The method

Erase the noun. Keep the evidence.

A temporary subtraction test for inherited categories.

Human knowledge is a map drawn on the cave wall.

Some lines mark real cliffs. Some mark useful trails. Some carry historical assumptions or compress several phenomena into one label. We erase one line temporarily and look at the territory again.

Not contrarianism: removing a word does not remove the observations, and it does not prove the category false. A line that the evidence forces us to redraw has survived the test.

The five questions

  1. What did we assume?

    Name the inherited category and the work we expect it to do.

  2. What did we remove?

    Set the noun aside without discarding measurements, mechanisms, constraints, or domain knowledge.

  3. What survived?

    Reconstruct only the relations and operations still required by the evidence.

  4. What changed?

    Record whether the boundary returned, moved, divided, weakened, or remained unresolved.

  5. What can we test?

    State a prediction, intervention, comparison, or kill condition.

Try it on one system

Open the full worksheet

Rules that keep the method honest

Keep native mathematics

Removing a noun is not permission to flatten domain-specific equations or discard established controls.

Permit incomparability

Two domains may share a signature while lacking a valid mapping. "Incomparable" is a legitimate result.

Redraw surviving boundaries

The method succeeds when it discovers the inherited category was already doing useful, evidence-backed work.