Proof Is Not Authorization

Proof is not authorization. Proof is a closed evidentiary chain for a specific claim under named conditions: reproducible checks, lineage, and a decision authority that can refuse when evidence is incomplete. That chain is claim-scoped and falsifiable. It answers whether the claim held. Authorization is a different act. A named human accepts the consequence and routes intent to the systems that already have authority to execute. Treating the closed chain as authorization ships a settled claim into plant work that nobody has accepted. Sync may surface proof beside Evidence and Verification when the claim, the conditions, the checks, and the lineage are named. Surfacing is still a read. A named human decides. Direct plant execute stays off. CMMS write-back is not a live product path.

A closed chain looks like permission. The claim is specific. The conditions are named. The checks can be repeated. The lineage is there. A decision authority could have refused because the evidence was incomplete, and did not. The meeting then treats the case as plant-execute: write the work order, clear the equipment to run, send the crew. The chain did none of that. It answered whether the claim held. It did not accept the consequence of the next action. It did not write that action into an authorized execution system.

Sync keeps that split on the signed-in Decision Case. A signed-in user completes the case in a fixed order: Question, Evidence, Recommendation, Human decision, Action, Verification, and Learning. Orville Davis states that order in Field Manual v0. The manuals index lives at /manuals. This essay is why a closed chain cannot be read as authorization. The Evidence chapter may hold the chain. The Verification chapter records named observation against the criteria the decision named. Neither chapter accepts the consequence, and neither chapter writes the work.

Proof answers whether a claim held

Assured Is Not Proven sits one step earlier. Assurance is time-bounded confidence that an asset stays in a known-good state under operating conditions, with ownership and monitoring. That standing claim can be probabilistic. Proof is the closed evidentiary chain for a specific claim under named conditions. This essay starts after that split has been kept. The record really can be proof. The claim can be specific. The conditions can be named. The checks can be reproducible. The lineage can be present. A decision authority can have been able to refuse when evidence is incomplete. That closed chain can be honest and still not be authorization.

Proof has a narrower object than the next action. It is about one claim, not about permission to act on every consequence of that claim. The claim held, or it did not, or the check was inconclusive. A sentence that only says the claim was proven does not say who accepted the consequence of executing the work, writing the work order, or clearing the equipment to run. Claim-scoped means the chain answers that claim. Falsifiable means a failed check would withdraw it. Neither property is a grant of authority.

Verification Is Not Optional states the gate for the check. The case stays open until named observation against named criteria is recorded as achieved, not_achieved, or inconclusive, with measured notes. Recording that check closes a claim about the past. It does not authorize the next action. A reproducible check is that observation, available to be run again against the same criteria. A label that says proven is not the observation, and the observation is not authorization.

Verified Is Not Assured keeps a verified stamp off standing confidence. A verified work package, inspection, or AI recommendation closes a claim about the past. Assurance is the standing claim that comes after. Proof is the closed chain for a specific claim. Each word can be true in its own tense. None of them is authorization to execute plant work.

Proof answers whether a claim held under named conditions. It does not accept the consequence of the next action.

Authorization is a separate named human decision

Authorization is not a property of the evidence. It is an act. A named human decides. That person accepts, rejects, escalates, or returns the recommendation, and in accepting takes the consequence of what follows. The closed chain can sit on the case while that act is still unrecorded. Until the act is recorded, the case is not authorized, however complete the chain is.

Human Decision Is Not Optional states the act. A named person accepts, rejects, escalates, or returns. The Human decision chapter records who accepted the consequence. That record is the decision authority for the claim on the case, including the right to refuse when evidence is incomplete. It is also the only place on the case where authorization is named. The chain does not fill a missing name. Proof without that act is a settled claim and an open decision.

The decision routes intent. It does not perform the write. Intent on the case says what the named person means to have done. Authorized execution systems are the systems that already accept the command: the CMMS that writes the work order, the system that records the isolation, the control path that starts or stops equipment. Routing intent means the named decision points at those systems. It does not make Sync one of them.

Action Is Not Execution keeps that write off the case. The Action chapter records intent. ACTION remains a locked disposition until authorized execution systems write the work order or isolate the equipment. A closed chain does not unlock ACTION. Sync does not write the work order. Sync does not clear equipment to run. Sync does not mark the case plant-execute.

Recommend Is Not Authorize is the split one step earlier than the human decision. A recommendation is an argument. Authorization is an act by a person who can be named. This essay is the split after the chain has closed. A proposal that cites proof is still a proposal. A closed chain that cites itself is still not the act. Proof is not authorization, and a recommendation is not authorization. They fail in different places. Collapsing them ships the same false permission.

A closed chain does not clear equipment to run

The failure mode is ordinary. The chain is shown beside a maintenance decision or a production decision, and the room treats proven as authorized. Write the CMMS work order. Clear the equipment to run. Return the asset to service. Release the hold. Treat the case as plant-execute. Each of those is a consequence. The chain does not accept it. The claim can have held. The authorization can still be unrecorded. The board looks settled because the proof word was allowed to stand in for the decision.

Green Is Not Go already refuses to treat a green tile as permission to run, clear, start, or leave equipment in service. A proof record painted on that tile is not a stronger green. It is a closed chain sitting on a display. Go still requires a named human decision. Clearing equipment to run is that decision, routed to the system that can carry it out. The color supplies neither the decision nor the write. The chain supplies neither one either.

Honesty Boundary Is Not Optional is the rule that keeps the words apart under the honesty and verification boundary. Sync states what was checked and what was not claimed. Calling proof authorization crosses that boundary. Treating a named decision as if it were only a closed chain, or treating the closed chain as if it were the decision, is the same confusion from the other side.

Treating proof as authorization ships a settled claim into plant work that nobody has accepted. The chain can be closed and the decision can still be open.

Surfacing proof beside Evidence and Verification is still a read

Sync may surface proof beside Evidence and Verification when the claim, the conditions, the checks, and the lineage are named. Surfacing is still a read. The screen can show the chain next to the evidence the case holds and next to the named observation the verification step stores. Showing it does not write a CMMS work order. Showing it does not clear equipment to run. Showing it does not treat the case as plant-execute. A read of a closed chain is still a read.

Evidence from the plant beats the chain. If the evidence on the case does not support the specific claim under the named conditions, the case refuses. If the evidence closes the chain and does not record a named human decision, the case may store the proof and must not store the case as authorized. The label does not fill the gap, and it does not close it. Sync refuses false precision. Sync refuses when evidence is insufficient.

Stage-1 evidence is the record held on the case. A live connector that pulls historian or control-system tags sits outside this edition. Simulated or seeded telemetry and assets are practice records. A practice record that says proven is not a customer plant release, and it is not authorization to execute.

What the Decision Case may store

Evidence may cite a closed chain when the claim, the conditions, the reproducible checks, and the lineage are named. That citation is a record that a specific claim held, or that it did not. It is not a record that plant work was authorized. The Evidence chapter publishes that step. A recommendation may say investigate because the chain is open, or because the chain is closed and authorization is still unrecorded. The proposal does not authorize the next action.

If the named person approves work, the case may store the intent. The intent is not execution. A proof label does not perform the write and does not accept the consequence the work was meant to carry. Authorized execution systems write the work order or the isolation. Sync does not write the work order. Sync does not mark an asset authorized. Sync does not write that state back. CMMS write-back is not a live product path. Direct plant execute stays off.

Verification asks whether the authorized action did what the decision named. The check is named observation against named criteria, stored as achieved, not_achieved, or inconclusive, with measured notes. That record can close a later claim about the action. It does not, by itself, authorize the action it is checking, and it does not turn an earlier proof into permission. A named human still decides.

Sync may surface proof beside Evidence and Verification when the claim, the conditions, the checks, and the lineage are named. Surfacing is still a read. Sync refuses false precision. Sync refuses when evidence is insufficient. A named human still decides. The chain stays an answer to whether the claim held. Authorization stays unnamed until a named human accepts the consequence and routes intent to authorized execution systems.

Learning keeps the closed case: achieved, not_achieved, or inconclusive, with measured notes. The Learning chapter keeps that record. It does not keep proof as authorization. A later question that cites a proven claim as if the work were already authorized is citing a closed chain. Sync must not auto-close, auto-authorize, or treat proven authorization as Learning credit.

Where the public statement lives

Field Manual v0 is the public contents of this loop. Start at the manuals index or open Sync Field Manual directly. Evidence may hold what the chain showed and which conditions it named. Human decision may hold who accepted the consequence. Action may hold the intent that decision routed. Verification may hold the named observation. Learning may hold achieved, not_achieved, or inconclusive, with measured notes. None of those steps is a proof label used as authorization. The Honesty boundaries keep this edition from treating a closed chain as permission to execute. Later editions can deepen a chapter. The spine stays in this order.

What this article is not claiming

This is an essay about the Decision Case order, not a customer case study. It names no plant, states no savings figure, and claims no prevented failure. It states no OEM limit and no operating threshold. It does not claim that a closed chain authorizes plant work, writes a CMMS work order, clears equipment to run, or keeps an asset known-good. It does not claim that Sync executes plant work. It does not claim CMMS write-back as a shipped product.

Stage-1 readiness means a signed-in user can complete the Decision Case — question, evidence, recommendation, human decision, action, verification, and learning — and Field Manual v0 describes that journey. Walking those steps is not authorization. The human decision step is where a named person accepts the consequence. This edition does not describe plant execute, a live connector tag pull, CMMS write-back, SMTP invite delivery, or automatic revocation of access on expiry as live. It does not describe Sync writing work orders, clearing equipment to run, marking a case plant-execute, starting equipment, releasing a hold, or controlling the plant. Simulated or seeded telemetry and assets are practice records. They are not live plant results. Self-guided onboarding is not claimed as a live product path.

Companion reading: Assured Is Not Proven on why standing confidence is not a closed chain, Verified Is Not Assured on why a verified record closes a claim about the past and is not assurance, Verification Is Not Optional on why the case stays open until the check is recorded, Green Is Not Go on why a green tile is not permission to run, Action Is Not Execution on why recorded intent is not plant execution, Honesty Boundary Is Not Optional on why the limit has to be stated, and Human Decision Is Not Optional on why a named person still has to accept, reject, escalate, or return. A Reliability Assessment asks whether the records can support a conclusion. A Strategic Pilot is a governed proof around one operating decision. The Human decision chapter records who accepted the consequence. The closed chain does not authorize the work.

The series continues with Authorization Is Not Accountability, on why a named human decision that accepts consequence and routes intent to authorized execution systems is still not accountability for outcomes after the work runs. Authorization answers who may start. Accountability is the continuing named ownership of results, exceptions, and learning after execution.

Accountability Is Not Closure states the split after that ownership. A named accountable human without a verified outcome leaves the operating loop open. Closure is the measured result recorded against that ownership, not named intent.