PSD.1 - Bound the Decision-Support Engagement and Authority Boundary

Type: DPF pattern body Status: Candidate

Primary working result: a bounded engagement question that names the decision-support return, its recipient, the holder or subject, the horizon, the receiving decision or Work, the choice owner and authority boundary, the current scope, the smallest missing value, and an observable condition for reopening.

PSD.1:0 - Use This When

Use this pattern when someone asks a decision-support performer to “help decide”, “structure the problem”, “recommend what to do”, or “say what should develop next”, but the engagement is not yet bounded well enough to produce a responsible result. The missing value may be the recipient, holder or subject, horizon, receiving decision, receiving Work, choice owner, authority, consequence, or kind of return expected.

Begin with the later decision or use that the decision-support result is meant to inform. Separate the performer who prepares decision support from the person or arrangement that receives it and from the Agent that owns the later choice. Then return one bounded engagement question or the exact value that is still missing.

The practical gain is an affordable start that works both before overt contest and inside a disputed situation. It prevents a workshop, model, score, or recommendation from silently becoming the decision, and it gives later problem-structuring or analytical Work a recoverable subject and stop.

Do not use PSD.1 merely because a chooser must select among an already-qualified live option set; use C.11 or the applicable domain choice rule directly. Use one holder-specific practice directly when it owns the whole question and no decision-support composition remains. A scientific, legal, ethical, safety, governance, or other specialist question stays with that practice unless its result must be returned into a separately governed decision-support engagement.

PSD.1:0.1 - Working Distinctions

Working term	Meaning in this pattern
decision-support engagement	The bounded situation in which a performer is expected to make a useful result available to a receiving decision or Work. The phrase alone establishes no Method, assignment, authority, WorkPlan, dated Work, service, or product.
decision-support performer	The actual or intended person, team, organization, tool-supported arrangement, or other capable System expected to prepare the return. A title, department, model, report, or tool output does not establish this performer or its authority.
recipient	The person or arrangement to whom the decision-support return is addressed for a named use. The recipient may or may not own the later choice.
holder or subject	The person, organization, System, arrangement, place, policy, capability, or other exact subject whose alternatives or consequences are under consideration. It is not inferred from the recipient’s identity.
choice owner	The Agent authorized to make, reject, defer, probe, or reroute the later decision. It may coincide with the recipient or performer only when that coincidence and authority are independently supported.
receiving decision or Work	The later choice or Work whose next action can change because of the decision-support return. A recommendation is one premise for that use, not the choice or Work itself.
bounded engagement question	The smallest decision-support question that makes the intended return, use, participants, scope, authority boundary, and reopen condition usable now. It is a practitioner result, not a universal account schema.

PSD.1:1 - Problem frame

Decision-support practice often begins before there is a stable “problem”. One engagement opens with several participant groups arguing about causes and values. Another opens with a calm request for advice whose recipient, subject, horizon, or later authority is still vague. Both can be prematurely forced into a workshop brief, model boundary, criteria table, or option set.

The primary concern here is the boundary of one decision-support engagement for one receiving use. The first useful result is not a complete problem formulation, alternative set, comparison, recommendation, plan, or decision. It is a question and authority boundary precise enough to say what result can be prepared now, what remains outside, and where a missing value must return.

Recognition is deliberately cheap: one unclear decision-bearing value is enough to open the pattern. Consequential use adds direct evidence for identity, assignment, permission, authority, source reliance, safety, law, ethics, and other claims that the engagement actually consumes. The short opening does not lower those assurance burdens.

PSD.1:2 - Problem

An unbounded engagement can expand into every concern, participant, model, consequence, and possible future. The performer gathers material without knowing which later action it can change. A premature boundary creates the opposite failure: the sponsor’s wording becomes the only formulation, the visible recipient is treated as the choice owner, or an analytical result is presented as authorization.

These failures compound. A favourite problem-structuring Method can be selected before the decision-support use is known. A list of stakeholders can stand in for affected Systems and material concerns. A model can hide excluded consequences. A recommendation can absorb holder-specific conclusions or take a choice that belongs elsewhere.

The missing discipline is to bind the engagement to one intended return and separately governed later use while keeping the boundary explicitly provisional. Contestation, plural formulations, affected participants, models, alternatives, values, uncertainty, and follow-up open only when they can change that return.

PSD.1:3 - Forces

Force	Tension
Affordable opening	A practitioner needs a first result quickly, while the most visible request may omit the value that makes any result usable.
Useful scope	A boundary is needed to act, while early closure can suppress affected participants, rival formulations, or decisive consequences.
Advice and choice	A performer must say something useful, while the later decision and its authority remain separate.
Stable use and revisable framing	The engagement needs a current question, while new evidence or participation can legitimately reopen it.
Plain roles and exact claims	Recipient, adviser, sponsor, analyst, and decider are useful words, while titles and labels establish no assignment, capability, or authority.
Domain ownership	Decision support must combine relevant premises, while holder, legal, ethical, safety, scientific, and governance practices retain their Methods and conclusions.
Recognition and assurance	One vague value can justify a cheap clarification, while high-consequence use requires stronger direct evidence and stops.

PSD.1:4 - Solution

Bind the engagement to the smallest receiving decision or Work that can use a decision-support return. Name the distinct participant positions, subject, horizon, expected return, current scope, and authority boundary. Return the bounded engagement question when those values are adequate; otherwise return the exact missing value instead of inventing it.

PSD.1:4.1 - Name the receiving use before the problem statement

Start with one ordinary sentence:

Make [kind of decision-support return] available to [recipient] about [holder or subject] for [receiving decision or Work] within [horizon]; [choice owner] retains the later choice.

Ask what action can change because of the return and what consequence makes delay, error, or omission material. If no receiving decision or Work can be named, distinguish exploratory inquiry, research, learning, communication, or another direct use rather than manufacturing a decision-support engagement.

The requested return can initially be a bounded question, participant-and-concern account, plural formulation set, model or Method-selection result, alternative gap, comparison, recommendation, blocker, or abstention. Do not require a recommendation when an earlier honest result is more useful.

PSD.1:4.2 - Separate performer, recipient, holder, and choice owner

Name each position separately even when one System may occupy two positions:

who is expected to prepare the decision-support return;
who will receive it and for what use;
which exact holder or subject is under consideration; and
who owns the later choice, deferral, probe, or reroute.

State any coincidence explicitly. A manager can receive advice without owning the board’s later decision. A team can perform analysis for its own authorized choice, but the analysis Work, recommendation result, and later ChoiceResult remain distinct. A tool, document, score, model, department, job title, or meeting creates none of these positions by visibility alone.

Recover assignment, capability, permission, responsibility, conflict, commitment, or authority only to the extent that the current use relies on it. Use their direct patterns for consequential claims. If the required choice owner or authority basis cannot be recovered, return that exact gap.

PSD.1:4.3 - Bound the subject, horizon, receiving Work, and consequence

State the holder or subject at the grain whose alternatives or consequences can change. Add the relevant configuration, place, population, operating condition, and interval only when they change the answer. State the receiving decision or Work and the horizon over which the return is expected to remain useful.

Name at least one consequence of getting the boundary wrong. This is a recognition cue, not a complete consequence analysis. When affected Systems, missing participants, duties, interests, or representation can change the engagement, continue with PSD.2. When several problem formulations imply materially different actions, continue with PSD.3. When included and excluded concerns need a deliberate scope decision, continue with PSD.4.

PSD.1:4.4 - Choose a provisional boundary and the next truthful branch

Record what is inside the engagement now, what remains outside, and what observation would move that boundary. Do not treat package order or the numbering of patterns as a practitioner sequence. Choose the next branch from the missing result:

Current situation	Next truthful branch
participants, concerns, consequences, or representation are incomplete	recover the participation and concern account through `PSD.2`
materially different formulations or intervention logics remain hidden	generate plural problem formulations through `PSD.3`
the engagement has no usable included/excluded scope and reopen basis	set and reopen the problem boundary through `PSD.4`
the engagement is bounded but needs a model, a choice or combination of problem-structuring Methods, or facilitation	use `PSD.5` for the model, `PSD.6` for Method selection or combination, or `PSD.7` for facilitated inquiry
the engagement is bounded and the live gap is alternatives, uncertainty, comparison, or recommendation	begin with the first missing result in `PSD.8`–`PSD.13`
another practice owns a premise that can reverse the return	inspect or request the smallest bounded result through `A.15.9`
an authorized chooser already has an adequate option set and comparison basis	use `C.11` or the direct domain choice rule; stop treating engagement framing as the live question

Several branches may be live together. Their results can proceed concurrently when their subjects and returns are clear. Co-occurrence does not turn the table into a workflow or make the engagement one Method.

PSD.1:4.5 - Record the bounded engagement question or exact stop

Use the smallest account that another participant or later use must recover:

Result position	Required content
intended return	The smallest decision-support result sought now, including blocker or abstention as valid branches.
recipient and use	Who receives the result, which decision or Work can change, and the horizon.
holder or subject	The exact subject, configuration, place, population, or interval needed by the current question.
participant positions	Decision-support performer and choice owner, plus only the assignments or authority facts the current use needs.
current boundary	Included concerns and consequences, important exclusions, and why that cut is usable now.
missing value or next branch	The exact participant, formulation, scope, source, alternative, evidence, authority, or other result still needed.
reopen condition	The smallest observable change that can alter the question, scope, authority boundary, or next branch.

A compact result may be one sentence and one stop:

For [recipient] and [receiving decision], prepare [return] about [subject] within [horizon]; [choice owner] retains the later choice. The engagement is blocked until [exact missing value] is recovered.

Do not add a role register, stage model, workshop plan, evidence pack, or account field merely because a larger case might need it.

PSD.1:4.6 - Reopen locally and preserve the later choice

Name the event that can invalidate the current boundary: a changed recipient, holder, horizon, receiving Work, authority relation, participant or concern, formulation, source, alternative family, evidence condition, consequence, or later decision. Reopen only the affected position and its dependent results.

A returned recommendation may inform the later decision, and follow-up may observe that later decision, but neither retroactively turns the decision-support performer into the choice owner. If the later choice changes the problem structure, use PSD.14 to determine which follow-up or reframing question is current.

PSD.1:4.7 - What changes in practice

The practitioner stops accepting “the problem” as a sufficient start. Within the first exchange, they can name the return, recipient, subject, horizon, later use, and choice owner—or return the exact missing value. Method selection, modeling, participation, alternative generation, and recommendation Work begin from that result instead of being used to discover silently what the engagement was supposed to accomplish.

PSD.1:5 - Archetypal Grounding

PSD.1:5.1 - Contested flood-pump decision

A municipal resilience office asks a decision-support team to “solve the pump problem” before the next flood season. Operations staff prefer another permanent pump; finance questions capital exposure; residents contest which districts and harms count; and the emergency investment board owns the later funding and deployment decision.

The team does not begin with a pump model or consensus workshop. It first returns this bounded engagement question:

For the municipal resilience office, prepare a recommendation for the emergency investment board’s pre-season funding and deployment decision about the district flood-pumping arrangement over the next flood season. Compare permanent, mobile, staged, and no-new-purchase branches only after affected districts, operating conditions, safety limits, and material dissent are recoverable; the board retains the later choice.

The boundary is useful but incomplete. PSD.2 is next because affected residents, operating groups, concerns, and consequence-bearing Systems can change the engagement. PSD.3 is also live because “insufficient capacity”, “unequal protection”, “maintenance fragility”, and “unsafe deployment” imply different interventions. The team stops before model or option selection if the board’s authority, the decision horizon, or an affected district remains unrecoverable.

PSD.1:5.2 - Development-direction advice without taking the choice

An engineering executive asks an advisory team, “What should we develop next for AI-assisted incident response?” The phrase does not yet say whether the holder is a person, team, organization capability, model, agent arrangement, or operating platform; whether the recipient or an investment committee owns the choice; or whether the horizon is the next release or the next two quarters.

After one clarification, the holder is the organization’s AI-assisted incident-response arrangement, the executive is the recipient, the investment committee owns the later portfolio choice, the horizon is two quarters, and the requested return is a source-qualified development-direction recommendation. The first result is:

For the executive’s two-quarter investment proposal, prepare a bounded recommendation about development directions for the AI-assisted incident-response arrangement; the investment committee retains the later choice. Return the exact missing human-capability, organization, operations, model-evaluation, safety, or authority result before using it as a premise.

No contest is assumed. With the engagement bounded, the practitioner may inspect external results through A.15.9 and begin with the first missing result among PSD.8–PSD.13. PSD.1 authorizes no training programme, organization change, model modification, purchase, WorkPlan, or choice.

PSD.1:5.3 - Cheap non-use when the choice is already local

A procurement board is the authorized chooser, has three qualified pump options, a current comparison basis, and no unresolved participant, formulation, scope, authority, or outside-practice premise. The live question is which option to select. Use C.11 or the board’s direct choice rule. Repeating PSD.1 would add no useful boundary.

PSD.1:6 - Bias-Annotation

Scope: decision-support engagements for a named receiving decision or Work. Lenses: Gov concerns participation and decision authority; Arch, the engagement boundary and dependent contributions; Onto/Epist, keeping the subject, claims, recommendation, and later choice distinct; Prag, a useful return at proportionate effort; Did, plain entry cues that preserve the distinction between cheap recognition and consequential assurance.

Recurring bias	Likely drift	Repair
sponsor-language bias	The sponsor’s first wording becomes the problem and scope.	Name the receiving use, affected consequence, exclusions, and reopen condition before accepting the formulation.
authority-by-visibility bias	The meeting chair, report author, model owner, or recipient is treated as the choice owner.	Recover the later decision and its authority basis separately.
Method-first bias	A familiar PSM, workshop, model, or scoring scheme determines the engagement boundary.	Return the bounded question first; select Methods only for the result that is actually missing.
stakeholder-list bias	A visible list is treated as complete participation and concern recovery.	Continue with `PSD.2` when affected Systems, representation, duties, interests, or dissent can change the return.
recommendation-as-choice bias	A confident recommendation, rank, or tool output is presented as authorization.	State the recipient, choice owner, next use, and stop at the recommendation boundary.
exhaustive-framing bias	Every possibly related concern is admitted before any useful result can be returned.	Choose the smallest current boundary and one observable reopen condition.

PSD.1:7 - Conformance Checklist

The opening names a recognizable receiving decision or Work and the smallest useful decision-support return.
Decision-support performer, recipient, holder or subject, and choice owner are separately recoverable.
Any coincidence among those positions is explicit and does not transfer authority by wording alone.
The holder or subject, horizon, relevant configuration, and at least one consequence of a wrong boundary are visible at the grain needed by the current use.
The current included and excluded scope and one observable reopen condition are stated.
Contestation opens PSD.2–PSD.4 only when participants, formulations, or scope can change the return; it is not a membership requirement for PSD.1.
A bounded direct-advice case can proceed to the first missing PSD.8–PSD.13 or external-result branch without a compulsory problem-structuring sequence.
Missing authority, subject, horizon, source, or receiving use produces an exact return rather than an invented value.
The result does not claim that a recommendation is the later choice, that planned Work occurred, or that an effect obtained.
The body supports both a contested-situation entry and development-direction advising without giving either entry a different meaning of engagement or authority.

PSD.1:8 - Common Anti-Patterns and How to Avoid Them

Anti-pattern	Repair
“First agree what the problem is.”	Preserve materially different formulations and ask which receiving decision each would change.
“The client is the decision maker.”	Separate recipient, holder, choice owner, and authority basis; name any supported coincidence.
“Run the workshop and the scope will emerge.”	State the intended return and provisional boundary before choosing the workshop Method.
“The model says option A, so that is the decision.”	Keep model result, recommendation, authority, and later `ChoiceResult` separate.
“We need every stakeholder before we can start.”	Return a bounded current question and use `PSD.2` to expose which missing participant or concern can change it.
“This is only advice, so authority does not matter.”	State who may prepare, receive, rely on, and choose whenever those differences change use or consequences.
“Use PSD.1 as stage one of every engagement.”	Bypass it when the receiving question and authority boundary are already adequate and another result is the live gap.

PSD.1:9 - Consequences

A decision-support engagement gains a usable first result before it gains heavy apparatus. Practitioners can stop on one missing value, route contestation to participation and plural-formulation Work, or proceed directly to alternatives and recommendation when the question is already bounded. Recommendations become easier to inspect because their recipient, subject, horizon, authority boundary, and next use were not reconstructed after the analysis.

The cost is explicit clarification at the start and occasional reopening later. Some requests will stop before analysis because no receiving decision, choice owner, horizon, or subject can be supported. Others will widen after affected participants or rival formulations become visible. Those are useful returns, not failed stages.

PSD.1:10 - Rationale

The engagement boundary is the smallest professional decision-support result that can coordinate problem structuring and later analysis without pretending that either has already occurred. Starting from a fixed problem statement is too narrow for contested or revisable framing. Starting from an unlimited situation is too broad to guide a return. Starting from a Method confuses the means with the receiving use.

Generic FPF patterns already govern outside-practice results, evidence, permission, agency, and choice. PSD.1 does not reproduce them. It contributes the field-specific opening that connects a separately governed later decision to a provisional problem-structuring and decision-support boundary, and it tells the practitioner which exact downstream question is now live.

PSD.1:11 - SoTA-Echoing

Practice question	Best-known line	Serious alternative or default	Defect overcome and pattern mutation	Source roles and limits	Reopen condition
How should a practitioner open an ill-structured or contested decision-support situation?	The best-known line is to bind inquiry to a named receiving use while keeping framing plural, contextual, and revisable before committing to one PSM, model, or formulation.	A fixed sponsor brief or Method-first workshop is the serious default; an unbounded “include everything” inquiry is the opposite default.	Both defaults hide the decision-support return: one naturalizes the first frame, the other cannot stop. Adapt: `PSD.1:4.1`–`4.4`, the flood-pump case, and the first seven checklist rows require a receiving use, provisional boundary, plural-formulation return, and local reopen before Method choice.	Smith and Shaw’s 2019 PSM review is a best-known-line candidate for the continuing PSM branch and its differing characteristics; Kogetsidis’s 2025 review is an application comparator. Kelly and Gero’s 2022 framing review, Litster, Cardoso, and Hurst’s 2024 systems-mapping study, and Nickel, Hurst, and Duimering’s 2024 contextual trade-off study support explicit candidate framing, contextual comparison, and revision. The reviews are non-exhaustive; the studies establish neither one framing ontology, universal Method, shared frame, prevalence, nor superiority.	Reopen if comparative practice shows a lower-effort opening that preserves the same receiving-use, plurality, authority, stop, and revision value, or if a source changes the conditions under which framing must be reopened.
How should decision-support advice remain useful without becoming the later choice?	The best-known current FPF line is result-first and authority-bounded: name the receiving decision, reuse or request only the outside-practice result it needs, return the recommendation as a premise, and let the authorized chooser produce the separate `ChoiceResult`.	Expert-report, approval, score, or model-output language treated as the decision is the serious default.	The default transfers authority by format and hides blockers. Adopt: `PSD.1:4.1`, `4.2`, `4.5`, the development-direction case, and the authority checks keep performer, recipient, holder, outside-result owner, and choice owner distinct.	Current `A.15.9` supplies the smallest bounded-result return and supplier/receiver authority split; current `C.11` supplies choice over an already-available option set. Neither pattern bounds a professional problem-structuring and decision-support engagement or selects its next field-specific branch.	Reopen if direct choice or outside-result guidance gains the complete engagement-boundary move at equal or lower effort, or if practitioner use shows that the current separation prevents rather than improves a necessary authorized combined role.

PSD.1:12 - Relations

PSD.2 receives a bounded receiving inquiry when participants, concerns, affected Systems, duties, interests, representation, or dissent can change the return.
PSD.3 receives the case when materially different problem formulations imply different boundaries, evidence, values, or interventions. PSD.4 receives the deliberate included/excluded scope and reopen decision.
PSD.5–PSD.7 receive model, PSM-selection, and facilitation questions only when those results are current. PSD.8–PSD.12 receive alternatives and comparison questions after the engagement is bounded.
PSD.13 prepares and returns a recommendation without authorizing the later decision. PSD.14 relates changed premises and later observations to continuation, adjustment, evidence refresh, or reframing.
A.15.9 supplies the inspect-reuse-or-request boundary for one outside-practice result. A.10 governs any evidence, source, currentness, or bounded reliance claim actually used.
C.11 governs the later choice over an already-available option set. A.13, assignment, capability, permission, responsibility, and authority patterns govern those actual claims when the engagement relies on them.
Both PSD-CARD-01 and PSD-CARD-02 begin from this same engagement and authority meaning. The entries add unlike extended-use examples; they neither redefine PSD.1 nor impose a traversal order.