Does God Know the Future? What Each Answer Costs
Does God Know the Future? What Each Answer Costs
Does God know the future? Yes, exhaustively and infallibly, and the live question in analytic theology is not whether He knows it but what each account of that knowledge has to pay for saying so. Open theism secures human freedom by denying that there is a settled future there to be known. Molinism secures providence with a class of truths about free creatures that nobody has managed to ground. The Many Beings Framework argues that neither bill is owed, because the argument that issues it carries a modal term across two frames that differ in kind.
Key takeaways
- The threat to freedom is an argument, not a mood. It runs in ten numbered steps, and the Stanford Encyclopedia of Philosophy reports a consensus that it is valid.
- Open theism buys freedom with foreknowledge. Pinnock, Sanders and Boyd hold that God knows every truth there is, and that undetermined future acts are not yet truths.
- Molinism buys providence with a grounding problem. Middle knowledge explains how God governs free creatures, and four centuries have not settled what makes those counterfactuals true.
- Frame-Indexed Classical Theism refuses the step that generates every bill. Where open theism gives up exhaustive knowledge and Molinism adds a moment inside God, the framework denies that what is settled in God's frame is necessary in ours.
Why does divine foreknowledge look like a threat to freedom?
Because of an argument that looks valid, not because of a vague unease. The case that divine foreknowledge destroys freedom has a name, theological fatalism, and it runs in steps a reader can check; the Stanford Encyclopedia of Philosophy sets them out in ten.
Take some act you think you will perform freely tomorrow. Yesterday God infallibly believed that you would perform it. Whatever happened in the past is now fixed, which the literature calls the Principle of the Necessity of the Past: nobody can now do anything about the fact that the milk was spilled. So it is now fixed that God held that belief yesterday.
Infallibility means that if He believed it, the act occurs. Necessity transfers across entailment, which is the Transfer of Necessity Principle. So it is now fixed that the act occurs. If it is fixed, you cannot do otherwise. And if you cannot do otherwise, by the Principle of Alternate Possibilities, you do not act freely.
Free — the complete book
Get Volume One free.
The complete book, not a sample. The fastest way to test whether the framework holds.
No spam. Unsubscribe in one click.
The Encyclopedia notes a consensus that the argument or something close to it is valid, which is why the debate has been a hunt for the false premise. Four premises are open to attack: that God infallibly believed it yesterday, that the past is necessary, that necessity transfers, and that freedom requires alternate possibilities. Nelson Pike is credited with putting the dilemma in the sharp modern form in 1965, and every position below is a choice about which of those four to reject.
Scripture gives the case its face. Jesus tells Peter what he will do before Peter does it:
"Truly I tell you," Jesus answered, "this very night, before the rooster crows, you will disown me three times." (Matthew 26:34, NIV)
Peter denies Christ that night. Whether he could have done anything else while standing in the courtyard is the question every position below must settle.
Volume One works this argument through from the beginning; it is free to download.
What does open theism say, and what does it give up?
Open theism says that the future is partly open, and that God knows it as such. He knows every truth there is to know, and on this view there are not yet truths about what undetermined creatures will freely do, so His omniscience is untouched by His not knowing them.
The position is best heard in the words of its own defenders. The Internet Encyclopedia of Philosophy states the thesis as the claim that because God loves us and desires that we freely reciprocate that love, He has made His knowledge of, and plans for, the future conditional upon our actions.
Open theists affirm, in the Encyclopedia's words, that God knows all the truths that can be known, and claim "there simply are not yet truths about what will occur in the 'open,' undetermined future." God "knows all the possible ways the world might go at any point in time, but He does not know the one way the world will go." The movement's landmark is The Openness of God: A Biblical Challenge to the Traditional Understanding of God (1994), by Clark Pinnock, Richard Rice, John Sanders, William Hasker and David Basinger.
John Sanders, in "An Introduction to Open Theism" (Reformed Review 60.2, 2007), calls his own account dynamic omniscience. God knows the past and present with exhaustive definite knowledge, and knows the future as partly definite, where He has resolved to act unilaterally, and partly indefinite, where free creatures have not yet acted. Providence here is general where the tradition makes it meticulous, and God's plan is a broad intention with the details left to be settled. The biblical case rests on the texts where God tests, waits and relents, Nineveh among them, where Jonah announces that the city will be overthrown and God does not bring the destruction He had threatened. Whether God changes His mind is a separate question with a treatment of its own.
Now the price, because a steelman that hides the cost is flattery. Exhaustive foreknowledge goes. Prophecy becomes projection from present conditions plus what God has resolved to do Himself, and texts of this kind carry an obvious strain:
"Remember the former things, those of long ago; I am God, and there is no other; I am God, and there is none like me. I make known the end from the beginning, from ancient times, what is still to come. I say, 'My purpose will stand, and I will do all that I please.'" (Isaiah 46:9–10, NIV)
The article on open theism works through its exegesis and its critics at length, and this piece builds forward from it.
The free volume walks the same ground at book length.
How does Molinism answer, and what does it cost?
By middle knowledge. Molinism holds that God knows, logically prior to His decision to create anything, what every possible free creature would freely do in every possible circumstance, and that He governs history by choosing which circumstances to actualise.
Named for the Jesuit Luis de Molina (1535–1600), whose Concordia appeared in 1588, the position arranges God's knowledge in three logical moments: natural knowledge of all necessary truths and possibilities, middle knowledge of counterfactuals of creaturely freedom, and free knowledge of the world He has decreed. The Internet Encyclopedia of Philosophy describes middle knowledge as prevolitional, meaning God has it without willing it. The counterfactuals take the form "If person S were in circumstances C, S would freely do X," and they stand prior to God's creative will, which uses them in deciding what to create. Thomas Flint's 1998 defence gave the position its strongest contemporary support, and William Lane Craig has defended it since.
Its strength is exactly what the Stanford Encyclopedia says it is. Unlike the other compatibilist answers, which show only that foreknowledge and freedom can coexist, Molinism "provides an account of how God knows the contingent future, along with a strong doctrine of divine providence." It also rejects the same premise this piece will reject, the transfer of necessity, which makes it a near neighbour of the position argued below.
The cost is the grounding objection, and it bites. What makes a counterfactual of freedom true before the creature exists? Gregory Boyd, arguing as an open theist, presses the trilemma hardest: God's will cannot ground it, since that is determinism under another name; the creature's later decision cannot ground it, since that decision did not exist an eternity ago; so the truths are, in his phrase, "caused by nothing and grounded in nothing," which he calls a type of eternal dualism, an eternal reality alongside God that God Himself did not create. Molinists answer that the counterfactuals are brute facts, about the worlds in which they hold or about the creatures they concern, and the exchange continues. The Molinism article gives the replies at length; what matters here is that four centuries have not closed it.
Volume One weighs Calvinism, Arminianism and Molinism against each other; get the free copy.
Is simple foreknowledge enough for providence?
Simple foreknowledge is divine foreknowledge with nothing behind it: God simply knows what will happen, with no middle knowledge to explain it and no decree causing it. Its critics reply that this buys nothing: knowledge that arrives with the actual world arrives too late to be used in shaping it.
The appeal of the view is easy to feel. It is the plainest reading of omniscience, it keeps God's knowing innocent of causing, and it adds no extra machinery. The instinct that events do not happen because God knows them is honoured directly.
The objection is equally real. The Stanford Encyclopedia records it as "one influential argument that open theists use against defenders of foreknowledge who do not also accept Molinism," namely that foreknowledge without middle knowledge is useless for divine providence. If God's knowledge of the actual world already includes everything He will do, He cannot consult it before deciding what to do. William Hasker made the charge and kept making it; the title of his 2009 paper in the Journal of the Evangelical Theological Society is the argument in six words: "Why Simple Foreknowledge Is Still Useless." David Hunt has defended the providential usefulness of foreknowledge without middle knowledge in a series of papers, and the argument is still live.
Can Ockham's soft facts rescue the past?
Ockhamism rejects the second premise. It holds that the necessity of the past does not cover the whole past, and that God's past beliefs, the divine foreknowledge at issue, belong to the part it does not cover.
The distinction is between hard and soft facts. A hard fact about the past is strictly about the past. A soft fact about the past, in the Encyclopedia's formulation, "is one that is in part about the future": that it was true yesterday that a man will post a letter tomorrow is a fact dated yesterday whose content reaches forward. William of Ockham called the necessity that attaches to the genuinely past necessity per accidens, accidental necessity, and Marilyn Adams revived the strategy in 1967. If God's forebeliefs are soft facts, they carry no accidental necessity, and the argument stops at its third step.
What has held the approach back is that nobody could draw the line. Adams's own criterion, that a fact counts as soft if it entails something about the future, proved too generous: on it no fact is a hard fact, since that a man sat yesterday entails that he will not sit tomorrow for the first time. The 1980s produced more refined analyses, from Freddoso, Hoffman and Rosenkrantz, Plantinga and others, and the refinements grew elaborate enough that the intuition they were meant to capture stopped being visible in them.
What did Boethius actually claim about eternity?
That God has no beliefs at a time at all. Boethius rejects the first premise rather than the second: it is a mistake to say that God believed anything yesterday, so what we call divine foreknowledge is not fore-knowledge in the first place.
His definition is the one the whole tradition quotes: "Eternity is the simultaneous and complete possession of infinite life" (Consolation of Philosophy, Book V, prose vi). All temporal events stand before the mind of God at once, in what the Stanford Encyclopedia calls an atemporal present in which God has a single complete grasp of all events in the entire span of time. Aquinas took the solution over and compared God's presence to each moment with the centre of a circle's presence to every point on the circumference. Eleonore Stump, Norman Kretzmann and Brian Leftow have carried the account into contemporary philosophy.
Boethius also supplied the distinction the rest of this piece depends on. "There are two kinds of necessities; one is simple: for instance, a necessary fact, 'all men are mortal'; the other is conditional; for instance, if you know that a man is walking, he must be walking." His example is worth the space: "take the sun rising and a man walking; while these operations are occurring, they cannot but occur: but the one was bound to occur before it did; the other was not so bound."
The cost is that timelessness does not obviously end the argument. Linda Zagzebski has shown that a structurally parallel argument can be run for timeless knowledge, where the key issue is not the necessity of the past but the necessity of the timeless realm: God timelessly knows the act; whatever is in the timeless realm is now-necessary; so the act is now-necessary. As she puts it, the timeless realm is as much out of our reach as the past. Others, from Pike and Wolterstorff to R. T. Mullins, press the different worry that timelessness sits badly with divine personhood.
Volume Three takes up time and eternity directly.
How does Frame-Indexed Classical Theism answer?
It keeps Boethius and refuses the transfer. God knows the future in the atemporal, complete apprehension Boethius described, and the inference from "settled in God's frame" to "necessary for the agent" is a category error.
Frame-Indexed Classical Theism (FICT) is a research program in analytic theology that resolves classical theological dilemmas by diagnosing them as artifacts of a single formal error. Where open theism removes the settled future, Molinism inserts a moment of knowledge before the decree, Ockhamism divides the past, and the Boethian eternity solution places God outside the sequence, the framework holds that all four answer a misasked question.
Its conclusions, without the derivation that belongs to the books, are three. Beings whose natures differ in kind have modes of apprehension that differ in kind, so God's frame, atemporal and simple, and ours, sequential and composite, are not two viewpoints on one standpoint. Predicates whose content involves time, possibility, agency or value carry their frame, and cross between frames analogically rather than univocally. The Many Beings Fallacy is the equivocation committed when such a term is carried across anyway.
"Potential" and "actual" are exactly such terms. In our frame the future is objectively potential: we are localised at a moment, later coordinates are not there to be read, and the act is open. In God's frame the same reality is objectively actual, because He does not wait for coordinates to arrive but sustains them all in one act. One reality, received under two modes, and the fatalist's argument needs the two to be one.
Take Peter back to the courtyard. He is the efficient cause of his denial, undetermined in his frame, able to keep silent. Because he denies Christ the proposition is true, and because God apprehends all truth in one eternal act, God knows it. That order, act to truth to knowledge, is close to the Encyclopedia's dependence solution.
A second relation runs the other way: God grounds that Peter and his freedom exist at all. One fixes which way the act goes, the other holds it in being; they never compete.
This is Boethius's conditional necessity, made precise. Given that God knows the act, the act occurs, which is necessity of the consequence. It never follows that the act is necessary in itself for the agent, which would be necessity of the consequent. The psalmist celebrates knowledge of this kind:
"You have searched me, Lord, and you know me. You know when I sit and when I rise; you perceive my thoughts from afar. You discern my going out and my lying down; you are familiar with all my ways. Before a word is on my tongue you, Lord, know it completely." (Psalm 139:1–4, NIV)
From afar, and before the word is on the tongue. This is a knower who never had to keep pace with us.
What the framework gives up is worth naming: the intuition that "possible," "open" and "now" mean one thing for every knower, and that intuition is not cheap. Anyone unwilling to part with it must pay one of the other prices instead, open theism's or Molinism's, and will find Frame-Indexed Classical Theism unavailable.
The free volume sets out the position step by step.
Where the rivals push back
Three replies deserve to be heard at full strength.
Zagzebski's is the sharpest. If the parallel argument runs in the timeless realm, indexing has bought nothing, because we can do no more about God's timeless knowing than about His past knowing. The framework grants the formulation and denies the same premise in it. That a state lies beyond our reach does not make what it entails necessary for us, and the reach of the timeless realm is precisely the question at issue, so it cannot be had as a free premise.
The open theist and the Molinist push back together on providence. If this is timeless foreknowledge under a new name, then Hasker's objection lands on it, and the framework has no more providential resource than simple foreknowledge does. The answer is that the objection pictures God knowing the world and then deciding what to do about it, which is a sequence inside God. In one eternal act, the knowing and the ordaining are the same act, and there is no moment at which the knowledge arrives too late.
The third reply is the one careful readers make first: that frame-indexing is a courteous way of declining to explain. The framework answers it by stating in advance which predicates carry their frame and which hold across frames unchanged, identity and non-contradiction and arithmetic among the latter, so a misapplication can be caught by anyone. It also names its own defeater here: if the permissive decree, God's decree to allow an act rather than to cause it, were shown to require middle knowledge after all, Molinism would be vindicated and this part of the argument would fall.
Volume One shows each step of that case.
How the positions compare
Does God know the future? Six answers stand, and each pays.
| Position | How God knows free acts | Gets right | Pays |
|---|---|---|---|
| Open theism | As possibilities only | Choices undetermined | Exhaustive foreknowledge |
| Molinism | Middle knowledge, pre-decree | Providence with freedom | Ungrounded counterfactuals |
| Simple foreknowledge | By knowing the actual world | Causes nothing | Providential usefulness |
| Ockhamism | As a soft fact | The past need not bind | An undrawable line |
| Boethian timelessness | In one atemporal present | No beliefs at a time | Zagzebski's parallel |
| Frame-Indexed Classical Theism | In one eternal act | Knowledge there, openness here | Modal univocity |
Frequently asked questions
If God knows the future, do I still have a choice?
Yes, and the reason is a distinction Boethius drew fifteen centuries ago. Given that God knows you will act, it is certain that you will act, which is conditional necessity. It does not follow that the act is necessary in itself, which would be simple necessity, and only simple necessity would remove your ability to do otherwise. Certainty is not compulsion, and a knower who does not cause what He knows takes nothing from the agent who does.
How does God know the future: by decree, or by seeing it?
Three answers compete. The Reformed answer is that God knows it because He decreed it. Simple foreknowledge says He knows it by knowing the actual world. Molinism says He knows it through middle knowledge of what each creature would freely do. Frame-Indexed Classical Theism gives a fourth: knowing and sustaining are not two acts in God but one, so He does not observe the timeline from a position along it and does not learn it from anywhere.
What is the difference between open theism and Molinism?
Both hold that creatures are free in the libertarian sense. They divide over the future. Molinism holds that God knows the future exhaustively, what every creature will do and would do in any circumstance. Open theism holds that future free acts are not yet there to be known, so God knows them as live possibilities. Open theists press the grounding objection against Molinism; Molinists reply that exhaustive foreknowledge is a price middle knowledge never has to pay.
Did God know Peter would deny Jesus?
Yes, and Jesus said so before the rooster crowed. What that settles is that the denial was certain; what it does not settle is that Peter was compelled. Nothing in the courtyard pushed him, no antecedent force removed the alternative, and knowledge is not a cause. Open theists read the prediction as God's reading of Peter's character under pressure; Molinists read it as middle knowledge of what Peter would do in that circumstance.
Is frame-indexing just a way of avoiding the question?
It would be, if it were applied wherever a dilemma appeared. The framework restricts it in advance: modal, temporal, agential and evaluative predicates carry their frame, and identity, non-contradiction and arithmetic do not, so an illegitimate use is detectable by anyone who checks. It also states what would break it. Show that God's permissive decree requires middle knowledge, and Molinism wins the argument outright; show that the future must be unsettled in itself, and open theism does. Start with Volume One if you want to test the steps.
Where to go from here
The five standing answers differ about what God knows. Frame-Indexed Classical Theism differs about something earlier: what "settled" can mean for a knower whose nature differs from ours in kind, and whether the word survives the journey between frames intact. Open theism and Ockhamism both assumed it does, and paid accordingly.
Volume One argues the free-will side of that case in full and is free to download. Volume Three carries the same reading into God and time, future contingents, tensed facts and Peter's denial, and Volume Four carries it into the texts where God appears to relent. The sovereignty of God article sets out how the major positions define the term they are all fighting over.
