Darwin published a book. Five hundred pages. One diagram. A single mechanism — natural selection — that explained the diversity of life without a designer. But Darwin could not hand his reader a test tube and say: run this experiment, watch selection happen, verify the mechanism with your own eyes. The evidence was in fossils, in finch beaks, in the distribution of species across islands — evidence the reader had to trust the author had collected correctly. This paper is different. The claim is not "natural selection explains the fossil record." The claim is: two operations — merge what is similar, prune what is not reinforced — produce centroids. Centroids perturb a shared field. Centroids precipitate across generations. Centroids that survive all three generations are VALUE. And the claim comes with six Python files. Zero dependencies. Three thousand lines. Anyone can download them. Anyone can point the instrument at their own stream — their own Bach, their own ECG, their own diplomatic record — and see structure emerge without being told what structure is. Darwin gave the world a theory. The code gives the world the theory, running.

Every scientific theory before this one was a description. The reader had to trust the author. Newton described gravity — but the reader could not hold the inverse-square law in their hand and watch an apple fall through it. Maxwell described electromagnetism — but the reader could not run the equations on their own telegraph wire. Darwin described natural selection — but the reader could not breed a new species in their living room to verify the mechanism. The theory was in the book. The evidence was in the world. The gap between them was the reader's trust.

This paper closes the gap. The theory is the code. The code is the evidence. Download six files. Install nothing. Feed the instrument a stream — any stream, from any domain, in any encoding. The instrument does not know what a heartbeat is, what a dominant chord is, what a diplomatic vote is. It knows two operations. Merge. Prune. And from those two operations, centroids emerge. Structure emerges. The reader does not need to trust the author. The reader can run the instrument on their own data and see for themselves.

This is not a small difference. It is the difference between the nineteenth century and the twenty-first. Darwin's readers waited decades for independent confirmation. Mendel's work was rediscovered after his death. Wegener died before plate tectonics was accepted. The gap between publication and verification was measured in careers — sometimes in lifetimes. The gap between this paper's publication and its verification is measured in the time it takes to run a Python script. The forward prediction on UN diplomatic voting — 2026-2028, falsifiable, three criteria — will be tested not by a generation of scientists but by the calendar. The answer arrives within three years. The reader can run the annual ideal-point pipeline on the 2026 data the moment it appears and see for themselves whether the displacement returned below 0.20 or stayed elevated.

This is what it means for a theory to be formal in the age of computation. Not formal in the sense of mathematical proof — formal in the sense of executable. The axioms are not sentences in a logic. They are lines of code. The deductions are not theorems. They are program traces. The evidence is not a table of numbers in a journal. It is the output of a script the reader can run on their own machine, on their own data, with their own parameters. The theory does not ask to be believed. It asks to be run.

Darwin gave the world a book that changed how we see ourselves. This paper gives the world six files that can be run by anyone. The difference is not that the claim is larger. The difference is that the evidence is executable. The reader does not need to trust that the centroids converge. They can watch them converge. They do not need to trust that the Codex preserves what survives. They can run three generations and count the anchors. They do not need to trust that the forward prediction is honest. They can check the verification criteria when 2028 arrives.

The theory is the code. The code is the evidence. The evidence is running. Anyone can see it. That is what makes this paper different from every paper that came before it. Not the claim. The form.