The greatest problem for theism (belief in God), as traditionally conceived, is thought to be the problem of evil or suffering. According to this argument, it is impossible for God, who is omnipotent, omniscient, and omnibenevolent [1], to create a world which contains evil, and because our world contains evil, it cannot have been created by God. This argument is traditionally formulated as follows:
Premise 1. God is omnipotent, omniscient, and omnibenevolent.
Premise 2. A being who is omnibenevolent would desire to create a world with no evil.
Premise 3. A being who is omniscient would know how to create a world with no evil.
Premise 4. A being who is omnipotent would be able to create a world with no evil.
Premise 5. God will always create a world with no evil [P1-P4].
Premise 6. Our world contains evil.
Conclusion. Our world was not created by God [P5-P6].
Solutions to the problem of evil - called “theodicies” - fall into two main categories. The first is called “Irenaean theodicy,” argues that the development of conscience requires the existence of evil, and that the good which results from this development is greater than the evil which causes it. This has since been developed further into what is now called “soul-making theodicy,” which argues that the formation and development of souls requires the existence of evil.
The second category is called “Augustinian theodicy,” which argues that the existence of free will requires the potential of evil, and that the existence of free will is a greater good than the evil which can potentially be caused by it. The modern form of this theodicy is Platinga’s free-will defense, which is probably the theodicy most commonly appealed to by theologians.
All of these theodicies essentially boil down to one basic premise: the idea that a world with evil ultimately results in a greater net good than a world without evil, because the resulting good is greater than the temporary evil, and therefore it is not inconsistent with an omnipotent, omniscient, and omnibenevolent God (if “omnibenevolence” is defined as “desiring the greatest possible good”). In this post, I will demonstrate that the only way for this to work is if all people will be saved and experience bliss in the afterlife. If our world was indeed created by an omnipotent, omniscient, and omnibenevolent God, universalism cannot but be true.
Abbreviations used
This post is going to get somewhat philosophical, so please try to bear with me. I’ll be using several abbreviations here as shorthand, so that the premises of my argument don’t become too long or confusing:
WECT: the set of all possible worlds in which eternal conscious torment is true; that is, in which (1) all of those who die in unbelief experience conscious suffering without end in the afterlife, and (2) at least one person dies in unbelief.
WAni: the set of all possible worlds in which annihilationism is true; that is, in which (1) all of those who die in unbelief are annihilated with no further conscious experience, and (2) at least one person dies in unbelief.
WNA: the set of all possible worlds in which there is no afterlife; that is, in which all people are annihilated with no further conscious experience at death.
WUni: the set of all possible worlds in which universalism is true; that is, in which either (1) all people experience bliss without end in the afterlife regardless of whether they died in unbelief, or (2) no person dies in unbelief.
Furthermore (although this probably goes without saying), I will also be using “God” as shorthand for a being who is omnipotent, omniscient, and omnibenevolent. Although there are many conceptions of God around the world which lack one or more of these factors, all three Abrahamic religions (Christianity, Judaism, and Islam) believe in this view of God, and this is how the problem of evil is traditionally formulated. Now, let’s get into the actual argument.
What kind of world would God create?
The argument for atheism from the problem of evil usually begins like this (as also described above):
P1 [2]. A being who is omnibenevolent would desire to create a world with no evil.
P2. A being who is omniscient would know how to create a world with no evil.
P3. A being who is omnipotent would be able to create a world with no evil.
C1. God will always create a world with no evil [P1-P3].
It is then argued that, because our world does indeed contain evil, it cannot have been created by God. Theodicies (solutions to the problem of evil) usually attack Premise 1, arguing that an omnibenevolent being would not necessarily desire to create a world with no evil, because a world which contains evil may ultimately result in a greater net good for every individual (either due to the “soul-making” process, or due to the existence of free will). Based on this, we can reformulate this argument like so:
P1. A being who is omnibenevolent would desire to create a world with the greatest possible net good for every individual.
P2. A being who is omniscient would know how to create a world with the greatest possible net good for every individual.
P3. A being who is omnipotent would be able to create a world with the greatest possible net good for every individual.
C1. God will always create a world with the greatest possible net good for every individual [P1-P3].
But what, exactly, would be “the greatest possible net good for every individual”? Because it is possible for God to create no world, in which case there would be net zero good for every individual (since they would not exist), the worst-case scenario is that the good and evil experienced by an individual perfectly balances out such that the net good experienced is zero. If, in a world, any single individual experiences more evil than good (net negative good), then God would simply not create that world, since a better world is possible for that individual (namely, no world) and God, being omnibenevolent, must always desire the greatest possible net good for every individual.
This can be summed up in the following logical argument:
P1. It is possible for God to create no world.
P2. If no world is created, the net good experienced by every individual is zero (because they would not exist).
P3. God will always create a world with the greatest possible net good for every individual [C1].
C2. In any world created by God, every individual will experience a net good of zero or greater.
Thus, from these arguments, we can see that - because of His omnipotence, omniscience, and omnibenevolence - it is not possible for God to create any world in which even one individual experiences more evil/suffering than good. If any individual in this world ultimately experiences more evil than good over the period of their existence, then this world cannot have been created by God, simply because it would not be among the set of worlds possible to be created by God.
Comparing infernalism, annihilationism, and universalism
Now that we have determined what a world created by an omnipotent, omniscient, and omnibenevolent God would look like, we can compare the set of worlds possible to be created by God with the three different theological positions (infernalism, annihilation, and universalism).
First of all, in the set of all possible worlds in which eternal conscious torment holds true (WECT), at least one person experiences suffering without end in the afterlife. The evil/suffering experienced by this individual (and any others who die in unbelief) would be infinite, and any good that this individual experienced during their lifetime would be negligible in comparison. Essentially, this person would experience a net good of negative infinity over the period of their existence.
P1. In WECT, at least one person’s experience of evil/suffering is infinite.
P2. In WECT, at least one person experiences net negative good (in fact, infinitely negative).
However, this conflicts with [C2]:
P3. In any world created by God, every individual will experience a net good of zero or greater [C2].
Because in all worlds within the set WECT, at least one individual experiences net negative good, it is impossible for God to create such a world. Creating no world at all would be more favorable to an omnibenevolent God, because in that case such an individual would experience net zero good. Thus, it is impossible for God to create a world in which eternal conscious torment occurs to one or more individuals. Or in other words:
C3. WECT is not within the set of worlds possible to be created by God.
The second position to be considered is annihilationism, the view that all of those who die in unbelief are annihilated with no further conscious existence. In the set of all possible worlds in which annihilationism holds true (WAni), at least one person is annihilated after dying, with no afterlife. Thus, in any world in which annihilationism holds true, the net good which at least one individual experiences is the same as the net good experienced during their lifetime, because they have no conscious experience afterward. This is the first premise of our next argument.
P1. In WAni, the net good which at least one individual experiences within this life is the extent of the net good which they experience fully (because they are annihilated afterward).
In such a world, it is only possible for every individual to experience net positive good over the period of their existence if all of those who are annihilated experience more good than evil during their lifetime.
P2. In WAni, every individual only experiences net positive good if all of those who are annihilated experience net positive good during their life.
Let us call the subset of worlds within WAni in which this occurs WAni-NPG, meaning “Annihilationism-Net Positive Good”). But because of [C2],
P3. In any world created by God, every individual will experience a net good of zero or greater [C2].
no worlds within the set WAni are possible to be created by God unless they are also within the subset WAni-NPG.
C4. Only the subset of worlds within WAni in which all of those who are annihilated experience net positive good within this life (WAni-NPG) is within the set of worlds possible to be created by God.
Although a world within WAni-NPG is theoretically possible, our world is clearly not within this set. In this subset of worlds, every person who dies in unbelief must also experience net positive good over their lifetime. However, this is clearly incompatible with our own observations of this world. For example, consider the six million Jewish victims of the Holocaust; few would doubt that the suffering which at least some of them experienced outweighed the good which they experienced throughout their life, and yet they died as non-Christians, and so would be annihilated if Christian annihilationism is true. This is a fairly extreme example, but there are certainly many other people who experienced more suffering than good throughout their lives and still died in unbelief. Thus, our world is not within the subset WAni-NPG.
Now, just for the sake of argument, let’s consider the set of all possible worlds in which there is no afterlife (WNA), i.e., in which the conscious experience of all individuals ends at death. This set of worlds can be described much the same as WAni, except for the fact that in WNA, all individuals are annihilated, rather than just some.
P1. In WNA, the net good which every individual experiences within this life is the extent of the net good which they experience fully (because they are annihilated afterward).
P2. In WNA, every individual only experiences net positive good if they also experience net positive good during their life.
Let us call the subset of worlds within WNA in which this occurs WNA-NPG, meaning “No Afterlife-Net Positive Good”).
P3. In any world created by God, every individual will experience a net good of zero or greater [C2].
C5. Only the subset of worlds within WNA in which every individual experiences net positive good within this life (WNA-NPG) is within the set of worlds possible to be created by God.
This suffers from the same problem as WAni-NPG; although such a world in which every individual experiences net positive good in this lifetime is technically possible, this does not occur in our world. No one would doubt that, sadly, there are at least some people in this world who experience greater evil/suffering than good throughout their life. Thus, although WNA-NPG is within the set of worlds possible to be created by God, our world is not within this set.
Finally, we can consider the set of all possible worlds in which universalism holds true (WUni). In WUni, every person experiences bliss in the afterlife without end. The infinite good which they ultimately experience outweighs any possible evil/suffering that they experienced during their lifetime, and so in any world in which universalism holds true, every individual will experience net positive good over the course of their existence. Thus, based on [C2], we can conclude that WUni is within the set of worlds possible to be created by God. This is summed up in the following logical argument:
P1. In WUni, every individual experiences net positive good.
P2. In any world created by God, every individual will experience a net good of zero or greater [C2].
C6. WUni is within the set of worlds possible to be created by God.
In summary, out of all of the theological options which we have considered (infernalism, annihilationism, universalism, and no afterlife), only three scenarios could have been created by a God who is omnipotent, omniscient, and omnibenevolent. These are WAni-NPG, WNA-NPG, and WUni. And since observation indicates that our world is not WAni-NPG or WNA-NPG - sadly, many of those who die (whether in unbelief or not) experience more suffering than good throughout their life - the only possible option is that our world is within the set WUni (that is, universalism holds true in our world).
P1. The only sets of worlds possible to be created by God, among the options considered, are WAni-NPG, WNA-NPG, and WUni [C3, C4, C5, C6].
P2. Our world is not within the subset WAni-NPG, nor WNA-NPG, based on simple observation.
C7. Our world is either within the set of worlds WUni, or is not created by God.
Or, in other words, the only way to solve the problem of evil in our world is if universalism is true. Our world can only have been created by a God who is omnipotent, omniscient, and omnibenevolent if every individual will be saved and experience bliss in the afterlife. Otherwise, at least one individual would be ultimately experience more suffering than good over the course of their existence, and creating no world at all would be more desirable to an omnibenevolent God than such a world.
Possible objections
As far as I can tell, there are only two possible philosophical objections to this argument:
First Objection. Perhaps the nature of God’s omnibenevolence is such that He does not desire the greatest good for every single individual, but the greatest net good overall.
This is possible, but would entail reframing the debate around the problem of evil, which has traditionally been about individual suffering. Furthermore, although such a definition of God’s omnibenevolence is possible in theory, it does not match up with God’s nature according to the Bible. Scripture tells us that God wills the best outcome (salvation) for every individual (1 Tim. 2:4, 2 Pet. 3:9), and that He loves every individual, even His own enemies (Matt. 5:43-48). Love works toward the greatest possible good of the individual, even if it includes temporary suffering (Lam. 3:31-33, Rom. 13:10, Heb. 12:5-11). Therefore, if we accept the Bible as a valid source of theological truth, we must conclude that God’s omnibenevolence extends to every individual person, rather than being some sort of cosmic karmic force, and so the above argument still holds true.
Moreover, even if it is true that God’s omnibenevolence only desires the greatest good overall, rather than the greatest good for every individual, this is still incompatible with the traditional Christian doctrine of eternal conscious torment. Based on statements like that of Matt. 7:13-14, Christians have traditionally believed that the majority of people will undergo “eternal conscious torment.” But if more than fifty percent of individuals undergo suffering without end, then the net good experienced throughout the universe as a whole must be negative - the good can never overtake the evil. An omnibenevolent God, even according to the modified definition of omnibenevolence, would never create such a world, because creating no world at all (in which the net good is zero) would be more desirable than such a world.
Second Objection. Perhaps the suffering experienced under eternal conscious torment is neutral or even desirable to an omnibenevolent God, because such suffering is primarily justice-oriented. In this case, it would still be possible for God to create a world in which eternal conscious torment holds true.
Again, this would require a fundamental re-framing of the problem of evil and of the definition of omnibenevolence. The “all-goodness” of a deity who desires the eternal, conscious suffering of any individual must be called into question. And furthermore, just as with the other objection, this ignores what the Bible itself says about God’s omnibenevolence; according to scripture, God “takes no pleasure in the death of the wicked” (Ezek. 33:11 cf. 18:23). The suffering of “the wicked”, then, is not of the sort which would be neutral or desirable to an omnibenevolent God.
But that’s a heresy/the Bible doesn’t teach it/universalism would make people want to sin!
While I don’t believe that any of these typical reactions to universalism are actually valid arguments anyway [3], the beauty of this argument is that it renders these other debates pointless. As far as I can see, it is only possible to solve the problem of evil if universalism is true. In order to deny universalism, one must also deny at least one of the three characteristic traits of God - omnipotence, omniscience, or omnibenevolence - or deny that God created this world altogether. Thus, if the Bible teaches against the salvation of all, as most Christians believe it does, then it does not present a self-consistent set of claims and should be rejected as a source of truth. There is no possible middle ground on this issue that I can see.
______________________________
[1] These terms are usually defined like so:
Omnipotence: being able to do everything which is logically possible
Omniscience: knowing every possible truth and falsehood
Omnibenevolence: desiring the greatest possible good in every circumstance
[2] For those who are unfamiliar with this philosophical shorthand, “P1” simply means “Premise 1,” and “C1” means “Conclusion 1.”
[3] See my other articles on this blog, many of which argue that the Bible actually does teach universalism.