Justified True Belief

Mapping the Landscape of Good Reasons for the Truth of Christianity

Christian Evidences

Evidence & Arguments for Christian Theism

Divinity of Christ

Biblical Evidence for High-Christology: Jesus is God

TBD

Old Testament Criticism

Evidence for the Reliability of the OT Bible

Common Objections

Objection Analyses to Christian Theism

World Religions

Critical Analyses of Non-Christian Religions

Philosophy

Phileō Sophia - to Love Wisdom.

Metaphysics: Free Will

Whether Our Choices Are Up to Us

What Free Will Is, and Why It Matters

(P1) Some people deserve praise or blame for what they do. + Some deeds earn praise and some earn blame, and the person who did them has it coming. That is a claim about what he is owed, which is a different thing from what blaming him might accomplish. (1) We know this more surely than we know the premises of any argument against it. - Think of a man who spends a year befriending an elderly widow, learns where she keeps her savings, and takes every penny. You do not need a philosophy degree to know he has earned something, and neither does she. - Now suppose somebody wants to argue that he has earned nothing at all. An argument for that verdict has to be built out of premises, and those premises, taken together, have to be more credible than the verdict they are meant to overturn. Here are the ones usually offered. • The physical world is causally closed, meaning that nothing outside the chain of physical causes ever reaches in and makes a difference to it. A choice is then one more physical event, with physical causes standing behind it. • Every event, choices included, was necessitated by what came before it, which is to say made to happen by it, the way each domino in a row is knocked over by the one behind it. And nothing made to happen by things in place before a man was born can be his own doing. • The self is a story the brain tells itself, and there is nobody home behind the story to run up a debt. - Hold each of those up against the widow and her savings, and it is not close. Every one of them is a serious thesis defended by serious people, and not one of them is anywhere near as certain as the thing it is being used to overturn. - Rejecting the verdict on those grounds is itself a piece of reasoning, and it follows a rule about which direction an argument runs. When a chain of reasoning ends somewhere you know to be false, the sane move is to go looking for the bad link. Philosophers do this all the time in every other field. Zeno of Elea offered clever proofs that motion is impossible, and people have answered him ever since by hunting for the flaw rather than by standing still. (2) Deserving blame and finding blame useful are two different things. - Philosophers have a name for the kind of deserving (P1) is about, and it is worth learning, because the whole quarrel turns on it. They call it basic desert. • "Desert" here is just the noun for deserving, as in getting your just deserts, and has nothing to do with sand. • "Basic" means that it rests on nothing further. A person deserves blame just for having done the wrong thing knowing what it was, and not because blaming him will reform him or warn anybody else. • The term was sharpened, as it happens, by philosophers who deny that anyone ever has any. - Picture a judge who knows the man in the dock is innocent and sentences him anyway, because the town is frightened and a conviction would calm everyone down. The sentence might well work. Everyone can see it would still be a crime, and the reason is that punishment has to be deserved before it can be useful. - Blame that is only a lever for changing behavior has stopped being blame. It has become management, which is how we handle a thermostat or a herd of cattle, and nobody thinks a thermostat has anything coming. (3) Our ordinary dealings with one another already assume desert. - Somebody steps on your hand on a crowded bus. Before you say a word you look up to see whether he meant it, and everything that happens next depends on the answer, even though the pain is exactly the same either way. - Resentment, gratitude, indignation, and forgiveness are responses to what a person did, and they collapse into nonsense if he did not do it. Nobody adopted them as policies because they produce good results. Nobody works out in advance that indignation will improve the neighborhood and then decides to feel it. • As it happens, the argument that did most to turn the field on this point in the twentieth century came from an unexpected quarter. You might expect it from people who think some choices are left unsettled. It came instead from the side that holds freedom survives even in a world where everything was settled in advance, so it cannot be waved away as special pleading by people with a stake in unsettled choices. - The sharpest test of all is one everybody has run, and it is an apology. If blame were only a tool for adjusting future behavior, an apology would be beside the point, since nothing about the past has changed and the lesson has already been learned. Yet nobody treats it as beside the point. We wait for it, we notice when it never comes, and when it does come something has been put right, because what an apology answers is a debt. (4) Scripture promises a judgment measured by what each person did. - The New Testament says so again and again, in plain words. • "He will render to each one according to his works" (Romans 2:6). • "I tell you, on the day of judgment people will give account for every careless word they speak" (Matthew 12:36). • "So then each of us will give an account of himself to God" (Romans 14:12). - Paul packs the whole premise into three words without meaning to. We must all appear before the judgment seat of Christ, he writes, "so that each one may receive what is due for what he has done in the body, whether good or evil" (2 Corinthians 5:10). He does not say what would be useful, or what would do the most good afterwards. His three words are "what is due." - Ezekiel has God saying he will judge Israel "every one according to his ways" (Ezekiel 18:30), and the prophets keep saying it because every warning they deliver stands on it. A warning shouted at somebody who can never be held to account is just theater. - In fairness, Scripture nowhere lays out a theory of the will. It speaks to its hearers as people who will answer for themselves, it confesses God as sovereign over everything, and it leaves the work of fitting those two together to us. • The famous commands, "choose life" (Deuteronomy 30:19) and "choose this day whom you will serve" (Joshua 24:15), are not by themselves evidence for one theory of the will over another, for the plain reason that every side in this dispute agrees that we choose. • So (P1) does not rest on the word "choose." It rests on the judgment. (5) The church said all of this before the argument had a name. - Justin Martyr, writing to a Roman emperor in the middle of the second century, put it as flatly as anyone has since. Punishments and rewards "are rendered according to the merit of each man's actions," and if instead all things happen by fate, "neither is anything at all in our own power" (First Apology 43). - Irenaeus, a generation later, saw the same thing from the other end. If some people had been made bad by nature and others good, "these latter would not be deserving of praise for being good, for such were they created; nor would the former be reprehensible, for thus they were made" (Against Heresies IV.37.2). - Justin also left the image worth keeping. God did not make man "like other things, as trees and quadrupeds, which cannot act by choice." Nobody praises an oak for growing straight. Therefore, at least some people genuinely deserve the praise or blame they get, in the plain sense that they have it coming.

(P2) Nobody deserves praise or blame for a deed that was not up to him. + This is the rule behind every excuse anyone has ever accepted. Where a deed was not the person's to control, blame slides off him, and everybody knows that long before anybody argues about it. (1) Every excuse we accept works by showing the deed was not fully the man's. - Aristotle wrote the rule down in the fourth century BC, and nobody has improved on it much. Praise and blame go to what a person does voluntarily; what is involuntary gets "pardon, and sometimes also pity." • The involuntary acts, on his account, are the ones done "under compulsion or owing to ignorance," and an act is compulsory when "the moving principle is outside" the man, "being a principle in which nothing is contributed by the person who is acting" (Nicomachean Ethics III.1). - Two families of excuse come out of that, and they work by different routes. • The first is compulsion, which is a failure of control. The deed ran through the man rather than out of him, the way a shove in a crowd sends you into the person in front. • The second is ignorance, which is a failure of knowledge. Knowledge is a second condition on responsibility, and a separate question (P3). - (P2) is about the first family only. When it says a deed was "up to him," it means the deed was in his control, and the phrase keeps that meaning every time it appears. - Aristotle's own example of compulsion has aged well. A man is carried somewhere "by a wind, or by men who had him in their power." Wherever he ends up, he did not go there. He was taken. - Run through the excuses of the first kind that you actually accept from people. He was pushed. Somebody stronger closed his hand around the knife. His arm jerked and he could not stop it. Then notice what never appears on any excuse list of either kind: nobody has ever been excused on the ground that the deed was too much his own. (2) The law of Moses separated the murderer from the manslayer. - Two neighbors go into the forest to cut wood, the axe head slips off the handle and kills one of them, and the survivor flees to a city of refuge and lives (Deuteronomy 19:4-5). Same corpse and same axe as a murder, and an entirely different verdict. The whole weight of that difference rests on what was up to the man. - Numbers puts the test in so many words. On one side is the man who "pushed him suddenly without enmity," or who "without seeing him dropped it on him," a man of whom the law says plainly that "he was not his enemy and did not seek his harm" (Numbers 35:22-23). On the other is the man who lay in wait. The law's question is never how bad the outcome was, but how much of the outcome was his. - Ezekiel then refuses to let the account be transferred to somebody else: "The son shall not suffer for the iniquity of the father, nor the father suffer for the iniquity of the son" (Ezekiel 18:20). Guilt does not travel, and the reason it does not travel is that it belongs to the one whose deed it was. • You may already be thinking of texts that pull the other way, where trouble falls on a whole household: the iniquity of the fathers visited on the children (Exodus 20:5), Achan's family destroyed along with him (Joshua 7:24-25). They are real, and they are hard. • What they describe is harm running down a family line, which is how families have always worked. What they never do is reckon a son guilty of his father's deed, and the law had already ruled that out: fathers are not put to death for their children, nor children for their fathers (Deuteronomy 24:16). (3) Courts still ask what was up to the man before they convict. - The law has a Latin name for the part of a crime that is not the deed itself. Mens rea means the guilty mind, and to convict a man of a serious crime you have to show not only that he did the thing but that he meant it, or knew what he was risking. Take the mental part away and what is left is an accident, which is a misfortune rather than a wrong. • There are exceptions, and they make the point rather than breaking it. A strict liability offense, such as selling alcohol to someone underage, convicts without any showing of a guilty mind at all. Those offenses are regulatory and carry small penalties, and lawyers have argued about them for as long as they have existed, for exactly the reason you would guess: convicting a man who did his honest best feels like an injustice. - A man with no history of illness has a seizure at the wheel and kills a cyclist. Anglo-American law calls this automatism, a movement of the body that the person never steered, and he is not sentenced for the death. Depending on the jurisdiction and on what caused the seizure, he is acquitted outright or he is examined and treated. Now change one fact, so that he knew he had epilepsy and drove anyway, and the verdict changes with it. The cyclist is just as dead in both stories. What differs is what was up to the driver. - Four more courtroom defenses look like four separate doctrines, and they turn out to be the same two families of excuse in legal dress. • Involuntary movement, where his arm was pushed or it spasmed, is the control family in its purest form. • Duress, where a threat left him no course anybody could be expected to refuse, is the control family at its edge. The law keeps that edge narrow: in English law duress is no defense at all to a charge of murder, however terrible the threat. • Infancy, where he was too young to grasp what he was doing, belongs to the knowledge family, since what the child lacks is the capacity to know. • Insanity belongs to whichever family the jurisdiction has chosen. The M'Naghten rules of 1843 ask whether the accused knew what he was doing and knew it was wrong, which is a test of knowledge. A court following the Model Penal Code, the model criminal statute the American Law Institute published in 1962 and many states have drawn on, also asks whether he could conform his conduct to the law, which is a test of control. (4) A deed can be up to a man indirectly, through a choice he made earlier. - A deed is directly up to a man when he chose it there and then. It is indirectly up to him when it followed from something earlier that he did choose. Confusing the two causes more trouble than almost anything else in this subject. - The drunk driver is the standard case. At the moment of the crash he could not steer straight, and he is guilty of the crash anyway, because the freedom he needed had already been spent three hours earlier, in the parking lot outside the bar, when he pocketed his keys and went in. Nobody thinks the second whisky bought him an excuse. - Blame also comes in degrees, and here the knowledge family does the grading (P3). The servant who knew what his master wanted and did not get ready "will receive a severe beating," while "the one who did not know" and did what deserved one "will receive a light beating" (Luke 12:47-48). The deed does not change between the two halves of that sentence. What changes is how much of it was the servant's. (5) A God who blamed people for what was never theirs would not be just. - Abraham's question assumes the premise and expects God to agree with it: "Shall not the Judge of all the earth do what is just?" (Genesis 18:25). He is asking God to be fair rather than merciful, and he is confident enough of the answer to bargain on it. - The judgment Scripture promises is "according to his works" (Romans 2:6), and a work that was never a man's own is not his work. Pull this premise out and the last judgment stops being a verdict and becomes a sorting, in which people receive outcomes for things that merely happened through them. Therefore, no praise or blame is deserved for a deed that was never the person's to control.

(C1) Therefore, some of what we do is up to us. + Put the two premises side by side and they leave exactly one place to land: some of what we do is up to us. The step is short enough to walk through in full. (1) The conclusion follows by a rule of logic called modus tollens. - A deductive argument works like a locked chain: if every link holds, the conclusion cannot slip out. There is no third option of granting the premises and shrugging at the conclusion. You either break a link or you take what comes. - The rule doing the work here is modus tollens, which is simpler than its name. If one thing would bring another with it, and the second thing is not so, then the first thing is not so either. If it rained last night the driveway is wet; the driveway is bone dry; so it did not rain. - Now fit the premises to it. (P2) says that if a deed was not up to the man, then he does not deserve praise or blame for it. (P1) says that some people do deserve praise or blame. Take one of them, the man who emptied the widow's savings. He deserves blame, so the second half of (P2) is false of him, and therefore the first half is false of him as well. His deed was up to him. (2) A premise has to be more plausible than its denial. - That is the standard for a deductive argument, and nothing more is required. Certainty is the wrong bar, since almost nothing outside mathematics can clear it, and demanding it is the usual way a real argument gets waved away. - Set (P1) beside its denial: nobody has ever deserved blame for anything, not the man who took the widow's savings, not anyone. - Set (P2) beside its denial: some people deserve blame for deeds that were never theirs to control, like the man carried off by the wind. - Neither denial is a thing people believe when they are not in the middle of defending a theory. (3) The conclusion claims a good deal less than it may appear to. - It does not say that everything we do is up to us, because a great deal of it is not. Your heartbeat is not, and neither is the family you were born into or the flinch when a door slams. Both premises were built from cases nobody argues about, the widow's thief on one side and the man carried off by the wind on the other. - It does not say what "up to us" requires. The premises leave that question open, and it is where the entire disagreement lives. - It does not say that determinism is false. Determinism is the claim that everything that happens, your choices included, was settled in advance by what came before. A thoughtful determinist who believes responsibility survives in such a world can sign this conclusion without flinching, and many do. - That is a strength of (C1) rather than a weakness. A conclusion both sides can accept is the only place an argument between them can start. Therefore, at least some of what human beings do is genuinely up to them.

(P3) Free will is the control that makes a deed up to the one who does it. + The control that makes a deed a person's own already has an ordinary name, and the name is free will. What the name does not tell us yet is what that control requires. (1) The phrase names the control, not a theory of how it works. - Free will is the power to make choices a person can be praised or blamed for. Philosophers call it the control condition: the part of responsibility that concerns a person's grip on what he did, as distinct from the part that concerns what he knew. - Think of two people standing beside a car. They can agree that the engine is what makes it go long before they agree whether the engine burns gasoline or runs off a battery. Naming the thing that does the work is a different job from saying how it works, and the name is useful precisely because both of them can agree on it first. - So defined, the phrase belongs to nobody in particular. Someone who holds that every choice was settled in advance can use it, and does, and so can someone who holds that some choices were not. That is exactly what lets the two of them disagree about one thing rather than talk past each other about two. (2) Two answers to "up to him in what sense" have run side by side since Aristotle. - The first is sourcehood: the choice came from the person himself rather than from something working on him from outside. Aristotle's version is still the crispest anyone has managed: "The acts whose moving principles are in us must themselves also be in our power and voluntary" (Nicomachean Ethics III.5). • Aristotle himself had no phrase for "free will." The vocabulary of a will that is free came later, worked out among the Stoics and inherited by the Latin church. What Aristotle supplies is the two questions, not the term. - The second is the categorical ability to do otherwise, which philosophers also call leeway. "Categorical" means unconditional: with the past and the laws of nature held exactly as they in fact were, more than one thing could have happened next. Picture a real fork in the road, as opposed to a road that only looked forked from the driver's seat. • In plainer words, leeway is an open door: another way the choice could have gone, there for the taking. • The principle built on leeway is worth learning by name, because the modern argument keeps coming back to it (D7). The principle of alternative possibilities says that a person is responsible for what he did only if he could have done otherwise. - Think of a letter with your signature at the bottom. The first question asks whether you wrote it at all, or took it down word for word from somebody standing behind your chair. The second asks whether you could have written a different letter. Those two questions usually get the same answer, but they are not the same question, and nearly every serious dispute in this field is a dispute about which one matters. • Of the two, the source comes first, and that order is kept everywhere downstream. The source is what makes a choice a person's own, and the open door is what a choice of that kind ordinarily looks like from the outside. A man can be the source of a choice with no other door open to him (D7). But nobody is the source of a choice that was made for him, however many doors stood open. (3) Responsibility needs knowledge as well as control. - A man hands his neighbor a glass of what he has every reason to believe is water, and it is poison. He handed it over, and handing it over was up to him, yet he is not a murderer, because he did not know. Nobody disputes this half of the picture. - Free will is the other half, the control half. The knowing half is real, but it is a different question, and keeping the two apart early saves a good deal of confusion later. (4) A free choice is made for reasons. - Hunger, habit, upbringing, advice, temperament, and the pull of what a person loves are all at work in a free choice, and none of them threatens freedom in the slightest. An influence is one thing and a determination is another, and the whole subject lives in the gap between them. - Your mother's advice weighs on a decision without making it for you. A hypnotist's command is a different sort of thing altogether, and everyone can feel the difference long before anyone can define it. (5) Freedom is not a power to do anything whatever. - A man who can swim still cannot swim in the middle of a desert. The ability is in him, and the opportunity is not always there. Keep those two apart and a surprising number of objections to free will fall over at once. - No one ever meant that a free man can choose to be seven feet tall, or to breathe under water, or to have been born in another century. The claim concerns which of the options actually before him he takes. Therefore, free will is the name for whatever control over his own actions a person must have in order to be praised or blamed for them.

(P4) A choice is either settled by everything that came before it or it is not. + Everything that came before a choice either settles which way the choice goes or it does not. There is no third thing for it to do. (1) Determinism is the claim that what came before settles everything that happens. - Set a row of dominoes on a kitchen table and tip the first one. The tenth domino's fall was settled before it happened, and a mind that knew the setup could have read the whole sequence off in advance. Determinism says the whole world is like that: the way it was a moment ago, together with the rules it runs by, was enough by itself to produce the next moment exactly as it came. - The conditions that do the settling need not be physical. Where the laws of nature do the work, it is physical determinism. Where God's decree does the work, it is theological determinism. Either way the thesis is the same: everything that comes to pass is causally necessitated by prior conditions, natural or supernatural. • "Prior" can mean logically prior rather than earlier on the clock, which leaves room for a timeless God settling things that happen in time. The claim is about what settles what, not about the order of the calendar. - "Settled" by what came before means causally necessitated by it: made to happen, the way each domino is knocked down by the one before it. The prior conditions have to be enough, all by themselves, to make the choice go the way it goes. • Nothing weaker will do, because a truth can make an outcome certain without making it happen. Suppose a fair coin will in fact land heads when it is tossed tomorrow. Then it is already true today that it will land heads, though nobody knows it yet, and that truth does not reach into tomorrow and turn the coin over. If a truth like that counted as settling the toss, every world would count as determined by definition, and the word would stop meaning anything. • Likelihood is not settling either. A prior state that makes a choice highly likely without making it certain has not settled it. A man who has ordered the fish every Friday for ten years will very likely order it again, and if his habit makes that likely without making it certain, his habit has not settled this Friday's order. • Nor is a guarantee that never acts. Picture a device planted in a man's brain that would force his choice if he wavered, and suppose he never wavers, so the device never has to do anything (D7). With the device in place only one outcome was possible, and still nothing before the choice made it go the way it went. • The likely choice and the guaranteed one therefore fall on the same side of the line as a choice with every door wide open: in each of them, nothing before the choice settled it. - Determinism is a claim about absolutely everything, which is why a single exception is fatal to it. If one person has ever made one choice that nothing before it settled, then determinism about human choices is false, whatever is true of the tides. (2) Whether freedom and determinism can both be true is a separate question. - Compatibilism is the thesis that they can. Notice how little it says. It does not say determinism is true, and it does not say that anybody is free, only that the two are not at war. - Incompatibilism says they are at war. If people are determined, then on this view they cannot be praised or blamed for anything they choose, and that is the end of it. - Two neighbors can argue for an hour about whether a man could be both a devoted father and a career soldier, without either of them ever saying whether this particular man is either one. The compatibility question is like that, and keeping it separate from the question of what is actually the case saves an enormous amount of crossed wire. (3) Two questions sort the field into four main positions. - Put two questions to any view, in order. Can freedom and determinism live together? And do we in fact have freedom? Four main positions come out of the answers, and every view you are likely to meet is one of them or a variant of one. - Libertarianism answers no, then yes: freedom is real, so determinism is false where choices are made. • The word has nothing whatever to do with politics; it comes from liberty. A free choice, on this view, is one that nothing settled but the person who made it, acting for reasons of his own. • In the ordinary case a choice like that shows up as a door he could have walked through and did not. The open door is its usual mark, and the source is the thing itself. - Hard determinism answers no, then no: determinism is true, so freedom is an illusion and nobody deserves anything. • Its best image comes from Spinoza, who denied that any will is free. Imagine a stone in flight that is given a mind just long enough to be aware of its own motion. It would be perfectly certain it was flying because it wanted to. - Free will skepticism also answers no to the second question, but it gets there without needing determinism: nobody is free and nobody is responsible. • That is what makes it the live version of the denial today. Suppose indeterminism is true, meaning that some things are genuinely not settled in advance. The skeptic says that is no help whatever, because a choice that nothing before it settled is not thereby anybody's doing. - Compatibilism, filled in, answers yes, then yes: we are free and responsible, and the world may be fully determined all the same. • The English philosopher Thomas Hobbes held it, and so did the American preacher and theologian Jonathan Edwards. When professional philosophers were last surveyed in large numbers, close to three in five of those who answered accepted some version of it. It is the majority view among philosophers, and anybody who treats it as a fringe position has not understood the disagreement. - None of this is new, and it helps to know how old it is. Josephus, writing for Roman readers in the first century, reports that the Jewish parties had already divided over exactly this. The Essenes said fate governs everything; the Sadducees denied fate altogether and held that "all our actions are in our own power"; the Pharisees held that some things are the work of fate and some are in our own power (Antiquities 13.171-173). The argument you are walking into was already old when the apostles were young. (4) Between settled and not settled there is no middle ground. - Nobody needs persuading of (P4), because it is an instance of the law of excluded middle, the rule that a statement and its denial cannot both be false. Either the prior conditions settled the choice or they did not. (P4) adds no new fact about the world. Its work is to split the question cleanly in two, so that the two accounts of freedom can be told apart. - The whole dispute comes down to a single yes-or-no question about a single choice. Hold every influence on it exactly as it was: the entire past, the man's makeup, his circumstances, and, if God is drawing on his heart, that too. Either all of it together settled the one choice he made, or it did not. - Picture it with a tape. Rewind to the second before he decided, with nothing whatever changed, and play it again. If what came before makes the choice go the way it goes, every replay comes out the same, and the choice was determined. If it does not, nothing before the choice settled it. Every position in the field is an answer to that question, paired with an answer about whether responsibility survives it. • In the ordinary run of choices, a choice that nothing before it settled also shows up in the replays as one that could have come out another way. That is the usual mark of such a choice, not the thing that makes it one (P3). • But the mark can be missing while the thing itself is there. Go back to the idle device in the man's head. With it in place no replay could end any other way, since in any replay where he began to waver the device would step in. Yet so long as it sits idle, as it did in the run that actually happened, nothing before the choice settled it (D7). • So the replay pictures the question without answering it. What answers it is whatever actually made the choice go the way it went. - An honest libertarian concedes a great deal here, and should. Much of the world runs like clockwork, or near enough: the tides, the chemistry, most of what goes on inside a body. His claim was only ever that some human choices are not settled in advance, never that the world as a whole is loose. Therefore, every choice a person makes was either settled by everything that came before it or it was not.

(C2) Therefore, human beings have free will, of either the compatibilist's kind or the libertarian's. + Human beings have free will, and it is one of two kinds, the compatibilist's or the libertarian's. Four claims are now in place, and together they leave those two live options and no more. (1) Put (C1) and (P3) together and free will follows. - Some of what we do is up to us (C1). The control that makes a deed a person's own is what "free will" names (P3). So human beings have free will, in the sense that matters to praise and blame. - Take the man who emptied the widow's savings and the neighbor who spent a Saturday fixing her roof. One has blame coming and the other has thanks, and so each of them had that control over what he did. - Which kind of control it is turns on (P4), and the split there is clean for a simple reason. A kind of control either survives determinism or it does not. • If the control we actually have is the sort a person could exercise even in a world where everything about him was settled in advance, it is the compatibilist's kind. • If it is the sort that requires that nothing before the choice settled it, it is the libertarian's. • There is no third kind. - Notice how little that second branch claims. Saying what libertarian freedom would require is not the same as saying anybody has it. Whether any human choice was in fact left unsettled by everything before it is a question (C2) leaves wide open. (2) Two positions are already out. - Hard determinism and free will skepticism both have to deny (P1), and denying it is expensive. On their view no punishment is ever deserved and no gratitude is ever earned. The man who robbed the widow has nothing coming, and neither does the stranger who pulls a child out of a river. The last judgment becomes a sorting rather than a verdict. - Their best defenders know the price, and that is the most telling thing about them. The serious books in that tradition give a great deal of their length to what should stand where desert stood: a forward-looking account of punishment, an account of love and meaning that asks nothing of desert, a replacement for the moral anger nobody can quite stop feeling. A substitution on that scale is an admission of how much (P1) was carrying. (3) Two hard questions are left standing. - First the philosophical one: whether the control responsibility needs can survive in a world where everything was settled in advance, and whether a choice that nothing before it settled could be anybody's doing at all. - Then the same question in theological dress: whether a God who knows every choice before it is made, and governs everything that happens, leaves the choosing to the one who does it. - (C2) decides neither, and pretending otherwise would spoil what it is for. What it decides is narrower and firmer: the denial of free will has been answered, and what remains is a family argument among people who agree that we are responsible for what we do. - It is the two people beside the car again (P3). They now agree that it runs, and what is left is the argument about what the engine burns. Therefore, human beings have free will, and the only question left open is which kind of freedom it is.

On the nature of free will and moral responsibility: Aristotle. Nicomachean Ethics. Translated by W. D. Ross. Oxford: Clarendon Press, 1908. Bourget, David, and David J. Chalmers. "Philosophers on Philosophy: The 2020 PhilPapers Survey." Philosophers' Imprint 23, no. 11 (2023). Finch, Alicia, and Ted A. Warfield. "The Mind Argument and Libertarianism." Mind 107, no. 427 (1998): 515-528. Fischer, John Martin. The Metaphysics of Free Will: An Essay on Control. Oxford: Blackwell, 1994. Fischer, John Martin, and Mark Ravizza. Responsibility and Control: A Theory of Moral Responsibility. Cambridge: Cambridge University Press, 1998. Frede, Michael. A Free Will: Origins of the Notion in Ancient Thought. Berkeley: University of California Press, 2011. Kane, Robert. The Significance of Free Will. New York: Oxford University Press, 1996. Lewis, C. S. Miracles: A Preliminary Study. London: Geoffrey Bles, 1947. Lewis, David. "Are We Free to Break the Laws?" Theoria 47, no. 3 (1981): 113-121. McKay, Thomas J., and David Johnson. "A Reconsideration of an Argument against Compatibilism." Philosophical Topics 24, no. 2 (1996): 113-122. Mele, Alfred R. Effective Intentions: The Power of Conscious Will. New York: Oxford University Press, 2009. Mele, Alfred R. Free Will and Luck. Oxford: Oxford University Press, 2006. Nagel, Thomas. Mortal Questions. Cambridge: Cambridge University Press, 1979. Nichols, Shaun, and Joshua Knobe. "Moral Responsibility and Determinism: The Cognitive Science of Folk Intuitions." Noûs 41, no. 4 (2007): 663-685. O'Connor, Timothy, and Christopher Franklin. "Free Will." In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, 2022. Plantinga, Alvin. Warrant and Proper Function. New York: Oxford University Press, 1993. Reid, Thomas. Essays on the Active Powers of Man. Edinburgh, 1788. Reppert, Victor. C. S. Lewis's Dangerous Idea: In Defense of the Argument from Reason. Downers Grove, IL: InterVarsity Press, 2003. Strawson, P. F. "Freedom and Resentment." Proceedings of the British Academy 48 (1962): 1-25. Timpe, Kevin. Free Will: Sourcehood and Its Alternatives. 2nd ed. London: Bloomsbury, 2013. van Inwagen, Peter. An Essay on Free Will. Oxford: Clarendon Press, 1983. van Inwagen, Peter. "Free Will Remains a Mystery." Philosophical Perspectives 14 (2000): 1-19. Widerker, David, and Michael McKenna, eds. Moral Responsibility and Alternative Possibilities: Essays on the Importance of Alternative Possibilities. Aldershot: Ashgate, 2003. Williams, Bernard. Moral Luck: Philosophical Papers 1973-1980. Cambridge: Cambridge University Press, 1981. Wolf, Susan. Freedom Within Reason. Oxford: Oxford University Press, 1990. On the theological questions: Craig, William Lane. "Free Will." Reasonable Faith, Question of the Week 320, 2013. Craig, William Lane. "Response to Paul Kjoss Helseth." In Four Views on Divine Providence, edited by Dennis W. Jowers. Grand Rapids: Zondervan, 2011. Crisp, Oliver D. Deviant Calvinism: Broadening Reformed Theology. Minneapolis: Fortress Press, 2014. Luther, Martin. The Bondage of the Will. Translated by J. I. Packer and O. R. Johnston. Westwood, NJ: Fleming H. Revell, 1957. Moreland, J. P. The Recalcitrant Imago Dei: Human Persons and the Failure of Naturalism. London: SCM Press, 2009. Pike, Nelson. "Divine Omniscience and Voluntary Action." Philosophical Review 74, no. 1 (1965): 27-46. Plantinga, Alvin. "On Ockham's Way Out." Faith and Philosophy 3, no. 3 (1986): 235-269. Stratton, Timothy A. Human Freedom, Divine Knowledge, and Mere Molinism: A Biblical, Historical, Theological, and Philosophical Analysis. 2nd ed. Eugene, OR: Wipf and Stock, 2026. Turretin, Francis. Institutes of Elenctic Theology. Edited by James T. Dennison Jr. Translated by George Musgrave Giger. Vol. 1. Phillipsburg, NJ: P&R, 1992. Wesley, John. "On Working Out Our Own Salvation." Sermon 85. In The Works of John Wesley, 3rd ed., edited by Thomas Jackson. London, 1872. Zagzebski, Linda Trinkaus. The Dilemma of Freedom and Foreknowledge. New York: Oxford University Press, 1991. Critic sources engaged: Bignon, Guillaume. Excusing Sinners and Blaming God: A Calvinist Assessment of Determinism, Moral Responsibility, and Divine Involvement in Evil. Eugene, OR: Pickwick, 2018. Frankfurt, Harry G. "Alternate Possibilities and Moral Responsibility." Journal of Philosophy 66, no. 23 (1969): 829-839. Frankfurt, Harry G. "Freedom of the Will and the Concept of a Person." Journal of Philosophy 68, no. 1 (1971): 5-20. Harris, Sam. Free Will. New York: Free Press, 2012. Helm, Paul. The Providence of God. Downers Grove, IL: InterVarsity Press, 1994. Hobbes, Thomas. Leviathan. London, 1651. Libet, Benjamin. "Unconscious Cerebral Initiative and the Role of Conscious Will in Voluntary Action." Behavioral and Brain Sciences 8, no. 4 (1985): 529-566. Pereboom, Derk. Free Will, Agency, and Meaning in Life. Oxford: Oxford University Press, 2014. Schurger, Aaron, Jacobo D. Sitt, and Stanislas Dehaene. "An Accumulator Model for Spontaneous Neural Activity Prior to Self-Initiated Movement." Proceedings of the National Academy of Sciences 109, no. 42 (2012): E2904-E2913. Soon, Chun Siong, Marcel Brass, Hans-Jochen Heinze, and John-Dylan Haynes. "Unconscious Determinants of Free Decisions in the Human Brain." Nature Neuroscience 11, no. 5 (2008): 543-545. Spinoza, Baruch. "Letter 58, to G. H. Schuller." In The Chief Works of Benedict de Spinoza, translated by R. H. M. Elwes. London: George Bell, 1883. Strawson, Galen. "The Impossibility of Moral Responsibility." Philosophical Studies 75, no. 1 (1994): 5-24. Legal sources: American Law Institute. Model Penal Code. Philadelphia: American Law Institute, 1962. M'Naghten's Case, 10 Cl. & F. 200 (H.L. 1843). Church documents: Canons of the Second Council of Orange (529). Council of Trent, Session VI: Decree on Justification, with its canons (1547). The Remonstrance (1610). Ancient sources: Irenaeus, Against Heresies IV.37. Josephus, Antiquities of the Jews 13.171-173. Justin Martyr, First Apology 43. Origen, Philocalia 23. Biblical citations: Genesis 18:25; 50:20; Exodus 4:21; 8:15, 32; 9:34; 20:5; Numbers 35:22-23; Deuteronomy 19:4-5; 24:16; 30:19; Joshua 7:24-25; 24:15; Ezra 1:1; Proverbs 16:9; 21:1; Isaiah 46:10; Jeremiah 17:9; Ezekiel 18:20, 30; 33:11; Daniel 4:35; Matthew 12:36; 23:37; Luke 7:30; 12:47-48; John 8:34; 19:11; Acts 2:23; 7:51; 17:30; Romans 1:24-28; 2:6; 3:11; 8:7; 9:18-19; 14:12; 2 Corinthians 5:10; Ephesians 1:11; 2:1.
+ Praise and blame are tools for shaping behavior. They work whether or not anybody "deserves" anything, so nothing about freedom follows from the fact that we use them.
1. Compatibilists themselves largely abandoned this answer. The position goes back to Hobbes, and it has an honorable history. On this view we blame and punish because doing so deters harm and reforms offenders, and that is justification enough. Nothing in it requires that the offender had it coming in any deeper sense. What is interesting is who gave it up. Over the second half of the twentieth century most compatibilists, the philosophers who hold that freedom and determinism could both be true at once (P4), stopped defending responsibility on those terms. They did it because of an argument from their own side, and the argument is simple. Resentment, gratitude, indignation, and forgiveness are not strategies. Nobody decides to feel grateful because gratitude encourages more favors. These responses are already answers to what somebody did, and an account of blame that leaves them out has changed the subject rather than explained it. A sharper version of the objection survives all that. You might say: of course we have these responses, but having them does not show they are warranted. Quite so. They are not free-floating feelings, though, and each one of them is about something. - Gratitude says a good was freely given. - Resentment says a wrong was freely done. - Forgiveness releases a debt that was really owed. Anyone who says the whole family of them is systematically mistaken has taken on a very large job: explaining how, wherever human beings have lived together, they have read a debt into a place where there was never one. 2. The modern version replaces desert rather than defending it. The strongest current form is not Hobbes's but the "quarantine" model, argued most carefully by Derk Pereboom. The proposal is to give up basic desert entirely, the idea that a wrongdoer has something coming purely for what he did (P1), and to handle dangerous people the way public health handles a dangerous infection. A typhoid carrier can be kept out of the kitchen without anybody blaming him for being ill. On the quarantine model a violent offender is detained in the same spirit: the detention looks only forward, restricts no more than it must, and carries no blame at all. The proposal is a serious one, and its defenders are not moral nihilists. But notice what kind of proposal it is. It is a replacement for desert, offered by people who see clearly that desert is going and that something has to stand where it stood. That is a concession about how much work desert was doing. 3. Desert is what sets a ceiling on what we may do to people. Desert is a limit as well as a license to punish, and the limit does more work than most people notice. - It is the reason a sentence can be too long. - It is the reason you cannot punish the innocent to make an example. - It is the reason a judge who fabricates a conviction to calm a frightened town has committed a crime even if the town does calm down. Quarantine has no such ceiling of its own. It has to borrow one from somewhere else, usually from a general principle that you restrict a person no more than you must, and a borrowed ceiling is exactly the kind that fails under pressure. If detaining a man who has done nothing would prevent more harm than detaining one who has, the forward-looking arithmetic says to detain him. 4. On the Christian account, judgment is a verdict and not a policy. "For we must all appear before the judgment seat of Christ, so that each one may receive what is due for what he has done in the body, whether good or evil" (2 Corinthians 5:10). The measure is what is due, not what will produce the best future. Concede the whole of the practical case. A determinist who gives up desert can keep prisons, keep deterrence, and keep every useful thing blame does. What he cannot keep is justice, and a God who handed out outcomes on those terms would be efficient rather than righteous. Abraham's question would have no answer: "Shall not the Judge of all the earth do what is just?" (Genesis 18:25).
+ "Up to him" just means he would have acted differently if he had wanted to. A determined man passes that test easily, so the premises are satisfied and no fork in the road is needed.
1. The conditional test gets something real right. Grant this much at once, because it is correct and it matters. If a man would have done otherwise had he chosen to, then nothing was blocking the door, nobody had tied his hands, and nothing stood between what he wanted and what he did. Philosophers call this the conditional analysis of "could have done otherwise," because it turns a claim about what he could have done into a claim about what would have happened if he had wanted to. It is the oldest compatibilist answer in the book, running from Hobbes through Hume to the twentieth century. Every account has to keep what the test gets right about failures to act: nobody is to blame for not doing what he could not have done even had he chosen to, unless he brought that on himself. A man in handcuffs fails the test. He could not have pulled the drowning child out of the pond even had he chosen to, and for whatever the handcuffs stopped, he is excused. 2. The trouble is a man whose wanting is itself not up to him. Picture someone with severe agoraphobia, a fear of leaving home so strong that he has not been outside in three years. Ask the conditional question. If he wanted to walk to the corner store, would he walk to the corner store? Yes. His legs work, the door opens, and nothing is stopping him. So he passes the test. Yet he is not free to leave, as everybody can see, and the reason is that his trouble lives in the wanting. The conditional analysis tests the machinery downstream of the want and never asks the question that matters, which is whether the want was his. Now widen the case. On determinism every want a person has was fixed by things he did not choose, running back through his upbringing and his genes to conditions that were in place before he was born. Everybody in that picture passes the conditional test perfectly, the agoraphobic along with everyone else, so passing it cannot be what makes anyone free. The agoraphobic, then, is no strange exception. He shows what the test cannot see in anyone, which is where the want came from. 3. Passing the test is not enough. The honest form of this objection does not claim that the conditional is all there is. It makes two narrower claims, and each deserves its own answer. - The conditional is the only requirement about alternatives that anyone has shown responsibility actually needs. - The categorical version, the real fork in the road (P3), has not been shown to be needed at all. The second claim can simply be granted. Nothing from (P1) to (C2) makes the fork a requirement: the source is what makes a choice a person's own, and the open door is only what such a choice ordinarily looks like (P3). The first claim, even if it is right, names at most a condition responsibility needs, and a condition a man needs is not one that makes him free. A car needs fuel, but a full tank does not tell you whether the car has an engine. The agoraphobic passes the test with room to spare and is not free, because the want that keeps him home is not his own doing. What the test leaves out, then, is the source of the want, and not some second requirement about alternatives. - Whether a choice came from the person or from something working on him is the question the test never asks, and granting the conditional does not dispose of it. - Whether a want produced entirely by what came before can still be the person's own is where the compatibilist and the libertarian part company, and the test cannot decide it for either of them. The accounts compatibilists offer today are built to answer exactly that question. 4. Most compatibilists no longer rest on the conditional test. The conditional analysis met the agoraphobic and cases like him decades ago, and the compatibilism you will actually meet today has moved past it. Its two main lines ask about the shape of the agent rather than about what he would have done in some other circumstance. - Reasons-responsiveness asks whether the man answers to reasons at all. He acts freely when the process he acts on would pick up a good reason and respond to it across a range of cases, so that he would not have taken the bribe had a police officer been standing at the door. - Hierarchical accounts look at a person's wants about his own wants, which is where the name comes from: one want sitting above another. A man acts freely when his will lines up with what he himself wants to want. That is why, on this account, the addict who hates his own craving is unfree, while the man who acts wholeheartedly is free. Both are better than the conditional analysis, and a serious compatibilist will put one of them on the table rather than the old test. Both also inherit a challenge of their own, which philosophers call the manipulation argument. Picture a man assembled to order by a team of neuroscientists, or a zygote designed so that thirty years later the grown man embezzles on schedule. He can satisfy either account to the letter, answering to reasons and wanting exactly what he wants to want. The question pressed against the compatibilist is this: why should being built by somebody be any worse for him than being produced by blind prior causes? That exchange is not finished, and nobody on either side pretends it is. 5. The premises themselves are not touched by this objection. Whatever "up to him" turns out to require, (P1) and (P2) stand. Some people deserve blame, and nobody deserves blame for what was not up to him. Both claims survive on any analysis of the phrase, whether the old conditional one or the two that replaced it. Suppose the conditional analysis is right. Then the deeds in question were up to those people in the conditional sense, and the conclusion holds with "up to us" read that way. That is why (C2) counts compatibilist freedom as one of the two live kinds rather than as a rival to be knocked down. In the end the objection is a proposal about what free will requires, which is a different thing from a reason to doubt that we have it. See also: Philosophy / Free Will: Doing What You Want Is Not the Same as Being Free Philosophy / Free Will: Wanting What You Want to Want Is Not Enough: The Hierarchical and Reasons-Responsive Accounts
+ If determinism is true, then everything you do follows from the distant past and the laws of nature, neither of which is up to you. So nothing you do is up to you, and (P1) is simply false.
1. This is the best argument that determinism leaves nothing up to us. The reasoning has a name, the Consequence Argument, and its most careful modern statement is Peter van Inwagen's. Informally, it runs in three steps. - If determinism is true, our acts are the consequences of the laws of nature and events in the remote past. - It is not up to us what went on before we were born, and it is not up to us what the laws of nature are. - So the consequences of those things, our present acts included, are not up to us either. Put plainly: nobody now alive had any say in how things stood in the age of the dinosaurs, or in whether gravity pulls, and on determinism those two together fixed what you had for breakfast this morning. That is a valid-looking argument from premises that look obviously true, and it has persuaded serious people. It is also the main reason for thinking freedom and determinism cannot both hold. 2. The Consequence Argument never says that determinism is true. Read it carefully and it is a conditional, an if-then. If determinism is true, nothing is up to us. About whether determinism is in fact true it says nothing at all. So the argument does not refute (P1) on its own. What it rules out is holding determinism, the argument's premises, and (P1) all at once. That leaves three ways out, and anyone who has (P1) in hand may take any of them. - Deny that determinism is true, in which case the argument's opening condition never fires. - Deny one of the argument's own premises, which is what compatibilists do, and have been doing in print since the early 1980s. - Deny that anybody is ever responsible for anything, which is what the objection wants and what (P1) refuses. Only the third touches (P1), and it is the one that has to be set beside the widow and her savings. Making that comparison is the move philosophers call tollensing the ponens, and the name is uglier than the idea. - An argument that runs forward, "if these premises, then this conclusion; these premises; so this conclusion," is a modus ponens. The Consequence Argument, as its defenders run it, is one. - The same if-then can be run backward, the way modus tollens runs (C1): if these premises, then this conclusion; but not this conclusion; so not all of these premises. - That is the whole move. When a valid argument ends somewhere you have better reason to reject than to accept, you keep the denial of the conclusion and give up a premise instead. Which premise to give up is a further question, and (C2) does not decide it. 3. The compatibilist reply to the Consequence Argument is a live one. The objection is usually delivered as though only somebody who rejects determinism could refuse it, and that is not so. The second way out has been open for forty years, and people who accept determinism have been taking it the whole time. The whole argument turns on what philosophers call a transfer rule, and the name says what the rule does: it carries powerlessness down a chain. Take an oak that has stood since before the Civil War. Nothing you can do affects the weather in the summer of 1850, and if that summer's weather fixed the width of one ring inside the oak, then nothing you can do affects the ring either. Put abstractly, if nothing you can do makes any difference to A, and A guarantees B, then nothing you can do makes any difference to B. It looks obvious. Two lines of attack have nonetheless done real damage to the argument. - The first is that the rule has counterexamples, cases where it carries you from true premises to a false conclusion, and they all work by adding. Hold a fair coin you have decided not to toss. You have no power to make it land heads, since a toss might come up tails, and for the same reason you have no power to make it land tails. Yet whether it lands on one face or the other is plainly up to you: just toss it. Two things you were powerless over, added together, made a thing you were not powerless over. The old rule quietly allowed that kind of adding, and that is where the trouble got in. • The reply from the other side is that the rule can be mended, and it was mended in print in 1998. The repaired version drops the adding and moves powerlessness along one chain at a time, and the known counterexamples do not touch it. So a compatibilist now has to aim at the mended rule, which is a good deal harder than pointing at the old one. - The second line reads the premises more carefully. "He could have done otherwise" does not require that he could have broken a law of nature. Say you kept your hand on the desk just now, and determinism is true. To say you could have raised it is to say only that, had you raised it, some earlier fact or some law would have been different from what it actually was. Being able to act in a way such that a law would have been broken is not the same as being able to break one. Nobody is claiming the powers of a stage magician who switches off gravity. Both lines are contested, and neither is a knockdown, but neither is silly either. They are the reason (C2) leaves compatibilist freedom standing as one of the two kinds instead of pretending the Consequence Argument has already closed the case. 4. Determinism is not something anybody has shown to be true. The objection is often delivered as though physics had decided the matter, and it has not. The textbook formulation of quantum mechanics, with its rule that a measurement resolves a range of possible outcomes into one, is indeterministic: on that reading the outcome of a single event is genuinely not fixed beforehand. There are fully deterministic interpretations as well, among them David Bohm's, in which particles are steered by a guiding wave, and Hugh Everett's many-worlds picture, in which every possible outcome occurs on some branch of reality. No experiment has yet told them apart. That cuts both ways, and it should be said plainly. Unsettledness in physics does not by itself give anyone free will, and no part of this case rests on quantum mechanics. The point is narrower than that. Determinism is a philosophical claim with respectable defenders and no experimental proof, and it is not the sort of thing that outweighs what we know about the widow. 5. The self-defeat reply works only in a much weaker form. Some answer the objection by turning it around. If determinism is true, they say, then the determinist was always going to believe determinism whatever the evidence was, so he cannot claim to have followed the evidence. That popular version fails, and it fails the way fatalism fails ("whatever will be will be, so why bother"), by holding the outcome fixed while changing everything that led up to it. Had the evidence been different, the prior conditions would have been different too, and he might well have believed something else. A pocket calculator is as determined as anything in the world and it is right every time, so being determined is no bar to tracking the truth. The calculator repays a second look, though. It gets sums right because somebody who could reason built it to, which makes it a poor model for a mind whose reliability, on a purely physical picture, has to arise with no reasoner anywhere behind it. What survives is narrower and harder to shake off. If a belief is fully accounted for by physical conditions reaching back before the believer was born, then whether it is also held because the reasons for it are good becomes a further question, and a further question needs an answer rather than an assumption. Determinists have given answers, mostly by way of accounts on which a physical process counts as reasoning when it reliably yields true beliefs. Whether that is enough belongs to a different argument, and neither (C1) nor (C2) needs it. See also: Philosophy / Free Will: If Determinism Is True, Nothing Is Up to Us NT / Moral and Rational: Free-Thinking Argument
+ A choice that nothing settled in advance is a fluke. If the reasons and the character and the whole prior state left the outcome open, then which way it went was a matter of luck, and luck is nobody's doing.
1. This is the strongest objection in the field. Grant the force of it. Rewind to the moment before the choice, hold everything fixed, play it forward a thousand times, and suppose the man takes the bribe in four hundred of those runs and refuses in the other six hundred. Nothing about him explains the difference between the runs. So what makes any particular run his doing rather than something that happened to him? The move has a name, or rather two of them. - Rewinding and replaying is van Inwagen's rollback argument. - A sharper version is Derk Pereboom's disappearing agent objection. It is aimed at accounts on which only events inside the person, his beliefs and desires and brain states, produce the choice, and the person himself never does. On those accounts, if the prior state of the agent does not settle the choice, then at the moment of decision there is nothing left to do any settling, and the agent has gone missing from his own act. Nobody serious pretends this is easy, and it is the reason a good many philosophers who reject determinism still reject libertarian freedom. 2. "Undetermined" and "uncaused" are not the same word. The objection quietly slides between them. A choice can fail to be necessitated by prior conditions and still be caused, if what causes it is not a prior condition but the person himself. That is the libertarian proposal in its strongest form, and it has a name: agent causation, the view that the person, rather than merely some state inside him, is what produces the choice. Think of the difference between a single radioactive atom, which decays at a moment nothing in particular brought about, and a woman who picks up the phone to call her sister because it is her sister's birthday. Nobody decayed the atom. The woman made the call, and she can tell you why. A random event is caused by nobody, while a free choice is made by somebody, for reasons he has in mind. The luck objection bites hardest where nothing at all produces the outcome, not even the person, and agent causation is the denial of exactly that. 3. Reasons explain a free choice without forcing it. Think of a woman weighing two job offers, one nearer her mother and one better paid. Both sets of reasons are real, and whichever she takes, she can say afterwards why she took it, and the explanation is true. What she cannot say is that the reasons left her no option. That is how reasons usually work in deliberation: they make a choice intelligible rather than inevitable. Whether an explanation of that kind is good enough, or whether only a condition that guaranteed the outcome would count as explaining it, is exactly what the two sides disagree about, and calling the undetermined choice a fluke assumes the answer instead of arguing for it. 4. Libertarians without agent causation have an answer too. Their answer starts from a harder case than the job offers, one where somebody is genuinely torn. A woman late for a meeting hears a scuffle in an alley and sees a man being robbed. Real reasons pull her both ways, and in that moment she is trying to do both things at once. Whichever way the effort comes out, it is an effort of hers that succeeded, and she cannot afterwards file the result under things that merely happened to her. A man who swings at a glass table and breaks it cannot put the break down to luck on the ground that his arm might have wavered, and she is in the same position. Willing both options and getting one of them is not the same as tossing a coin between them. 5. The same problem waits on the determinist's side of the room. Ask the determinist the same question about the woman weighing two job offers. Why did she take that job? Because the state she was in settled it. And why was she in that state? Because an earlier one settled that, and so on back past her birth. Neither picture ends in a place where the person plainly authored himself. The difference is that on the libertarian account there is somebody at the point of decision doing the settling, and on the determinist account the settling was finished before he arrived. The luck objection is answerable, and answering it properly is real work rather than a slogan. 6. (C2) survives even if this objection wins outright. Suppose the luck objection is decisive and no undetermined choice could ever be anybody's doing. That would be a reason to strike out one of the two kinds of freedom, not a reason to deny that anyone is free. (P1) and (P2) would stand untouched, (C1) would still follow, and the control that makes a deed a person's own would have to be the kind that survives determinism. Taken by itself, the luck objection is an argument for compatibilism. It becomes an argument for free will skepticism only when it is joined to the claim that a determined choice is not up to us either. Put the two together and the whole weight falls where it belongs, on (P1), which is where the widow and her savings are waiting. See also: Philosophy / Free Will: An Undetermined Choice Is Not a Lucky One, and Responsibility Does Not Require Self-Creation Philosophy / Free Will: The Agent, Not Merely the Agent's States, Settles the Choice
+ Nobody chose the character his choices come out of. To be responsible for what you do you would have to be responsible for how you are, and for whatever made you that way, all the way down, and nobody is.
1. This objection is aimed at everybody, not only at libertarians. Galen Strawson's Basic Argument runs in four steps, and each one looks reasonable on its own. - What you do flows from how you are. - So to be truly responsible for what you do, you would have to be truly responsible for how you are, at least in the respects that produced the deed. - But to be responsible for how you are, you would have had to choose to be that way, and that choice would itself have flowed from how you already were, which you would in turn have had to be responsible for. - The demand keeps backing up a step and never stops, and making yourself from scratch is impossible, so nobody is ever truly responsible for anything. Put in ordinary terms: you snapped at your brother because you are short-tempered. To answer for the snap you would have to answer for the temper, and to answer for the temper you would have to answer for whatever gave it to you, and so on back to the cradle. The force of it is real. This argument does not care whether determinism is true. It cuts against compatibilists and libertarians alike, and it is the reason free will skepticism is a live position held by careful people rather than a pose. 2. The argument proves far too much. The conclusion is not "human responsibility is diminished" but "responsibility is impossible for any being whatsoever, necessarily." No agent of any kind, in any world that could ever have been, however he was put together, could be responsible for anything, because the condition the argument sets is one that nothing could satisfy. A being who had simply always existed would fail it too, since he never chose the nature he has always had. It is like a club whose rules admit only people who were members before they joined. Nobody fails that test through bad luck, because nobody could ever pass it. When an argument delivers a conclusion of that scope from premises about ordinary human limits, the right response is to look hard at the premises rather than to accept a result that nothing could ever count against. 3. The step that fails is the demand for self-creation. It is the third step that has to be refused: the claim that to be responsible for how you are, you must have chosen to be that way. Responsibility does not require that you made yourself. It requires that the deed came from you, that you and not something working on you from outside are the one who settled it (P3). Think of a man who inherited a quick temper from his father, and knows it, and has spent twenty years learning to leave a room instead of shouting. He did not choose the temper, and nobody imagines he did. What he chose, a thousand separate times, was what to do in the first three seconds. Nobody watching that man thinks the unchosen temper makes the twenty years of work meaningless, and if the Basic Argument says it does, the Basic Argument has lost contact with the thing it is about. 4. Scripture holds people responsible for lives they did not design. The Bible's picture of a person is nothing like a self-made man. We are born into a family, a nation, a body, and a fallen race, and none of it was chosen. Scripture then addresses that same person as accountable, and it varies the account by what he had: the servant who knew and did not act gets the heavier sentence, and the one who did not know gets the lighter (Luke 12:47-48). That is the distinction the Basic Argument flattens. Responsibility comes in degrees and tracks what was available to the person, and a picture that requires either total self-authorship or total excuse fits neither Scripture nor anything anyone actually believes about their own children. 5. Raising the stakes does not repair the argument. Galen Strawson aims the argument hardest at desert of the sort that could make an eternal reward or punishment fair, and that is the version a Christian answer can least afford to duck. Take it at full weight. If nobody is ever truly responsible, a final judgment is not merely mistaken; it is the largest injustice ever committed, and the size of the stakes is exactly why the question matters. Four things can be said in reply, and all four are needed. - What hangs on the argument cannot make the third step true. Suppose self-creation really is required before anybody deserves anything. Then gratitude for a kindness is as groundless as a verdict at the last day, since both rest on the same footing. Suppose it is not required. Then the verdict is no more groundless than the gratitude. The reasoning runs the same distance in both directions, and no argument is made valid by the weight of what depends on it. - Judgment in Scripture is not a bare ledger. It is graded by what a person had and knew (Luke 12:47-48), and the message the whole book is carrying is that the debt is paid by someone else rather than collected. A judgment pictured as a machine weighing deeds against a standard nobody chose is not the picture the New Testament gives. - The judge knows everything the argument says we cannot see: the temper inherited, the fear learned at six years old, the door that was never really open. That is the difference between a verdict that is roughly fair and one that is exactly right, and it is the reason a Christian can hold both that desert is real and that nobody is judged for what he never had. - Galen Strawson concedes the thing that matters most here. On his own account the conviction he is attacking is not a doctrine somebody argued us into. It arrives with the experience of choosing, it is rooted deeply enough that he expects most people never to give it up, and he does not treat philosophers as exempt from it. An argument that nobody can live inside, resting on a demand for self-creation that nobody made until the argument needed it, is an argument whose premises are the place to look. See also: Philosophy / Free Will: Nobody Chose Their Own Character, and Everyone Is Still Responsible: The Basic Argument
+ Brain recordings show activity building toward a decision hundreds of milliseconds before the person is aware of deciding. The brain has already chosen, and consciousness is told afterwards.
1. The finding is real. Benjamin Libet asked subjects to flex a wrist whenever they felt like it, and to note the position of a spot on a clock face at the moment they first became aware of the urge. Electrodes taped to the scalp picked up a slow rise in electrical activity, the "readiness potential," beginning several hundred milliseconds before the reported moment of awareness. That is about a third of a second, roughly the length of a blink. Later work using brain imaging reported that the upcoming choice between two buttons could be predicted, at rates modestly above chance, seconds before the subject reported deciding. None of this is fraud or bad science, and the results have replicated. An answer that questions the data has not read the papers. 2. What was measured is not what the argument needs. Flexing a wrist whenever you feel like it is not deciding anything. There is nothing to weigh and no reason favoring one moment over another, and the subject has already made the only real decision in the room, which was to cooperate with the experiment. That matters because the conclusion on offer is about deliberation: about a man deciding whether to testify honestly, or how to answer a letter from his brother. Nothing in the laboratory tested a decision of that kind, and the leap from a twitching wrist to a moral choice is the weakest joint in the whole argument. 3. The readiness potential may not be a decision at all. A later and well supported model, published by a team of neuroscientists in 2012, says the readiness potential is nothing but ordinary background fluctuation in brain activity. Neural activity is always drifting up and down, like the surface of water that never sits perfectly still. When the task is "move whenever you feel like it," the movement tends to happen at a moment when the drift has wandered high enough to cross a threshold. Line up many trials by the movement and average backwards from it, and you get the familiar rising curve even though nothing was building toward anything. The model fits the data at least as well as Libet's reading of it. It has not replaced that reading by agreement, and nobody should be told it has: the argument about what the readiness potential actually is has run for more than a decade and is running still. What it has taken away is anybody's right to say the data speak for themselves. The experiment that was supposed to show the brain deciding early may have been showing noise. 4. Libet himself did not draw the conclusion he is quoted for. He held that the conscious agent retains a veto in the remaining fraction of a second, and he wrote about it at length. Whatever one makes of that, the researcher at the center of the story did not think his own data abolished free will, and the popular version of his experiment is one he spent years correcting. 5. A prediction slightly better than a coin toss is not a settled future. The imaging studies that report predicting a choice in advance typically do so at around sixty percent accuracy against a fifty percent baseline. That is a real effect, and it is worth explaining. It is also a very long way from showing that the outcome was fixed, and the gap between "somewhat predictable" and "determined" is precisely the gap the objection needs to close and does not. 6. The objection assumes that only a consciously initiated act is the person's own. Suppose every finding is granted at full strength, and the brain begins to move toward the flexion before the subject notices any urge. Why should that show the person did not do it? The assumption underneath the objection is that a deed is mine only if my conscious awareness started it, and that assumption excuses far too much to be usable. A pianist plays a run of sixteen notes faster than she can form a thought about any one of them. A driver brakes before he has registered the child. Nobody concludes that the music and the braking were done by somebody else. A person is not a small conscious observer perched on top of a brain, receiving bulletins. A person is the whole living creature, awareness and habit and body together, and the moment the observer is told is not the moment the deed is done. Notice which way this cuts. A compatibilist can take the entire Libet literature on board without flinching, since nothing in it says the movement came from outside the man. At its very strongest the research is a claim about the timing of awareness, while (P1) is a claim about desert, and those are different subjects. See also: Philosophy / Free Will: The Brain Scans Have Not Found the Decision Before the Decider
+ A man can be responsible when he could not have done otherwise. Put a device in his brain that would force the choice if he wavered; he makes it on his own, the device never fires, and he is responsible with no alternative open.
1. Frankfurt's case has moved a great many philosophers. Harry Frankfurt's 1969 example is the most influential thought experiment in the field. In the version most people meet now, a neuroscientist, usually called Black, wants Jones to vote a certain way and implants a device that will intervene if Jones shows signs of choosing otherwise. Jones chooses that way on his own, the device stays dormant, and it seems plain that Jones is responsible for a choice he could not have avoided making. If that is right, then responsibility does not require the ability to do otherwise, and the principle of alternative possibilities (P3) is false. Concede at once that this is a powerful argument and that most compatibilists now lean on it. The strongest form of the objection presses one step further. The classic reason for thinking determinism rules out responsibility is that it closes every door (D3). If responsibility never needed an open door, then closing all of them costs responsibility nothing, and the libertarian has lost his main reason for denying determinism. 2. Frankfurt's case leaves the source untouched. Free will is the control that makes a deed the person's own (P3). On the libertarian account, a choice is free when nothing settled it but the person who made it, for reasons of his own. The open door is the ordinary mark of such a choice, not its definition. As the story is told, Jones meets that condition completely. The device did nothing at all. The vote came from Jones, for his reasons, and nothing outside him settled it. The one thing the story takes from him is the open door, which is why libertarians who define freedom by the source accept a version of the case. Some who define freedom by the source deny that Frankfurt's own story closes every door. On their reading Jones could still have tried to vote the other way, though he could never have pulled it off, and he was the source of his vote either way. The claim that a man can be the source of a choice with no door open at all does not hang on Jones, though. It rests on the rebuilt versions of the case, which are still argued over, and on Pilate's case from the Gospels, which has no device in it at all. 3. Whether Jones kept a door open is disputed. The first standing reply to the case is a dilemma. For the device to act in time it needs an early sign that Jones is about to waver, and either that sign is reliably connected to the choice or it is not. - If it is, then the earlier state already fixed the later one. The case has quietly assumed that Jones's choice was determined, and anyone who doubts that a determined man is responsible has been given no reason to think Jones is. - If it is not, then Jones could still have gone the other way without ever showing the sign, too late for the device to stop him, and an alternative was open to him after all. The second reply is what philosophers call the flicker of freedom. Even in the leanest versions of the case, it says, Jones keeps some alternative, if only the alternative of beginning to waver and being overridden. You might expect a determinist to welcome Frankfurt's case, since it seems to help his side. Yet one of the people who press the flicker reply is the most thorough recent defender of theological determinism, who declines to rest any part of his case on Frankfurt cases because he holds that the flicker reply defeats them. Frankfurt's defenders have not stood still, and they have an answer to each. - Against the dilemma they build what are called buffer cases. The sign the neuroscientist watches for is not a cause of anything. It is simply a first step Jones cannot skip on the way to the other option, rather as a man cannot walk out of a room without first turning toward the door. Block that one step and every route to the alternative is closed in advance, while the choice Jones actually makes stays unsettled until he makes it. - Against the flicker they ask what a flicker is worth. An alternative so thin that the agent could never have reached anything different by taking it is not the kind of alternative anyone thought responsibility required. The exchange has been running since the 1990s, and neither side has swept the board. 4. Pilate's case has no device for either reply to reach. Both replies feed on the device: the dilemma on its early sign, the flicker on the wavering it stands ready to catch. One defender of free will grants that replies of this kind seem "cogent against human examples of prevention or intervention." A God who knows what a man would freely do in any circumstance, the knowledge theologians call middle knowledge (D9), needs neither sign nor device, though. The same defender makes the point with Pilate, the Roman governor who tried Jesus. - Suppose God had resolved that if Pilate, once in the judgment seat, would let Jesus go, Pilate would never be put in the seat at all. Now Pilate is in the seat. So under the circumstances he could not have let Jesus go, since if he had been going to, he would not be there. Yet, in that defender's words, "he freely sent Jesus to the cross, since nothing determined him to do so." - There is no sign here for the dilemma to work on. God's knowing is reliable because it conforms to the choice, not because it causes it (D8). Nothing stands ready to step in, so there is no overridden wavering for a flicker to live in. - What shuts the door is God's resolve, not his knowing, and it shuts it in one sense only: the resolve and Pilate's letting Jesus go could not both have happened. Nothing stopped Pilate from letting Jesus go, and nothing took away his power to do it. The resolve settles where he stands, not what he does there. Jesus still counts Pilate's part a sin, if the lesser one (John 19:11). - God need not arrange things that way, and in the ordinary run of choices the door stands open. The case does assume that God knows what Pilate would freely do before deciding where to put him, and that knowledge has serious critics as well as serious defenders. 5. A winning Frankfurt case would buy less than it is credited with. Suppose the rebuilt cases carry the day. Then an open door is not something responsibility needs, and the libertarian account in (P4) never said it was. What that account asks is that the person be the source. That is why the strongest form of the objection does not go through. It needs a premise it never states: that closed doors were the only harm determinism could do. The libertarian who defines freedom by the source never granted that. For him the trouble with determinism is that something other than the person settles every choice he makes, and Jones, whose choice nothing outside him settled, is no evidence either way about that. Three things about the debate since Frankfurt point the same way. - Frankfurt's followers drew the lesson that a choice is judged by how it actually came about, not by the roads not taken. Determinism is a claim about exactly that: that conditions in place before the man was born were enough to produce his choice. - The argument that a choice produced that way is not the man's own never mentions a door. It is the manipulation argument (D2), the man built to order who embezzles on schedule. It is still argued over, and no Frankfurt case, however it comes out, can decide it. - The accounts compatibilists built in Frankfurt's wake still ask for something in the neighborhood of an alternative, such as a man who would not have taken the bribe had a police officer been standing at the door. That is a condition on how the choice came about, though, not a second road he could have taken. After Frankfurt, then, both sides argue about where a choice came from. What no version of the case can show, without assuming it at the start, is that a choice settled before it was made is free. 6. Nothing in the case disturbs either premise. (P1) says some people deserve praise or blame, and Jones deserves it. (P2) says nobody deserves it for a deed that was not up to him, and the vote was up to Jones, since he made it and the device did not. The case is about what "up to him" requires, and it gives no reason at all for thinking that nothing is. See also: Philosophy / Free Will: Black the Neurosurgeon: Freedom Never Depended on the Open Door Philosophy / Free Will: God Knows What You Would Freely Do, and Plans Around It
+ God knew every choice you would make before the world was created. If he knew it, it was going to happen, and if it was going to happen you were never able to do anything else.
1. The inference commits a named mistake in logic. Two sentences here look almost alike and say quite different things. - True: necessarily, if God knows that you will do something, you will do it. The word "necessarily" governs the whole if-then. Given the knowing, the doing follows, and there is no way to have the one without the other. - Not true: if God knows that you will do something, then you will necessarily do it. Here "necessarily" has slid off the if-then and landed on the last part by itself, where it now says that the deed could not have been otherwise. The second does not follow from the first, and the same slide is easy to spot in an everyday case. Necessarily, if a man is a bachelor, he is unmarried. It does not follow that a bachelor is necessarily unmarried, as though he could never walk down the aisle. Medieval theologians had names for the two things being confused, the necessity of the consequence and the necessity of the consequent. Today the mistake is called the modal fallacy, "modal" because it concerns the modes of necessity and possibility. All that follows from God's knowing is that you will do it, not that you had to. 2. Foreknowledge depends on the future, not the future on foreknowledge. Suppose a forecaster is infallible, never once wrong in a thousand years. She says it will rain on Thursday, and on Thursday it rains. Did her forecast make the rain fall? Obviously not. Her being right depended on the weather; the weather did not depend on her being right. Knowledge of what the knower does not himself bring about is like that all the way down: it conforms to its object. Your knowing that a friend is left-handed does not make him left-handed, and the reliability of the knower changes nothing about the direction that dependence runs. Origen made the point in the third century, and Christians have been giving it ever since: a thing does not happen because God foreknows it; God foreknows it because it is going to happen. The objection gets its grip by picturing foreknowledge as a cause, a kind of advance pressure on the future. It is not a cause. It is knowledge, and nothing about a proposition's being known changes how the thing it is about came to be. 3. The hard version of this objection makes no such mistake. A well-read objector will not argue that way at all. He will run the argument through the fixedness of the past instead, which is the version philosophers have worked on since Nelson Pike put it in print in 1965. - Eighty years ago God believed that you would read this sentence today. - What happened eighty years ago is fixed. Nobody now has any power over it, and no fact about the past is up for revision. - If you were able to do otherwise now, you would be able to make it the case that God believed otherwise then, and that would be a power over the past. - So you are not able to do otherwise. Nothing in that turns on confusing the two necessities. Everything turns on the fixedness of the past, which nobody has ever disposed of quickly. Two replies have carried the weight for centuries, and both are live. - God's past belief is not a fixed fact about the past in the sense the argument needs. Philosophers sort facts about the past into two kinds. A hard fact is over and done with, settled entirely by what happened at the time: the Battle of Hastings was fought in 1066, and that is that. A soft fact is one whose truth still waits on what happens later, the way "the prophet spoke truly in 700 BC" depends on what happened in the first century. Nobody has power over a hard fact and nobody needs any. What a free agent has over the soft one is a harmless sort of backward reach: had he chosen otherwise, God would always have believed otherwise. - God may not be located in the past at all. If his knowing is not at a time, there is no earlier moment whose fixedness could reach forward and close anybody's options, and the argument loses its first step. The first reply sometimes gets waved away as a convenience invented to rescue foreknowledge. It is nothing of the kind. It was already true eighty years ago that you would read this sentence today, and if every fact about the past were fixed in the way the argument requires, that plain truth would abolish freedom all by itself, with God left out of the story altogether. Hardly anyone accepts that, which means hardly anyone is without the distinction already. Neither reply is free of difficulty, and the literature on both is large. What does not wait on that dispute is where the burden sits. The fixity argument runs only if God's past belief is fixed in the way the Battle of Hastings is fixed, and that is the objector's to establish rather than to help himself to. The version of the objection most people meet is invalid before it starts. 4. Denying God's foreknowledge is not the way out. Some Christians answer this objection by denying that God knows future free choices at all, a view called open theism. That trade should be refused. God is the one "declaring the end from the beginning" (Isaiah 46:10), and a God who had to wait and see is not the God of the prophets. How foreknowledge of a free choice is possible, and how God's knowledge fits with his government of the world, are real questions with long answers. Whether his knowing makes everything fated has a short one. Fatalism, the view that what happens was going to happen whatever anyone did, does not follow from God's knowing, and no appeal to mystery is needed to say so. See also: Philosophy / Free Will: God's Knowing What You Will Do Does Not Make You Do It Philosophy / Free Will: The Future Is Not Open to God: Answering Open Theism Philosophy / Free Will: God Does Not Look Ahead: How Foreknowledge Is Possible
+ Scripture says God "works all things according to the counsel of his will" and that none can stay his hand. If God ordains everything that comes to pass, then no choice of yours was ever open.
1. These texts should be read at full strength. Five of them carry the weight of the objection, and every one says what the objector says it says. - "In him we have obtained an inheritance, having been predestined according to the purpose of him who works all things according to the counsel of his will" (Ephesians 1:11). - "My counsel shall stand, and I will accomplish all my purpose" (Isaiah 46:10). - Of the God who does as he wills among the inhabitants of the earth, Nebuchadnezzar says that "none can stay his hand or say to him, What have you done?" (Daniel 4:35). - "The heart of man plans his way, but the Lord establishes his steps" (Proverbs 16:9). - Proverbs elsewhere pictures a king's heart as a channel of water in God's hand, turned wherever he pleases (Proverbs 21:1). No answer that trims those is worth anything. They teach a God whose purposes are not at risk, who governs all things down to the last detail, and who is not waiting on anybody. That is the Christian doctrine of providence, God's actual government of everything that happens, and it is not negotiable. 2. The texts say that God governs everything, not how he does it. Here is the distinction the objection passes over. That God accomplishes all his purpose is one claim. That he accomplishes it by determining each human choice is a second claim, and the first does not contain the second. Consider a God who knows in full what each person would freely do in any situation he might be placed in, and who chooses which world to create with all of that in view. Theologians call the first half of that middle knowledge, so named because it sits between God's knowledge of everything that could happen and his knowledge of what actually will. A God who has it works all things according to the counsel of his will in the strictest sense. Nothing escapes him and nothing surprises him. That is a reading of the same texts on which every one of them stays true and no choice is settled for the chooser. The king's heart is the hardest of the five, because it does speak to how. Read it closely and it says that God turns a particular heart, when he wills, to serve a purpose he has announced, which is what the record shows him doing with Cyrus (Ezra 1:1). Nobody on either side of the dispute denies that God can move a man. What the verse does not say is that every heart is moved so, in everything it does, and that is the claim the objection needs. The determinist reading satisfies these texts too, and saying so costs nothing. That is the whole point: a text that comes out true on either reading cannot be used to decide between them. 3. Scripture repeatedly puts the plan and the guilt in one sentence. Joseph tells his brothers, "you meant evil against me, but God meant it for good" (Genesis 50:20). The brothers are not excused by the second clause. Peter preaches that Jesus was "delivered up according to the definite plan and foreknowledge of God" and in the same breath that his hearers "crucified and killed" him "by the hands of lawless men" (Acts 2:23). Whatever those verses teach about providence, they refuse the inference the objection wants. The definite plan does not make the lawless men anything other than lawless, and the text puts the two facts side by side without apology or explanation. 4. Scripture also shows things happening that God does not want. Four more texts sit in the same Bible, and they pull the other way. - "How often would I have gathered your children together as a hen gathers her brood under her wings, and you were not willing!" (Matthew 23:37). - Stephen tells the council that "you always resist the Holy Spirit" (Acts 7:51). - Luke records that "the Pharisees and the lawyers rejected the purpose of God for themselves" (Luke 7:30). - "As I live, declares the Lord God, I have no pleasure in the death of the wicked" (Ezekiel 33:11). Any account of providence has to hold both sets of texts at once, and the Reformed tradition, the Calvinist branch of Protestantism, has a careful answer that deserves to be named rather than gestured at. It distinguishes two senses in which God is said to will: his decretive will, what he has determined shall come to pass, and his preceptive will, what he commands and approves. On that reading Jerusalem's children were not gathered because it was not decreed, while the command and the grief over its refusal are entirely sincere. The distinction is old and was not invented for this quarrel. In fact every account of providence needs something like it, since even a reading on which no choice is settled must explain how God can command what he knows will not be done. What the distinction does not do is decide the question, and its best defenders do not claim otherwise. It shows that the two sets of texts can be held together on a determinist reading. They can also be held together on a reading where what God decrees includes what free creatures would do. Both leave every verse standing, and the sovereignty texts therefore cannot carry the objection by themselves. 5. The hardening texts do not close the question either. The strongest scriptural form of this objection is not Ephesians 1:11. It is Pharaoh, and Romans 9. God says before the plagues begin that he will harden Pharaoh's heart (Exodus 4:21). Paul draws the general conclusion that God has mercy on whom he wills and hardens whom he wills, then puts the reader's protest into words: if that is so, why does God still find fault, since no one can resist his will (Romans 9:18-19)? Anyone who has not felt the weight of that has not read it. Three things are true of those texts at once, and an honest reader keeps all three. - Exodus divides the hardening between them. God hardens Pharaoh, and Pharaoh also hardens his own heart (Exodus 8:15, 32; 9:34), and the narrative alternates rather than choosing. - Paul's answer to the protest he has just voiced is not an account of the mechanism. It is a rebuke of the question and a statement of the potter's right over the clay, and a reading that makes it more than that is supplying what the text withholds. - Hardening in Scripture is most often judicial. It falls on people who have already refused, in the pattern where God gives men up to what they have chosen (Romans 1:24-28). None of that turns Romans 9 into a comfortable chapter, and it is not meant to. The narrow point is the one that bears on (C2). The texts say that God hardens, and they do not say whether the hardening settles a choice Pharaoh would otherwise have settled himself or seals him in a choice he has already made. The question they raise is the question (P4) puts. See also: Philosophy / Free Will: What the Sovereignty Texts Say, and What They Do Not Philosophy / Free Will: Theological Determinism and the Authorship of Sin Philosophy / Free Will: God Knows What You Would Freely Do, and Plans Around It
+ Scripture says the fallen will is dead in sin and enslaved to it. A dead man does not choose, so whatever freedom Adam had, nobody has it now.
1. The doctrine behind these texts is not in dispute. The texts are every bit as strong as the objection says they are. - Paul tells the Ephesians that "you were dead in the trespasses and sins" (Ephesians 2:1). - Jesus says that "everyone who practices sin is a slave to sin" (John 8:34). - Paul writes that "the mind that is set on the flesh is hostile to God," and as for submitting to God's law, "indeed, it cannot" (Romans 8:7). - "The heart is deceitful above all things, and desperately sick" (Jeremiah 17:9). Nobody turns to God on his own steam. Far from being a Reformed distinctive, that is the common inheritance of the whole church, and the names that say so are not the ones people expect. The Second Council of Orange said it in 529, the Council of Trent said it in the sixteenth century, the Remonstrants (the Dutch Protestants who founded the Arminian tradition) said it, and John Wesley said it. Grace comes first, always, and anyone who denies that has left the faith rather than won an argument. 2. The bondage in question is a bondage toward God. Read what the texts are about. They concern a man's ability to turn to God, to please him, to submit to his law. They are not about whether he chose to insult his neighbor this morning. Luther, whose The Bondage of the Will is the hardest book in this tradition, drew that line himself. He granted free choice in what he called the things below a man, the ordinary business of living, and kept his denial for the things above him. So the bondage runs toward God, while (P1) and (P2) concern deeds like the taking of the widow's savings, where nobody on any side denies that the man did it. Luther should not be made to say more than he said. He wanted the phrase "free choice" dropped from Christian speech altogether, and he was harsher about the human will than almost anyone since. The concession is his all the same, and it is the shape of his denial that matters: what he denied was a power in a man to stand well before God. 3. Scripture holds the fallen accountable. This is decisive and simple. The same Bible that describes the slavery pronounces the judgment. "He will render to each one according to his works" (Romans 2:6) is written about people described a chapter later as having none who seeks for God (Romans 3:11). If bondage removed responsibility, the judgment texts would be unjust and the commands would be cruel. Paul's word in Athens, that "now he commands all people everywhere to repent" (Acts 17:30), is spoken to people God had not yet changed, and God says of those same people that he takes no pleasure in their death but wants them to "turn back" and live (Ezekiel 33:11). Scripture never treats this inability as an excuse, and that fact is evidence both sides have to account for. 4. The grace that draws a person can also be refused. Here the common ground runs out, and saying so plainly is better than smoothing it over. Grace comes first, and grace can be refused. That is a real position with real opponents, it crosses a line the Reformation drew, and nothing is gained by dressing it up as agreement. The answer on the other side is a good one. It distinguishes two callings: an outward call that comes through preaching and can certainly be refused, and an inward call, the Spirit's own work in a heart, which cannot. On that reading Stephen's charge that "you always resist the Holy Spirit" (Acts 7:51) describes the first and says nothing at all about the second. Three things stand against it. - The distinction is a theological construction rather than one the text draws. Stephen's charge is that they resist the Holy Spirit, not that they resist a preacher, and Luke writes of men who "rejected the purpose of God for themselves" (Luke 7:30), which is as inward a description as anybody could ask for. - Jesus's lament over Jerusalem puts God's wanting and the outcome on opposite sides of one sentence: "How often would I have gathered your children together as a hen gathers her brood under her wings, and you were not willing!" (Matthew 23:37). Whatever else is going on there, something God wanted did not happen, and a human unwillingness is the reason the sentence gives for it. A familiar answer separates the ones who were unwilling, the leaders being addressed, from the children they kept away. Grant the reading. Christ still wills a gathering that does not occur, and the failure is still laid at a human door. - God states his own preference about the wicked, and it is not the outcome they get. He takes no pleasure in their death, and he calls them to "turn back" and live (Ezekiel 33:11). A call whose failure the caller had already settled is a strange thing to grieve over. None of this makes the other side foolish, and its defenders have a reply to each of the three. What it does mean is that the resistible reading is a reading of the texts rather than a retreat from them, and that it is where most of Christendom has stood: the Eastern churches, Rome since Trent, whose canons on justification condemn the claim that a man's will cannot refuse its consent when God calls and stirs it, and the Wesleyan and Arminian traditions after them. 5. Election and perseverance are left exactly where they were. Neither the basis of election nor whether the saved can fall away is touched by any of this. Those belong to the doctrine of salvation, and a person can hold that ordinary choices are free and still take any of the historic views on them. Some Reformed philosophers do exactly that. See also: Philosophy / Free Will: The Bondage of the Will Is Not the Denial of Free Will Philosophy / Free Will: Ought Implies Can: What a Command Presupposes
+ Free will has been defined as whatever control responsibility requires. On that definition the conclusion is true by definition and tells us nothing, and the interesting question has been dodged rather than answered.
1. The definition is neutral on purpose. Grant the complaint's first half entirely: the definition is chosen on purpose. "Free will" names the control condition for moral responsibility and nothing more (P3), and it is used that way precisely because both sides of the argument can use it. It is the definition a leading recent defender of theological determinism adopts in so many words, noting that the phrase serves determinists and those who deny determinism alike. A trial works the same way. Prosecution and defense agree on what counts as murder before they argue about whether the defendant committed one, and nobody calls that agreement a dodge. An argument that started from a definition only libertarians accept would prove nothing to anybody else, and the charge most often laid against libertarians is exactly that: that they load indeterminism into the word and then act surprised when determinism turns out to be incompatible with it. Refusing to do that is the opposite of dodging. 2. Neither premise is true by definition. (P1) says some people deserve praise or blame. That is a contested claim about the world, and a substantial body of philosophers denies it. (P2) says nobody deserves praise or blame for a deed that was not up to him. That is also contestable, and Hobbes's descendants have contested it. The conclusion that we have free will is only as strong as those two, and neither is true by definition. A definition can name what an argument establishes; it cannot establish it. All the work is done by (P1) and (P2), and an argument that was empty would not have opponents who spend whole books denying (P1). 3. The narrow conclusion is exactly what a foundation should deliver. What (C2) establishes is that the denial of free will is false, that two named positions are out, and that the remaining disagreement is between people who agree we are responsible. That is not everything, and it was never going to be. Think of it as a survey stake rather than a house. It fixes the boundary of what is being argued about, so that both sides are arguing about the same thing. A dispute in which each side uses the key phrase to mean something different is not a dispute at all, and a great deal of what passes for debate on this subject is precisely that. 4. The interesting question is named rather than dodged. (P4) puts it in one line: either everything before a choice settled it or it did not. (C2) says that the answer to that question is what divides compatibilist freedom from libertarian freedom, and that one of the two is the kind we have. Naming the question is not answering it. There is also no way to answer it well before it has been named. See also: Philosophy / Free Will: Read This First: Thesis
+ The Bible is full of commands to choose, but that proves nothing, because every side of this dispute agrees that people choose. You still have to show why choosing requires that nothing before it settled the choice.
1. The objection is correct. This is the sharpest thing the other side says, and its best defenders put it this way: pointing out that there is choice everywhere in the Bible decides nothing, because both sides affirm choice, and what is needed is a reason why choice requires that the past left it unsettled. Exactly so. A determinist affirms that people choose, deliberate, weigh options, and decide. He says those choices are determined. So "choose life" (Deuteronomy 30:19) and "choose this day whom you will serve" (Joshua 24:15) are common property, and neither side can claim them. 2. Neither premise rests on the word "choose." (P1) rests on desert, the claim that praise and blame are sometimes owed, which Scripture teaches in the judgment texts rather than the choosing texts. "He will render to each one according to his works" (Romans 2:6) is a text about receiving, not about choosing. (P2) rests on the excusing principle, which is common ground with anybody who has ever accepted an excuse and is written into the law of Moses at Numbers 35:22-23. Strike every "choose" out of the Bible and both premises stand where they stood. 3. What Scripture does supply is a pattern the objection has to account for too. Four things in the text are evidence, and they are the evidence (P1) and (P2) draw on. - Its hearers are addressed throughout as people who will answer for themselves. - Judgment is varied by what a person knew (Luke 12:47-48). - Guilt is not allowed to pass from father to son (Ezekiel 18:20). - God states his displeasure at what people in fact do (Ezekiel 33:11). A determinist can account for all four. That is the honest position, and pretending otherwise would be the mirror image of the mistake this objection names. What he cannot do is read them off as support for his own view either, and that leaves the question where it belongs, in philosophy and theology rather than in a word search of the Bible. 4. Whether a free choice must be unsettled by the past has to be won by argument. The question the objection asks is the right one, and three further questions carry the weight of answering it. - Whether a determined agent can be the source of his own act. - Whether "he could have done otherwise" means the conditional thing (what he would have done had he wanted to) or the categorical thing (a real fork in the road), and whether responsibility needs either. - What happens to responsibility when somebody else builds the agent. None of that is decided by (C2), and none of it is smuggled into (C2). The conclusion (C2) reaches is one the objector himself accepts, which is a strange thing to complain about. See also: Philosophy / Free Will: Doing What You Want Is Not the Same as Being Free Philosophy / Free Will: Wanting What You Want to Want Is Not Enough: The Hierarchical and Reasons-Responsive Accounts Philosophy / Free Will: A Manipulated Agent Is Not Free, and a Determined Agent Is a Manipulated One
+ The whole case rests on the intuition that the man who robbed the widow has something coming. Intuitions have been badly wrong before, they flip depending on how the case is described, and the one you call obvious is the very thing in dispute.
1. The strongest form of this objection is not a debating point. Put it the way its best defenders put it. Moral intuitions have endorsed slavery, honor killing, and the burning of heretics, so feeling obvious is no guarantee of anything. Worse, the intuitions in this particular dispute are unstable. Ask people in the abstract whether anybody in a fully determined universe could be morally responsible, and most say no. Ask the same people about a particular man in that same universe who murdered his wife and children, and most say he is fully responsible. Psychologists have run those studies, and the split holds up across repetitions. An argument that opens with "you already know he deserves it" is helping itself to the thing that was supposed to be under examination. That is a real challenge, and an answer that just retold the widow story more loudly would deserve to be ignored. 2. Every argument has to start from something it did not argue for. There is no reasoning about morality from nowhere, and the premises on the other side are not self-evident either. Causal closure is a working commitment of physical science rather than one of its findings, and the claim that a determined agent can deserve nothing is itself an intuition, with a long history of thoughtful people rejecting it. There is also a standard being assumed here without argument: that a starting point is guilty until it can be proved from something else. Nobody works that way. What we are directly aware of, the way a memory presents itself or the way a deed strikes us as vile, is evidence as it stands, and we reason from it to what best explains it until something explains it better. That is how every one of us treats memory and eyesight, and there is no separate and harsher rule kept in reserve for moral perception. So the question is never "intuition or argument." It is which starting points hold steadiest under pressure. Set the two candidates side by side. - That the man who spent a year befriending a widow in order to rob her has earned blame. - That no human being has ever deserved anything, and the only difference between that man and a driver whose brakes failed is a difference in what we should do next, never a difference in what either of them is owed. Anybody who genuinely finds the second steadier than the first should follow it where it goes. Almost nobody does, and the philosophers who defend it are the first to report how hard the belief is to shake even in themselves. 3. The intuitions that were overturned were overturned by other intuitions. Slavery was not refuted by physics. It was refuted by pressing a moral conviction people already held, that a human being is not a tool, against a practice that contradicted it, and by exposing the special pleading that had kept the two apart. That is how moral knowledge corrects itself, and naming the method tells you what a refutation here would have to look like. It would have to show that the desert intuition conflicts with a moral conviction steadier than itself, or that it survives only by an exception nobody would grant in any other case. This objection does neither. It observes that intuitions are fallible, which is true of every source of belief anybody has, eyesight included, and which by itself defeats nothing in particular. 4. The framing studies show something narrower than they are used for. Take the finding at full strength: abstract questions about determinism draw one answer, concrete cases draw another, and the gap is wide and repeatable. What follows is that people are unreliable at applying a metaphysical thesis they were handed thirty seconds ago, which is not surprising. What does not follow is that the concrete judgment is the mistaken one. The data read at least as naturally the other way, as showing that the abstract question is the one people answer badly, because on hearing the word "determined" they picture a man in chains rather than the thesis defined in (P4). Those studies are evidence about how human beings process a hard idea. They are not a verdict on which of the two answers is true, and the researchers who ran them do not claim they are. 5. (P1) does not need intuitions to be infallible. The standard is the one every deductive argument works to: a premise earns its place when its denial is harder to believe than it is (C1). Not certainty, and not an intuition that could never fail. That leaves one question, and it is not whether the desert intuition might conceivably be wrong. It is which of the two you would stake a life on: some people have it coming, or nobody has ever had anything coming. The first is not a hunch to be dropped the moment a theory asks for it. It is among the most widely shared and most stubbornly held judgments human beings make, and Scripture writes it into the last day: each one receives "what is due for what he has done in the body, whether good or evil" (2 Corinthians 5:10). See also: Philosophy / Free Will: We Have Good Reason to Believe We Are Free NT / Moral and Rational: The Moral Argument
+ (P2) is false, and everybody's practice shows it. Two equally reckless drivers, one of whom hits a child while the other gets home safely, are not blamed the same, though the difference between them was pure luck.
1. The phenomenon is real. Grant it at once, because the cases are everywhere once you start looking, and they run well beyond the two drivers. Philosophers have catalogued them under the heading of moral luck, meaning the ways a person's blameworthiness seems to depend on things outside his control. - Luck in how it turns out: the drunk who gets home and the drunk who kills a child face different charges, different sentences, and different faces in the mirror. - Luck in circumstance: a man who would have informed on his neighbors, had he lived in the wrong country in the wrong decade, never had to find out, and nobody blames him for what he never did. - Luck in who you are: nobody chose his temper, his intelligence, or his capacity for courage. The third is the demand that a man author his own character, and it fails for the reason every demand for self-creation fails, which is that it sets a condition nothing could meet. The first two are the ones aimed at (P2), and they are not a trick. The law grades by outcome too, and attempted murder is punished more lightly than murder, where the whole difference is sometimes a surgeon's skill. 2. Outcome changes how much a man answers for, not whether he answers. Look hard at the two drivers. Both are blameworthy, and both were blameworthy the moment they turned the key; nobody says the one who got home did nothing wrong. What the dead child changes is how much the man must answer for, while (P2) is a claim about whether he must answer at all: nobody deserves blame for a deed that was never his to control. Both men controlled the drinking and the driving. Neither controlled the child stepping off the curb. So the honest description of the case is not "we blame a man for what was not up to him." It is "we blame a man for what was up to him, and we measure the blame partly by a result that was not." Whether that measuring is justified is a genuine and unresolved question. It is not a counterexample to the premise as written, and (C1) needs no stronger premise than the one written. 3. The discomfort these cases produce is itself evidence for (P2). This is the part the objection cannot afford. Nobody finds these cases comfortable. Judges, juries, and ordinary people all feel the pull of the thought that the two drivers were equally at fault, and the reason they feel it is that they already accept the control principle and can see it straining. Consider what the objection would have to say if (P2) were simply false. It would have to say the cases are unremarkable, no more troubling than charging more for a larger loaf. That is not how anybody experiences them, including the philosophers who first wrote the puzzle down. A principle that explains why a puzzle is a puzzle is in better shape than a rival that has to deny the puzzle is there. 4. The law has never been comfortable with it either. Anglo-American law does grade by outcome, and its own scholars have criticized it for doing so for the better part of a century, on precisely the ground that the two drivers are morally alike. The arguments for keeping the grading are practical ones. - Results are visible where intentions are not. - A system that punished every reckless driver as a killer would be unworkable to administer. - The community's response to a death is not something the law can simply ignore. Those are good reasons of policy. What they are not is the claim that the man who got home deserves less, and the defenders of outcome grading generally concede that much, which is why the argument has never ended. A practice its own practitioners defend on grounds of workability is thin evidence against a principle they admit they are straining. See also: Philosophy / Free Will: Nobody Chose Their Own Character, and Everyone Is Still Responsible: The Basic Argument

Public Theology

Theological Analyses of Societal Issues

Theology of the Family

Where Faith and Family Intersect

Biographies

Notable Works & Great Quotes from Key Figures

Contemporary Voices

1960 AD – present
Playlist
Justified True Belief