Every Contradiction
Has a Parent.

Einstein did not discover that the ether was absent. He discovered that the equations did not need it, and removed it — and the removal is the single event from which every paradox in this act descends.

Lorentz had the same transformations first, with a medium underneath them and rods that shortened because they were moving through something. Einstein's version threw the something away and kept the arithmetic. On every measurement made since, the two agree. That fork was never closed by experiment — it was closed by preference, and it is still standing open at the far end of this page.

What follows from the removal is bookkeeping about the word now. Two clocks, no medium, and no fact about which of them is at rest: the twin who comes home younger needs an asymmetry the theory just abolished, so it gets smuggled back through acceleration. The pole fits in the barn and does not fit in the barn. Only the two-way speed of light is ever measured; the one-way speed is a convention you choose, and the choosing is invisible in the results.

Then Paul Ehrenfest spun a disk, and the bookkeeping broke. A rigid body in rotation is not a puzzle about observers — the rim is contracted and the radius is not, so C/2R ≠ π, and no choice of frame repairs it. This one is a contradiction, and Einstein took it seriously enough that it walked him out of flat geometry altogether. The rigid body was the casualty. The continuum was the replacement, and the continuum is Act Two.

In 1965 Penrose proved that the infinity at the centre is not an artefact of bad coordinates. It is generic. A theory had produced, from its own machinery, a theorem stating where it stops being true.

Einstein disliked singularities and spent real effort trying to argue them out of physical existence. He lost that argument to his own field equations. The same theorem class, extended backwards, puts one at the beginning of the universe — so the two boundaries of cosmic history are both places where the description hands in its notice.

Then the horizon got a temperature. Hawking's 1974 calculation made black holes radiate, and the radiation came out exactly thermal — carrying no imprint of what fell in. A pure state goes in; a mixed state comes out; unitarity, the one thing quantum mechanics never negotiates, is violated. That is a contradiction between the two best theories physics owns, and it has stood for fifty years.

Every attempt to close it charges a fee. Entangle the late radiation with the early radiation to save unitarity, and monogamy of entanglement forces a wall of hard quanta at the horizon — which incinerates the equivalence principle, the seed idea of the whole theory. Quantise gravity canonically and time itself drops out of the equation: Ĥ|ψ⟩ = 0, the wavefunction of the universe does not evolve, nothing happens. And in 1949 Gödel handed Einstein a birthday present — an exact rotating solution containing closed timelike curves. Einstein's published reply worried, in print, about whether such solutions ought to be ruled out on physical grounds. Nothing in the theory rules them out.

The cat was never a mascot. Schrödinger built it in 1935 as an objection — a burlesque, his word — to show that if you take linearity seriously it walks straight out of the laboratory and into the furniture. The field adopted the objection as a logo and kept the premise.

Einstein never accepted it, and the thing he refused was not the arithmetic — it was the ontology. His EPR argument was an offer: either the description is incomplete, or the world is nonlocal. Pick. Thirty years later John Bell turned that offer into a number, and the number went to experiment, and the experiments came back saying no local account reproduces these correlations. The measured world is nonlocal or it has no definite values before you look. There is no third door.

Bell also said, flatly, what most of the field would only mutter: that ordinary quantum mechanics is unprofessionally vague and ambiguous, working for all practical purposes while leaving the central question unanswered. He asked why the pilot-wave picture — an existing, working, nonlocal, deterministic account — was left out of the textbooks, and got no satisfying answer.

Hugh Everett removed the collapse law entirely. It works, and the bill is every outcome happening in a branch of its own. Bohr's circle dismissed him, his thesis was cut down before publication, and he left physics for defence analysis and never came back. Sixty years on the problem sharpened rather than dissolved: Frauchiger and Renner showed in 2018 that if quantum theory applies to observers who use quantum theory, rational agents derive contradictory facts about a single event. The theory cannot consistently describe its own use.

Both halves of this act are true at once, and that is what makes it the strangest chapter in the census. Nothing humans have ever built predicts better. Nothing humans have ever built asks to be believed less by the people who built it.

Dirac would not sign off on renormalisation. He called the procedure not sensible mathematics — infinities discarded arbitrarily because the answer comes out right if you discard them that way. Feynman, who won a Nobel Prize for the machinery, called it a dippy process and a shell game, and said the fact that it works has not stopped it from being hocus-pocus. Neither of them was being modest. They were pointing at the continuum.

Then the numbers. Nineteen to twenty-six free parameters — masses, mixing angles, couplings — every one measured, not one derived. Three generations of matter for no reason. Yukawa couplings spanning a factor of a million with no pattern. The Higgs mass held at 125 GeV by a cancellation running to one part in 10³⁴. The QCD angle θ, which has no business being small, measured below 10⁻¹⁰. And the vacuum's own energy, which the theory computes and the sky contradicts — by about 10¹²⁰, the worst quantitative miss in the history of the subject.

Pauli set the tone for the whole century of it in 1930. Needing to save conservation of energy in beta decay, he invented a particle he believed could never be detected, and wrote that he had done a terrible thing. He was right about the terrible thing and wrong about the detection — and the neutrino came back in the end to break the model that inherited it, because the Standard Model says the neutrino is massless and the oscillations say it is not.

This act's entries are not contradictions. They are something a physicist can find harder to live with: a structure so accommodating that no measurement can disagree with it.

The mathematics is real and often beautiful, and the programme has paid out elsewhere — holography, black-hole entropy counting, tools that migrated into nuclear and condensed matter physics. What it has not produced, after four decades, is one distinctive number at one reachable energy. The string scale sits near Planck. The superpartners that the superstring requires did not appear at the LHC, so the breaking scale is pushed up again, as it has been pushed up before, with no principle setting where it stops.

Susskind's response was to take the landscape seriously and go anthropic: with 10⁵⁰⁰ vacua available, ours is the one compatible with observers, and that is the whole of the explanation. Gross has spent years refusing that trade, on the ground that it converts physics into a story about our own presence. Both are looking at the same fact. Neither has a way to break the tie from inside the theory.

Two structural bills sit under all of it. Perturbative strings are defined on a fixed background, which is precisely what general relativity says a theory of gravity must not need — and the non-perturbative definition that would fix this does not exist yet. And the corner of the theory that is actually under control is anti-de Sitter: negative cosmological constant, the wrong sign for the universe we live in.

Milgrom's law does something no dark-matter halo does: it predicts individual rotation curves from the visible matter alone, one number for the whole population, and it keeps being right about galaxies it was never fitted to.

The number is a₀ ≈ 1.2 × 10⁻¹⁰ m/s², and it was set by hand. Worse — or better, depending on where you think physics is going — it lands suspiciously close to cH₀/2π, a cosmological quantity that has no business appearing in the dynamics of a single galaxy. MOND cannot say why. It offers a force law, not a derivation of its own scale, and a scale you cannot derive is a scale someone else's theory will have to explain.

Then the failures, and they are honest ones. Clusters still need roughly twice the mass MOND can supply — the missing-mass problem is reduced, not removed. The Bullet Cluster shows lensing mass sitting away from the gas, which is the textbook argument against modifying the force law alone. Internal dynamics depend on the external field a system happens to sit in, so the strong equivalence principle goes. And there is no cosmology in it: no acoustic peaks, no nucleosynthesis, no structure growth.

The hardest blow was a single event. In August 2017 a neutron-star merger arrived as gravitational waves and as light, and the two arrived together to within a part in 10¹⁵. The relativistic completions of MOND had bought their extra dynamics with extra fields, and those fields made gravity travel at a speed other than c. One measurement, one afternoon, and that whole family of theories was ruled out.

Fritz Zwicky counted the Coma cluster in 1933, found the galaxies moving far too fast for the light they showed, and named what was missing dunkle Materie. The field ignored him for forty years, largely because he was difficult and because the measurement was hard to believe.

Vera Rubin made it impossible to ignore. Spiral after spiral, the rotation curves went flat where they should have fallen, and the discrepancy was not a marginal effect but a factor of ten. She said the ratio of dark to light matter is about the same as the ratio of our ignorance to our knowledge, and she was not being modest either.

Forty years of direct detection have now been run — deeper and deeper underground, tonne-scale targets, the sensitivity improving by orders of magnitude — and the result is null. Twenty-seven per cent of the universe is inferred from gravity and nothing else. Sixty-eight per cent more is a name attached to a number, w ≈ −1, with no microphysics behind it. And the same vacuum energy that Act Four could not compute is here again, off by 10¹²⁰, and additionally tuned to become important at roughly the moment we happen to be looking.

The tensions have not softened with better data — they have hardened. The local distance ladder says H₀ ≈ 73; the microwave background, run through the model, says ≈ 67. The gap is past five sigma and neither side has found the error. Lensing surveys find less structure than the CMB requires. On small scales the cold-dark-matter simulations predict cusps where cores are observed, swarms of satellites that are not there, and the satellites that are there sit in thin co-rotating planes that random debris has no business forming. BBN predicts three times the lithium-7 the old stars carry. JWST keeps finding galaxies too massive and too settled too early.

Inflation was invented to repair the horizon and flatness problems, and Alan Guth's idea did the job so elegantly that it became standard within a decade. Paul Steinhardt helped build it and now argues in public that as practised it cannot be falsified — the measure problem is unsolved, eternal inflation produces every outcome somewhere, and the initial conditions the inflaton itself needs are as special as the ones it was hired to explain. A fix that needs fixing.

Run the seven acts backwards and a pattern falls out that has nothing to do with anyone's cleverness. Every contradiction in this census is the price of a premise, and the premises are all removals or inventions made a century ago for reasons that were good at the time.

SDT does not hold them. There is no wavefunction to collapse, so there is nothing for a measurement problem to be about. There is no continuum without a floor, so a collapse has somewhere to stop and the infinities have nowhere to grow. There is no vacuum carrying zero-point energy, so there is no quantity that can be 10¹²⁰ wrong. There is no attraction sourced by mass, so there is no missing mass — the shadow does the holding and the shadow is cast by what is already there. And there is a frame: the lattice, and the tick, and a now that is the same count everywhere. The present that special relativity abolished is back, so the block universe, the conventional simultaneity and the disappeared time have nothing to stand on.

Not by construction. That phrase should be struck wherever it appears, including in the census this page walks. You cannot build a theory that dodges contradictions — there is no such move available to anybody. You can only run the work and read what comes back. Nothing was suppressed either: the exclusions here are exclusions of inputs — no G, no M, no wavefunction, no dark sector, no free parameter to fit — and refusing yourself inputs does not protect you from contradictions, it makes them far easier to find. Fewer knobs, fewer places to hide.

The count is zero because zero fell out. Hydrogen closes. The deuteron binds at 2.200 against a measured 2.224 MeV with no fitted scale. The closure schedule prints 2, 8, 20, 28, 50, 82, 126 from two closed forms and then keeps going to 184. A clock at the Sun's surface runs slow by exactly the depth it sits at, to three parts in ten thousand. A galaxy turns. At no point does one result turn round and contradict another, and at no point does the machinery forecast the place where it stops being true. No paradox exists because no paradox exists. None fell out. There was never one there to remove.

The hunt was run properly: the framework was pushed at the places it is thinnest, and what came back was debts — unfinished derivations — not contradictions. Those are on the next screen, named, with what each one owes.