The Foreknowledge Problem

God is omniscient. He is capable of knowing everything that can be known by a maximally great being. If everything includes what is going to happen in the future, then it follows that he can know the future.

If God knows the future, then he cannot be wrong. That’s because, unlike you and I, God has infallible beliefs. He cannot get things wrong. So, whatever he knows about the future must be correct. We cannot surprise him by doing something he doesn’t expect.

The problem of divine foreknowledge is produced by noticing that God has infallible beliefs about what you and I will do tomorrow. Indeed, God has infallible beliefs about what every human person will do in advance of them doing it. But wait. How can God not be wrong about what you and I will do tomorrow if we haven’t decided yet? Aren’t we supposed to be freely performing our actions?

If, in the past, God has infallible beliefs about me sitting here writing this post this morning, and he cannot be wrong, then how could I do otherwise? I can’t change those beliefs any more than I could change what I did yesterday. No one can change the past. If those beliefs cannot be changed and cannot be wrong, then it seems there isn’t any way for me to avoid writing this post. The same is true for any human action. But that means our actions are not freely performed, and we do not possess free will. But surely you and I do possess free will. I could have done something else this morning. Such reasoning leaves us with a dilemma: either we don’t possess free will or God does not know what humans will do in the future.

Of course, the problem depends on what one means by free will. As I explained in the last post, one common idea of human freedom is that you and I are free if and only if there is no antecedent condition that determines what we will do. We call such a view libertarian or incompatibilist free will.

The foreknowledge problem suggests God’s omniscience and libertarian human freedom are incompatible because it seems there must be some antecedent condition that determines what we will do. For how else would God know?

For those who embrace divine (or theistic) determinism, there is no difficulty. On this view, God knows what will occur because he determines what will occur. God has meticulous providence. Every event down to the last movement of each and every molecule is decreed by God in advance and given God’s decision and his omnipotence, every event he decrees will happen and could not happen otherwise.

On this view, God knows what we will do because he has decreed our actions. He decided in advance that I will write this post this morning. His knowledge of what I am doing isn’t based on somehow seeing I will do it, but on his knowledge of his own decisions about what I will do.

If God determines all our actions, in what sense could they be free? Many determinists hold to a view of human freedom called compatibilist free will. Human free will is compatible with determinism. For example, on classical compatibilism, human beings are free if and only if (i) an action is desired by the agent (ii) the agent is not coerced (against his or her desires). Hence, God can know the action in advance, and it can be freely performed.

Divine determinists may breathe a sigh of relief on this one, but they have their own problems which don’t plague the non-determinist. We will come to those problems in a later post.

However, if one wants to maintain both the view that God is omniscient and that human agents have libertarian free will, there are a number of options. In what follows, I will examine four prominent solutions. Somewhat confusingly, such views have the name, compatibilist views, since they suggest libertarian free will is compatible with divine foreknowledge. This kind of compatibilism is not to be confused with compatibilist free will (free will is compatible with determinism). On that topic compatibilists about divine foreknowledge and free will hold to incompatibilism about free will and determinism!

In what follows I will try to bring out what is so intractable about the problem by showing the weaknesses of the solutions.

Open Theism

The simplest solution is to deny that God has foreknowledge of the kind in question. For example, according to open theists, God has probabilistic knowledge of the future because he has comprehensive knowledge of the past and present. But he doesn’t have the kind of infallible knowledge of the future the problem of divine foreknowledge assumes. Indeed, not being able to know the future actions of free creatures isn’t a limitation on God’s knowledge since it is logically impossible to know what they will be. Hence, God retains omniscience, and we can possess libertarian free will.

Before offering a criticism of the open view, an important methodological point for theologians can be made. It should be noted that the view does in fact solve the problem. Redefining omniscience in such a way allows the proponent to hold to both the doctrine of divine omniscience and human libertarian free will. However, being able to solve a philosophical problem isn’t all we are interested in.

For any solution offered, it must not turn out to be inconsistent with what Scripture teaches. For Christian thinkers, what God tells us trumps what we might find to be most rationally defensible. That’s not to say we must jettison reason all together, but we may find we have to adopt a position that isn’t as rationally defensible as we’d like. But far better to adopt a solution that is consistent with Scripture but has other problems or tensions than to hold a view that is inconsistent with Scripture no matter how well it stands up to rational scrutiny.

Having said that, we can turn to the main theological objection to the open view. The problem is that God makes quite a number of predictions about world events that are contingent on the free choices of human beings. They aren’t merely probabilistic predictions. Rather they are guaranteed events. Since, according to the open view, God only has probabilistic knowledge of the future, his predications cannot be promises. They are fallible. Holding to the open view limits predictions to what God can do without involving any particular human activity. But many predictions are about what God does in response to human activity (see predictions in the books of Daniel and Zechariah for examples). Hence, the open view solves the problem but is inconsistent with the Bible’s teaching about future events. This is a good reason to reject it.

Boethius’ Solution

Boethius (480 – 524) was born in Rome in 480 AD. During his life, he translated Aristotle’s works from Greek to Latin, served under King Theodoric, but fell out of favor and was imprisoned. From prison he wrote The Consolation of Philosophy, in which he offers a solution to the problem of divine foreknowledge.

Boethius’ answer was to say that God exists ‘outside’ time. Although we experience temporal succession—one thing happening after another—God doesn’t. To him, all things are in an eternal present. He sees all things happening at once: “Eternity is the whole and perfect possession, all at once, of endless life” (The Consolation of Philosophy, 5.6).

To grasp the kind of perspective God has on this view, imagine taking an old film strip, the kind made out of celluloid, and laying it out flat. Now imagine standing at a very high point and being able to see every frame of the movie all at once. That kind of view of the movie is analogous to God’s view of the world. He sees all its events at once. Whereas we experience one moment and then another just as we watch the movie one frame after another, God can see everything that happens all at once like watching the whole movie in one glance.

On this view, God can know the actions of his creatures. Hence, the view amply preserves the doctrine of divine omniscience. But can our actions be free in the libertarian sense? Unfortunately, it seems not. Merely changing God’s perspective from seeing into the future to seeing everything in an eternal present doesn’t evade the problem. Recall that the problem is that God infallibly believes what we will do in the future. Hence, we must not be able to do otherwise without making some of God’s beliefs false (i.e. God does not have omniscience). The problem is that God’s beliefs are fixed prior to the action. Once they are fixed, there is nothing anyone can do to change them.

The Boethian view merely adjusts the time at which God’s beliefs are fixed. Instead of them being fixed before we act, they are fixed timelessly. But just because there is a difference in when God’s beliefs are fixed, it doesn’t follow that there is any less of a problem. As Linda Zagzebski puts it, “If there is no use crying over spilt milk, there is no use in crying over timelessly spilt milk either” (Linda Zagzebski, “Recent Work on Divine Foreknowledge and Free Will,” p. 52). Hence, the Boethian view does not offer a solution to the problem.

Simple Foreknowledge Solution

According to the simple foreknowledge solution, God simply knows the future in virtue of being able to perceive the future. Just as you and I can look out of the window and see what’s happening, God can simply look into the future and see what will happen. If this is how it works, then God is merely observing our freely performed actions. They can both be free (in the libertarian sense), and God can know them.

There are a few problems with the simple foreknowledge view, the most intractable of which suggests that if it is true, then God cannot deliberate over what he does in the future. Let me explain.

The simple foreknowledge view relies on a sequence. First, God decides to create a world. He cannot see what actions free creatures will perform in advance of actually creating those creatures. But once he creates the world, he can see into the future. So, he creates a world and, as a consequence, knows everything that will happen in the future.

But now consider something we take to be true about God. God deliberates over what he will do in response to what human beings do. For example, we sin, and God decides to save us through the sacrificial death of Jesus Christ, the Son of God. To decide to do something requires being able to refrain from doing it. God couldn’t have been forced to save us. He didn’t have to do it. He does it out of his great love and it is an act of grace.

But now consider whether God can deliberate over his own actions in response to ours. On the simple foreknowledge view it appears not. For God knows what will happen in the future only once he creates the world. Once he does so, he perceives what we will do, but he also perceives what he will do as well. Just as he discovers our free actions, he discovers his own future actions. But if God cannot deliberate about his action in advance, then he cannot decide to intervene in human affairs. He merely finds that he will do so as a result of creating the world. But such a view of God is highly implausible. Hence, the simple foreknowledge view is implausible.

The other notable problem with the simple view is that once God sees what we will do in the future, he has those infallible beliefs. Since he has those beliefs prior to the actions they are about, the problem remains unsolved. Hence, the simple view also fails to solve the problem.

Molinist Solution

Suppose, then, that God does have a way of deliberating over what happens in the future without determining our actions. How might that work?

One view put forward by a Jesuit monk, Luis de Molina, suggests that God has three kinds of knowledge. First, he possesses natural knowledge, knowledge of all necessary and possible truths. He knows what can’t be false, such as two plus two equaling four, and what could be true, such as possibly, Ben Holloway being an astronaut.

The second kind of knowledge God possesses is his free knowledge. According to his free knowledge, he knows what will happen in the actual world. He knows, for example, that Ben Holloway is not an astronaut, but instead teaches philosophy.

Importantly for the solution, God also possesses middle knowledge, according to which God has knowledge of what any possible free creature would do in any given situation in every possible world in which those free creatures exist. More specifically, he knows what you and I would do in every possible situation. Since he can know what we would do in any possible situation, he can know our future actions prior to creating the world.

So, the logical sequence is different to the simple foreknowledge view. First, God surveys all possible worlds. Knowing what we would do in any of them, he chooses one possible world that pleases him and creates it.

As far as I can tell at this present time, Molinism is the most favourable view for a proponent of libertarian free will to hold. There are many eminent defenders, and it seems to offer a solution which can hold both divine omniscience and libertarian free will together.

However, it is not without difficulty. One difficulty has to do with how God can know what we would do in any situation. Supposedly, he knows us well enough to know how we will behave in any situation. For example, William Lane Craig writes,

Why could one not maintain that God, in virtue of his exhaustive knowledge of every single possible person, so that he discerns even what we would freely do in any circumstance in which he might place us, knows which counterfactuals concerning human freedom are true? (William Lane Craig, The Only Wise God, 145).

By analogy, Craig offers an example of the kind of knowledge he has of his wife. He knows his wife so well that he knows what choices she would make if she were under certain circumstances. Surely God knows us well enough to do the same with much more certainty.

The question is not solved, however, because we want to know what Craig knows about his wife that gives him reason to think he is correct in his prediction of what she would do in certain circumstances. Those reasons will include past experiences, observations of what she does usually, and facts about the circumstances. Crucially, anything he knows about his wife’s actions will come down to knowledge of the antecedent conditions that bring about her choice.

However, as John Feinberg argues, if one replies by saying that God knows what we will do because he knows all the conditions that will cause our actions, that is just to affirm determinism. After all, if he knows what I will do because he knows all the antecedent conditions such that, given those conditions, I must perform those actions, then we have determinism. On the other hand, if one replies that he ‘just knows’ what we would do, that isn’t an answer. It doesn’t explain how God knows.

Conclusion

For many, libertarian free will is a given. If they are also committed to the biblical doctrine of divine omniscience and not prepared to embrace open theism, they believe that there must be a workable solution. I have not covered all the solutions offered but tried to show how they work and what must be done to achieve one. I have also tried to bring out the difficulty of the problem. In my own view, the problem is not solvable. Nor can I square the open view with Scripture. Hence, I hold to a form of divine determinism and free will that is compatible with God determining all human actions. That isn’t the only reason I hold that view, but it is a contributing factor.

Further Reading

For an excellent introduction to the subject, I recommend Divine Foreknowledge: Four Views edited by James Beilby and Paul Eddy. Many of the criticisms of libertarian solutions are presented in No One Like Him by John Feinberg. I find his question “how would God know that?” the most difficult to solve for Molinists. For comments about God’s beliefs being fixed and, hence, unchangeable, see Linda Zagzebski, “Recent Work on Divine Foreknowledge and Free Will,” in The Oxford Handbook of Free Will (Oxford University Press, 2005).