Do I have a coin in my pocket? On the face of it, it would seem silly for you to either think that the answer is yes or that the answer is no. You have no evidence either way. And without evidence, you should neither believe that there is a coin in my pocket nor deny it. You should be agnostic. What the case of the coin teaches us is that we need a reason to deny something. Denial isn't the ‘default position.’ Scepticism is.
Now consider the case of God. A theist is someone who believes that there is a God. An atheist is someone who denies that there is a God. An agnostic is someone who does neither. The agnostic neither believes that God exists nor believes that God does not exist. The agnostic is simply that: agnostic on the issue. Now this doesn't mean that the agnostic doesn't care about the issue. She might care a great deal. An agnostic is not necessarily passive or bored or uninterested. An agnostic is simply uncommitted.
The case of God is like the case of the coin: without evidence one way or the other, we shouldn't believe that there is a God, nor should we believe that there isn't a God. Agnosticism—not atheism—is the default position. We need evidence for theism to be a reasonable option, and we need evidence for atheism to be a reasonable option. Theists come along and present proofs, arguments, and reasons that attempt to pull us off our agnosticism towards theism. Atheists come along and do the same thing in a different direction.
There are a number of things that atheists can leverage as evidence for the non-existence of God. By far the most famous case for the non-existence of God is what people usually call the problem of evil. The idea, very quickly, is that God's existence seems incompatible with the existence of evil, and since there is evil in the world, there must not be a God. Hence we have good evidence for atheism. Let me first say something about what I mean by ‘God’ and what I mean by ‘evil’.
The problem of evil is not a problem for any old divine being. Showing that there is evil in the world wouldn't, for example, be considered very strong evidence against the existence of Zeus. That's because Zeus isn't a very nice guy, and so it wouldn't be surprising that a world he created would include not very nice things. However, God is supposed to be a perfect being. This perfection extends both to his moral character and to his power. So, by ‘God’ I mean the perfectly good, perfectly powerful being described in the great monotheisms of the world. It's that being that is targeted by the problem of evil.
And by ‘evil’ I mean any feature of the world that is bad. Pain and suffering are easy examples. If -other things being equal -the world would have been better without a certain thing, then that thing is evil on this very broad definition. And when people offer arguments from evil, they may have many different features of evil in mind. For some people, the very existence of any evil whatsoever constitutes good evidence against theism. For others, it's not the existence of evil, per se, but the existence of a certain type of evil. For example, perhaps the existence of God is compatible with minor evils like scraped knees but not horrific evils like rape or genocide. And still others insist that it's neither the mere existence nor the type of evil but the distribution of the evil that is good evidence against God. For example, if only evil people suffered evil things, perhaps this would be compatible with theism, but the fact that the innocent suffer is evidence against theism. Finally, certain people might think that it's the amount of evil that generates the strongest evidence against the existence of God. From here on, I'll use ‘evil’ as a general term covering all of these various aspects.
With these clarifications in hand, consider a very simple version of an argument from evil for atheism:
-
1. If God exists, then there is no evil in the world.
-
2. There IS evil in the world.
-
3. Therefore, God does not exist.
This argument is valid, which means if both of the assumptions are true, then the conclusion follows as a matter of logic. It's not possible for both of the assumptions to be true and the conclusion to be false. So is this a good argument?
No. Philosophers largely agree that this simple version of the argument from evil is implausible. The problem lies in the first assumption. It's easiest to introduce the problem by way of analogy. Suppose a father takes his son to the local hospital for an inoculation. The injection causes the child both fear and pain -two evils on our broad definition. Would this occasion be evidence that the father is a bad person? Of course not. We think that even though the father deliberately caused his son fear and pain, these evils are justified because of the long-term good that the inoculation secured. In other words, the evil in question is not pointless. The evil is necessary for some end result, and the end result is good enough that it compensates for the badness of the evil.
Might God be in a similar position to the father? Might certain evils be necessary for some end result and the end result be good enough that it compensates for the badness of the evil? The answer is plausibly yes. Take a simple example: building character. There is something really good about building one's own character. Not just showing up on the scene with your character already complete, but actively pushing and stretching yourself to develop character traits of your choosing. But this requires both an opportunity to fail and some hardships along the way. If you are to develop courage, you need to experience at least some times of fear and distress to do so. And this seems one of many examples of goods whose existence logically requires evils or at least the possibility of evils.
Now you might object in the following way: ‘Look, God is supposed to be perfect in power. That means he can do anything. So even if it's impossible to get a certain good without a certain evil, God can do what is impossible. And so this version of the argument is sound after all.’
There is a long-standing debate among theists about whether it even makes sense to say that God can do impossible things. But fortunately, we don't have to settle this dispute here in order to see that this objection fails. Either God can do impossible things or he cannot. If he cannot, then the first assumption of this argument is mistaken for reasons just pointed out. If he can do the impossible, then the argument fails for a different reason. On this horn of the dilemma, it's no longer impossible for God to create a world filled with evils even despite the fact that he is perfectly good and perfectly powerful. In other words, the argument from evil assumes at the outset that certain things are impossible for a being like God. That's the whole point of the argument: it would be impossible to get a world like this if it were governed by God. But once we allow that God can do the impossible, all bets are off.
Fortunately, we can easily adjust the argument to make it stronger. Here's a more nuanced version that dodges the objection about some evils being necessary for certain goods:
-
1. If God exists, then there is no pointless evil in the world.
-
2. There IS pointless evil in the world.
-
3. Therefore, God does not exist.
This version of the argument avoids the weakness of the first. It's not just any evil that constitutes evidence against the existence of God. It's pointless evil that constitutes evidence against the existence of God. A pointless evil is an evil that is not necessary to secure some great compensating good or prevent some worse evil. This argument grants that a sample of the evils on earth might be necessary for some compensating good. But surely not all of them are. A great many of the evils that we experience on earth seem absolutely pointless. Examples of people who suffer such pointless evils are not hard to come by: kidnapped children, earthquake survivors, people with cancer, victims of the Holocaust. Surely there is no God who would allow all of that that. And since it seems that there are many evils of this sort, there probably really are at least some. And so the second assumption is true as well.
So what can the theist say by way of reply? Historically, there have been three main options. The first is to deny the second assumption. Philosophers like St. Augustine and Spinoza think that there is no evil in the world. This position seems untenable to most contemporary thinkers. I think this response to the argument is plausible only if you deny that there are moral facts. If you truly think that nothing is right or wrong, good or bad, then, of course, the second assumption of this argument is mistaken.
The second main option is to deny the first assumption on the grounds that all of the evils in our world, despite appearances to the contrary, actually serve a greater purpose. This approach is typically called offering a theodicy or an explanation of the evils on earth. The history of philosophy is littered with various theodicies. Evil has been said to be necessary for the existence of good, knowledge of the good, free will, moral responsibility, character-building, etc. I won't say anything here about this approach other than to note that if some theodicy or set of theodicies is correct, then there are no pointless evils in the world, and this argument has a false assumption.
The third main option is the sceptical option, and it's this move that I want to focus on for the remainder of the essay. The sceptical option says that we are in no epistemic position to make a call on whether or not the evils in our world are pointless. And if the sceptical option is reasonable, then we should withhold belief about assumption number two. So whereas the theodicist wants to deny assumption two, the skeptic merely wants to be agnostic about assumption two. When the skeptic is also a theist, this response to the argument from evil has come to be known as sceptical theism.
Sceptical theism has a long history in Western philosophy. The portion Jewish and Christian scripture that deals most closely with the problem of evil is the story of Job. Job was an upright man who lost everything: his family, his wealth, and his reputation. The solution to the problem—if there is one—is that Job isn't capable of fathoming the ways of God:
Can you find out the deep things of God? Can you find out the limit of the Almighty? It is higher than heaven—what can you do? Deeper than Sheol—what can you know? (Job 11:7–8)
This same sort of scepticism can be found among philosophers in the early modern era. For example, when Descartes wrestles with a species of the problem of evil, he reasons as follows:
As I reflect on these matters more attentively, it occurs to me first of all that it is no cause for surprise if I do not understand the reasons for some of God's actions; and there is no call to doubt his existence if I happen to find that there are other instances where I do not grasp why or how certain things were made by him. For since I now know that my own nature is very weak and limited, whereas the nature of God is immense, incomprehensible and infinite, I also know without more ado that he is capable of countless things whose causes are beyond my knowledge…there is considerable rashness in thinking myself capable of investigating the impenetrable purposes of God. (Meditations on First Philosophy, AT 55, Cottingham 1984 pp. 38–39)
John Locke echoes this move:
I think it a very good Argument, to say, the infinitely wise God hath made it so: And therefore it is best. But it seems to me a little too much Confidence of our own Wisdom, to say, I think it best, and therefore God hath made it so… (Essay, Book I, Chapter IV, §12, Nidditch 1975, pp. 90–1)
And of course there is Hume, who perhaps puts the sceptical portion of sceptical theism most cleverly of all:
The great source of our mistake in the subject of God, and of the unbounded ‘license to suppose’ that we allow ourselves, is that we silently think of ourselves as in the place of the supreme being, and conclude that he will always behave in the way that we would find reasonable and acceptable if we were in his situation. (Enquiry Concerning Human Understanding §11)
Adding to this historical importance is the fact that sceptical theism has seen something of a renaissance in the last 30 years or so. Contemporary philosophers are invoking scepticism as a response to arguments from evil and doing so in ever more sophisticated ways. On the other hand, many contemporary philosophers are also offering objections to this sort of sceptical response and trying to make clear the costs of scepticism. I will offer a survey of both sides of the debate.
What can be said for sceptical theism? What reason is there for thinking that it is true and that this is the proper response to arguments from evil? I will sketch four different reasons that have been offered by contemporary sceptical theists. First, some sceptical theists have offered analogies. Suppose you were watching a chess match between two world-class champions. Unless you are a chess master yourself, you likely won't understand most of the moves made by the champions. But it would be silly to reason as follows: ‘I don't see a reason for that particular move, therefore there is no good reason for that move.’ Similarly, if one of the champions sacrifices a chess piece, it would silly to respond as follows: ‘I don't see a compensating good that can come from that sacrifice. Therefore, there probably is no compensating good that can come from that sacrifice.’ The fact of the matter is that the chess master's grasp of the game is so far above ours that we are in no position to make these kinds of judgement calls.
But then, says the sceptical theist, such is our situation vis-à-vis God. Just as a novice can't comment on whether the sacrifice of a particular chess piece is pointless, so, too, the average human can't comment on whether a particular evil in our world is pointless. And this is a reason to be sceptical about the second assumption in the argument from evil.
Second, philosophers have introduced concepts like the so-called butterfly effect to bolster our scepticism about whether any given evil is pointless. The so-called butterfly effect is when a seemingly small event leads to a much more significant outcome. For example, the flutter of a butterfly's wings in North America might start an air movement that results in a snow storm in the UK. Our knowledge of the physical world has shown us that predicting consequences over the long-term is a notoriously difficult proposition. To take one example, consider the Non-Identity Problem, a philosophical problem first made prominent in the 1980s by the British philosopher Derek Parfit. The Non-Identity Problem is the problem of determining who will exist in the future given that future existence is contingent on contemporary events. To take a prominent example from the literature, whether Lady Churchill went to sleep on her back or her stomach after making love to her husband would affect whether Winston Churchill or someone else was born 9 months later, and this, in turn, would have surely affected the outcome of WWII which would have surely affected who would exist in the year 2015.
Given this complexity, some philosophers urge caution in assuming that any particular evil is pointless. Remember: if an evil was necessary for a compensating good down the road, then it's not a pointless evil. And since the long-term consequences of our actions and various states of affairs are so hard to predict, epistemic humility requires that we refrain from classifying current events as ultimately pointless.
Third, philosophers have offered what are called sensitivity constraints on evidence. The most famous comes from an American philosopher named Stephen Wykstra who argues for a condition on when the non-existence of evidence constitutes evidence of non-existence. Wykstra thinks that a lack of evidence for something is evidence that there really is no such thing only if it's also true that if there were evidence for it, we would likely be aware of it. For example, suppose I ask you if there are any elephants in this room. You could look around and - seeing no elephants - conclude that there are no elephants in the room. This makes sense because if there would have been an elephant, you would very likely see it. So your non-evidence counts as evidence. But suppose I ask you if there are any germs in this room. You could look around and -seeing no germs - conclude that there are no germs in the room. But this doesn't make sense, and the epistemic principle in question explains why. If there were germs in the room, your evidence would look exactly the same to you. So in this case, your non-evidence does not count as evidence.
Hopefully the connection to sceptical theism is now perfectly clear. The question is whether the goods that might compensate for the evils we experience are more like elephants or germs. Some philosophers have argued that they are more like germs. Even if they existed, we are unlikely to be aware of compensating goods in a wide range of cases, and so the fact that we can see no compensating good for an evil does not license the conclusion that there is no such good. Hence we have no reason to endorse the second assumption of the argument from evil.
Fourth, and finally, philosophers like the American Philosopher Michael Bergmann, have invoked inductive scepticism in defense of sceptical theism. The basic idea is similar to the last point about sensitivity. It is reasonable to make an inductive inference from what we know to what we don't only when we have good reason for thinking that our sample is representative. For example, if I tell you that all thirty of my water samples of the Thames River are polluted, should you conclude that it is likely that the Thames River is polluted? Well, no. You'll want to know where my samples were taken from. If all thirty came from the same exact location just downstream of London, then you shouldn't make any broad inferences on the basis of that data. However, if I tell you that the thirty samples came from the entire length of the river at randomly assigned locations, then this seems like pretty good evidence that the Thames River is polluted.
So how does this connect to sceptical theism? Consider the range of goods you are familiar with, the range of evils you are familiar with, and the range of connections between goods and evils that you are familiar with. Are you sure that your sample is representative in all three cases? Some philosophers think that you should doubt this. Maybe there are a lot more types of goods, evils or connections between the two than we can fathom. And unless we have a reason to think that our sample is representative, we shouldn't make any inductive inferences about whether the evils in our world are truly pointless.
Needless to say, sceptical theism has its detractors, both inside and outside of theism. What can be said against sceptical theism? What reason is there for thinking that it is false or an improper response to arguments from evil? I will sketch three different objections that contemporary philosophers have offered to sceptical theism.
First, some philosophers are concerned that sceptical theism seems to entail a view called ethical consequentialism. Ethical consequentialism is the view that right and wrong are solely functions of the consequences of actions. If an action has good enough consequences, then it will be the right thing to do. In slogan form: the ends justify the means.
Suppose someone says that the death of a child in a car accident is a pointless evil. The sceptical theist responds by pointing out that -for all we know -this death was necessary to secure some great good or stave off some really bad evil. But this seems to commit the sceptical theist to the view that anything is, in principle, morally permissible so long as it produces enough good in the end. And that sounds like ethical consequentialism.
Now whether this implication is a bad thing depends on one's view of ethics. In point of fact, a great many professional philosophers are utilitarians which means that they endorse a species of ethical consequentialism. But other philosophers have been convinced that there are absolute moral prohibitions -things that are wrong no matter the consequences. If there are absolute moral prohibitions, then a new form of the argument from evil could be constructed showing that at least some of the evils that occur are violations of these absolute moral prohibitions. If that could be done, sceptical theism would be impotent as a response.
Second, sceptical theism might backfire on theists. Remember, sceptical theists are sceptical about our ability to determine whether any apparently pointless evil is genuinely pointless. God is so far above our level, that we're unable to grasp his ways. But then won't this scepticism ‘bleed over’ into the life of a theist? Take one of the most famous arguments for theism, the argument from design. The idea is roughly that our world is so perfectly designed that it must have had a powerful and provident creator. But, to take the sceptical quote from Hume seriously, when we imagine what the world would look like if it were created by a powerful and provident creator, aren't we silently thinking of ourselves as in the place of the supreme being, and concluding that he will always behave in the way that we would find reasonable and acceptable if we were in his situation? Maybe we should be agnostic about what a good world would look like, and if so, we should be agnostic about a crucial assumption in the argument from design.
Or to take a more practical example, how do we know that any of the world's scriptures are authentic or any religious experience veridical? After all, if the ends justify the means, then how are we to trust any putative divine revelation? If we say that it is an absolute prohibition that God deceive us (as Descartes held), then the earlier objection to sceptical theism applies. On the other hand, if lying is morally permissible given good enough consequences, then sceptical theism seems to undercut the basis for our trust in God's communications with humans.
Third, and finally, taking sceptical theism seriously seems to impose a kind of moral paralysis on the part of the skeptic. An example makes this clear. Suppose you are hiking in the forest and come upon a child drowning in a pond. Should you save her? You might think the initial answer is a resounding yes: you have a moral obligation to aid the drowning child. But suppose your sceptical theism kicks in. It's possible that there is a good reason to let her drown. After all, for all you know, this child could be the next Hitler! Now that may sound far-fetched, but remember: God's ways are not our ways, and compensating goods and evils are hard for us to see. So the mere fact that we can see no good reason to let the child drown doesn't make it likely that there is no such reason. Sceptical theists seem to be stuck: no matter how they reason about the drowning child case, there is no satisfactory explanation for what one ought to do. And that is a high price to pay to avoid the argument from evil for atheism.
Of course, sceptical theists have offered responses to each of these objections. But surveying these responses is beyond the scope of the present project which has been to offer and motivate a clear form of an argument from evil, explain the sceptical response to the argument, and offer reasons both for and against taking this response seriously.
