Assumptions of the Divine Hiddenness Argument
John Schellenberg has presented an argument noteworthy in several respects. One interesting respect is that his “divine hiddenness” argument is a philosophically interesting innovation in a debate that has raged for millennia. Innovation in philosophy, especially an interesting innovation, is not an easy task, but Professor Schellenberg has accomplished it. The divine hiddenness argument (DHA) runs:
- If there is a perfectly loving God, all creatures capable of explicit and positively meaningful relationship with God who have not freely shut themselves off from God are in a position to participate in such relationships. And,
- No one can be in a position to participate in such relationships without believing that God exists. So,
- If there is a perfectly loving God, all creatures capable of explicit and positively meaningful relationship with God who have not freely shut themselves off from God believe that God exists. And,
- It is not the case that all creatures capable of explicit and positively meaningful relationship with God who have not freely shut themselves off from God believe that God exists: there is nonresistant nonbelief. So,
- It is not the case that there is a perfectly loving God. And,
- If God exists, God is perfectly loving. Therefore,
- It is not the case that God exists.
A key idea of the DHA is that a perfectly loving being would desire the best for its beloved. Another key idea is that a deep relationship or friendship with God would constitute a very great good for creatures. So, God, if he exists, would desire that each creature enjoy the benefit of a deep friendship with him. Of course, very little is said about what an “explicit and positively meaningful relationship with God” consists in. And this dearth of detail may be important, as much hangs on what such a relationship would be.
In what follows I argue that interest and innovation aside, Schellenberg’s argument is unsound. My argument for this judgment proceeds along three paths. The first two paths are but short sketches for thinking that the DHA is unsound. My third objection is developed in greater detail. Before looking at the objections it will be useful to note several assumptions required by the DHA.
Assumptions of the Divine Hiddenness Argument
The DHA has, like all arguments, assumptions or unstated premises. One assumption of the DHA is the proposition that:
A1. The probability that God exists, given the available supporting evidence, is significantly greater than one-half, if God exists.
Importantly, proposition (A1) is supposed to be a necessary truth, a proposition true in all possible circumstances. We might symbolize (A1), using standard notation, and employing as placeholders G for God exists, E for the evidence indicating that God exists, this way:
∧[G → P(G/E) >> 0.5]
Notice that if (A1) were false, Schellenberg’s argument would fail. If it were possible that the probability that God exists, given the available supporting evidence, were equal to, or nearly so, one-half, then it would not be a necessary truth that that probability had to be significantly higher than one-half (assuming the existence of God). In symbols, (A1) is false if
◊[P(G/E) ≈ 0.5 & G]
Proposition (A1) makes it clear that no middling probability assignment for God is allowed if the DHA succeeds.
A second assumption of the DHA is that absolute evidentialism is true. Absolute evidentialism (AE) is the thesis that:
for all persons S and propositions p and times t, S ought to believe that p at t if the evidence renders p more likely than not at t; and S ought not believe that p if the evidence does not render p more likely than not at t.
A famous anecdote involving Bertrand Russell vividly captures the absolute evidentialist attitude: having been asked what he would say to God if after death he were to find himself before the divine throne, Russell answered, “not enough evidence God, not enough evidence.” That the DHA assumes AE can be seen by remembering that the argument is erected upon the alleged consequences of the notion of perfect divine love; a love without limits, or defects. According to Schellenberg, God would ensure that each competent creature is exposed to evidence sufficient to generate theistic belief. The level of evidence would be so high that only an irresponsible disregard could produce nonbelief. Put another way, the DHA requires that each person has strong reason to believe. A Pascalian, of course, would point out that that is what we in fact do have. It is in the interest of each person to form the belief that God exists. Schellenberg does not countenance that response, assuming that the divine insurance would be purely evidentiary and not pragmatic in nature.
A third assumption is that doxastic voluntarism is false. According to doxastic voluntarism, believing is a direct act of the will, with which propositions we believe under our immediate control. A basic action is an action that a person intentionally does, without doing any other basic action. Jones’ moving of her finger is a basic action, since she need not perform any other action to accomplish it. Her handing the book from Smith to Brown is not basic, since she must intentionally do several things to accomplish it. According to doxastic voluntarism, forming a belief is in some cases a basic action. We can will, directly and voluntarily, what to believe and the beliefs thereby acquired are freely obtained and are not forced upon us. In short, one can believe at will. Schellenberg rightly assumes that doxastic voluntarism is false. But of course even if doxastic voluntarism is false, it does not follow that we have no control over our beliefs. The falsity of doxastic voluntarism is compatible with our having indirect or roundabout control over our beliefs. So, while we lack direct control of our beliefs, we do have indirect or roundabout freedom over our beliefs.
A fourth assumption is that God, being perfectly loving, loves universally and equally; that every human is beloved and, as a consequence, is a recipient of equal treatment on the part of God. It is a common claim of theists that God is perfectly good, and by that they mean not just that God perfectly loves, but that God is perfectly just. God’s love, then, would have to be calibrated to that degree compatible with the other properties essential to divine perfection. Divine love may not have the consequence Schellenberg assumes if that consequence is incompatible with divine justice. Schellenberg’s assumption blithely ignores a venerable theological tradition populated with names like Paul, Augustine, Thomas, Luther, and Calvin, which asserts that divine love is constrained by divine justice. This tradition holds that grace is necessary for one to appreciate the evidence in support of theism; without grace, one will not believe. But grace is a divine gift of which justice precludes a universal distribution. If this tradition were correct, premise (6) would be false as understood in the sense necessary for the validity of the DHA.
The first objection focuses on premise (2) of the DHA. In support of (2) Schellenberg asserts that:
For the belief that God exists is obviously and necessarily one of the aforementioned conditions of being in a position to exercise one’s capacity for relationship with God–how can I hear God speak to me or consciously experience divine forgiveness and support or feel grateful to God or experience God’s loving presence and respond to it in love and obedience and worship if I do not believe that there is a God?
So, according to (2), belief is required to enjoy a deep relationship or friendship with God. But there is good reason to doubt this. To see why, consider the distinction between belief and acceptance. Accepting a proposition, unlike believing, is an action that is characterized, in part, by one’s assenting to the proposition, whether one believes it or not. One accepts a proposition when she assents to its truth and employs it as a premise in her deliberations. What is it to believe a proposition? Believing a proposition is being disposed to feel that it is probably the case. Belief and acceptance typically converge, but they can diverge, since one can believe a proposition that one does not accept. For example, think of the gambler’s fallacy. One might believe that the next toss of the coin will very probably come up Tails, since it has been Heads on the previous seven tosses. Nevertheless, one ought not to accept that the next toss must come up Tails, or that the probability that it will is greater than one-half. Acceptance, unlike believing, is an action that is under our direct control. If one accepts a proposition, one can also act upon it. Acting upon a proposition is behaving as though it were true. The two-step regimen of accepting a proposition and acting upon it is a common way of inculcating belief in that proposition. And, importantly, there is no hint of self-deception tainting the process.
The relevance of this distinction is that one can accept that God exists, even if one does not believe that God exists. Since acceptance is under our direct control, one can choose to accept, even if one cannot choose to believe. Indeed, God, if he exists and perfectly loves, may value acceptance, since God would know that doxastic voluntarism is false. Keeping in mind that one way to inculcate a belief is by accepting the proposition and acting upon it, one might think that acceptance is an action that God, if he exists, would value. Of course, much hangs on just what an “explicit and positively meaningful relationship with God” is. If we anthropomorphize that idea, I suspect we’ll have one result; and if we don’t, I suspect we’ll have a different result. In any case I know of no good reason for thinking that, if God were to value acceptance, acceptance would preclude one from a deep relationship with God. If this is correct, premise (2) is false. And if (2) is false, the DHA is unsound. While more needs to be said here, we move on to the second objection.
The second objection is directed toward Schellenberg’s absolute evidentialism. According to AE, it is wrong to form beliefs or to preserve beliefs which lack the support of adequate evidence. To inculcate a belief on the basis of a pragmatic argument is wrong, whether morally or cognitively, according to AE.
But as argued in my opening case, AE cannot be sustained. If AE were true, it would be necessarily true. There are, however, possible situations in which taking steps to form or maintain a belief lacking adequate evidence is morally obligatory. And since no one is irrational in doing her moral duty, AE is false. Think of it like this. Suppose you’re married, and you’ve been confronted with evidence that your spouse is a bank robber. Knowing well your spouse you have reason to believe that your spouse has been mistakenly accused of the crime. Weighing the evidence, pro and con, in as disinterested a manner as you can, you find that you have just about as much reason to doubt your spouse’s innocence as you do to affirm it. Although the evidence is balanced, you do not suspend judgment on the matter. Remembering your vow to love and cherish, you take steps to maintain the belief that your spouse is innocent by continuing to accept her innocence. If more con-evidence were to become available, you’re prepared to concede; but until then and as long as the evidence is at worst balanced, you aspire to honor your vow by maintaining the belief that your spouse is innocent. If AE were true, you’d be wrong not to disabuse yourself of the belief that your spouse is innocent. But no one could justly charge you with irrationality or with immorality in this circumstance. AE is, therefore, false.
Since AE is false, arguably no one would be wrong, in certain circumstances, to form beliefs on the basis of a pragmatic argument. A Pascalian holds that it is permissible to form a theistic belief on the basis of a pragmatic argument when one finds oneself with as much reason to believe as not to believe. If the Pascalian is right, then everyone has overwhelming reason to inculcate theistic belief, since, as I showed in the opening case, the Jamesian wager is a dominance argument–depending on how the world turns out, taking steps to form the belief that God exists may be in your best interest, and doing so never renders you worse off than any other action open to you.
Why is the falsity of AE relevant? Schellenberg dismisses the response that nonbelievers are in fact culpable for nonbelief. While it is true that the evidentiary situation may be ambiguous, it is manifest that the pragmatic situation is conclusively tilted toward theistic belief. All persons have overwhelmingly good reason to accept that God exists, and to inculcate theistic belief. This is a point clear enough for all to see. A common way of trying to elude this point is via an unfounded allegiance to AE. An allegiance that proclaims, perhaps arrogantly, not enough evidence God, not enough evidence–despite the fact that there is abundant reason to believe.
An old joke may aid in developing this objection: a devout Calvinist is trapped on top of his house surrounded by rising flood waters. His neighbor from the north comes by in a canoe and tells the Calvinist to climb in. “No, I will wait on the succor of the Lord” he says. Later, as the waters rise, his neighbor from the south floats by in a boat and tells the old man to climb in. “No, I wait upon God to rescue me” the old man answers. As the waters rise even higher, a neighbor from the west arrives on a barge and implores the old man to climb aboard. “No, I wait upon the Lord” the Calvinist replies again. Soon the old man is swept away by the flood and drowns. Finding himself postmortem before God, the old Calvinist asks God, “Lord what happened? I faithfully waited for your rescue.” God says to him, “Did you not see the three boats I sent?”
The target of objection three is assumption (A1):
A1. The probability that God exists, given the available supporting evidence, is significantly greater than one-half, if God exists.
Keep in mind that (A1) is, allegedly, a necessary truth. The only reason Schellenberg provides for thinking that perfect divine love would necessarily ensure that all persons are presented with strong evidence if God exists is an analogy with human parents. It appears that Schellenberg thinks it is obvious that it is impossible that both God exist and that the pro and con evidence be roughly equal. But there is good reason to doubt that this is a necessary truth. To see this, consider what we will call “the Story”:
Suppose God exists and desires that humans choose to enter into a relationship with Him. God desires, that is, that humans accept Him as a vital concern in their lives. Moreover, since belief is a passive state over which one has no direct control, God would not present one with evidence sufficient to elicit theistic belief, since such “automatic belief” would not preserve the free choice to align oneself with God. What God values is the initial choice to freely accept, the freedom to choose to align oneself with God, and the effort to try to bring about belief, the free inculcation of belief. God would present reason sufficient to motivate one to choose to accept God, but God would not expose one to strong evidence, since he desires the decision to accept to be as unfettered as possible. Presenting a religiously ambiguous creation God preserves the freedom of both acceptance and the inculcation of belief.
Suppose it’s not clear whether two propositions, P and Q, are logically compatible. One way of showing that they are is to come up with a third proposition, R, which is itself a possible proposition and is consistent with both P and Q, and to conjoin R to P (or to Q). If the conjunction (P & R) entails Q, then the conjunction (P & Q & R) is possible. And if (P & R & Q) is possible then so too is (P & Q). And, hence, P and Q are compatible.
In the Story I conjoined the proposition that God exists with various propositions about belief and acceptance and about God valuing free acceptance and the free inculcation of belief. It follows from my conjunction that we would expect the evidence that God exists to be as likely as not. For our purposes, of course, the Story need only be possible. Even if the Story is far-fetched, it is far from inconceivable. Perhaps the Story is false; it may well be. But as long as it is not necessarily false, the Story serves its point, since it implies that there could be a situation in which both God exists and the evidence does not render the existence of God significantly greater than one-half. If the Story is possible, (A1) is false:
◊[P(G/E) ≈ 0.5 & G] → ~∧[G → P(G/E) >> 0.5]
Is the Story possible? Clearly it is, since, for one thing, it entails no contradiction. And if (A1) is false, Schellenberg’s argument is unsound.
Why might God value the conjunctive state of affairs of free acceptance and free inculcation? Keeping in mind that it only need be possible that God values these, support for the possibility of that valuation can be gleaned in at least a couple of ways. One way would build upon the recognition that belief ebbs and flows with one’s grasp (whether reliable or not) of the evidence. Perhaps God would value acceptance as a kind of protection for the believer, since one can control one’s acceptances, even if one cannot directly control one’s beliefs. According to this idea, God, desiring that no one would be harmed by an erosion of belief caused by their grasp (whether reliable or not) of the evidence, provides strong reason to accept even in the absence of strong evidence. A second way builds upon what we might call the Celebrity Case:
Suppose you are a rich and famous celebrity. You know that among your entourage are many who associate with you just because you’re rich and famous. You seek, however, true friends. You realize that celebrity gets in the way of establishing a deep relationship as the lure of wealth, power, and fame lead people away from you as a person and toward your celebrity. To establish deep friendships requires that you try to find persons ignorant of your celebrity, or indifferent to it, who will like you for who you are, regardless of your celebrity status.
In a situation like this it makes sense for one to hide her celebrity, as she seeks friends. What is important is that an appropriate foundation is laid, which will support the superstructure of a deep friendship. As a celebrity might hide that fact about herself, perhaps God has a similar reason to hide certain facts about himself in order that an appropriate foundation might be laid that’ll support a deep and free relationship.
One might object that there’s a big difference between God the creator and a celebrity. The celebrity thinks of friendships as a good for herself; while God would not, since it is the good of the creation that’s important. Moreover, in the celebrity case, what’s hidden are facts about wealth and fame, but with God the fact allegedly hidden would be existence, and how could hiding divine existence be good for God’s creatures?
In response to this objection think of a teacher preparing his students for a standardized exam the results of which will determine the student’s life-chances (admission to the best schools say). While the teacher should certainly not provide the answers by showing the students a purloined copy of the exam, s/he should provide three things: enough information to prepare the students for the exam, the motivation to try their best, and the requisite skills to apply what they’ve learned in original ways (as opposed to being “taught to the test”). S/he should do this for the good of the students, since otherwise s/he is harming them for life. In like manner a theist could hold that God obscures his existence to preserve both the freedom to accept and the freedom to inculcate belief, while at the same time providing enough evidence of his existence such that it is as likely as not, and strong reason to motivate the effort to inculcate saving-belief. Divine hiddenness could be good for God’s creatures by preserving the dual freedoms of acceptance and inculcation, which are necessary for establishing a deep, free and genuine relationship.
I suspect that something very much like the story is true. Mentioning that hunch, however, is needlessly extravagant as truth in this context is overkill. All I need is that the Story is possible, and enough has been said to make manifest that possibility. Since the Story is possible, (A1) is false. And with (A1) false, the DHA is unsound, since premise (2) is false. Premise (2) asserts:
No one can be in a position to participate in such relationships without believing that God exists.
But if the Story is possible, it may be that every human is in a position to freely accept that God exists, and to freely take steps to try to bring about the belief that God exists, without the occurrent belief that God exists. A deep and meaningful relationship with God may require, for all I know, that the requisite belief is earned through free acceptance and taking steps to inculcate that belief, rather than just finding oneself saddled with it. Clearly enough, since the Story is possible, the DHA is unsound. Religious uncertainty does not provide a coup de grâce to theistic belief, by tipping the scales decisively toward skepticism.
Continue the Debate
 I owe this point to Joel Pust.
 The symbol P(x/y) should be read: the probability of x given that y. The symbol x >> y should be read: x is significantly greater than y. The symbol x << y should be read: x is significantly less than y. The symbol x ^ y should be read: either x or y. The symbol ∧(x) should be read: It is necessarily true that x. The symbol x → y should be read: if x, then y. The symbol ◊(x) should be read: It is possibly true that x. The symbol x ≈ y should be read: x is almost equal to y.
 Schellenberg comments on a different aspect of this tradition in his important book, Divine Hiddenness and Human Reason (Ithaca, NY: Cornell University Press, 1993): 74-82.
 Schellenberg, “What Divine Hiddenness Reveals, or How Weak Theistic Evidence is Strong Atheistic Proof.”
 My development of this distinction owes much to the discussion in William Alston, “Belief, Acceptance, and Religious Faith” in Faith, Freedom, and Rationality, eds. Jeffrey Jordan & Daniel Howard-Snyder (Lanham, MD: Rowman & Littlefield, 1996): 3-27.
 By “anthropomorphize” I mean understanding the relationship with God as in all relevant respects being the same as human relationships.
 Richard Gale formulates a similar argument contra evidentialism. See his The Nature and Existence of God (Cambridge: Cambridge University Press, 1993): 357.
 I owe a variant of the “teacher case” to Doug Stalker.
 This debate has been about the appropriateness of theistic belief given religious uncertainty. Importantly, belief, while the attitude usually favored, is not the only attitude which could serve as a foundation for theistic commitment. One might embrace theism because one accepts it, or even out of hope.
 I thank Paul Draper, Michael Murray, Joel Pust, and Doug Stalker for their comments.
Copyright ©2008 Jeffrey Jordan. The electronic version is copyright ©2008 by Internet Infidels, Inc. with the written permission of Jeffrey Jordan. All rights reserved.