What should be discarded when an experiment cannot score its answers honestly?
https://github.com/research-farm/construct/blob/main/notes/walkthrough/16_EPISTEMIC_FRAME_CHECK.md
A free-text scorer may appear deterministic while missing the decision expressed by the text. A rule that sees ship inside do not ship cannot distinguish an action from its negation. Adding more phrases may repair known cases without repairing the boundary.
Separate question from apparatus
When the scoring surface cannot survive fresh review, the experiment should stop. That closes the instrument, not necessarily the inquiry that motivated it.
The question may return with a different answer surface, oracle, or experimental geometry. Its earlier lineage should remain visible so the new instrument inherits the failure rather than repeating it.
Questions to carry forward
Did the research question fail, or only the way it was asked? Can the scorer recognize ordinary paraphrase, negation, and revocation? How many repairs should be allowed before review becomes accommodation?
A blocked experiment can be a completed result when it identifies what cannot yet be measured.