Temporal Calibration

Preamble

Temporal Calibration studies how an acceptance edge moved before. It treats history as evidence for a trajectory, then keeps the future estimate visibly provisional.

---

The Edge Moves

Ideas age strangely. A pattern that once looked too alien can become obvious after the audience, tools, language, or incentives change. A pattern that once felt advanced can become background noise.

This proxy compares time slices inside a domain: accepted-novel examples, rejected too-early examples, ordinary defaults, and the conditions that made the boundary move.

History Can Overfit The Future

The danger is clean extrapolation. A language model can make history look like a line, then extend the line into a future that never arrives.

Temporal Calibration needs real corpora, careful normalization, evidence, confidence, and a falsifiable test. The output names which conditions would need to change before the idea becomes more acceptable.

Prediction Becomes A Test

The branch lands as a trajectory hypothesis with cited examples, uncertainty, and a testable forward claim.

It earns trust when a historical pattern helps choose a live experiment while leaving room for the world to contradict the line.