Remove the stage, and see what is left.

For a hundred years, physics has staged the world inside a smooth continuum of space and time — a fixed arena on which fields ripple and particles move. The arena is assumed. Everything else is placed upon it.

IRMT asks a more unsettling question. What if the arena is not fundamental at all? What if space, distance, geometry — even the passage of time — are not the ground floor of reality, but patterns that emerge from something more primitive underneath?

The proposal is austere. At the bottom there are no coordinates, no distances, no particles. There is only a discrete web of events, related to one another by nothing more than order — which came before which — and the information carried along the links between them. From the statistics of that web, the theory contends, the familiar universe must be reassembled.

Four things, and nothing more.

The substrate is built from a small vocabulary. Its discipline is that each piece is defined precisely, and none of it presumes the world it is meant to explain. The labels we attach to events are bookkeeping only; physics is whatever survives when the labels are scrambled.

Trace the smallest closed loops in the web — the theory calls them causal diamonds — and compare the two ways information can travel around them. The mismatch is a single number that does not depend on how you labelled anything. Out of such mismatches, IRMT argues, the curvature of spacetime and the forces of nature are supposed to appear.

Five ways the world must be recovered.

If everything familiar is emergent, then every familiar structure has to be earned back — derived as a large-scale, statistical pattern of the network, not smuggled in by hand. IRMT divides that labour into five sectors. Each map below is generated from the audit ledger itself: a chain of what is assumed, what is proved, and where the argument still has to travel.

Where quantum probability comes from — and why the correlations between distant measurements can be strong, but not arbitrarily strong. Alongside the ceiling itself, the Born rule now has a route of its own: rank-one weights extended to rank-general projectors, then to full effects and mixed states. What remains open is the last step down — deriving the update axioms from the substrate rather than assuming them.

The forces of nature as the twisting of transports around loops. A second route now runs beneath the first: from two austere axioms, a minimum five-dimensional carrier, a balanced 3|2 split, and from there a charge lattice, one anomaly-free family, and a unique Higgs representation. The observed group SU(3)×SU(2)×U(1) remains a target rather than a result, and chirality selection is still open.

Gravity as geometry, and geometry as counting. This is now the most developed sector, not the least. An independent coframe, a Palatini elimination of the connection, and one jointly varied action yield exact finite Ward and Bianchi identities and a local recovery of Einstein–Cartan gravity. The whole structure closes on a single number — and that number is still measured, not derived.

Why the electron, muon, and tau have the masses they do — and why the quarks are harder. The charged-lepton construction closes on its own declared domain, through the Koide relation and a stationary phase. The quark problem has since been moved: mass is no longer read off a pole but is the restricted Hessian of one common action, and the next gate is to build that operator in its retarded form. Nothing numerical is claimed here.

The universe as a growing network, and a testable fingerprint. The background now follows from the same coframe action — Friedmann and Raychaudhuri equations, and exactly two tensor polarizations travelling at the speed of light. The large-angle suppression survives as one bounded signal from a calibrated kernel, not a derived prediction; the routes that promised more were closed by the program itself.

The rarest thing here is the bookkeeping.

What sets this program apart is not a bold claim but a refusal to make one loosely. Every statement in the specification wears a tag declaring exactly how much it is worth — whether it is a definition, a proven theorem, an imported result, a number from a computer, an educated guess, or an open problem. Failed approaches are not deleted; they are kept on the record so no one wanders down them twice.

The result is a document you can audit. For any conclusion, a reader can ask: what was assumed, what was proved, what was merely fitted — and what still has to be done.

These are the working ranges recorded in the ledger, not measurements. The program deliberately declines to publish a single headline number: recent releases note that overall percentages are left unrecalibrated, because a sector can advance a long way without any gate actually closing. By the program’s own binary standard the number of completed sectors is still zero. That candor is the point.

Sixteen papers, eight records, one frozen core.

The papers came one at a time, and each was a test — a result proposed, attacked, and, where it failed, revised or retired in the open. Read in order, Papers I through XVI are the trail of how the theory reached its present form. What they arrived at is gathered, corrected, and kept current in the Core Specification: the single authoritative statement of the theory. The papers show the road travelled; the specification is where it now stands.

“Not endorsement. Attack.”

The author is not asking to be believed. He is asking to be tested — for the sharpest readers to find the weakest joints in the argument and press on them. A theory that survives that is worth something. A theory that fears it is not.

• →The axioms beneath the quantum-probability route.
• →The gates that protect the construction from collapse.
• →The floors on manifold-likeness and chirality.
• →The CMB kernel, where theory meets real data.