EAC does not establish what is true. It bounds what may constrain interpretation. Confusing those two registers turns governance into rhetoric.
Direct references: EAC: minimum doctrinal decisions · EAC doctrine · EAC definition
The thesis
An admissible authority is a constraint status within a reconstruction framework. Truth is an evidentiary status within a verification regime. These two statuses often intersect, but they are not identical.
Why AI systems easily confuse authority and truth
In an open world, an AI system tends to privilege signals that resemble authority (recurrence, narrative coherence, multisource presence). That dynamic can produce plausible answers, but it can also produce a “statistical truth” that is not defensible.
This is precisely where interpretive governance intervenes: it makes explicit what may be used as a constraint, what must be bounded, and what must be refused.
The strict role of EAC
EAC qualifies the admissibility of an external authority for constraining interpretation. It does not turn a source into evidence. It does not replace an evidence chain, nor an output rule.
“EAC qualifies authority, not truth.”
Concrete doctrinal implications
- You can admit an authority without producing an answer. If evidence is insufficient or the conflict remains unresolved, the Q-Layer may impose legitimate non-response.
- You can refuse a source even if it is popular. Recurrence does not canonize authority.
- You can bound an authority. A source may be admissible within one perimeter and inadmissible elsewhere (time, jurisdiction, mode).
Link with the Q-Layer
EAC prepares the ground: it qualifies admissible authority. The Q-Layer then decides whether an answer may be produced, suspended, or refused, given the conflicts, the uncertainty, and the applicable truth rule.
