There is a popular theistic argument meant to explain why God permits even a slim evidential basis for atheism. In “Godfellas,” an episode of the cartoon show Futurama, Bender the robot is given the chance to be God for a group of people, and learns the need for God to balance the interest in helping his creatures with the interest in preserving their freedom. Bender either interferes in their lives, with disastrous results, or does nothing at all, with disastrous results. The episode concludes with the real God claiming that ideally a deity’s creatures shouldn’t be able to tell that they have divine assistance so that the deity might even appear not to exist.
An informal statement of the argument behind this sentiment is found in a Toronto Sun column by Michael Coren. In “A Matter of Faith …” he writes: “God does not give us absolute proof because this would work against our free will. He gives us just enough evidence so that we can find Him and just enough to reject His existence if that is our desire.” Coren then adds: “Don’t blame the God in whom you do not believe because He asks you to work just a little before you have faith.”
This argument is an application of the Freewill Defense against the Argument from Evil. The assumption is that atheism is bad and needs to be reconciled with God’s good purpose. The solution is that atheism is an unintended consequence of people’s freedom. In particular, it’s plain why most Christians need to believe that some voluntary and laudatory work–including mental or spiritual work such as wanting to have faith–has to be done by people for them to be saved. Otherwise, people’s escape from punishment in hell would be predetermined by God’s grace; but if reward in heaven could be caused entirely by God rather than at least to some extent by people’s own effort, so too might punishment in hell be caused by God’s pulling of puppet strings. In this way, the analogy between a divine and a human parent would break down, since human parents typically don’t punish their children for no good reason; if people have no genuine freedom to control their actions, no reward or punishment of them could be morally justified. Then the Christian would be left with mysticism, with the view that God is beyond our comprehension. Thus there would be no basis for a religious institution.
So if atheism itself–the belief that God doesn’t exist–is taken to be a sin deserving of everlasting punishment in hell, perhaps even the root of many other sins, the Christian needs to regard atheism as voluntary. To merit punishment, the atheist needs voluntarily and sinfully to reject the evidence of God’s existence, which evidence would have convinced this same atheist were he or she to have freely cultivated a more-susceptible frame of mind.
Special Pleading and the Objectivity of Epistemic Justification
Were Coren’s claim about the voluntariness of belief that God exists to be extended to a general claim about the justification of beliefs, the evidence which justifies beliefs would generally have to be ambiguous and subjective. The evidence would appear strong or weak depending on whether the evidence were considered by the right or by the wrong sort of person. Presumably, “right” and “wrong” here would mean, respectively, virtuous and vicious. For example, were the evidence considered by an arrogant, untrusting, bitter person, the evidence would lead only to an unjustified belief. And since the point of justifying a belief, of supporting a belief with reasons, is roughly to confirm the correspondence between the belief and a certain fact, the unjustified belief would be regarded as likely false.
Now, the evidence against this general view of the rational justification of belief is overwhelming, since many different kinds of people share the same beliefs, including hundreds of thousands of mundane beliefs such as that the sky is blue and that people usually have ten fingers. Surely whether most beliefs are justified doesn’t depend on whether the believer is a good or a bad person. A person may lack many virtues and still have many justified beliefs.
On the other hand, some minimal virtues do seem required for a person to have any justified beliefs, such as the virtues of intellectual curiosity and of what might be called ambition. Without the desire to succeed, for example, a person wouldn’t be interested in checking the evidence or in possessing accurate beliefs, and thus the person wouldn’t think about objections or alternative explanations, in which case the person’s beliefs might all be rationally unjustified even if ultimately they were true. But this doesn’t support the extension of Coren’s point, since in the case of most beliefs, the virtues needed are so minimal as to be hardly worthy of praise. After all, the worst murderers in history would have possessed these virtues, since they couldn’t have successfully carried out their evil acts were they to have been uninterested in having true beliefs, say, about where their victims were or whether their weapon would suffice for the task. The kind of virtues that support the rational process of evaluating evidence for mundane beliefs are so minimal as to be innate features of the human mind rather than praiseworthy dispositions or genuine virtues.
For this reason, Coren’s claim might be a case of special pleading. In most cases, a belief is rationally justified if the belief is based on sufficient evidence. Call this kind of justification epistemic, since it has to do with knowledge. For example, a belief might have strong inductive support, in which case the belief is justified by the objective strength of the empirical evidence. The strength or weakness of this evidence has to be objective to account for the fact that most beliefs, especially mundane ones, are both universally held and also justifiable by each believer regardless of the believer’s character traits. Both Hitler and Jesus surely believed that humans usually have ten fingers, that some things should be eaten and some not eaten, that putting a limb in fire causes pain. And both believers could have supplied a skeptic with adequate reasons that justify these and a million more mundane beliefs. Such beliefs are justified because the evidence has certain objective properties, namely ones that reliably indicate that the belief agrees with the facts. Moreover, the evidence will be such that an action based on the belief will have a greater chance of succeeding than were there no such evidence in the belief’s favour. So Coren needs a reason why the belief that God exists is special in being subjectively justified, given that the justification of most beliefs is objective, depending just on the quality of the evidence which all kinds of people can equally well appreciate.
Coren might respond that the question of God’s existence is special, given something like Paul Tillich’s definition of “religious faith” (theistic belief) as ultimate concern. For two reasons this definition would be of no help to Coren. First, the object of ultimate concern, God, would have to be subjective. Some people are ultimately concerned about God, the heavenly parent, and others are ultimately concerned about money, the god of mammon. The belief that God exists would be just the belief that whatever a person is ultimately concerned about exists. The monotheist would then need a reason why one god is true while all others are false, or why one rather than any other should be worshipped.
Second, this subjectivist definition of theistic belief exacerbates the problem of special pleading, since the question is whether there is such a thing as subjective justification of a belief, given that epistemic justification is normally objective. The problem is that whether most of our beliefs are justified depends on objective features of the evidence, since all different kinds of people equally well possess both the beliefs and the ability to appreciate the reasons supporting the beliefs. At least, this is so for ordinary empirical beliefs. Then the question is whether there is another kind of justification for special beliefs, namely subjective justification which depends crucially not on the objective properties of the evidence, but on the qualities of the person evaluating the evidence. The forgoing response is that one such special belief is religious faith, which is the belief that whatever ultimately concerns a person exists. This response confirms that the belief is special because of its subjectivity, but doesn’t show that this special kind of belief has its own kind of justification. The contrary claim would be that this subjective belief based on ultimate concern, which is consistent with mere wishful thinking, is unjustified precisely because of the lack of objective evidence. Again, this is a problem for a Catholic like Michael Coren, because without some way of justifying theistic belief, there would be no reason for God to regard atheism itself as a sin worthy of punishment in hell, and the comparison of the divine parent to a human one would break down, reducing theism to mysticism and undermining religious institutions.
The Reduction of Epistemic to Ethical Justification
A better response to the objection about special pleading would be that theistic belief is special because those qualities of the believer which are relevant to whether the belief is justified and praiseworthy are the ethical ones of virtues and vices. This would explain how theistic beliefs differ from ordinary beliefs. What I called the “virtue” of ambition is just the survival instinct to succeed, but this instinct isn’t virtuous precisely because it’s innate and involuntary. But the belief that God exists might be supported by genuine virtues, whereas atheism might be supported by vices. This might also explain subjective justification, by linking it to the ethical evaluation of character. Just as a virtuous action might be said to inherit its praiseworthiness from the character of the person performing the action, so too might the belief that God exists be a mental or spiritual action of a sort, and the fruit of a virtuous character; on the other hand, atheism might be the rotten fruit of vice and therefore blameworthy.
One obvious problem with this response is that there seems to be insufficient evidence warranting the claim that theists are virtuous whereas atheists are vicious. There are many indicators of a person’s character, and whether the person believes that God exists doesn’t seem to be a reliable one. The theist would have the burden of showing that atheists are generally less virtuous than theists. This is often assumed about atheists, but to my knowledge there has been no study done worthy of the task of rationally supporting the generalization. In granting that everyone sins and that theists often sin in the name of their religion, many theists thereby grant that there are no such correlations between virtue and theism, and between vice and atheism.
Another problem with the second response is that while the ethical standard can be used to evaluate a person’s character and behaviour, this standard seems irrelevant to the question of a belief’s rational justification. This is because the justification of a belief, as opposed to the justification of a person’s character or behaviour, is a component of knowledge and is supposed to point to the truth. A justified belief has a better chance of being true than an unjustified one, which is why people are interested in supporting their beliefs with reasons and with a consideration of the evidence. But whether a person is genuinely virtuous seems to be no reliable indicator of whether the person’s beliefs are true, or whether they agree with the facts. At best there might be some accidental correlation between having a certain character and having justified or unjustified beliefs. Even were theists typically virtuous, which they are not, this wouldn’t show that the belief that God exists ought to be evaluated in ethical terms, that is, by the standard that seems appropriate only for the practical evaluation of a person’s character and deeds. Just as an accidental correlation between an alignment of the planets and people’s behaviour wouldn’t call for the reduction of astronomy to astrology, so too the accidental correlation between theistic belief and virtue wouldn’t call for the reduction of the standard of epistemic justification to the ethical standard.
The problem, then, is that epistemic justification isn’t logically connected to the ethical standard. The former has to do with evidence supporting a belief and thus with the right to call the belief likely true. The latter standard has to do with preferred human behaviour. Given a realist, commonsense theory of truth, a preference doesn’t affect the given state of the world, although a preference can lead to the world becoming a certain way. The ethical standard reflects our collective desire to live in a certain sort of society, whereas a belief’s epistemic justification reflects the likelihood of the belief’s objective agreement with the relevant fact. The agreement between a belief and a fact may hold in spite of anyone’s preference for the belief to be false.
Let’s return to Coren’s application of the Freewill Defense: “God does not give us absolute proof because this would work against our free will. He gives us just enough evidence so that we can find Him and just enough to reject His existence if that is our desire.” Coren then adds, “Don’t blame the God in whom you do not believe because He asks you to work just a little before you have faith.” So if a person prefers that there be no God, the person regards the cup as half empty, as it were, and becomes an atheist. Since God is not to be blamed, the fault lies with the atheist, and presumably with the atheist’s desire not to be ruled by God. The assumption is that atheism is bad and thus that blame is appropriate. On a Christian view, the atheist deserves to be punished for being an atheist. Ultimately, the punishment is not for the lack of theistic belief, since this is just the rotten fruit of the tree, the tree being the atheist’s sinful desire, which is to say the atheist’s vice. For example, the preference that there be no God might amount to the arrogant assumption that humans are capable of governing their own affairs.
The primary defect of this argument is the confusion between epistemic and ethical justification. Even if the atheist were sufficiently guilty to warrant punishment in hell, such that the atheist’s character and deeds were condemnable on ethical grounds, this wouldn’t show that atheism itself, the belief that God doesn’t exist lacks rational, epistemic justification. The judgment of an atheist as a person is logically independent of the judgment of atheistic belief. To say otherwise is to commit the genetic fallacy, since the origin of atheistic belief in the atheist’s sin would have no bearing on the rational validity of the belief itself. The question of a belief’s epistemic justification rests on an entirely different standard from the standard which determines whether the atheist’s character and behaviour can be ethically justified for the atheist to avoid punishment. The epistemic “standard” is just the objective evidence which points to the relevant fact with which a belief can be in agreement. But the ethical standard is thoroughly prescriptive, not descriptive. Thus, the argument also commits the naturalistic fallacy of failing to appreciate the category distinction between what is the case and what ought to be the case.
This reduction of one standard to the other is fallacious, given the commonsense objective view of truth. Were the theist to claim that atheistic belief is nothing more than a sinful act which deserves punishment, the theist would assume a subjective view of truth–at least with respect to whether God exists. Then the proper response to atheism wouldn’t be that God will really punish atheists for their atheism, but that the theist would prefer that atheists be punished for that reason. On the subjective view of truth, faith matters, because the “object” or target of theistic belief is only that which ultimately concerns a person, which depends on which person is concerned. So too, on the subjectivist view, the question of a belief’s epistemic justification might be nothing more than the question of a person’s ethical justification. After all, there would be no objective relation between belief, evidence, and the facts, since were truth subjective there would be no objective facts.
Michael Coren’s argument, then, is self-contradictory. On the one hand, he assumes as a Catholic that atheism is a sin which deserves to be punished. So far the epistemic standard of a belief’s rational justification isn’t applied to atheism. But Coren also says, of course, that it’s God himself, not some figment, who presents the evidence in an ambiguous way. Thus, Coren also assumes that the standard of epistemic justification is in play. God really does exist, after all, and he really will punish atheists in hell. But atheism itself isn’t exactly irrational; on the contrary, the evidence of God’s existence is ambiguous by divine design, and the decisive factor leading to atheism isn’t the nature of the evidence itself, but the atheist’s sinful desire to be an atheist. This latter move on Coren’s part conflicts with the application of the epistemic standard of justification to atheism, since now Coren blames atheism not on the nature of the relevant facts or on the state of the evidence, but on sin, which requires the application of the ethical standard. On the one hand, atheism is just a sin; on the other hand, God really does exist. Thus, on the one hand, atheism is unethical; on the other hand, atheism is rationally unjustified and false.
It might seem that Coren entirely rejects the epistemic standard of justification, if not the objective view of truth, since Coren says that both theism and atheism are supported by preferences rather than by the unambiguous nature of the evidence. But I think this would be a mistaken interpretation. Theism begins with the virtuous desire to have faith, and thus with a technically irrational leap beyond the evidence, but then God is found by means of a personal relationship. Faith is eventually supported with personal religious experience, but this experience is evidence subject to the epistemic standard of justification. Thus, theism may begin irrationally, but once supported with personal experience of God, theism is epistemically justifiable, according to what I think is the best interpretation of Coren’s Christian position. The guiding idea, familiar to readers of the New Testament, is about searching for God and only eventually finding him.
Thus the conflict is between the theistic expectation about the objective state of the world, and the moralistic psychoanalysis of the atheist which supposedly reveals that atheism rests on a sinful desire. Given the objective views of truth and of justification, any sinful desire on the atheist’s part is irrelevant to whether atheism is true or epistemically justified. What matters is the nature of the evidence.
Were the evidence as ambiguous as Coren suggests, it wouldn’t follow that agnosticism is the only rational position. This is because both theism and atheism might be justified to some extent. The evidence would be mixed, and the rational response would be to draw an inference to the best explanation. As empiricist philosophers of science from Carnap to Kuhn have said, this inference depends on certain pragmatic values which determine a theory’s overall merit. Assuming that the evidence is mixed, and that both theism and atheism are equally hobbled by their incompleteness (by the fact that neither theory can perfectly account for all of the evidence), the appeal is then made to such values as the theory’s simplicity, fruitfulness, and compatibility with established knowledge. Whichever theory provides the best explanation, given these pragmatic values, is selected as superior. This is where Coren might object that certain pragmatic values are sinful. In any case, separate objective arguments could be offered for or against each of the values needed to choose the best explanation, in which case the question of God’s existence could conceivably be resolved in an objective and rational manner, that is, without resorting to agnostic indecision or appealing to the ethical standard of justification and to the threat of eternal punishment.
The reason Coren’s argument contradicts itself is that Coren applies both the ethical and the epistemic standards of justification to atheism, and yet the standards conflict with each other. For example, the claim that the epistemic standard of justification applies to a belief implies that the ethical standard ought not to be applied to the belief. In other words, the commitment to the epistemic standard of justification (understood in the commonsense way I’ve laid out) is coupled with the need to avoid committing the genetic and naturalistic fallacies. But evaluating the belief in ethical terms, taking the belief to be part of the believer’s character or else to be some sort of action, commits both of these fallacies, given that the belief’s epistemic justification depends on the objective evidence.
A Speculative Diagnosis
I’ll end with a diagnosis of this defect of Coren’s application of the Freewill Defense. On the one hand, the modern Christian needs to conform to the epistemic standard of justification which takes truth and evidence to be objective, because this is the scientific standard, and the success of the scientific methods is obvious. This isn’t to say that scientific knowledge guarantees social progress in applying this knowledge. Still, the technological applications show that there is currently a tremendous amount of scientific knowledge. Crucially, the unifying pragmatic value which governs the scientist’s choice of a theory of nature is the value of control over nature. Technological applications of scientific knowledge don’t logically demonstrate the truth of a scientific theory, but the applications do confirm the theory’s power. The value of power over nature is compatible with an objective view of truth, since if the facts of nature are objects rather than subjects with moral rights, these objects present so many opportunities to satisfy the desire to control them. For example, the scientific study of hurricanes assumes that a hurricane is an objective process, one which can be both known and controlled by applying that theory which is the best explanation in so far as the theory satisfies certain pragmatic criteria.
On the other hand, the modern Christian is committed to an ancient cosmology which prevailed when the potential for human control over the world wasn’t as appreciated as it is now. According to most versions of theism, the person who has ultimate power over the world is God, and God will eventually reveal this power by judging all of us and punishing those who displease him. Again, the threat of a show of force over atheists doesn’t logically demonstrate the falsehood of atheistic belief. But given the pragmatic value of power, if the threat of overwhelming force against nonbelievers can be impressed upon people, theism might be accepted as the best explanation on pragmatic grounds. Thus, the difference between the objective and the theistic ethical standards of justification is that the former assumes the interest in people’s control over nature whereas the latter assumes the interest in control over people. These interests are incompatible, because nature can be controlled only by people, whereas controlled, terrified and submissive people can’t control anything.
Even if God were a person, God wouldn’t control nature in the sense of using facts about the world to survive, so much as he would sustain the existence of nature. And although there are some theistic traditions which command human stewardship of Earth, theism still constrains this stewardship by balancing it with the obligation to submit to a higher power. If the unrestrained exploitation of nature can be supported at all by theism, this must be by a hollow version according to which atonement can be made for the concomitant pride by such superficial shows of subservience to God as verbal acts of contrition. This is the way theism can be gutted to serve the primary task of eliminating natural threats to our survival, which we eliminate by embedding ourselves in an artificial environment.
Coren’s argument is caught between these incompatible interests. On the one hand, he has to assume there are objective facts about God, because the success of science has made denial of the objective view of truth unwise. Thus, the question of God’s existence should be determined by the nature of the evidence, and whatever could be learned about God could be used to control God as much as possible. In effect, this is what is assumed by the person who prays to God. On the other hand, the very notion of God is prescientific, which is to say that theism assumes that God’s power over people is more important than people’s power over nature. This is because theism developed prior to the modern awe-inspiring demonstrations of human power by the application of scientific knowledge. Theism therefore commits the theist to being more interested in submitting to God than in having nature submit to science. I think this is the root of the main defect with Coren’s theodicy.