Religious Epistemology

"Enlightenment critiques of the reasonableness of religious belief point to defects not so much in religious belief as in the conceptions of knowledge uncritically adopted as the basis of these critiques. Maybe religious knowledge looks dubious because we have the wrong idea about what it is to know something and how we know what we know."
--C. Stephen Evans and Merold Westphal, Christian Perspectives on Religious Knowledge.


In Religious Epistemology we encounter the intersection of traditional epistemology with the newly burgeoning field of Philosophy of Religion. One of the most rapidly developing areas of Philosophy of Religion has been the exploration of the epistemic status of religious truth claims, their rationality, and warrant.

Positivism and the Presumption of Atheism

We may gain an appreciation of how Religious Epistemology has changed by casting a backward glance at questions facing the previous generation of philosophers. As mid-twentieth century philosophers of religion struggled under the pall of Logical Positivism, they were forced to defend the very meaningfulness of their claims against the attacks of positivists and their philosophical ilk. Positivists championed a Verification Principle of meaning, according to which an informative sentence, in order to be meaningful, must be capable in principle of being empirically verified. Since religious statements like “God exists” or “God loves the world” were, in their opinion, incapable of being empirically verified, positivistic philosophers held them to be literally meaningless, as if one had asserted, “’twas brillig, and the slythey toves did gyre and gimble in the wabe.” Under criticism, the Verification Principle underwent a number of changes, including its permutation into the Falsification Principle, which held that a meaningful sentence must be capable in principle of being empirically falsified. The fate of religious language was thought to be no brighter under falsificationism than under verificationism, as became evident at a famous Oxford University symposium on 'Theology and Falsification' held in 1948.

At the symposium Antony Flew borrowed a story told several years earlier by John Wisdom concerning two explorers who came upon a patch of flowers in a jungle clearing. One explorer was convinced that the flowers were tended by a gardener. In the ensuing days, however, despite the explorers’ every effort to find him, no gardener was ever detected. To save his hypothesis, the one explorer was progressively forced to qualify his original hypothesis to the point that the hypothesized gardener must be invisible, intangible, and undetectable. To which the other finally replied, “Just how does what you call an invisible, intangible, eternally elusive gardener differ ... from no gardener at all?”[1] The gardener in the story is obviously a symbol of God, the putatively invisible, intangible, eternally elusive Creator of the world.

Now we would all agree with Flew that the explorer’s original gardener hypothesis had suffered death by a thousand qualifications. But why? The evident answer is that the hypothesis, like nineteenth century physics’ aether hypothesis, had become increasingly ad hoc, or contrived to fit the data, which counts against its being the best explanation of the facts. But Flew maintained that the problem lay in the fact that anything which would count against an assertion must be part of the meaning of that assertion. Since nothing is allowed to count against the gardener- or God-hypothesis, that hypothesis therefore asserts nothing. On Flew’s view the God-hypothesis is not false but simply meaningless.

Flew’s theory of meaning was clearly mistaken. The very fact that the two explorers in the story could disagree about the merits of the undetectable gardener-hypothesis (or that Flew’s colleagues on the panel understood the story’s ending!) shows that the explorer’s statement was meaningful. The extraordinary ad hoc-ness of the hypothesis counted against its truth, not its meaningfulness.

In general, verificationist analyses of meaning ran into two insuperable problems: (1) The verification/falsification principle was too restrictive. It was quickly realized that on such theories of meaning vast tracts of obviously meaningful discourse would have to be declared meaningless, including even scientific statements, which the principle had aimed to preserve. (2) The principle was self-refuting. The statement “In order to be meaningful, an informative sentence must be capable in principle of being empirically verified / falsified” is itself incapable of being verified or falsified. Therefore, it is by its own lights a meaningless statement—or, at best, an arbitrary definition which we are free to reject. The inadequacies of the positivistic theory of meaning led to the complete collapse of Logical Positivism during the second half of the twentieth century, helping to spark not only a revival of interest in Metaphysics but in Philosophy of Religion as well. Today’s Flew’s sort of challenge, which loomed so large in mid-century discussions, is scarcely a blip on the philosophical radar screen.

Similarly, another philosophical relic is the much-vaunted presumption of atheism. At face value, this is the claim that in the absence of evidence for the existence of God, we should presume that God does not exist. Atheism is a sort of default position, and the theist bears a special burden of proof with regard to his belief that God exists.

So understood, such an alleged presumption seems to conflate atheism with agnosticism. For the assertion that “God does not exist” is just as much a claim to knowledge as is the assertion that “God exists,” and therefore the former requires justification just as the latter does. It is the agnostic who makes no knowledge claim at all with respect to God’s existence, confessing that he does not know whether God exists or does not exist, and so who requires no justification. (We speak here only of a “soft” agnosticism, which is really just a confession of ignorance, rather than of a “hard” agnosticism, which claims that it cannot be known whether or not God exists; such a positive assertion would, indeed, require justification.) If anything, then, one should speak at most of a presumption of agnosticism.

In fact, when one looks more closely at how protagonists of the presumption of atheism use the term “atheist,” one discovers that they are sometimes defining the word in a non-standard way, synonymous with “non-theist," which would encompass agnostics and traditional atheists, along with those who think the question meaningless. As Flew confesses,

the word ‘atheist’ has in the present context to be construed in an unusual way. Nowadays it is normally taken to mean someone who explicitly denies the existence ... of God ... But here it has to be understood not positively but negatively, with the originally Greek prefix ‘a-’ being read in this same way in ‘atheist’ as it customarily is in ... words as ‘amoral’ .... In this interpretation an atheist becomes not someone who positively asserts the non-existence of God, but someone who is simply not a theist.[2]

Such a re-definition of the word “atheist” trivializes the claim of the presumption of atheism, for on this definition, atheism ceases to be a view, and even babies, who hold no views at all on the matter, count as atheists! One would still require justification in order to know either that God exists or that He does not exist.

Other advocates of the presumption of atheism continued to use the word in the standard way and so recognized their need of justification for their claim that atheism is true; but they insisted that it was precisely the absence of evidence for theism that justified their claim that God does not exist. Thus, in the absence of evidence for God, one is justified in the presumption of atheism.

The problem with such a position is captured neatly by the aphorism “Absence of evidence is not evidence of absence.” For example, in theoretical physics entities are frequently postulated for which there is (as yet) no evidence, but that absence of evidence in no way justifies one in thinking that such entities do not exist. To give an illustration, it has become commonplace in astrophysical cosmology to postulate an early inflationary era in the expansion of the universe in order to explain such features of the universe as its flat space-time curvature and large scale isotropy. Unfortunately, by the very nature of the case, any evidence of such an era will have been pushed by the inflationary expansion out beyond our event horizon, so that it is unobservable. But woe be to the cosmologist who asserts that this absence of evidence is proof that inflation did not take place! At best we are left with agnosticism.

Now clearly there are cases in which the absence of evidence does constitute evidence of absence. If some one were to assert that there is an elephant on the quad, then the failure to observe an elephant there would be good reason to think that there is no elephant there. But if someone were to assert that there is a flea on the quad, then one’s failure to observe it there would not constitute good evidence that there is no flea on the quad. The salient difference between these two cases is that in the one, but not the other, we should expect to see some evidence of the entity if in fact it exists. Thus, the absence of evidence is evidence of absence only in cases in which, were the postulated entity to exist, we should expect to have some evidence of its existence. Moreover, the justification conferred in such cases will be proportional to the ratio between the amount of evidence that we do have and the amount of evidence that we should expect to have if the entity existed. If the ratio is small, then little justification is conferred on the belief that the entity does not exist.

Again, the advocates of the presumption of atheism recognized this. Michael Scriven, for example, maintained that in the absence of evidence rendering the existence of some entity probable, we are justified in believing that it does not exist, provided that (1) it is not something which might leave no traces, and (2) we have comprehensively surveyed the area where the evidence would be found if the entity existed. But if this is correct, then our justification for atheism depends on (1) the probability that God would leave more evidence of His existence than what we have and (2) the probability that we have comprehensively surveyed the field for evidence of His existence That puts a different face on the matter! Suddenly the presumer of atheism, who sought to shirk his share of the burden of proof, finds himself saddled with the very considerable burden of proving (1) and (2) to be the case.

The debate among contemporary philosophers has therefore moved beyond the facile presumption of atheism to a discussion of the so-called “Hiddenness of God” --in effect, a discussion of the probability or expectation that God, if He existed, would leave more evidence of His existence than what we have. One’s perspective on this issue cannot but be influenced by one’s assessment of the project of natural theology (see chapters 27-28). For if one is convinced that God has left pretty convincing evidence of His existence, then one is apt to be skeptical that we should expect to see much more evidence of His existence than that which we have. Scriven, in the end, held that we are justified in rejecting the existence of some entity only if the claim that it exists is wholly unsupported, that is to say, there is no particular evidence for it and not even general considerations in its favor. By this criterion, Scriven advocated that we remain merely agnostic, rather than disbelieving, even about such entities as the Loch Ness monster and the Abominable Snowman! But surely any unprejudiced observer will discern as much evidence for God as for the Loch Ness monster.

Unsatisfied with the evidence we have, some atheists have argued that God, if He existed, would have prevented the world’s unbelief by making His existence starkly apparent (say, by inscribing the label “Made by God” on every atom or planting a neon cross in the heavens with the message “Jesus Saves.” But why should God want to do such a thing? As Paul Moser has emphasized, on the Christian view it is actually a matter of relative indifference to God whether people believe that He exists or not. For what God is interested in is building a love relationship with us, not just getting us to believe that He exists. Even the demons believe that God exists—and tremble, for they have no saving relationship with Him (James 2.19). Of course, in order to believe in God, we must believe that God exists. But there is no reason at all to think that if God were to make His existence more manifest, more people would come into a saving relationship with Him. Mere showmanship will not bring about a change of heart (Luke 16.30-31). It is interesting that, as the Bible describes the history of God’s dealings with mankind, there has been a progressive “interiorization” of this interaction with an increasing emphasis on the Spirit’s witness to our inner selves (Romans 8.16-17). In the Old Testament God is described as revealing Himself to His people in manifest wonders: the plagues upon Egypt, the pillar of fire and smoke, the parting the Red Sea. But did such wonders produce lasting heart-change in the people? No, Israel fell into apostasy with tiresome repetitiveness. If God were to inscribe His name on every atom or place a neon cross in the sky, people might believe that He exists; but what confidence could we have that after time they would not begin to chafe under the brazen advertisements of their Creator and even come to resent such effrontery? In fact, we have no way of knowing that in a world of free creatures in which God’s existence is as obvious as the nose on your face that more people would come to love Him and know His salvation than in the actual world. But then the claim that if God existed, He would make His existence more evident has little or no warrant, thereby undermining the claim that the absence of such evidence is itself positive evidence that God does not exist.

Religious Belief without Warrant

One of the presuppositions underlying the original discussions of the presumption of atheism was theological rationalism or, as it has come to be known, evidentialism. According to this view, religious belief, if it is to be justified, must have supporting evidence. Thus, Scriven asserted that if someone claims that “theism is a kind of belief that does not need justification by evidence,” then there must be “some other way of checking that it is correct besides looking at the evidence of it;” but that cannot be right because “any method of showing that belief is likely to be true is, by definition, a justification of that belief, that is, an appeal to reason.”[3] Here Scriven equates holding a belief justifiably with being able to show that belief to be true, and he assumes that an appeal to reason to justify a belief involves providing evidence for that belief. Both of these assumptions have been vigorously challenged by contemporary epistemologists.

A number of thinkers have argued that one can have pragmatic justification for holding a belief, wholly apart from that belief’s being epistemically justified, or knowledge, for the person holding it. Following Alvin Plantinga, let us refer to epistemic justification as warrant, that property which serves to transform mere true belief into knowledge. Proponents of pragmatic arguments aim to show that we are sometimes within our rights in holding beliefs for which we have no warrant. A pragmatic argument seeks to provide grounds for holding a particular belief because of the benefits to be had from holding that belief. Jeff Jordan has helpfully distinguished two types of pragmatic arguments: truth-dependent and truth-independent arguments. A truth-dependent argument recommends holding a belief because of the great benefits to be gained from holding that belief if it should turn out to be true. A truth-independent argument recommends holding a belief because of the great benefits to be gained from holding that belief whether or not it turns out to be true.

The most celebrated and oft-discussed, truth-dependent, pragmatic argument is Pascal’s Wager, the brainchild of the French mathematical genius Blaise Pascal. Pascal argued, in effect, that belief in God is pragmatically justified because we have nothing to lose and everything to gain from holding that belief. Although Pascal’s Wager can be formulated in a number of ways, one way to understand it is by constructing a pay-off matrix exhibiting the expected utility of one’s choices relative to the truth of the belief that God exists:

  I. God exists II. God does not exist
i. I believe A. Infinite gain minus finite loss B. Finite loss
ii. I do not believe C. Finite gain minus infinite loss D. Finite gain

Pascal reasons that if I believe that God exists and it turns out that He does, then I have gained heaven at the small sacrifice of foregoing the pleasures of sin for a season. If I believe and it turns out that God does not exist, then I gain nothing and have suffered the finite loss of the pleasures of sin I have foregone. On the other hand, if I do not believe and it turns out that God does, in fact, exist, then I have gained the pleasures of sin for a season at the expense of losing eternal life. If I do not believe and it turns out that there is no God, then I have the finite gain of the pleasures afforded by my libertine lifestyle.

Now according to a principle of decision theory called the Expected Utility Principle, in order to maximize the utility or benefit of my choices, I should multiply each of the mutually exclusive outcomes by the probability of each of the two states' obtaining and add these products together. Make that choice having the highest expected utility. In Pascal’s Wager the odds of states (I) and (II) are assumed to be even (the evidence for and against God’s existence is of exactly the same weight). So, letting W0 stand for infinity and n for any natural number, we may calculate the utility of choices (i) and (ii) as follows:

i. (A x 0.5) + (B x 0.5) = (W0 x 0.5) + (~n x 0.5) = W0

ii. (C x 0.5) + (D x 0.5) = (~W0 x 0.5) + (n x 0.5) = ~W0

In other words, choice (i) has infinite gain and choice (ii) infinite loss. Thus, it is clear that belief has greater expected utility than unbelief. Therefore, even in the absence of preponderant evidence for theism, we should believe in the existence of God.

Two principal objections have been raised against the Wager argument. First, in standard decision theory infinite utilities cannot be handled. In particular, since division of infinite quantities is prohibited in transfinite arithmetic, it makes no sense to speak of W0 x 0.5. But this problem is easily resolved: simply substitute for W0 any arbitrarily high finite quantity, and it will still swamp the lower quantity n representing our finite loss or gain.

Rather the truly serious objection to Pascal’s Wager is the so-called “Many gods” objection. A Muslim could set up a similar pay-off matrix for belief in Allah. A Mormon could do the same thing for his god. In other words, state (II) God does not exist is actually an indefinitely complex disjunction of various deities who might exist if the Christian God does not. Thus, the choice is not so simple, for if I believe that the Christian God exists and it turns out that Allah exists instead, then I shall suffer infinite loss in hell for my sin of associating something (Christ) with God.

There are two possible responses to this objection. First, in a decision-theoretic context we are justified in ignoring states which have remotely small probability of obtaining. Thus, I need not concern myself with the possibility that, say, Zeus or Odin might exist. Second, we could try to limit the live options to the two at hand or to a tractable number of alternatives. This may have been Pascal’s own strategy. The Wager is a fragment of a larger, unfinished Apology for Christian theism cut short by Pascal’s untimely death. As we look at other fragments of this work, we find that although Pascal disdained philosophical arguments for God’s existence, he embraced enthusiastically Christian evidences, such as the evidence for Christ’s resurrection. It may be that he thought on the basis of such evidence that the live options could be narrowed down to Christian theism or naturalism. If the alternatives can be narrowed down in this way, then Pascal’s Wager goes through successfully.

A good example of a truth-independent, pragmatic argument for theistic belief may be found in William James’ classic essay “The Will to Believe.” Written in response to W.K. Clifford’s ringing pronouncement that it is wrong, always, everywhere, and for anyone to believe anything on insufficient evidence, James wants to show that we are sometimes pragmatically justified in willing to believe something in the absence of evidence of its truth. In cases in which we have no preponderant evidence in favor of a belief, we may resort to pragmatic considerations, James insists, if and only if the belief is for us a genuine option, that is to say, a choice which is living, momentous, and forced. A living choice is one which presents to me a belief to which I can give genuine assent. A choice is momentous if a great deal hangs on it, it presents to me a rare opportunity, and its consequences are irreversible. Finally, a choice is forced if there is no option of remaining indifferent, if to not choose to believe is, in effect, to choose to not believe. James held that religious belief meets these criteria. Moreover, he was convinced that religious belief is beneficial in this life, regardless of its promises for the next. His studies had convinced him that religious believers are more balanced, happier, and more virtuous people than unbelievers. Regardless of religion’s truth, then, religious belief is beneficial and, in view of such benefits, pragmatically justified.

Warrant without Evidence

The evidentialist might insist that while pragmatic arguments show that holding certain beliefs, including religious beliefs, is beneficial and therefore prudent, nevertheless that does not show that holding such beliefs is epistemically permissible, that one has not violated some epistemic duty in believing without evidence. One of the most significant developments in contemporary Religious Epistemology has been so-called Reformed Epistemology, spearheaded and developed by Alvin Plantinga, which directly assaults the evidentialist construal of rationality. Plantinga’s epistemology developed gradually over the course of three decades, but he has articulated it fully in a monumental three volume series, Warrant: The Current Debate (1993), Warrant and Proper Function (1993), and Warranted Christian Belief (2000). Here we can limn only the broad outlines of his theory.

Plantinga distinguishes between what he calls de facto and de jure objections to Christian belief. A de facto objection is one aimed at the truth of the Christian faith; it attempts to show that Christian truth claims are false. By contrast a de jure objection attempts to undermine Christian belief even if Christianity is, in fact, true. Plantinga identifies three versions of the de jure objection: that Christian belief is unjustified, that it is irrational, and that it is unwarranted. Plantinga’s aim is to show that all such de jure objections to Christian belief are unsuccessful, or, in other words, that Christian belief can be shown to be unjustified, irrational, or unwarranted only if it is shown that Christian beliefs are false. There is thus no de jure objection to Christian belief independent of a de facto objection.

Plantinga endeavors to show this by developing a model or theory of warranted Christian belief, that is to say, an account of how it is that we know the truth of various Christian truth claims. On behalf of his model Plantinga claims, not that it is true, but that (i) it is epistemically possible, that is to say, for all we know, it may be true, (ii) that if Christianity is true, there are no philosophical objections to the model, and (iii) if Christianity is true, then something like the model is very likely to be true. So Plantinga sets for himself two projects, one public and one Christian: (1) to show that there is no reason to think that Christian belief lacks justification, rationality, or warrant (apart from presupposing the falsehood of Christian belief) and (2) to provide from a Christian perspective an epistemological account of warranted Christian belief.

Consider, then, the de jure objection to religious belief, for instance, to the belief that God exists. According to the evidentialist, even if it is true that God exists, one is unjustified or irrational in believing that God exists unless one has evidence supporting that belief. For according to the evidentialist, one is rationally justified in believing a proposition to be true only if that proposition is either foundational to knowledge or is established by evidence that is ultimately based on such a foundation. According to this viewpoint, since the proposition God exists is not foundational, it would be irrational to believe this proposition apart from rational evidence for its truth. But, Plantinga asks, why cannot the proposition God exists be itself part of the foundation, so that no rational evidence is necessary? The evidentialist replies that only propositions that are properly basic can be part of the foundation of knowledge. What, then, are the criteria that determine whether or not a proposition is properly basic? Typically, the evidentialist asserts that only propositions that are self-evident or incorrigible are properly basic. For example, the proposition The sum of the squares of the sides of a right triangle is equal to the square of the hypotenuse is self-evidently true. Similarly, the proposition expressed by the sentence "I feel pain" is incorrigibly true, since even if I am only imagining my injury, it is still true that I feel pain. Since the proposition God exists is neither self-evident nor incorrigible, then according to the evidentialist it is not properly basic and therefore requires evidence if it is to be believed. To believe this proposition without evidence is therefore irrational.

Now Plantinga does not deny that self-evident and incorrigible propositions are properly basic, but he does demand, “How do we know that these are the only properly basic propositions or beliefs?” He presents two considerations to prove that such a restriction is untenable: (i) If only self-evident and incorrigible propositions are properly basic, then we are all irrational, since we commonly accept numerous beliefs that are not based on evidence and that are neither self-evident nor incorrigible. For example, take the belief that the world was not created five minutes ago with built-in memory traces, food in our stomachs from the breakfasts we never really ate, and other appearances of age. Surely it is rational to believe that the world has existed longer than five minutes, even though there is no way to prove this. The evidentialist’s criteria for properly basicality must be flawed. (ii) In fact, what about the status of those criteria? Is the proposition Only propositions that are self-evident or incorrigible are properly basic itself properly basic? Apparently not, for it is certainly not self-evident nor incorrigible. Therefore, if we are to believe this proposition, we must have evidence that it is true. But there is no such evidence. The proposition appears to be just an arbitrary definition—and not a very plausible one at that! Hence, the evidentialist cannot exclude the possibility that belief in God is also a properly basic belief.

And, in fact, Plantinga thinks that belief in God is properly basic, not only with respect to justification but also with respect to warrant. For Plantinga justification involves obedience to one's epistemic duties or possession of a sound noetic structure of beliefs, whereas warrant is that property which converts mere true belief into knowledge when possessed in sufficient degree. Plantinga thinks that the theist is not only within his epistemic rights in believing in God without evidence, but that he actually knows apart from evidence that God exists. In order to show that such a view is tenable, Plantinga introduces his epistemological model of religious belief. Quoting John Calvin’s teaching that “There is within the human mind, and indeed by natural instinct, an awareness of divinity .... a sense of divinity which can never be effaced is engraved upon men’s minds” (Institutes I.iii.1, 3), Plantinga proposes that “there is a kind of faculty or cognitive mechanism, what Calvin calls a sensus divinitatis or sense of divinity, which in a wide variety of circumstances produces in us beliefs about God.”[4] Plantinga also speaks of the sensus divinitatis as “a disposition or set of dispositions to form theistic beliefs in various circumstances or stimuli that trigger the working of this sense of divinity.”[5] Just as perceptual beliefs like “There is a tree” are not based on arguments from more basic beliefs but arise spontaneously in me when I am in the circumstances of a tree’s appearing to be there, so the belief “God exists” arises spontaneously in me when I am in appropriate circumstances, such as moments of guilt, gratitude, or awe at nature’s grandeur, as a result of working of the sensus divinitatis. Plantinga emphasizes that God’s existence is not inferred from such circumstances—such an argument would be manifestly inadequate--; rather the circumstances form the context in which the sensus divinitatis operates to produce a basic belief in God. Thus, belief in God is not arbitrary; it is grounded by the appropriate circumstances and so is properly basic. Hence, if such a model of theistic belief is true, the theist whose belief is produced in the described way violates no epistemic duty in believing and so is justified in believing that God exists.

But does he know that God exists? We often are justified in holding beliefs that turn out to be false (for example, the object I thought was a tree turns out to be a paper-maché simulation). Is our belief that God exists not merely justified, but warranted, and therefore knowledge? That all depends on what warrant is. In the first volume of his warrant trilogy, Plantinga surveys and criticizes all major theories of warrant which are offered by epistemologists today, such as deontologism, reliablism, coherentism, and so forth. Fundamentally, Plantinga’s method of exposing the inadequacy of such theories is to construct thought experiments or scenarios in which all the conditions for warrant stipulated by a theory are met and yet in which it is obvious that the person in question does not have knowledge of the proposition which he believes because his cognitive faculties are malfunctioning in forming the belief. This common failing suggests that rational warrant inherently involves the notion of the proper functioning of one’s cognitive faculties. But this raises the troublesome question, what does it mean for one’s cognitive faculties to be “functioning properly”? Here Plantinga drops a bomb into mainstream epistemology by proposing a peculiarly theistic account of rational warrant and proper functioning, namely, that one’s cognitive faculties are functioning properly only if they are functioning as God designed them to. He summarizes the conditions for warrant as follows:

This view can be stated to a first approximation as follows: S knows p if (1) the belief that p is produced in S by cognitive faculties that are functioning properly (working as they ought to work, suffering from no dysfunction), (2) the cognitive environment in which p is produced is appropriate for those faculties, (3) the purpose of the module of the epistemic faculties producing the belief in question is to produce true beliefs (alternatively, the module of the design plan governing the production of p is aimed at the production of true beliefs), and (4) the objective probability of a belief’s being true, given that it is produced under those conditions, is high.[6]

Although he adds various subtle qualifications, the basic idea of Plantinga’s account is that a belief is warranted for a person just in the case his cognitive faculties are, in forming that belief, functioning in an appropriate environment as God designed them to. The more firmly such a person holds the belief in question, the more warrant it has for him, and if he believes it firmly enough, it has sufficient warrant to constitute knowledge. With respect to the belief that God exists, Plantinga holds that God has so constituted us that we naturally form this belief under certain circumstances; since the belief is thus formed by properly functioning cognitive faculties in an appropriate environment, it is warranted for us, and, insofar as our faculties are not disrupted by the noetic effects of sin, we shall believe this proposition deeply and firmly, so that we can be said, in virtue of the great warrant accruing to this belief for us, to know that God exists.

So, Plantinga maintains, if his model is true, theistic belief is both justified and warranted. So is theistic belief warranted? That all depends on whether or not God exists. If He does not, then theistic belief is probably not warranted. If He does, then Plantinga thinks that it is. For if God exists, then He has created us in His image, He loves us, and He desires that we know and love Him.

And if that is so, the natural thing to think is that he created us in such a way that we would come to hold such true beliefs as that there is such a person as God .... And if that is so, then the natural thing to think is that the cognitive processes that do produce belief in God are aimed by their designer at producing that belief. But then the belief in question will be produced by cognitive faculties functioning properly according to a design plan successfully aimed at truth: it will therefore have warrant.[7]

The bottom line is that the question of whether belief in God is warranted is at root not epistemological, but metaphysical or theological. The question “can’t be settled just by attending to epistemological considerations; it is at bottom not merely an epistemological dispute, but an ontological or theological dispute.”[8] It follows that there is no de jure objection to theistic belief independent of the de facto question of whether theism is true.

But if there is no de jure objection to theistic belief, what about specifically Christian beliefs? How can one be justified and warranted in holding to Christian theism? In order to answer this question, Plantinga extends his model to include not just the sensus divinitatis but also the inner witness or instigation of the Holy Spirit.

The extended model postulates that our fall into sin has had disastrous cognitive and affective consequences. The sensus divinitatis has been damaged and deformed, its deliverances muted. Moreover, our affections have been skewed, so that we resist what deliverances of the sensus divinitatis remain, being self-centered rather than God-centered. God in His grace needed to find a way to inform us of the plan of salvation which He has made available, and He has chosen to do so by means of (i) the Scriptures, inspired by Him and laying out the great truths of the Gospel, (ii) the presence and action of the Holy Spirit in repairing the cognitive and affective damage of sin, thereby enabling us to grasp and believe the great truths of the Gospel, and (iii) faith, which is the principal work of the Holy Spirit produced in believers’ hearts. When a person is informed of the great truths of the Gospel, the Holy Spirit produces in him, if he is willing, assent to these truths. The internal instigation of the Holy Spirit is therefore “a source of belief, a cognitive process that produces in us belief in the main lines of the Christian story.”[9] In Plantinga’s view the internal instigation of the Holy Spirit is the close analogue of a cognitive faculty in that it, too, is a belief-forming “mechanism.” As such the beliefs formed by this process meet the conditions for warrant: (i) they are produced by cognitive processes functioning properly, (ii) the environment in which we find ourselves, including the cognitive contamination wrought by sin, is the cognitive environment in which this process was designed to function, (iii) the process is designed to produce true beliefs, and (iv) the beliefs produced by it, namely, the great truths of the Gospel, are in fact true, so that the process is successfully aimed at producing true beliefs.

Therefore, one can be said to know the great truths of the Gospel through the instigation of the Holy Spirit.

Because we know the great truths of the Gospel through the Holy Spirit’s work, we have no need of evidence for them. Rather they are properly basic for us, both with respect to justification and warrant. Plantinga therefore affirms that “according to the model, the central truths of the gospel are self-authenticating,”[10], that is to say, “They do not get their evidence or warrant by way of being believed on the evidential basis of other propositions.”[11]

Once again, then, Plantinga concludes if Christianity is true, then it probably has warrant in a way similar to the way described in the model. For if Christian belief is true, then we have fallen into sin and are in need of salvation.

Furthermore, the typical way of appropriating this restoral is by way of faith, which, of course involves belief ... in the great things of the gospel. If so, however, God would intend that we be able to be aware of these truths. And if that is so, the natural thing to think is that the cognitive processes that do indeed produce belief in the central elements of the Christian faith are aimed by their designer at producing that belief.[12]

A great deal more could be said about Plantinga’s religious epistemology, such as his claim that it constitutes a sort of theistic argument, since no naturalistic account of warrant and, in particular, proper functioning is forthcoming, or his claim that naturalism cannot be rationally affirmed, since the naturalist can have no confidence that his cognitive faculties produce true beliefs as opposed to beliefs merely conducive to survival; but what has been said provides a general picture of his epistemology.

What might be said by way of assessment of Plantinga’s religious epistemology? It will be recalled that Plantinga embarked on two projects, a public one and a private, Christian one. His public project was to show that there is no reason to think that Christian belief lacks justification, rationality, or warrant, apart from presupposing Christianity’s falsehood. The most common objection lodged against the success of this public project is that it leads to a radical relativism. If belief in God or Christianity can be properly basic, then just any belief, it is alleged, such as Linus’ belief in the Great Pumpkin, can also be properly basic. Now with respect to justification, this allegation is, in fact, true. We can imagine someone placed in circumstances in which belief in the Great Pumpkin would be justified for him in a properly basic way. For example, perhaps Linus’ parents assured him of the existence of the Great Pumpkin, just as some normally trustworthy parents assure their children of the reality of Santa Claus. Since beliefs grounded in testimony are on Plantinga’s analysis properly basic, it follows that Linus' belief in the Great Pumpkin is in such circumstances properly basic with respect to justification. But this admission is for Plantinga inconsequential. It does not imply that bizarre beliefs like belief in the Great Pumpkin are properly basic for normally situated adults. In order to be properly basic with respect to justification, a belief must be appropriately grounded in the circumstances, and for most people belief in the Great Pumpkin is not. More importantly, belief in the Great Pumpkin is not, in any case, implied by Plantinga’s epistemology to be properly basic with respect to warrant. Just because certain beliefs are properly basic in respect to warrant in no way implies that any arbitrarily selected belief is also warranted in this way. In Linus’ case the cognitive environment is not appropriate because he is being lied to, and therefore his belief is unwarranted. Thus, even if on Plantinga’s theory Linus is within his rational rights in believing in the Great Pumpkin, it does not follow that he knows that the Great Pumpkin exists.

But, it has been objected, if Christian epistemologists can legitimately claim that their beliefs are properly basic, then any community of epistemologists, for example, voodoo epistemologists, can also legitimately claim that their beliefs are properly basic, no matter how bizarre they might be. Plantinga calls this objection the “Son of Great Pumpkin” objection. Again, he freely admits that the allegation is correct with respect to justification. We can easily imagine circumstances under which voodoo epistemologists could legitimately claim that their beliefs in voodoo are justified in a basic way. The important question is whether they could legitimately claim that voodoo beliefs are properly basic with respect to warrant. The answer to that question, says Plantinga, will depend on what we mean by “legitimately.” If we mean merely “justifiably,” then once again Plantinga freely admits that they could; but he sees no relativistic consequences flowing from this admission. Being justified is just too easy a state to attain to be of much significance. Voodoo epistemologists under the influence of a native narcotic might well be within their rational rights in thinking that voodoo beliefs are properly basic with respect to warrant. But no relativistic conclusion follows from that. Do we mean by “legitimately,” then, “warrantedly"? If so, then nothing in Plantinga’s model implies that the voodoo epistemologists are warranted in their claim. On the contrary, insofar as voodoo beliefs are incompatible with the innate sensus divinitatis, Plantinga’s model implies that voodoo epistemologists cannot be warranted in claiming that their beliefs are properly basic with respect to warrant. Thus, Plantinga’s model does not lead to relativism.

Interestingly, Plantinga does concede that practitioners of other theistic religions could, like the Christian, argue with equal cogency that, say, a Muslim version of Plantinga’s model is epistemically possible, philosophically unobjectionable given the truth of Islam, and probably warranted in a way similar to that described in the model if Islam is true. But such a conclusion does not support relativism. It merely shows that there is no de jure objection to other theistic faiths independent of de facto objections to them. Perhaps even more significantly, it is not the case that such a concession holds for any belief set; in particular it does not hold, in Plantinga’s view, for naturalism. For if naturalism is true, then it is not likely that our belief-forming mechanisms are reliable, since they are not aimed at truth but are merely selected for survival. Thus, although adherents of other theistic religions could cogently argue on behalf of their religions what Plantinga argues on behalf of Christian belief, the same cannot be said for what, in the Western academic world, is the main alternative to Christian belief today.

What, then, of Plantinga's private project? How well has he fared in providing an epistemological account of Christian belief from a Christian perspective? Here reservations need to be expressed. The aim of this project is to show that if Christianity is true, then Plantinga’s extended epistemological model or something like it is very probably true. Oddly, Plantinga’s argument in support of this contention is surprisingly thin. All of the intricate machinery concerning proper functioning, cognitive environment, design plan, and so forth, along with the nuanced descriptions of the sensus divinitatis and internal instigation of the Holy Spirit play no role in this argument. In fact, all we get is about a paragraph or two, quoted above, that if God existed, then He would want us to know Him and so would provide a means to do so. Thus, if Christianity is true, it is likely to be warranted. But with this conclusion the Christian evidentialist would enthusiastically concur, adding, “Therefore, it is very likely that God has provided evidence of His existence which is such that all culpable persons can draw a warranted inference that God exists.” At most, then, Plantinga’s argument shows that

1. If (Christian) theistic belief is true, then it is warranted.

which is neutral with regard to an evidentialist or Plantingian model. Now Plantinga also asserts that

2. If (Christian) theistic belief is true, the model or something similar is correct.

is probably true; but he seems to provide no argument in support of this assertion. Reflecting on (1) and (2), one wonders if Plantinga intends to infer (2) from (1) with the help of the assumed premise.

3. If (Christian) theistic belief is warranted, the model or something similar is correct.

By Hypothetical Syllogism, (2) follows validly from (1) and (3). Now Plantinga argues at length against a proposition deceptively similar to (3), namely,

4. If (Christian) theistic belief is warranted, then belief in the model or something similar is warranted.

Plantinga is quite clear that our Christian theistic belief can be warranted, but that we might not be warranted in believing the model. Indeed, for most Christians (who have never read Plantinga), their Christian theistic belief is warranted, and yet they have no warrant for believing in Plantinga’s model, of which they have not even heard. But while Plantinga rejects (4), it seems that he should find (3) unobjectionable. Indeed, (3) would seem to be the crucial premise differentiating Plantinga’s view from those of evidentialists, authoritarians, and so on. But so far as we can tell, Plantinga provides no argument at all in support of (3).

In the absence of any philosophical argument for (3), the Christian, in assessing the worth of Plantinga’s model as an account of how believers are warranted in their Christian belief, will turn to Scripture and Christian experience in order to size up the model by its approximation to their deliverances. When we do so, however, the model would seem to be in need of important modifications.

Take, first, Plantinga’s postulation of a sensus divinitatis. It is worth noting that Plantinga seriously misinterprets Calvin on this score. When the French Reformer spoke of an innate sense of divinity, he meant an awareness of God, just as we speak of a sense of fear, or a sense of foreboding, or a sense of being watched. But Plantinga takes him to mean a cognitive faculty akin to our sense of sight, or sense of hearing, or sense of touch. Nothing in Calvin supports the idea that we have a special inborn cognitive mechanism that produces belief in God. Now, as Plantinga reminds us, the model is Plantinga’s, not Calvin’s. But when we turn to Scripture, neither do we find there any such suggestion of a special faculty of the soul which is designed to produce belief in God. In fact, we do not find in Scripture even any unambiguous support for Calvin’s weaker notion of an innate awareness of divinity (John 1:9 would be an exegetical stretch). What about the appeal to Christian experience? The difficulty here is that it is impossible to distinguish experientially between an inborn sensus divinitatis and the testimonium Spiritu sancti internum (inner witness of the Holy Spirit). The Scripture does teach that the Holy Spirit works in people’s hearts to bring conviction of Christian truth claims, both in the case of the unbeliever (John 16.7-11) and the believer (Romans 8.15-16; I John 2.20, 26-27; 3.24; 4.13; 5.6-10), and any awareness one might experience of God can be as plausibly ascribed to the Spirit’s work as to an inborn sense of divinity. Thus, Scriptural teaching inclines against the postulation of a sensus divinitatis, and Christian experience does not require it.

What shall we say about Plantinga’s doctrine of the internal instigation of the Holy Spirit? Certainly, Scripture teaches that there is such a witness. But nothing in Scripture supports Plantinga’s surprising claim that the witness of the Spirit is given only in response to human sin and fallenness. Given that sin quenches the work of the Holy Spirit, it would be surprising if Adam, had he not fallen into sin, would not have enjoyed the fullness and fellowship of the Holy Spirit. In fact, we have compelling grounds for rejecting Plantinga’s view in light of the life of our Lord, who, though sinless, was nonetheless led and inspired by the Holy Spirit in his ministry, just as the Old Testament judges and prophets had been in theirs.[13] Plantinga’s construal of the witness of the Holy Spirit results from his doctrine of the sensus divinitatis as a cognitive faculty functioning properly prior to the fall but then damaged by the noetic effects of the fall—a doctrine that finds no support in Scripture.[14]

Moreover, Plantinga’s understanding of the instigation of the Holy Spirit as a belief-forming process analogous to a cognitive faculty is surely suspicious. It is as though there were a faculty outside myself which forms beliefs in me. But since this faculty or process is not mine, not being part of my cognitive equipment, then it cannot literally be true that “I have believed in God,” which contradicts both Scripture and experience. Certainly, the belief is formed in me, but I am not the one who formed it, and, therefore, I have not truly believed. For this reason, it seems preferable to construe the testimonium Spiritu Sancti internum either literally as a form of testimony and thus its deliverances as properly basic or else as part of the circumstances which serve to ground belief in God and the great truths of the Gospel and thus again the beliefs formed in the context of the Spirit’s witness as properly basic. In either case, it is we, employing the soul’s ordinary, God-given capacities, who, in response to the Spirit’s testimony or in the circumstances of experiencing His prevenient convicting and drawing, come to believe in God and the great things of the Gospel.

Such a modified model seems better suited than Plantinga’s original model to serve Christians as an account of how Christian belief is warranted. Nonetheless, it is still so close to Plantinga’s approach that he seems correct that if Christian belief is true, his model or something very similar is likely to be correct.

Checklist of Basic Terms and Concepts

de facto objection
de jure objection
Expected Utility Principle
Falsification Principle
hiddenness of God
inner witness (or instigation) of the Holy Spirit
Logical Positivism
Pascal’s Wager
pragmatic argument
pragmatic justification
presumption of atheism
properly basic
sensus divinitatis
theological rationalism
truth-dependent argument
truth-independent argument
Verification Principle


[1] Antony Flew, R. M. Hare, and Basil Mitchell, “Theology and Falsification,” in New Essays in Philosophical Theology, ed. Antony Flew and Alasdair McIntyre (New York: Macmillan Co., 1955), p.96.

[2] A Companion to Philosophy of Religion, ed. Philip Quinn and Charles Taliaferro (Oxford: Blackwell, 1997), s.v. “The Presumption of Atheism,” by Antony Flew.

[3] Michael Scriven, Primary Philosophy (New York: McGraw-Hill, 1966), p.99.

[4] Alvin Plantinga, Warranted Christian Belief (Oxford: Oxford University Press, 2000), p.172.

[5] Ibid., p.173.

[6] Alvin Plantinga, "A Defense of Religious Exclusivism," in Philosophy of Religion, 3d ed., ed. Louis Pojman (Belmont, Cal.: Wadsworth, 1998), p.529.

[7] Plantinga, Warranted Christian Belief, pp.188-89.

[8] Ibid., p.190.

[9] Ibid., p.206.

[10] Ibid., p.261.

[11] Ibid., p.262.

[12] Ibid., p.285.

[13] See the extended treatment by James D. G. Dunn, Jesus and the Spirit (London: SCM, 1975).

[14] Plantinga inconsistently portrays the sensus divinitatis as both operative today and part of the narrow image of God which was destroyed, not merely damaged, in the Fall (Plantinga, Warranted Christian Belief, pp.204-05).