The Preface Copernicus Never Wrote

In the spring of 1543, a book came off a printing press in Nuremberg that would eventually move the Earth. It was called On the Revolutions of the Heavenly Spheres, and its author, a Polish churchman and astronomer named Nicolaus Copernicus, had spent decades arguing something that sounded insane: that the Earth is not the still center of everything, but a planet, spinning on its axis and racing around the Sun.

Copernicus was dying as it printed. The story goes that a copy was placed in his hands on the last day of his life. He may never have read it. And it is a good thing, in a way, that he didn’t get to read the first page, because the first page betrayed him.

Right at the front, before Copernicus’s own words, sat a short unsigned note “to the reader.” It was calm and reassuring. It told anyone who was alarmed by this Earth-moving nonsense to relax. The astronomer, it explained, is not really claiming the Earth moves. He is just offering a handy piece of math. Listen to how it puts it:

For it is not necessary that these hypotheses should be true, or even probable; but it is enough if they provide a calculus which fits the observations.

Read that again, because it’s doing something sneaky. It says: don’t ask whether this picture is true. Ask only whether it works — whether the numbers come out right when you predict where Mars will be next winter. The note goes further and pulls the rug out from under the whole science: “let no one expect anything in the way of certainty from astronomy, since astronomy can offer us nothing certain.” Whatever Copernicus thinks he has found, the preface says, he has not found a fact about the world. He has found a convenient way to keep the books.

Here is the thing most people don’t know. Copernicus did not write that preface. He never saw it. It was slipped in, without his name and without his consent, by a Lutheran theologian named Andreas Osiander, who was helping shepherd the book through the press and got nervous about the storm it would bring. And it flatly contradicts the man himself. A few pages later, in his own dedication to Pope Paul III, Copernicus says plainly that he “attributes certain motions to the terrestrial globe” — that the Earth really does move — and that a scholar’s “loving duty is to seek the truth in all things.” He knew exactly how outrageous it sounded. He wrote that “the scorn which I had to fear on account of the newness and absurdity of my opinion almost drove me to abandon a work already undertaken.” He published anyway, because he thought it was true.

So the book opens with a quiet little war between two men. Osiander says: it’s only convenient. Copernicus says: it’s the truth. Both of them, it turns out, had hold of something real, and the disagreement between them is one of the deepest and strangest things you can notice about human knowledge.

What you already believe about truth

Before we go anywhere, let’s get clear on what you think right now. Because you have an opinion about the Osiander–Copernicus fight already, even if you’ve never heard of it, and your opinion is almost certainly this: Osiander was cheating. He was watering down a real discovery to avoid a fight. The Earth either goes around the Sun or it doesn’t, and it does, and Copernicus was right, and the mealy-mouthed preface was a dodge.

Underneath that reaction is a picture of truth that feels like plain common sense. It goes like this. Truth is out there. It is a feature of the world, not of us. The world is a certain way, whether we like it or not, and a statement is true when it matches the way the world is. We don’t invent truths; we find them, the way you find a coin under a couch cushion. The coin was there before you looked.

And the natural flip side of that picture is just as strong: if you didn’t find something out in the world, if you just decided it, then it isn’t really truth at all. It’s a convention. A handshake. An agreement. Driving on the right side of the road is a convention. So is calling this animal a “dog” instead of a chien. So are the rules of chess. These things are fine and useful, but nobody thinks they’re true the way “the Earth orbits the Sun” is true. They’re just things we agreed to. We could have agreed to the opposite, and driven on the left, and called dogs chiens, and let bishops move in straight lines. Conventions are arbitrary. Truths are not.

So you end up with two bins, and everything goes in one or the other. Bin one: real truths. Found, not made. Forced on us by the world. Not up to us. Bin two: mere conventions. Made, not found. Chosen by us. Arbitrary. Could’ve been otherwise. And the whole point of the distinction is that the bins don’t overlap. If a thing is in bin two, it is by that very fact not in bin one. To call something “a convention” is to say it is not a truth.

This is a good, clean, sensible way to carve up the world. Almost everyone carries it around. And it is about to spring a leak.

The turn: a truth you can’t discover and can’t just invent

Around the year 1900, a French mathematician named Henri Poincaré, often called the last person to understand all of the mathematics of his day, sat down to ask a simple-sounding question. Is the geometry you learned in school true?

You remember the geometry. Parallel lines never meet. The angles of a triangle add up to a straight line. The shortest path between two points is a straight line. This is Euclid’s geometry, laid down around 300 BC, and for over two thousand years people treated it as the clearest example of solid, necessary, discovered truth. If anything was a fact about space itself, this was.

Then, in the 1800s, mathematicians did something that should have been impossible. They built other geometries. In one, the angles of a triangle add up to less than Euclid says. In another, more. In these strange spaces, parallel lines curve toward each other, or away. And here was the shock: these other geometries were not sloppy or broken. They were perfectly consistent. You could not catch them in a contradiction. Poincaré even sketched a “fourth geometry” of his own where, he says, “a real straight line may be perpendicular to itself” — and it held together fine.

So now there’s a problem. If Euclid’s geometry is the true one, the way you always assumed, then the others must be false. But they can’t be caught being false. Nothing inside them ever goes wrong. So which is the true geometry of space?

Poincaré’s answer is the hinge of this whole essay:

The geometrical axioms are therefore neither synthetic à priori intuitions nor experimental facts. They are conventions.

Strip the jargon. “Synthetic à priori intuitions” is Immanuel Kant’s phrase for truths built into the mind, truths you can see must be so without ever checking. “Experimental facts” are things you go out and measure. Poincaré is saying Euclid’s rules are neither. You don’t know them just by thinking, and you don’t discover them by measuring. They are conventions — things we settled on. And then he drops the hammer:

What, then, are we to think of the question: Is Euclidean geometry true? It has no meaning. We might as well ask if the metric system is true, and if the old weights and measures are false… One geometry cannot be more true than another; it can only be more convenient.

Your two bins just collapsed into each other. Poincaré took the single most famous example of a discovered, necessary, forced-on-us truth. Euclid, the gold standard, the thing your teacher swore was just how space works. And he put it in bin two. He called it a convention. He said asking whether it’s true is as silly as asking whether the metric system is true, or whether inches are “false.”

(History would play a joke on Poincaré here, and the joke only helps him. He guessed we would always keep Euclid’s geometry, because it is the simplest. We didn’t. A few years later Albert Einstein handed physics a curved, non-Euclidean spacetime, and the bending of starlight as it passed the Sun, measured during a 1919 eclipse, bore him out. Even the convenient choice can change when the world insists on it.)

But wait. He is not saying geometry is a joke, or that math is made up, or that anything goes. That is the part everyone gets wrong, and it is exactly the part that matters. Poincaré is not demoting Euclid to arbitrary nonsense. He is telling you there is a third bin you didn’t know you had. There is a kind of thing that is chosen, like a convention, and yet is not arbitrary at all — a thing you pick, but are not free to pick however you like. It is a truth you cannot discover and cannot merely invent. And once you see that this third bin exists, you cannot unsee it, because it turns out to hold a shocking amount of what we call knowledge.

That is where Osiander and Copernicus were both right. Osiander was right that the moving Earth was, in one sense, a chosen frame, a way of drawing the map that made the numbers behave. Copernicus was right that it was not arbitrary — that there was a real sky under the maps, and it would get a vote. They were fighting because each of them saw one half of the third bin, and neither had the whole thing.

Now, the honest reader has a fair objection at this point, and we should say it out loud instead of hiding it. “Fine,” you say, “maybe a few abstract things like geometry live in this weird third bin. But that’s a special case. Most truths are still just plain facts. You’ve found one strange exception, not overturned the whole picture.” Good. That is the right worry. So let’s put it under pressure, and watch the third bin get bigger, and bigger, until it starts to swallow things you were sure were solid ground.

Movement one: the door swings open

The first thing that happens when you admit that some truths are chosen is that a very old voice starts whispering: then why not all of them?

That voice belongs, more than to anyone, to Michel de Montaigne, the French essayist writing in the 1570s and 80s. Montaigne had read the ancient Greek skeptics, the followers of a philosopher named Pyrrho, and he described their favorite move with obvious delight. The Pyrrhonist, he says, will take any claim you make and calmly argue the reverse, not because he believes the reverse, but to show you that your certainty is a bluff:

If you postulate that snow is black, they argue on the contrary that it is white. If you say that it is neither one nor the other, it is up to them to maintain that it is both.

The point is not really about snow. The point is that for any belief you hold, a clever enough person can build a case against it, and you will not be able to prove them wrong to their satisfaction. Montaigne pushes this at the grandest scale he can find, and — this is almost eerie — he reaches for Copernicus himself. Writing just decades after that Nuremberg printing, Montaigne says the Earth-moving idea has been “grounded so well” that it now runs the calculations smoothly. And then:

And who knows whether a third opinion, a thousand years from now, will not overthrow the preceding two?

He means: people were once sure the heavens moved and the Earth stood still. Now here comes Copernicus, sure of the opposite. What makes us so confident that a future thinker won’t sweep both views away? He hammers the warning home with the ancient astronomer Ptolemy, who had mapped the limits of the known world and gotten them wrong. It would have seemed like madness, Montaigne says, to doubt Ptolemy’s geography a thousand years ago — “it was heresy to admit the existence of the Antipodes,” to believe there was land and people on the far side of the globe. Then explorers sailed over the horizon and found a whole new world Ptolemy never dreamed of. The experts were certain. The experts were wrong. Why should today’s experts be any different?

This is the door swinging open, and it swings the way you didn’t want it to. We started with a narrow, technical point about geometry. Montaigne widens it into a lifestyle. If Euclid was a convention, and Ptolemy’s confident world-map got overturned, and any claim can be argued both ways, then maybe the whole business of “truth” is just the opinions that happen to be fashionable right now, holding their ground until the next fashion knocks them over. Montaigne even names the machine that installs our beliefs in the first place, and it isn’t reason. It’s habit — “a violent and treacherous schoolmistress,” he calls her, who teaches us to find our own customs obvious and everyone else’s bizarre, when they are all equally accidents of where we happened to be born.

You can feel the floor tilting. Let’s let it tilt further.

Movement two: truth with an expiration date

Montaigne says our truths might be temporary. A novelist can show you that they are — that the graveyard of expired certainties is real, and crowded, and that everyone buried there was as sure as you are now.

In the second epilogue of War and Peace, Leo Tolstoy stops telling the story and starts arguing with the historians. Look, he says, at how they judge a figure like Alexander I, the Russian tsar who fought Napoleon. Historians praise some of his choices and condemn others. But on what scale? Their idea of what is “good for humanity.” And that idea, Tolstoy notices, will not sit still:

every year and with each new writer, opinion as to what is good for mankind changes; so that what once seemed good, ten years later seems bad, and vice versa.

Then he springs the trap. The historian sits in judgment on the tsar, certain he knows what the tsar should have done. But:

even if we assume that fifty years ago Alexander I was mistaken in his view of what was good for the people, we must inevitably assume that the historian who judges Alexander will also after the lapse of some time turn out to be mistaken in his view of what is good for humanity.

Sit with how sharp that is. The judge is standing on the exact same ground as the man he’s judging. He is certain, and the man he condemns was certain, and in fifty years someone will condemn him with equal certainty, and be condemned in turn. There is no seat in the courtroom that isn’t itself on trial. Truth, on this view, comes with an expiration date stamped on it, and you can never read your own.

The playwright George Bernard Shaw put the same thought more bluntly. History, he wrote, “is always out of date.” Children are taught the safely settled version of the past and fed convenient falsehoods about the present: in his example, they “are taught history about Washington, and told lies about Lenin.” Every generation thinks it has finally cleared away the myths and reached the facts. Every generation is the myth the next one clears away.

Now the reader who came in believing in plain, found, out-there truth is in real trouble. We have gone from “geometry is a convention” to “maybe everything is opinion” to “and the record of the past proves that even our most confident truths curdle into errors within a lifetime.” The skeptic seems to be winning. If you stopped the essay here, you would walk away believing that truth is just a story your era tells itself, and that Osiander’s shrug — it’s not true, it’s only convenient — was the honest adult position all along.

But we are not going to stop here, because the skeptic has been allowed to cheat. He has been treating every kind of claim as if it were the same kind of claim. And it isn’t. This is where the fall stops, and the climb back begins.

Movement three: not everything belongs in the same bin

Here is the trick the skeptic pulled, and once you see it you can’t stop seeing it. He lined up “the Earth orbits the Sun” next to “this custom is disgusting” next to “she is beautiful” next to “murder is wrong,” and he waved his hand over the whole row and said: look how people disagree, look how it all changes, it’s all just opinion. But that row is a con. Those claims do not behave alike at all.

Plato caught this 2,400 years ago. In the dialogue Phaedrus, Socrates points out that some words drop the same picture into every mind, and some words start a fight:

When any one speaks of iron and silver, is not the same thing present in the minds of all?… But when any one speaks of justice and goodness we part company and are at odds with one another and with ourselves?

Say “iron” and every person in the room is thinking of the same stuff. Say “justice” and the room splits. Socrates’ point is that these are two different kinds of subject. There’s a class of things where agreement comes easily because the world is pinning us all to the same answer, and a class where it doesn’t. The skeptic’s whole trick was to smear these two classes together. Pull them apart and his evidence falls in half. Yes, people disagree wildly about beauty and justice. No, they do not disagree about whether iron is heavier than air.

And look what happens to the skeptic’s best examples once you sort them. Take beauty, which Montaigne loved to use as proof that everything is relative. He piled up the evidence gleefully: standards of beauty are different everywhere. “The Indies paint it black and dusky,” he writes, with wide flat noses; in Peru “the biggest ears are the fairest,” and people stretch them on purpose; elsewhere they blacken their teeth, or prize a tiny forehead. Beauty, clearly, is in bin two. It’s chosen, local, a custom. And Montaigne even tells you how he knows it’s a convention:

If there was any natural prescription for it, we should recognize it in common, like the heat of fire.

That is a beautiful line and it gives away the whole game — in your favor, not his. Montaigne is admitting there’s a test. If something were really out there in the world, everyone would feel it the same, “like the heat of fire.” Fire is hot for the Inca and the Belgian alike; nobody has a local opinion about touching a flame. So the variation in beauty is real evidence that beauty is at least partly up to us. Fine. But that same test, turned around, tells you that the things which don’t vary — that fire is hot, that iron sinks, that the Sun does not orbit a stationary Earth — are the ones the world is pinning down. The skeptic’s own yardstick separates the chosen from the found. He just never turned it around.

This is worth making concrete, and there’s a perfect old fight for it. In Plato’s dialogue Cratylus, two men argue about names. Hermogenes takes the convention side, and states it as cleanly as anyone ever has:

There is no name given to anything by nature; all is convention and habit of the users.

Change a slave’s name, he says, and the new name is just as correct as the old. Words are pure agreement. His opponent, Cratylus, takes the opposite extreme: names are true by nature, there is a real, correct name for each thing, “the same for Hellenes as for barbarians,” and a wrong name isn’t really that thing’s name at all. So you have the two bins personified. Hermogenes: it’s all convention. Cratylus: it’s all discovered truth.

And Socrates, in the middle, refuses both of them. He grants Hermogenes that of course we chose these particular sounds — there’s nothing in the barking animal that demands the noise “dog.” But he presses Cratylus’s worry too: if names were pure convention, with no answering to anything real, then you couldn’t say anything false, and you couldn’t say anything true either, and language would just be noise, “an unmeaning sound like the noise of hammering at a brazen pot.” A name, Socrates decides, is like a picture. The paint is our choice. But a picture can still be a good likeness or a bad one, a true portrait or a false one, and that part is not up to us. “We must find,” he says, “some other principle of truth in images, and also in names.” Some other principle. A third bin. Chosen materials, but a real fit or misfit with the world underneath. Twenty-four centuries before Poincaré, the same shape.

So the skeptic’s mistake has a name now. He sorted nothing. He threw geometry and dinner customs and the moving Earth into one pile and called the pile “opinion.” But some of those things are pinned by the world and some are handed to taste, and a few strange ones — geometry, names, the frame you measure motion in — are the genuinely interesting cases that are both. Which raises the question we’ve been dancing around. What is that “both”? What is it for a thing to be chosen and constrained at the same time? For that, we need to look hard at one small, boring, world-changing object.

Movement four: relative is not the same as arbitrary

There is a metal bar. For a long time it lived in a climate-controlled vault outside Paris, made of platinum and iridium, and it was the meter. Not a copy of the meter. The meter. If you wanted to know how long a meter was, the final answer was: as long as that bar. Every ruler and tape measure on Earth traced its authority back to that one lump of metal.

Now ask Poincaré’s favorite question about it. Is the meter true? Did somebody discover that a meter is exactly that long? Obviously not. Nobody found the meter under a cushion. We made it up. The bar’s length is a choice, and we have changed our minds and re-chosen it more than once. Today a meter is defined as the distance light travels in one 299-millionth-and- change of a second. You cannot do that to a discovered fact. You can’t hold a committee and redefine how far the Earth is from the Sun. So the meter is in bin two. It’s a convention, plain as day.

And yet. Try to actually use an arbitrary meter and watch what happens. If I decree that my personal meter is however long my arm happens to be today, my measurements stop agreeing with yours, my bridge doesn’t line up with your bridge, and the whole point of measuring collapses. The meter is chosen, but it is not chosen freely. It has a job to do — make the world’s measurements agree, stay put so a beam cut in Tokyo fits a slot machined in Detroit — and the job rules out almost every possible choice. That is why we went to the trouble of pinning it to the speed of light, which doesn’t drift and doesn’t warp and is the same for everyone. The choice is real. The constraint is also real. Both at once.

This is exactly what Poincaré said, in the sentence people skip when they quote him. Having called the principles of physics conventions, he immediately slams the brakes:

This convention, however, is not absolutely arbitrary; it is not the child of our caprice. We admit it because certain experiments have shown us that it will be convenient.

Not the child of our caprice. Caprice means whim, doing whatever you feel like. Poincaré is carving out the third bin precisely: a convention that is chosen but not capricious, picked by us but guided by hard facts about what actually works. The word we’ve been missing this whole time is the one that fixes everything: relative is not the same as arbitrary.

An arbitrary thing could be anything; nothing constrains it. A relative thing is fixed once you say relative to what. The meter is relative to a purpose — measuring, agreeing, building — and once you name the purpose, most “choices” are ruled out instantly. Which side of the road to drive on is genuinely arbitrary; left works as well as right, and some countries prove it. But whether to let your meter be your arm-length is not a free choice at all, even though the meter is “just a convention.” The skeptic in movement one won his scary conclusions by sliding from “chosen” straight to “arbitrary,” from “relative to us” straight to “up for grabs.” They are not the same move. Almost nothing important is arbitrary. A great deal is relative. And relative things have right and wrong answers the moment you fix the frame.

Now go back to Copernicus, because his fight with Osiander is subtler than “realist beats coward,” and the honest version is better. Here is the part the triumphant story leaves out: in 1543, the world had not yet voted. Copernicus’s system was not actually more accurate than Ptolemy’s. It still ran on perfect circles stacked on circles, with no ellipses; it did not predict the planets any better; and it carried a real problem of its own. If the Earth truly races around the Sun, the fixed stars should shift back and forth over the year, and no one could detect the shift. No one would, until 1838. By the evidence you could actually gather in 1543, choosing the Sun-centered frame over the Earth-centered one was genuinely undecided. This is not a knock on Copernicus. It is the whole point standing up in the flesh: heliocentrism began its life exactly where Osiander put it, in the zone of chosen hypotheses the evidence does not yet settle. Osiander was right about 1543. And the instrumentalist reading he gave — treat it as a handy calculation, don’t claim it’s true — has never lacked serious defenders. The historian Pierre Duhem built a whole case that “save the appearances, don’t overreach” is the more careful stance, precisely because the evidence so often underdetermines the theory.

Where Osiander went wrong was not about his own moment. It was about the future. He treated “the evidence doesn’t settle it yet” as if it meant “there is nothing here to settle, only convenience, forever.” And that is the skeptic’s slide once more: from “undecided now” to “merely conventional always.” Because the world does keep voting. The phases of Venus, which Galileo saw through a telescope in 1610, killed the old Earth-centered scheme outright. Kepler traded the circles for ellipses and the numbers finally came clean. Newton supplied the single force that ran the whole machine. Bit by bit the frontier between chosen and found moved, and heliocentrism crossed it, from a convenient hypothesis into a fact you’d be a crank to deny. Copernicus’s confidence was a bet, not a proof he could have shown you in 1543. It happened to be a winning bet, because there was a real sky under the maps, casting real votes, on its own schedule. That is the third bin in motion: not a fixed border with truth safe on one side, but a line the world is always redrawing as its evidence comes in.

Which leaves exactly one question standing, the biggest one, the one the skeptic will fall back on when everything else is stripped away. Okay — but what stops this from being turtles all the way down? If truths are chosen frames, isn’t “the world constrains our choices” itself just another chosen frame? Can’t the skeptic say the whole thing, top to bottom, is convention — including your precious constraint? This is his last move. It’s also where he finally cuts his own throat.

Movement five: the branch the skeptic is sitting on

The skeptic’s final position is the biggest one he can take: everything is convention. Every truth, all the way down, is just something we agreed to; there is no bedrock, no world pinning anything, only frames resting on frames resting on nothing. It sounds unbeatable, because whatever fact you name to stop the slide, he says “that’s a convention too.”

The trouble is that this position destroys itself, and the great books have been pointing at exactly how for a very long time.

Start with the cleanest version, from Plato’s Theaetetus. There was a famous teacher named Protagoras who said that “man is the measure of all things” — meaning whatever seems true to you is true, for you, and that’s all truth ever is. “What seems to a man, is to him.” It’s the ancient ancestor of “everything is opinion.” Socrates lets him have it, then quietly turns the knife:

the best of the joke is, that he acknowledges the truth of their opinion who believe his own opinion to be false.

Follow it slowly, because it’s fatal. Protagoras says every opinion is true. But most people believe Protagoras is wrong. By his own rule, their opinion is true. So by his own rule, he is wrong. The claim “everything is just opinion” is refuted the instant anyone disagrees with it, because the claim itself grants the truth of the disagreement. It has no floor to stand on. It kicks away its own floor as its first act.

A determined relativist has one escape here, and it is worth naming. He can refuse to say “every opinion is true” flat out, and say instead “every opinion is true for the one who holds it,” relativizing every claim so the contradiction never quite closes. Fine. But watch what it costs him. Once “true” always means “true-for-me,” he can no longer tell you that you are wrong about anything, ever — not about the well, not about the Earth, not about him. His grand thesis stops being a claim about the world and shrinks to a report on the weather inside his own head. He has saved the doctrine from contradiction by draining it of every use. Nobody can live in it, and nobody tries.

Aristotle drives at the same target from the side of action instead of logic, and it’s the most down-to-earth argument in all of philosophy. You say you believe nothing is really true or false, that it’s all the same? Fine, says Aristotle:

why does a man walk to Megara and not stay at home, when he thinks he ought to be walking there? Why does he not walk early some morning into a well or over a precipice, if one happens to be in his way?

The total skeptic does not, in fact, stroll off cliffs. He steps around the well like everyone else. His feet believe in a real world with real edges even while his mouth says there isn’t one. You cannot actually live the “everything is opinion” life; the moment you get hungry, or dodge a bus, you are betting your body on truths you claim not to have.

The careful skeptic has a reply ready, the one we met back with the Pyrrhonists: I don’t step around the well because I believe in some absolute truth. I step around it because that’s how things appear, and I follow appearances without swearing to them. True enough. But push on why the appearance is worth following. You trust the look of the well because, again and again, wells that look solid are solid, and the ones you ignored broke your leg. The appearance earns its authority only by answering to a real well underneath. Strip out the real well and there is no reason the appearance should keep you alive — yet it does, reliably, which is the tell that something real is anchoring it. The skeptic gets to skip the metaphysics. He does not get to skip the well.

And notice, while we are here, that Aristotle has a second blade. Not all wrong answers are equally wrong. As he puts it, “he who thinks four things are five” is not “equally wrong with him who thinks they are a thousand.” One is a little off and one is wildly off, and that comparison only means anything if there is a real answer both are missing by different distances. Being more or less wrong exists only if there is a true north to be off from. As Aristotle says, “there must be some truth to which the more true is nearer.”

Now here is the part that clinches it, and it comes from the very people you’d expect to be on the skeptic’s side. The radicals themselves refuse to go all the way. Poincaré — the man who called Euclid a convention, who started this whole avalanche — turns around and personally strangles the “everything is convention” position. He gives it a name, Nominalism, and calls it a fantasy:

Some philosophers have generalised far too much. They have thought that the principles were the whole of science, and therefore that the whole of science was conventional. This paradoxical doctrine, which is called Nominalism, cannot stand examination.

Why can’t it stand? Because, Poincaré explains, no matter how much of science you show to be chosen convention, you never reach the bottom. Peel away every convention and there is always a residue of raw fact left over, the stubborn experimental stuff the conventions were built to organize. “It is clear,” he writes, “that however far this decomposition may be carried, laws will always remain.” There is a floor. He should know; he mapped the whole building. And earlier, in the same breath as all his talk of conventions, he had already said the thing the skeptic can’t afford to hear: “Experiment is the sole source of truth.” The world gets a vote. The world always gets a vote. That is the constraint that is not itself a convention.

Even Friedrich Nietzsche, who went further toward “truth is a useful fiction” than almost anyone, runs into the same wall. His famous, dangerous line is that “the falseness of a judgement is to us not necessarily an objection to a judgement” — that some of our most useful beliefs might be false, and we keep them because “to recognize untruth as a condition of life” is the brave thing. Bracing stuff. But notice what it smuggles in. To say a fiction is useful, that it “preserves life,” you have to help yourself to a real life that really gets preserved, a real world in which some fictions work and others get you killed. The moment Nietzsche says a belief is life-serving, he has leaned on a reality doing the serving-or-not. Kick truth out the front door and it climbs back through the window wearing the word “useful.” You cannot even state radical skepticism without standing on the solid ground you are denying.

So the third bin is safe, and its floor is real. We can hold, without contradiction, all of the following at once, though it sounded impossible at the start: Euclid’s geometry is a convention, not a discovered fact. The meter is chosen, not found. Which frame you measure motion in is up to you. And the world constrains those choices so hard that most of them are wrong, that some maps get thrown away, that “the more true is nearer” some real truth, and that anyone who denies all of it will still, when the time comes, step around the well. The line between what we choose and what the world hands us is real. It moves — Euclid crossed it one way, the moving Earth crossed it the other — but it is never erased, and it is not itself one of the things we get to choose. That last fact is the bedrock the skeptic keeps standing on while he insists there’s no such thing as standing on anything.

Back to Nuremberg

Return, now, to the dying man and the stolen preface.

For centuries, Osiander’s note was read as cowardice — a theologian defanging a dangerous truth to keep the peace. And there’s something to that; he hid his name, and he put words in a dead man’s book. But we can read it now with different eyes. Osiander had actually noticed something real and deep. In 1543, the moving Earth genuinely was a chosen frame, a way of drawing the heavens that made the math come clean, and the evidence that would decide the matter did not yet exist. About his own moment, he was right, and the calculating-tool reading he gave had honest defenders then and has them now.

His mistake was not about 1543. It was about forever. He treated “the evidence hasn’t settled this yet” as if it meant “there is nothing here to settle, only convenience,” and so he slid from a chosen frame to a merely convenient one, from “undecided” to “not the kind of thing that can be true.” He forgot that the world keeps voting, that relative is not arbitrary, that a line which has not yet been drawn is still going to be drawn. And it was. Venus, Kepler, Newton — vote after vote came in, and they came in for Copernicus.

Which is why Copernicus, insisting on plain truth against a preface he never authorized, was right about the deeper thing. He could not have proved it to you in 1543. He staked his reputation and the last of his life on a bet: that his picture was not a bookkeeping trick but was closer to what is. The two men were not really enemies. They were the two halves of an answer neither could complete alone — the frame is chosen, and the choice is not free. Put their names together and you get the shape of nearly all human knowledge: a made thing that answers to an unmade world, on the world’s own schedule.

The map on your wall is not the territory, and it was never trying to be. Every map distorts; every one is a choice, drawn for a purpose, and could have been drawn otherwise. But there is a territory. And that is the whole reason some maps get folded up and thrown away.