# Theological Fatalism Objection Defeater ## Intro The objection: God knew a thousand years ago what you would do at 3:14 today. Nobody can change the past. So your 3:14 was settled before you existed, and you are not free. This is the serious form of the foreknowledge objection, and it deserves better than the answer it usually gets. It does **not** claim that God's knowing causes your action. Whoever presses it well will happily grant that knowing is not causing, because the argument does not need it. What it needs is that God's ancient belief is a fact of the past, that the past is unpreventable, and that unpreventability passes down the entailment chain to your afternoon. The defeat has three layers. The argument is valid, so one premise has to go, and there are three independently motivated candidates. The transfer principle it rests on fails wherever the later fact explains the earlier one, which is exactly the case here. And most decisively, God does no work in the argument at all: swap His belief for the plain prior truth of the proposition and the conclusion still follows, which means the objector has not produced an argument against God but a general puzzle about time and truth that they own as much as anyone. ## Cheatsheet **30-second reply.** The argument never says God's knowing causes anything, and I am not going to answer it that way. It says the past is fixed and passes its fixity down to whatever it entails. But fixity does not travel that direction. God believed it a thousand years ago *because* of what you do today; the belief depends on the act, not the act on the belief. And notice God is not doing any work here. Drop Him and run the same argument on the mere fact that "you will do it" was already true a thousand years ago. Same conclusion. That is Aristotle's sea battle, and it has nothing to do with theism. So either the reasoning kills free will for atheists too, or it is broken. Either way it is not an argument against God. **Fast facts.** - Canonical modern statement: Nelson Pike, 1965. Not a new objection and not a fringe one. - The engine is **accidental necessity**, the fixity a fact acquires simply by being past. Not logical necessity, and not causation. - The godless twin is **logical fatalism**, Aristotle's sea battle, *De Interpretatione* 9, roughly 350 BC. - [Acts 2:23](/codex/acts-2-23/) asserts foreknowledge and culpability in one sentence, in the same clause structure, without apology. - Six standard replies, each with live defenders: eternalism, Ockhamism, Molinism, open theism, compatibilism, dependence. **Counter-moves.** 1. **Refuse the causal frame in both directions.** Do not "defend" against a causal claim they did not make; it signals you are answering a weaker argument. 2. **Name the load-bearing premise out loud** (transfer of necessity) and make them defend it, rather than trading intuitions about the whole. 3. **Run the reductio early.** The no-God version of the argument is the fastest way to show the objection is not about God. 4. **Offer a primary reply, keep the rest as fallbacks.** Cycling between eternalism and Ockhamism under pressure looks evasive; pick one and hold the others as a cumulative posture you name in advance. 5. **Make them state what freedom they require.** If they require the ability to do otherwise, the argument is live and the replies above apply. If they will accept the compatibilist account, the objection dissolves immediately and the debate is a different one. **Concessions to grant.** - The argument is **valid**. Say so. Denying validity is a losing position and costs credibility. - The scope-fallacy reply, on its own, is **not sufficient** against the careful version. Grant that too. - The eternalist reply carries **real metaphysical commitments** about time and is not free. - The hard-and-soft-fact distinction is **genuinely hard to formalize**. Fischer and Adams landed real hits there. - Nobody has an answer that all Christian philosophers accept. This is an open question inside the household, not a settled one. **Closing line.** You have shown me that if a fact about the future was ever settled, and settledness travels backward from entailment, freedom is gone. That is true with God and true without Him. So the thing you have found is not a problem with God. It is a problem with your transfer principle, and it was waiting for you at home. ## In full Theological fatalism argues that essential divine omniscience is logically incompatible with libertarian freedom. The canonical form (Pike 1965) proceeds from four claims: God is infallible; God held a specific belief at a past time about a present act; past facts are accidentally necessary, that is, unpreventable now; and accidental necessity transfers across entailment. Since infallible belief entails truth, the act inherits the unpreventability of the past belief, and the agent cannot do otherwise. This defeater does not contest validity. It contests three premises with independent motivation, argues that the transfer principle fails for dependence-ordered pairs, and closes with a reductio: the argument's structure survives the deletion of God, generating logical fatalism, which almost no one accepts and which nobody treats as evidence against theism. The full survey of the six response families, with the pressure each faces, is at [Theological Fatalism](/codex/theological-fatalism/). ## Argument structure | # | Premise | |---|---| | **P1** | The objection's serious form makes no causal claim, so answering it with the knowing-is-not-causing distinction misses it entirely. | | **P2** | The argument is valid, so the dispute is over premises; three of its premises are independently contestable, and each has a developed defense. | | **P3** | The necessity of the past does not reach God's beliefs, either because God holds no beliefs *at* times (eternalism) or because such facts are soft (Ockhamism). | | **P4** | The transfer of necessity fails where the earlier fact depends on the later one, which is precisely the relation between God's belief and the act. | | **P5** | The argument proves too much: with God deleted, the same structure yields logical fatalism, so it is not an argument against theism. | | **C** | **Divine foreknowledge does not defeat human freedom, and the fatalist argument does not constitute evidence against God.** | ## Form Defensive, in five moves. P1 is framing and prevents the most common own-goal. P2 establishes where the debate actually is. P3 and P4 are the substantive defeaters, offered disjunctively (any one suffices) and cumulatively (they reinforce). P5 is a reductio on the argument's own structure and is the only move that transfers the burden completely, since it shows the objector holds the same puzzle whether or not God exists. The conclusion is modest by design: it defeats the objection without claiming to settle which of the six accounts of foreknowledge is correct. --- ## P1, The serious form makes no causal claim ### Affirmative case 1. **Pike says so explicitly.** The 1965 paper is built to avoid the causal charge, precisely because Boethius and Aquinas had already answered it. The argument's premises mention belief, entailment, and pastness. No premise mentions causation, production, or explanation. 2. **The argument runs on a fact about time, not about power.** Its work is done by accidental necessity, the fixity a fact acquires simply by being past. That fixity is not a causal notion; it is the observation that no agent's later action is a candidate for making an earlier fact not have obtained. 3. **The causal reply is available and cheap, which is the tell.** If the objection reduced to "knowing causes", it would have died in the sixth century with [Boethius](/codex/boethius/). Its survival in professional philosophy for sixty years is evidence that something else is carrying it. ### Anticipated objections 1. *"Then why do all the popular versions say God's knowing locks it in?"* 2. *"You are being pedantic. Fixed is fixed, whatever the mechanism."* ### Rebuttals 1. Because most popular versions are the weak one, and the weak one really is answered by [Foreknowledge vs Causation](/codex/foreknowledge-vs-causation/) and defeated at [Foreknowledge Is Causation Objection Defeater](/codex/foreknowledge-is-causation-objection-defeater/). Sorting them is not pedantry; it decides which reply is honest. Answering the strong version with the weak version's answer is the single most common failure in this exchange, and an informed objector will call it. 2. Not pedantic, load-bearing. If the fixity were causal, the theist would owe an account of how God's belief exerts force on a future act, which would be hard. Since it is not causal, the theist owes an account of unpreventability instead, which is a claim about modality that turns out to be much more vulnerable. Naming the mechanism tells you where to push. --- ## P2, The argument is valid, so the fight is over premises ### Affirmative case 1. **Concede validity immediately.** From an infallible past belief, the necessity of the past, and unrestricted transfer, the conclusion follows. Denying this wastes the exchange and signals unfamiliarity. 2. **Three premises are contestable, not one.** The claim that God holds beliefs at times (denied by eternalism), the claim that His beliefs are hard facts about the past (denied by Ockhamism), and the transfer principle (denied by the dependence reply). Each has a serious literature. 3. **The theist needs only one.** The objection is a proof of incompatibility. A single defensible failure point defeats a proof, and there is no requirement to establish which reply is true. ### Anticipated objections 1. *"Having several answers means you have none. Pick one."* 2. *"Each of those replies has known problems, so a pile of damaged replies is not a defense."* ### Rebuttals 1. Multiple defeaters for a single proof is the normal state of a healthy debate, not a weakness. But the criticism has a fair core, so answer it honestly: name a primary reply and hold the others as fallbacks, rather than switching whenever one is pressed. Switching under fire is the vice; having options is not. 2. True of each and false of the set. The objection must show every premise secure; the theist must show one insecure. Even substantial damage to each individual reply leaves the proof unestablished, and a proof that fails to establish its premises establishes nothing. --- ## P3, The necessity of the past does not reach God's beliefs ### Affirmative case 1. **Eternalism: there is no past time at which God believed anything.** On the classical account God's knowledge is a single timeless act, "the whole, simultaneous and perfect possession of interminable life" ([Boethius](/codex/boethius/), *Consolation* V.6). Aquinas: one watching from a height sees the whole road at once, and the seeing compels no step (*Summa Theologiae* I q.14 a.13). The premise that fixes a divine belief a thousand years ago simply has no referent. 2. **Ockhamism: the belief is a soft fact.** Facts wholly about a past time are accidentally necessary. Facts about a past time whose truth depends on later times are not. "God believed at T1 that S does X at T2" is in the second class, because whether it is true turns on what happens at T2. So the fixity principle does not apply to it, and refraining is not a power to change the past. 3. **The distinction is not ad hoc.** Soft facts are everywhere and pre-theoretical: "in 1900, the twentieth century's worst war was still 39 years away" is a claim about 1900 nobody thinks was fixed in 1900. Ockham did not invent the category to save foreknowledge; he applied an existing one. ### Anticipated objections 1. *"Zagzebski's necessity of eternity. If God's timeless knowledge is unalterable, rerun the argument on eternity."* 2. *"Nobody can state the hard-and-soft distinction precisely."* (Adams; Fischer) 3. *"Eternalism requires a B-theory of time, which is contested and has theological costs of its own."* ### Rebuttals 1. Fair, and it is the best objection to reply 1. But the rerun needs its own transfer principle for eternal facts, and that principle is *less* plausible than the temporal one, because the intuitive pull of "you cannot change the past" comes precisely from pastness. Strip out pastness and the principle is left asserting bare unalterability, which is what the argument was supposed to prove rather than assume. It also runs straight into P4: the timeless knowledge depends on the act. 2. Granted, and it is why this defeater does not rest on Ockhamism alone. Note the shape of the concession, though: the difficulty is in *formalizing* the distinction, not in doubting that soft facts exist. An objection that says "your true distinction resists precise statement" does not license treating a soft fact as hard. 3. Partly true; the natural home is a B-theory or the Stump-Kretzmann eternal-temporal simultaneity account, and the costs are real (see [A-Theory vs B-Theory of Time and Divine Foreknowledge](/codex/a-theory-vs-b-theory-of-time-and-divine-foreknowledge/)). This is why eternalism is offered here as one option rather than the answer. An A-theorist should lead with P4. --- ## P4, Transfer fails where the earlier fact depends on the later ### Affirmative case 1. **Unpreventability is not closed under entailment across dependence.** The intuition behind "you cannot change the past" is that later actions are not candidates for making earlier facts obtain or not obtain. But that intuition presupposes independence. Where the earlier fact obtains *because of* the later one, the later one is exactly the sort of thing the earlier fact answers to. 2. **The dependence runs the right way here.** God believed at T1 that S does X because S does X. Nobody, including the fatalist, thinks the explanation runs the other direction; that would be the causal claim P1 already showed they disavow. So the pair is dependence-ordered, and the transfer step is applied to the one case where it is least warranted. 3. **What S has is a counterfactual power, and that is enough.** S cannot change God's past belief. S can act such that, had she acted otherwise, God would always have believed otherwise. Libertarian freedom requires the ability to do otherwise, not the ability to alter history. Plantinga's "On Ockham's Way Out" (1986) develops this; Merricks ("Truth and Freedom", 2007) presses the dependence form hardest. 4. **The principle overgenerates when tested elsewhere.** Any past fact that depends on a present free act inherits the same problem: that a prophecy was recorded, that a bet was placed, that a truth was true. If the transfer principle really applied, it would make all of them unpreventable, which is the reductio at P5. ### Anticipated objections 1. *"Counterfactual power is not real power. You have relabeled the problem."* 2. *"This is Ockhamism with new vocabulary, so it inherits Ockhamism's problems."* 3. *"Dependence does not matter. Fixed facts are fixed however they got that way."* ### Rebuttals 1. Then say what "real power" is. If it means the power to make the past not have been, no one has ever had it and no account of freedom has ever required it, so the objection proves that nothing is free rather than that foreknowledge is the problem. If it means the power such that things would have gone otherwise had one chosen otherwise, that is precisely what is on offer here. 2. They are cousins and they can be run together, but they are separable. Ockhamism classifies the belief-fact as soft and exempts it from the fixity principle. The dependence reply grants the classification argument entirely and attacks the transfer principle instead. An objection that defeats the hard-soft taxonomy leaves the dependence reply standing. 3. This is the crux, so make them defend it rather than assert it. The claim needs an argument, because the fixity intuition is generated by cases of independent past facts (Caesar, the Rubicon) and is then extended to a dependence-ordered case without further support. Extending an intuition past the conditions that generated it is exactly what needs justification here. --- ## P5, The argument proves too much ### Affirmative case 1. **Delete God, keep the structure.** Suppose that a thousand years ago the proposition *S does X at T2* was already true. Past truth is as past as any other past fact. Apply the fixity principle and unrestricted transfer, and it follows that S cannot do otherwise. No deity appears anywhere in the derivation. 2. **This is a known ancient result.** It is Aristotle's sea battle (*De Interpretatione* 9) and the medieval problem of future contingents. It is a problem about bivalence and time, and it long predates the Christian framing of it. 3. **So the objector must take one of two branches.** If the argument is sound, libertarian freedom is impossible with or without God, and the objection is no longer about theism. If it is unsound, because prior truth removes nobody's freedom, then the theological version is unsound too, since an infallible knower adds only a knower to a truth that was already there. 4. **Nothing in the argument uses divine attributes.** Omnipotence is unused, goodness is unused, and even omniscience is used only to license the entailment from belief to truth, which prior truth supplies for free. When an objection to theism does not use theism, it is not an objection to theism. ### Anticipated objections 1. *"I accept the fork. I am a determinist, and this is one more reason."* 2. *"There is a disanalogy: a bare truth is not a belief held by an agent."* 3. *"Fine, but you still have not told me how foreknowledge and freedom fit together."* ### Rebuttals 1. Then the objection is retired as an objection, and say so plainly. A determinist who runs theological fatalism is offering the theist a conclusion the determinist already held for unrelated reasons. The exchange should move to whether determinism is true, where the theist has arguments and where God's knowledge is no longer the topic. This is the best outcome available and it is worth naming as a win. 2. Correct, and it cuts the wrong way for the objector. Adding a believer *weakens* the fatalist's case, because a belief has a truthmaker and depends on what it is about, which activates P4. A free-floating truth is harder for the theist to handle, not easier. The disanalogy makes the godless version the stronger one. 3. Two different projects. Defeating an incompatibility proof does not require producing a positive model, and pretending otherwise inverts the burden. But the positive models exist and are worth naming: eternalism, Ockhamism, Molinism, compatibilism, and the dependence account, surveyed at [Theological Fatalism](/codex/theological-fatalism/). Offer one; do not accept the demand that you must. --- ## Live-cite kit **Scripture.** - [Acts 2:23](/codex/acts-2-23/), delivered over "by the predetermined plan and foreknowledge of God" and nailed to a cross "by the hands of godless men." Both halves, one sentence, no tension acknowledged. The single best verse for this exchange. - [Acts 4:27-28](/codex/acts-4-27-28/), the same structure applied to Herod, Pilate, and the nations. - [Luke 22:34](/codex/luke-22-34/), the denial foreknown to the hour, followed by real guilt and real restoration ([John 21:15-17](/codex/john-21-15-17/)). A foreknowledge that excused would not need a restoration scene. - [Isaiah 46:9-10](/codex/isaiah-46-9-10/), declaring the end from the beginning; the foreknowledge side stated at maximum strength. - [Deuteronomy 30:19](/codex/deuteronomy-30-19/), choose life; the responsibility side stated at maximum strength, by the same author who wrote the foreknowledge texts. **Scholarly.** - Nelson Pike, "Divine Omniscience and Voluntary Action", *Philosophical Review* 74 (1965). Cite the opponent's best source by name; it earns the room. - Linda Zagzebski, *The Dilemma of Freedom and Foreknowledge* (Oxford, 1991), and her Stanford Encyclopedia entry "Foreknowledge and Free Will", the standard neutral map. - Alvin Plantinga, "On Ockham's Way Out", *Faith and Philosophy* 3 (1986). - Trenton Merricks, "Truth and Freedom", *Philosophical Review* 116 (2007), the dependence reply in its sharpest form. - Eleonore Stump and Norman Kretzmann, "Eternity", *Journal of Philosophy* 78 (1981), the technical account of eternal-temporal simultaneity. - William Lane Craig, *The Only Wise God* (Baker, 1987), the accessible Molinist treatment. - Aristotle, *De Interpretatione* 9, for the reductio. **Aphorism.** - "God does not know the future because He causes it; He knows it because it is there to be known." - "You cannot change the past. You were never asked to. You were asked to act, and the past answers to what you do." - "Take God out of the argument and it still runs. Whatever it caught, it did not catch Him." ## Tactical notes - **Ask one question first: does the objection depend on God's knowing causing anything?** If yes, you are facing the popular version, go to [Foreknowledge Is Causation Objection Defeater](/codex/foreknowledge-is-causation-objection-defeater/) and finish quickly. If no, you are facing Pike and should say so, by name. Correctly identifying which argument you are in is worth more than any single rebuttal. - **Do not open with eternalism.** It is the most famous reply and the most metaphysically expensive, and a prepared objector has the Zagzebski counter ready. Open with the reductio at P5, which costs nothing and reframes the whole exchange, then take the dependence reply at P4 as your primary. - **Concede validity early and audibly.** It buys the credibility you need for the premise-level work, and objectors who expect a bad answer are disarmed by a good concession. - **Watch for the switch to determinism.** Many objectors are determinists who do not think foreknowledge is a special problem at all. The moment they accept the P5 fork, the topic has changed and the fatalism objection is spent. Name that out loud rather than continuing to defend against a retired argument. - **Do not claim the church has a settled answer.** It does not, and saying so is both true and disarming. What the church has is several accounts, any one of which breaks the proof. - **Refuse the demand for a mechanism.** "Explain exactly how God knows future free acts" is not a burden the objection creates. Defeating an incompatibility proof requires showing a premise unsupported, nothing more. ## See also - [Theological Fatalism](/codex/theological-fatalism/), the concept hub with the six response families and their pressures - [Foreknowledge Is Causation Objection Defeater](/codex/foreknowledge-is-causation-objection-defeater/), the defeater for the causal version of the objection - [Foreknowledge vs Causation](/codex/foreknowledge-vs-causation/), the epistemic-versus-causal distinction - [God Is Surprised Objection Defeater](/codex/god-is-surprised-objection-defeater/), the texts pressed against exhaustive foreknowledge - [Molinism](/codex/molinism/), middle knowledge as a positive model - [Counterfactuals of Freedom](/codex/counterfactuals-of-freedom/), possible worlds against feasible worlds - [Open Theism](/codex/open-theism/), the option that gives up exhaustive foreknowledge, and what it costs - [Compatibilism](/codex/compatibilism/), the reply that denies freedom requires alternative possibilities - [Libertarian Free Will](/codex/libertarian-free-will/), the freedom the objection targets - [A-Theory vs B-Theory of Time and Divine Foreknowledge](/codex/a-theory-vs-b-theory-of-time-and-divine-foreknowledge/), the metaphysics behind the eternalist reply - [Boethius](/codex/boethius/), the *Consolation* V source of the eternal-now reply - [Modal Logic](/codex/modal-logic/), necessity operators and scope - [Divine Providence](/codex/divine-providence/), the doctrine downstream - [Arguments](/codex/arguments/), the master index
## Common questions this page answers **Q: How do I answer someone who says God's foreknowledge means I have no free will?** First find out which argument they are making. If they think God's knowing *causes* the act, answer with the epistemic-versus-causal distinction and you are done. If they grant that knowing is not causing and argue instead from the fixity of the past, you are facing the serious version. For that one, the fastest reply is that the argument works with God removed: if it were sound, the mere prior truth of "you will do it" would abolish freedom too, which is Aristotle's sea battle. Then press the premise that unpreventability transfers from God's belief to your act, since His belief depends on your act rather than the reverse. **Q: Is the theological fatalism argument valid?** Yes, and you should concede that immediately. From an infallible past belief, the necessity of the past, and unrestricted transfer of necessity, the conclusion follows. The entire dispute is over whether those premises are true, and three of them are contestable with serious literature behind each. Denying validity is a losing position and costs you credibility with anyone who knows the argument. **Q: What is the strongest Christian answer to the foreknowledge dilemma?** There is no consensus, which is worth saying honestly. The two strongest current candidates are the dependence reply, that unpreventability does not transfer backward across a relation where the earlier fact depends on the later one, and the reductio, that the argument works without God and so is not an objection to theism. The classical replies of Boethian eternalism, Ockhamist soft facts, and Molinist middle knowledge remain live, each with real defenders and real critics. **Q: Doesn't the Bible settle this?** It settles the constraint, not the mechanism. Scripture asserts exhaustive foreknowledge and full human responsibility side by side, sometimes in a single sentence ([Acts 2:23](/codex/acts-2-23/)), without ever treating the pair as a difficulty. So any account that saves one half by discarding the other is out, which rules out both hard determinism and a foreknowledge-free God. Within that constraint, Scripture does not adjudicate between the philosophical models. **Q: If I could have done otherwise, would God have been wrong?** No. Had you done otherwise, God would have believed otherwise. His belief tracks what you do; it is not a prior commitment your action might falsify. The mistake is picturing God's belief as fixed first and your act as having to match it. The order of explanation runs the other way, which is exactly why the fatalist's transfer principle fails. **Q: Does open theism solve the problem?** It dissolves it, at a price most of the tradition judges too high. If God does not know future free acts, no past belief exists to be fixed and the argument never starts. But specific fulfilled prophecy, the certainty of God's promises about the end, and texts such as [Isaiah 46:9-10](/codex/isaiah-46-9-10/) and [Psalm 139:4](/codex/psalms-139-4/) all become difficult. The other five replies keep exhaustive foreknowledge and break the argument elsewhere, so open theism is not the only exit. See [Open Theism](/codex/open-theism/).