Six-Stage Pipeline

Summary

The six-stage pipeline is the proposed operational architecture of Standing and Answerability Ethics. It identifies directed claims and their content and maps every feasible option. At Stage 1, an asserted ground relation is recorded as established, absent on the current record, or unresolved; unresolved records identify whether the source is factual evidence, the boundary of a recognized ground, or a candidate new ground. Absence is an affirmative finding after reasonable inquiry, not the default when proof fails; missing or controlled evidence that could materially change the status leaves the relation unresolved. Only established claims populate the claim roster. Stage 4 applies the relevant entry filters; it records no eligible option if none remains and otherwise compares the eligible pre-classification profiles. Stage 5 conditionally audits every eligible option for the nonempty forced-set condition, records operational selection where applicable, and fixes the claim classifications for the selected or taken act; Stage 6 determines residue. Each stage receives its normative inputs from earlier stages, so a desired verdict cannot enter as a description of the facts or as an unstated comparison rule.

The pipeline does not make the operational system derived. It locates where its defended, proposed, and open rules do their work. Where Stage 4 fails to establish a top option or top set, the pipeline records that complete output and its source rather than manufacturing completion downstream. Positive pairwise relations remain recorded but do not by themselves establish a top result.

The sequence

Stage 1: identification

From the facts and Claim Grounds, audit each asserted directed relation and construct the claim roster: who holds what claim against whom, and on which ground. Standing identifies possible holders; it does not itself populate the roster.

Stage 1 records one of three ground statuses:

An unresolved record identifies its source as factual-evidential where the criteria of a recognized ground are fixed but a material fact is uncertain; recognized-ground-boundary where the material facts are fixed but the scope or threshold of an existing ground is unsettled; or candidate-new-ground where a relation may ground a claim but is not covered by an established ground.

Candidate-new-ground status requires more than a grave interest, preference, causal connection, or imaginable moral relation. The record must identify a possible holder and addressee, the relation between them, provisional claim content, the reason that relation could generate a directed obligation, why the recognized grounds do not already decide the case, and what would establish, subsume, or defeat the candidate. These requirements keep the roster open without letting an asserted consideration enter as a claim by being called plausible.

Mere failure to prove an asserted relation does not establish absence. Where missing evidence could materially change the status, the record remains unresolved from factual evidence unless independent evidence supports absence. An actor or institution cannot obtain an absence finding by withholding, destroying, or making inaccessible evidence it controls. Existing risk, holding, account, role, and controlled-route evidential duties govern nonproduction where their own conditions are independently established; this classification rule does not bootstrap any of them from the candidate ground.

These rules state one condition rather than two. Nonestablishment is probative in proportion to the sensitivity of the inquiry: a search that would have found the ground had it been present licenses an absence finding, and a search that would not licenses nothing. That is why absence requires affirmative inquiry rather than mere failure to prove, and why evidence withheld by the party controlling it blocks an absence finding instead of supporting one — a sensitive search whose result was suppressed is not a search that came up empty. The rule fixes when a negative observation is informative. It does not allocate a burden, and an unresolved record is not a finding that the ground is absent.

An unresolved ground is not a partial claim and does not occupy a fourth claim status downstream. Its nonestablishment also does not prove that the consideration is claimless. A conflict omitted here cannot be repaired by quietly adding its interests during comparison, while a live unresolved relation cannot be converted into clean absence by omission.

Inquiry does not need a candidate claim in order to be permissible. Voluntary or theoretical inquiry can proceed for epistemic reasons, and the proposed taxonomy itself withholds an absence finding until reasonable inquiry has occurred. A directed duty to investigate can arise independently from established due-care, holding, record, role, enacted-commitment, or claim-infrastructure relations. Genuine Reason supplies its own precautionary discipline where uncertainty concerns whether a someone is present. Each route supports inquiry through its own warrant; none establishes the candidate ground, converts uncertainty into a claim, or supplies enforcement authority.

This taxonomy concerns the ground of a directed relation. Uncertainty about whether a candidate being is a someone is a distinct Stage 1 party-roster question governed by Genuine Reason, and uncertainty about whether a fixed later individual presently holds a claim remains the temporal question stated at Open Questions. Neither is reclassified as a candidate new ground.

Stage 2: content

For each established claim, specify what full satisfaction would require from its ground, the holder's good, and the situation. The claim is also typed as burden-type, which can in principle be left unsatisfied under an adequate justification, or constraint-type, which invokes the party-preserving limit at Possession.

For an unresolved ground, Stage 2 may state conditional content: what the claim would require if the ground were established. Conditional content does not establish the ground or enter a claim into the roster.

This stage fixes content before options are ranked. It prevents the apparent strength of an option from shrinking the claims it would fail.

Stage 3: mapping

For every feasible option, describe which established claims it satisfies and leaves unsatisfied, in what degree, and with which gravity factors. Where a ground remains unresolved, map the admissible branches needed to show how its establishment or absence would alter the option profiles. Mapping is descriptive. It includes delay and continuation where they are feasible, because either can impose irreversible costs, but it expresses no verdict about ground status, eligibility, or comparative priority.

Stage 4: comparison

Stage 4 output reuse. For a nonempty eligible set, Stage 4 produces one pre-classification set-level output. If Stage 5 establishes a forced set, Necessity reuses that output for operational selection. No classification-sensitive second comparison occurs without another independently warranted comparison rule.

The applicable entry rules, including Possession within its scope, filter the feasible map to the morally eligible option set. If the set is empty, Stage 4 records no eligible option and stops before comparison. Necessity then performs no selection and generates no necessity-required status. Comparison Discipline compares only a nonempty eligible set. Its inputs include which established claims each option satisfies or leaves unsatisfied, degree, gravity, irreversibility, duration, cumulative position, probability, and every other feature admitted by the governing rule. The profiles do not contain the labels justified override or wronging.

An unresolved Stage 1 record cannot be cited, counted, silenced, or ranked as though it were a claim. Comparison may continue conditionally across its admissible branches. An unconditional output requires either a separately warranted rule for handling the uncertainty or the same eligible set and complete output across the branches relevant to the decision. Even where such a rule supplies an operational output, the branch-specific profiles remain attached and the rule supplies no claim classification. Otherwise the pipeline remains unresolved from Stage 1; a Stage 4 fallback cannot supply the missing ground determination.

Stage 4 produces one complete output for the eligible set: a uniquely best option, an affirmative co-best set where the validity discipline is satisfied, or failure to establish a top option or top set with its source recorded. Pairwise relations remain recorded but do not by themselves establish the complete output. Adequate Justification applies the result holder by holder at the warrant carried by the governing comparison rule.

Stage 5: classification and forced-set audit

For any option under examination, a claim has one of three statuses:

• satisfied where the option satisfies the claim at its fixed content;
• justified override where the option leaves the claim unsatisfied but an adequate justification is available to its holder;
• wronging where the option leaves the claim unsatisfied and no adequate justification is available to its holder.

These are the only claim-level statuses. Where an unsatisfied claim is left by a positively established top result — a uniquely best option, or a certified affirmative co-best set — whose holder-relative vindicatory effect is unsettled, it remains unresolved whether the claim is a justified override or a wronging; that uncertainty is not a fourth status. Where instead no top option or top set is established, the classification is not suspended: a band-clearing claim left unvindicated by the ensuing selection is presumptively a wronging under the burden-of-justification rule at Adequate Justification, and a presumptive wronging satisfies the forced-set conjunct below while it stands.

Only an established claim can receive one of these statuses. Where its ground remains unresolved, the branch in which the ground is established may be classified conditionally, but the current record establishes neither satisfaction, override, nor wronging for that asserted relation. An unresolved ground therefore cannot establish a forced-set conjunct.

A forced set is established only where the eligible set is nonempty and every eligible option contains at least one established wronging. Where forced-set status matters, audit each eligible option to one of three results:

1. At least one wronging is established. That option satisfies its conjunct in the universal forced-set test, and auditing that option may stop.
2. Every relevant claim is conclusively classified and none is a wronging. The eligible set is conclusively not forced, and the forced-set audit may stop.
3. No wronging is established, but one or more claim classifications capable of determining whether the option contains a wronging remain open. That option's conjunct in the universal test remains open. The set cannot be established as forced on this record, although a different option conclusively shown to contain no wronging would establish that the set is not forced.

These are option-audit results, not new claim-level statuses.

Complete classification of an unselected option is unnecessary once one wronging is established under it. Complete classification of every relevant claim may be necessary to establish that an option contains no wronging. Only if every eligible option reaches the first result is the set forced. The Stage 4 output-reuse rule then governs downstream dataflow.

After an act is selected or taken, its claim-level classifications are recorded as its actual profile. One act may satisfy some claims, justifiably override others, and wrong still others. An act is justified only where every claim it leaves unsatisfied is a justified override. Operational selection and the proposed necessity requirement remain separate; the four composite act labels at Necessity summarize the resulting profile without replacing its claim-level classifications. Stage 6 assigns residue from established classifications.

Stage 6: residue

From established classifications, determine the secondary claims specified by Residue. A conditional branch can state what residue would follow if its ground and classification were established; it generates no present residue merely from the unresolved record. Justification can change the kind of residue. It cannot make an unsatisfied claim disappear from the record.

Two boundary controls

Case: In the first case, C publishes accurate criticism of P's proposal without deception, undertaking, control, or another special relation. Other parties then choose a rival proposal through exercises they did not owe P, and P loses the opportunity. In the second, A validly promises P to return P's borrowed tool unchanged by Friday and does so on Friday.

Verdict: The first setback does not enter Stage 1 as an imposition claim against C on the stated facts, because it consists solely in protected exercises. The second case does enter the roster as P's exercise-grounded claim against A, but the performed return satisfies it. Neither case reaches Stage 6 with residue: the first because the described ground is absent, the second because the identified claim was satisfied.

Machinery: The protected-exercise filter controls the first Stage 1 decision. No independently identified claim requires C or the choosing parties to exercise differently toward P in the respect producing the setback. Exercise supplies the claim in the second; Stage 2 fixes timely return as its content, and Stage 5 records satisfaction. Responsiveness applies only after a claim is left unsatisfied.

Cost: The controls depend on deliberately clean facts. Deception, damage, delay, induced reliance, structural control, or another ground governing the relevant exercise would change the roster or satisfaction result. The absence of an imposition claim and the satisfaction of an exercise claim establish no general conclusion about every moral reason in either case.

A ground-status battery

Case: Four records reach Stage 1. First, C publishes accurate criticism of P's proposal under the clean facts stated in the preceding case. Second, records do not establish whether operator A discharged the contaminant found in P's well, although discharge by A would satisfy the recognized imposition ground; A exclusively controls the discharge logs and has not produced them. Third, a supplier receives almost all its revenue from one buyer under fixed facts of unilateral term-setting and practical dependence, but the open boundary of holding does not determine whether the buyer holds the supplier. Fourth, A and P have maintained a long reciprocal friendship whose practice includes showing up for each other's major losses, but they made no promise, induced no specific reliance, exercised no control, and created no urgent rescue relation. P argues that this relation itself makes A's attendance at P's parent's funeral owed to P as a friend rather than valuable only impersonally.

Verdict: The criticism case is absent on the current record. The discharge case is unresolved from factual evidence. The supplier case is unresolved from recognized-ground-boundary uncertainty. The friendship case qualifies as candidate-new-ground: it identifies P as possible holder, A as addressee, friendship as the relational basis, attendance as the claim object, and reciprocal recognition as the reason the putative obligation would be directed to P. The status does not establish that friendship grounds the claim. None of the three unresolved records enters the claim roster. Conditional mapping may continue, but no unresolved record can be weighed, classified, given residue, or used to satisfy claim-protection for enforcement.

Machinery: Imposition fixes the consequence of the missing discharge fact; A's nonproduction prevents an absence finding without itself establishing discharge or an imposition claim. Holding supplies the recognized ground while leaving its commercial boundary open. The nonexhaustive ground roster leaves conceptual space for a relationship ground without adopting one. The friendship candidate would be subsumed if a valid exercise, enacted commitment, or specific reliance were established, and defeated if friendship supplies reasons but no directed claim absent one of those relations. Stage 1 preserves each source of uncertainty; Stages 2 and 3 may map its establishment branch, while the later stages receive only established claims or explicitly conditional profiles.

Cost: The taxonomy requires judgment about reasonable inquiry and whether a proposed new relation is live rather than merely asserted. It can leave action-guiding analysis conditional where a binary roster would appear decisive. Refusing it restores that apparent decisiveness at the cost of allowing factual, criterial, or taxonomic nonestablishment to look identical to clean absence.

An initiation boundary battery

Case: An already existing patient validly authorizes clinician A to implant a reversible device whose monitoring and side effects the patient must carry, and A accepts responsibility for follow-up care. Radiologist B later reads one scan under a limited consultation and neither changes the device nor assumes its care. Covering clinician C says that they will "look after the monitoring" for a weekend, but the fixed evidence does not determine whether C assumed responsibility for the continuing condition or only agreed to perform one bounded task.

Verdict: A's conduct instantiates initiation even if the implantation and care are fully justified; the authorization is the existing patient's exercise concerning treatment, not authorization by a resulting holder to be originated. Authorization changes the claim structure without making the ground a wronging classifier. B does not instantiate initiation on the stated facts, although error or another relation could generate a different claim. C's case is unresolved at the recognized-ground boundary. It does not enter the roster unconditionally until the scope of the assumed responsibility is fixed.

Machinery: Initiation covers knowingly beginning the condition and deliberately assuming responsibility for it. Exercise records the existing patient's authorization and its scope separately. Newness bars transferring that description to procreation, where no resulting holder exists beforehand to perform an exercise. Stage 1 distinguishes A's established ground, B's clean absence on the stated initiation theory, and C's unresolved boundary without treating medical involvement as sufficient for all three.

Cost: The ground reaches permissible and beneficial undertakings, so identifying it cannot function as a verdict of wrongness. Its assumption branch also requires evidence about the role actually undertaken; preserving that boundary can leave responsibility conditional where occupational labels or helpful conduct would make attribution easier.

Stage discipline

The sequence blocks five recurrent shortcuts:

• a social benefit cannot enter Stage 1 without a holder and ground;
• a ground not yet established cannot enter the roster as a claim, while a live unresolved ground cannot be treated as affirmatively absent;
• an option cannot determine the content of the claims against it;
• a factual map cannot conceal a comparison rule in its descriptions;
• necessity-required status under the proposed forced-set rule cannot double as comparative vindication, an adequate answer to any holder the act wrongs, authorization, institutional jurisdiction, or enforcement authority.

An institutional process may distribute the stages among different actors. It remains answerable for preserving the sequence, the evidence, and the points at which judgment entered.

Limits

• The pipeline is architecture, not an algorithm. It does not supply the facts, gravity judgments, comparison rules, or institutional competence needed to run it.
• The Stage 1 taxonomy does not establish a candidate new ground, impose a general directed duty to investigate one, or decide how much inquiry is reasonable. It requires enough inquiry to support an absence classification; further investigation can remain voluntary, epistemically warranted, or independently owed through another claim, role, or institutional relation.
• Every downstream result inherits the limits of the comparison discipline.
• Feasibility can itself be contested. Excluding an option as unavailable is part of mapping and must not serve as an unreviewable verdict.
• Refusing the pipeline restores an operational gap; it does not weaken the same classifications into informal equivalents.

Related pages

• Standing and Answerability Ethics
• The Holder's Good
• Adequate Justification
• Possession
• Comparison Discipline
• Necessity
• Residue