Divine Hiddenness, Greater Goods, and Accommodation

J.L. Schellenberg argues that one reason to think that God does not exist is that there are people who fail to believe in Him through no fault of their own. If God were all loving, then He would ensure that these people had evidence to believe in Him so that they could enter into a personal relationship with Him. God would not remain ‘hidden’. But in the world, we actually do find people who fail to believe that God exists, and their nonbelief does not seem to be due to their resisting God. I argue that if there are valuable goods brought about by God’s hiddenness, then even if each of those goods might obtain without hiddenness, God would have a sufficient reason for remaining hidden so long as enough of those goods would be made sufficiently more valuable because of God’s hiddenness. If this is the case, then the existence of ‘nonresistant nonbelievers’ in the actual world does not entail that God does not exist.


Introduction
Does the fact that there are people who, through no fault of their own, fail to believe that God exists count as evidence for the nonexistence of an all loving God? J.L. Schellenberg argues that if an all-loving God existed, then one of the greatest goods for His creatures would be to have a close personal relationship with Him and that God would ensure that all of His creatures would be in a position to enter into such a relationship just by choosing. But this requires that those creatures first believe that God exists. The fact that there are 'nonresistant nonbelievers' is therefore evidence that an all-loving God does not exist.
Schellenberg often employs what he calls the 'accommodation strategy' to reject attempts by critics to argue that God might temporarily allow nonresistant nonbelief for the sake of some resulting good. The accommodation strategy demonstrates that proposed goods are possible within a relationship with God. I argue that the accommodation strategy is rendered ineffective if many goods result from temporary nonresistant nonbelief and if at least some of those goods are sufficiently greater outside a relationship with God.
The structure for this paper is as follows. First, I outline the argument as presented by Schellenberg and discuss the support for its key premise. Then, I discuss how 'greater goods' responses are supposed to respond to that premise. After that, I introduce my strategy: the 'many goods response' and how it might overcome some of the objections commonly aimed at greater goods responses, including Schellenberg's 'accommodation strategy'. This is followed by a discussion of several objections to the many goods response. The final two sections discuss how various goods might fit into my proposed strategy; some of these goods have already been proposed in the literature while the final good is one that I propose for the first time here. If enough goods which fit into my strategy plausibly exist, the argument from divine hiddenness will fail.

The Problem of Divine Hiddenness
The problem of Divine Hiddenness is presented by Schellenberg (2007) in the following form: (1) Necessarily, if God exists, anyone who is (i) not resisting God and (ii) capable of meaningful conscious relationship with God is also (iii) in a position to participate in such relationship (able to do so just by trying).
(2) Necessarily, one is at a time in a position to participate in meaningful conscious relationship with God only if at that time one believes that God exists. Therefore, (3) Necessarily, if God exists, anyone who is (i) not resisting God and (ii) capable of meaningful conscious relationship with God also (iii) believes that God exists [from (1) and (2)]. But, (4) There are (and often have been) people who are (i) not resisting God and (ii) capable of meaningful conscious relationship with God without also (iii) believing that God exists. Therefore, (5) God does not exist [from (3) and (4)].
The inference from (3) and (4) to (5) is valid, and so one must deny one of these premises in order to maintain that God exists. But the inference from (1) and (2) to (3) is also valid, and so one cannot deny (3) without denying one of those premises as well. Much of the criticism of Schellenberg's formulation focuses on the support for premise (1).
Early on, Schellenberg supported premise (1) by arguing that if a perfectly loving God were to exist, He would desire the highest well-being for His creatures. For all creatures that are capable of personally relating to God, their highest well-being would include a personal relationship with Him. So it seems that if a loving God exists, He desires to have a personal relationship with every creature that is capable of entering into one with Him. He would do everything He could to make that relationship possible (Schellenberg 1993). Later, Schellenberg backs up the claim that the highest well-being for creatures would include a personal relationship with God. He argues that, if a perfect and all loving God would create finite creatures at all, He would do so in a way that reflects His relationalpersonal love. If He creates, 'the creativity God exercises in producing finite creatures must itself be expressed in pursuit of value realizable in and through and by them' (Schellenberg 2007). And so the value God would seek in and through and by finite creatures must include relational-personal value. 1

Greater Goods?
Few theists would deny that that a relationship with God would be a very great good. 2 Nevertheless, 'greater goods' style responses attempt to show that there is some good for the sake of which God would be willing to sacrifice a personal relationship with some of His creatures for a period of time. Greater goods style responses address the support for premise (1) by attempting to demonstrate that an all-loving God could consistently desire the best for His creatures 3 while simultaneously allowing them to be nonresistant nonbelievers. If this is true, and there is no other reason to think (1) is true, then (1) is left groundless and we cannot conclude that God does not exist. 4 Here are two examples of greater goods style responses. Richard Swinburne (1998) suggests that the opportunity to make morally significant free choices might be a greater good for which nonresistant nonbelief is necessary. If a creature has too great an awareness that God exists, then the desire to please God (or the fear that God will punish wrong behaviour) will make it too easy to do the right thing, and thus her decision will not be morally significant. 1 Schellenberg also argues that an all-loving God would desire a personal relationship with every capable creature for its own sake. But it seems to me that, if God is omnibenevolent, God's own desire would only come into play in the case of a tie. It seems clear that if some person were better off not knowing that God exists (at least for a time), God would not reveal Himself to that person (at that time) even though He desires a personal relationship with that person for its own sake. One might wonder how anyone might be 'better off' not knowing about God; this will be addressed below. 2 It's likely that many non-theists also maintain that a relationship with God would be a very great good. 3 Sometimes, these goods contribute to the well-being of individual creatures; sometimes, they contribute to the total value of the world as a whole. 4 In order for a 'greater goods' response to succeed, it must meet at least three criteria. First, the proposed good must outweigh the evil it justifies. Second, the proposed good must require, necessarily, that that evil obtains (or that some other evil equally as bad as or worse than that evil obtains) in order to come about. If the response on offer is a defense, then a third criterion for success is that the good must plausibly obtain for all we know. There is also a common fourth criterion given. It is commonly held that God cannot impose an evil on some innocent person S for the sake of some good that does not benefit S. I address this in note 9. These criteria are addressed in Vandergriff (2016).
Daniel Howard-Snyder (1996) suggests that God wants everyone to respond appropriately to Him and that He would be justified in remaining hidden from certain people until they are ready to respond appropriately. Schellenberg (2007) utilizes his 'accommodation strategy' to demonstrate that proposed goods, such as those mentioned above, can plausibly be accommodated within a personal relationship with God. If this can be shown, then these goods cannot count as a sufficient reason for God to allow nonresistant nonbelief. It would seem that, in order for a proposed good to overcome the accommodation strategy, the existence of nonresistant non-belief must be a necessary condition for acquiring that good (Schellenberg 1993). For, if God might acquire it in a way that does not require nonresistant nonbelief, He would do so. But Daniel Howard-Snyder (1996) points out, 'if permitting [nonresistant] nonbelief is no worse than any other means for some significantly greater good to obtain', then God might reasonably allow for nonresistant nonbelief to obtain in order to acquire that greater good.
I suggest that a proposed good might overcome the accommodation strategy if that good is somehow more valuable given temporary nonresistant nonbelief. If nonresistant nonbelief sufficiently increases the value of some good, then it will constitute a sufficient reason for God to allow for nonresistant nonbelief. 5 But Schellenberg has noted that a personal relationship with God would be the greatest good, and so no single possible good would be sufficiently good for God to allow it. Travis Dumsday (2010) notes two possible strategies to get around this. First, if there is some great good which requires-at some point in time-the existence of nonresistant nonbelief, and if that great good is a necessary condition for having a personal relationship with God, then God would allow for nonresistant nonbelief. Second, Dumsday suggests that a cumulative case might be the most promising strategy and that there may be many goods which result from nonresistant nonbelief for the sake of which God would allow nonresistant nonbelief. It is his second strategy that I aim to promote.

The 'Many Goods' Response
Rather than identifying a single good as a proposed reason for why God might allow for nonresistant nonbelief, the 'many goods' response [MGR] admits that no single good may be sufficient. Instead, the MGR holds that the multitude of goods brought about by nonresistant nonbelief together provide a sufficient reason for God to allow for nonresistant nonbelief.
Contrast this with the problem of evil, where 'greater goods' responses require that a multitude of evils must all be justified. Greater goods responses to the problem of evil must either require many goods to account for the many evils that occur, or they require a small number of very valuable goods which can justify all the evil there is. 6 Either way, weighing the evils we know about against the goods that arise from those evils can be somewhat complicated. The many goods response to the argument from divine hiddenness is plausibly a bit simpler. This is because nonresistant nonbelief is a single evil (if it can be properly called an evil), and multiple goods may arise from a single instance of evil without that evil increasing. 7 This makes weighing value just a bit simpler. If many goods are brought about by a single evil, then it is easier to see how they might provide a sufficient reason for allowing that particular evil. This is how it might look. Let 'W' denote a world that includes nonresistant nonbelief, and let 'W′' denote a world that has no nonresistant nonbelief, but is identical to W in all other relevant respects. 8 If W has a total value which is greater than W′, then God has a sufficient reason to bring about W rather than W′. 9 It should be noted that each individual good proposed in the MGR might not be essential to the success of the MGR. Even if a critic is able to demonstrate that some particular good does not qualify as contributing to the overall value brought about by nonresistant nonbelief, she will not have demonstrated that the MGR fails. In order to show that the MGR fails, critics must demonstrate that it is not plausible to think that the total value of all the proposed goods is sufficient to 6 Or some combination of the two. Perhaps one very great good can account for a large amount of the evil that occurs, but not all, while some other lesser goods account for the remaining instances of evil. And of course perhaps 'greater goods' responses themselves only account for a portion of the evil that occurs, while the rest can be accounted for in some other way. But I am not primarily concerned with the problem of evil here. 7 The point I'm making here is not that there is only one kind of hiddenness. Even though there are many kinds of hiddenness, as demonstrated by Schellenberg (2007), the reason why each of kind of hiddenness seems at odds with the existence of God is the same: each kind of hiddenness precludes (or at least impedes) the possibility of a personal relationship with God. Each person, at some given time, can only be deprived of a personal relationship with God once (by contrast, one person may experience several different kinds of evil at once). 8 An anonymous reviewer has pointed out that there may be problems with framing the issue in this way. Because of the butterfly effect, even slight changes to a world might ripple outward to create massive differences. And whether or not a world has nonresistant nonbelief seems to be much more than just a slight difference. Because of this, the 'relevant respects' in which W and W′ are identical must be very few (as the differences created by nonresistant nonbelief would be many). I make the neat comparison between W and W′ for convenient explanation, but it is not essential to the point I am making. As we will see, the point I am making is simply that if enough goods arise from some instance of nonresistant nonbelief, and if that set of goods cannot come about any other better way, and if their combined value outweighs the value lost due to nonresistant nonbelief, then God would be justified in allowing for that instance of nonresistant nonbelief. On the scale of worlds, rather than single instances, it might be very difficult for us to compare worlds. But so long as God knows all the details of a world before He creates, the butterfly effect will not pose a problem for His choice of a world. 9 A critic might argue that comparing the total value of two worlds is insufficient to show that God would create one world over the other. She might argue that God would ensure that the overall well-being of each individual who experienced nonresistant nonbelief would be greater because of that nonresistant nonbelief (or at least no worse). It seems to me that, for those who are persuaded of this conclusion, an analogous argument to mine might be made. Each occurrence of W and W′ might reasonably be substituted by 'H' and 'H'. H refers to an individual creature's life which includes at least a period of nonresistant nonbelief. H′ refers to that same creature's life if it contains no nonresistant nonbelief (all else being equal). If, for each and every H, H is on the whole more valuable than H′, then God would permit H.
Alternatively, a critic might argue that, for the MGR to be successful, it must be demonstrated that all W worlds are more valuable than all W′ worlds or that each W world is more valuable than its corresponding W′ world. This would be a more difficult argument to make. However, it doesn't seem to me that this is a requirement for the MGR to be successful. If God would create only the best possible world, then it must only be demonstrated that the best possible W is greater than the best possible W′ (it must also be demonstrated that the greatest possible W is the actual world, but that would move beyond the problem of divine hiddenness). outweigh the value lost by nonresistant nonbelief. 10 That is, critics must show that the total value of W is, based on the goods we have identified, not plausibly higher than the total value of W′. So while the total value brought about by the various goods which result from nonresistant nonbelief may be sufficient for God to allow nonresistant nonbelief, there may be some goods which contribute to the overall value but are not necessary for God to allow nonresistant nonbelief.
Since the MGR admits that a relationship with God would be the greatest good, there is no longer the burden of identifying one single good for which God would be willing to sacrifice a relationship. This means that when a potential good is proposed, it does not need to be discarded when its value is inevitably judged to be lower than the value of a relationship with God. Any good at all that is brought about by nonresistant nonbelief will contribute to the overall value of W. This means that the task of identifying goods might actually be relatively simple. There might be many ordinary goods that result from nonresistant nonbelief, and they might seem insignificant taken individually, but together they might 11 be significantly valuable. Of course, the success of the MGR will be simpler to determine if a few very significant goods which result from nonresistant nonbelief can be identified.
Another advantage of the MGR is that nonresistant nonbelief does not have to be a logically necessary condition for bringing about any of the goods. Even if for each good identified for the MGR there is some possible way to attain that good without allowing for nonresistant nonbelief, the MGR may still succeed. So long as nonresistant nonbelief is the only way for God to acquire a sufficient number of goods, the MGR will succeed. The MGR will also succeed if all of the goods in question might be acquired by some means other than nonresistant nonbelief, but that other means causes a greater amount of evil than nonresistant nonbelief. Again, God has a sufficient reason for allowing nonresistant nonbelief if the total value of W is greater than the total value of W′.
This general advantage can be applied more specifically to J.L. Schellenberg's accommodation strategy. With 'greater goods' style responses that appeal to only one good, if a proposed good can be demonstrated to be possible to acquire within a relationship with God, then it cannot count as a sufficient reason for God to allow nonresistant nonbelief. For the value of a relationship with God in conjunction with that good will be much greater than the proposed good on its own. But the accommodation strategy is not so devastating against the MGR. Even if some particular good might be possible to acquire within a relationship with God, it might still count in favour of W, rather than W′. The good in question, it must be admitted, will only count if it is in some way greater given nonresistant nonbelief. If there is a good which can be acquired both in W and in W′, and its value is the same in both worlds, then it does not count in favour of either world. However, even if the particular good in question is slightly greater in W than it is in W′, it will count in favour of W. Given this, it is possible that the MGR will succeed even if every good identified can also be realized within a relationship with God. If each good is possible to acquire in both W and W′, but a sufficient number of those goods are sufficiently greater in W than they are in W′, then the total value of W might still be greater than the total value of W′. This is true, even keeping in mind that W′ includes the value added by relationships with God that do not obtain in W, and keeping in mind that the value of those relationships cannot be exceeded by any single good.
If the goods identified can be demonstrated to have increased value if there is nonresistant nonbelief, then the MGR may succeed even if every good is possible within a relationship with God. 12 Of course, goods which cannot be accommodated for within a relationship with God make the task of weighing goods less complicated, but they don't necessarily contribute more value to the overall value of a world. There may be some goods which are only possible given the existence of nonresistant nonbelief, but if they have a very low value then they may still contribute less value than a good that can be accommodated, if that good has a significantly higher value given nonresistant nonbelief.
For the MGR to be successful, it must be demonstrated that the total value of W is plausibly higher than W′. 13 If there are good reasons to think that W is plausibly more valuable than W′, then the critic of the MGR must give reasons to think that W is not plausibly more valuable than W′. If the total values of W and W′ are too close to call, then it seems to me that it is not implausible to maintain that W is more valuable than W′, but neither is it implausible to maintain that W′ is more valuable than W. If this is the case, then argument from divine hiddenness will result in a stalemate.

Objections to the MGR
The many goods response is not without challenges. A proponent of the Problem of Divine Hiddenness might argue that the value of a relationship with God is so great that the total value necessary to outweigh it cannot be achieved even by a great number of other goods. The value of a relationship with God might either be infinite, in which case it would be impossible for any number of goods to outweigh it, 14 or it might be so much greater than any other good that it is highly unlikely that a great enough number of goods exist to outweigh it.
I know of no argument to suggest that the value of a relationship with God would be infinite, 15 and it seems, prima facie, to be finite, given that only the goodness of God Himself is infinite. It is reasonable to suppose that a relationship between God and a human would be of finite value on account of the human's imperfection. But this is no reason to suppose that the criticism fails. Even if the value of a relationship with God is not infinite, it might still be so great that there simply aren't enough goods from nonresistant nonbelief to outweigh the value of a relationship with God. 16 To assess this criticism, two varieties of nonresistant nonbelief must be considered. First, a person might fail, through no fault of her own, to ever believe in the existence of God [permanent nonresistant nonbelief]. Second, a person might fail, through no fault of her own, to believe in the existence of God for a temporary period of time, but eventually come to form the belief that God exists [temporary nonresistant nonbelief]. Temporary nonresistant nonbelief allows for the possibility that the person in question might eventually choose to enter into a relationship with God, 17 while permanent nonresistant nonbelief does not allow for this.
If some person is a permanent nonresistant nonbeliever, then she can never enter into a relationship with God. 18 It is reasonable to think that in this case the value of the relationship lost would be so great that the goods gained from the individual's nonresistant nonbelief would be insufficient to outweigh that lost value. 19 However, if a person is a temporary nonresistant nonbeliever, and she eventually comes to believe that God exists, then she might possibly enter into a relationship with God. If she does enter into a relationship with God sometime after she comes to believe, then it seems reasonable that that the total value lost by her temporary nonresistant nonbelief would not be so great that it could not be outweighed by many goods. 20 That is, the total value of W might still be greater than the total value of W′ if the nonresistant nonbelief within W is temporary.
I do not suggest that the total value of W is necessarily lower than the total value of W′ if W contains permanent nonresistant nonbelief. But unlike temporary nonresistant 16 An anonymous reviewer argues that, even if a personal relationship with God is not infinitely good, it might be that being connected to an infinite good (God) is more valuable for one than any other set of goods that doesn't include a connection to the infinite good. In response, I point out that the argument from divine hiddenness relies on the specific good of a personal relationship with God being unsurpassably valuable. A personal relationship with God is one way of being connected to God. But, if God existed, then it seems probable that many created goods provide one with a connection to God, even if one is a nonresistant nonbeliever. Probably, these provide one with a lesser or looser connection to God than a personal relationship with God, but I go on to explain how the MGR might address this issue further down. 17 T.J. Mawson (2012) suggests that every person in the world will eventually enter into a relationship with God, though many people may only do so after they have already died. 18 This idea may be challenged by Andrew Cullison (2010), where he argues that two persons might have a personal relationship with each other even if one does not believe that the other exists. If Cullison's argument succeeds then it only creates further problems for the argument from divine hiddenness. 19 It is not implausible to suggest that the value lost from permanent nonresistant nonbelief could be potentially infinite, especially if Schellenberg is correct (see note 12). 20 One might argue that a personal relationship with God is so great at all times, so that God could have no sufficient reason to allow even temporary nonresistant nonbelief. This strikes me as implausible. While I offer no knockdown argument, consider the following. Suppose God remains hidden from Smith for a day. It seems plausible to me that God might have reasons to remain hidden from Smith-perhaps, for example, Smith learns an important lesson about desiring God during that time. There might also be other benefits from that time. In the long run, a day is not very long for God to remain hidden from Smith, and Smith has the rest of His life to enjoy a personal relationship with God. If Smith's day of nonresistant nonbelief enhances his lifelong personal relationship with God, long after that day is over, then that day's lost value might be outweighed. But even a lifetime of hiddenness seems very short relative to an eternity of enjoying a personal relationship with God. And if there are goods which only come about because of one's lifetime of nonresistant nonbelief and which enhance one's eternal relationship with God, it seems that God would be justified in allowing for that lifetime of nonresistant nonbelief. nonbelief, it is not obvious that permanent nonresistant nonbelief actually occurs. Even if many nonresistant nonbelievers die without deathbed conversions, they may become postmortem believers. 21 If permanent nonresistant nonbelief is incompatible with the existence of God, then God's existence would simply entail that permanent nonresistant nonbelief does not occur. 22 So unless one can demonstrate that permanent nonresistant nonbelief actually occurs, or very likely occurs, the mere fact that it is incompatible with the existence of God does not pose a threat to theism. 23 One way to object to the MGR would be to hold that, plausibly, various goods cannot actually be compared to one another in the way I have been assuming, such that the total value of worlds themselves cannot be compared to one another in this way. If the total value of worlds cannot be compared in the manner I have been suggesting, then, as pointed out by Klaas J. Kraay (2011), this 'threaten[s] the rationality of God's choice of a world'. If God's choice is not rational, then there is no explanation for why He would create one world instead of another. 24 The problem of incomparability goes beyond a discussion of divine hiddenness. For this reason, and because of space limitations, I cannot respond adequately to this objection. I move forward on the assumption that all worlds are comparable. For it may be that any reason to think that they are not comparable 25 proves only that humans are incapable of making accurate judgments of value. But this does not mean that an omniscient being would be incapable of making such judgments. 26 A third way one might object to the MGR is also related to God's ability to compare the values of W and W′. In order for God to choose to create either W or W′ based on 21 This idea is compatible with Mawson's account of universalism (see note 11), but my position might not necessarily entail universalism. It might be that everyone becomes a believer post-mortem but can freely choose whether or not to enter into a personal relationship with God. However any account like this will need to be expanded and defended and that is outside the bounds of this paper. 22 Theists within the Christian tradition should not be surprised if permanent nonresistant nonbelief does not occur. The Bible indicates that there will be a time when 'every knee shall bow to [God] and every tongue shall confess to God' and 'each of us will give an account to God' (Romans 14:11-12, ESV). Of course, what happens to each person after God reveals Himself will be controversial, and much too big an issue to address here. 23 An anonymous reviewer objects that this response begs the question for the theist, since I say that the existence of a perfectly loving God entails that it is not a problem that some people die as nonresistant nonbelievers-but whether or not a perfectly loving God exists is our present concern. But all I am arguing is that we have no way of knowing whether or not there are permanent nonresistant nonbelievers. The best evidence that some permanent nonresistant nonbelievers exist is atheism (especially naturalism), and the best evidence that no permanent nonresistant nonbelievers exist is theism. Thus, we should remain agnostic about the existence of permanent nonresistant nonbelievers until we determine whether atheism or theism is very probable. But this means that we cannot use the existence of permanent nonresistant nonbelief in an independent argument for atheism. 24 This is not only a problem for the MGR. Any greater goods style response must face this difficulty. Furthermore, it seems that the argument from divine hiddenness itself relies on at least a certain level of comparability between worlds. 25 Two reasons are mentioned by Kraay. First, worlds might be judged by different criteria so that there would be multiple ways to rank worlds. But if this is the case, then the worlds are incomparable. Second, suppose that there are two worlds, neither of which surpasses the other in value. If a minor improvement to one world does not increase its value such that it is more valuable than the other world, then those two worlds are incomparable. 26 My thanks to an anonymous reviewer who adds that, even if there is no non-arbitrary standard for comparing worlds and goods, this does not mean that one world or good cannot be preferred over others. The example given by the anonymous reviewer is the goods of a sunset, a chocolate bar, and a child's health. Even if these three goods cannot be compared by a common metric, it seems clear that the child's health should be preferred. My claim is that God might have insights like this at the level of possible worlds. their values, He must know what those values are prior to creating. But whether God can know all the minute details of the world He creates-especially if that world includes creaturely libertarian free will-is controversial. While some, such as Molinists, think that God can know everything about a world with libertarian free will prior to creating, others reject this. It would be a virtue for the success of the MGR not to depend on the truth of a controversial position like this. 27 While it is well outside the bounds of this paper to argue that God could know everything about W and W′ prior to creating, I think it is important to highlight this concern. But even if the MGR, as I have outlined it, does assume this controversial position, it seems plausible that a reply similar to the MGR might be possible even if God cannot know everything about the world He creates prior to creating it. Even if, for example, open theism is true, God might know the set of possible worlds that might obtain, given the initial conditions He puts in place. As long as God knows that there are many goods which can come about only given nonresistant nonbelief, 28 He might be able to determine that the potential reward is greater than the potential risk. 29 But a complete, satisfying outline of this strategy is also outside the bounds of this paper.
The final objection to MGR I'll address is that, if it succeeds, it creates a new problem for theism. If there are many goods that are brought about by nonresistant nonbelief, then why is the number of believers so high? 30 Why doesn't God desire for there to be universal nonresistant nonbelief? Two points address this concern. First, since no single good is valuable enough on its own for God to sacrifice the good of a relationship, God could only justify sacrificing a relationship with a person if it would allow for a number of goods which together were sufficiently valuable. Second, many of the goods might be possible only if some believers exist. To see how this might be the case, I now turn to address some potential goods which plausibly contribute to the MGR.

Some Potential Goods
The goods listed in this section do not exhaust the proposed goods within the literature, and I address them only to demonstrate how goods can be assessed as part of the MGR. 31 I provide only a very rough sketch of each good, and there may be problems with each good that goes unmentioned here, much less adequately defended against. But through this discussion, I hope it will be clear that these goods might be incorporated into the MGR, and together overcome Schellenberg's accommodation strategy. 32 One suggested benefit of permitting nonresistant nonbelief is that it makes it possible for people to have morally significant free choices (Swinburne 1998). The idea is that a 'deep awareness of the presence of God' gives humans too much incentive to choose good actions rather than bad ones. Fear of punishment or hope of earning God's favour would make the good choice easy, and thus would require no strength of will on the part of the individual. God would make Himself hidden so that humans could have a real opportunity to exercise their will. While there is not enough space to discuss the difficulties of this strategy as a 'greater goods style response' in length, Schellenberg (2007) argues that this good can be accommodated for within a relationship with God. He also points out that this strategy relies on each person having a 'deep awareness of the presence of God', while the argument from divine hiddenness claims only that God would ensure that all humans have evidence sufficient for belief. He expresses doubt that having evidence sufficient for belief would really inhibit a person's ability to make morally significant free choices. 33 Even if Schellenberg's response here succeeds, there may be a way to incorporate this good into the MGR. If the value of some moral choice is to any degree greater when chosen by a nonresistant nonbeliever than it is when chosen by a believer, then this contributes value to W. It seems plausible that there are at least some cases in which a morally significant free choice is more valuable, and in which choosing the good demonstrates a greater strength of will, when it is chosen by a person who does not even have evidence sufficient for belief in the existence of God. Suppose, for example, that Smith and Jones are both filing their taxes. They both know that they can cheat and get away with it, without anyone getting hurt. But Smith has evidence sufficient to believe that God exists while Jones does not. It is intuitive that it will require much less strength of will for Smith to cheat than it will for Jones. Smith can remind himself 'God is watching', and easily do the right thing. Jones cannot be so sure, and his doubt might even play a role in making the deliberation more difficult; as he considers doing the right thing he knows the only person who might find out is God, but he might think 'I don't even know if God exists'. If Jones does the right thing, God's hiddenness will have allowed him to exercise a greater strength of will than Smith has. While this is not sufficient on its own for a 'greater goods' style response, it still contributes to the MGR. 32 However, I will not conclude whether or not the MGR ultimately succeeds in defeating the argument from divine hiddenness. This is a project that may ultimately take many more years of discussion and publication, as potential new goods continue to be proposed year by year. 33 An anonymous reviewer suggests there is 'another possible problem with the claim that a deep awareness of God would make it too easy to choose the morally right action'. For example, if the right action requires a very great sacrifice on the part of the person who is deeply aware of God's existence, it might still be very difficult to do the right thing. But even if this is true, it only means that doing the right thing in certain instances will be just as difficult whether or not one knows God exists. In order for the good of morally significant free will to contribute to the MGR, it only needs to be that nonresistant nonbelief allows for enough instances in which one must exercise a greater strength of will to do the right thing than one would otherwise have to. For more on the factors which might contribute to the strength of will needed to act rightly, see Murray (1993).
A second benefit of allowing for nonresistant nonbelief is that it makes it possible for some humans to have the responsibility of telling other humans about God's existence (Swinburne 1998). Schellenberg (2007) argues that this good, too, might be accommodated for within a relationship with God. Since there may be many significant truths about God, God could give some believers the responsibility of teaching other believers these truths. Travis Dumsday (2010) argues that, even if this good can be accommodated for, the responsibility of telling another person that God exists is the greatest responsibility (and privilege) a person could have. If he is correct, then this particular good contributes significantly to the MGR. Furthermore, this good requires the existence of believers to have this responsibility, so it helps to explain why there would be so many believers, despite the benefits brought about by nonresistant nonbelief.
Another reason God might allow for temporary nonresistant nonbelief is because He wants His creatures to respond appropriately to learning of His existence: by pursuing a relationship with Him, rather than rejecting Him, and by pursuing that relationship for the right reasons (Howard-Snyder 1996). For this reason, God might remain hidden from someone until she is ready to respond in the right way. While this may not provide a sufficient reason on its own for God to remain hidden, it seems that W likely contains a greater number of people who respond appropriately to God's existence than W′, since in W God is able to wait until the opportune moment to reveal Himself to individuals..
One final benefit from nonresistant nonbelief suggested in the literature is that it allows for the possibility of genuine self-sacrifice (Cullison 2010). Unless someone sincerely believes that there is no God, she cannot sacrifice her life without considering the hope (however small it may be) that she will continue to exist in the afterlife. But unless she sacrifices her life knowing that she is truly giving everything up, her sacrifice cannot be genuine. While this does not provide a sufficient reason on its own for God to prefer W over W′, it plausibly increases the value of W, and, to that extent, contributes to the MGR. 34

A New Good
I suggest that there is another good-making feature of W that W′ lacks. In W, there can be a greater variety of personal testimony about entering into a relationship with God. Personal testimonies are unique to each individual, but they can be divided into at least four main categories. First, there are people who have always believed in God, and whose relationship with God has grown since they were small children. Second, there are people who culpably resist God for a period of time until they finally accept God. Third, there are some who first have a relationship, and then fall away for a time, before returning to God. 35 The final category includes nonresistant nonbelievers who do not know that God exists at all until something happens which allows them to believe.
It is plausible that an individual's personal testimony can be good for both the individual and for the community to which an individual shares her testimony. 36 First, for many individuals, whatever their testimonies may be, they connect with those testimonies in a way that brings them each value. Testimonies are unique to each person, and many believers acknowledge that they would not be the same people without them. Thus, reflecting on their own testimonies encourages them and gives them the sense that they are part of God's plan. And it may be that God does have a plan for testimonies. It may be that each person's story of how she entered into a relationship with God is the best story for that individual. If God knew this from eternity, then He would do what he could to create a world where each person lives the story of coming to know God that is best for her.
Personal testimonies also provide communal goods. Many individual members of the community may benefit from one person's testimony, and that person's testimony will also benefit the community as a whole. Individual members who hear a person's testimony are able to empathize and know and understand what it would be like to be that person. This allows individuals to understand the many ways God works, and to perhaps appreciate their own testimonies in a new light. The community itself is then strengthened as people are able to understand each other in deeper ways. These testimonies may also encourage the members of the community. It's plausible that the community will grow stronger and flourish if there are a variety of testimonies and they make a practice out of sharing those testimonies.
A testimony which includes a period of nonresistant nonbelief may be particularly valuable. If a person is such that the best story for her is one which includes nonresistant nonbelief, then she will appreciate this story when she reflects on it and shares it with others. She will see it as a valuable part of God's plan, and may exalt Him because of it. It may also cause her to reflect on how, even while she did not know of His existence, God was watching over her and loving her. Further, her story will be valuable to the community of which she is a part. If she shares her story, then the members of the community will understand in a new way what it would be like to not know of God's existence. This will allow them to appreciate their own stories in a new way, and be in awe of God's intricate plan. They will understand her and themselves better, and so the community will grow stronger.
There are a couple of possible objections here. First, one might wonder why God would require there to be actual cases of nonresistant nonbelief, rather than fictional stories. If W′ contains fictional stories of nonresistant nonbelievers, then the believers in that world might have the communal benefits of those stories; they might be able to understand what it would be like to be a nonresistant nonbeliever, and appreciate their own stories in a new light. However, even though this good might in some way be accommodated within a relationship with God, it is clear that it would be more valuable if there were actual nonresistant nonbelievers. Only in this case could there be the benefits for the individual. 37 Furthermore, 36 Personal testimonies may also provide goods for God Himself. God is exalted in a person's testimony, and each testimony is a testament to God's perfect plan. An all-loving God would likely not see this alone as a sufficient reason for a person having some particular testimony, but it is still an added benefit of a person's testimony. However, one might object that God is unsurpassable and unchangeable, and therefore, nothing could possibly benefit Him. So whether or not personal testimonies can possibly bring value to God will partly depend on one's view of God. 37 This assumes that the fictional benefits for the fictional nonresistant nonbeliever do not contribute actual value.
it seems intuitive that a story which is factual has more power than a fictional story, and so the benefits to the community would be greater if there were actual nonresistant nonbelievers.
A second objection is that the stories of resistant nonbelievers would be enough to gain these benefits, so nonresistant nonbelief would not be necessary. While it is true that there is value in the testimony of a person who once resisted God but finally came to accept Him, this only means that God might desire for there to be some people with this testimony as well. It does not mean that the testimony of a former nonresistant nonbeliever would not be valuable. That added value is all that is needed to contribute to the MGR. Furthermore, nonresistant nonbelief might be a more desirable state of affairs than resistant nonbelief, since nonresistant nonbelievers do not sin in their nonbelief. If God knows some individual would resist belief in Him, He might withhold from that individual evidence sufficient for belief in Him. 38 As a third objection, one might remain unconvinced that the variety of personal testimonies found in a world without nonresistant nonbelief would be of lesser value than the variety of personal testimonies found in a world which includes nonresistant nonbelief. Surely, there would still be a great diversity of personal stories about coming to know God. What makes the personal testimony of noresistant nonbelievers so important? 39 I have a couple of points that address this concern. First, I think that if humans understand-to some extent-what it would be like to be apart from God, we can gain a new appreciation for not being apart from God. The personal testimonies of nonresistant nonbelievers allow many people to gain this new appreciation without all of them having been nonresistant nonbelievers themselves. Second, if there were never any nonresistant nonbelievers, it's plausible that the practice of sharing personal testimonies would itself be less common. After all, these are our stories about coming to know God, and if most people always knew God, then we would be less likely to make a practice out of talking about always having known God.

Conclusion
In this article, I have briefly sketched a strategy for overcoming Schellenberg's accommodation strategy, and therefore the problem of divine hiddenness. If there are enough goods brought about by nonresistant nonbelief, then the value of a world containing nonresistant nonbelief might be greater, all else being equal, than the value of a world without nonresistant nonbelief. A response of this kind simplifies the task of identifying goods because any good brought about by nonresistant nonbelief, no matter how valuable it is, contributes to the value of that world on the whole. Even goods which might be realized in a world with no nonresistant nonbelief may count towards the many goods response if they are somehow more valuable because of nonresistant nonbelief. 40 38 Daniel Howard-Snyder (1996) argues in a similar vein. 39 My thanks to an anonymous reviewer for pointing out my need to address this concern. 40 Thanks to those at the 2015 Canadian Philosophical Association Annual Congress, the 66th Annual Northwest Philosophy Conference, and the 1st IIFs-UNAM Philosophy Graduate Conference who provided valuable feedback on earlier drafts of this piece. Thanks as well to those who offered their thoughts when I presented this paper to the Toronto Philosophy of Religion Work-In-Progress Group. I especially want to thank Klaas Kraay for guidance on the earliest drafts of this piece. Finally, thanks to the anonymous reviewers for their thorough comments.