Brenner Operators

Separate conceptually blended categories into distinct causal roles so you can reason cleanly. Distinguish program from interpreter, message from machine, specification from execution.

Change the problem's coordinates/representation so that (a) structure becomes obvious and (b) rival hypotheses make different predictions. Especially: reduce dimensionality from 3D to 1D when possible.

Convert an explanatory narrative into a concrete decision procedure. Ask: "If this were true, what would I see? How would I get hold of the information to test this?"

Identify properties that remain meaningful when implementation details are unknown. Find what survives coarse operations, what must hold regardless of specifics.

Derive what patterns are *forbidden* under each hypothesis, then design cheap tests that probe those forbidden patterns. Prefer experiments that kill whole model families over experiments that merely accumulate supportive data.

Change the experimental organism/system until the decisive test becomes cheap, fast, and unambiguous. Treat the Tree of Life as a component library to be raided.

Use biological amplification mechanisms (selection, replication, dominance, abundance) to make signals large and robust. Let biology do the work.

Build what you need rather than waiting for infrastructure. Don't let missing tools define your pace or hypothesis space.

Import patterns from unrelated fields. Maintain "fresh eyes" by resisting expert entrainment. The best people to push a science forward often come from outside it.

Find contradictions in the current model. Paradox is not a nuisance—it's a beacon pointing to missing production rules or level confusions.

Preserve a high-coherence core model while isolating anomalies. Don't let exceptions collapse a coherent framework prematurely, but don't hide them either. Put them in an appendix and resolve them later.

Move out of phase with fashion. Work half a wavelength ahead or behind. Avoid crowded priors and industrialized midgames.

Discard hypotheses the moment they fail. Don't fall in love with theories—treat them as mistresses to be discarded once the pleasure is over.

Calculate actual numbers. Stay imprisoned within physical constraints. Use scale as a hard prior to filter impossible cartoons.

When the "real" experiment is hard, first run a cheap pilot that would kill the key alternative. De-risk before committing months of work.

Before elaborate inference, directly observe. Collapse inferential chains by looking.

Before concluding a hypothesis is wrong, verify that the intervention actually worked. Distinguish "won't" from "can't."