SUMMARY: A contradiction ledger is a compact record of claims that cannot all be true, evidence that points in different directions, and questions that remain unresolved.
Most operational systems are designed to close items. Tickets move to resolved. Reports receive a final assessment. Meeting notes convert disagreement into action points. Closure is necessary, but it can also erase the exact tension that future reviewers need when conditions change.
The ledger does not preserve every argument. It preserves contradictions with decision value: two sources describing incompatible events, a policy that conflicts with observed behavior, a metric that improves while the underlying service deteriorates, or a confident summary built from evidence that remains incomplete.
The Minimum Entry
Claim pair: state the conflicting propositions without blending them into vague compromise.
Evidence boundary: link the strongest support for each claim and name what neither source can establish.
Operational consequence: record which decision changes if one claim becomes more credible.
Review trigger: set a date, event, or new evidence threshold that will reopen the entry.
Keep Contradiction Useful
Assign an owner without assigning a preferred outcome. Separate contradiction from accusation. Close an entry only when evidence resolves the conflict, the decision becomes irrelevant, or the uncertainty is formally accepted as a continuing constraint.
Use the ledger beside synthetic consensus to prevent repetition from masquerading as resolution. It also pays down verification debt by making unresolved evidence visible before later decisions inherit it as certainty.
Operator rule: do not force agreement when preserving the disagreement creates a better future decision.