Justified True Belief

(●) A slippery slope predicts that a first step leads by an unstoppable chain to disaster, without arguing the chain. + Some chains are real, and a person who warns about one is doing us a favour. The fallacy is asserting the chain instead of arguing it. (1) The shape of the move is a chain with only its two ends showing. - Mrs. Adair asks the town council for a bench outside the library. The reply: allow a bench and next it will be a bus shelter, then a coffee kiosk, and in ten years the green will be a car park. And so, no bench. - The reasons someone offers are the premises, and the thing they are meant to support is the conclusion. Here the conclusion is "no bench", and the only premise is "a bench leads to a car park", which nobody defended. Patrick Hurley's textbook files the move among the fallacies of weak induction, where the premises give the conclusion some support but nowhere near enough. - Give each of four links generous odds, seven chances in ten, and the whole chain runs less than one time in four. A slope feels certain because nobody multiplied. (2) It tempts because some slopes are real. - Every parent knows the first exception tends to become the rule. Eugene Volokh's study of the slope in law lists the ways a first step does ease the second: it lowers the cost of going further, changes what people are used to, and hands the next campaigner a precedent. The feeling of the ground tilting is honest. - What misleads is that the feeling does the arithmetic for us, and does it badly. A fair warning inspects the links one at a time. The fallacy skips the inspection and keeps the alarm. (3) The name, and the three cousins. - The pictures are older than the label: the thin end of the wedge, and the camel that wanted its nose in the tent and ended by owning it. In 1908 F. M. Cornford gave the academic forms mock-solemn names, the Principle of the Wedge and the Principle of the Dangerous Precedent, and drew from the second the moral that "nothing should ever be done for the first time". The label itself is twentieth-century, and Douglas Walton's standard 1992 monograph sorts the family into four, three mechanisms and a fourth that combines them. - The causal slope is the domino: one event sets off the next. "Legalise the lottery and the casinos will follow." - The precedent slope is the wedge, decisions rather than events: "allow this exception and you must allow the next case like it." Courts reason this way on purpose, so this cousin is legitimate as often as not. - The sorites, Greek for "heap", runs the slope through a vague word. Eubulides posed it in the fourth century BC: one grain is not a heap, and adding one never makes one, so there are no heaps. "One more won't hurt", said forty times, is the same move. Therefore, a slippery slope is a chain with its middle links missing, offered as though every one of them were in place.

(●) Simple example: "If I let you stay up ten minutes, next it will be an hour, and then you'll never sleep again." + It is ten past bedtime, and the parent hears the sentence come out of their own mouth. (1) What is claimed and what is offered. - The claim is that there will be no extra ten minutes. The support is a chain, ten minutes tonight, an hour next week, a child who never sleeps again, with only the two ends shown. (2) Why the support does no work. - Nothing was said about why ten minutes leads to an hour. Try the swap test: put "one more story" in place of "ten minutes" and the chain runs to the same disaster in the same words, which means it was not built from the facts of this case. - What would bear on the claim is the record: what happened last time, and whether this child treats exceptions as precedents. Some do. That is an argument, one link long. (3) What the reply was really doing. - It was refusing without giving the real reason, probably a good one (school in the morning). A slope is restful to commit because it sounds like foresight and asks nothing of the person committing it.

(●) A critic says: "Once you allow that God can act in nature, science is finished, because any result could be a miracle." + A chemist says this at a barbecue, and means it kindly: she is worried that a friend who believes in miracles has, without noticing, given up the ground her laboratory stands on. (1) What is being argued. - The claim is that belief in miracles and the practice of science cannot live in the same head. The support is a chain: allow one miracle, and no experiment can be trusted, because any odd result might be God's doing; and if no experiment can be trusted, there is no science. - The worry has a real seed, and it should be granted. Science does run on the assumption that nature behaves on Tuesday as it did on Monday. David Hume built his case against miracles on that same assumption in the Enquiry: unbroken experience is what establishes a law of nature, he argued, so the proof against any exception is as strong as experience can make it. (2) Where the fallacy sits. - The link that does the work, from "God can act" to "any result could be a miracle", is asserted and never argued. A miracle, on the believer's own definition, is a rare exception, and an exception only makes sense against a rule. C. S. Lewis's Miracles makes the point with Joseph, who was troubled by Mary's pregnancy precisely because he knew perfectly well where babies come from. Belief in a miracle, far from ignoring the regularity of nature, depends on it. - The chain also had three and a half centuries to run, and it did not. Robert Boyle, Isaac Newton and Michael Faraday all believed the God of the Bible had acted in history, and all went on running experiments as though nature would behave, because they expected it to. (3) The sound version. - The real question is whether a believer has a principled reason to expect regularity, and whether allowing miracles hands him a science stopper: an explanation that ends inquiry the moment a result surprises him, "God did it", and the notebook closes. Alvin Plantinga answers both in Where the Conflict Really Lies. The laws of nature describe how the world goes when God is not acting specially, so a miracle is not a law failing. And a Christian expects regularity because a faithful God made the world, which means nothing in his theology licenses reading a miracle into a titration that came out wrong. - The critic can then press the fair question: how do you tell a real exception from a bad experiment? That is a question of method with an answer, because a miracle claim is a historical claim, checked with the historian's tools rather than the chemist's. (4) How to answer it. - "Every scientist who has ever lived assumes nature usually behaves, and so do I. A miracle only counts as one because nature usually does. Let's talk about how you would tell a real exception from a bad result, because that is a question I have to answer, and so do you." See also: Natural Theology: Evolutionary Argument Against Naturalism (EAAN) CE / Resurrection: Minimal Facts Argument (Habermas)

(●) A Christian says: "Question the age of the earth and you will end up denying the resurrection." + The same move from our side of the table, and it is said out of love for Scripture, which is exactly why it needs taking apart with care. (1) What is being argued. - The claim is that a Christian must hold one particular reading of the days in Genesis 1. The support is a chain: read the days as anything other than ordinary days and you are reading Scripture non-literally; once one passage is read that way there is no principle to stop; and so the resurrection goes too. - The strongest form should be granted. Interpretive habits do travel, and church history has scholars who began by reinterpreting one text and ended by setting aside the creed. The popular apologist who says this has usually watched it happen to someone. (2) Where the fallacy sits. - The link "there is no principle to stop" is asserted, and there is a principle. It is the oldest one in reading, genre: the kind of writing a text is, which tells you how to take it. A psalm, a parable and a census list are all in the Bible, and nobody reads them the same way. - Orthodox Christians differ over Genesis 1 precisely because they differ over what kind of text it is. Some read the days as ordinary days; some read each day as a long age (the day-age reading); some take the week as a literary arrangement of God's work by topic rather than a timetable (the framework reading). All three are held by people who confess the resurrection every Sunday. Nobody differs over the Gospels on that point. Luke opens by appealing to eyewitnesses and earlier accounts, and dates his story by the rulers then reigning (Luke 1:1-4; 3:1-2), and reading a psalm as poetry has never obliged anyone to read Luke as poetry. - So the chain jumps from a judgment about one book to a judgment about a quite different book, and never notices the jump. The same slope runs the other way in a critic's mouth, "admit Genesis is history and you must reject geology, then astronomy, then medicine", and the shape is identical. (3) The sound version. - Underneath the slope is a real point about method: read Scripture consistently. The consistent method is to read each text according to the kind of text it is, and to let the text itself show you which kind that is. That is an argument about method, not a prediction about falling, and it can be made for Genesis 1 on Genesis 1's own evidence. - What the speaker owes the room is that argument. Christians who reach different answers there can all hold the tomb empty, because the Gospels are a different case. (4) How to answer it, when a fellow Christian says it. - "The question is what kind of writing Genesis 1 is, and faithful Christians differ on that. The Gospels are a different kind of writing, and that is why the people you are worried about, who read Genesis differently from you, still believe he rose." See also: CE / NT Criticism: Literary/Historical Character of the Gospels CE / NT Criticism: The Gospels: Real History

(●) A charge that misses: "Saying that without God everything is permitted is a slippery slope. Atheists don't go on to permit everything." + Ivan Karamazov never says the sentence in those words. In Dostoevsky's novel other characters report his view that way. The charge reads it as a forecast that has already failed. (1) What the charge claims. - The moral argument says that if God does not exist, objective moral values do not exist: values that hold whether or not anyone agrees with them. The charge takes that as a prediction: remove God and people slide into permitting anything. It then points to the many atheists who keep their promises and raise their children well, and concludes that the slope never happened. (2) Why the charge misses. - The sentence is a claim about grounding, not about behaviour. To ground something is to be what makes it the case: the foundation makes the house stand, whatever the tenants do. The claim is that on atheism nothing makes cruelty wrong, whatever anyone goes on to do about cruelty. "If the bank fails, your savings are no longer backed" is not a prediction that you will go on a spending spree. - William Lane Craig, the argument's best-known defender, says in Reasonable Faith exactly what it does not claim: not that atheists cannot know right from wrong, not that they cannot live good lives, and not that belief in God is needed to behave well. He separates what makes moral values real (he calls this moral ontology) from how we come to know them (moral epistemology), and the argument lives entirely in the first. - Here is the line between the fallacy and its legitimate cousin. Douglas Walton's verdict is that a slope argued step by step, with the likelihood of each link weighed, is a reasonable argument, and Volokh shows that in law it is often the right one, because precedent really does move. The test to carry away is a question: has each link been argued, or only the first step and the last? Paul's chain in 1 Corinthians 15 passes, because he states every link: "And if Christ has not been raised, then our preaching is in vain and your faith is in vain." (1 Corinthians 15:14). - A grounding claim has no links at all. It is one step, a fact and what makes it a fact, so the test does not even apply to it. (3) What the charge would need to be true. - It would need the argument to be a forecast about societies, and the concession is that Christians often make it one. "Take God out of the schools and watch the crime rate" is a slippery slope unless the sociology is done link by link, and it usually is not. Defenders of the moral argument should keep their distance from that sentence; Dostoevsky's book is a story about one family, not a statistical claim. (4) How to answer it. - "You are right that atheists are often good, and the argument agrees with you. It is not saying you will behave badly. It is asking what makes cruelty wrong if there is no one it is wrong before. That is a question about what grounds morality, and I would like your answer to it." See also: Natural Theology: The Moral Argument