Justified True Belief

(●) No True Scotsman rescues a claim about a group by redefining the group to exclude whoever breaks it. + A claim about a whole group meets a member who breaks it, and rather than give up the claim the speaker gives up the member. One word does the work, "true" or "real", slipped in front of the group's name. (1) The definition changes when the evidence does. - Hamish opens his Sunday paper and reads that a man in Sidcup has committed a terrible crime. "No Scotsman would do such a thing," he says. The next Sunday the paper reports an Aberdeen man who has done worse, and Hamish does not budge. "No true Scotsman would do such a thing." The whole fallacy lives in that one word. - That Aberdeen man is a counterexample, a member of the group who breaks the claim made about it. Hamish's first sentence was about the world, and Aberdeen refuted it. His second looks the same but is really about how Hamish uses a word, and Aberdeen cannot touch it. He has kept the sentence and thrown away its content. - Two words help here. The reasons someone offers are the premises, and what they support is the conclusion. Hamish's premise was supposed to rest on the facts about Scotsmen, and when the facts turned he redefined instead. - Logicians file the move among the fallacies of presumption, arguments that quietly assume what they were meant to prove. The assumption here is that Scotsmen are spotless by definition. (2) It tempts because groups really do have standards. - We say "a real friend would have called" all day long, and we mean something by it. Friendship has content, and a person can fall short of it. So "true" feels as though it points at something, and sometimes it does. - The trouble is the timing. Nobody said that being a Scotsman meant not committing crimes until a Scotsman committed one, and that gives the game away. A standard written after the counterexample, for no purpose but to keep it out, measures nothing. (3) The name, and the family it belongs to. - Antony Flew, for decades Britain's best-known atheist philosopher, caught the move in God and Philosophy (1966) and named it in Thinking About Thinking (1975), where the Scotsman with the paper has no name. What matters in the name is timing: Hamish's definition moved after Aberdeen, never before. - An ad hoc rescue (Latin ad hoc, "for this") patches a claim for no reason except to save it. "My horoscope said I would meet a stranger, and I did not, but only because I stayed in." No True Scotsman is that move applied to membership. - Moving the goalposts changes what counts as winning after the shot is taken. "Show me a transitional fossil." "Here is one." "Now show me the one between that and the next." - A persuasive definition, Charles Stevenson's term from 1938, keeps a word's warm glow while changing what it refers to. "True art is whatever moves me" borrows the prestige of "art" for a private taste. Every "no true Scotsman" runs on one. Therefore, No True Scotsman never has to lie about the facts. It changes what the words mean after the facts are in, so the facts no longer count.

(●) Simple example: "No Scotsman puts sugar on his porridge." "My uncle Angus does." "No true Scotsman does." + The textbook version, and a sugar bowl is easier to laugh at than a crime. The shape is exactly Hamish's. (1) What is claimed and what is offered. - The claim is about every Scotsman's breakfast. The reply is a counterexample: a Scotsman by birth and passport, with his spoon in the sugar. That is the evidence the claim invited, and it is fatal to it. (2) Why the rescue does no work. - Angus has not stopped being Scottish. What has changed is the word. Try the swap: whatever Angus puts on his porridge, salt, whisky or ketchup, the same reply is waiting. A claim that survives every counterexample by the same trick was never about Scotsmen. - What would have worked is to give up the generalisation gracefully ("most Scotsmen, then"), or to show that Angus's Scottishness was already in doubt before breakfast. (3) What the reply was really doing. - It was protecting a picture of Scotland rather than describing one. That is what the fallacy is for: keeping a flattering picture of our own side, or a damning one of theirs, without paying what a counterexample charges.

(●) A critic says: "Stalin and Mao were not really acting as atheists. Communism was a religion." + Someone has just said that the twentieth century's worst killers were atheists, and the reply is that they were not really atheists at all. Christians make the same move about the Inquisition, and both times a real point is buried under a bad one. (1) What is being argued. - The claim is that atheism carries no blame for the gulags and the famines. The support is a reclassification: communism had prophets, scriptures, heresy trials and an infallible leader, so it was really a religion, and its crimes go on religion's ledger. - The published form is Christopher Hitchens's chapter on totalitarianism in God Is Not Great (2007). The communist regimes, he argues, did not abolish religion so much as replace it, with the party standing where the church had stood. North Korea, whose dead founder was made eternal president, he treats as a theocracy, a state ruled in the name of a god. The resemblance is real. (2) Where the fallacy sits. - Stalin was an atheist by every test anyone would have accepted the day before the question came up. He denied that God exists, he left the seminary that trained him without ordination, and his state closed churches and shot priests as policy. - To say he was "not really" an atheist because he behaved like a believer is Hamish reading about Aberdeen. The definition has quietly moved from "denies that God exists" to "denies that God exists and behaves well", and nothing prompted the move except the body count. - Try the mirror. "The Inquisitors were not really Christians, because Christianity is love." Every critic knows how that lands, and the two sentences have the same shape. - "No real scientist believes in God" is the same fallacy in a lab coat. A geneticist who prays is a scientist by any test stated in advance. (3) The sound version. - There is a real point here, and it should be granted in full: atheism as such entails nothing about ethics. "There is no God" is a denial rather than a creed, and it commands purges no more than "there is no Loch Ness monster" does. - What ordered Stalin's murders was Marxism-Leninism, a whole system of belief of which atheism was one plank. That is fair, and it is a different claim from saying Stalin was not an atheist. - So the honest conversation is about what did the killing rather than whose column it goes in. What did the killing was a belief that licensed force, held by men who answered to nobody, and that combination turns up in theocracies and atheist states alike. It is the thing worth arguing about. (4) How to answer it. - "I will grant that atheism by itself tells nobody to kill anybody. But Stalin was an atheist by any definition either of us would have given yesterday. Can we agree he was one, and then talk about what actually drove him?" See also: Natural Theology: The Moral Argument

(●) A Christian says: "No true Christian would have supported the Inquisition." + Critics wait for this sentence, because it costs a believer more credibility than almost anything else he can say. A faith that keeps its record clean by deleting the embarrassing members has kept no record at all. (1) What is being argued. - The claim is that the Inquisition's torture and burnings say nothing about Christianity. The support is that anyone who did such things was not really a Christian, whatever he said or sang. - The verse usually behind it is 1 John 2:19, "They went out from us, but they were not of us; for if they had been of us, they would have continued with us." The doctrine is real: some who wear the name do not belong to Christ, and Jesus said so himself (Matthew 7:21). What has gone wrong is the use. (2) Where the fallacy sits. - The Inquisitors were baptised, ordained, and acting on the church's authority. Spain's first Grand Inquisitor, Tomás de Torquemada, was a Dominican friar appointed with the pope's approval. By every descriptive test anyone would have offered the day before, he was a Christian. - The word "true" arrived after the counterexample, just as it did with Hamish. Nobody proposed "never tortures a heretic" as part of what "Christian" means until a Christian tortured a heretic. Meanwhile the monks who built that century's hospitals stay on the rolls. The membership line moves in one direction only, whenever the record hurts. - C. S. Lewis saw the trap in the preface to Mere Christianity. The word "Christian" must keep its plain descriptive sense, he argued, meaning someone who accepts the common Christian doctrines. Otherwise it goes the way of "gentleman", which once meant a man with a coat of arms and now means a man the speaker likes. When such a man lives unworthily of the doctrine, Lewis wrote, it is much clearer to call him a bad Christian than to deny he is one. (3) The sound version. - Christianity has a standard, fixed long before the Inquisition: the creeds, the Sermon on the Mount, and Jesus's own test, "You will recognize them by their fruits" (Matthew 7:16). The honest use of it is to own the record and judge it, not to strike the offenders off. The Inquisition counts as a Christian failure precisely because the standard condemning it is Christianity's own. - Owning the record also means sizing it honestly. David Bentley Hart argues that the popular death tolls for the Spanish Inquisition are wildly inflated, and that the institution served the crown as much as the church. Rodney Stark makes a contested case that the Crusades began as a defensive war. Whatever one makes of them, those are the right kind of argument: they argue about what Christians did, and never pretend the men who did it were not Christians. (4) How to answer a fellow Christian who says it. - "They were Christians, and they did it, and Jesus told us what to make of that: by their fruits. Let's judge them by our own standard instead of pretending they were never ours." See also: Common Objections: Did God Command Genocide?

(●) A charge that misses: "Saying the prosperity gospel is not Christianity is just No True Scotsman. You can't disown every group that embarrasses you." + The charge sounds like a fair application of Hamish's lesson, and sometimes it is. Whether it is depends on one question: when was the standard stated? (1) What the charge claims. - A Christian has said that a preacher selling private jets as God's reward, or a sect preaching hatred in Jesus's name, is not Christianity. The charge says this is Hamish and Aberdeen all over again: a group cut loose the moment it embarrasses the speaker. (2) Why the charge misses. - The fallacy is redefinition after the fact, and an appeal to a definition fixed in advance is what a definition is for. The Nicene Creed was written in 325, and the Sermon on the Mount three centuries before that. Both were on the table sixteen centuries before anyone preached that faith buys a jet. "That teaching contradicts the creed" applies a standard rather than inventing one. - The test to carry away: was the criterion on the table before the counterexample arrived, or made up to answer it? Hamish fails it, because "does not commit crimes" appeared only after Aberdeen. The creed passes, because its date is on the document. The test cuts both ways, too. A critic is entitled to call a communist state atheist: "denies that God exists" was on the table in advance as well, and no Christian gets to move that line. - Flew himself drew this line. His complaint was never against definitions, but against turning a claim the facts could refute into a tautology, a sentence true by the meaning of its words alone ("all bachelors are unmarried"). A standard that could in principle be failed, and was stated before the test, is no tautology, and judging conduct against it is how every tradition polices itself. (3) What the charge would need to be true. - It would need to show that the standard was invented for the occasion, or comes out only when the record hurts. On the second count the charge partly lands, and the concession is owed: Christians reach for the creeds when the prosperity preacher comes up and forget them when the Inquisitor does. - Notice what the creed and the Sermon actually do, though. They condemn what the prosperity preacher teaches, and they do nothing to remove the Inquisitor from the church, because by the creed he was in it. A standard fixed in advance is for judging members, never a door for showing them out. - So the line is narrower than either side would like. "That teaching is not Christian doctrine" can be a plain application of the creed. "That person was never a Christian" is almost always Hamish talking. (4) How to answer it. - "Fair challenge. Here is the creed, dated 325, and here is the Sermon on the Mount. If the preacher contradicts those, I am applying a standard, not inventing one. And I will hold myself to it: it condemns what the Inquisitors did, and it does not let me pretend they were never in the church." See also: Philosophical Theology: The Two Errors Nicaea Refused