DUAL-OBSERVATION COMPARATOR
Independent Comparison Against Common Origin
Class: Measurement Primitive
Abbreviation: DOC
Version: 1.2
Domain Scope: Verification, Calibration, Alignment, Audit
Purpose
The Dual-Observation Comparator is a domain-neutral measurement primitive for detecting divergence between independently produced observations of a common reference.
DOC does not determine which observation is true.
It determines whether independently derived observations remain sufficiently congruent for the requested operation to proceed.
Its governing principle is simple:
Where adjustment depends upon observation, agreement should not be manufactured by allowing one observation to produce the other.
Common Origin
Every DOC comparison begins with a declared reference coordinate.
From that common origin, two observations are produced independently.
\[ O_1(R),\quad O_2(R) \]
where \(R\) is the declared reference and \(O_1\) and \(O_2\) are independently produced observations.
Independence matters.
Two copies of the same calculation do not constitute independent observation merely because they produce two outputs.
Likewise, one observer reformulating its own conclusion does not create corroboration.
DOC therefore requires sufficient separation between observation paths to make comparison meaningful.
Comparison
The observations must be rendered into a mutually comparable structure before evaluation.
DOC then compares the declared fields without supplying semantic reconciliation of its own.
Its function is:
\[ DOC(O_1(R),O_2(R)) \rightarrow S \]
where \(S\) is a declared comparison state.
The comparator does not alter either observation in order to obtain agreement.
Comparison States
READY
The observations are congruent within the declared tolerance applicable to the comparison.
DRIFT
The observations remain comparable, but consequential asymmetric deviation has been detected.
DISAGREE
The observations produce materially incompatible structures or results.
UNRESOLVED
The observations cannot presently support a valid comparison because information, comparability, independence, or measurement sufficiency is missing.
I would definitely add UNRESOLVED. We just found the same problem with Kraken. A constitutional system shouldn't force “I cannot determine this” into either agreement or disagreement.
Effect
READY permits the requested downstream operation to continue, subject to any other applicable requirements.
DRIFT suspends any operation whose validity depends upon the disputed alignment.
DISAGREE prevents the disputed comparison from being used as a sufficient basis for adjustment.
UNRESOLVED preserves the existing reference condition until sufficient comparison becomes possible.
None of these states determines substantive truth.
They determine the condition of the comparison.
Reconciliation
DOC does not reconcile its own failures.
That is important.
A DRIFT, DISAGREE, or UNRESOLVED result may be referred to the appropriate office, process, or human authority for investigation.
The underlying observations may then be corrected, supplemented, independently reproduced, or otherwise resolved.
DOC may subsequently be run again.
The earlier result remains part of the measurement record.
Reconciliation may resolve divergence. It may not erase the fact that divergence occurred.
Constitutional Boundary
DOC may determine:
whether two observations are comparable;
whether their declared structures are congruent;
whether deviation exceeds an established tolerance;
whether sufficient independence exists for the comparison.
DOC may not determine:
what an observation ought to have said;
which observer deserves greater authority;
whether a disputed proposition is substantively true;
whether governance should change;
or whether disagreement should be suppressed for operational convenience.
Congruence is evidence. It is not sovereignty.
