Home » Library » Modern Library » Nonbelief vs. Lack of Evidence

Nonbelief vs. Lack of Evidence

Two Atheological Arguments (1998)

Theodore M. Drange

Abstract: Here are two atheological arguments, called the “Lack-of-evidence Argument” (LEA) and “the Argument from Nonbelief ” (ANB). LEA: Probably, if God were to exist then there would be good objective evidence for that. But there is no good objective evidence for God’s existence. Therefore, probably God does not exist. ANB: Probably, if God were to exist then there would not be many nonbelievers in the world. But there are many nonbelievers in the world. Therefore, probably God does not exist. Reasons are given for saying that although LEA is not totally implausible, ANB is a stronger atheological argument than it is.


The claim is sometimes made that there is no good argument or evidence for God’s existence. Probably most philosophers would agree with that. Even theists sometimes say such things as “God is hidden” or “the world is ambivalent or ambiguous (as between being governed by God or being totally natural).” Whether such a statement is made in terms of “hiddenness” or “ambivalence” or “ambiguity,” it runs counter to Saint Paul’s (general-revelation) idea, expressed in Rom. 1:20, that “God’s invisible qualities — his eternal power and divine nature — have been clearly seen, being understood from what has been made.” So if it is a statement made by Christian writers at all, they would not be Calvinists or evangelical Christians but rather Christians of a more liberal persuasion.

One example of such a writer is John Hick. After putting forward what he takes to be opposing considerations that are equal or balanced, Hick concludes as follows:

It seems, then, that the universe maintains its inscrutable ambiguity. In some aspects it invites whilst in others it repels a religious response. It permits both a religious and a naturalistic faith, but haunted in each case by a contrary possibility that can never be exorcised. Any realistic analysis of religious belief and experience, and any realistic defense of the rationality of religious conviction, must therefore start from this situation of systematic ambiguity.[1]

Although Hick in effect acknowledges God’s hiddenness, he never views that very fact as a grave problem for theism itself. Hick thinks that if God were to exist, then for certain reasons it would be only natural and expected for him to be hidden from humanity.

Perhaps the idea of divine hiddenness does not seem problematic to Hick because it appears in Scripture (Ps. 44:24: “Why do you hide your face and forget our misery and oppression?” and Isa. 45:15: “Truly you are a God who hides himself, 0 God and Savior of Israel”). These passages seem to conflict with the one from Romans, above. However, divine hiddenness has been recognized as posing a problem for theism by various writers down through the centuries, including Blaise Pascal and Sören Kierkegaard.[2] In recent times, one of the earlier writers to mention it is Ronald Hepburn, who wrote:

One might be tempted to see in that ambivalence a vindication of atheism. For how could such an ambiguous universe be the work of perfect love and perfect power? Could this be a way to love and express love, to leave the loved one in bewildering uncertainty over the very existence of the allegedly loving God? Would we not have here a refined weapon of psychological torture? That is: if the situation is ambivalent, it is not ambivalent; since its ambivalence is a conclusive argument against the existence of the Christian God.[3]

Another writer who has seen God’s hiddenness as a problem for theism is Terence Penelhum, who says:

Theologians should … welcome, not discourage, attempts to provide intellectual support for belief in God. If they conclude, in the face of the long record of failures, that they have to agree with those philosophers who dismiss such attempts as hopeless, they should judge this result to be what it is — a great pity, and a significant theological problem. For what reason could there be for unbelievers always to have reasonable grounds for their hesitations?[4]

In a later work, Penelhum pins the problem down a bit more, for he applies it to faiths which appeal to “a God who wishes us to respond to him.” He says the following:

If the world really is ambiguous, there are some faiths for which this is a problem. Put simply, how can it be that a God who wishes us to respond to him reveals his reality only in ways that it is possible conscientiously not to recognize? If one or more of the theistic faiths is true, why is it not unambiguously true? This problem may well have adequate answers within those traditions where it can arise, but it is not an inconsiderable one, especially if the advocate of such a tradition rejects appeals to evidence or any other form of natural theology.[5]

Two other writers who have wrestled with the “hiddenness problem,” ones who use the term “hiddenness” instead of “ambivalence” or “ambiguity,” are Thomas Morris[6] and Robert McKim. McKim says the following:

Part of the cost of divine hiddenness is uncertainty and profound disagreement about the existence and nature of God. Another important part … is this. God, if God exists, is worthy of adoration and worship. A worshipful and respectful attitude is appropriate towards a morally good and wise creator… But if many creatures are in the dark about the existence and nature of the creator, then the appropriate human responses are, at least, made more difficult than they otherwise might be. So another part of the cost of divine hiddenness is its contribution to the large scale failure of human beings to respond to God in the right way. It inhibits the fulfillment of many duties. And there are further costs. The profound disagreements about God, and more broadly the profound disagreements between religious groups, sometimes play a role in promoting social conflict…. And the mystery surrounding God provides opportunities for charlatans and frauds to pose as experts on the nature and activities of God, and for religious authorities to acquire and exercise power and control over others.[7]

Despite the recognition by these writers that there is a grave “hiddenness problem,” none of them declares the problem insoluble or employs it in an attempt to construct a sound (or strong) atheological argument (i.e., an argument for God’s nonexistence).


So far as I know, the first writer to approach the topic in the latter way, i.e., in terms of an argument for God’s nonexistence, is J. L. Schellenberg, who says the following:

Why, we may ask, would God be hidden from us? Surely a morally perfect being — good, just, loving — would show himself more clearly. Hence the weakness of our evidence for God is not a sign that God is hidden; it is a revelation that God does not exist.'[8]

When the matter is put in this way, then the problem should not be called “the problem of God’s hiddenness,” which would normally be taken to presuppose God’s existence, but “the problem of the lack of good evidence for God’s existence.” One way to bring out the problem is by formulating an argument based on it which aims to establish God’s nonexistence. Let us refer to such an argument as the Lack-of-evidence Argument, or LEA for short. It is the argument suggested by the quotation above from Schellenberg, and may be formulated as follows:


(A) Probably, if God were to exist, then there would be good objective evidence for that.

(B) But there is no good objective evidence for God’s existence.

(C) Therefore, probably God does not exist.

The basic datum appealed to here is not the fact of people’s nonbelief in God, but, as Schellenberg puts it, the alleged fact of “the weakness of our evidence for God.” It should be pointed out that the argument here formulated is not the one defended at great length by Schellenberg in his book. That argument is closer to what I call the Argument from Nonbelief, but is different from it in certain important respects. I shall not be discussing Schellenberg’s own atheological argument in the present essay.

I certainly agree that LEA’s premise (B) expresses a fact, but, unlike the mere fact of people’s nonbelief, it is not one which can be established by simply taking a survey. Furthermore, it is a debatable issue: many theists, probably the majority of them, claim to have strong evidence for God’s existence. Although some of them think of God as being hidden, most of them do not. So in that respect, LEA can be attacked in a way not applicable to the Argument from Nonbelief.

The Argument from Nonbelief, abbreviated as ANB, is my own invention. It is an atheological argument that differs both from LEA and from the Argument from Evil. One way to formulate it is as follows:


(A) Probably, if God were to exist, then there would not be many nonbelievers in the world.

(B) But there are many nonbelievers in the world.

(C) Therefore, probably God does not exist.

The term “nonbelievers” as it appears here can be taken in various ways. Let us take it to refer to nontheists. Since that class includes not only atheists and agnostics but also deists, pantheists, Buddhists, Hindus, and countless other individuals throughout our planet who do not believe in a single Supreme Being, it actually contains close to half the earth’s population. That in itself is enough to establish the truth of ANB’s premise (B), which brings out my point above. Whereas LEA’s premise (B) is highly controversial and is a place at which the argument might be attacked, ANB’s premise (B) is a well-established empirical truth, given a suitable definition for the term “nonbelievers.” But the main force of each argument lies in its premise (A) and so that should be the focus of our attention.

Two questions to be addressed are the following:

1. Which of the two arguments, LEA or ANB, is the stronger one, and why?

2. Does either of them establish God’s nonexistence beyond a reasonable doubt?

I begin with question 1.


It should first be noted that there is another version of LEA which has been discussed in the philosophical literature. It is the argument that the absence of good objective evidence for God’s existence itself makes it reasonable to deny that God exists. A certain methodological principle is assumed here: that the burden of proof is always on one who claims the existence of something and if that burden is not fulfilled then it is reasonable to claim that the thing in question does not exist. The main drawback to such a line of thought is that there is no good support for the methodological principle in question. It is not a principle observed in scientific research. For example, scientists do not deny the existence of, say, tachyons (faster-than-light particles) simply because no good evidence has been produced that they exist. And the same is true for other entities postulated in other hypotheses. Scientists do not reason to the nonexistence of the postulated entity merely from the current absence of positive evidence for its existence. The burden-of-proof principle is therefore not one employed in the sciences.

Perhaps there are contexts in which it would be reasonable to infer the nonexistence of something simply from an absence of evidence that the thing exists, but it is not a mode of reasoning that is generally valid. To reason that way in the case of God might be charged with being an evasion of honest toil (such toil consisting of a proof of God’s nonexistence). It also smacks of the informal fallacy known as Argument from Ignorance. That is the fallacy of arguing that something is so simply from an absence of evidence that it is not so, or that something is not so simply from an absence of evidence that it is so. Certainly things may exist even if there is currently no good evidence for their existence. In order for it to be reasonable to deny a thing’s existence, there needs to be some reason to think that if it were to exist, then by now we would have found good evidence of that fact. But that is the first premise of LEA, as formulated above. My point, then, is that the version of LEA that is being considered here is preferable to the older version mentioned above and naturally supersedes it. Once the methodological principle appealed to in the older version is abandoned, one needs a step such as LEA’s premise (A) in order to infer God’s nonexistence from a lack of evidence for his existence. Let us, therefore, take a close look at the soundness of LEA, as given above.

Despite the dissent from many theists, I shall here assume that LEA’s premise (B) is true. Whether or not there is good objective evidence for God’s existence is a huge philosophical problem. It is discussed in practically every introductory philosophy course. But in this essay I want to focus, instead, on LEA’s premise (A), so let us look more closely at that one.

Why believe that, probably, if God were to exist then there would be good objective evidence for his existence? The reasons can be divided into two sorts. First, there is the idea contained in ANB: if God were to exist, then he would want people to believe in him, and so would probably do things to bring that about. However, that in itself does not yield LEA’s premise (A), for there are other ways for God to bring about belief than by means of good objective evidence. One way, for example, would be by the direct implantation of belief in people’s minds. Another way, emphasized by Schellenberg, is for God to reveal himself to people by means of private religious experiences. In order for LEA’s premise (A) to be established, it would need to be shown that if God were to exist and were to cause people to believe in him, then he would probably do so by means of objective evidence rather than by any other method. But that would be very hard to show, which is certainly a weakness in the given line of reasoning. It seems, then, that to try to support LEA’s premise (A) by appeal to ANB is not at all promising.

The second sort of reason that might be given for premise (A) is that there are specific divine attributes, apart from any desire on the part of God to reveal himself to humanity, which imply the existence of the type of evidence in question. To put it metaphorically, God is the kind of being that would “leave clear tracks.” Well, what divine attributes might they be? Perhaps the attribute of “being the ruler of the universe” would be a possible candidate. It might be claimed that if A is a ruler over B, then B would probably have good objective evidence of that fact. Is that, indeed, a reasonable claim?

Well, consider human societies. No doubt most subjects have good evidence for the existence of their rulers. But there may be exceptions. Some societies exist in remote places covered largely by wilderness. For example, there are tribes in the Amazon rain forests whose members have not been visited by agents of the government which has jurisdiction over those regions. The same is true of parts of Africa, Asia, and Australia, as well as some large islands such as Borneo and New Guinea. Although those people are governed by others who dwell far off, they do not have any good evidence of that. It could be that way with the planet Earth. Maybe God has not as yet gotten around to exerting his control over us in a way that would present us with good evidence of such control. We are like the wilderness tribesmen who are unaware of what is really going on in the nations of which they are citizens. God has governmental control over us but we have not as yet acquired evidence of such control. Maybe such evidence will come to be supplied to us in some way in the future. The fact that we do not as yet have evidence of being ruled does not entail that we will never acquire such evidence or that we are not being ruled. I conclude that the divine attribute of being a ruler of the universe can not be appealed to as support for LEA’s premise (A).

As I run through the various attributes usually ascribed to God, I am not able to find any that entail or presuppose that if God were to exist then there would be good objective evidence of that fact. My conclusion is that, not only is there much controversy surrounding LEA’s premise (B), but there are also serious problems confronting its premise (A). Since those problems do not arise for ANB, this is excellent reason to prefer ANB to LEA. I would say that to view our lack of evidence for God’s existence as direct support for atheism, as in LEA, might possibly be an approach to the topic worth pursuing, but it is definitely a line of thought different from, and in my opinion inferior to, that contained in ANB.


With regard to ANB, the version of it formulated above may initially not seem plausible, for there is no very obvious reason why God, should he exist, would want people to believe in him. Schellenberg brings in the divine attribute of omnibenevolence and suggests that if God were “perfectly loving,” then he would necessarily do something to bring about such belief. He says the following:

A being who did not seek to relate himself to us explicitly in this life — who elected to remain elusive, distant, hidden, even in the absence of any culpable activity on our part — would not properly be viewed as perfectly loving.[9]

It is a controversial matter whether an elusive or hidden deity could still be “perfectly loving.” But I think that Schellenberg is right that if a loving God were to exist then he would not deprive us of the awareness of his existence. In the quotation given above, Ronald Hepburn suggests that the absence of good evidence of God’s existence has left us in a state of “bewildering uncertainty” and that such a situation appears to be “a refined weapon of psychological torture,” which obviously would not be employed or permitted by a perfectly loving deity. Robert McKim suggests that because of our lack of good evidence for God, “appropriate responses are made more difficult” and “it inhibits the fulfillment of many duties.” It would be counterproductive for God to permit that. McKim also mentions what he calls “further costs”: social conflict in the form of holy wars and religious persecution, and opportunities for charlatans and religious authorities to take advantage of people in their state of ignorance. God could do much to prevent such harm simply by revealing himself to humanity in a clear and unambiguous manner. That, then, is good reason to deny that an elusive or hidden deity might still love humanity greatly.

Another reason based on the assumption that God is omnibenevolent pertains to those religious doctrines that predict the damnation of all nonbelievers. It seems totally obvious that if God loves us and nonbelief would automatically get us damned, then God must want us to be believers rather than nonbelievers. I have asked Christians who accept such doctrines why God does not clearly reveal himself to nonbelievers and thereby eliminate their nonbelief, but have never gotten any intelligible answer to that question. There seems nothing clearer than that a God who loves us would provide us with a revelation of his existence if such revelation would keep us from being damned.

It should be noted here that the sort of divine revelation in question need not take the form of objective evidence. It could be brought about by private religious experiences provided by God to everyone on earth. If there is any defect in people here, it is a kind of ignorance, and the cure for such ignorance need not take the form of objective evidence. There are other ways by which God might make people aware of the truth, including even the possibility of direct implantation of the relevant beliefs. Thus, if God is perfectly loving, then he would want to eliminate humanity’s ignorance, but he need not accomplish that by providing good objective evidence. So, the considerations introduced in the above arguments would support only ANB’s premise (A); not LEA’s premise (A).

However, John Hick would reject such arguments. According to him, God could be perfectly loving and still want to preserve a certain “epistemic distance” from humanity. First of all, Hick does not believe that nonbelievers will be damned, so the second argument would not be relevant for him. And secondly, he wants to apply a kind of free-will defense to ANB, arguing that for God to reveal himself clearly to people would interfere with their free will. This is a totally bogus idea, and I have devoted a chapter of my book Nonbelief and Evil to a refutation of it.[10] People do not have their free will interfered with by becoming enlightened about facts, so Hick is clearly misguided on this point. We have good reason to maintain that if God is a being who loves humanity, then he would want people to be aware of his existence, in which case there would not be many nonbelievers in the world. ANB’s premise (A) is thus well supported.

Still further support for the premise appeals to the idea that God is a deity who desires something from humanity, which is an idea that is expressed in the Bible and widely accepted in Western religions. God is conceived of as wanting humanity to love him and to worship him. In the particular case of Christianity, it is said that God wants people to accept his son as their Lord and Savior. But all of this requires that people be believers, and so that is further reason to infer that God must want people to be believers rather than nonbelievers. Presumably if he were to exist, then he would have done something to prevent there from being as many nonbelievers in the world as there actually are, which establishes the truth of ANB’s premise (A). I take this further support to be the most powerful argument there is for that premise.

However, when God is understood in a general way, quite apart from any particular religion, there is no reason to assume that he has to be “perfectly loving” toward humanity or that he wants anything in particular from it. The idea that God loves humanity and wants that love to be reciprocated, which is an idea prominent in Western religions, is crucial to ANB’s premise (A). If that idea is rejected, then it is hard to see how there could be any good support for the premise. I would say, then, that ANB does not clearly establish the nonexistence of God in general, taken in a totally unrestricted way. It can be used in debate against a Christian, Muslim, or Jew, but probably not against people of other religions. Whether there is any good argument for the nonexistence of God in general, taken in a totally unrestricted way, is an interesting question, but beyond the scope of the present essay.

Also beyond the scope of this essay is the question whether ANB, when applied in the context of Western religions, is a stronger argument than the Argument from Evil. In my book I argue at length that it is. There is excellent reason to think that the deity described and accepted by those religions would be more upset by the existence of many nonbelievers than by the occurrence of much earthly suffering.

My conclusion is that, although LEA resembles ANB and is not a hopelessly weak argument, it nevertheless has defects not possessed by ANB. Overall, ANB is the very best support there is for the nonexistence of the deity appealed to by the Western religions. More atheists should make use of it in their debates with theists.


I would say that LEA does not prove God’s nonexistence beyond a reasonable doubt. But what about ANB? Does ANB do so? My answer is that it depends on how “God” is defined and what assumptions are made about God. ANB has no application whatever to God in general, defined simply as the omnipotent and eternal creator of the universe. However, if God is taken to love people greatly and to strongly desire that they love and worship him in return, then that provides excellent support for ANB’s premise (A). In such a context, yes, ANB proves God’s nonexistence beyond a reasonable doubt. Furthermore, if it is assumed that all nonbelievers will be damned, and God is taken to be a perfectly loving being, then, again, ANB’s premise (A) must be true and ANB can be taken to prove God’s nonexistence beyond all reasonable doubt.

At this point, theists sometimes bring in what I call the Unknown-purpose Defense (or UPD). When applied to ANB, UPD is the idea that God exists and although he loves humanity and strongly desires a reciprocal response from humanity, yet, for some unknown reason, he permits many people to be unaware of his existence. Some advocates of UPD even go so far as to maintain that all the millions of nonbelievers in the world will be damned and yet God permits that to happen for some reason unknown to us. I deal with this idea in chapter 11 of my book. UPD only brings up a logical possibility. It does not bring up anything which it would be reasonable to believe. We can here apply the old principle “If it walks like a duck and quacks like a duck, then it’s a duck.” Although there is the logical possibility that the thing is not a duck, nevertheless, it is, beyond all reasonable doubt, a duck. Similarly with God. ANB provides excellent evidence that God, defined a certain way, does not exist. That may not prove God’s nonexistence beyond all possible doubt but it is enough to prove it beyond a reasonable doubt.

But now we come to a more difficult question. What if God is defined as loving humanity but not as desiring any particular response from people, and what if no assumption is made about the salvation or damnation of nonbelievers? How strong an argument would ANB be in such a context? Well, the first argument in support of premise (A), above, would still be relevant even if the last two would not apply. If God loves us, then he must want us to have awareness of his existence, considering the great benefits that such awareness would provide. Nevertheless, the support for ANB’s premise (A) would in that case be weaker and it would be unclear whether or not ANB could still be said to prove God’s nonexistence beyond all reasonable doubt. I am undecided on that matter and see point to both sides. I leave it to the reader to make a judgment there.


[1] John Hick, An Interpretation of Religion (New Haven, Corm.: Yale University Press, 1989), p. 124.

[2] For an excellent discussion of Pascal’s and Kierkegaard’s treatments of the hiddenness problem, see J. L. Schellenberg, Divine Hiddenness and Human Reason (Ithaca, N.Y: Cornell University Press, 1993), chap. 6.

[3] R. W. Hepburn, “From World to God,” Mind 72 (1963): 50; reprinted in Basil Mitchell, ed., The Philosophy of Religion (Oxford: Oxford University Press, 1971), p. 178.

[4] Terence Penelhum, God and Skepticism (Dordrecht, Netherlands: Reidel, 1983), p. 158.

[5] Terence Penelhum, Reason and Religious Faith (Boulder, Colo.: Westview Press, 1995), p. 132.

[6] Thomas V. Morris, “The Hidden God,” Philosophical Topics 16 (1988): 5-11; revised and reprinted as chapter 6 of Making Sense of It All (Grand Rapids, Mich.: Eerdmans, 1992), pp. 85-108.

[7] Robert McKim, “The Hiddenness of God,” Religious Studies 26 (1990): 142.

[8] J. L. Schellenberg, Divine Hiddenness and Human Reason (Ithaca, N.Y: Cornell University Press, 1993), p. 1.

[9] Schellenberg, Divine Hiddenness, pp. 28-29.

[10] Theodore M. Drange, Nonbelief and Evil: Two Arguments for the Nonexistence of God (Amherst, N.Y: Prometheus Books, 1998, forthcoming), chap. 5.

Theodore M. Drange is Professor of Philosophy at West Virginia University.

“Nonbelief vs. Lack of Evidence” was originally published in Philo and is copyright © 1998 by the Society of Humanist Philosophers. The electronic version is copyright © 1998 Internet Infidels, Inc. with the written permission of the Society of Humanist Philosophers.

all rights reserved