F.16 - Worked-Example Template (Cross-Domain)
“Show one claim, the actual values and relations that make it true or false, and enough evidence for a reader to replay the example.” Type: Architectural pattern.
Status: Stable
Builds on: E.10.D1 Recovering What “Context” Means in Use; F.1 for the source cut; F.0.1, F.2, F.3, and F.17 for exact source-local meaning when needed; F.7 for an optional comparison surface; F.9 only for actual semantic relations and bounded-use claims; F.10 for windows; F.15 for checks. Coordinates with. F.12 when a case evaluates promise acceptance; A.10 for evidence use; B.3 only for an actual named assurance claim; C.16.P and A.6.RCD when indicator or proxy wording hides a missing relation; E.13 only for optimized or decision-driving proxies; the direct Part C pattern for every illustrated subject; and the A.3, A.15, A.6.1, and B.1.5 method-and-work stack.
F.16:1 - Intent & applicability
Intent. Provide a one-page worked-example shape that starts with a recognizable question and practical gain, names the actual values and obtaining relations, exposes exact sources and evidence, and leaves a clear boundary.
Optional lexical cells and comparison tables help navigation only when they reduce reader effort.
Use this when. A pattern needs a compact example that crosses disciplines, source schemes, design-time and run-time boundaries, or several relation families and prose alone would hide the hand-offs.
Do not use this when. A shorter direct example already shows the claim. F.16 does not require a cross-source relation, table row, lexical cell, system-role claim, window, or SoD unless the case contains one.
Scope. The template organizes the explanation presented to the reader.
F.16:2 - Problem frame
Cross-domain examples fail when:
- The claim is missing. Sources and facts are listed, but the reader cannot tell what is being demonstrated.
- Words replace subjects. Process, role, service, or execution is used without identifying which value or relation the word denotes.
- Relations hide in prose. “Basically the same”, “implements”, “uses”, or “governs” conceals several different relation families.
- A table is asked to prove. Co-placement or row membership becomes sameness, permission, or evidence.
- Evidence and limits disappear. The page reaches a result or status without showing the evaluation, observation, source, loss, or boundary that supports it.
F.16:3 - Core idea (didactic)
A useful worked example is a compact case with nine visible parts:
- Question and gain — the recognizable practical problem and what changes when it is solved.
- Claim — one sentence that can be accepted, rejected, or qualified.
- Actual subjects — the values and entities to which the worked claim applies, with their relevant kinds and relation positions.
- Direct patterns — the patterns that define, constrain, or test each substantive value or relation.
- Source basis — exact sources and editions and the passages or interpretation schemes that matter.
- Obtaining relations — each named directly, with participants, direction, and limits.
- Evidence, result, and reliance — what supports the claim, what the evaluation returned, if one occurred, and what remains uncertain.
- Optional aids — F.17 cells when a local meaning needs a stable address and an F.7 table when comparison needs a display; their presence establishes no fact.
- Boundary and checks — where the example stops and the few checks a reader can replay.
F.16:4 - Minimal vocabulary
- Worked claim — the exact proposition demonstrated by the page.
- Actual subject — the value or entity to which a substantive claim applies.
- Obtaining relation — a relation supported for these exact participants; its appearance on the page does not create it.
- Source basis — exact source and edition, passage, and effective reference scheme needed by the claim.
- Evidence basis — the observations, source claims and independently established direct relations supporting the conclusion, represented in A.10’s descriptive evidence-provenance account with its bounded-reliance disposition. A separately current B.3 assurance claim retains its own result and limits.
- Evaluation result — when the case evaluates a promise or criterion, the result of applying its rule to the selected inputs, on the result scale that rule declares. Recover an exact A.6.1 application or dated evaluation Work when the case claims it or the governing evaluation rule requires it.
- Receiving use — the concrete decision, explanation, comparison, or action for which the result is used.
- SchemeSenseCell — the exact F.17 coordinate
<ReferenceScheme by value, LocalExpression, LocalSenseClaim>. A separate address resolves that coordinate when the receiving use needs it. - Comparison table — an optional F.7 display of exact entries and already obtaining relations.
- Window and separation of duties — F.10 or F.14 constraints included only when time, population, phase, or separation of duties changes the case.
F.16:5 - The one-page Worked-Example Canvas
Each item is a thought the reader needs, not a mandatory form field.
-
Title, working situation, and gain. Name the situation in ordinary language and say what practical error or decision the example resolves.
-
Worked claim. State one bounded claim. Example: “June availability is evaluated from observations of the defined service-delivery Work, not from the approved runbook.”
-
Actual subjects. Name the values and entities the claim concerns, distinguishing its subject, relation participants and result.
-
Direct pattern route. Beside each substantive claim, cite the pattern that defines, constrains, or tests the value or relation used by the claim. This prevents one vague word or generic Bridge from taking over several relation families.
-
Exact source basis. Cite the source and edition and, where local wording matters, the expression, claim, and effective scheme. Add an F.17 cell only if the exact local meaning needs a stable address for this use.
-
Obtaining relations. For each relation, name its exact participants, direction, basis, and loss. Use F.9 when the example tests a proposed relation between local meanings with different interpretation bases. Assert a Bridge only if its predicate holds.
Use the defining or testing pattern for MethodDescription membership, enactment, performed Work, measurement, evidence, fulfilment, kind, assignment, publication, transformation, and any load-bearing indicator relation. Use C.16.P to recover the proposed indicator relation. If no rule defines or tests it, return A.6.RCD
missing-governor. When a rule exists, state what the case evidence supports under it. -
Evidence, result, and limits. Show the independently established direct relations by which observations or sources support the claim. A.10 describes that path and classifies reliance on it for the bounded use; the cited basis must meet the evidence requirement of the claim and its direct rule.
When evaluation is part of the case, show the selected inputs, evaluation rule, result and any declared result scale. Recover the evaluation Work, enacted Method and exact application and result binding when the case claims them or its governing rule requires them, including the performance-history branch of A.2.3/F.12 promise evaluation.
Add F.10 status and a durable C.2.1 verdict assertion only for their selected uses. Use B.3 only for an actual named assurance claim. State the consequential limits of the conclusion.
-
Optional comparison surface. If two or more local meanings or source claims are hard to compare in prose, add one compact F.7 table. State the receiving-use conclusion separately. Omit the table when it adds ceremony but no clarity.
-
Micro-narrative and checks. Walk briefly from the working situation through the actual relations to the result. End with the direct subject checks that let the reader replay the claim, and add a non-use boundary only when F.19’s grounded-contribution test requires it. Add F.15 only when a finite naming slice needs its static or regression checks.
Memory rule. If the case cannot fit on one page or slide, reduce it to one claim or split it into linked examples. When shortening the case, keep the evidence or relation that makes its claim intelligible.
F.16:6 - Invariants
- Claim first. The example states one recognizable claim and practical gain before its architecture.
- Actual subjects. World, claim, evidence, status, and use assertions attach to actual values, not lexical cells or rows.
- Direct relations. Every substantive relation cites the pattern that defines or tests it; F.9 is conditional on a real local-meaning relation, and the word proxy never supplies an indicator relation.
- Optional aids. F.17 cells and F.7 tables appear only when they reduce reader effort; their presence or layout does not supply evidence.
- Source precision. Source and edition and the effective scheme are explicit where wording or interpretation changes the claim.
- Time and object distinctions. MethodDescription, Work, observation, and output remain distinct; windows are stated when they change the result.
- Agency precision. Every asserted dated Work uses the performer’s recovered A.13 core and independent A.15.1 admission. The core includes its obtaining assignment; add F.6 only when the worked claim needs precise assignment-bound attribution. Recover a separately needed local-kind or description claim under its own rule.
- Evidence, evaluation, and boundary. The page shows why the claim is supported. An evaluation retains its inputs, rule, result and applicable scale; any claimed or required evaluation Work, exact application, status and durable verdict assertion remain distinct. State the consequential limits of the result.
- Didactic parsimony. Every item changes the worked answer; optional machinery is omitted when it does not.
F.16:7 - Optional comparison panel
When a table helps, use a small F.7 panel:
| Named comparison or use | Exact local claims or cells | Obtaining relations | Loss and boundary | Evidence | Separate conclusion |
|---|
The panel may also be a two-source contrast with no relation asserted. A source name becomes a column only when the comparison benefits from it. A row does not make entries the same, and the page does not compute a row-wide permission or assurance score.
F.16:8 - Worked micro-example
F.16:8.0 - A complete constructed availability case
Question and gain. Does the supplied June observation support the service promise’s 99.9% availability threshold? The example separates the measurement and acceptance result from a runbook’s approval.
Inputs and direct rules. This is a constructed case, not a report of observed service performance. Take as independently admitted inputs the service-delivery Work Delivery-June and evaluation Work Evaluation-July-1, with their performers’ A.13 cores and A.15.1 admissions already established.
The exact A.2.3 promise clause Availability-June concerns the delivery Work’s outcome. Its acceptance specification uses all 43,200 minutes in the 30-day June window, no exclusions, and the rule availability = availableMinutes / 43200; Met means at least 0.999 and NotMet means less. These two values are the declared result scale.
Observation and basis. The supplied C.16 observation Availability-Trace-June directly concerns the promised available/unavailable characteristic of Delivery-June. Under its stipulated complete-coverage measurement basis, exactly the first 40 minutes are unavailable and the remaining 43,160 minutes are available.
The observation’s subject, characteristic, window, unit and complete coverage are premises of this case; an alarm log without that measurement basis could not replace it. No distinct indicator relation or F.9 correspondence is needed.
Application and result. During Evaluation-July-1, the evaluator enacts the declared evaluation Method through the A.6.1 application Availability-Application-June with availableMinutes=43160, totalMinutes=43200 and threshold=0.999. The application calculates 43160/43200 = 0.999074074..., compares it with 0.999, and returns Met on the declared acceptance scale.
If a separately recovered F.10 rule maps Met to RequirementStatus=Satisfied for this clause and window, that status can be reported; the application result alone is already the worked conclusion.
Evidence and bounded use. The descriptive A.10 account cites the exact observation-subject/measurement basis, promise-outcome fit, evaluation application and result-binding facts. Those independently established premises supply the evidence demanded by this acceptance rule, so pass is limited to explaining this June result. There is no separately named assurance claim, future-availability conclusion or permission claim.
Replay and boundary. The available and unavailable minutes sum to 43,200. The 40 unavailable minutes are below the permitted 43.2, and the computed ratio exceeds 0.999. With 50 unavailable minutes the same rule would return NotMet.
If complete coverage were missing, this worked evidence basis would not support either conclusion; return the rule’s missing-evidence question instead of inventing coverage or a pass.
F.16:8.1 - Applying the canvas to a multi-source availability case
The following is an application sketch. Supply the named source passages, exact relations, observations and result before claiming it is a completed worked example.
Title and situation. An alarm log does not by itself prove monthly uptime. Operations has an approved runbook and a month of IEC task and alarm logs; a service report must judge the exact ITIL promise-content claim.
Worked claim. June uptime is judged from admissible observations of the promised service outcome over the stated population and window. Alarm and command records may contribute evidence only through explicit relations and coverage limits.
Actual subjects and routes. The ITIL promise content and its promise-use, delivery, and fulfilment relations use A.2.3; the service-delivery Work and separate evaluation Work use A.15.1; the exact observations, availability characteristic, scale, and values use C.16; A.6.1 identifies the evaluation application and result binding; F.12 supplies the evaluation shape; A.10 describes the independently established evidence relations and qualifies the bounded reliance; and B.3 applies only when an actual named assurance claim is current.
To recognize the runbook as a MethodDescription under A.3.2, identify its one admitted Method as its exact subject and a substantive claim about that Method as a way of doing. Name the edition here only when it changes the evaluation result or replay.
Source basis. Cite the ITIL edition and promise passage, IEC edition and task and alarm passages, observation source and procedure, and any source-local meaning needed to interpret availability or alarm.
Relations and limits. State which observations concern which Work. First ask whether the observation and its measurement model directly concern the promised availability characteristic. If they do, use C.16 and A.10 and add no proxy.
If alarm-state intervals instead indicate a distinct unavailable-service characteristic, name both participants and the pattern that defines or tests that relation, with covered modes and blind spots. Use C.16.P to recover the relation and stop at A.6.RCD missing-governor when no such rule exists.
If the case uses an indicator, apply E.13 only when that indicator is optimized or drives a target, incentive, gate, release argument, reputation signal, repair, or decision. F.9 is needed only if the exact local meanings of alarm state and unavailable service are themselves related.
Result. A System performs evaluation Work, enacts the evaluation Method, and applies the declared availability rule to June’s in-scope observations. The A.6.1 application binds those inputs and returns a result on the declared acceptance scale.
Map the result to RequirementStatus=Satisfied or RequirementStatus=Violated only through the exact F.10 rule. If the evidence is inadequate, use EvidenceStatus=Inconclusive and leave RequirementStatus=Pending, or return the exact local result declared by the scale.
Create a verdict episteme only if another use needs it. State whether the promise was met, was not met, or cannot yet be judged from the evidence. The approved runbook establishes none of these results or statuses.
Checks. Actual subjects; a defining or testing pattern for each relation; direct-measurement-before-proxy; evaluation Work, application and result binding; declared result scale; separate EvidenceStatus and RequirementStatus; matching window and population; the indicator’s limit when one is used; and no row-created fact.
F.16:9 - Relations
Builds on:
- Use F.1 for the exact source cut and F.0.1 and F.17 for each local claim or durable address needed by the case.
- Use F.7 only for a readable comparison surface and F.9 only for relations that obtain between exact local meanings, together with the separate bounded-use claim.
- Use E.10.D1 to replace vague context wording with the source, scheme, scope, model use, working situation, comparison basis, or other value that changes the case.
- Use F.10 for status claims and their windows. Use each direct subject rule for the worked checks; add F.15 only for the example’s finite naming slice.
Coordinates with: A.2 with C.3 for a local system-role kind, A.2.1 for an assignment, F.4 only for a separately needed kind description, and F.6 only for precise performed-Work attribution; A.10 for evidence use; B.3 only when an actual named assurance claim is current; and the direct Part C pattern for every illustrated subject and relation.
Constrains: A cross-domain example may use this canvas or a faithful reduction. It names claim and gain first, actual values and relations next, and optional aids last. It never treats a lexical cell, table row, page layout, or generic Bridge as proof.
F.16:10 - Didactic distillation
“Start with one practical question and one claim. Name the actual values involved and the pattern that defines each relation. Cite the exact source and evidence. Use a local-meaning cell only when you need a stable address; use a comparison table only when it makes the argument easier to read; use F.9 only when a real relation between local meanings is part of the case. State the result, its limits, and a few checks. The page displays the reasoning—it does not create it.”
F.16:11 - Anti-patterns & remedies
| # | Anti-pattern | Symptom | Why harmful | Remedy |
|---|---|---|---|---|
| AP-1 | Architecture before question | The page opens with cells and routes but no recognizable problem. | Reader cannot tell what changes in practice. | Lead with situation, gain, and worked claim. |
| AP-2 | Mandatory row | Every example must span two sources and contain a Concept-Set row. | Ceremony replaces the simplest adequate explanation. | Make the table optional; use direct prose when enough. |
| AP-3 | Row-created sameness | Entries are “the same for this claim” because they share a row. | Layout becomes an ontological assertion. | State the actual relation and evidence, or show a contrast. |
| AP-4 | Cell as subject | A RoleDescription, promise, Work, observation, result, or status is anchored to one cell as though the cell were that value. | Lexical address replaces the subject. | Name the actual value; cite a cell only for local wording. |
| AP-5 | Generic Bridge | All cross-domain relations are labelled Bridges. | MethodDescription membership, description use, enactment, indicator, evidence, assignment, and fulfilment collapse. | Use each defining or testing pattern; recover unsupported indicator wording through C.16.P and A.6.RCD, and reserve F.9 for local-meaning relations. |
| AP-6 | Global trigger word | Role, function, process, or service appears without recovering its subject. | Polysemy hides objects and relations. | Apply E.10 and F.0.1 and then choose precise plain wording. |
| AP-7 | Design-time and run-time blur | A design description is narrated as Work. | Plan and occurrence collapse. | Use F.11, A.3, and A.15 and state the actual relation. |
| AP-8 | Edition haze | Source name lacks the edition that controls meaning. | The example cannot be replayed. | Name the source, edition, and relevant passage. |
| AP-9 | Evidence silence | The result appears without observations, source claims, or reliance basis. | Confidence cannot be assessed. | Show evidence use and consequential limits; add a non-use boundary only when F.19 requires the grounded correction. |
| AP-10 | Ontologist’s shorthand | Predicate notation replaces an ordinary explanation. | Precision becomes inaccessible to the cold reader. | Lead with plain language; retain compact notation when it clarifies a needed relation or calculation without replacing the explanation. |
F.16:12 - Cross-domain application sketches
These sketches select claims and routes for a later worked case. They become replayable examples only after their exact local meanings, subject facts, source passages, observations and results are supplied.
F.16:12.1 - OWL class and FCA formal concept
Situation and claim. A product catalogue uses an OWL class named Pump and an FCA formal concept with a similar label. The example asks whether their covered products may be compared for one catalogue query.
Recover both exact local claims. State the actual relation, if any, between their extensions for this data set and the difference between OWL subclass semantics and FCA lattice order. A small comparison table may display the evidence and limits. Similar labels and table co-placement do not establish identity or class inclusion.
F.16:12.2 - System role and RBAC role
Situation and claim. The same System performs Work under a local operator system-role assignment and also has permissions grouped by an RBAC role. The claim is that these are different subjects even when both use the word role.
Recover the local system-role kind under A.2 with C.3 and the exact assignment under A.2.1. For claimed dated Work, recover the performer’s A.13 core and independently admit the Work under A.15.1; add F.6 only for precise attribution through that same obtaining assignment. F.4 supplies a kind description only when one is needed.
Use the access-control pattern for the permission grouping, E.10.ROLE and F.0.1 for the trigger word, and the direct rule for any separation-of-duties constraint. A table row or shared spelling establishes neither sameness nor disjointness.
F.16:12.3 - Method description, Work, and service promise
Situation and claim. A build MethodDescription contains a target duration; dated build Work occurs; observations report actual duration; a service promise is evaluated for a calendar week.
Use A.3.2 for MethodDescription membership, A.15 and B.1.5 for delivery Work and any claimed evaluation Work or Method enactment, C.16 for observations and measured values, and A.2.3 with F.12 for promise evaluation. Use A.6.1 when the exact evaluation application and result binding are current; use A.10 for a descriptive account of independently established evidence relations and bounded reliance, and B.3 only for an actual named assurance claim.
Cite the MethodDescription edition only if it changes the result or replay, and state separately whether the Work used it. A comparison table may show promised and observed values, but it establishes neither Work, result, status, nor a verdict episteme.
F.16:13 - Safe reasoning moves
-
Recognize. State the working situation, failure, and practical gain.
-
Bound. Write one claim and one receiving use.
-
Name actual subjects. Replace vague words with the Systems, epistemes, Methods, Work, claims, observations, values, kinds, or assignments involved in the claim.
-
Route relations. Name the defining or testing pattern for each relation. For an alleged proxy, first test direct measurement; if a distinct indicator relation is needed and no rule supplies it, return A.6.RCD
missing-governor. -
Recover local wording. Cite the source, edition, and scheme; add an F.17 cell only when the receiving use needs a stable local-meaning address.
-
Show evidence and evaluation. Recover the direct evidence-use relations, their descriptive A.10 account and its bounded-reliance disposition.
When evaluation occurs, show its inputs, rule and supported result with any declared scale. Recover Work, Method and exact application/result bindings when claimed or required by the governing rule; add F.10 status or a durable verdict episteme only for their selected uses. Use B.3 only for an actual named assurance claim.
-
Use aids sparingly. Add an F.7 table only when it lowers reading cost; never infer from layout.
-
State limits. Include direction, loss, window, population, uncertainty, or non-use boundary that changes the conclusion.
-
Replay. Run focused checks for the case’s failure risks, including result and status separation when evaluation is present.
F.16:14 - Acceptance tests
F.16:14.1 - Static conformance
- SCR-F16-S01 (recognition). Situation, failure, gain, and worked claim are clear to a cold reader.
- SCR-F16-S02 (actual subjects). No cell, row, label, or record substitutes for an actual value.
- SCR-F16-S03 (direct relations). Every substantive relation names its exact participants and cites the pattern that defines or tests it; use A.6.RCD
missing-governorfor a proposed indicator when no rule defines or tests its relation. - SCR-F16-S04 (conditional F.9). F.9 appears when the worked claim needs a test of semantic correspondence between local meanings with different interpretation bases. Any asserted Bridge must meet its predicate.
- SCR-F16-S05 (source and evidence). Exact sources and editions and the A.10 evidence-use basis are visible where they affect the claim; B.3 appears only for an actual named assurance claim.
- SCR-F16-S06 (optional aids). Cells and tables are omitted when they do not reduce reader effort; their presence or layout is not treated as evidence.
- SCR-F16-S07 (time and agency). Windows, Work, System, assignment, and attribution are explicit when material.
- SCR-F16-S08 (one-page parsimony). Every included item changes the answer; a larger case is split without losing its basis.
- SCR-F16-S09 (evaluation architecture). When evaluation is present, its inputs, rule, supported result and applicable result scale are recoverable. Any claimed or rule-required performing System, evaluation Work, enacted Method and exact application/result binding satisfy their direct patterns; any selected EvidenceStatus, RequirementStatus or durable verdict assertion remains separate.
F.16:14.2 - Regression
- RSCR-F16-E01 (edition drift). Changed editions reopen only affected local claims and relations; no silent rewrite.
- RSCR-F16-E02 (relation change). A changed relation updates dependent conclusions and limits, not unrelated parts of the page.
- RSCR-F16-E03 (sense split). A changed local claim leaves no ambiguous F.17 address or table entry.
- RSCR-F16-E04 (window change). Changed cadence or population creates an explicit new evaluation boundary.
- RSCR-F16-E05 (plain-language guard). Repairing ontology does not replace readable sentences with avoidable symbolic or bureaucratic language.
F.16:15 - Migration notes
- Refactor source tours. Recover the practical question and one claim; keep only sources that change that answer.
- Replace mandatory rows. Retain a table only if it improves comparison; move substantive relations and conclusions back to their direct statements.
- Replace cell subjects. Name actual kinds, Systems, Work, claims, observations, values, and evidence; leave cells as optional wording addresses.
- Split generic Bridges. Restore MethodDescription membership, description use, enactment, performed-Work, measurement, evidence, reliance, fulfilment, kind, assignment, publication, or transformation claims as applicable. Add an indicator relation only when a defining or testing pattern supplies it; otherwise retain the blocker.
- Repair trigger words. Recover what role, function, process, service, or context denotes in each sentence before rewriting it.
- Keep the gain. After every ontological repair, reread the page as a cold practitioner and simplify any wording that became harder without gaining precision.
F.16:16 - Teaching variants
- Single-source direct case. One source, one actual value, one obtaining relation, one boundary; no table or F.9 relation.
- Two-source comparison. Two exact local claims, an optional F.7 row, and one separately supported conclusion.
- Triangulated evidence case. Promise, Work, and observation sources joined by their direct fulfilment, measurement, and evidence relations.
- Difference lesson. Two trigger-word uses kept distinct, with no relation asserted unless one obtains.
- Window primer. The same promise and Work evaluated under different explicit windows to show why result, status-use, and verdict-episteme identity must be checked separately.
F.16:17 - Didactic checklist
- Recognizable situation, failure, and gain?
- One bounded worked claim and receiving use?
- Actual values named?
- Defining or testing pattern named for every substantive relation, with direct measurement checked before any proxy?
- Exact sources and editions and evidence where material?
- F.17 cell or F.7 table only if it lowers reading cost?
- If the example tests a relation between local meanings with different interpretation bases, does it distinguish the F.9 test result from any asserted Bridge?
- Evaluation Work, application result, declared scale, status family, optional verdict episteme, window, population, agency, loss, and uncertainty visible where they change the answer?
- Plain language retained after ontological repair?
F.16:18 - Closing distillation
A useful worked example is one replayable argument: situation and gain → claim → actual subjects → direct relations → exact sources and evidence → result and limits → a few focused checks. Cells and tables are optional reading aids. They never replace the values, relations, evidence, or judgement that make the example true.