God, Atheism and Incompatibility: The Argument from Nonbelief (2001)

Philip Kuchar


The Argument from Nonbelief against God's existence (ANB) has been used by a number of writers, such as Theodore Drange [1] and J. L. Schellenberg [2], to show that the mere existence of nonbelievers or the presence of sufficient evidence for nonbelief in God's existence is incompatible with God given a certain description of him. Against some of Drange's criticisms I will defend a more or less Schellenbergian version of the argument. I consider the differences between the two formulations to be relatively minor--the thrust of both versions of the ANB is the same--and therefore in this presentation I won't get into a detailed comparison.[3]

The ANB seems to me very important and persuasive, and goes to the heart of doubt regarding God's existence. The ANB is essentially a form of self-awareness on the skeptic's part. Just by understanding what rational nonbelief is, the skeptic realizes that there should not be this sort of doubt caused by a lack of clear evidence as to whether God exists. God should want to eliminate the uncertainty, in this case to forestall unnecessary and unfair interference in the formation of a parent-child relationship. God, like any parent, would have to find his children's inculpable nonbelief in his existence outrageous. And yet the nonbelief exists, which leads to a stark incompatibility. One must go: either there is no rational nonbelief and the evidence for theism is unambiguous or God, properly defined, does not exist.

A Formulation of the ANB

SRICP = sane, rational, informed and otherwise capable people

(1) If some SRICP have rational nonbelief in God's existence, then it is not the case that all SRICP have an opportunity to love God.

(2) Some SRICP have rational nonbelief in God's existence.

(3) If God exists, God loves the SRICP, is averse to deceit, and is omnipotent and sovereign over the universe.

(4) If God loves the SRICP, is averse to deceit, and is omnipotent and sovereign over the universe, then all SRICP have an opportunity to love God.

(5) If God exists, then all SRICP have an opportunity to love God. [(3), (4) transitive]

(6) It is not the case that all SRICP have an opportunity to love God. [(1), (2) modus ponens]

(7) God does not exist. [(5), (6) modus tollens]

Elements of the ANB

The ANB's focus on the incompatibility between a certain deity and certain of his created beings, specifically the "SRICP"

The "SRICP" include intelligent extra-terrestrials, certainly the vast majority of humans and possibly intelligent terrestrial non-humans, such as dolphins and chimpanzees. The ANB focuses, however, on the second category since only in that group can we confirm the presence of rational nonbelief. "Informed" means an awareness specifically of representative portions of the evidence for and against God's existence, and "otherwise capable" means the ability specifically to love God. The SRICP are those individuals fully able to love God and ready to hold theistic belief; thus they are sentient, sane, emotional and rational beings, and informed of the evidence for theism. A mentally challenged person's nonbelief doesn't pose the same sort of challenge to theism, but were God to exist we would expect all the SRICP to lack any unfair obstacle to their love of God, such as rational atheism, just because they are otherwise so well equipped to be spiritual theists.

God's desire for the SRICP's acknowledgement and affection

This statement underlies (4). Loving concern for someone presupposes a desire to have the individual return the love. Love of an individual without any interest in this love's reciprocation would be considered defective and better described as a form of voyeurism. Assuming God's love for humankind is not just a matter of spying on us, he should want his affection reciprocated. Moreover, according to the major religions human happiness is defined in terms of a "right relationship" with or trust in God. Given God's loving concern for the SRICP and the theistic view that human happiness depends on a tapping into the source of Life, God, he too would be aware of this condition and therefore would desire us to love him in return for our own sake if not for his. Needless to say, the threat of hell would only add dramatically to God's concern for our welfare. God's desire for the SRICP's reciprocation follows as well from the common theistic analogy of God as a parent figure. If God is our parent, it follows that he should desire our loving appreciation of his parenting efforts.

In his article, "Nonbelief as Support for Atheism," Drange objects to the claim that God would necessarily want his children to love him, on the grounds that there are cases of human parents who love their children from a distance, without wanting their children's reciprocation. This form of love, however, is defective. God would lack the sorts of defects that might compel a parent not to want to meet her child or form a relationship with her. God's "moodiness" (one of Drange's counter-examples) would conflict with God's sovereignty. A moody person lacks control over herself let alone the universe. Likewise, God's fear of repercussions from the forming of a relationship with his children would conflict with his sovereignty, since a sovereign deity would be in full control of his own plan.

Drange's other concern is that we might be incapable of producing valuable belief in God's existence because of the deity's transcendence. God's transcendence could make both our direct perception of him dangerous or impossible, and indeed, as Drange says, our theism so confused as to be irrelevant to God. The problem with Drange's point is that it virtually begs the question against the theist by postulating an unintelligible deity. A God so transcendent that we could not possibly believe in his existence wouldn't be analogously personal and we couldn't speak intelligibly about him. Anyone who might believe in such a "God" would be virtually indistinguishable from an atheist. Most if not all theists believe God is personal and at least partly intelligible such that our understanding of God wouldn't be automatically misguided or irrelevant to him. Drange thinks it would still make sense to call a wholly transcendent deity "loving," but that doesn't follow.

I note that there are other ways of generating the contradiction between what God wants of his children and what they can supply at the optimum level, other than by stipulating God's desire for a relationship. God might want his children to obey certain laws, and yet rational doubt may exist regarding these laws' authority. God would then have to remove this doubt by making plain the justification of these laws. Were rational doubt then actually to obtain, God's non-existence would follow. The ANB here formulated follows only one track for revealing the incompatibility.

The generation by God's love, honesty and sovereignty of the SRICP's opportunity to love God

This point is found in (4). Given the point above that God's love would involve an interest in the SRICP's reciprocation, God's love of the SRICP would provide a strong motivation for him to produce and allow for their opportunity to reciprocate.

Does God love everyone or just the elect? Calvinist answers create more problems than they solve. Double predestination means that the condemned souls were created not out of love but hatred, something fit more for a creative demon than a benevolent parent. Why would a loving, fatherly God create a number of children without ever intending to love them, or make their inevitable destination a state of everlasting misery? Why wouldn't such a God create only the elect, a number of children whom he loves? What is the meaning of "punishment" in this context, given that the condemned never have any hope of pleasing God, and fail by design? Double predestination means that God treats the condemned persons as means rather than ends, perhaps to provide the elects' satisfaction or some venting of divine frustration. Moreover, there is no apparent distinction between the behaviour of the "elect" and that of the "condemned group," no empirical grounds for drawing such a sweeping dichotomy, no differences between any two sets of people plain and extreme enough to warrant the view that some are created out of love to enter heaven, and the rest out of inexplicable anger to enter hell. Given these sorts of problems I assume in the ANB that God's love for his children is universal not selective.

On the other hand, an acceptance of double predestination refutes the ANB by denying (4) and (5). Were God to lack love of those SRICP who possess rational nonbelief in his existence, we could no longer claim in the above way that God's existence entails an opportunity for all the SRICP to love God. God might give some people overwhelming evidence for his existence and other people none at all, and this could be made consistent with God's character were the latter properly defined. Such a definition of God, however, involving the predication of this sort of selective "love" would make God much less intelligible and analogous to a human parent, and the theist who were to accept double predestination would be virtually indistinguishable from an atheist, as mentioned above.

Drange would object to (3) and claim that God need not be thought of as perfectly loving. (See Objection B in his above article.) Some theists conceive of God as occasionally violent and vindictive. Yet these theists usually claim that God permits violence and genocide not that he causes them out of a failure of love. Such theists would not take evidence of suffering and destruction to negate (3), the claim that if God exists he loves the SRICP. Were someone to believe that the creator of the universe does indeed cause suffering and death out of a failure or lack of love and an abundance of hatred, she would not obviously be a theist or a believer in "God," since the "creator" in her case would be closer to a demon. Is there a theist who uses the term "God" to refer to an evil supreme being, a devil? If so, such a theist could object to (3). However, such a theist would likely not live long enough to make the objection since suicide would be her most logical course. This is one of the stubborn objections to Schopenhaur's system in which he claims that the "Will" is evil not good. Even were God benevolent most of the time with only occasional outbreaks of moral neutrality or malevolence, this would still warrant the SRICP's doubt as to whether God would keep his promises or act benevolently at any given time. Such doubt would be justifiable even were the theist to do evil in expectation of reward from the malevolent deity, since such a deity could renege or cause suffering by a whim regardless of the SRICP's obedience. Such would be the devil-worshipper's peril.


Moving on to the second divine attribute given in (3), the importance of God's hatred of deception is that such an attribute would disallow negligence, manipulation, reneging on a promise, and dishonesty on God's part, which might otherwise lead to the SRICP's obstacle. If God were to tolerate deception he might mischievously conceal facts from the SRICP necessary to their theism's rationality. Is it conceivable that God should lie? In the ANB the lie under consideration is whether God would intentionally deprive us of clarity regarding the evidence of his existence, whether by supplying misleading, ambiguous information or no favourable evidence at all. In both cases God would be guilty of a lie of omission, which differs from an unintentionally deficient presentation of evidence. A limited, fallible being might easily intend fully to disclose her case and yet overlook key pieces of evidence or misstate facts. With God, however, given his omnipotence (the third personal attribute in (3)), this sort of failure would be impossible.

The other stipulated attribute is sovereignty, God's unchallenged, perfect control over his Creation, and power to carry out any logically possible wish. This includes the absence of any genuine rivals to God or significant unintended consequences of his plans. In this case, God's sovereignty explains how he could ensure that there are no obstacles such as rational nonbelief, by way of carrying through with his benevolent motivation and interests in honesty and full disclosure relative to the SRICP's needs.

God's obligation to eliminate a decisive obstacle imposed on the SRICP that precludes their love of him

This is stated in (5). One clear reason as to why God would have to remove the obstacle of rational atheism in particular is because of the necessary connection between affection for someone and belief in the person's existence. This claim underlies (1), that nonbelief would preclude the SRICP's opportunity to love God. Affection for someone is logically impossible without belief in the person's existence. "Sarah believes that John exists" is included in the meaning of "Sarah has affection for John." This is why the inculpable lack of theistic belief would be a decisive and overwhelming obstacle to a person's love of God.

Should God have only an interest in removing rational nonbelief or an "obligation" to do so? Ordinarily God might owe his children nothing, but given the penalty of eternal punishment depending on our response to God, as reported in a number of major religions, being the sentencer God should have a duty to give the defendants a fair opportunity to produce the appropriate response to him.

The only way for God to remove rational nonbelief, the decisive obstacle to the SRICP's love of God, would be for him to improve the evidence so that the SRICP, using all their critical and interpretive powers, could not in good-faith reach a negative conclusion regarding God's existence. God could, of course, also change the SRICP's critical and interpretive powers, leaving the evidence as it is or even worsening it, but that would involve creating a new species and entail perhaps a degree of incompetence on God's part which would contradict God's sovereignty. The ANB stipulates, however, that God wants the "SRICP" to love him, a certain group of organisms as they are. By recreating their brains to find any piece of evidence highly favourable to theistic belief, God would no longer receive the favour of the SRICP with their previous character and ability. Thus God would have to work on the side of the evidence rather than the SRICP.

As to whether God would remove justified atheism, aside from the above considerations regarding his three stipulated attributes he should have another reason. Assuming that God's love is analogous to human love, it's reasonable to believe that God should be appalled by his children's rational doubt regarding his very existence. The following analogy dramatizes this consequence. Imagine that Johnny is Mary's son whom she loves. Johnny doesn't love his mother, though. In fact he behaves as though she doesn't exist. A parent might tolerate this were the behaviour a temporary stage of pretend doubt on Johnny's part. Johnny would then eventually grow tired of his stunt and reveal to his mother the admiration and respect he secretly felt all along. But what if Johnny weren't faking? What if Johnny were genuinely to believe his mother doesn't exist? To account for this bizarre scenario, imagine that Johnny exists in the same universe as everyone else, but that whenever his mother comes into his perceptual range or is mentioned by someone else, Johnny shifts unknown to anyone into a different dimension such that he has no experience of Mary. And worse, imagine that Mary appreciates the fact that Johnny genuinely believes that she doesn't exist. In Mary's own home Johnny will go about his daily routine ignorant of Mary's presence and all the ways she helps her son. Johnny simply accepts the dinner he finds presented at the table, but he doesn't thank his mother. For some infuriating reason--from her perspective--Johnny is barred through no fault of his own from loving his mother and even from seriously entertaining the idea that she exists in the first place. To make the situation worse still, imagine that Mary knows that were Johnny to continue to remain ignorant of his mother, he would find himself in never-ending agony.

I submit that this situation would be utterly intolerable to Mary. Who could remain sane when faced with such an outrage? How could Mary not desperately scream at her son every moment of every day, working as hard as possible to save him from disaster, not to mention fulfill her own natural desire for a family bond? Turning to God, he certainly knows that he himself exists! Many people also believe that he exists. And yet a great many other people have rational doubt that there is a personal Creator. Is this the best God can do in removing rational atheism? This is not a matter of vanity. It would not be vain of Mary to expect and to desire her son's acknowledgement that she exists, and to want to form a relationship with her own son. Vanity is a matter of excessive self-glorification. God's desire for our acknowledgment of him should be at the very least the recoiling from the nightmarish alternative described in the case of Mary and Johnny. The results of God's failure to reveal himself to us would include the stifling of his fatherly concern, and an absurd voyeuristic situation in which a significant number of God's children justifiably doubt that he even exists and yet God with fatherly interest in his children looks on, watching us eat, sleep, and often kill each other in his house. If God is at all like a human parent he should be forced by the threat of this outrage to make his existence clear to us, to come out from his hiding place. God's love for us combined with our justified skepticism regarding his existence would cause him intolerable agony, which would impede his sovereign control. His aversion to deception would remove a number of possible reasons for his lack of outrage at our inculpable ignorance and confusion.

(5) can be challenged, however, by pointing to a higher concern that might justify God's withholding of information from us. The theist has two options at this point. First, she can refrain from specifying the higher concern and appeal instead to mystery. Second, the theist can specify the possible higher divine concern, and explain how the negative outcome would outweigh or negate the gains were God by removing the possibility of rational atheism to grant everyone an equal opportunity to respond favourably to him. The mere possibility of an unspecified sufficient reason on God's part for the lack of clear and overwhelming evidence for theism is simply the weakest possible evidence for theism, and falls far short of the theist's burden of proof. The evidence remains ambiguous and nonbelief rational in spite of the appeal to mystery or some unspecified reason why nonbelief might exist in God's universe.

Regarding the second option, one possible higher concern might be to keep our freedom intact and thus our love of God genuine. According to a number of theists, overwhelming or unambiguous evidence of God's existence, whether directly or indirectly transmitted would deprive us of our freedom and specifically of our ability to choose to accept or reject God's offer of a relationship with him. Instead we would be forced to accept God, which is not what God wants.

There are several replies to this objection. First, "love of God" would have to be a special case, since the mere recognition of an individual's existence is not usually understood as trivializing someone's love of the individual. If "love of God" were a special case, however, the term would have a special meaning, which would have to be defined. Why should a person's certitude regarding God's existence disqualify her love of God, whereas a person's certitude regarding the existence of a human mate, parent, child or friend never has any bearing whatsoever on this love's value except to make the affection possible, like the presence of brain matter in the loving individual?

Second, the notion that love is freely chosen is suspect. Do children, for example, choose to love their parents? Or is their loving concern and appreciation forced upon them by the circumstances of growing up with kind, successful caretakers, and having a sense of thankfulness and obligation? In so far as love is an emotion, whom a person loves is at least not significantly a matter of choice. What may be a matter of choice is the effort a person makes to remain in certain circumstances for the love to develop. For example, a wife might choose to remain with her husband although she no longer loves him, just in case her feelings for him might return. Yet this is a gamble that by putting herself into a certain situation her feelings could reoccur, not a choice to cause those feelings. The difference is between preparing for something or adding a necessary precondition, and causing the thing itself. Remaining in the same physical area as the husband is a necessary precondition of her feelings' return, but whether her feelings will return is not a matter of her will power. The overriding cause of a strong emotion is not exactly known to modern science, but such an emotion is not simply willed into existence. On the contrary, many kinds of love are felt in spite of the person's will and even against his or her best judgment. What value would love have if its support were ultimately a mere matter of will power, and could be shut on or off like a light switch? Furthermore, there is some doubt as to whether, in the example given above, the wife's choice to remain with her husband itself is free. There would seem to be several other determining factors, such as her fond memories of her husband and the intensity of her past feelings for him, her husband's current behaviour, financial considerations, embarrassment by the prospect of divorce, and the welfare of her children.

What all of this means is that our feelings of appreciation and affection for God would not be based on our freewill but on a number of factors not chosen by us, such as the enormous difference in status between the Creator and the created being, God's mystical dimensions, and the insanity of choosing hell over heaven. The objection, therefore, that God would want to protect our freedom so that our love for him might be genuine is based on the false premise that freewill has much to do with affection, respect, and thankfulness. Were God overwhelmingly to reveal his facticity every sane person would very probably return God's love, which is exactly what God should want. We would not be free sanely or rationally to reject God at this point. This contradicts, of course, the reason why the Christian theist maintains that love is a matter of will power, which is that God's judgment allegedly depends on our decision to accept or reject him. If we do not choose to love or reject God, reward for or punishment of us becomes incoherent. On the other hand, given only ambiguous evidence of God's existence theistic belief would be more a matter of will power, the choice to have "faith." In any case, since the necessary preconditions of love (including the overwhelmingly justified belief in the beloved person's existence) are not simply willed into existence, God should have no interest in protecting our freedom relative to his desire for us to reciprocate his affection. Therefore, our freewill would not be God's higher concern when choosing how clearly to relate his existence to us.

Third, according to some theodicies there was in fact an individual who not only had indirect but direct confirmation of God's existence, and yet was able to reject God and be held responsible for this rebellion. Was Satan free in rejecting God or was he forced to do so, say by his evil nature or divine decree? If the latter how could he be fairly punished? Contradicting what was just said about irrelevance of freewill to the presence of positive feelings, we might ask whether Satan exercised his freewill in rejecting God, switching off his affection. If so, the myth seems hardly intelligible. Even if freewill were somehow relevant to affectionate obedience, the myth of Satan's rebellion allows for both the possession of overwhelmingly justified theism and the culpable rejection of God. Satan both knew that God exists and yet rejected him. Therefore we should be able to do likewise, and so indubitable evidence of God's existence would be irrelevant to our ability to choose to accept or reject God.

Fourth, there is a difference between God (a) revealing the fact of his existence and (b) making a personal appearance. Conceivably God could make his existence overwhelmingly clear to everyone without physically appearing before anyone. After all, even a personal appearance doesn't amount to indubitable evidence of the person's existence. Perhaps the apparent individual is an automaton, a twin, a clone, a person with a sophisticated mask, or an hallucination. How high should the standard be for clarity regarding someone's existence? One definition might be: an unambiguous sign the genuine denial of which would be strong evidence of the doubter's insanity. A personal appearance could easily make a person's existence obvious in this sense, although under some circumstances reasonable doubt might be warranted. However, the more powerful and metaphysically awesome an individual, the greater the consequences of her personal manifestation for the observer. Were God to appear personally before the SRICP, they might explode, turn into goo, become deranged, or at least lose their freedom to reject God in any way because of this transcendent experience. Alternatively God might not even be able to register to our senses, since God wouldn't be physical. God could enhance our senses, but any enhancement sufficient to provide adequate conditions for perception of God would likely involve the organism's total transformation, in which case we would no longer be considering God's appearance to us but rather to a new kind of entity altogether.

However, the increase in danger from the more terrifyingly mysterious entity's personal manifestation is matched by this entity's increase in options for indirect revelation. Direct sensation of God would be potentially more debilitating for the recipient, but God should have available vast possibilities for indirect yet equally decisive pieces of evidence of his existence. Even the range of our imagination could not encompass God's options in this area. God could devise a sound deductive argument guaranteeing his existence. Alternatively God could provide circumstantial evidence pointing strongly to his existence, such as unmistakable traces of design in the universe. Yet there is no sound deductive proof of God's existence. The ontological argument, for example, is fallacious. And although there is evidence in nature that can be assigned some circumstantial weight in favour of God's existence, such as finely tuned physical laws to allow for life and the existence of moral laws, none of this is unambiguous or beyond sane or rational doubt. There are alternative explanations, such as natural selection, that account for these facts in atheistic terms. God's existence need not be invoked by all SRICP to explain any single fact about the universe as we now understand it.

Fifth, there is a difference between God revealing his existence and clarifying his praiseworthy character. Yet only the latter would seem likely to reduce our freedom to reject God. The ANB claims that we lack clear evidence even of God's existence. Were God to eliminate rational doubt that the universe is the product of a personal Creator, he would still be able to maximize our freedom to respond positively or negatively to him by leaving the evidence of his character ambiguous. That way at least the initial overwhelming obstacle to universal love of God would be removed without endangering the value of our response to him.

Sixth, there is reason to doubt the value or even the possibility of "love" of God given only ambiguous evidence of God's existence. Take the analogy of a child who, for some reason, is uncertain as to whether her parents ever existed. Perhaps, she believes, she was spontaneously created without any human process of procreation. She does have, however, some ambiguous clues to the existence of her human parents, but never enough to warrant certainty or even necessarily a strong belief in that direction. Could this child "love" her parents after deciding on the basis of the ambiguous evidence that her parents probably exist? Assuming the child's sanity and rationality, the reason for the evidence's ambiguity would have to be the parents' absence and lack of a recognized impact on the child, both of which are required for the formation of the child's emotional bond with her parents. If the parents were so hidden that the child would have only unclear evidence as to their existence let alone their character, what could be the basis of the child's love for the parents? If God provides his children with only unclear indirect evidence of his existence, how could he expect us genuinely to love him or confidently to accept the authority of his laws?

The difference is between the value of love based on "faith" and that of love based on the preponderance of the evidence. Ordinarily people respond positively to each other based on an appreciation of the individual's praiseworthy qualities and of course certainty regarding the person's existence. But let's take a more virtuous form of love: affection for someone in spite of the person's lack of attractive qualities and the presence of severe personal flaws. Imagine a saintly individual who learns to overlook someone's faults despite possessing overwhelming evidence of them. That seems to me a much more meaningful test of someone's capacity to love than to ask for someone's affection for an individual given only ambiguous evidence of the individual's existence. In the latter case the imagination rather than the loving person's strength of character and sympathy would likely have to compensate for the lack of evidence or attractive qualities, and supply sufficient content to solidify an emotional attachment. This sort of "love" based on hope and circular certainty would seem to be freely available precisely because of its insubstantiality and lack of genuine emotional depth, whereas ordinarily love of an individual is constrained by the individual's evident attractive qualities or the strength of the lover's character.

(See Hebrews 11:1, "faith is being sure of what we hope for and certain of what we do not see." Sometimes this is taken as saying that faith itself is evidence of an invisible world, since otherwise there would be no innate need to believe in spite of the evidence's insufficiency. This "need to believe" argument presupposes that we can trust in the justification of our need to trust, most likely because God himself created in us that need. This is a circular account of faith.)

Another possible higher concern that might account for the evidence's ambiguity would be for the SRICP to develop their character by dealing precisely with ambiguous evidence for theism. In so far as this claim appeals to the importance of freedom in this character development, the claim may be subject to the above objections. This claim also appears to be an updated version of Pascal's Wager and is thus subject to a host of other objections. God intentionally gives us unclear evidence of his existence so that we can gamble on whether he exists and act appropriately without the crutch of overwhelming evidence for theism. Our souls are forged by the conflict between our doubts and our faith, and the more faith we exercise the more mature we are deemed by God. I won't rehash all the problems with Pascal's Wager, but I will point out that this wager's justification of theism is itself entirely a matter of faith. This form of the Wager stipulates that the evidence for theism is unclear. Therefore we cannot know if we are meant to increase or decrease our faith. Perhaps God will respect us more if we favour our doubts and err on the side of caution by embracing atheism. Perhaps the atheist rather than the theist passes the maturity test.

Rational nonbelief as a negative judgment necessitated by an optimum evaluation of the evidence

This is the definition assumed in (1) and (2). Schellenberg makes use of the distinction between "inculpable" and "culpable" nonbelief. "Inculpable" is a term that becomes relevant to the ANB given the assumption that hell awaits the nonbeliever because of a divine judgment of the individual. Inculpable nonbelief is something that could not deserve punishment. One example of inculpable belief is what I've called "rational" atheism, which I take to be (a) an authentic rejection of some proposition, and (b) a forced reaction to the evidence relative to the person's critical powers when optimally employed. A violation of (a) would be a belief that a person only superficially possesses in which case the belief is caused by self-deception, a partial, lazy use of her reasoning abilities, or some other factor such as drug use, depression or preoccupation. The second condition, (b), would be contradicted by a frivolous, imaginative belief that doesn't follow from a fair examination of both sides of the issue. Rational nonbelief is inculpable in that the person cannot help but possess the belief. The rational nonbeliever cannot force herself to believe otherwise than as dictated by her understanding of the evidence. Punishment of the rational atheist would be as unfair as punishment of someone for rationally rejecting belief in the existence of unicorns.

What is an example of culpable belief? Some theists provide an answer with an alternative account of atheistic belief. According to them atheism begins as a culpable, free choice to reject the evidence for God's existence, and becomes sinful as the person's pride interferes with God's attempt to convince her of atheism's falsity. This scenario denies that there is rational nonbelief in God's existence. The nonbeliever is culpable because at some point she can indeed choose to believe or not to believe, and criminally chooses the latter. What evidence is offered in support of this view of atheism? After all, at least most nonbelievers claim the opposite, that their rejection of theism is based on a consideration of the evidence not personal hatred of God despite their knowledge that God exists. This is an overt presuppositionalist attack against the nonbeliever despite the nonbeliever's own testimony as to the cause of her atheism, an attack that reveals the theist's double standard. Whereas the theist often asks others to trust her personal testimony as to her private religious experiences and the testimony of, say, the New Testament authors, she doesn't extend the same courtesy to her opponent. Were the theist's scenario as to the formation of nonbelief accurate we might expect empirical evidence in the scenario's favour. Is there any empirical data suggesting an intensification of stubbornness that overshadows any examination of the evidence on the part of long-time nonbelievers? I'm not aware of any such data.

The person who is free to pick theism or atheism is agnostic not an atheist. But according to the agnostic, there is no non-arbitrary, meaningful choice between theism and atheism since both equally lack sufficient evidence to warrant belief. Certainly the nonbeliever claims to be compelled by her examination of the evidence to lack belief in God's existence. Freewill has as much to do with this nonbelief as it does with the quadriplegic's belief that she lacks all her limbs. In both cases the facts assert themselves and compel a certain belief relative to the person's capacity for evaluation. This is not to say that emotions such as stubbornness or pride don't enter into the evaluation at all, but there is no evidence that the nonbeliever's initial rejection of theism is entirely willful, as though she could choose to embrace theism despite her doubts as to something as fundamental to theism as God's existence. And although the long-term nonbeliever may naturally gain increased comfort with her view, there is no evidence that her nonbelief is ever based primarily on pride or hatred of God or that she would be incapable of accepting theism given sufficient evidence of God's existence. Nonbelievers claim that far lesser evidence than a personal appearance by God would convince them of theism's truth, such as the plain, uncontroversial fulfillment of a religious prophecy, the witnessing of a miracle, or the unique, inspired and enlightened behaviour of believers throughout history. Some sort of hard evidence should be offered to counter the nonbeliever's testimony, as opposed to an unfalsifiable presupposition of atheism's irrationality and purely emotional, sinful basis.

In rejecting the theistic account of sinful nonbelief I affirm that nonbelief is inculpable and rational. The distinction between inculpable and culpable nonbelief is, at least, clarified by the different accounts of atheistic belief offered by the nonbeliever and the believer. But is the distinction valid? The notion of culpable nonbelief seems coherent but factually groundless in the atheist's case. We sometimes punish people for their beliefs rather than their actions. The Church used to punish people for "heresy," certain governments still punish their citizens for holding wrong beliefs, and criminals are sometimes given a worse sentence if their crime is determined to depend on "hatred." Here the belief itself is considered culpable because the holding of the belief is thought to amount to the carrying of a loaded weapon, something that tends to cause violence to others. Regardless of whether the notion of a thought crime is justified in any particular case, the notion itself seems coherent.

A more accurate example of culpable belief than atheism would be a racist individual who chooses to consider the Jews an essentially inferior "race" despite the lack of evidence. The belief is genuinely held but on the basis of trust in a leader or some sort of rationalization the individual chooses to maintain, as opposed to a rational examination of the evidence. Racist belief may sometimes be based on an understanding of the evidence, but the free and in this case offensive and likely dangerous choice to accept a certain belief as a means to political power or the venting of rage is at least conceivable. The difficulty is in finding evidence to support this uncharitable account of someone's belief, as opposed to presupposing the attack's justification. In the case of racism, the evidence is overwhelmingly against the notion of "essential superiority of race," whereas in the case of theism the evidence is at best unclear and the evidential support for the view that atheism is culpable is correspondingly weak. Culpable belief cannot be based on a fair understanding of the evidence, and therefore if the evidence is complicated enough to warrant a certain interpretation, that interpretation is inculpable. Culpable belief is necessarily non-rational, criminal (at least according to one legal standard), and based on faith, the imagination, or some mental disturbance that doesn't yet amount to a legal defense.

Moving to a related issue, Drange considers the qualification regarding "rational" nonbelief irrelevant. All nonbelief should be repellent to God, not only the justified kind. Even if the theist could show that all nonbelief is culpable or sinful, the ANB would still function since God should want to alleviate even willful, misguided or evil nonbelief. While God's unconditionally merciful response to our sin is conceivable, this is a weaker, more strained form of the ANB. God seems more obviously duty-bound to extinguish reasonable doubt regarding his existence than correcting the obstacles that we ourselves erect between God and us through sin. On the other hand, I have no objection to forming an Argument from Culpable Nonbelief. However, the core of the ANB should certainly not be placed in God's duty to eliminate nonbelief in general, since that would entail that God should force inanimate objects to carry theistic belief. The heart of the ANB is not in the mere presence of nonbelief, but in the presence of a certain kind of nonbelief, one which God is duty-bound to rectify, whether this be inculpable or culpable nonbelief. Does God have an equal duty to eliminate both culpable and inculpable nonbelief? Perhaps, but as stated above Schellenberg's formulation trades broadness of scope for intuitiveness. A perfectly loving parent would indeed want to help even her misbehaving child, yet such a parent would surely want to help much more those of her children who fail to succeed in some way through no fault of their own, who fail, say, because of the education system's defects. Here we could easily speak of the parent's obligation and not just her desire to rescue her innocent children from ruin.

There is a related problem. Is the ANB when formulated with an emphasis on rational nonbelief a separate demonstration of God's non-existence or does the ANB beg the question by assuming that there is in fact nonbelief and that therefore of course God does not exist. I assert, in fact, in (2) that rational nonbelief does in fact exist, and indeed Schellenberg's type of formulation would be impossible without this declaration. But if rational nonbelief exists, is that not itself a sufficient demonstration of God's nonexistence? Given (2), couldn't we formulate a brief disproof such as, "There is rational nonbelief in God's existence. Therefore God does not exist," making the ANB redundant?

The key question, though, concerns the level of nonbelief's rationality claimed in the ANB to exist. Were this rational nonbelief itself overwhelming, there really would be no need for an ANB. Instead nonbelievers could simply appeal to the arguments that cause such a high level of nonbelief, and take those arguments as sufficient proof of atheism. The ANB as Schellenberg formulates it, however, does not assume that such a high level of rational nonbelief exists, and neither does my formulation. I claim only that the evidence of God's existence is ambiguous enough to allow for inculpable, "rational" nonbelief, not that the evidence is unambiguously in favour of atheism such that theism is always insane or irrational. This nonbelief does indeed have its own force against theism, being supported by various atheological arguments. The ANB adds to this force by extending it with a targeted deduction. While reasonable nonbelief based, say, on arguments from evil and the multiplicity of religions might fail to show decisively that God doesn't exist, the ANB is a deduction that uses the presence of this relatively well-justified nonbelief, and therefore all the atheistic arguments supporting the nonbelief, as a decisive disproof of God's existence. By itself reasonable nonbelief amounts to some probability against God's existence, but when taken with certain premises and stipulations about God's attributes the nonbelief gains a much more powerful thrust, indeed certitude that by way of incompatibility with any level of rational nonbelief a certain type of God does not exist.

The rationality of atheism

This, (2), is the assertion that rational rather than sinful nonbelief does in fact exist, contrary to our expectations given the existence of God as stipulated. What evidence is there to believe this nonbelief exists? As mentioned above, we have (a) the testimony of nonbelievers themselves who have existed for many centuries. We also have (b) the testimony of many theists who likewise grant that nonbelief is rational given the evidence's ambiguity. There is also (c) the theistic emphasis on faith or trust, which would be impossible were theism indubitable, (d) the disagreement of theists themselves on many core elements of their faith, (e) the lack of empirical evidence for and testability of theistic supernatural claims, and the theist's failure to meet the higher-than-average burden of proof warranted by the extraordinariness of these claims, (f) the undermining of religious institutions' authority regarding matters of fact by the success and openness of non-theistic scientific theories, and (g) the unlikelihood that evolution would produce organisms capable of arriving so easily and quickly at any transcendent, ultimate truth. The lines of argument could be further multiplied.

Yet some theists assert that (2) is false, since God has in fact revealed beyond any doubt that he exists. Specifically, the bible is said to confirm the details of the gospel, as well as its consistency and majesty, and God is claimed to have incarnated himself in the human form of Jesus Christ to confirm for us his love, his empathy for our plight and his solution. Both demonstrate for us indirectly at least the fact of God's existence. But how persuasive is the evidence for these two claims? Is there unambiguous evidence that Jesus was God in human form or that the bible is divinely inspired? Are these claims decisive and overwhelmingly supported by the evidence such that atheism is irrational, insane, or evil? If so, then premise (2), that there does in fact exist rational if not obviously correct atheistic belief, is false and the ANB as formulated above is unsound.

Aside from the above seven answers against theistic claims in general, the following specific doubts about the bible and the Incarnation are warranted. What evidence is offered in support of the claim that the bible is "God's Word"? Is the bible miraculously harmonious? Are there no inconsistencies or contradictions in the bible? How much interpretation is required to make the bible perfectly consistent, and why shouldn't God the author have been able to write a book that doesn't require massive amounts of interpretation in the first place? Why don't all optimum interpretations of the bible by bible scholars agree? How does it help non-Greek and Non-Hebrew speaking people for the bible to have been written in Hebrew and Greek? After all, sophisticated debates about particular theistic doctrines very often depend on differences over the translation of a single word or phrase. How does it help to have lost the first manuscripts of the gospel narratives so that we can't be certain what was originally said? What is so miraculous anyway about a library whose volumes were specifically chosen out of a number of incompatible runners-up because of the canonical ones' relative consistency concerning key issues? What is the value of hearsay testimony regarding any claim, especially a metaphysically unfamiliar miracle claim? What empirically supported theory of memory could account for a stable transmission of accurate historical details about Jesus' life and death over a period of decades, especially when there were no computational (rigidly accurate and non-creative) storage devices or rigorous scientific standards for history-telling, and when the midrashic technique of using ambiguous Scriptural stories found in the Jewish holy texts to find meaning in contemporary confusing events and even in colouring the description of these events was commonplace among Jewish writers?

Likewise, what evidence is offered in favour of Jesus' divinity? The above objections counter the appeal to the bible. How persuasive is the theist's testimony regarding a personal relationship with God? Such testimony is contradicted by the lack of personal experiences with invisible persons among nonbelievers. Personal religious experiences are also untestable and excessively subjective. Is such personal testimony overwhelming evidence for outsiders, given that the theist has the high burden to prove that she prays to a living yet transcendent being rather than to empty space? Are Christians so different from everyone else? Is their behaviour overwhelming evidence of their alleged personal connection to God? Are they wiser, happier, or more successful than nonbelievers? There is no direct, hard evidence today that someone two thousand years ago is in fact the second member of a tripartite deity. And the indirect evidence is grossly subjective and based substantially on faith, the will to believe regardless of the evidence. None of this comes close to establishing the necessary irrationality of atheism, which is required to refute the ANB's fourth premise.

How strong must God's "decisive," "unambiguous" evidence for theism be? This is actually the wrong question. The question is not whether we can state precisely the standards for acceptance of circumstantial evidence regarding some publicly verifiable proposition by all SRICP. After all, we might not presently be clever or powerful enough to define or utilize such a standard. If such a standard exists, however, God should certainly be able to fulfill it. Yet it's clear that unless all nonbelievers are liars, self-deceived, or profoundly evil, this standard--assuming it exists--has not in fact been met by God for all SRICP at this or any other time. I see no reason to believe that the existence of such a standard is impossible. Meeting this standard appears, rather, to be a practical challenge, perhaps only temporarily out of our reach. For example, although most specialists agree on some fundamental points in their respective disciplines, there is always a minority of apparently intelligent, sane and informed individuals who hold out for an alternative interpretation. This is true in all the sciences, and far more so in the humanities. Different disciplines will of course have their own standards for evaluation of the evidence to match their subject matter.

In the case of verifying someone's existence, perhaps we should look to the detective's standard. Among all detectives who are charged to locate an individual and demonstrate her existence, is there some universal verification standard, besides producing the individual herself, that is acceptable to all rational people? What sort of indirect evidence would be considered indubitable or impervious to refutation? Is there a foolproof technological process, such as DNA tests? Yet as we learn more about the genetic code, we will likely find ways to produce false DNA readings or in some way to deceive the detectors. Moreover, what sort of uniform educational process would have to be in place to allow everyone to understand the evidence let alone accept it? Again, all of these appear to be empirical challenges not logical bans. There is nothing logically contradictory about such a standard, and thus an omnipotent being with infinite options should, if she wanted to, be able to meet this standard and convince all rational individuals of some proposition's truth using only indirect, circumstantial evidence.

God's nonexistence

This, (7), is the ANB's conclusion. If God exists he would eliminate responsible, inculpable nonbelief, an overwhelming obstacle and unfair to his children's love for him. This sort of nonbelief, however, has not been eliminated and in fact exists. Therefore God does not exist. The mere presence of responsible atheism, the rational lack of theistic belief, counts as decisive evidence against a fatherly God's existence. There should not be this sort of ambiguity and confusion in a world supposedly created and run by a loving divine parent. Either he exists but is far different than most theists believe, and like a deadbeat dad God doesn't desire anything from us but perhaps some spectacle for his voyeuristic interests. Or else God doesn't exist at all and therefore can't intervene or reassure us that our theistic leanings are in fact warranted. We hear no exasperated call from our horrified heavenly Father, no adequate sign of his outrage by the absence of clear evidence for his existence in the home he supposedly built for his children, simply because there is no divine parent, the universe is not our home, and we are no one's children.

[1] Theodore Drange, "Nonbelief as Support for Atheism" < http://www.bu.edu/wcp/Papers/Reli/ReliDran.htm >, 1998, spotted June 14, 2001.

[2] J. L. Schellenberg, Divine Hiddenness and Human Reason, 1993.

[3] For Drange's detailed examination of the ANB see his book, Nonbelief and Evil, 1998, and his essay, "The Arguments from Evil and Nonbelief" < http://www.infidels.org/library/modern/theodore_drange/aeanb.html >, 1996, spotted June 14, 2001.