PHY.10:10 - Architectural Rationale
B.5.TC compares how theoretical accounts answer a working question. This method develops a physical consequence of that comparison: how a preparation or interaction can expose the disputed dependence. PHY.9 supplies an observable readout; the statistical or mathematical design uses that physically realizable relation.
Equal unperturbed responses can conceal different retained state. Changing a drive or reversing an interaction therefore supplies more than another sample of the old observation. It asks the accounts how the physical system responds to a different operation. That response can also reveal how the system might be controlled or used.
The test retains the apparatus and preparation because they participate in the consequence being compared. This makes an unsuccessful comparison informative: a return can change the subject account, the experimental construction or their mathematical description, according to which premise failed.