The concept of the theistic God does not itself include a promise of eternal happiness for believers. Nevertheless, many theistic religious systems do assume a guarantee, and historically this guarantee has been offered as a key incentive for belief in God. But is a divine guarantee of eternal happiness possible? The free will defense to the problem of evil states that suffering is the result of the free actions of humans, and thus God could not have created a world with free creatures who would be guaranteed to do only good. I will argue that the same must also apply to free creatures in an afterlife, and that there, too, God cannot guarantee that free creatures will do only good.
This problem derives from the free will defense and as such may be called the eschatological problem of free will. I will begin by briefly stating the free will defense to the problem of evil. After presenting the eschatological problem of free will, I consider several possible solutions to this problem. I conclude by discussing implications of this argument on theism in general. For simplicity, the notion of the afterlife used throughout this discussion will be restricted to an eternal and ideally happy or heavenly realm. Questions relating to other possible realms which are less than happy, such as hell, limbo, or purgatory, will not be considered.
In its most simple form, the problem of evil is as follows:
According to premise one, if God is all good, then he would have the desire to prevent evil. If he is all knowing, then he would know where evil exists and how to eliminate it. If he is all powerful, then he would have the ability to eliminate it. Premise two states that evil exists, and is usually understood to mean the suffering that results from both natural and human causes. Since premise two is clearly true, then this suggests that God, defined in the above theistic terms, does not exist.
The free will defense attempts to reconcile the existence of God with the presence of evil by attributing at least human-caused suffering to the willful acts of free creatures. For, once God created creatures with free wills, then those creatures, and not God, are responsible for the suffering they cause. Further, the free will defense maintains that it would be logically impossible for God to determine the willful acts of free creatures and that God cannot force any free creature to willfully choose to act benevolently, and to never act malevolently. It is then impossible for God to guarantee that a free creature would always use his free will for good purposes.
The free will defense amounts to a voluntary limitation of power on God's part to allow for the possibility of free human actions. Thus, the free will defense attacks the problem of evil by redefining God's power so that it excludes control over the activities of free creatures. Whether the free will defense is successful in preserving God's goodness and accounting for the presence of all evil cannot be explored here. What is significant, though, is that the free will defense shows that the creation of genuinely free creatures precludes any guarantee on God's part that such creatures will always act benevolently.
The question then arises: does a heavenly afterlife include free creatures?
If so, then an eschatological problem of free will emerges:
The conclusion to the above argument can, of course, be blocked by rejecting either of the two premises. However, the traditional theist has compelling reasons for accepting these premises and it is precisely for that audience of traditional theists that the above argument is intended. Premise one assumes that humans have free wills, an assumption which the theist has strong reasons for making. Denying free will would undermine the free will defense, and would consequently undermine the strongest attack on the problem of evil. Further, denying free will reintroduces an additional problem discussed by 18th century determinists: if determinism is true, then all human actions causally trace back to God and would make God directly responsible for evil human actions.
Premise one also stipulates that a person with free will on earth retains free will in the afterlife. For, if an agent with a free will on earth is stripped of his will in heaven, then he has lost an essential feature of his identity, and serious questions are raised about that person's continued identity in heaven. Also, if free will were not retained in heaven, questions could be raised about whether a realm of free creatures is inherently better than a realm of automata. And this reintroduces the problem of evil into the free will defense by questioning why a benevolent God would create an inherently inferior world with free creatures who do evil over an inherently superior world of automata who are programmed to do good.
The theist also has good reason to accept premise two, which is the basis of the free will defense itself. Premise two is also analytically true, given the definition of "freedom." A free act is one in which the agent alone has decisive control over whether that act is done. By contrast, God can guarantee the course of an agent's actions only if God, and not the agent, has decisive control over that action.
What follows, then, is that God cannot guarantee that free creatures in an afterlife will act benevolently. The most that God can guarantee is that God himself will act benevolently in a heavenly realm, but God cannot speak for the behavior of other free creatures in that realm. The type of malevolence that might be inflicted depends on the metaphysical nature of the realm and the type of creatures inhabiting it. Suppose, for example, that the afterlife involved a community of free agents who have new bodies. If those bodies were capable of experiencing physical pain, then they would be vulnerable to physical suffering which other free agents might impose on them. Suppose, on the other hand, that heaven consisted of only spiritual or mental entities. In comparison to God's infinite greatness, our condition would be that of finite mental entities with limited mental qualities. These mental vulnerabilities, then, could possibly be exploited by free creatures to produce mental suffering. This might involve the suffering which results from deceit, degradation, malicious taunting, and competition. Augustine, for example, identifies pride as the fundamental cause of evil, and pride is thought to be an attribute which a purely spiritual entity could possess.
Even if each creature were isolated to be protected from possible malevolent actions of other free creatures, the isolated person, as a free agent, would have it within his power to be self-destructive and inflict himself with suffering. If the individual could experience physical pain by means of possessing a new body, then physical masochism is in the realm of possibility. If he exists as only a mental entity, then self-degradation could result in mental suffering. It would be exceedingly difficult to explore the complete range of heavenly configurations to show that none preclude the possibility of suffering. However, it is sufficient to see that realms containing free creature with finite abilities will be ones where suffering is at least possible.
Like the problem of evil, the eschatological problem of free will involves a concise argument with premises that should be compelling to many traditional theists. Also like the problem of evil, the success of this argument hinges on its ability to resist solutions. Consequently, the remainder of this essay will discuss possible solutions to the eschatological problem of free will.
Several simple solutions to this eschatological problem of free will must immediately be dismissed. First, it does not help to argue that God's omnipotence allows him to guarantee proper behavior in heaven. "Omnipotence," as traditionally understood, involves the ability to do all things which are logically possible. But, as defenders of the free will defense recognize, it is logically impossible for any agent to determine the willful acts of other free creatures. Thus, such an action would be outside the realm of possibility for even an omnipotent being. Second, it would not resolve the eschatological problem of free will to suggest that God could monitor the acts of all heavenly creatures and expel them from heaven just as they were about to do evil. The expulsion itself would violate the guarantee of eternal happiness to those who were expelled. Third, it would not help to argue that an eternal realm lacks the lures which would tempt people to be evil (for example, hormones or insufficient supplies of material goods). For, even without such earthly motivations, an agent could still freely choose to do evil, so long as it is within the realm of possibility.
Finally, the problem cannot be quickly resolved by arguing that God foresees the future and knows before hand that the residents of heaven will in fact choose to behave morally. This solution rests on an ambiguity with the word "guarantee." In the most favorable circumstances, foreknowledge would only offer a nominal assurance of eternal happiness which would be contingent on the decisions of the free creatures in heaven. Theoretically, this is a report which any finite time traveler could issue.(1) It remains, though, that it would not be within God's power to prevent free creatures in a heavenly realm from being evil if they chose. Thus, God cannot guarantee that he will prevent free heavenly creatures from doing evil.
Other possible solutions to the eschatological problem of free will are more interesting and deserve closer examination. James Sennett has recently argued for a notion of human freedom which is such that (1) humans have free will on earth, (2) humans develop character traits on earth by exercising free will, and (3) humans lack free will in heaven, but retain the character traits developed on earth.(2) Sennett's suggestion circumvents the eschatological problem of evil by denying premise one, that humans retain free will in heaven. For, if Jones lacks free will while in heaven, then God can guarantee that the causes of Jones's actions are such that they never result in Jones performing an evil action.
However, the weak link in Sennett's argument is whether the Jones in heaven would be essentially the same person as the Jones on earth in view of Jones's loss of freedom in heaven. According to Sennett, Jones's identity would be retained since he would retain the freely acquired character traits from his earthly status. But retaining only such freely acquired character traits is not sufficient for continued identity. Suppose, for example, that humans were designed in such a way that for 30 years of our lives on earth we had free wills, and for the remainder of our lives on earth we lacked free wills. Further, during the first 30 years we would use our free will to developed character traits, but for the rest of our life we acted only on the basis of the character traits which we developed. Within this context, suppose that Jones spent the first half of his life developing criminal character traits. At any time within the first half of his life Jones could voluntarily reverse his criminal behavior, but for the second half of his life his criminal behavior would be a matter of necessity. That is, Jones would be accidentally evil for the first 30 years, but essentially evil for the remainder of his life.
This contrast of Jones's early and later identities is accentuated if for the first 30 years Jones lived in an inner city environment where criminal behavior was normal, but for the second part of his life was in an otherwise crime-free and utopian environment. Although this scenario is not one which would have a direct counterpart in a heavenly realm, it does suggest that, even when retaining our freely developed character traits, loss of freedom constitutes a loss of identity.(3)
Another solution to the eschatological problem of evil might be that seeing God face to face would provide an adequate incentive for us to be good, and this would be a virtual guarantee of eternal happiness. Unlike life on earth, in heaven we would be in the immediate presence of God and would not reject his goodness. This solution is especially strong when heaven is understood as the absorption of the individual into God, as with the Hindu Atman-Brahman. Here, the individual would be identified with God so that the free will of the individual would be the same as the free will of God. Accordingly, God's goodness would be our goodness, and God's free choices would be our free choices. If God is eternally happy, then we too would be eternally happy.
However, the theistic notion of God does not allow for the absorption of the individual within God, either in this life or in the afterlife. According to traditional Western theism, God is transcendent, wholly other, and distinct from his creatures. Consequently, none of God's creatures, either in this life or the next, can experience God directly. Even in heaven, we would have an incomplete and imperfect understanding of God, and that understanding of him could never go beyond our own limitations. Our incentive to emulate God's goodness, then, would be restricted by our limited capacity to fully comprehend God (as is the case on earth). Thus, even seeing God face to face would not provide a virtual guarantee that we will be good.
The Hebrew allegory of the fallen angels illustrates this problem. The allegory relates how a community of angels, who were free rational creations of God, rejected God for reasons of pride and ambition. On a symbolic level, this story may be seen as a metaphor for any free and rational creature's inability to fully comprehend and appreciate the infinite goodness of his creator. This symbolizes how even our best communion with the wholly other is not a sufficient incentive to behave in a Godly fashion, and will not keep us from evil.(4)
Similar considerations led Kierkegaard to conclude that the first century eyewitnesses of Christ had no special belief advantage over those of us who live two thousand years later. An act of faith is required in both cases. Eyewitnesses of Christ needed to accept on faith that the charismatic leader performing miracles before them was divine, and not simply a magician. By the same reasoning, inhabitants of heaven are also faced with a choice. The free creature stands before a power which is greater than the creature himself. That power claims to be God, and claims he will provide for the creature's needs for eternity. How does the creature know that the power is indeed God, rather than a powerful fallen angel, or simply an illusion? Any proof God might give would then become a subject of choice -- to either accept or reject that proof. In making this choice, there is no guarantee that a free creature will make a choice in favor of God and his goodness.
Another possible solution to the eschatological problem of free will is that residents of heaven consist only of people who were believers on earth and thus have had prior screening on earth. This would assure that all inhabitants of heaven are team players. Although not a full guarantee, this would provide good reasons for expecting residents of heaven to act morally. However, this solution assumes that believers on earth are fundamentally different kinds of people than nonbelievers, and have a special set of virtues. This, though, is an empirical question about the respective virtues of believers and nonbelievers, and little indicates that believers are more moral, less malicious, and more trustworthy than nonbelievers. Hume argued to the contrary that he actually suspected that a person was a rascal if that person had strong religious beliefs.
But even if residents of heaven did pass a preliminary moral examination on earth, a more important factor would be consistency of action. For, if on balance person A was more moral than person B, this would only be an indication of A's future proper behavior if A were consistently moral on earth. To be an effective indicator of consistent behavior in an eternal realm, one's behavior on earth, for a mere 80 years, would have to be almost flawlessly consistent. The phenomenon of religious backsliding shows that believers are not necessarily more consistent in their beliefs than nonbelievers are consistent in their disbelief. Perhaps even a believer's own religious conversion -- from nonbelief to belief -- would count against him, since this would show a radical shift in belief attitude. At this point, the present solution would lack appeal for most theists since the litmus test of consistency would be too strict for most humans to pass, including, perhaps, the theist himself.
A final possible solution to the eschatological problem of free will is to recognize that the problem is flawed with an overly earth-like view of heaven. That is, the discussions so far have construed heaven within the category of time and involved human-type interaction and human-type thoughts. When construed in these ways, it is not surprising that the concept of heaven might mirror the evils on earth. The true nature of heaven might be so other-worldly that earth-like descriptions are inherently misleading. In response, the earth-like construel of heaven in the above rests on three fundamental presumptions about heaven: inhabitants will have free wills, their essential identities will be retained, and God will continue to be distinct from his creations. Abandoning any or all of these would make the notion of heaven more other-worldly, and perhaps resolve the eschatological problem of free will. However, as argued above, the theist would resist abandoning any of these presumptions. For the sake of argument, however, suppose that there was an other-worldly construel of heaven which did not abandon the above three presumptions, and was such that there was a virtual guarantee that all creatures freely chose to do good. A consequence of this possibility is that the present world is not the best possible world God could create which contained free creatures. As a result, the free will defense would fail, and the problem of evil would re-emerge.
If the above reasoning is correct and God cannot guarantee eternal happiness, what consequences does this have for traditional theism? First, if no guarantee of eternal happiness is possible then one's motivation to adopt a particular theistic system might change. For Pascal, the chance at eternal happiness in heaven outweighs any temporary happiness we might gain by rejecting God on earth. The reasonable person, then, should wager in favor of belief in God. However, even if God exists and a heavenly realm awaits believers, the absence of a guarantee of eternal happiness makes a wager in favor of belief in God significantly less reasonable. There are further implications for religious proselytizers, for, the absence of a guarantee would raise ethical questions about those who advertise eternal happiness assurance as an incentive for conversion.
Second, the eschatological problem of free will also has consequences for the free will defense. If the reasoning of the eschatological problem of free will holds, then it can be seen as the unfortunate logical extension of the free will defense itself. This might incline one to reject the free will defense and, more significantly, the contention that humans have free wills. If, on the other hand, an effective solution can be found to the eschatological problem of free will, then that same solution is likely to undermine the free will defense, as seen in several of the proposed solutions above. In either case, the eschatological problem of free will constitutes an attack on the free will defense.
1. Even this nominal assurance offered through foreknowledge has problems. First, this is only an assurance if in fact such a guarantee has already been given by God. If no guarantee has been given, then no conclusion can be drawn about the privileged information God would have through foreknowledge. From the standpoint of natural theology, there is no reason to believe such a guarantee has been given. From the stand point of revealed religion, given problems of selecting and interpreting revealed texts, it may also not be clear whether a guarantee that appears in a sacred text is in fact a guarantee from God. Instead, apparent guarantees may simply reflect the desires of the authors of the texts. A second problem with this nominal guarantee is that this raises a classic problem about the compatibility of free will and divine foreknowledge: if God knows beforehand what we will do, then it is beyond our power to do otherwise, hence we have no freedom.
2. James F. Sennett, "Is There Freedom in Heaven?", delivered at the American Philosophical Association eastern division meeting, December, 1992.
3. Consider also a situation where a master switch would control the free wills of all humans. When the switch is on, we would have free wills, and when it is off we would mechanically act on the basis of the character traits developed earlier. Suppose that a switch operator randomly threw the switch every few weeks, and that we were aware each time it was on or off. Questions would be raised about whether on-mode people could be held responsible for their off-mode actions and vice versa. We would most likely see ourselves as having two distinct identities, and feel that the switch operator was tampering with a fundamental aspect of who we are.
4. This message is seen again in the allegory of the fall of Adam, and in the crucifixion of Christ by people who were acquainted with him.