear to be relatively clear sources of immediate religious beliefs, e.g., religious experience, acceptance of the explicit statements of Scripture, and rational intuition. The problem is not so much the internalist demand for grounds, though we often forget why we believe what we believe. Rather, the problem is located in the demand that the subject possess justified beliefs about the adequacy of his grounds of belief, or even the weaker demand that the adequacy of the grounds be the sort of thing the cognizer could come to determine just upon reflection. This saddles justification, and hence knowledge, with a rather stringent requirement that most religious people are not able to meet.

Alston locates several crucial problems with internalist accounts of justification, as a necessary condition for knowledge. Two are worth noting here.

First, most forms of internalism necessarily restrict the range of knowledge, including immediate knowledge by the sorts of constraints that are placed on justification.[38] In its most radical form, justifiers are restricted to other justified beliefs of the subject, in which case there are no immediately justified beliefs at all. Of course, the more plausible forms of internalism make no such restriction. But they nonetheless impose others that are equally implausible. If, for instance, justification depends in part on beliefs about the adequacy of grounds, or even the mere accessibility of the adequacy of the grounds, many paradigmatic cases of belief will fail to be justified. And since internalist justification is an element of knowledge, the range of knowledge will be more restricted than is plausible. Small children do not have the ability to make determinations about the adequacy of the grounds, and yet we attribute knowledge to them. Most people do not have adequacy beliefs, and they would be hard pressed to acquire them upon reflection. But it certainly looks like such people have knowledge nonetheless. Our perceptual beliefs in particular are formed too quickly and spontaneously for there to be any adequacy beliefs worked into such knowledge. So internalism seems to run contrary to paradigm cases of knowledge.

Second, according to Alston, an equally pervasive general problem facing nearly all internalist theories is their lack of truth-conductivity.[39] A basic question for any concept of epistemic justification is how to think of the positive relationship between one’s belief and the truth goal of believing. While internalist theories employ the notion of likelihood of truth to forge the positive connection between belief and the truth goal of believing, the notion of probability is internalized. If the adequacy of a ground is understood in terms of a high probability of being true and yet adequacy is something cognitively accessible, then justification requires that the likelihood of truth be cognitively accessible. Hence it must be made relative to the believer’s perspective. So for the internalist the likelihood of some belief being true is a matter of epistemic probability. It is a probability relative to a person’s evidence and deductive/inductive standards, ideally correct evidence and standards.[40] But the epistemic probability of the belief that p on grounds G (even if high) is quite distinct from what we might call the factual probability of the belief that p on G. The latter refers to the propensity or tendency of beliefs like p in fact being true when held on grounds like G. Certain law-like features of the world dictate what is probable here. Probability is, we might say, externalized. It is a function of grounds or processes being such that they would normally yield a high proportion of true beliefs, an empirical matter not introspectively accessible. Since beliefs can be epistemically probable and fail to be factual probable, Alston maintains that a justified belief is one that is rendered factually probable by its grounds. [41]

So for these reasons and others, Alston concludes that internalism is not a plausible account of justification or knowledge. But if evidentialism is based on internalism, it becomes just as implausible.

IV. William Alston’s Model of Immediate Knowledge of God

While the implausibility of internalist assumptions tends to undermine the evidentialist position and hence objections to the immediate knowledge of God, the prospects for the latter may be seen in a more favorable light once we take on a very different theory of knowledge. In contrast to internalism there are a variety of externalist theories of knowledge. Given that externalism stands in contrast to internalism, externalist theories – stated in the negative - involve a rejection of various internalist requirements for warrant. Put otherwise, externalism locates warrant in epistemically relevant conditions that are not cognitively accessible to the subject. According to reliabilism, one of the main forms of externalism, a belief is warranted just if it is the product of a reliable belief forming process. On another version of externalism, a belief is warranted only if it produced by properly functioning cognitive faculties. “Reliability” and “proper function” both concern empirical aspects of our psychology that are not introspectively accessible. We can’t determine whether these conditions obtain or not from the inside. Nor do externalist theories require that the subject form any beliefs about reliability or proper function.[42] Hence, externalism represents a significant departure from the internalist understanding of knowledge. But with this departure come interesting implications for the immediate knowledge of God.

A. Alston’s Reliabilist Epistemology of Theistic Belief

In several essays William Alston had developed an externalist account of justification according to which a belief is justified just if it based upon an internally accessible adequate ground.[43] While this sounds internalist, it is not, at least not entirely. Alston thinks that justification has one necessary internalist constraint, namely that beliefs be based on grounds that are accessible to the cognizer. But the allegiance to internalism ends there. The “adequacy” of a ground is understood in externalist terms. An adequate ground is a ground that is sufficiently indicative of the truth of the belief. More precisely, an adequate ground is one that renders the belief highly factually probable (as explained above). Nor does Alston require any higher-level requirements, e.g., that the subject believes or is able to determine that his grounds are adequate.

This view of justification is applied to belief in God in Alston’s book Perceiving God in which he presents a case for the immediate justification of religious beliefs on the basis of religious experience. Alston argues that the putative experiential awareness of God can be construed as significantly analogous to our perception of objects in the physical environment since it exhibits many of the essential generic features found in sensory perception. If there is a perception of God, howbeit a perception typically lacking sensory qualia, religious experience cannot be reduced to a realm of subjective feelings. God will be presented to consciousness in much the same way that objects in the physical environment are presented to consciousness through sensory perception. The presentation is direct, and thus it is not mediated by the perception of other things. Moreover, beliefs formed by way of the perception of God are, like sensory perceptual beliefs, formed immediately, not arrived at inferentially or based on other beliefs.[44]

The putative perception of God yields beliefs to the effect that God is doing something in relation to the subject (e.g., forgiving, loving), or that God has perceivable property (e.g., goodness, power). Alston calls these M-beliefs, “M” for manifestation. But rather than view the formation of such beliefs as an individual and isolated phenomenon, Alston emphasizes construing the formation of such beliefs a socially established doxastic practice. We find that the formation and sustenance of most of our beliefs (e.g., introspective, sensory perceptual, memorial, inferential beliefs) is intimately related to a range of dispositions to form and hold such beliefs in a wide range of contexts. Alston refers to these dispositions or habits of belief formation as doxastic practices. When such dispositions are socially guided and shared, they are socially established. Alston argues that M-belief formation can be construed a socially established doxastic practice since it shares many of the essential features of standard doxastic practices, especially sense perception.[45] (Alston calls this practice MP for “mystical practice,” or alternatively CMP, for “Christian mystical practice” where the MP practice is specifically located in the Christian tradition).

Alston argues that considerations of general epistemology lead us to conclude that it is rational to engage in socially established doxastic practices whose belief outputs are free from significant and massive internal and external contradiction, as well as that receive significant self-support. That is, following the lead of Thomas Reid, all socially established doxastic practices must be regarded “innocent until proven guilty,” a theme we noted earlier in Wolterstorff. The main argument, then, depends heavily on showing that MP satisfies the criteria of rational acceptability that apply to all socially established doxastic practices. But central paradigm here to which theistic perception is compared is the practice of sensory perceptual belief formation. Given the similarities between sensory perception and theistic perception, to the extent to which we think of our perceptual beliefs as prima facie immediately justified (and Alston addressed the various objections to this), we have a basis for thinking of the non-sensory perceptual awareness of God as a ground for prima facie immediately justified beliefs about God. I say prima facie here because justification on such grounds can be overridden by sufficient reasons to the contrary. A good part of Alston’s argument, then, aims at considering the plausibility of the application of epistemological insights regarding our sensory perceptual doxastic practice to MP and answering various objections this application.

Perceiving God does not directly enter into the question of the knowledge of God, but the extension of Alston’s case for immediately justified theistic belief to immediate knowledge of God is rather straightforward.[46] While Alston has reservations about whether justification is necessary for knowledge,[47] he does think that justified true belief is sufficient for knowledge (at least where justification is understood in Alston’s sense), and that the bulk of human knowledge, including much religious knowledge, will fall into this category. For instance, if justified, cases of theistic beliefs formed on the basis of the non-sensory perceptual awareness of God, will satisfy Alston’s the constraint of being based on internally accessible grounds, even if the adequacy condition is not accessible. But if such beliefs are true, then (barring Gettier qualifications) they will count as knowledge, immediate knowledge in particular. But Alston recognizes that a belief produced by a sufficiently reliable belief forming process, where there is nothing like a ground present, will also count as knowledge.

The way of pure reliabilism opens up further ways of thinking of the immediate knowledge of God, as Alston himself has recognized and suggested. Consider cases where a person simply finds himself believing certain religious doctrines, but he cannot specify any ground for such a belief. Whilst there may be theological explanations for the formation of the belief, he may not be aware of such explanations. Consider the prominent Reformed idea that many Christian beliefs are produced by the inward testimony of the Holy Spirit or the influence of divine grace. We may think of such models as involving God directly causing people to hold certain Christian beliefs, perhaps in conjunction with the testimony of Scripture or the gospel preaching. But in that case, it is reasonable to suppose that with the formation of such beliefs (initially at any rate) a person would typically lack an awareness of the process by which such beliefs were formed. They would have nothing in the way of internally accessible grounds on which the belief is based, just a sort of impulse to believe. The same thing can be true in the case of theistic beliefs simpliciter. Consider theistic intuitionism examined in the previous chapters. Some of the Reformed thinkers surveyed regarded theistic belief as a sort of rational intuition of the mind. In some sense, God’s existence is just seen to be true, in a way similar to the intuition of a apriori truths or the knowledge of our own existence. There is nothing in the way of grounds in these cases, but reliabilism can provide an epistemological framework that makes these ideas coherent and credible. If knowledge of God depended solely on a belief’s being produced by some reliable process of belief formation, then there is no further requirement that the subject hold any beliefs about the process being reliable or any other sort of adequacy assumptions. The internalist sting of evidentialism is gone.

B. The Truth of the Reliability Model

Why suppose, though, that Alston’s model is true? That is, why suppose that theistic beliefs are produced by a reliable process or that a doxastic practice such as MP is a reliable ground for the formation of theistic beliefs?

Alston thinks that the question is a difficult one to resolve. The locus of the problem has little to do with M-beliefs in particular, but with a certain stubborn epistemic fact we encounter in examining all fundamental or basic sources of belief. Consider several of such sources: introspection, sense perception, memory, deduction, and induction. Alston argues that all attempts to show by way of argument that these socially established doxastic practices are reliable runs into a problem of circularity. While such arguments do not utilize the conclusion <doxastic practice Q is reliable> as a premise, that conclusion is in a sense presupposed in the premises. There is a commitment to the conclusion as a presupposition of taking ourselves to be justified in holding the premises. Assuming the reliability of a source of belief in giving an argument for its reliability Alston calls epistemic circularity. For instance, Alston argues that attempts to show that the sense perception is reliable are infected with epistemic circularity, since they appeal explicitly or implicitly to premises that rely on sense perception for their justification. While Alston leaves open the possibility that the practice of M-belief formation might not suffer the same fate, he is content to build his case on the assumption that it does. The important point to note is that if we are rational to engage in the sensory perceptional doxastic practice even though it cannot be shown to be reliable in a non-circular way, there is no decent objection to taking the practice of M-belief formation to be reliable simply because it cannot be shown as such in a non epistemically circular manner.

Hence, we must settle for something less impressive at this juncture. If we cannot show that the practice of forming M-beliefs is a reliable doxastic practice, we can nonetheless conclude that it is in a sense rational to suppose that this practice is reliable. But if there is no non-epistemically circular argument for the reliability of such a practice, how can it be rational to suppose that it is? Well it certainly isn’t rational to suppose this in the sense that we have solid grounds or sufficient reasons for taking the practice to be reliable. It isn’t rational in that sense. As Alston notes, “I only claim that we cannot be faulted on grounds of rationality for forming and evaluating beliefs in the ways we normally do (absent any sufficient overriding considerations), since there are no alternatives that commend themselves to rational reflection as superior.”[48] Alston is targeting a weaker sense of rationality, what he calls practical rationality. Such rationality does not imply any likelihood of truth for the belief that MP is reliable, but it practical rationality still sustains an important connection to the matter of reliability.

The practical rationality of engaging in any particular doxastic practice is a function in part of how firmly rooted such a practice is in our lives, it’s being embedded in wider spheres of practice, widely distributed, and the failure of there being any genuine alternatives open to us. Of course, the rationality of engaging in any doxastic practice is prima facie. It can be overridden by sufficient reasons to the contrary in the form of massive or persistent internal or external contradiction. Alston, of course, aims to show that potential candidates here against MP do not go through. Similarly, the prima facie rationality of any practice, MP included, can be strengthened by various degrees of self-support. And Alston attempts to show why MP satisfies this condition. And what of the connection to reliability? According to Alston, although showing it (practically) rational to engage in MP does not amount to showing that MP is reliable, he does take it that he has shown that it is rational to suppose that MP is reliable. Judging any doxastic practice to be rational implies that one is committing oneself to the reliability of that practice. In holding any belief p, one is committed to p being true. But since doxastic practices are responsible for the formation of one’s belief, the commitment to the truth of one’s beliefs within any doxastic practice implies a similar commitment to the practice’s yielding mostly true beliefs, hence being reliable. Of course, to say that one is committed to the reliability of some doxastic practice Q does not mean that one judges or believes that Q is reliable. It simply means is that it would be irrational for one to believe the practice to be rational and yet deny the reliability of Q (or withhold the belief that Q is reliable upon considering the matter). Hence, by virtue of showing that it is rational to engage in MP, Alston has shown that it is rational to suppose that MP is reliable.[49] This is about as far as we can go in terms of any external validation of MP.

C. The Theological Validation of Immediate Knowledge of God

While the prospects of showing that MP is a reliable source of theistic belief, in a way that avoids circularity, faces insuperable difficulties, we may raise a question here that bears on the larger context of this chapter and the next. We are considering the philosophical case for and defense of immediate knowledge of God for the sake of making determinations about the status of natural theology as a source of knowledge of God. It is possible to raise this question from within a theistic or Christian viewpoint. We have already considered how some of the objections to natural theology covered in other chapters rest on theological premises. But as Alston himself has argued, the considerations of epistemic circularity imply that whether we have knowledge in any particular sphere is a matter that can only be settled from within that sphere, by relying on what is presumably known in that domain itself.[50] And while this generally raises the specter of epistemic circularity, it may nonetheless prove useful for our purposes. For we may pose the question as follows: if we assume the truth of theism, does Alston’s model carry a likelihood of being true as an account of how God is naturally known?

I think Alston’s model becomes plausible if we suppose that there is a God, as that at least puts theistic perceptions in the general metaphysical ballpark as it were. If one supposes that there is no God, while people could have perceptions of what they take to be God, beliefs about God that issue from such perceptions will be false. If theism is true, then at least one rather basic presupposition for the truth of M-beliefs will be satisfied. But I don’t see that the mere truth of theism makes it very probable that MP is reliable. Suppose that God is not the sort of being presupposed in putative perceptions of him. For instance, if God is not good, perceptions of Him as such will obviously be false. If God has no relationship to human persons, then putative perceptions of God doing things vis-à-vis the cognizer will be false. Alston’s model will gain a greater degree of plausibility, though, once more specific theological assumptions are adopted that tell us something of what God is like and whether and how God is involved in the lives of human persons. Here we must obviously consult other sources of theological beliefs. But plenty present themselves. Consider the role of Scripture in the Christian tradition in informing us about the nature of God and divine-human interaction. Here there is a compelling case to suppose that God interacts with people from many different walks of life in a multiplicity of different ways. We need only recall the discussion of general and special revelation in chapter one to clearly see this. God has not only directly communicated with people through supernatural means, but He also makes his presence and activity in our lives known in more ordinary, though diverse ways. The broad features of MP would seem to comport well with some of these theologically sanctioned views of what God is like and how He interacts with humans.

Matters are no doubt more complicated. Alston, for instance, recognizes that MP varies in different cultures. And even in any particular tradition there will be a need to separate genuine from counterfeit perceptions of God. It is interesting to note that many theological traditions have spent considerable time developing criteria to accomplish this. The truth of Alston’s model would not require the uncritical acceptance of all claims to direct perceptions of God, anymore than the reliability of the sensory perceptual doxastic practice commits one to regarding all putative sensory perceptions as veridical. The important point here is that support for the contention that MP is reliable can in principle be achieved by consulting our other sources of theological beliefs. And while this bars an extrafaith case for the knowledge of God, it does not prevent an intrafaith reflective consideration of the nature of the mode(s) by which God is known. The theological sources naturally do not inform us much about epistemological details, but they do provide necessary theological premises. These premises, in conjunction with Alston’s general epistemology, render his model very plausible indeed. Thus there is a potentially strong case for the theological validation of the reliability of MP and thus the immediate knowledge of God.

V. Alvin Plantinga’s Model of Immediate Knowledge of God

A. Plantinga, Warrant, and Proper Function

Like Alston, Plantinga’s case for the immediate knowledge of God rests in large part on a rejection of the internalist assumptions typically involved in the evidentialist requirement and the articulation of a distinctly externalist epistemology. Unlike Alston, Plantinga thinks that knowledge requires something more than the satisfaction of a reliability constraint. According to Plantinga, knowledge is sufficiently warranted true belief; and 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.

Proper function[51] implies 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 specifications relevant for warrant are truth-oriented, they are specifications for that segment of our 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 for the purpose of acquiring true beliefs and avoiding false ones.[52] Proper function isn’t enough though. Proper function must take place in the right sort of cognitive environment, the one for which our faculties were designed. There must be the right sort of fit between our cognitive systems and the external environment.[53] Moreover, the design plan must be a good one vis-à-vis the truth goal of believing. There must be a high objective statistical probability that a belief produced by a certain faculty is true. Hence, a reliability constraint is necessary. Lastly, the degree of warrant is a function of the degree of belief. Given the satisfaction of the warrant conditions, the more firmly a belief is held, the more warrant it will have.

Plantinga’s warrant epistemology 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.[54] Nor does Plantinga require any higher-level requirements to the effect that the subject believe or come to see upon reflection that his faculties are functioning properly. Viewed this way warrant entails a kind of rationality, proper function rationality. But there are two forms of proper function rationality. Internally rational beliefs are beliefs formed by proper functioning faculties “downstream from experience.” That is to say they involve the proper response to whatever is given to us by way of sensuous and or doxastic experience. Internal rationality is also a function of a certain degree of coherence between beliefs. Internal rationality also involves whatever is required by proper function in regards to epistemic responsibility, and thus overlaps with deontological rationality. By contrast, external rationality is proper function in the origin or source of our experiences. Since warrant requires proper function, it clearly requires internal and external proper function rationality.

B. Warrant, Theism, and Natural Knowledge of God

How does Plantinga’s warrant epistemology lend itself to an account of the immediate knowledge of God?

In Warranted Christian Belief Plantinga applies the proper function model to theistic and Christian belief. Here I will only consider its application to theistic belief. Plantinga presents a model for how theistic belief could have warrant in a basic way. The model draws on insights found in Aquinas and Calvin. In chapter four we considered Plantinga’s interpretation of Calvin’s treatment of the natural knowledge of God in the beginning of the Institutes. This interpretation of Calvin, of course, represents the sort of model of belief in God that Plantinga himself proposes, with a little bit of inspiration from Aquinas’s statement, “to know in a general and confused way that God exists is implanted in us by nature.”[55] According to Plantinga’s A/C model (“A/C” for Aquinas/Calvin), we have a natural disposition to form various beliefs about God in a wide range of experiential circumstances. Following Calvin, Plantinga refers to this disposition as the sensus divinitatis, a natural mechanism, faculty, or process that yields theistic beliefs in the appropriate circumstances. While theistic belief has this experiential tethering, the circumstances do not involve the direct perception of God, as on Alston’s model.[56] Rather they are largely publically observable phenomena: starry night sky, the majestic grandeur of the mountains, beauty of a small flower, etc. We do not, of course, infer theistic beliefs from these observations by way an argument of some sort. Much like the formation of sensory perceptual beliefs, theistic beliefs arise naturally or spontaneously in these circumstances. Thus, like sensory perceptual beliefs, theistic beliefs formed by way of the sensus divinitatis are basic. The model further takes it that our cognitive systems have been designed by God. While other faculties designed to provide us with true beliefs about ourselves and the world in which we live, the purpose of the sensus divinitatis is to furnish us with true beliefs about God.

Plantinga argues several points regarding the A/C model. First, if this model is true, then theistic belief easily has warrant in a basic way. Since the purpose of the sensus divinitatis is to produce true beliefs about God, theistic belief originating from the sensus divinitatis would (typically at any rate) be the result of properly functioning, truth-aimed cognitive faculties or processes. Moreover, we are in the right sort of cognitive environment. On the model, God designed our cognitive faculties to operate in the world in which we find ourselves. Hence, the sort of conditions that elicit theistic belief are themselves part of our cognitive design plan. Hence, the model, if true, easily satisfies the conditions of warrant. [57] Secondly, while Plantinga does not claim that the model is true (though he believes this), he does claim and argue that it is epistemically possible. In other words, the model is logically consistent with what we know, that is, with what the various participants in the epistemological discussion, theists and non-theists, agree on. To argue this point, Plantinga considers a broad range of potential and actual objections to the mode, and answers them. Third, it follows from the previous claims that it is epistemically possible that theistic belief is warranted in basic way. This is, of course, the main thesis that Plantinga argues.

Lastly, while Plantinga does not argue for the truth of the model, he does argue that the truth of the model follows (or is at least likely) given the truth of theism. And hence if theism is true it is at least likely that theistic belief is warranted. How does that argument go? Well, if one thinks of humans as created by God, one will not think of theistic belief as the product of belief forming processes aimed at something other than truth, much less as the product of some intellectual defect. If God exists and has created human persons in His image for the purpose of entering into communion with Him, then it is natural to think that God desires us to hold true beliefs about Him and our duties to Him. So He would probably create us in such a way that we can achieve this cognitive goal. This seems likely if theism is true. But then it is also likely that the faculties that produce theistic belief have been aimed at this end by their designer. In that case, theistic belief is the product of cognitive faculties functioning properly according to a design plan successfully aimed at truth.[58]

By contrast, Plantinga also claims that if theism is false, then it is likely that theistic belief does not have warrant in a basic way. A few considerations will suffice here. Since, according to Plantinga, a false belief cannot have warrant sufficient, with true belief, for knowledge, the falsity of theism would entail that it cannot have enough warrant for knowledge. But Plantinga thinks that it couldn’t have any warrant at all. If false belief has warrant, then it is usually because it is produced by a faculty functioning at the limit of its capacity. I look into the distance and see what appears to be my Siamese cat Suisa. But it’s really the neighbor’s cat from the around the block. At such and such a distance, my visual processes reach a limit of effectiveness. The resultant belief need not lack warrant altogether. After all, my belief is produced by properly functioning, truth-aimed cognitive faculties, and the general environment seems right.[59] But if there is no God, then our noetic establishment would lack any truth-aimed faculty or process for belief in God. There would simply be no sensus divinitatis. Hence, no relevant faculty could be functioning at its limit since no such faculty exists at all. At best, it might be as Freud suggests. Basic theistic belief would be the product of properly functioning non truth-aimed cognitive faculties, presumably aimed at preserving psychological health in the face of the hostile forces of one’s environment. But such beliefs still fail to be warranted.

While Plantinga’s arguments may strike some as a mild victory at best, his argument is significant on at least two grounds. First, from the vantage point of apologetics, if Plantinga’s arguments are sound, it follows that there can be no objection to the positive epistemic status of theistic belief independent of objections to the truth of theism. Warranted Christian Belief draws a contrast between de jure and de facto objections to theism and Christian belief. The former concerns the rightness, propriety, of theistic belief, whereas the latter concerns the truth of theistic belief. Plantinga explains that traditionally philosophers have desired to develop de jure objections quite independent of de fact objections. But this separation of epistemology and metaphysics is implausible as Plantinga sees it. How we think of knowledge, what we know and how we it, depends in large part on metaphysical assumptions at the outset. This is especially the case where knowledge is understood along the lines of proper function. Secondly, from the vantage of the Christian faith, since the truth of the model follows from theism, Plantinga has provided a model for how Christians can think about the natural knowledge of God. Here we find a connection with Alston’s arguments about the validation of our basic sources of belief. It must be an inside job. Plantinga concurs. But while this may beg the question from an extrafaith perspective, it can be wielded in an intrafaith context to explain how it is that human persons acquire their natural knowledge of God. If Plantinga is correct about how it is that we in fact form theistic belief, as well as about the epistemological implications of theism, then it looks like there is an intrafaith case for immediate natural knowledge of God. On that note it is only natural to ask what place, if any, natural theology can be afforded within such a model.

VI. Conclusion

The case for immediate knowledge of God, as it is developed by Reformed epistemology, has several facets. On the one hand, it involves a critique of the epistemological assumptions that ground objections to the immediate knowledge of God. Secondly, while it aims to show that basic theistic belief is rational or justified (in the internal and deontological senses), it recognizes that there are inescapable limitations when it comes to showing that theistic belief, even if true, can constitute knowledge, immediate or otherwise. Most importantly, Alston and Plantinga have each brought into sharp focus the metaphysical questions inextricably woven into epistemological fabric, for both our knowledge of God and the external world. This carries with it implications for the intrafaith assessment of how it is that God is known, and in particular whether God can be immediately known. But having considered Reformed epistemology’s case for immediate knowledge of God, we can now critically engage this model and consider the extent to which it actually undermines the epistemic function of natural theology.

---

[1] Here we need two important clarifications. First, evidentialism in the present sense should not be confused with what is commonly evidentialist apologetics. The latter advocates the use of theistic arguments to defend the faith and show that theism is reasonable. This is not the same as requiring evidence for one’s belief to be rational, justified, or constitute knowledge. For more on this distinction, see chapter eleven and my paper “Reformed Epistemology and Christian Apologetics” (Forthcoming, Religious Studies). Secondly, I refer to the denial of immediate knowledge of God in long-standing epistemological traditions (plural) because there are important distinctions between the sort of evidentialism one finds in medieval philosophy and the variety of evidentialisms proximately connected to the Enlightenment tradition. The sense in which medieval philosophers like Aquinas were evidentialists is closely connected to an understanding of knowledge as scientia, knowing the cause of a thing by way of a demonstrative syllogism. This is clearly compatible with affirming the immediate knowledge of God in other senses that have loomed large since the Enlightenment. Moreover, epistemological traditions that draw on Enlightenment insights are thick with other epistemic concepts such as rationality and justification whose place in the medieval tradition is far from clear. So we should probably distinguish between pre-modern and modern evidentialism. The focus in this chapter will be on the latter with respect to rationality and knowledge. For views on the nature of evidentialism in relation to the medieval period, see Nicholas Wolterstorff, “The Migration of Theistic Arguments: From Natural Theology to Evidentialist Apologetics,” in Rationality, Religious Belief, and Moral Commitment: New Essays in the Philosophy of Religion, ed. Robert Audi and William Wainwright (Ithaca: Cornell University Press, 1986); Arvin Vos, Aquinas, Calvin, and Contemporary Protestant Thought (Grand Rapids, MI: Eerdmans Publishing Company, 1985), chapter three; Anthony Kenny, What is Faith? Essays in the Philosophy of Religion (New York: Oxford University Press, 1992), p. 43; Brian Davies, The Thought of Thomas Aquinas (Oxford: Clarendon Press, 1993), chapter two; Norman Kretzmann, The Metaphysics of Theism (Clarendon Press, 1998), pp. 1-53.

[2] The designation "Reformed epistemology” stems from similarities between the philosophical thesis of immediate knowledge of God and ideas about belief in God in the Reformed theological tradition. Although Plantinga and Wolterstorff belong to this theological tradition, Alston and others who advocate the proper basicality of theistic belief do not. The designation "Reformed epistemology" (coined by Plantinga to described to contemporary movement in philosophy of religion) is perhaps an unhappy and misleading one, but it is no more problematic than popular alternatives like "theistic foundationalism" and "anti-evidentialism." For a comparison of the contemporary Reformed epistemology movement with the Reformed theological tradition, see John Frame, “The New Reformed Epistemology” in Frame, The Doctrine of the Knowledge of God (Phillipsburg, N.J.: Presbyterian and Reformed, 1987), pp. 382-400.

[3] Again, the focus here is modern and contemporary evidentialism: e.g., John Locke, David Hume, W.K. Clifford, Brand Blanshard, A.J. Ayer, Bertrand Russell, Anthony Flew, J.L. Mackie. Some evidentialists are thesists; others are not. Moreover, as I will explain below, rationality (or justification) is distinct from knowledge in several respects. The evidentialist requirement may be and has been imposed on both or only one of these epistemic desiderata.

[4] John Greco writes, “Reformed epistemologists argue that natural theology is not necessary for theistic knowledge.” See “Is Natural Theology Necessary for Theistic Knowledge,” in Rational Faith: Catholic Responses to Reformed Epistemology, ed. Linda Zagzebski (Notre Dame: Unviersity of Notre Dame Press, 1993), p. 168. Hugo Meynell and Patrick Lee express a similar concern for the marginalization of natural theology in the same volume.

[5] [JE] designates “justification evidentialism.” Later in the chapter I will distinguish this form of evidentialism from evidentialism with respect to knowledge. Here I also ignore two variations on the evidentialist requirement. Some versions of evidentialism require only that a person have evidence for theistic belief, not necessarily base theistic belief on this evidence. Other versions of evidentialism incorporate an idea introduced by Locke, namely that belief or assent must be proportioned to the evidence. We may, for the moment, pass over these variations.

[6] William Alston has pointed out all three of these motivations for [SE]. See Alston, “What’s Wrong with Immediate Knowledge,” in Alston, Epistemic Justification (Ithaca: Cornell University Press, 1989), especially pp. 62-64, 70-72. See also Alston’s discussion of the immediate justification of perceptual beliefs in Perceiving God (Ithaca: Cornell University Press, 1991), pp. 77-93.

[7] Naturally, the matter is a bit more complicated. In the case of perceptual beliefs, for instance, information about how the cognizer is situated (in place and time) and positioned (in distance) vis-à-vis the perceived object seems to play an important role in the formation of the belief. Many different objects look similar and can only be discriminated by assumptions about the setting. While standing on a beach, I am appeared to oceanly and form the belief that “the Pacific Ocean is vast.” But from where I am standing, it could just as well be the Atlantic Ocean. It is only because I believe that I am located on a beach on the west coast, as opposed to the east coast, that I form one belief about the Pacific. Similarly, my belief that a particular object is a car rather than a mailbox seems to presuppose beliefs about the relative distance of objects from me. It might be argued, then, that even perceptual beliefs (one of the more plausible candidates for immediate justification) are based on other beliefs, from which they receive their justification in part. But if a ground is something of which the subject is aware, a crucial question is whether cognizers are aware of such doxastic factors as grounds in the formation of perceptual beliefs. This does not always seem to be case. Moreover, we can plausibly account for the relevance of contextual beliefs in other ways. The justification of perceptual beliefs is prima facie. Other beliefs can override or undermine immediately justified perceptual beliefs. It doesn’t follow that such beliefs factored into the basis of the belief. Secondly, if I am asked to provide reason for supposing that I see the Pacific Ocean, I must have recourse to other beliefs. Perhaps some of these allow me to show the adequacy of my perceptual grounds. But the use of beliefs to defend a belief, as well as the reflective awareness of the adequacy of grounds, need not factor into the actual basis of believing. Moreover, it is quite plausible to suppose that overtime contextual information is internalized in such a way that cognizers are able to go directly from perceptions to the target belief as a kind of perceptual skill, in much the same way that the rules of proper driving are learned, internalized, and thereafter take the form of driving skills. See Alston on “contextual beliefs” in Perceiving God, pp. 88-91.

[8] Alston raises this particular objection. See Alston, “The Role of Reason in the Regulation of Belief,” in Rationality in the Calvinian Tradition, ed. Hendrik Hart, Johan Van der Hoeven, and Nicholas Wolterstorff (Lanham, MD: University Press of America, 1983), p. 139.

[9] Robert Audi develops this point his The Structure of Justification, p. 127.

[10] Plantinga, “Coherentism and the Evidentialist Objection to Belief in God,” in Rationality, Religious Belief, and Moral Commitment, pp. 110-112; cf. Plantinga, “Reason and Belief in God” in Faith and Rationality, ed. Nicholas Wolterstorff and Alvin Plantinga (Notre Dame: University of Notre Dame Press, 1983), p. 52.

[11] The following exposition is based on Plantinga’s arguments in “Reason and Belief in God,” in Faith and Rationality.

[12] Plantinga, “Reason and Belief in God,” p. 62.

[13] This point was made by William Alston in an early criticism of Plantinga’s article, “Alvin Plantinga’s Epistemology of Religious Belief.”

[14] Plantinga initially considered this in his “Coherentism and the Evidentialist Objection to Belief in God,” in Rationality, Religious Belief, and Moral Commitment, pp. 109-38.

[15] A good account of the concept of coherence is found in Laurence Bonjour, The Structure of Empirical Knowledge (Cambridge, MA: Harvard University Press, 1985).

[16] This foreshadows what will become Plantinga’s later theory of warrant and proper function. See section V below.

[17] This also provides an alternate way of thinking about evidentialism. Instead of demanding that a person’s theistic belief be based on evidence, the requirement could be that a person must have adequate evidence for their theistic belief (whether of the coherentist sort or not), thereby emphasizing the relation of evidential support, not the basis relation. This could be derived from the general principle that rationality simpliciter involves a person’s merely having evidence.

[18] Properly speaking, blame here attaches directly to activities in which he engaged that had as a result the formation of a particular belief. If beliefs are not within our voluntary control (as surely seems to be the case), it may be improper to attach obligations and duties to believings or what beliefs we form. In that case, the relevant obligations attach to actions or behaviors that are within our control and which influence the sorts of beliefs we end up forming. See Alston, “The Deontological Conception of Justification,” in Alston, Epistemic Justification, pp. 115-152.

[19] Indeed, if coherentism was correct as a criterion for proper basicality, it is hard to see how theistic belief could then be excluded from being properly basic. Surely man who take belief in God as properly basic would not thereby be displaying any defect in their noetic structure. Cf. Plantinga, “Evidentialism and the Coherentist Objection to Belief in God,” pp. 127-128.

[20] Anthony Kenny raises this complaint in What is Faith? (New York: Oxford University Press: 1992), p. 13. Cf. Plantinga’s concluding comments: “Thus far we have found no reason at all to exclude belief in God from the foundations. So far we have found no reason at all for believing that belief in God cannot be basic in a rational noetic structure.” (“Is Belief in God Rational?” in Rationality and Religious Belief, ed. C.F. Delaney (Notre Dame: University of Notre Dame Press, 1979), p. 000).

[21] Plantinga, Warranted Christian Belief (New York: Oxford University Press, 2000), pp. 100-102, 177-178.

[22] An initial problem is the involunatary nature of belief. This implies that duties will apply not directly to our believings but to activities that influence are beliefs. It is this understanding with which I am working in the text. Epistemic responsibility is perhaps a better description.

[23] The account in this paragraph is based on the case Plantinga presents in Warranted Christian Belief, pp. 99-102, though the context there is Christian belief in particular. Plantinga also thinks that the ease with which theistic belief can be shown to be deontologically rational proves that this is not the most plausible sort of objection to the rightness or propriety of theistic belief. The better objection lies in the neighborhood of non-defectiveness or proper function. See Plantinga, Warranted Christian Belief, pp. 102, 108. Also see below, section V.

[24] See Wolterstorff, "Can Religious Belief be Rational if it Has No Foundations" in Faith and Rationality, pp. 135-186.

[25] For a detailed account of Locke’s epistemology, see Wolsterstorff, John Locke and the Ethics of Belief (Cambridge: Cambridge University Press, 1996). Pages 60-86 are particularly important to the concepts of obligation and entitlement to belief.

[26] Though not a Reformed epistemologist, Richard Swinburne also defends a similar thesis, with similar implications for the rationality of belief in God. According to Swinburne, the principle of credulity is a principle of rationality that asserts, roughly, that all things being equal, things are as they appear to be. (Swinburne contrasts with another principle of rationality, the principle of testimony, which is equivalent to what Reid called the principle of credulity). “This Principle,” writes Swinburne, “leads to the view that the person with justified beliefs, the rational person, is the credulous person; she believes everything until she has a reason not to do so” (Swinburne, Epistemic Justification (Oxford: Clarendon Press, 2001), p. 142). In other words, we are epistemically right to begin where we are, with the beliefs we have, even though subsequent considerations may defeat this prima facie rationality. This is essentially the point being made by Plantinga and Wolterstorff. For Swinburne’s account of the principle of credulity, see his The Existence of God (Oxford: Clarendon Press, 198), pp. 254-271, and Epistemic Justification, chapter five.

[27] The actual formulation has a couple of other qualifications. If a person has adequate reason to give his belief, but is justified in not recognizing this fact, he is rationally justified in his belief. And it is also necessary that the subject not be rationally obliged to think that he has an adequate reason to cease holding the belief. See Wolterstorff, “Can Belief in God be Rational. . . ,” pp. 163-169.

[28] This is true both with respect to Reformed epistemology and evidentialists. Alvin Plantinga, Warranted Christian Belief; William Alston, “On Knowing that We Know: The Application to Religious Knowledge,” in Christian Perspectives on Religious Knowledge, ed. C. Stephen Evans and Merold Westphal (Eerdmans, 1993); Alston, “Knowledge of God,” in Faith, Reason, and Skepticism, ed. Marcus Hester (Philadelphia, Temple University Press, 1991); Evans, “Empiricism, Rationalism, and the Possibility of Historical Religious Knowledge,” in Christian Perspectives on Religious Knowledge, ed. Evans and Westphal; Evans, “Evidentialist and Non-Evidentialist Accounts of Historical Religious Knowledge,” International Journal for Philosophy of Religion 35 (1994), pp. 153-182. But evidentialism with respect to knowledge has become an increasingly emphasized position, e.g., Steve Wykstra, “Toward a Sensible Evidentialism” in Philosophy of Religion: Selected Readings, 2^nd edition, ed. William Rowe and William Wainwright (San Diego: Harcourt Brace Jovanovich, Publishers, 1989), pp. 427-437; and John Greco, “Is Natural Theology Necessary for Theistic Knowledge,” in Rational Faith, ed. Linda Zagzebski, pp. 168-198.

[29] I say “typically” because we sometimes use the terms rational and justified in a way that is not so closely connected to a positive connection to truth. Some beliefs are practically, morally, or prudentially rational or justified. This is so when they are conducive not necessarily to believing what is true, but believing in such a way that certain practical or non-truth goals can be achieved, survival, happiness, etc.

[30] Of course, the affirmation of epistemic evidentialism is logically consistent with the denial of evidentialism with respect to deontological justification or rationality. One may quite sensibly maintain that a theistic may be within her intellectual rights in believing in God in a basic way, but that this belief, even if true, does not qualify as knowledge, absent the satisfaction of some evidentialist requirement.

[31] Gettier, “Is Justified True Belief Knowledge?” Analysis 23 (1963), pp. 121-123. The up-shot of the Gettier problem is that knowledge must not be accidentally true belief. Gettier argued that that the conjunction of justification (construed in an evidentialist or deontological sense) and true belief is logically consistent with a person’s believing what is true as a matter of epistemic luck. I look at my clock and it says 2:30. I believe it is 2:30. It is 2:30. So my belief is true. It seems that I have good evidence for it. My clock has always been reliable in the past. But in fact my clock broke the previous day at precisely the time at which I am looking at it today. I hold a true belief by sheer epistemic serendipity. It is counterintuitive to suppose that I know the time of day. For more on Gettier, see chapter ten.

[32] It isn’t merely that I can be within my intellectual rights in believing that I had Honeycomb for breakfast, but I can believe this with warrant and know it (immediately), even if it isn’t self-evident, evident to the senses, or about my current states of consciousness. Similarly, at least some of the (true) propositions we accept on testimony are warranted. However, unlike rationality classical foundationalism, the sting of self-referential incoherence is not as painful for classical foundationalism with respect to warrant. The classical foundationalist can admit that his belief in classical foundationalism is – pending an argument – unwarranted. But owing to distinction between justification and warrant he can nonetheless recommend classical foundationalism as justified, a weaker positive epistemic status.

[33] I shall examine defeasibility theories in greater detail in a subsequent chapter, as they have consequences for whether the noetic effects of sin undermine natural knowledge of God.

[34] Clearly the stronger form of internalism maintains that the adequacy of a ground, not just the ground itself, is something that is cognitively accessible. As we will see below, this implies that adequacy will be spelled out in terms of certain logical relations between propositions that can in principle be introspectively accessed or grasped upon reflection. For this reason, warrant ordinarily involves not merely an adequate ground, but this adequacy is something that the subject can in principle come to see upon reflection. Hence, for any warranted belief B, the subject can in principle tell upon reflection, though not necessarily infallibly, whether B is warranted.

[35] See Alston, “Knowledge of God” in Faith, Reason, and Skepticism, and “On Knowing that We Know” in Christian Perspectives on Knowledge.

[36] To some extent we also see a connection with the deontological tradition here. Presumably among our epistemic duties will be the general duty to conduct our intellectual life in such a way that as far as we can see we are maximizing true beliefs and minimizing false beliefs. But this would seem to require access to the adequacy of the grounds. My believing p on the basis of q does not constitute intellectual dutifulness unless I am justified in supposing that q provides the appropriate sort of support for p. I must be justified in believing that q adequately supports p, otherwise I would not be conducting my intellectual life in such a way that relates positively to the epistemic goal of believing. So the internalist idea of access to the adequacy or efficacy of one’s grounds of belief is a reasonable implication of both evidentialism and deontologism. For more on this, see Plantinga, Warrant: The Current Debate (New York: Oxford University Press, 1993), chapter one.

[37] Even those forms of internalism that maintain that a belief’s being warranted is a function at least in part of its being made probable by the subject’s other (warranted) beliefs would be consistent with either partial immediate knowledge or merely psychologically basic beliefs.

[38] Alston, “Knowledge of God,” pp. 18-20.

[39] Alston, ibid., pp. 24-25.

[40] Internalists may disagree about whether it is sufficient for justification that a proposition p be probable relative to S’s evidence and what he (justifiably) takes to be correct inferential standards or whether it is also required that the standards in fact be correct. It is also possible to regard the latter as affording a stronger degree of justification than the latter, but perhaps a degree necessary for knowledge. There is also a distinction between whether the probability of h on evidence e relative to correct inductive standards is being assessed by beings with limited or unlimited logical capacities for determining the logical consequences of propositions and all relevant possibilities and necessities. See Swinburne, Epistemic Justification, chapter three.

[41] See Alston, “How to Think about Reliability” Philosophical Topics 23 (1995).

[42] Some do, however, require that the subject not believe that his belief was produced in an unreliable manner or was produced by cognitive malfunction. For more on the discussion on “defeaters” in the following chapter

[43] The primary essays here are “Internalism and Externalism in Epistemology” and “An Internalist Externalism” in Alston, Epistemic Justification, pp.185-226, 227-245.

[44] Subject to the qualification, to be explored in the next chapter, that some of cases of M-beliefs are partially based on both the experiential awareness of God and other beliefs. Such cases represent partially immediate beliefs.

[45] There are several important details that I pass over here, some of which I will consider in the next chapter.

[46] Alston does briefly address this epistemic extension of his model in Perceiving God, pp. 284-285.

[47] Alston’s argument against the necessity of justification for knowledge is argued in “Justification and Knowledge,” in Alston, Epistemic Justification (Ithaca: Cornell University Press, 1989), pp. 172-182. See chapter seven for a consideration of one of Alston’s arguments here.

[48] Alston, Perceiving God, p. 168.

[49] See Alston, Perceiving God, pp. 178-180. Alston’s argument at this juncture involves a kind of “multilevel view” that distinguishes between a weak sense of rationality (attributed to suppositions about the reliability of doxastic practices) and a stronger sense of justification (attributed to the outputs of the doxastic practice if it is reliable). As Alston recognizes (p. 182), it is also possible to state the argument in a simpler fashion solely in terms of a weaker notion of justification. The argument would then be: (i) a belief is prima facie justified (in a non truth conducive sense) just if it is stems from a socially established doxastic practice that is not discredited by its total output, (ii) beliefs formed on the basis MP satisfy the conditions stipulated in (i), therefore (iii) beliefs formed on the basis of MP are prima facie justified. Alston attempts this sort of weaker approach in “Christian Experience and Christian Belief” in Faith and Rationality, pp. 103-134. In this earlier essay, Alston argues that M-beliefs and perceptual beliefs can be affirmed as prima facie justified in a “weak normative” sense on the basis of their respective doxastic practices. This weak normative sense of justification is simply a matter of permisson to believe p unless one has adequate reason to suppose that p is false (or unreliably formed). It stands in contrast to strong normative justification, which is a matter of being obligated to refrain from believing p unless one has adequate reasons for supposing that p is true or reliably formed). While Alston finds this earlier approach permissible, the virtue of the approach in Perceiving God, as he sees it, is that it strikes a connection to reliability by showing it rational to take MP, and other doxastic practices, as reliable.

[50] Alston develops this point in considerable detail in “On Knowing that we Know” in Christian Perspectives on Religious Knowledge, especially pages 32-39.

[51] By “proper function” Plantinga understands an absence of cognitive dysfunction, impairment, disorder, or pathology in some person’s holding B. Moreover, a particular doxastic state might be the product not of malfunction but of properly functioning cognitive faculties in some way overridden by certain emotional states (e.g., anger, ambition, lust). Warrant requires properly functioning cognitive powers not impeded, inhibited, or overridden by emotions of a certain type. I will understand “proper function” throughout to include these qualifications.

[52] It is, of course, not necessary that all of a person’s cognitive faculties be functioning properly. Only the faculties responsible for the production or sustenance of the belief in question need be functioning properly, and even the need not be functioning properly over the entire range of their operation.

[53] Since the publication of Warrant and Proper Function Plantinga has fine-tuned this aspect of warrant to address a variety of counterexamples to his warrant theory. In short, the necessary adjustment requires a distinction between maxi and mini environments. The former is the large- scale environment for which our cognitive faculties have been designed, whereas the mini-environment involves highly localized circumstances in the context of the formation of specific beliefs. The degree of warrant needed for knowledge requires a favorable minienvironment. See Warranted Christian Belief, pp. 156-161. See also chapter ten for a more detailed discussion of this.

[54] Similarly, I cannot tell upon reflection whether the reliabilist constraint is satisfied, nor whether I am in the environment for which my cognitive faculties were designed.

[55] Aquinas, Summa theologiae I, q.2, a.1, ad 1.

[56] For comparisons of Alston’s M-beliefs and Plantinga’s theistic beliefs, see Alston, Perceiving God, pp. 195-197 and Plantinga, Warranted Christian Belief, pp. 180-184.

[57] Plantinga recognizes that certain qualifications are needed here. For instance, due to the noetic effects of sin, there are varying degrees of malfunction and impedance of the sensus divinitatis. See Plantinga, Warranted Christian Belief, chapter seven and chapter ten of the present work.

[58] See Plantinga, Warranted Christian Belief, pp. 188-190.

[59] Though note that the problem here overlaps with considerations of how the minenvironment may go wrong. While warrant simpliciter requires the right maxi-environment, the degree of warrant needed for knowledge requires the right minienvironment. See footnote no. 52.