OBJ-008: “Non-unitary intervention breaks physics”

Target claim(s)

Any claim that divine action introduces non-unitary dynamics or violates standard conservation constraints.

The objection (steelman)

“If you allow non-unitary ‘grace’ terms, you can explain anything. Also, non-unitary changes typically violate conservation/information constraints or conflict with well-tested physics. This is a license to cheat.”

Why it seems compelling

Unconstrained intervention destroys predictability and invites post hoc patching.

Reply (logic-first)

• The framework must treat intervention as an explicitly typed channel (not a free miracle knob).
• The intervention channel must:
  • specify when it activates (preconditions),
  • specify what it changes (state variables),
  • specify what it preserves (invariants), and
  • generate distinctive, testable signatures with confound controls (T8).
• If those conditions can’t be met, the intervention claim should be downgraded to interpretation (not part of the formal layer).

What would change my mind

• If repeated attempts to specify falsifiers fail and the claim only ever retreats to vagueness, classify it as non-empirical and remove from the equation layer.