Master Art of Translation | The Meeting Minutes, Decisions and Action-Item Fidelity System — How Discussed, Proposed, Approved, Assigned, Deferred and Completed Preserve Outcomes

How do you translate meeting minutes, meeting decisions, action items and agenda notes accurately across languages without turning a discussion into an approval or a suggested task into an obligation? Readers searching for meeting minutes translation, translate minutes of meeting, action item versus decision, meeting summary translation, motion carried meaning, approved subject to conditions and multilingual meeting notes template need more than fluent equivalents. They need a dependable way to show exactly what the meeting considered, what it decided, what remains open and who is responsible for the next step.

The hidden problem is that a meeting record contains several different kinds of statements, even when the same sentence appears in ordinary language. An agenda item is not a decision; a proposal is not necessarily adopted; a seconded motion is not automatically carried; agreement in principle is not unconditional approval; a task assignment is not evidence of completion; and an approved set of minutes is not an endorsement of everything once discussed. Translation fails when it smooths these boundaries into a confident narrative. A reader may then spend money, release a document, close a safety issue or assign work on the strength of a decision that never happened.

This world-facing guide provides a repeatable translation method for minutes, decisions, motions, meeting outcomes, action trackers and follow-up notes. It begins with search intent, reveals the documentary mechanics, supplies terminology tests, explains quality risks and works through original fictional case files in multiple settings. The goal is to preserve procedural status, speaker and body authority, exact action scope, owner, due date, preconditions and evidence of closure while keeping the target text natural and usable. For the larger discipline, begin with Master Art of Translation | The Complete System for Moving Meaning Between Languages; this chapter concentrates only on the distinctive fidelity problems of meeting records.

• First identify the record: agenda, draft notes, approved minutes, verbatim transcript, decision register or action log.
• Separate five forces: raised, proposed, decided, assigned and verified; they cannot be treated as interchangeable.
• Record the decision boundary: adopted text, scope, conditions, dissent, abstentions or unresolved questions only when the source establishes them.
• Preserve accountability: action wording, named owner or role, deadline, evidence required and remaining approval.
• Test the translated version: someone who did not attend should not infer powers, promises or outcomes the meeting did not create.

1. The reader’s problem: minutes can be grammatically correct and operationally false

Meeting records are unusually vulnerable to over-translation because their short phrases do heavy procedural work. “For discussion” can become “for approval” in a hurried summary. “The proposal was supported” may refer to individual speakers, not the meeting as a body. “It was agreed that a revised paper would be brought back” approves a future review process, not necessarily the proposal inside that paper. A fluent sentence can be dangerous when it collapses the action state. The translator therefore protects what is authorised, not merely what reads smoothly.

Imagine a multilingual project review. The chair says the team has reviewed three options and wants a revised risk comparison. The notes mention strong enthusiasm for Option B but do not record a decision. A target-language summary saying “Option B was chosen” has moved the project from evidence gathering to execution. The mistake might not look like a vocabulary error, yet it can change procurement, staffing, budgets and downstream communications. That is why the first diagnostic question is not “Does this sound natural?” but “What can the reader legitimately do after reading it?”

The audience matters. A participant may remember the qualifications; a new team member sees only the record. An auditor reads minutes as evidence; an action owner reads them as instructions; a translator or interpreter may need to check whether a proposal was formal. The same source must preserve its limits for each audience. A good translation does not quietly repair a vague source by inventing unanimity, a due date, a responsible individual or a permission that the source never stated.

Meeting minutes do not have a single universal legal form. Working groups, public bodies, clubs, research committees, boards, remote teams and standards panels use different voting customs and records. This guide explains meaning and documentary accuracy; it does not prescribe any jurisdiction’s rules. Before translating formal decisions, obtain the relevant meeting procedure or governance rules, especially where words such as motion, quorum, abstention, resolution, ratification, consent and minute approval carry technical consequences.

2. A map of eight documentary objects

Agenda: an ordered intention about what a meeting will consider. An item included for discussion may never reach a decision. A note describing planned consideration must not be promoted into an accomplished act.

Contemporaneous notes: observations written during a meeting. They may include fragments, incomplete speaker labels and tentative wording. Their status is normally lower than a reviewed formal record unless a particular process says otherwise.

Draft minutes: a proposed record subject to correction. The word draft is a status of the record, not proof that the underlying meeting decisions were provisional. Conversely, a draft note may not conclusively establish that a formal motion passed.

Approved minutes: a record accepted under the relevant procedure as an accurate account of the meeting. Approval of minutes concerns the record. It does not automatically renew every previous decision or authorise every action discussed in the meeting.

Decision register: a structured list identifying a decision, its authorised scope, date, actor and qualifications. Translation must keep each field attached to the same decision and avoid merging adjacent decisions.

Action log: a task register with a responsible owner, due or review date, current state and evidence. Assignment, progress, completion and verification are distinct events even when compressed into one status label.

Verbatim transcript: a record of speech. Spoken assertions, jokes, false starts and contested statements remain contributions rather than adopted outcomes unless the process explicitly shows adoption.

Summary or briefing note: a selective digest prepared for a particular audience. A digest should disclose its summarising nature and must not pretend to reproduce the authority or full detail of approved minutes.

3. Why procedural force, not word-for-word resemblance, determines fidelity

Every relevant sentence carries several coordinates: actor (who said or authorised it), act (discussed, proposed, decided or assigned), object (what document or activity), status (open, pending, adopted, complete), scope (which people or cases it applies to), conditions (what must happen first), time (meeting date, effective point, due date), and evidence (what source proves the status). Losing just one coordinate may transform the record.

A useful mental model is a small state machine. An item may be scheduled; it may be introduced; evidence may be considered; a proposal may be tabled; the authorised body may vote or decide through another agreed procedure; tasks may be assigned; implementation may be reviewed; and the item may be closed. Not all organisations follow exactly this sequence. It is a diagnostic model, not a rule. A translation must never jump to a later state simply because that later state would be helpful or expected.

The lexical surface can be deceptively mild. “Noted” may mean merely acknowledged, formally recorded, or—as in certain institutional formulae—an act with specified consequences. “Supported” can mean endorsed personally or adopted institutionally. “Agreed” can report a consensus, a narrow procedure or a binding decision depending on context. “Carried” may mark a successful motion within one system but sound like physical transport elsewhere. For each term, read the surrounding motion, vote or decision fields before selecting the target-language verb.

Minutes have a dual function. They are an institutional memory of past conversation and a prospective map of future actions. A translation may be readable as historical narrative while misleading as an instruction sheet. Review therefore proceeds in both directions: ask what the text says happened, and separately ask what the reader is now authorised or asked to do. Any mismatch between those two readings deserves investigation.

This is also a problem of attribution. “Several members favoured a trial” does not mean “the committee approved a trial”. The first is attributed sentiment; the second is a collective decision. Translators should identify which verbs belong to speakers, which belong to a minute-taker’s summary and which belong to the constituted decision-making body. Where the original fails to specify an actor, a faithful target must often preserve the ambiguity and flag it for clarification instead of assigning one.

4. The source-to-target evidence envelope

Before drafting, collect the original agenda, approved or draft minutes, related motion text, decision register, task log and any later correction notice. Confirm the intended translation audience and whether the translated item is informational or itself part of an authorised operational workflow. Do not assume a video recording outranks signed minutes, or that a later email automatically amends a resolution. These are questions for the responsible body under its own procedures.

Build a one-page evidence envelope: original language, target language, document version, meeting identifier, meeting date, record status, source custodian, authority of the meeting, currentness date, known disputes, defined abbreviations and an explicit list of unavailable evidence. The translator does not need to decide institutional law, but does need to know which document is being rendered. This step prevents accidental merging of fragments from different meetings or revisions.

A bilingual term sheet should record candidate translations for agenda, minutes, motion, resolution, consent, seconded, carried, lost, deferred, abstained, assigned, action owner, outstanding, closed and verified. Add a sentence-level example for each, because a dictionary headword is insufficient. Formal equivalents may vary significantly by organisation and language variety. Preference must follow the reader’s actual document system, not a convenient universal thesaurus.

Keep an uncertainty register. For every disputed clause, quote the source fragment, describe its possible interpretations, identify the consequence of choosing wrongly and specify the evidence required to resolve it. A question such as “Did the body approve the revised budget or only authorise a revised budget paper?” is better than a vague request to “clarify minutes”. The purpose is to make source uncertainty visible to the person able to answer it.

Do not invent an official multilingual version where none exists. A translation made for comprehension should be clearly described as such. If an organisation needs a certified, approved or authoritative version, determine who can confer that status. The translator’s linguistic accuracy does not independently create legal or organisational authority.

5. The twelve-stage repeatable method for translating minutes and action logs

1. Establish the requested use

Ask whether the reader needs to understand past discussion, execute actions, archive evidence, prepare a formal translated record or quote an institutional decision. These are different risk profiles. A concise summary may serve orientation, but task execution demands the full owner, scope, status and conditions. Write the audience and intended use into the translation brief before choosing sentence style.

Output a reader-action statement such as “read for background only” or “use as the approved implementation record”. If the source custodian cannot confirm that status, mark the point unresolved rather than treating circulation as proof.

2. Identify the record layer

Check whether the source is an agenda, transcript, draft minute, approved minute, working note, decision extract or action tracker. Look for version numbers, approval notes and later corrections. A heading that says “Minutes” is not enough if the footer says draft and the action table is still open to revision.

Preserve a visible record label in the target. Where a document combines draft narrative with final decisions, handle each segment according to its evidence rather than forcing a single overall label.

3. Segment by meeting item

Separate each agenda identifier and any nested sub-item. A decision on Item 5.2 must not become approval of Item 5.1 or of a whole programme. Maintain numbered references even when the target language normally prefers a different heading order.

For each segment create a mini ledger with original ID, exact subject, source sentence, decision state, owner and evidence. Do not merge minutes from parallel committees simply because their subjects are similar.

4. Identify the acting body

Distinguish what a participant said from what the chair summarised and what the group formally decided. Check whether a subcommittee can recommend but not approve. “The panel recommended” may mean a referral to a board; “the board adopted” is the later act.

If authority is missing, avoid adding titles such as “approved by management”. Use a neutral actor or an editorial query, depending on the agreed conventions.

5. Trace the procedural verb

Underline discussed, noted, proposed, supported, seconded, moved, voted, carried, rejected, referred, deferred, ratified, authorised, tasked and closed. Determine which state each verb actually asserts. Words that appear synonymous in general prose may be non-equivalent in minutes.

Draft a plain-language semantic gloss before selecting the target-language form: “A proposal existed; no adoption is recorded,” for example. This gloss is a check, not text to insert automatically.

6. Extract objects and quantities

Capture each named paper, document revision, amount, threshold, option, procurement package, project phase, population and timeframe. A decision to approve a pilot for ten sites is not approval to launch everywhere. A task to evaluate a budget is not authority to spend it.

Check tables as well as sentences. Values can drift between languages when cells wrap, headings are reordered or footnotes no longer attach to the correct figure.

7. Mark qualifications

Highlight if, subject to, in principle, pending, except, without commitment, provided that, within budget, for review, for information and for recommendation. The qualifier may be the only reason a decision is limited. Treat each one as a separate controlled field.

Test a reader who skims only the translated decision headline: would that reader still see the qualifier? Put the restriction close to the relevant verb, not hidden in a distant footnote.

8. Map the action ledger

Record deliverable, owner or team, destination, due date, dependencies, acceptance authority and status. “Prepare a revision” and “publish the revision” are not the same action. “By the next meeting” needs a meeting identifier, not a guessed calendar date.

If the source has no named owner or date, preserve that absence and ask the custodian. Adding a plausible owner is a substantive alteration, not a helpful translation.

9. Choose target terminology

Use a documented glossary matched to the organisation’s process. Test terms in full clauses. In some languages a target word for “resolution” denotes a formal legal instrument; in others it can mean a generic solution. The context tells you whether the source really grants that degree of formality.

Keep recurring labels consistent across agenda, minutes and action log while allowing different syntax where it improves naturalness. Consistency should protect meaning, not force awkward literalism.

10. Draft with controlled directness

Prefer explicit actor and outcome where the source supplies them. Preserve passive voice when agency is genuinely unknown or institutionally irrelevant. Do not turn “was considered” into “the committee agreed”, or “to be returned” into “has been approved for return”.

For headings, use labels such as Discussion, Decision, Action and Outstanding Question, but only if the source supports that segmentation. A translated heading cannot manufacture a missing vote.

11. Review the states independently

Run separate passes for decision authenticity, action ownership, timing, conditions, quantities, attribution and version status. Each pass answers a concrete question; relying on general fluency review allows serious procedure errors to look polished.

Where practical, compare the target against a structured bilingual decision/action ledger rather than asking the reviewer to discover every commitment in an uninterrupted block of prose.

12. Release with traceable exceptions

Document unresolved queries and the approved disposition. Provide the final target in its requested format while retaining source identifiers and translation version. Clarify whether the output is a convenience translation, approved bilingual minutes or another recognised document type.

A final checkpoint asks whether the translation invites any action not authorised in the source. If the answer is uncertain, flag it before operational use rather than silently upgrading the text.

6. Forty terminology distinctions that prevent false decisions

Use the following glossary as a diagnostic map rather than a universal bilingual dictionary. The meaningful unit is often the entire clause, including the authorising body and qualifier. For specialised governance language, the approved organisational glossary and relevant procedure take precedence over general word similarity.

1. agenda item

Functional meaning: listed for consideration. Translation trap: An item heading is a plan to discuss, not the record of what occurred. Diagnostic: Check agenda number and actual minute outcome. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

2. for information

Functional meaning: circulated to inform. Translation trap: Readers may wrongly infer endorsement from inclusion. Diagnostic: Check for a separate recorded decision. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

3. for noting

Functional meaning: submitted to be noted. Translation trap: Noting may be procedural rather than approving. Diagnostic: Ask what the minute says was done with the paper. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

4. for approval

Functional meaning: submitted to seek approval. Translation trap: The phrase describes purpose, not result. Diagnostic: Look for adopted text and competent actor. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

5. discussed

Functional meaning: considered in conversation. Translation trap: Discussion can include strong support but no decision. Diagnostic: Compare with resolution or decision field. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

6. noted

Functional meaning: acknowledged or recorded. Translation trap: It does not always authorise a proposed action. Diagnostic: Consult the institution’s minute formula. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

7. raised

Functional meaning: introduced an issue. Translation trap: A concern being raised does not establish a finding. Diagnostic: Preserve speaker perspective when known. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

8. proposed

Functional meaning: suggested for adoption. Translation trap: The proposal may be rejected, deferred or unvoted. Diagnostic: Look for an outcome recorded separately. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

9. moved

Functional meaning: formally introduced a motion. Translation trap: A motion can exist without passage. Diagnostic: Check motion text and governing procedure. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

10. seconded

Functional meaning: supported motion consideration. Translation trap: A second is not an affirmative vote on merits. Diagnostic: Preserve the relationship to the motion. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

11. carried

Functional meaning: motion adopted under process. Translation trap: Do not translate as physical carrying. Diagnostic: Confirm recorded result and formal context. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

12. lost

Functional meaning: motion not adopted. Translation trap: It is not a missing document. Diagnostic: Retain any voting evidence if stated. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

13. abstained

Functional meaning: declined to vote for or against. Translation trap: It is not an absent participant or negative vote. Diagnostic: Keep counts in the correct column. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

14. unanimous

Functional meaning: without dissent in stated voting set. Translation trap: Unanimity cannot be inferred from no objections in prose. Diagnostic: Check attendance and decision procedure. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

15. consensus

Functional meaning: shared decision procedure or support. Translation trap: It need not be the same as unanimous formal voting. Diagnostic: Use process-aware language and evidence. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

16. in principle

Functional meaning: general support with limits. Translation trap: It does not necessarily permit execution. Diagnostic: Preserve the subsequent approval condition. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

17. conditional approval

Functional meaning: adoption limited by stated conditions. Translation trap: The condition is part of the operative decision. Diagnostic: Place condition beside approved activity. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

18. ratified

Functional meaning: formally confirmed earlier act. Translation trap: It is different from first proposing an act. Diagnostic: Check which earlier act is being confirmed. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

19. referred

Functional meaning: sent to another person or body. Translation trap: Referral is not endorsement of the substance. Diagnostic: Preserve destination and purpose of referral. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

20. deferred

Functional meaning: decision postponed. Translation trap: It is not rejection or cancellation. Diagnostic: Record trigger for future consideration. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

21. tabled

Functional meaning: formally put forward or postponed depending on usage. Translation trap: Meaning differs sharply between parliamentary varieties. Diagnostic: Use procedural evidence, never assume. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

22. withdrawn

Functional meaning: proposal removed from consideration. Translation trap: Withdrawn is not necessarily voted down. Diagnostic: Record who withdrew and at what stage. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

23. approved minutes

Functional meaning: accepted record of prior meeting. Translation trap: Approval of minutes is not blanket approval of contents as new proposals. Diagnostic: Keep approval object explicit. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

24. amended

Functional meaning: text changed under procedure. Translation trap: Amendment may affect a motion or the record itself. Diagnostic: Identify which object was amended. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

25. action item

Functional meaning: task arising from meeting. Translation trap: The task may be assigned while the proposal remains unapproved. Diagnostic: Record instruction and approvals separately. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

26. action owner

Functional meaning: person or role accountable for task. Translation trap: Owner does not always mean final approver. Diagnostic: Track delivery versus acceptance role. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

27. due date

Functional meaning: target deadline stated by source. Translation trap: A future review date may be different. Diagnostic: Avoid inventing dates from relative language. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

28. for review

Functional meaning: submitted for examination. Translation trap: Review is not final approval. Diagnostic: Keep reviewer’s authority distinct. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

29. for sign-off

Functional meaning: awaiting an approving act. Translation trap: Submission for sign-off does not mean signed off. Diagnostic: Identify pending actor and condition. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

30. outstanding

Functional meaning: still open or unresolved. Translation trap: An overdue item and an open item are not identical. Diagnostic: Check evidence for both status and time. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

31. carried forward

Functional meaning: kept for a later meeting. Translation trap: It may retain its current owner or require new decision. Diagnostic: Do not silently reset deadline. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

32. in progress

Functional meaning: work underway. Translation trap: Activity is not completion. Diagnostic: Keep completion evidence and status distinct. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

33. completed

Functional meaning: assigned work reported done. Translation trap: It may still await independent acceptance or verification. Diagnostic: Ask what completion criterion applied. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

34. verified

Functional meaning: evidence checked against criterion. Translation trap: Verification is different from self-reported completion. Diagnostic: Preserve verifier and evidence if stated. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

35. closed

Functional meaning: administratively ended in tracker. Translation trap: Closure may mean resolved, superseded, cancelled or accepted. Diagnostic: Check closure reason and remaining obligations. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

36. quorum

Functional meaning: minimum participation for valid acts under rules. Translation trap: A quorum count is jurisdiction- and body-specific. Diagnostic: Do not infer validity when quorum not stated. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

37. apologies

Functional meaning: regrets sent by absent invitee. Translation trap: Not all non-attendance is an apology. Diagnostic: Separate invited, present and absent. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

38. present for item

Functional meaning: attended only part of meeting. Translation trap: Participation at a particular vote may differ from meeting total. Diagnostic: Keep partial attendance boundaries. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

39. deciding vote

Functional meaning: vote with specified procedural effect. Translation trap: May refer to tie-breaking or another defined mechanism. Diagnostic: Consult specific rules before naming. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

40. minutes circulated

Functional meaning: record distributed. Translation trap: Circulation is not necessarily approval. Diagnostic: Check actual approval status. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

41. without objection

Functional meaning: no objection recorded in a given procedure. Translation trap: Do not automatically call it unanimity. Diagnostic: Confirm decision mechanism and attendees. In a bilingual record, record the source expression, its immediate action object and the chosen target expression together, so later reviewers can see why the term was selected. Where a term has procedural significance, preserve that significance rather than importing a seemingly stronger everyday synonym.

7. Translation choices across languages: equivalence requires the whole clause

In Spanish, se propuso describes something being proposed, while se aprobó reports approval. They are not interchangeable just because both can occur after a successful conversation. An English source saying “the committee discussed the proposal” should not be turned into an approval construction. A translator must also establish who approved what, whether the decision was conditional and whether the source is presenting the result as a formal resolution.

French offers a comparable high-risk distinction between proposé and approuvé. The construction approuvé sous réserve de can express approval subject to something, but the target must explain the actual condition when ambiguity would matter. The French lexical choice alone cannot identify whether the condition must be completed before publication, before implementation or before a future vote. That boundary remains a documentary fact to preserve.

In German, besprochen can describe something discussed, whereas beschlossen can signal a decision reached. These forms often appear close together in narratives of committee work. A speaker can say that a plan was discussed positively without meaning that it was adopted. Check the surrounding language for the relevant deciding body and record of the result; do not rely on an isolated machine-translated verb.

Chinese meeting records may distinguish 已讨论 (already discussed) from 已批准 (already approved). Yet the translator must be especially attentive to subjects and omitted agents; a sentence without an explicit actor cannot justify inserting “the board” in the target if several bodies could have made the decision. Record qualifiers such as 待批准 (pending approval) when they form part of the source, and do not replace them with a completed-action label.

Japanese may distinguish a matter having been 検討した (considered or examined) from 承認した (approved). These are brief illustrative forms, not universal replacements in every governance context. A translated sentence should capture the decision-making act, the appropriate level of politeness and the relevant body. If a record is merely a working memo, avoid adding the force of a formal resolution absent from the source.

Portuguese proposto and aprovado present a similar proposal/approval distinction. However, even correct terminology is inadequate if the source says “approved for consultation” and the target implies “approved for deployment”. The object of approval must remain exact. A complete translation review checks the whole clause across the source and target languages rather than marking word equivalents correct one at a time.

These are selected educational contrasts, not certified translations of a specific document. Regional governance conventions, institution-specific terms and formal procedures vary. For operational or legally significant minutes, a qualified translator and appropriate institutional reviewer should settle the final target wording. The invariant is the procedural distinction; a language may encode it using verbs, aspect, particles, passive constructions, headings or explicit explanatory phrases.

8. Thirty-six failure modes with a targeted repair

1. Discussion upgraded to adoption

Failure: A summary says the group “approved” an option when the source records debate only. Correction: Restore the discussion verb, show that no formal outcome was recorded and ask the custodian for the approved minute before permitting action. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

2. Proposal upgraded to commitment

Failure: A participant’s recommendation is translated as the organisation’s promise. Correction: Keep the proposer’s attribution, the proposal status and any outstanding decision. Identify who has authority to commit resources. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

3. Seconding treated as passage

Failure: A seconded motion is reported as carried despite an absent vote result. Correction: Preserve seconding as a procedural step, not the final decision; seek the vote outcome. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

4. Silence invented as unanimity

Failure: The record says “no concerns were raised” but the translation says “unanimously adopted”. Correction: Treat lack of recorded objection as evidence of the discussion only unless the procedure defines an adoption event. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

5. Conditional approval stripped

Failure: An approval “subject to security validation” appears as immediate authorisation. Correction: Reattach the condition to the same decision and test whether action can occur before validation. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

6. Approval object expanded

Failure: Approval of a draft communications plan becomes approval of a launch. Correction: Repeat the exact approved object and ensure the translation does not infer unstated implementation authority. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

7. Referral mistaken for endorsement

Failure: A paper sent to a subcommittee is rendered as accepted by the main body. Correction: Preserve destination, purpose and pending decision stage. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

8. Deferral mistaken for rejection

Failure: A delayed decision becomes “proposal failed”. Correction: State that determination was postponed, with the next evidence or meeting if known. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

9. Withdrawal mistaken for defeat

Failure: The proposer withdrew a motion but the target says it was voted down. Correction: Record withdrawal and its actor; do not manufacture voting outcomes. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

10. Motion amendment omitted

Failure: The source adopted an amended motion but the target repeats the original. Correction: Locate the final text voted upon and mark omitted or superseded wording clearly. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

11. Minute correction confused with changed policy

Failure: A correction to last month’s minutes is presented as a new policy reversal. Correction: Differentiate correcting the record from changing an existing substantive decision. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

12. Action owner invented

Failure: The original action says “to be allocated”; a translator inserts a department name. Correction: Preserve unassigned status and flag the gap for the secretariat. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

13. Deliverable broadened

Failure: “Prepare a feasibility note” is translated as “implement the feasibility project”. Correction: Keep the output artifact and authority boundary distinct from implementation. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

14. Due date inferred

Failure: “Before next review” becomes a specific calendar date that was never agreed. Correction: Retain the relative reference and request the actual review date for operational use. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

15. Completion confused with verification

Failure: An action marked completed is reported as audited and accepted. Correction: Track self-report, submission, review and sign-off as separate states. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

16. Closed interpreted as successful

Failure: A tracker closes an item after cancellation but translation says successfully finished. Correction: Preserve the actual reason for closure. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

17. Wrong actor credited

Failure: A recommendation by an adviser is attributed to a decision-making board. Correction: Keep the initiating speaker, advisory body and authority distinct. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

18. Abstentions added to opposition

Failure: Voting results move abstainers into the against column. Correction: Preserve all original vote categories and counts. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

19. Partial attendance hidden

Failure: Members present only for opening remarks are treated as voters on a later motion. Correction: Retain time and item boundaries and do not infer participation. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

20. Agenda title mistaken for outcome

Failure: “Approval of revised guidelines” appears as evidence that guidelines were approved. Correction: Read the actual minute body, not only the agenda heading. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

21. Draft minutes labelled final

Failure: A draft record is published in another language without its status marking. Correction: Keep draft visible and warn that the record may be corrected. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

22. Appendix detached

Failure: A decision refers to Annex C, but the translation omits or renames the annex. Correction: Preserve reference IDs and verify that readers can locate the exact attachment. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

23. Negation swallowed

Failure: “No further action authorised” becomes “further action authorised”. Correction: Run a focused negation and scope audit separate from general fluency. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

24. Budget amount misattached

Failure: A spending ceiling from one agenda item is assigned to another. Correction: Reconcile tables and source references line by line. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

25. Disagreement collapsed

Failure: One dissenting concern disappears from a consensual-sounding summary. Correction: Preserve meaningful disagreement when it is part of the authorised record. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

26. Quorum status assumed

Failure: The translation states the meeting was valid without source evidence of quorum. Correction: Do not pronounce procedural validity; retain the source’s statement or lack of one. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

27. Date and timezone shifted

Failure: An action due at a named local time is reworded with an unverified offset. Correction: Keep original time data and check against the dedicated time-zone method where necessary. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

28. AI draft mistaken for the minute

Failure: Machine-generated recap prose is presented as approved minutes. Correction: Label provenance and approval status; verify against authorised records. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

29. Resolution title implies broader powers

Failure: “Resolution on pilot evaluation” becomes a blanket implementation mandate. Correction: Separate the title from exact operative clauses and their boundaries. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

30. Suggested owner promoted to assignee

Failure: The chair asks who could take a task and the target names the first volunteer as accountable. Correction: Require explicit assignment evidence and capture any accepted ownership. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

31. Outstanding risk presented as resolved

Failure: A risk is described as mitigated because an action was assigned. Correction: An action commitment is not evidence the underlying risk has been controlled. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

32. Action count merged

Failure: Two tasks with separate owners become a single joint task. Correction: Retain discrete deliverables, identifiers, owners and dates. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

33. Circulation interpreted as ratification

Failure: The minutes were emailed to members, but the target says they were ratified. Correction: Record circulation as distribution and any later approval as a separate event. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

34. Status date invented

Failure: A tracker says “completed” without timestamp and the target inserts the meeting date. Correction: Avoid retroactive chronology; report only known time attributes. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

35. Recommendations mistaken for requirements

Failure: A working group recommends tests but the translated notes command every team to conduct them. Correction: Keep the governing force and recipient exact. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

36. No decision rewritten as indecision

Failure: The group deliberately resolved not to choose yet, but the target depicts confusion. Correction: Preserve deliberate procedural postponement when recorded. The reviewer should locate the precise source phrase, identify who spoke or acted, classify the state and write one sentence describing what the target text might wrongly lead a reader to do. A translation that is stylistically elegant still fails if a reader could take an unauthorised step. Record the correction against the same meeting item identifier, so the remedy can be audited and carried into later editions.

9. Extended worked diagnostics: 192 original fictional meeting-record cases across 24 settings

Every following source statement is invented for teaching. It is not an actual minute, legal finding, medical instruction, safety approval, procurement award or organisational policy. The corrected line is a controlled English meaning statement: it teaches what a target-language translation must preserve, not a ready-made translation into every language. Each case distinguishes the tempting incorrect reading from the source-supported reading, then gives a review method, downstream consequence and transfer exercise.

Workshop 1: Digital accessibility committee

The committee must not confuse an enthusiastic demonstration with accessibility acceptance. The working file concerns a keyboard-navigation accessibility checklist. The immediate question is whether the group may extend testing to the public beta within pilot features only. The key unresolved safeguard is completion of assistive-technology checks. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 1.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a keyboard-navigation accessibility checklist and consider whether to extend testing to the public beta for pilot features only. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved extend testing to the public beta.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to extend testing to the public beta for pilot features only; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Digital accessibility committee, the object is a keyboard-navigation accessibility checklist and the operational boundary is pilot features only. The failure matters because a user-facing feature could remain inaccessible. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to test results for low-vision navigation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 1.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a keyboard-navigation accessibility checklist and raised a user-facing feature could remain inaccessible. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that extend testing to the public beta.” Meaning-preserving controlled rendering: “Members discussed a keyboard-navigation accessibility checklist, raised a user-facing feature could remain inaccessible, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Digital accessibility committee, the object is a keyboard-navigation accessibility checklist and the operational boundary is pilot features only. The failure matters because a user-facing feature could remain inaccessible. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to test results for low-vision navigation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 1.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to extend testing to the public beta for pilot features only; the chair referred it for assessment against completion of assistive-technology checks.” Misleading rendering to reject: “The proposal to extend testing to the public beta was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to extend testing to the public beta for pilot features only was referred for assessment against completion of assistive-technology checks; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Digital accessibility committee, the object is a keyboard-navigation accessibility checklist and the operational boundary is pilot features only. The failure matters because a user-facing feature could remain inaccessible. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to test results for low-vision navigation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 1.4 — Formal motion with vote postponed

Fictional source record: “A motion to extend testing to the public beta for pilot features only was seconded. The chair deferred the vote until test results for low-vision navigation could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to extend testing to the public beta was seconded but the vote was deferred pending review of test results for low-vision navigation.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Digital accessibility committee, the object is a keyboard-navigation accessibility checklist and the operational boundary is pilot features only. The failure matters because a user-facing feature could remain inaccessible. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to test results for low-vision navigation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 1.5 — Qualified decision with limited scope

Fictional source record: “The committee approved extend testing to the public beta for pilot features only, subject to completion of assistive-technology checks. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a keyboard-navigation accessibility checklist immediately.” Meaning-preserving controlled rendering: “Approval covered extend testing to the public beta for pilot features only only, and depended on completion of assistive-technology checks; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Digital accessibility committee, the object is a keyboard-navigation accessibility checklist and the operational boundary is pilot features only. The failure matters because a user-facing feature could remain inaccessible. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to test results for low-vision navigation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 1.6 — Assigned preparation versus final approval

Fictional source record: “the accessibility lead will prepare a revised keyboard test report by 20 December 2026 and submit it to the release manager. No publication or final sign-off was recorded.” Misleading rendering to reject: “the release manager approved and published a revised keyboard test report.” Meaning-preserving controlled rendering: “the accessibility lead was assigned to prepare a revised keyboard test report by 20 December 2026 for submission to the release manager; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Digital accessibility committee, the object is a keyboard-navigation accessibility checklist and the operational boundary is pilot features only. The failure matters because a user-facing feature could remain inaccessible. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to test results for low-vision navigation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 1.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to extend testing to the public beta was deferred pending test results for low-vision navigation. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to extend testing to the public beta was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to extend testing to the public beta until test results for low-vision navigation became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Digital accessibility committee, the object is a keyboard-navigation accessibility checklist and the operational boundary is pilot features only. The failure matters because a user-facing feature could remain inaccessible. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to test results for low-vision navigation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 1.8 — Reported completion pending verification

Fictional source record: “the accessibility lead marked a revised keyboard test report complete and submitted the draft. Verification by the release manager against test results for low-vision navigation remains outstanding.” Misleading rendering to reject: “the release manager verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the accessibility lead reported completion and submission of a revised keyboard test report; verification by the release manager remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Digital accessibility committee, the object is a keyboard-navigation accessibility checklist and the operational boundary is pilot features only. The failure matters because a user-facing feature could remain inaccessible. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to test results for low-vision navigation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 2: Scientific publishing editorial board

A correction process changes a scholarly record and should not be implied by mere editorial discussion. The working file concerns a correction notice for a research article. The immediate question is whether the group may publish the correction within the article’s data caption. The key unresolved safeguard is verification by the corresponding author and editors. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 2.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a correction notice for a research article and consider whether to publish the correction for the article’s data caption. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved publish the correction.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to publish the correction for the article’s data caption; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Scientific publishing editorial board, the object is a correction notice for a research article and the operational boundary is the article’s data caption. The failure matters because readers may cite an erroneous figure. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the final caption proof, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 2.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a correction notice for a research article and raised readers may cite an erroneous figure. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that publish the correction.” Meaning-preserving controlled rendering: “Members discussed a correction notice for a research article, raised readers may cite an erroneous figure, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Scientific publishing editorial board, the object is a correction notice for a research article and the operational boundary is the article’s data caption. The failure matters because readers may cite an erroneous figure. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the final caption proof, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 2.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to publish the correction for the article’s data caption; the chair referred it for assessment against verification by the corresponding author and editors.” Misleading rendering to reject: “The proposal to publish the correction was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to publish the correction for the article’s data caption was referred for assessment against verification by the corresponding author and editors; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Scientific publishing editorial board, the object is a correction notice for a research article and the operational boundary is the article’s data caption. The failure matters because readers may cite an erroneous figure. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the final caption proof, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 2.4 — Formal motion with vote postponed

Fictional source record: “A motion to publish the correction for the article’s data caption was seconded. The chair deferred the vote until the final caption proof could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to publish the correction was seconded but the vote was deferred pending review of the final caption proof.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Scientific publishing editorial board, the object is a correction notice for a research article and the operational boundary is the article’s data caption. The failure matters because readers may cite an erroneous figure. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the final caption proof, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 2.5 — Qualified decision with limited scope

Fictional source record: “The committee approved publish the correction for the article’s data caption, subject to verification by the corresponding author and editors. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a correction notice for a research article immediately.” Meaning-preserving controlled rendering: “Approval covered publish the correction for the article’s data caption only, and depended on verification by the corresponding author and editors; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Scientific publishing editorial board, the object is a correction notice for a research article and the operational boundary is the article’s data caption. The failure matters because readers may cite an erroneous figure. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the final caption proof, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 2.6 — Assigned preparation versus final approval

Fictional source record: “the managing editor will prepare the corrected caption and citation notice by 14 January 2027 and submit it to the editor-in-chief. No publication or final sign-off was recorded.” Misleading rendering to reject: “the editor-in-chief approved and published the corrected caption and citation notice.” Meaning-preserving controlled rendering: “the managing editor was assigned to prepare the corrected caption and citation notice by 14 January 2027 for submission to the editor-in-chief; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Scientific publishing editorial board, the object is a correction notice for a research article and the operational boundary is the article’s data caption. The failure matters because readers may cite an erroneous figure. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the final caption proof, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 2.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to publish the correction was deferred pending the final caption proof. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to publish the correction was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to publish the correction until the final caption proof became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Scientific publishing editorial board, the object is a correction notice for a research article and the operational boundary is the article’s data caption. The failure matters because readers may cite an erroneous figure. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the final caption proof, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 2.8 — Reported completion pending verification

Fictional source record: “the managing editor marked the corrected caption and citation notice complete and submitted the draft. Verification by the editor-in-chief against the final caption proof remains outstanding.” Misleading rendering to reject: “the editor-in-chief verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the managing editor reported completion and submission of the corrected caption and citation notice; verification by the editor-in-chief remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Scientific publishing editorial board, the object is a correction notice for a research article and the operational boundary is the article’s data caption. The failure matters because readers may cite an erroneous figure. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the final caption proof, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 3: Museum interpretation council

The conservation record and visitor text serve different functions and must not be conflated. The working file concerns a bilingual object label for a travelling collection. The immediate question is whether the group may approve the revised label within the travelling exhibition gallery. The key unresolved safeguard is confirmation of provenance wording. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 3.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a bilingual object label for a travelling collection and consider whether to approve the revised label for the travelling exhibition gallery. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved approve the revised label.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to approve the revised label for the travelling exhibition gallery; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Museum interpretation council, the object is a bilingual object label for a travelling collection and the operational boundary is the travelling exhibition gallery. The failure matters because a label might imply an ownership claim not established. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the accession file and provenance notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 3.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a bilingual object label for a travelling collection and raised a label might imply an ownership claim not established. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that approve the revised label.” Meaning-preserving controlled rendering: “Members discussed a bilingual object label for a travelling collection, raised a label might imply an ownership claim not established, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Museum interpretation council, the object is a bilingual object label for a travelling collection and the operational boundary is the travelling exhibition gallery. The failure matters because a label might imply an ownership claim not established. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the accession file and provenance notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 3.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to approve the revised label for the travelling exhibition gallery; the chair referred it for assessment against confirmation of provenance wording.” Misleading rendering to reject: “The proposal to approve the revised label was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to approve the revised label for the travelling exhibition gallery was referred for assessment against confirmation of provenance wording; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Museum interpretation council, the object is a bilingual object label for a travelling collection and the operational boundary is the travelling exhibition gallery. The failure matters because a label might imply an ownership claim not established. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the accession file and provenance notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 3.4 — Formal motion with vote postponed

Fictional source record: “A motion to approve the revised label for the travelling exhibition gallery was seconded. The chair deferred the vote until the accession file and provenance notes could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to approve the revised label was seconded but the vote was deferred pending review of the accession file and provenance notes.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Museum interpretation council, the object is a bilingual object label for a travelling collection and the operational boundary is the travelling exhibition gallery. The failure matters because a label might imply an ownership claim not established. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the accession file and provenance notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 3.5 — Qualified decision with limited scope

Fictional source record: “The committee approved approve the revised label for the travelling exhibition gallery, subject to confirmation of provenance wording. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a bilingual object label for a travelling collection immediately.” Meaning-preserving controlled rendering: “Approval covered approve the revised label for the travelling exhibition gallery only, and depended on confirmation of provenance wording; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Museum interpretation council, the object is a bilingual object label for a travelling collection and the operational boundary is the travelling exhibition gallery. The failure matters because a label might imply an ownership claim not established. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the accession file and provenance notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 3.6 — Assigned preparation versus final approval

Fictional source record: “the collections curator will prepare a sourced label proof by 9 February 2027 and submit it to the collections director. No publication or final sign-off was recorded.” Misleading rendering to reject: “the collections director approved and published a sourced label proof.” Meaning-preserving controlled rendering: “the collections curator was assigned to prepare a sourced label proof by 9 February 2027 for submission to the collections director; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Museum interpretation council, the object is a bilingual object label for a travelling collection and the operational boundary is the travelling exhibition gallery. The failure matters because a label might imply an ownership claim not established. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the accession file and provenance notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 3.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to approve the revised label was deferred pending the accession file and provenance notes. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to approve the revised label was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to approve the revised label until the accession file and provenance notes became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Museum interpretation council, the object is a bilingual object label for a travelling collection and the operational boundary is the travelling exhibition gallery. The failure matters because a label might imply an ownership claim not established. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the accession file and provenance notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 3.8 — Reported completion pending verification

Fictional source record: “the collections curator marked a sourced label proof complete and submitted the draft. Verification by the collections director against the accession file and provenance notes remains outstanding.” Misleading rendering to reject: “the collections director verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the collections curator reported completion and submission of a sourced label proof; verification by the collections director remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Museum interpretation council, the object is a bilingual object label for a travelling collection and the operational boundary is the travelling exhibition gallery. The failure matters because a label might imply an ownership claim not established. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the accession file and provenance notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 4: Research ethics working group

A readability recommendation cannot amend the authorised research protocol by itself. The working file concerns a participant-information sheet. The immediate question is whether the group may test an alternative wording within a small readability pilot. The key unresolved safeguard is review of consent comprehension. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 4.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a participant-information sheet and consider whether to test an alternative wording for a small readability pilot. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved test an alternative wording.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to test an alternative wording for a small readability pilot; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Research ethics working group, the object is a participant-information sheet and the operational boundary is a small readability pilot. The failure matters because participant consent language could misstate what is voluntary. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the approved protocol and change log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 4.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a participant-information sheet and raised participant consent language could misstate what is voluntary. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that test an alternative wording.” Meaning-preserving controlled rendering: “Members discussed a participant-information sheet, raised participant consent language could misstate what is voluntary, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Research ethics working group, the object is a participant-information sheet and the operational boundary is a small readability pilot. The failure matters because participant consent language could misstate what is voluntary. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the approved protocol and change log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 4.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to test an alternative wording for a small readability pilot; the chair referred it for assessment against review of consent comprehension.” Misleading rendering to reject: “The proposal to test an alternative wording was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to test an alternative wording for a small readability pilot was referred for assessment against review of consent comprehension; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Research ethics working group, the object is a participant-information sheet and the operational boundary is a small readability pilot. The failure matters because participant consent language could misstate what is voluntary. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the approved protocol and change log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 4.4 — Formal motion with vote postponed

Fictional source record: “A motion to test an alternative wording for a small readability pilot was seconded. The chair deferred the vote until the approved protocol and change log could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to test an alternative wording was seconded but the vote was deferred pending review of the approved protocol and change log.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Research ethics working group, the object is a participant-information sheet and the operational boundary is a small readability pilot. The failure matters because participant consent language could misstate what is voluntary. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the approved protocol and change log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 4.5 — Qualified decision with limited scope

Fictional source record: “The committee approved test an alternative wording for a small readability pilot, subject to review of consent comprehension. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a participant-information sheet immediately.” Meaning-preserving controlled rendering: “Approval covered test an alternative wording for a small readability pilot only, and depended on review of consent comprehension; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Research ethics working group, the object is a participant-information sheet and the operational boundary is a small readability pilot. The failure matters because participant consent language could misstate what is voluntary. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the approved protocol and change log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 4.6 — Assigned preparation versus final approval

Fictional source record: “the research coordinator will prepare an anonymised comprehension report by 16 March 2027 and submit it to the ethics chair. No publication or final sign-off was recorded.” Misleading rendering to reject: “the ethics chair approved and published an anonymised comprehension report.” Meaning-preserving controlled rendering: “the research coordinator was assigned to prepare an anonymised comprehension report by 16 March 2027 for submission to the ethics chair; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Research ethics working group, the object is a participant-information sheet and the operational boundary is a small readability pilot. The failure matters because participant consent language could misstate what is voluntary. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the approved protocol and change log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 4.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to test an alternative wording was deferred pending the approved protocol and change log. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to test an alternative wording was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to test an alternative wording until the approved protocol and change log became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Research ethics working group, the object is a participant-information sheet and the operational boundary is a small readability pilot. The failure matters because participant consent language could misstate what is voluntary. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the approved protocol and change log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 4.8 — Reported completion pending verification

Fictional source record: “the research coordinator marked an anonymised comprehension report complete and submitted the draft. Verification by the ethics chair against the approved protocol and change log remains outstanding.” Misleading rendering to reject: “the ethics chair verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the research coordinator reported completion and submission of an anonymised comprehension report; verification by the ethics chair remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Research ethics working group, the object is a participant-information sheet and the operational boundary is a small readability pilot. The failure matters because participant consent language could misstate what is voluntary. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the approved protocol and change log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 5: Open-source release council

The meeting’s enthusiasm is not equivalent to the repository’s release approval. The working file concerns release notes for a software security patch. The immediate question is whether the group may publish the patch announcement within the staging release. The key unresolved safeguard is successful regression and dependency tests. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 5.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review release notes for a software security patch and consider whether to publish the patch announcement for the staging release. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved publish the patch announcement.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to publish the patch announcement for the staging release; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Open-source release council, the object is release notes for a software security patch and the operational boundary is the staging release. The failure matters because users may install a build not cleared for release. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the versioned build hash, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 5.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed release notes for a software security patch and raised users may install a build not cleared for release. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that publish the patch announcement.” Meaning-preserving controlled rendering: “Members discussed release notes for a software security patch, raised users may install a build not cleared for release, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Open-source release council, the object is release notes for a software security patch and the operational boundary is the staging release. The failure matters because users may install a build not cleared for release. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the versioned build hash, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 5.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to publish the patch announcement for the staging release; the chair referred it for assessment against successful regression and dependency tests.” Misleading rendering to reject: “The proposal to publish the patch announcement was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to publish the patch announcement for the staging release was referred for assessment against successful regression and dependency tests; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Open-source release council, the object is release notes for a software security patch and the operational boundary is the staging release. The failure matters because users may install a build not cleared for release. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the versioned build hash, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 5.4 — Formal motion with vote postponed

Fictional source record: “A motion to publish the patch announcement for the staging release was seconded. The chair deferred the vote until the versioned build hash could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to publish the patch announcement was seconded but the vote was deferred pending review of the versioned build hash.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Open-source release council, the object is release notes for a software security patch and the operational boundary is the staging release. The failure matters because users may install a build not cleared for release. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the versioned build hash, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 5.5 — Qualified decision with limited scope

Fictional source record: “The committee approved publish the patch announcement for the staging release, subject to successful regression and dependency tests. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of release notes for a software security patch immediately.” Meaning-preserving controlled rendering: “Approval covered publish the patch announcement for the staging release only, and depended on successful regression and dependency tests; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Open-source release council, the object is release notes for a software security patch and the operational boundary is the staging release. The failure matters because users may install a build not cleared for release. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the versioned build hash, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 5.6 — Assigned preparation versus final approval

Fictional source record: “the release maintainer will prepare a signed staging-test summary by 28 November 2026 and submit it to the release steward. No publication or final sign-off was recorded.” Misleading rendering to reject: “the release steward approved and published a signed staging-test summary.” Meaning-preserving controlled rendering: “the release maintainer was assigned to prepare a signed staging-test summary by 28 November 2026 for submission to the release steward; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Open-source release council, the object is release notes for a software security patch and the operational boundary is the staging release. The failure matters because users may install a build not cleared for release. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the versioned build hash, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 5.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to publish the patch announcement was deferred pending the versioned build hash. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to publish the patch announcement was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to publish the patch announcement until the versioned build hash became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Open-source release council, the object is release notes for a software security patch and the operational boundary is the staging release. The failure matters because users may install a build not cleared for release. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the versioned build hash, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 5.8 — Reported completion pending verification

Fictional source record: “the release maintainer marked a signed staging-test summary complete and submitted the draft. Verification by the release steward against the versioned build hash remains outstanding.” Misleading rendering to reject: “the release steward verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the release maintainer reported completion and submission of a signed staging-test summary; verification by the release steward remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Open-source release council, the object is release notes for a software security patch and the operational boundary is the staging release. The failure matters because users may install a build not cleared for release. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the versioned build hash, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 6: Humanitarian logistics coordination group

A trial shipment must not become permission for an unrestricted distribution. The working file concerns a relief-package packing manifest. The immediate question is whether the group may distribute revised packing instructions within one test shipment. The key unresolved safeguard is reconciliation with warehouse counts. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 6.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a relief-package packing manifest and consider whether to distribute revised packing instructions for one test shipment. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved distribute revised packing instructions.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to distribute revised packing instructions for one test shipment; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Humanitarian logistics coordination group, the object is a relief-package packing manifest and the operational boundary is one test shipment. The failure matters because volunteers could dispatch the wrong supplies. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the depot count discrepancy report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 6.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a relief-package packing manifest and raised volunteers could dispatch the wrong supplies. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that distribute revised packing instructions.” Meaning-preserving controlled rendering: “Members discussed a relief-package packing manifest, raised volunteers could dispatch the wrong supplies, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Humanitarian logistics coordination group, the object is a relief-package packing manifest and the operational boundary is one test shipment. The failure matters because volunteers could dispatch the wrong supplies. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the depot count discrepancy report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 6.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to distribute revised packing instructions for one test shipment; the chair referred it for assessment against reconciliation with warehouse counts.” Misleading rendering to reject: “The proposal to distribute revised packing instructions was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to distribute revised packing instructions for one test shipment was referred for assessment against reconciliation with warehouse counts; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Humanitarian logistics coordination group, the object is a relief-package packing manifest and the operational boundary is one test shipment. The failure matters because volunteers could dispatch the wrong supplies. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the depot count discrepancy report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 6.4 — Formal motion with vote postponed

Fictional source record: “A motion to distribute revised packing instructions for one test shipment was seconded. The chair deferred the vote until the depot count discrepancy report could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to distribute revised packing instructions was seconded but the vote was deferred pending review of the depot count discrepancy report.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Humanitarian logistics coordination group, the object is a relief-package packing manifest and the operational boundary is one test shipment. The failure matters because volunteers could dispatch the wrong supplies. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the depot count discrepancy report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 6.5 — Qualified decision with limited scope

Fictional source record: “The committee approved distribute revised packing instructions for one test shipment, subject to reconciliation with warehouse counts. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a relief-package packing manifest immediately.” Meaning-preserving controlled rendering: “Approval covered distribute revised packing instructions for one test shipment only, and depended on reconciliation with warehouse counts; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Humanitarian logistics coordination group, the object is a relief-package packing manifest and the operational boundary is one test shipment. The failure matters because volunteers could dispatch the wrong supplies. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the depot count discrepancy report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 6.6 — Assigned preparation versus final approval

Fictional source record: “the logistics coordinator will prepare a signed manifest comparison by 12 December 2026 and submit it to the programme director. No publication or final sign-off was recorded.” Misleading rendering to reject: “the programme director approved and published a signed manifest comparison.” Meaning-preserving controlled rendering: “the logistics coordinator was assigned to prepare a signed manifest comparison by 12 December 2026 for submission to the programme director; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Humanitarian logistics coordination group, the object is a relief-package packing manifest and the operational boundary is one test shipment. The failure matters because volunteers could dispatch the wrong supplies. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the depot count discrepancy report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 6.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to distribute revised packing instructions was deferred pending the depot count discrepancy report. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to distribute revised packing instructions was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to distribute revised packing instructions until the depot count discrepancy report became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Humanitarian logistics coordination group, the object is a relief-package packing manifest and the operational boundary is one test shipment. The failure matters because volunteers could dispatch the wrong supplies. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the depot count discrepancy report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 6.8 — Reported completion pending verification

Fictional source record: “the logistics coordinator marked a signed manifest comparison complete and submitted the draft. Verification by the programme director against the depot count discrepancy report remains outstanding.” Misleading rendering to reject: “the programme director verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the logistics coordinator reported completion and submission of a signed manifest comparison; verification by the programme director remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Humanitarian logistics coordination group, the object is a relief-package packing manifest and the operational boundary is one test shipment. The failure matters because volunteers could dispatch the wrong supplies. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the depot count discrepancy report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 7: Educational assessment design panel

Changing rubric language can change scoring decisions even without changing numbers. The working file concerns a translated marking rubric. The immediate question is whether the group may replace one ambiguous scoring descriptor within the pilot examination task. The key unresolved safeguard is moderation across independent markers. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 7.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a translated marking rubric and consider whether to replace one ambiguous scoring descriptor for the pilot examination task. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved replace one ambiguous scoring descriptor.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to replace one ambiguous scoring descriptor for the pilot examination task; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Educational assessment design panel, the object is a translated marking rubric and the operational boundary is the pilot examination task. The failure matters because students could be graded using an unapproved criterion. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the moderation notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 7.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a translated marking rubric and raised students could be graded using an unapproved criterion. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that replace one ambiguous scoring descriptor.” Meaning-preserving controlled rendering: “Members discussed a translated marking rubric, raised students could be graded using an unapproved criterion, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Educational assessment design panel, the object is a translated marking rubric and the operational boundary is the pilot examination task. The failure matters because students could be graded using an unapproved criterion. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the moderation notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 7.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to replace one ambiguous scoring descriptor for the pilot examination task; the chair referred it for assessment against moderation across independent markers.” Misleading rendering to reject: “The proposal to replace one ambiguous scoring descriptor was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to replace one ambiguous scoring descriptor for the pilot examination task was referred for assessment against moderation across independent markers; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Educational assessment design panel, the object is a translated marking rubric and the operational boundary is the pilot examination task. The failure matters because students could be graded using an unapproved criterion. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the moderation notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 7.4 — Formal motion with vote postponed

Fictional source record: “A motion to replace one ambiguous scoring descriptor for the pilot examination task was seconded. The chair deferred the vote until the moderation notes could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to replace one ambiguous scoring descriptor was seconded but the vote was deferred pending review of the moderation notes.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Educational assessment design panel, the object is a translated marking rubric and the operational boundary is the pilot examination task. The failure matters because students could be graded using an unapproved criterion. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the moderation notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 7.5 — Qualified decision with limited scope

Fictional source record: “The committee approved replace one ambiguous scoring descriptor for the pilot examination task, subject to moderation across independent markers. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a translated marking rubric immediately.” Meaning-preserving controlled rendering: “Approval covered replace one ambiguous scoring descriptor for the pilot examination task only, and depended on moderation across independent markers; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Educational assessment design panel, the object is a translated marking rubric and the operational boundary is the pilot examination task. The failure matters because students could be graded using an unapproved criterion. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the moderation notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 7.6 — Assigned preparation versus final approval

Fictional source record: “the assessment specialist will prepare a double-marking comparison by 7 January 2027 and submit it to the assessment director. No publication or final sign-off was recorded.” Misleading rendering to reject: “the assessment director approved and published a double-marking comparison.” Meaning-preserving controlled rendering: “the assessment specialist was assigned to prepare a double-marking comparison by 7 January 2027 for submission to the assessment director; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Educational assessment design panel, the object is a translated marking rubric and the operational boundary is the pilot examination task. The failure matters because students could be graded using an unapproved criterion. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the moderation notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 7.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to replace one ambiguous scoring descriptor was deferred pending the moderation notes. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to replace one ambiguous scoring descriptor was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to replace one ambiguous scoring descriptor until the moderation notes became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Educational assessment design panel, the object is a translated marking rubric and the operational boundary is the pilot examination task. The failure matters because students could be graded using an unapproved criterion. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the moderation notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 7.8 — Reported completion pending verification

Fictional source record: “the assessment specialist marked a double-marking comparison complete and submitted the draft. Verification by the assessment director against the moderation notes remains outstanding.” Misleading rendering to reject: “the assessment director verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the assessment specialist reported completion and submission of a double-marking comparison; verification by the assessment director remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Educational assessment design panel, the object is a translated marking rubric and the operational boundary is the pilot examination task. The failure matters because students could be graded using an unapproved criterion. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the moderation notes, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 8: Digital privacy notice committee

A communications change must not be mistaken for evidence of operational deletion. The working file concerns a retention-policy summary for service users. The immediate question is whether the group may issue the clearer notice within a single service workflow. The key unresolved safeguard is review of deletion and access rights. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 8.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a retention-policy summary for service users and consider whether to issue the clearer notice for a single service workflow. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved issue the clearer notice.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to issue the clearer notice for a single service workflow; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Digital privacy notice committee, the object is a retention-policy summary for service users and the operational boundary is a single service workflow. The failure matters because readers could think their data have already been erased. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the retention schedule and service map, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 8.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a retention-policy summary for service users and raised readers could think their data have already been erased. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that issue the clearer notice.” Meaning-preserving controlled rendering: “Members discussed a retention-policy summary for service users, raised readers could think their data have already been erased, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Digital privacy notice committee, the object is a retention-policy summary for service users and the operational boundary is a single service workflow. The failure matters because readers could think their data have already been erased. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the retention schedule and service map, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 8.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to issue the clearer notice for a single service workflow; the chair referred it for assessment against review of deletion and access rights.” Misleading rendering to reject: “The proposal to issue the clearer notice was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to issue the clearer notice for a single service workflow was referred for assessment against review of deletion and access rights; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Digital privacy notice committee, the object is a retention-policy summary for service users and the operational boundary is a single service workflow. The failure matters because readers could think their data have already been erased. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the retention schedule and service map, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 8.4 — Formal motion with vote postponed

Fictional source record: “A motion to issue the clearer notice for a single service workflow was seconded. The chair deferred the vote until the retention schedule and service map could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to issue the clearer notice was seconded but the vote was deferred pending review of the retention schedule and service map.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Digital privacy notice committee, the object is a retention-policy summary for service users and the operational boundary is a single service workflow. The failure matters because readers could think their data have already been erased. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the retention schedule and service map, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 8.5 — Qualified decision with limited scope

Fictional source record: “The committee approved issue the clearer notice for a single service workflow, subject to review of deletion and access rights. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a retention-policy summary for service users immediately.” Meaning-preserving controlled rendering: “Approval covered issue the clearer notice for a single service workflow only, and depended on review of deletion and access rights; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Digital privacy notice committee, the object is a retention-policy summary for service users and the operational boundary is a single service workflow. The failure matters because readers could think their data have already been erased. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the retention schedule and service map, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 8.6 — Assigned preparation versus final approval

Fictional source record: “the privacy programme lead will prepare an annotated notice draft by 30 January 2027 and submit it to the data governance officer. No publication or final sign-off was recorded.” Misleading rendering to reject: “the data governance officer approved and published an annotated notice draft.” Meaning-preserving controlled rendering: “the privacy programme lead was assigned to prepare an annotated notice draft by 30 January 2027 for submission to the data governance officer; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Digital privacy notice committee, the object is a retention-policy summary for service users and the operational boundary is a single service workflow. The failure matters because readers could think their data have already been erased. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the retention schedule and service map, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 8.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to issue the clearer notice was deferred pending the retention schedule and service map. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to issue the clearer notice was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to issue the clearer notice until the retention schedule and service map became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Digital privacy notice committee, the object is a retention-policy summary for service users and the operational boundary is a single service workflow. The failure matters because readers could think their data have already been erased. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the retention schedule and service map, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 8.8 — Reported completion pending verification

Fictional source record: “the privacy programme lead marked an annotated notice draft complete and submitted the draft. Verification by the data governance officer against the retention schedule and service map remains outstanding.” Misleading rendering to reject: “the data governance officer verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the privacy programme lead reported completion and submission of an annotated notice draft; verification by the data governance officer remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Digital privacy notice committee, the object is a retention-policy summary for service users and the operational boundary is a single service workflow. The failure matters because readers could think their data have already been erased. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the retention schedule and service map, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 9: Cultural festival programme team

Approval of a subtitle sample does not certify every event in a programme. The working file concerns accessible subtitle files for recorded talks. The immediate question is whether the group may release the revised subtitle package within two recorded sessions. The key unresolved safeguard is review by speakers and accessibility reviewers. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 9.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review accessible subtitle files for recorded talks and consider whether to release the revised subtitle package for two recorded sessions. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved release the revised subtitle package.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to release the revised subtitle package for two recorded sessions; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Cultural festival programme team, the object is accessible subtitle files for recorded talks and the operational boundary is two recorded sessions. The failure matters because audiences could receive an incorrectly attributed statement. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the speaker-approved transcript, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 9.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed accessible subtitle files for recorded talks and raised audiences could receive an incorrectly attributed statement. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that release the revised subtitle package.” Meaning-preserving controlled rendering: “Members discussed accessible subtitle files for recorded talks, raised audiences could receive an incorrectly attributed statement, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Cultural festival programme team, the object is accessible subtitle files for recorded talks and the operational boundary is two recorded sessions. The failure matters because audiences could receive an incorrectly attributed statement. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the speaker-approved transcript, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 9.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to release the revised subtitle package for two recorded sessions; the chair referred it for assessment against review by speakers and accessibility reviewers.” Misleading rendering to reject: “The proposal to release the revised subtitle package was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to release the revised subtitle package for two recorded sessions was referred for assessment against review by speakers and accessibility reviewers; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Cultural festival programme team, the object is accessible subtitle files for recorded talks and the operational boundary is two recorded sessions. The failure matters because audiences could receive an incorrectly attributed statement. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the speaker-approved transcript, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 9.4 — Formal motion with vote postponed

Fictional source record: “A motion to release the revised subtitle package for two recorded sessions was seconded. The chair deferred the vote until the speaker-approved transcript could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to release the revised subtitle package was seconded but the vote was deferred pending review of the speaker-approved transcript.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Cultural festival programme team, the object is accessible subtitle files for recorded talks and the operational boundary is two recorded sessions. The failure matters because audiences could receive an incorrectly attributed statement. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the speaker-approved transcript, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 9.5 — Qualified decision with limited scope

Fictional source record: “The committee approved release the revised subtitle package for two recorded sessions, subject to review by speakers and accessibility reviewers. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of accessible subtitle files for recorded talks immediately.” Meaning-preserving controlled rendering: “Approval covered release the revised subtitle package for two recorded sessions only, and depended on review by speakers and accessibility reviewers; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Cultural festival programme team, the object is accessible subtitle files for recorded talks and the operational boundary is two recorded sessions. The failure matters because audiences could receive an incorrectly attributed statement. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the speaker-approved transcript, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 9.6 — Assigned preparation versus final approval

Fictional source record: “the captions editor will prepare a synchronised subtitle proof by 23 February 2027 and submit it to the programme producer. No publication or final sign-off was recorded.” Misleading rendering to reject: “the programme producer approved and published a synchronised subtitle proof.” Meaning-preserving controlled rendering: “the captions editor was assigned to prepare a synchronised subtitle proof by 23 February 2027 for submission to the programme producer; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Cultural festival programme team, the object is accessible subtitle files for recorded talks and the operational boundary is two recorded sessions. The failure matters because audiences could receive an incorrectly attributed statement. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the speaker-approved transcript, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 9.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to release the revised subtitle package was deferred pending the speaker-approved transcript. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to release the revised subtitle package was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to release the revised subtitle package until the speaker-approved transcript became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Cultural festival programme team, the object is accessible subtitle files for recorded talks and the operational boundary is two recorded sessions. The failure matters because audiences could receive an incorrectly attributed statement. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the speaker-approved transcript, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 9.8 — Reported completion pending verification

Fictional source record: “the captions editor marked a synchronised subtitle proof complete and submitted the draft. Verification by the programme producer against the speaker-approved transcript remains outstanding.” Misleading rendering to reject: “the programme producer verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the captions editor reported completion and submission of a synchronised subtitle proof; verification by the programme producer remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Cultural festival programme team, the object is accessible subtitle files for recorded talks and the operational boundary is two recorded sessions. The failure matters because audiences could receive an incorrectly attributed statement. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the speaker-approved transcript, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 10: Library classification consortium

Terminology harmonisation cannot erase legitimate distinctions between archival categories. The working file concerns a new multilingual subject heading. The immediate question is whether the group may adopt the candidate term within a limited discovery catalogue. The key unresolved safeguard is a collision test with older records. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 10.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a new multilingual subject heading and consider whether to adopt the candidate term for a limited discovery catalogue. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved adopt the candidate term.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to adopt the candidate term for a limited discovery catalogue; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Library classification consortium, the object is a new multilingual subject heading and the operational boundary is a limited discovery catalogue. The failure matters because researchers could retrieve unrelated works. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the conflicting heading report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 10.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a new multilingual subject heading and raised researchers could retrieve unrelated works. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that adopt the candidate term.” Meaning-preserving controlled rendering: “Members discussed a new multilingual subject heading, raised researchers could retrieve unrelated works, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Library classification consortium, the object is a new multilingual subject heading and the operational boundary is a limited discovery catalogue. The failure matters because researchers could retrieve unrelated works. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the conflicting heading report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 10.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to adopt the candidate term for a limited discovery catalogue; the chair referred it for assessment against a collision test with older records.” Misleading rendering to reject: “The proposal to adopt the candidate term was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to adopt the candidate term for a limited discovery catalogue was referred for assessment against a collision test with older records; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Library classification consortium, the object is a new multilingual subject heading and the operational boundary is a limited discovery catalogue. The failure matters because researchers could retrieve unrelated works. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the conflicting heading report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 10.4 — Formal motion with vote postponed

Fictional source record: “A motion to adopt the candidate term for a limited discovery catalogue was seconded. The chair deferred the vote until the conflicting heading report could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to adopt the candidate term was seconded but the vote was deferred pending review of the conflicting heading report.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Library classification consortium, the object is a new multilingual subject heading and the operational boundary is a limited discovery catalogue. The failure matters because researchers could retrieve unrelated works. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the conflicting heading report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 10.5 — Qualified decision with limited scope

Fictional source record: “The committee approved adopt the candidate term for a limited discovery catalogue, subject to a collision test with older records. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a new multilingual subject heading immediately.” Meaning-preserving controlled rendering: “Approval covered adopt the candidate term for a limited discovery catalogue only, and depended on a collision test with older records; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Library classification consortium, the object is a new multilingual subject heading and the operational boundary is a limited discovery catalogue. The failure matters because researchers could retrieve unrelated works. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the conflicting heading report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 10.6 — Assigned preparation versus final approval

Fictional source record: “the metadata librarian will prepare a search-result audit by 17 December 2026 and submit it to the standards convenor. No publication or final sign-off was recorded.” Misleading rendering to reject: “the standards convenor approved and published a search-result audit.” Meaning-preserving controlled rendering: “the metadata librarian was assigned to prepare a search-result audit by 17 December 2026 for submission to the standards convenor; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Library classification consortium, the object is a new multilingual subject heading and the operational boundary is a limited discovery catalogue. The failure matters because researchers could retrieve unrelated works. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the conflicting heading report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 10.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to adopt the candidate term was deferred pending the conflicting heading report. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to adopt the candidate term was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to adopt the candidate term until the conflicting heading report became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Library classification consortium, the object is a new multilingual subject heading and the operational boundary is a limited discovery catalogue. The failure matters because researchers could retrieve unrelated works. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the conflicting heading report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 10.8 — Reported completion pending verification

Fictional source record: “the metadata librarian marked a search-result audit complete and submitted the draft. Verification by the standards convenor against the conflicting heading report remains outstanding.” Misleading rendering to reject: “the standards convenor verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the metadata librarian reported completion and submission of a search-result audit; verification by the standards convenor remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Library classification consortium, the object is a new multilingual subject heading and the operational boundary is a limited discovery catalogue. The failure matters because researchers could retrieve unrelated works. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the conflicting heading report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 11: Transit information group

An approved design illustration is not necessarily approval to install signage. The working file concerns a multilingual platform-wayfinding graphic. The immediate question is whether the group may install the amended sign within one interchange pilot. The key unresolved safeguard is accessibility and emergency-route review. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 11.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a multilingual platform-wayfinding graphic and consider whether to install the amended sign for one interchange pilot. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved install the amended sign.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to install the amended sign for one interchange pilot; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Transit information group, the object is a multilingual platform-wayfinding graphic and the operational boundary is one interchange pilot. The failure matters because travellers could be sent to a closed exit. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the updated station-layout drawing, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 11.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a multilingual platform-wayfinding graphic and raised travellers could be sent to a closed exit. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that install the amended sign.” Meaning-preserving controlled rendering: “Members discussed a multilingual platform-wayfinding graphic, raised travellers could be sent to a closed exit, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Transit information group, the object is a multilingual platform-wayfinding graphic and the operational boundary is one interchange pilot. The failure matters because travellers could be sent to a closed exit. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the updated station-layout drawing, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 11.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to install the amended sign for one interchange pilot; the chair referred it for assessment against accessibility and emergency-route review.” Misleading rendering to reject: “The proposal to install the amended sign was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to install the amended sign for one interchange pilot was referred for assessment against accessibility and emergency-route review; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Transit information group, the object is a multilingual platform-wayfinding graphic and the operational boundary is one interchange pilot. The failure matters because travellers could be sent to a closed exit. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the updated station-layout drawing, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 11.4 — Formal motion with vote postponed

Fictional source record: “A motion to install the amended sign for one interchange pilot was seconded. The chair deferred the vote until the updated station-layout drawing could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to install the amended sign was seconded but the vote was deferred pending review of the updated station-layout drawing.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Transit information group, the object is a multilingual platform-wayfinding graphic and the operational boundary is one interchange pilot. The failure matters because travellers could be sent to a closed exit. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the updated station-layout drawing, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 11.5 — Qualified decision with limited scope

Fictional source record: “The committee approved install the amended sign for one interchange pilot, subject to accessibility and emergency-route review. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a multilingual platform-wayfinding graphic immediately.” Meaning-preserving controlled rendering: “Approval covered install the amended sign for one interchange pilot only, and depended on accessibility and emergency-route review; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Transit information group, the object is a multilingual platform-wayfinding graphic and the operational boundary is one interchange pilot. The failure matters because travellers could be sent to a closed exit. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the updated station-layout drawing, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 11.6 — Assigned preparation versus final approval

Fictional source record: “the passenger information lead will prepare a route-map sign-off copy by 8 April 2027 and submit it to the operations controller. No publication or final sign-off was recorded.” Misleading rendering to reject: “the operations controller approved and published a route-map sign-off copy.” Meaning-preserving controlled rendering: “the passenger information lead was assigned to prepare a route-map sign-off copy by 8 April 2027 for submission to the operations controller; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Transit information group, the object is a multilingual platform-wayfinding graphic and the operational boundary is one interchange pilot. The failure matters because travellers could be sent to a closed exit. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the updated station-layout drawing, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 11.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to install the amended sign was deferred pending the updated station-layout drawing. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to install the amended sign was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to install the amended sign until the updated station-layout drawing became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Transit information group, the object is a multilingual platform-wayfinding graphic and the operational boundary is one interchange pilot. The failure matters because travellers could be sent to a closed exit. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the updated station-layout drawing, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 11.8 — Reported completion pending verification

Fictional source record: “the passenger information lead marked a route-map sign-off copy complete and submitted the draft. Verification by the operations controller against the updated station-layout drawing remains outstanding.” Misleading rendering to reject: “the operations controller verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the passenger information lead reported completion and submission of a route-map sign-off copy; verification by the operations controller remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Transit information group, the object is a multilingual platform-wayfinding graphic and the operational boundary is one interchange pilot. The failure matters because travellers could be sent to a closed exit. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the updated station-layout drawing, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 12: Manufacturing quality council

A closed action entry does not automatically establish that products are conforming. The working file concerns a supplier traceability matrix. The immediate question is whether the group may use the updated inspection checklist within the next sample batch. The key unresolved safeguard is independent calibration verification. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 12.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a supplier traceability matrix and consider whether to use the updated inspection checklist for the next sample batch. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved use the updated inspection checklist.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to use the updated inspection checklist for the next sample batch; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Manufacturing quality council, the object is a supplier traceability matrix and the operational boundary is the next sample batch. The failure matters because a missing batch reference could prevent a product recall. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the instrument calibration record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 12.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a supplier traceability matrix and raised a missing batch reference could prevent a product recall. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that use the updated inspection checklist.” Meaning-preserving controlled rendering: “Members discussed a supplier traceability matrix, raised a missing batch reference could prevent a product recall, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Manufacturing quality council, the object is a supplier traceability matrix and the operational boundary is the next sample batch. The failure matters because a missing batch reference could prevent a product recall. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the instrument calibration record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 12.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to use the updated inspection checklist for the next sample batch; the chair referred it for assessment against independent calibration verification.” Misleading rendering to reject: “The proposal to use the updated inspection checklist was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to use the updated inspection checklist for the next sample batch was referred for assessment against independent calibration verification; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Manufacturing quality council, the object is a supplier traceability matrix and the operational boundary is the next sample batch. The failure matters because a missing batch reference could prevent a product recall. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the instrument calibration record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 12.4 — Formal motion with vote postponed

Fictional source record: “A motion to use the updated inspection checklist for the next sample batch was seconded. The chair deferred the vote until the instrument calibration record could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to use the updated inspection checklist was seconded but the vote was deferred pending review of the instrument calibration record.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Manufacturing quality council, the object is a supplier traceability matrix and the operational boundary is the next sample batch. The failure matters because a missing batch reference could prevent a product recall. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the instrument calibration record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 12.5 — Qualified decision with limited scope

Fictional source record: “The committee approved use the updated inspection checklist for the next sample batch, subject to independent calibration verification. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a supplier traceability matrix immediately.” Meaning-preserving controlled rendering: “Approval covered use the updated inspection checklist for the next sample batch only, and depended on independent calibration verification; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Manufacturing quality council, the object is a supplier traceability matrix and the operational boundary is the next sample batch. The failure matters because a missing batch reference could prevent a product recall. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the instrument calibration record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 12.6 — Assigned preparation versus final approval

Fictional source record: “the quality systems engineer will prepare a measured sample inspection report by 2 December 2026 and submit it to the quality director. No publication or final sign-off was recorded.” Misleading rendering to reject: “the quality director approved and published a measured sample inspection report.” Meaning-preserving controlled rendering: “the quality systems engineer was assigned to prepare a measured sample inspection report by 2 December 2026 for submission to the quality director; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Manufacturing quality council, the object is a supplier traceability matrix and the operational boundary is the next sample batch. The failure matters because a missing batch reference could prevent a product recall. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the instrument calibration record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 12.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to use the updated inspection checklist was deferred pending the instrument calibration record. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to use the updated inspection checklist was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to use the updated inspection checklist until the instrument calibration record became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Manufacturing quality council, the object is a supplier traceability matrix and the operational boundary is the next sample batch. The failure matters because a missing batch reference could prevent a product recall. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the instrument calibration record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 12.8 — Reported completion pending verification

Fictional source record: “the quality systems engineer marked a measured sample inspection report complete and submitted the draft. Verification by the quality director against the instrument calibration record remains outstanding.” Misleading rendering to reject: “the quality director verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the quality systems engineer reported completion and submission of a measured sample inspection report; verification by the quality director remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Manufacturing quality council, the object is a supplier traceability matrix and the operational boundary is the next sample batch. The failure matters because a missing batch reference could prevent a product recall. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the instrument calibration record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 13: Climate data standards forum

A terminological change can alter the interpretation of long-term datasets. The working file concerns a dataset field-definition note. The immediate question is whether the group may adopt the new precipitation label within the next experimental dataset. The key unresolved safeguard is cross-provider unit reconciliation. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 13.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a dataset field-definition note and consider whether to adopt the new precipitation label for the next experimental dataset. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved adopt the new precipitation label.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to adopt the new precipitation label for the next experimental dataset; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Climate data standards forum, the object is a dataset field-definition note and the operational boundary is the next experimental dataset. The failure matters because analysts could combine incompatible measurements. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the observed unit mismatch table, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 13.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a dataset field-definition note and raised analysts could combine incompatible measurements. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that adopt the new precipitation label.” Meaning-preserving controlled rendering: “Members discussed a dataset field-definition note, raised analysts could combine incompatible measurements, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Climate data standards forum, the object is a dataset field-definition note and the operational boundary is the next experimental dataset. The failure matters because analysts could combine incompatible measurements. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the observed unit mismatch table, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 13.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to adopt the new precipitation label for the next experimental dataset; the chair referred it for assessment against cross-provider unit reconciliation.” Misleading rendering to reject: “The proposal to adopt the new precipitation label was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to adopt the new precipitation label for the next experimental dataset was referred for assessment against cross-provider unit reconciliation; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Climate data standards forum, the object is a dataset field-definition note and the operational boundary is the next experimental dataset. The failure matters because analysts could combine incompatible measurements. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the observed unit mismatch table, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 13.4 — Formal motion with vote postponed

Fictional source record: “A motion to adopt the new precipitation label for the next experimental dataset was seconded. The chair deferred the vote until the observed unit mismatch table could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to adopt the new precipitation label was seconded but the vote was deferred pending review of the observed unit mismatch table.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Climate data standards forum, the object is a dataset field-definition note and the operational boundary is the next experimental dataset. The failure matters because analysts could combine incompatible measurements. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the observed unit mismatch table, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 13.5 — Qualified decision with limited scope

Fictional source record: “The committee approved adopt the new precipitation label for the next experimental dataset, subject to cross-provider unit reconciliation. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a dataset field-definition note immediately.” Meaning-preserving controlled rendering: “Approval covered adopt the new precipitation label for the next experimental dataset only, and depended on cross-provider unit reconciliation; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Climate data standards forum, the object is a dataset field-definition note and the operational boundary is the next experimental dataset. The failure matters because analysts could combine incompatible measurements. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the observed unit mismatch table, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 13.6 — Assigned preparation versus final approval

Fictional source record: “the data dictionary editor will prepare a schema compatibility report by 15 February 2027 and submit it to the steering committee chair. No publication or final sign-off was recorded.” Misleading rendering to reject: “the steering committee chair approved and published a schema compatibility report.” Meaning-preserving controlled rendering: “the data dictionary editor was assigned to prepare a schema compatibility report by 15 February 2027 for submission to the steering committee chair; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Climate data standards forum, the object is a dataset field-definition note and the operational boundary is the next experimental dataset. The failure matters because analysts could combine incompatible measurements. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the observed unit mismatch table, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 13.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to adopt the new precipitation label was deferred pending the observed unit mismatch table. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to adopt the new precipitation label was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to adopt the new precipitation label until the observed unit mismatch table became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Climate data standards forum, the object is a dataset field-definition note and the operational boundary is the next experimental dataset. The failure matters because analysts could combine incompatible measurements. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the observed unit mismatch table, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 13.8 — Reported completion pending verification

Fictional source record: “the data dictionary editor marked a schema compatibility report complete and submitted the draft. Verification by the steering committee chair against the observed unit mismatch table remains outstanding.” Misleading rendering to reject: “the steering committee chair verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the data dictionary editor reported completion and submission of a schema compatibility report; verification by the steering committee chair remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Climate data standards forum, the object is a dataset field-definition note and the operational boundary is the next experimental dataset. The failure matters because analysts could combine incompatible measurements. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the observed unit mismatch table, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 14: Volunteer coordination network

A request for volunteers is not yet an accepted duty assignment. The working file concerns a revised multilingual shift handbook. The immediate question is whether the group may pilot the new roster instructions within weekend outreach volunteers. The key unresolved safeguard is confirmation of rest-break wording. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 14.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a revised multilingual shift handbook and consider whether to pilot the new roster instructions for weekend outreach volunteers. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved pilot the new roster instructions.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to pilot the new roster instructions for weekend outreach volunteers; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Volunteer coordination network, the object is a revised multilingual shift handbook and the operational boundary is weekend outreach volunteers. The failure matters because volunteers may assume an unsafe duty schedule. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the roster feedback log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 14.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a revised multilingual shift handbook and raised volunteers may assume an unsafe duty schedule. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that pilot the new roster instructions.” Meaning-preserving controlled rendering: “Members discussed a revised multilingual shift handbook, raised volunteers may assume an unsafe duty schedule, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Volunteer coordination network, the object is a revised multilingual shift handbook and the operational boundary is weekend outreach volunteers. The failure matters because volunteers may assume an unsafe duty schedule. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the roster feedback log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 14.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to pilot the new roster instructions for weekend outreach volunteers; the chair referred it for assessment against confirmation of rest-break wording.” Misleading rendering to reject: “The proposal to pilot the new roster instructions was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to pilot the new roster instructions for weekend outreach volunteers was referred for assessment against confirmation of rest-break wording; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Volunteer coordination network, the object is a revised multilingual shift handbook and the operational boundary is weekend outreach volunteers. The failure matters because volunteers may assume an unsafe duty schedule. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the roster feedback log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 14.4 — Formal motion with vote postponed

Fictional source record: “A motion to pilot the new roster instructions for weekend outreach volunteers was seconded. The chair deferred the vote until the roster feedback log could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to pilot the new roster instructions was seconded but the vote was deferred pending review of the roster feedback log.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Volunteer coordination network, the object is a revised multilingual shift handbook and the operational boundary is weekend outreach volunteers. The failure matters because volunteers may assume an unsafe duty schedule. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the roster feedback log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 14.5 — Qualified decision with limited scope

Fictional source record: “The committee approved pilot the new roster instructions for weekend outreach volunteers, subject to confirmation of rest-break wording. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a revised multilingual shift handbook immediately.” Meaning-preserving controlled rendering: “Approval covered pilot the new roster instructions for weekend outreach volunteers only, and depended on confirmation of rest-break wording; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Volunteer coordination network, the object is a revised multilingual shift handbook and the operational boundary is weekend outreach volunteers. The failure matters because volunteers may assume an unsafe duty schedule. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the roster feedback log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 14.6 — Assigned preparation versus final approval

Fictional source record: “the volunteer coordinator will prepare a revised shift-card sample by 26 November 2026 and submit it to the safeguarding officer. No publication or final sign-off was recorded.” Misleading rendering to reject: “the safeguarding officer approved and published a revised shift-card sample.” Meaning-preserving controlled rendering: “the volunteer coordinator was assigned to prepare a revised shift-card sample by 26 November 2026 for submission to the safeguarding officer; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Volunteer coordination network, the object is a revised multilingual shift handbook and the operational boundary is weekend outreach volunteers. The failure matters because volunteers may assume an unsafe duty schedule. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the roster feedback log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 14.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to pilot the new roster instructions was deferred pending the roster feedback log. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to pilot the new roster instructions was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to pilot the new roster instructions until the roster feedback log became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Volunteer coordination network, the object is a revised multilingual shift handbook and the operational boundary is weekend outreach volunteers. The failure matters because volunteers may assume an unsafe duty schedule. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the roster feedback log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 14.8 — Reported completion pending verification

Fictional source record: “the volunteer coordinator marked a revised shift-card sample complete and submitted the draft. Verification by the safeguarding officer against the roster feedback log remains outstanding.” Misleading rendering to reject: “the safeguarding officer verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the volunteer coordinator reported completion and submission of a revised shift-card sample; verification by the safeguarding officer remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Volunteer coordination network, the object is a revised multilingual shift handbook and the operational boundary is weekend outreach volunteers. The failure matters because volunteers may assume an unsafe duty schedule. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the roster feedback log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 15: Customer service policy team

A phrase that sounds helpful can create expectations beyond approved policy. The working file concerns a refund-message template. The immediate question is whether the group may publish the revised customer response within one product support queue. The key unresolved safeguard is legal review of eligibility wording. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 15.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a refund-message template and consider whether to publish the revised customer response for one product support queue. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved publish the revised customer response.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to publish the revised customer response for one product support queue; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Customer service policy team, the object is a refund-message template and the operational boundary is one product support queue. The failure matters because customers might rely on an incorrectly promised refund. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the approved eligibility matrix, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 15.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a refund-message template and raised customers might rely on an incorrectly promised refund. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that publish the revised customer response.” Meaning-preserving controlled rendering: “Members discussed a refund-message template, raised customers might rely on an incorrectly promised refund, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Customer service policy team, the object is a refund-message template and the operational boundary is one product support queue. The failure matters because customers might rely on an incorrectly promised refund. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the approved eligibility matrix, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 15.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to publish the revised customer response for one product support queue; the chair referred it for assessment against legal review of eligibility wording.” Misleading rendering to reject: “The proposal to publish the revised customer response was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to publish the revised customer response for one product support queue was referred for assessment against legal review of eligibility wording; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Customer service policy team, the object is a refund-message template and the operational boundary is one product support queue. The failure matters because customers might rely on an incorrectly promised refund. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the approved eligibility matrix, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 15.4 — Formal motion with vote postponed

Fictional source record: “A motion to publish the revised customer response for one product support queue was seconded. The chair deferred the vote until the approved eligibility matrix could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to publish the revised customer response was seconded but the vote was deferred pending review of the approved eligibility matrix.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Customer service policy team, the object is a refund-message template and the operational boundary is one product support queue. The failure matters because customers might rely on an incorrectly promised refund. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the approved eligibility matrix, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 15.5 — Qualified decision with limited scope

Fictional source record: “The committee approved publish the revised customer response for one product support queue, subject to legal review of eligibility wording. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a refund-message template immediately.” Meaning-preserving controlled rendering: “Approval covered publish the revised customer response for one product support queue only, and depended on legal review of eligibility wording; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Customer service policy team, the object is a refund-message template and the operational boundary is one product support queue. The failure matters because customers might rely on an incorrectly promised refund. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the approved eligibility matrix, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 15.6 — Assigned preparation versus final approval

Fictional source record: “the service content manager will prepare a sample response set by 18 January 2027 and submit it to the policy owner. No publication or final sign-off was recorded.” Misleading rendering to reject: “the policy owner approved and published a sample response set.” Meaning-preserving controlled rendering: “the service content manager was assigned to prepare a sample response set by 18 January 2027 for submission to the policy owner; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Customer service policy team, the object is a refund-message template and the operational boundary is one product support queue. The failure matters because customers might rely on an incorrectly promised refund. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the approved eligibility matrix, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 15.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to publish the revised customer response was deferred pending the approved eligibility matrix. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to publish the revised customer response was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to publish the revised customer response until the approved eligibility matrix became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Customer service policy team, the object is a refund-message template and the operational boundary is one product support queue. The failure matters because customers might rely on an incorrectly promised refund. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the approved eligibility matrix, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 15.8 — Reported completion pending verification

Fictional source record: “the service content manager marked a sample response set complete and submitted the draft. Verification by the policy owner against the approved eligibility matrix remains outstanding.” Misleading rendering to reject: “the policy owner verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the service content manager reported completion and submission of a sample response set; verification by the policy owner remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Customer service policy team, the object is a refund-message template and the operational boundary is one product support queue. The failure matters because customers might rely on an incorrectly promised refund. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the approved eligibility matrix, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 16: Cross-institution research data team

The meeting may approve testing without approving the data transfer itself. The working file concerns a secure dataset-access checklist. The immediate question is whether the group may trial the proposed access flow within one approved collaboration. The key unresolved safeguard is completion of access-control testing. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 16.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a secure dataset-access checklist and consider whether to trial the proposed access flow for one approved collaboration. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved trial the proposed access flow.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to trial the proposed access flow for one approved collaboration; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Cross-institution research data team, the object is a secure dataset-access checklist and the operational boundary is one approved collaboration. The failure matters because external researchers may receive access too early. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the data-use agreement and audit report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 16.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a secure dataset-access checklist and raised external researchers may receive access too early. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that trial the proposed access flow.” Meaning-preserving controlled rendering: “Members discussed a secure dataset-access checklist, raised external researchers may receive access too early, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Cross-institution research data team, the object is a secure dataset-access checklist and the operational boundary is one approved collaboration. The failure matters because external researchers may receive access too early. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the data-use agreement and audit report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 16.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to trial the proposed access flow for one approved collaboration; the chair referred it for assessment against completion of access-control testing.” Misleading rendering to reject: “The proposal to trial the proposed access flow was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to trial the proposed access flow for one approved collaboration was referred for assessment against completion of access-control testing; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Cross-institution research data team, the object is a secure dataset-access checklist and the operational boundary is one approved collaboration. The failure matters because external researchers may receive access too early. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the data-use agreement and audit report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 16.4 — Formal motion with vote postponed

Fictional source record: “A motion to trial the proposed access flow for one approved collaboration was seconded. The chair deferred the vote until the data-use agreement and audit report could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to trial the proposed access flow was seconded but the vote was deferred pending review of the data-use agreement and audit report.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Cross-institution research data team, the object is a secure dataset-access checklist and the operational boundary is one approved collaboration. The failure matters because external researchers may receive access too early. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the data-use agreement and audit report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 16.5 — Qualified decision with limited scope

Fictional source record: “The committee approved trial the proposed access flow for one approved collaboration, subject to completion of access-control testing. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a secure dataset-access checklist immediately.” Meaning-preserving controlled rendering: “Approval covered trial the proposed access flow for one approved collaboration only, and depended on completion of access-control testing; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Cross-institution research data team, the object is a secure dataset-access checklist and the operational boundary is one approved collaboration. The failure matters because external researchers may receive access too early. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the data-use agreement and audit report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 16.6 — Assigned preparation versus final approval

Fictional source record: “the project data manager will prepare a permissions verification record by 11 March 2027 and submit it to the data-sharing committee. No publication or final sign-off was recorded.” Misleading rendering to reject: “the data-sharing committee approved and published a permissions verification record.” Meaning-preserving controlled rendering: “the project data manager was assigned to prepare a permissions verification record by 11 March 2027 for submission to the data-sharing committee; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Cross-institution research data team, the object is a secure dataset-access checklist and the operational boundary is one approved collaboration. The failure matters because external researchers may receive access too early. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the data-use agreement and audit report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 16.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to trial the proposed access flow was deferred pending the data-use agreement and audit report. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to trial the proposed access flow was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to trial the proposed access flow until the data-use agreement and audit report became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Cross-institution research data team, the object is a secure dataset-access checklist and the operational boundary is one approved collaboration. The failure matters because external researchers may receive access too early. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the data-use agreement and audit report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 16.8 — Reported completion pending verification

Fictional source record: “the project data manager marked a permissions verification record complete and submitted the draft. Verification by the data-sharing committee against the data-use agreement and audit report remains outstanding.” Misleading rendering to reject: “the data-sharing committee verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the project data manager reported completion and submission of a permissions verification record; verification by the data-sharing committee remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Cross-institution research data team, the object is a secure dataset-access checklist and the operational boundary is one approved collaboration. The failure matters because external researchers may receive access too early. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the data-use agreement and audit report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 17: Procurement evaluation panel

Shortlisting, recommending and awarding are separate decision stages. The working file concerns an evaluation scoring summary. The immediate question is whether the group may shortlist the revised specification within a preliminary market review. The key unresolved safeguard is confirmation of conflict-of-interest declarations. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 17.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review an evaluation scoring summary and consider whether to shortlist the revised specification for a preliminary market review. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved shortlist the revised specification.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to shortlist the revised specification for a preliminary market review; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Procurement evaluation panel, the object is an evaluation scoring summary and the operational boundary is a preliminary market review. The failure matters because bidders might believe a contract was awarded. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the panel’s scored comparison, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 17.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed an evaluation scoring summary and raised bidders might believe a contract was awarded. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that shortlist the revised specification.” Meaning-preserving controlled rendering: “Members discussed an evaluation scoring summary, raised bidders might believe a contract was awarded, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Procurement evaluation panel, the object is an evaluation scoring summary and the operational boundary is a preliminary market review. The failure matters because bidders might believe a contract was awarded. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the panel’s scored comparison, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 17.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to shortlist the revised specification for a preliminary market review; the chair referred it for assessment against confirmation of conflict-of-interest declarations.” Misleading rendering to reject: “The proposal to shortlist the revised specification was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to shortlist the revised specification for a preliminary market review was referred for assessment against confirmation of conflict-of-interest declarations; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Procurement evaluation panel, the object is an evaluation scoring summary and the operational boundary is a preliminary market review. The failure matters because bidders might believe a contract was awarded. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the panel’s scored comparison, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 17.4 — Formal motion with vote postponed

Fictional source record: “A motion to shortlist the revised specification for a preliminary market review was seconded. The chair deferred the vote until the panel’s scored comparison could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to shortlist the revised specification was seconded but the vote was deferred pending review of the panel’s scored comparison.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Procurement evaluation panel, the object is an evaluation scoring summary and the operational boundary is a preliminary market review. The failure matters because bidders might believe a contract was awarded. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the panel’s scored comparison, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 17.5 — Qualified decision with limited scope

Fictional source record: “The committee approved shortlist the revised specification for a preliminary market review, subject to confirmation of conflict-of-interest declarations. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of an evaluation scoring summary immediately.” Meaning-preserving controlled rendering: “Approval covered shortlist the revised specification for a preliminary market review only, and depended on confirmation of conflict-of-interest declarations; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Procurement evaluation panel, the object is an evaluation scoring summary and the operational boundary is a preliminary market review. The failure matters because bidders might believe a contract was awarded. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the panel’s scored comparison, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 17.6 — Assigned preparation versus final approval

Fictional source record: “the procurement secretary will prepare a screened evaluation note by 4 February 2027 and submit it to the procurement authority. No publication or final sign-off was recorded.” Misleading rendering to reject: “the procurement authority approved and published a screened evaluation note.” Meaning-preserving controlled rendering: “the procurement secretary was assigned to prepare a screened evaluation note by 4 February 2027 for submission to the procurement authority; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Procurement evaluation panel, the object is an evaluation scoring summary and the operational boundary is a preliminary market review. The failure matters because bidders might believe a contract was awarded. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the panel’s scored comparison, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 17.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to shortlist the revised specification was deferred pending the panel’s scored comparison. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to shortlist the revised specification was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to shortlist the revised specification until the panel’s scored comparison became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Procurement evaluation panel, the object is an evaluation scoring summary and the operational boundary is a preliminary market review. The failure matters because bidders might believe a contract was awarded. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the panel’s scored comparison, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 17.8 — Reported completion pending verification

Fictional source record: “the procurement secretary marked a screened evaluation note complete and submitted the draft. Verification by the procurement authority against the panel’s scored comparison remains outstanding.” Misleading rendering to reject: “the procurement authority verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the procurement secretary reported completion and submission of a screened evaluation note; verification by the procurement authority remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Procurement evaluation panel, the object is an evaluation scoring summary and the operational boundary is a preliminary market review. The failure matters because bidders might believe a contract was awarded. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the panel’s scored comparison, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 18: Public-interest communications review

Safety messages need confirmed operational instructions, not extrapolated minutes. The working file concerns a multilingual evacuation diagram. The immediate question is whether the group may distribute the new instruction sheet within a community awareness pilot. The key unresolved safeguard is validation against current routes and signage. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 18.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a multilingual evacuation diagram and consider whether to distribute the new instruction sheet for a community awareness pilot. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved distribute the new instruction sheet.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to distribute the new instruction sheet for a community awareness pilot; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Public-interest communications review, the object is a multilingual evacuation diagram and the operational boundary is a community awareness pilot. The failure matters because people might follow an unverified evacuation route. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the latest safety map and validation record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 18.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a multilingual evacuation diagram and raised people might follow an unverified evacuation route. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that distribute the new instruction sheet.” Meaning-preserving controlled rendering: “Members discussed a multilingual evacuation diagram, raised people might follow an unverified evacuation route, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Public-interest communications review, the object is a multilingual evacuation diagram and the operational boundary is a community awareness pilot. The failure matters because people might follow an unverified evacuation route. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the latest safety map and validation record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 18.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to distribute the new instruction sheet for a community awareness pilot; the chair referred it for assessment against validation against current routes and signage.” Misleading rendering to reject: “The proposal to distribute the new instruction sheet was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to distribute the new instruction sheet for a community awareness pilot was referred for assessment against validation against current routes and signage; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Public-interest communications review, the object is a multilingual evacuation diagram and the operational boundary is a community awareness pilot. The failure matters because people might follow an unverified evacuation route. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the latest safety map and validation record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 18.4 — Formal motion with vote postponed

Fictional source record: “A motion to distribute the new instruction sheet for a community awareness pilot was seconded. The chair deferred the vote until the latest safety map and validation record could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to distribute the new instruction sheet was seconded but the vote was deferred pending review of the latest safety map and validation record.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Public-interest communications review, the object is a multilingual evacuation diagram and the operational boundary is a community awareness pilot. The failure matters because people might follow an unverified evacuation route. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the latest safety map and validation record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 18.5 — Qualified decision with limited scope

Fictional source record: “The committee approved distribute the new instruction sheet for a community awareness pilot, subject to validation against current routes and signage. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a multilingual evacuation diagram immediately.” Meaning-preserving controlled rendering: “Approval covered distribute the new instruction sheet for a community awareness pilot only, and depended on validation against current routes and signage; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Public-interest communications review, the object is a multilingual evacuation diagram and the operational boundary is a community awareness pilot. The failure matters because people might follow an unverified evacuation route. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the latest safety map and validation record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 18.6 — Assigned preparation versus final approval

Fictional source record: “the safety communications officer will prepare a route-checked leaflet proof by 6 December 2026 and submit it to the emergency planning lead. No publication or final sign-off was recorded.” Misleading rendering to reject: “the emergency planning lead approved and published a route-checked leaflet proof.” Meaning-preserving controlled rendering: “the safety communications officer was assigned to prepare a route-checked leaflet proof by 6 December 2026 for submission to the emergency planning lead; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Public-interest communications review, the object is a multilingual evacuation diagram and the operational boundary is a community awareness pilot. The failure matters because people might follow an unverified evacuation route. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the latest safety map and validation record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 18.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to distribute the new instruction sheet was deferred pending the latest safety map and validation record. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to distribute the new instruction sheet was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to distribute the new instruction sheet until the latest safety map and validation record became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Public-interest communications review, the object is a multilingual evacuation diagram and the operational boundary is a community awareness pilot. The failure matters because people might follow an unverified evacuation route. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the latest safety map and validation record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 18.8 — Reported completion pending verification

Fictional source record: “the safety communications officer marked a route-checked leaflet proof complete and submitted the draft. Verification by the emergency planning lead against the latest safety map and validation record remains outstanding.” Misleading rendering to reject: “the emergency planning lead verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the safety communications officer reported completion and submission of a route-checked leaflet proof; verification by the emergency planning lead remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Public-interest communications review, the object is a multilingual evacuation diagram and the operational boundary is a community awareness pilot. The failure matters because people might follow an unverified evacuation route. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the latest safety map and validation record, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 19: Literary translation collective

Editorial discussion about tone is not author consent. The working file concerns a bilingual reading-programme introduction. The immediate question is whether the group may publish a revised author biography within the online preview programme. The key unresolved safeguard is consent to the biographical edits. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 19.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a bilingual reading-programme introduction and consider whether to publish a revised author biography for the online preview programme. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved publish a revised author biography.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to publish a revised author biography for the online preview programme; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Literary translation collective, the object is a bilingual reading-programme introduction and the operational boundary is the online preview programme. The failure matters because an author may appear to endorse words they did not approve. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the corrected biographical source, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 19.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a bilingual reading-programme introduction and raised an author may appear to endorse words they did not approve. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that publish a revised author biography.” Meaning-preserving controlled rendering: “Members discussed a bilingual reading-programme introduction, raised an author may appear to endorse words they did not approve, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Literary translation collective, the object is a bilingual reading-programme introduction and the operational boundary is the online preview programme. The failure matters because an author may appear to endorse words they did not approve. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the corrected biographical source, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 19.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to publish a revised author biography for the online preview programme; the chair referred it for assessment against consent to the biographical edits.” Misleading rendering to reject: “The proposal to publish a revised author biography was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to publish a revised author biography for the online preview programme was referred for assessment against consent to the biographical edits; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Literary translation collective, the object is a bilingual reading-programme introduction and the operational boundary is the online preview programme. The failure matters because an author may appear to endorse words they did not approve. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the corrected biographical source, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 19.4 — Formal motion with vote postponed

Fictional source record: “A motion to publish a revised author biography for the online preview programme was seconded. The chair deferred the vote until the corrected biographical source could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to publish a revised author biography was seconded but the vote was deferred pending review of the corrected biographical source.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Literary translation collective, the object is a bilingual reading-programme introduction and the operational boundary is the online preview programme. The failure matters because an author may appear to endorse words they did not approve. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the corrected biographical source, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 19.5 — Qualified decision with limited scope

Fictional source record: “The committee approved publish a revised author biography for the online preview programme, subject to consent to the biographical edits. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a bilingual reading-programme introduction immediately.” Meaning-preserving controlled rendering: “Approval covered publish a revised author biography for the online preview programme only, and depended on consent to the biographical edits; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Literary translation collective, the object is a bilingual reading-programme introduction and the operational boundary is the online preview programme. The failure matters because an author may appear to endorse words they did not approve. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the corrected biographical source, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 19.6 — Assigned preparation versus final approval

Fictional source record: “the literary editor will prepare an author-approved biography proof by 21 January 2027 and submit it to the programme coordinator. No publication or final sign-off was recorded.” Misleading rendering to reject: “the programme coordinator approved and published an author-approved biography proof.” Meaning-preserving controlled rendering: “the literary editor was assigned to prepare an author-approved biography proof by 21 January 2027 for submission to the programme coordinator; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Literary translation collective, the object is a bilingual reading-programme introduction and the operational boundary is the online preview programme. The failure matters because an author may appear to endorse words they did not approve. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the corrected biographical source, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 19.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to publish a revised author biography was deferred pending the corrected biographical source. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to publish a revised author biography was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to publish a revised author biography until the corrected biographical source became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Literary translation collective, the object is a bilingual reading-programme introduction and the operational boundary is the online preview programme. The failure matters because an author may appear to endorse words they did not approve. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the corrected biographical source, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 19.8 — Reported completion pending verification

Fictional source record: “the literary editor marked an author-approved biography proof complete and submitted the draft. Verification by the programme coordinator against the corrected biographical source remains outstanding.” Misleading rendering to reject: “the programme coordinator verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the literary editor reported completion and submission of an author-approved biography proof; verification by the programme coordinator remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Literary translation collective, the object is a bilingual reading-programme introduction and the operational boundary is the online preview programme. The failure matters because an author may appear to endorse words they did not approve. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the corrected biographical source, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 20: Remote technical documentation team

Updating documentation differs from deploying the underlying interface. The working file concerns an application-programming interface migration note. The immediate question is whether the group may merge the revised documentation within one developer preview environment. The key unresolved safeguard is verification of deprecated endpoint examples. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 20.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review an application-programming interface migration note and consider whether to merge the revised documentation for one developer preview environment. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved merge the revised documentation.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to merge the revised documentation for one developer preview environment; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Remote technical documentation team, the object is an application-programming interface migration note and the operational boundary is one developer preview environment. The failure matters because developers could act on an unsupported migration path. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the version-specific API test log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 20.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed an application-programming interface migration note and raised developers could act on an unsupported migration path. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that merge the revised documentation.” Meaning-preserving controlled rendering: “Members discussed an application-programming interface migration note, raised developers could act on an unsupported migration path, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Remote technical documentation team, the object is an application-programming interface migration note and the operational boundary is one developer preview environment. The failure matters because developers could act on an unsupported migration path. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the version-specific API test log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 20.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to merge the revised documentation for one developer preview environment; the chair referred it for assessment against verification of deprecated endpoint examples.” Misleading rendering to reject: “The proposal to merge the revised documentation was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to merge the revised documentation for one developer preview environment was referred for assessment against verification of deprecated endpoint examples; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Remote technical documentation team, the object is an application-programming interface migration note and the operational boundary is one developer preview environment. The failure matters because developers could act on an unsupported migration path. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the version-specific API test log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 20.4 — Formal motion with vote postponed

Fictional source record: “A motion to merge the revised documentation for one developer preview environment was seconded. The chair deferred the vote until the version-specific API test log could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to merge the revised documentation was seconded but the vote was deferred pending review of the version-specific API test log.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Remote technical documentation team, the object is an application-programming interface migration note and the operational boundary is one developer preview environment. The failure matters because developers could act on an unsupported migration path. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the version-specific API test log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 20.5 — Qualified decision with limited scope

Fictional source record: “The committee approved merge the revised documentation for one developer preview environment, subject to verification of deprecated endpoint examples. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of an application-programming interface migration note immediately.” Meaning-preserving controlled rendering: “Approval covered merge the revised documentation for one developer preview environment only, and depended on verification of deprecated endpoint examples; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Remote technical documentation team, the object is an application-programming interface migration note and the operational boundary is one developer preview environment. The failure matters because developers could act on an unsupported migration path. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the version-specific API test log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 20.6 — Assigned preparation versus final approval

Fictional source record: “the technical writer will prepare a tested code-example appendix by 3 March 2027 and submit it to the documentation maintainer. No publication or final sign-off was recorded.” Misleading rendering to reject: “the documentation maintainer approved and published a tested code-example appendix.” Meaning-preserving controlled rendering: “the technical writer was assigned to prepare a tested code-example appendix by 3 March 2027 for submission to the documentation maintainer; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Remote technical documentation team, the object is an application-programming interface migration note and the operational boundary is one developer preview environment. The failure matters because developers could act on an unsupported migration path. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the version-specific API test log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 20.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to merge the revised documentation was deferred pending the version-specific API test log. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to merge the revised documentation was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to merge the revised documentation until the version-specific API test log became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Remote technical documentation team, the object is an application-programming interface migration note and the operational boundary is one developer preview environment. The failure matters because developers could act on an unsupported migration path. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the version-specific API test log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 20.8 — Reported completion pending verification

Fictional source record: “the technical writer marked a tested code-example appendix complete and submitted the draft. Verification by the documentation maintainer against the version-specific API test log remains outstanding.” Misleading rendering to reject: “the documentation maintainer verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the technical writer reported completion and submission of a tested code-example appendix; verification by the documentation maintainer remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Remote technical documentation team, the object is an application-programming interface migration note and the operational boundary is one developer preview environment. The failure matters because developers could act on an unsupported migration path. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the version-specific API test log, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 21: International archives digitisation panel

An edited transcript should expose uncertainties rather than invent certainty. The working file concerns a transcription accuracy protocol. The immediate question is whether the group may introduce a double-check sampling scheme within a small batch of damaged manuscripts. The key unresolved safeguard is review of uncertain readings. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 21.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a transcription accuracy protocol and consider whether to introduce a double-check sampling scheme for a small batch of damaged manuscripts. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved introduce a double-check sampling scheme.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to introduce a double-check sampling scheme for a small batch of damaged manuscripts; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For International archives digitisation panel, the object is a transcription accuracy protocol and the operational boundary is a small batch of damaged manuscripts. The failure matters because uncertain historical text could be presented as verified fact. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the source-image uncertainty register, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 21.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a transcription accuracy protocol and raised uncertain historical text could be presented as verified fact. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that introduce a double-check sampling scheme.” Meaning-preserving controlled rendering: “Members discussed a transcription accuracy protocol, raised uncertain historical text could be presented as verified fact, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For International archives digitisation panel, the object is a transcription accuracy protocol and the operational boundary is a small batch of damaged manuscripts. The failure matters because uncertain historical text could be presented as verified fact. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the source-image uncertainty register, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 21.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to introduce a double-check sampling scheme for a small batch of damaged manuscripts; the chair referred it for assessment against review of uncertain readings.” Misleading rendering to reject: “The proposal to introduce a double-check sampling scheme was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to introduce a double-check sampling scheme for a small batch of damaged manuscripts was referred for assessment against review of uncertain readings; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For International archives digitisation panel, the object is a transcription accuracy protocol and the operational boundary is a small batch of damaged manuscripts. The failure matters because uncertain historical text could be presented as verified fact. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the source-image uncertainty register, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 21.4 — Formal motion with vote postponed

Fictional source record: “A motion to introduce a double-check sampling scheme for a small batch of damaged manuscripts was seconded. The chair deferred the vote until the source-image uncertainty register could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to introduce a double-check sampling scheme was seconded but the vote was deferred pending review of the source-image uncertainty register.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For International archives digitisation panel, the object is a transcription accuracy protocol and the operational boundary is a small batch of damaged manuscripts. The failure matters because uncertain historical text could be presented as verified fact. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the source-image uncertainty register, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 21.5 — Qualified decision with limited scope

Fictional source record: “The committee approved introduce a double-check sampling scheme for a small batch of damaged manuscripts, subject to review of uncertain readings. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a transcription accuracy protocol immediately.” Meaning-preserving controlled rendering: “Approval covered introduce a double-check sampling scheme for a small batch of damaged manuscripts only, and depended on review of uncertain readings; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For International archives digitisation panel, the object is a transcription accuracy protocol and the operational boundary is a small batch of damaged manuscripts. The failure matters because uncertain historical text could be presented as verified fact. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the source-image uncertainty register, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 21.6 — Assigned preparation versus final approval

Fictional source record: “the archive transcription lead will prepare a sampled transcription comparison by 5 January 2027 and submit it to the archival custodian. No publication or final sign-off was recorded.” Misleading rendering to reject: “the archival custodian approved and published a sampled transcription comparison.” Meaning-preserving controlled rendering: “the archive transcription lead was assigned to prepare a sampled transcription comparison by 5 January 2027 for submission to the archival custodian; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For International archives digitisation panel, the object is a transcription accuracy protocol and the operational boundary is a small batch of damaged manuscripts. The failure matters because uncertain historical text could be presented as verified fact. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the source-image uncertainty register, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 21.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to introduce a double-check sampling scheme was deferred pending the source-image uncertainty register. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to introduce a double-check sampling scheme was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to introduce a double-check sampling scheme until the source-image uncertainty register became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For International archives digitisation panel, the object is a transcription accuracy protocol and the operational boundary is a small batch of damaged manuscripts. The failure matters because uncertain historical text could be presented as verified fact. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the source-image uncertainty register, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 21.8 — Reported completion pending verification

Fictional source record: “the archive transcription lead marked a sampled transcription comparison complete and submitted the draft. Verification by the archival custodian against the source-image uncertainty register remains outstanding.” Misleading rendering to reject: “the archival custodian verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the archive transcription lead reported completion and submission of a sampled transcription comparison; verification by the archival custodian remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For International archives digitisation panel, the object is a transcription accuracy protocol and the operational boundary is a small batch of damaged manuscripts. The failure matters because uncertain historical text could be presented as verified fact. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the source-image uncertainty register, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 22: Multilingual product safety committee

A design proposal has no power to remove an existing warning. The working file concerns a warning-label layout. The immediate question is whether the group may approve an alternative pictogram within the prototype package only. The key unresolved safeguard is independent comprehension testing. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 22.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a warning-label layout and consider whether to approve an alternative pictogram for the prototype package only. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved approve an alternative pictogram.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to approve an alternative pictogram for the prototype package only; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Multilingual product safety committee, the object is a warning-label layout and the operational boundary is the prototype package only. The failure matters because consumers might miss a critical warning. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the approved hazard phrase glossary, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 22.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a warning-label layout and raised consumers might miss a critical warning. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that approve an alternative pictogram.” Meaning-preserving controlled rendering: “Members discussed a warning-label layout, raised consumers might miss a critical warning, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Multilingual product safety committee, the object is a warning-label layout and the operational boundary is the prototype package only. The failure matters because consumers might miss a critical warning. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the approved hazard phrase glossary, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 22.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to approve an alternative pictogram for the prototype package only; the chair referred it for assessment against independent comprehension testing.” Misleading rendering to reject: “The proposal to approve an alternative pictogram was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to approve an alternative pictogram for the prototype package only was referred for assessment against independent comprehension testing; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Multilingual product safety committee, the object is a warning-label layout and the operational boundary is the prototype package only. The failure matters because consumers might miss a critical warning. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the approved hazard phrase glossary, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 22.4 — Formal motion with vote postponed

Fictional source record: “A motion to approve an alternative pictogram for the prototype package only was seconded. The chair deferred the vote until the approved hazard phrase glossary could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to approve an alternative pictogram was seconded but the vote was deferred pending review of the approved hazard phrase glossary.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Multilingual product safety committee, the object is a warning-label layout and the operational boundary is the prototype package only. The failure matters because consumers might miss a critical warning. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the approved hazard phrase glossary, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 22.5 — Qualified decision with limited scope

Fictional source record: “The committee approved approve an alternative pictogram for the prototype package only, subject to independent comprehension testing. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a warning-label layout immediately.” Meaning-preserving controlled rendering: “Approval covered approve an alternative pictogram for the prototype package only only, and depended on independent comprehension testing; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Multilingual product safety committee, the object is a warning-label layout and the operational boundary is the prototype package only. The failure matters because consumers might miss a critical warning. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the approved hazard phrase glossary, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 22.6 — Assigned preparation versus final approval

Fictional source record: “the product safety writer will prepare an interpreted pictogram test by 27 February 2027 and submit it to the compliance reviewer. No publication or final sign-off was recorded.” Misleading rendering to reject: “the compliance reviewer approved and published an interpreted pictogram test.” Meaning-preserving controlled rendering: “the product safety writer was assigned to prepare an interpreted pictogram test by 27 February 2027 for submission to the compliance reviewer; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Multilingual product safety committee, the object is a warning-label layout and the operational boundary is the prototype package only. The failure matters because consumers might miss a critical warning. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the approved hazard phrase glossary, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 22.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to approve an alternative pictogram was deferred pending the approved hazard phrase glossary. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to approve an alternative pictogram was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to approve an alternative pictogram until the approved hazard phrase glossary became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Multilingual product safety committee, the object is a warning-label layout and the operational boundary is the prototype package only. The failure matters because consumers might miss a critical warning. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the approved hazard phrase glossary, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 22.8 — Reported completion pending verification

Fictional source record: “the product safety writer marked an interpreted pictogram test complete and submitted the draft. Verification by the compliance reviewer against the approved hazard phrase glossary remains outstanding.” Misleading rendering to reject: “the compliance reviewer verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the product safety writer reported completion and submission of an interpreted pictogram test; verification by the compliance reviewer remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Multilingual product safety committee, the object is a warning-label layout and the operational boundary is the prototype package only. The failure matters because consumers might miss a critical warning. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the approved hazard phrase glossary, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 23: Disaster-preparedness data consortium

A legend label can change the perceived operational readiness of supplies. The working file concerns a resource-availability dashboard legend. The immediate question is whether the group may release the revised legend within a restricted rehearsal dashboard. The key unresolved safeguard is alignment with field inventory updates. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 23.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a resource-availability dashboard legend and consider whether to release the revised legend for a restricted rehearsal dashboard. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved release the revised legend.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to release the revised legend for a restricted rehearsal dashboard; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Disaster-preparedness data consortium, the object is a resource-availability dashboard legend and the operational boundary is a restricted rehearsal dashboard. The failure matters because responders could infer equipment is available when it is merely requested. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the stock-status reconciliation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 23.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a resource-availability dashboard legend and raised responders could infer equipment is available when it is merely requested. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that release the revised legend.” Meaning-preserving controlled rendering: “Members discussed a resource-availability dashboard legend, raised responders could infer equipment is available when it is merely requested, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Disaster-preparedness data consortium, the object is a resource-availability dashboard legend and the operational boundary is a restricted rehearsal dashboard. The failure matters because responders could infer equipment is available when it is merely requested. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the stock-status reconciliation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 23.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to release the revised legend for a restricted rehearsal dashboard; the chair referred it for assessment against alignment with field inventory updates.” Misleading rendering to reject: “The proposal to release the revised legend was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to release the revised legend for a restricted rehearsal dashboard was referred for assessment against alignment with field inventory updates; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Disaster-preparedness data consortium, the object is a resource-availability dashboard legend and the operational boundary is a restricted rehearsal dashboard. The failure matters because responders could infer equipment is available when it is merely requested. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the stock-status reconciliation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 23.4 — Formal motion with vote postponed

Fictional source record: “A motion to release the revised legend for a restricted rehearsal dashboard was seconded. The chair deferred the vote until the stock-status reconciliation could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to release the revised legend was seconded but the vote was deferred pending review of the stock-status reconciliation.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Disaster-preparedness data consortium, the object is a resource-availability dashboard legend and the operational boundary is a restricted rehearsal dashboard. The failure matters because responders could infer equipment is available when it is merely requested. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the stock-status reconciliation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 23.5 — Qualified decision with limited scope

Fictional source record: “The committee approved release the revised legend for a restricted rehearsal dashboard, subject to alignment with field inventory updates. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a resource-availability dashboard legend immediately.” Meaning-preserving controlled rendering: “Approval covered release the revised legend for a restricted rehearsal dashboard only, and depended on alignment with field inventory updates; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Disaster-preparedness data consortium, the object is a resource-availability dashboard legend and the operational boundary is a restricted rehearsal dashboard. The failure matters because responders could infer equipment is available when it is merely requested. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the stock-status reconciliation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 23.6 — Assigned preparation versus final approval

Fictional source record: “the incident-data coordinator will prepare an inventory mapping check by 10 December 2026 and submit it to the response planning director. No publication or final sign-off was recorded.” Misleading rendering to reject: “the response planning director approved and published an inventory mapping check.” Meaning-preserving controlled rendering: “the incident-data coordinator was assigned to prepare an inventory mapping check by 10 December 2026 for submission to the response planning director; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Disaster-preparedness data consortium, the object is a resource-availability dashboard legend and the operational boundary is a restricted rehearsal dashboard. The failure matters because responders could infer equipment is available when it is merely requested. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the stock-status reconciliation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 23.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to release the revised legend was deferred pending the stock-status reconciliation. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to release the revised legend was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to release the revised legend until the stock-status reconciliation became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Disaster-preparedness data consortium, the object is a resource-availability dashboard legend and the operational boundary is a restricted rehearsal dashboard. The failure matters because responders could infer equipment is available when it is merely requested. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the stock-status reconciliation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 23.8 — Reported completion pending verification

Fictional source record: “the incident-data coordinator marked an inventory mapping check complete and submitted the draft. Verification by the response planning director against the stock-status reconciliation remains outstanding.” Misleading rendering to reject: “the response planning director verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the incident-data coordinator reported completion and submission of an inventory mapping check; verification by the response planning director remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Disaster-preparedness data consortium, the object is a resource-availability dashboard legend and the operational boundary is a restricted rehearsal dashboard. The failure matters because responders could infer equipment is available when it is merely requested. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the stock-status reconciliation, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Workshop 24: Multilingual community advisory circle

Feedback recorded during consultation does not equal an adopted policy. The working file concerns a plain-language public consultation summary. The immediate question is whether the group may circulate revised consultation questions within one trial listening session. The key unresolved safeguard is independent review for accessibility and fairness. For readers of translated minutes, this setting illustrates why apparently modest verb choices can activate a later operational step. In every record below, the named object, responsible person and review gate are part of the meaning—not decorative details.

Case 24.1 — Agenda entry without outcome

Fictional source record: “Agenda item: Review a plain-language public consultation summary and consider whether to circulate revised consultation questions for one trial listening session. Presented for discussion; no decision sought in this session.” Misleading rendering to reject: “The group approved circulate revised consultation questions.” Meaning-preserving controlled rendering: “The agenda scheduled discussion of whether to circulate revised consultation questions for one trial listening session; it did not record approval.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on planned consideration versus completed decision. A heading predicts a conversation; it does not prove the conversation produced authority. For Multilingual community advisory circle, the object is a plain-language public consultation summary and the operational boundary is one trial listening session. The failure matters because readers might think a consultation result was already settled. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the agenda heading and the actual decision section. The current materials also refer to the anonymised feedback report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What recorded later result, if any, changed the status from agenda item to decision? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 24.2 — Discussion and concerns without agreement

Fictional source record: “Members discussed a plain-language public consultation summary and raised readers might think a consultation result was already settled. Several participants wanted further analysis; no resolution or vote was recorded.” Misleading rendering to reject: “Members reached agreement that circulate revised consultation questions.” Meaning-preserving controlled rendering: “Members discussed a plain-language public consultation summary, raised readers might think a consultation result was already settled, and did not record a decision.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on contribution versus collective determination. Speakers may favour a proposal without constituting a decision of the body. For Multilingual community advisory circle, the object is a plain-language public consultation summary and the operational boundary is one trial listening session. The failure matters because readers might think a consultation result was already settled. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the attributed discussion notes and absence of an operative outcome. The current materials also refer to the anonymised feedback report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Could a participant’s view have been mistaken for an institutional outcome? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 24.3 — Proposed action and next-stage referral

Fictional source record: “A proposal was made to circulate revised consultation questions for one trial listening session; the chair referred it for assessment against independent review for accessibility and fairness.” Misleading rendering to reject: “The proposal to circulate revised consultation questions was adopted and implemented.” Meaning-preserving controlled rendering: “A proposal to circulate revised consultation questions for one trial listening session was referred for assessment against independent review for accessibility and fairness; adoption was not recorded.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on proposal and referral versus adoption and execution. A referral can be an affirmative decision about process without being approval of the referred substance. For Multilingual community advisory circle, the object is a plain-language public consultation summary and the operational boundary is one trial listening session. The failure matters because readers might think a consultation result was already settled. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the proposer’s wording, chair’s referral and next decision gateway. The current materials also refer to the anonymised feedback report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which body must act next, and did the current meeting exercise that authority? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 24.4 — Formal motion with vote postponed

Fictional source record: “A motion to circulate revised consultation questions for one trial listening session was seconded. The chair deferred the vote until the anonymised feedback report could be reviewed.” Misleading rendering to reject: “The motion was carried following a second.” Meaning-preserving controlled rendering: “The motion to circulate revised consultation questions was seconded but the vote was deferred pending review of the anonymised feedback report.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on seconding versus passage. Seconding ordinarily moves a motion forward in the process; it cannot substitute for the missing voting result. For Multilingual community advisory circle, the object is a plain-language public consultation summary and the operational boundary is one trial listening session. The failure matters because readers might think a consultation result was already settled. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against motion text, seconding entry and any later result. The current materials also refer to the anonymised feedback report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Is there a separate result entry showing carried, lost or withdrawn? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 24.5 — Qualified decision with limited scope

Fictional source record: “The committee approved circulate revised consultation questions for one trial listening session, subject to independent review for accessibility and fairness. No wider rollout was authorised.” Misleading rendering to reject: “The committee approved full deployment of a plain-language public consultation summary immediately.” Meaning-preserving controlled rendering: “Approval covered circulate revised consultation questions for one trial listening session only, and depended on independent review for accessibility and fairness; no wider rollout was authorised.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on conditional and scoped authority versus blanket implementation. The condition and limitation belong to the decision itself, not to optional background information. For Multilingual community advisory circle, the object is a plain-language public consultation summary and the operational boundary is one trial listening session. The failure matters because readers might think a consultation result was already settled. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the operative decision clause, scope and condition. The current materials also refer to the anonymised feedback report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What action would be premature if the stated condition remained unsatisfied? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 24.6 — Assigned preparation versus final approval

Fictional source record: “the community liaison will prepare a feedback and response matrix by 19 February 2027 and submit it to the facilitation chair. No publication or final sign-off was recorded.” Misleading rendering to reject: “the facilitation chair approved and published a feedback and response matrix.” Meaning-preserving controlled rendering: “the community liaison was assigned to prepare a feedback and response matrix by 19 February 2027 for submission to the facilitation chair; final approval remained pending.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on task assignment versus acceptance and publication. An assigned deliverable creates responsibility to prepare something, not permission to release it. For Multilingual community advisory circle, the object is a plain-language public consultation summary and the operational boundary is one trial listening session. The failure matters because readers might think a consultation result was already settled. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against action row, owner, date, intended receiver and sign-off log. The current materials also refer to the anonymised feedback report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: What evidence would distinguish preparation, submission, acceptance and publication? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 24.7 — Decision postponed for missing evidence

Fictional source record: “Decision on whether to circulate revised consultation questions was deferred pending the anonymised feedback report. The proposal remains open for a future meeting.” Misleading rendering to reject: “The proposal to circulate revised consultation questions was rejected.” Meaning-preserving controlled rendering: “The group postponed a decision on whether to circulate revised consultation questions until the anonymised feedback report became available; it did not reject the proposal.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on postponement versus substantive rejection. Delay can preserve an option for later consideration rather than determining its merits. For Multilingual community advisory circle, the object is a plain-language public consultation summary and the operational boundary is one trial listening session. The failure matters because readers might think a consultation result was already settled. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against the deferral note and specified evidence gate. The current materials also refer to the anonymised feedback report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Has the required evidence arrived, and which later meeting may decide? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

Case 24.8 — Reported completion pending verification

Fictional source record: “the community liaison marked a feedback and response matrix complete and submitted the draft. Verification by the facilitation chair against the anonymised feedback report remains outstanding.” Misleading rendering to reject: “the facilitation chair verified and formally closed the implementation.” Meaning-preserving controlled rendering: “the community liaison reported completion and submission of a feedback and response matrix; verification by the facilitation chair remained outstanding.” These three lines show a typical translation diagnosis: the source creates one degree of authority, the bad version grants another, and the repair states the narrow outcome in language a reader can act upon safely.

Mechanism: This case turns on reported completion versus verified acceptance. Completion reported by an action owner does not prove that acceptance, verification or closure has occurred. For Multilingual community advisory circle, the object is a plain-language public consultation summary and the operational boundary is one trial listening session. The failure matters because readers might think a consultation result was already settled. A translator who smooths the record into an uncomplicated success story would remove the distinction between consideration, organisational authority and evidence. Instead, retain the actual procedural verb, its object, any negative statement and the person or body whose further act is required.

Document-level check: Compare the source against task register, submission artefact and independent acceptance evidence. The current materials also refer to the anonymised feedback report, which must be treated as an evidential item rather than assumed to be a completed approval. Check the original item number, document version, any table heading and all qualifications. If the source says nothing about a vote, date or approval, do not fabricate one for target-language fluency. If the same short term can mean several things in the target language, prefer an explicit clause or record a query for the responsible custodian.

Reader diagnostic: Ask: Which named acceptance test or signature would justify changing the status to verified? Then imagine a remote reader seeing only the translated decision extract. Could that person reasonably release a document, spend a budget, announce a result, remove an existing safeguard or close an action before the original record allowed it? If yes, the translation is not yet safe. The repair must survive both a linguistic review and an action-consequence review. The answer is not to make every sentence longer, but to keep the one critical distinction visible where the reader expects the decision to be.

Transfer exercise: Rewrite the corrected sentence as (a) a concise bilingual decision-register entry, (b) a plain-language meeting summary, and (c) an action-log note where relevant. For each version, preserve the same state and authority boundary. Tag the issue as status, actor, scope, condition, time or evidence; mark any detail not established by the source as unknown rather than inventing it. A reviewer should be able to reconstruct exactly what happened from any of the three versions without needing to guess the meeting’s intentions.

10. The decision register and action register should not be combined blindly

A decision register answers “What did the authorised body resolve, on what topic, under what scope and conditions?” An action register answers “What work must be performed, by whom, by when and how will completion be assessed?” Their records may point to each other, but the relationship is not one-to-one. One decision can create several tasks; one task can gather evidence for a decision not yet made; and a completed action can close without proving the policy question was settled.

Consider a pilot project. The group may decide to commission feasibility research, assign a writer to draft an options paper, and schedule a later decision. The decision to commission research is real. The proposal being researched remains undecided. In translation, an action table with “prepare revised specification” must not be carried into an executive summary as “new specification approved”. Preserve cross-references such as Decision D-17 authorises Task A-31, but Task A-31 does not itself establish Decision D-18.

For high-stakes work, keep separate fields for decision ID, operative words, authority, decision date, scope, conditions, source reference and amendment trail; and separate fields for action ID, deliverable, owner, due or review date, dependency, progress state, submission evidence, acceptance authority and closure reason. These are practical editorial fields, not a claim that every organisation must use the same software. The objective is to make it difficult for translation to move meaning invisibly between incompatible columns.

When a source uses a combined decision/action paragraph, parse it into its logical parts before writing the target. For example, “It was agreed to review the safety test and the coordinator will circulate recommendations” contains an agreement about review and a task about circulation. It may not contain an agreement to adopt recommendations. A translation that leaves the two parts distinguishable is more useful than a polished sentence implying a full programme was approved.

Do not split an institution’s authorised minute into new formal records without agreement from the custodian; parsing is a private analytic technique and the final document must follow its requested format. When summaries are allowed, clearly label them as summaries and retain traceability to the original approved record.

11. Translation memory, terminology databases and machine translation are aids, not decision makers

Translation memory can improve the consistency of recurring headings such as Agenda, Decisions, Matters Arising and Action Owner. It cannot tell whether the underlying motion passed in this meeting. A once-correct translation for “approved” may be dangerously misleading when a new sentence says “submitted for approval”. Segment matching works on textual similarity; procedural equivalence requires source and context analysis. Review the full sentence, its neighbouring fields and the item’s history before accepting a high-match suggestion.

A terminology database should record the meaning boundary, not just one word on each side. For each governance term, include an approved example, forbidden confusions and a note about the organisation’s own rules. “Seconded” should have a warning that it is not equivalent to “carried”; “closed” should specify the closure reason; “noted” should give examples of passive acknowledgement versus any procedural formula used by the source body. This type of context reduces false friends in even skilled human translation.

Machine translation and generative text systems can produce attractive meeting summaries that hallucinate a decision where the transcript only suggests one. Automatic extraction may assign tasks to the person who mentioned them rather than the person explicitly assigned. Human verification is therefore especially important for motions, voting results, responsibility, conditions, dates and completion claims. If an AI-assisted draft is used, document the approved source and let an accountable reviewer compare each operative statement.

ISO 11669:2024 provides general guidance for translation-project specifications, needs analysis, risk assessment and project communication; it is not a substitute for a decision register. ISO 17100:2015 addresses core translation-service processes and resources, while ISO 5060:2024 gives guidance for evaluating translation output using error-based analysis. The Multidimensional Quality Metrics framework supplies a vocabulary for classifying accuracy, terminology and fluency issues. The procedure in this article is an educational application of those broad quality principles, not an ISO-prescribed meeting-minutes checklist.

Structured task standards can help communicate task status, but they must not be confused with meeting authority. For instance, IETF RFC 5545 defines task states such as NEEDS-ACTION, IN-PROCESS, COMPLETED and CANCELLED in its VTODO component. Those labels describe a task object in a calendaring system; they do not themselves demonstrate that a meeting adopted the underlying proposal or that a completed task passed independent acceptance. Keep the record-layer distinction visible even when data are moved between applications.

12. Thirty-two short diagnostic drills and answer keys

Drill 1: A chair says “I hear no further questions.”

Question: Does the formula mean the meeting has formally voted, or only that discussion has ended? Model answer: Do not translate it as unanimous adoption without a recorded decision mechanism. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 2: A report title says “Budget approval”.

Question: Is the title a requested action or a completed decision? Model answer: Look for the operative minute and exact amount. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 3: An attendee says “Sounds good to me”.

Question: Is it individual support, a formal motion or an authorised body decision? Model answer: Keep speaker attribution unless adoption is established. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 4: The draft minutes use “resolved” in a narrative paragraph.

Question: Does the organisational convention make this a formal resolution? Model answer: Seek the decision wording and body authority. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 5: A task row shows “Done” with no attached evidence.

Question: Does “Done” mean submitted, accepted, deployed or simply closed? Model answer: Preserve the reported state and ask for acceptance evidence. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 6: A motion is approved for a trial only.

Question: Does the translated headline mention the trial limit? Model answer: Move the limitation next to approval. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 7: The body refers a draft to another committee.

Question: Has the first body endorsed the substance? Model answer: Translate referral without implying adoption. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 8: A vote lists five in favour, one against and two abstentions.

Question: Have all three vote categories survived translation? Model answer: Audit the counts and category labels individually. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 9: A decision is made “in principle”.

Question: Is money authorised now or subject to a later gate? Model answer: State the remaining condition explicitly. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 10: The chair asks for a revised proposal at the next meeting.

Question: Is the revised proposal already agreed? Model answer: Separate a task to prepare a paper from approval of its contents. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 11: An item is moved to “any other business”.

Question: Was it decided, introduced for information or merely scheduled? Model answer: Do not infer outcome from agenda placement. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 12: The agenda says “to note the report”.

Question: Does noting amount to acceptance of findings? Model answer: Use the meeting’s actual minute and procedural rules. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 13: The secretary corrects a previous meeting date.

Question: Does correction to the minute reverse any substantive decision? Model answer: Keep the record correction and policy act separate. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 14: The action owner delegates data collection.

Question: Did responsibility transfer formally or only day-to-day execution? Model answer: Preserve assigned accountability unless the source changed it. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 15: The next meeting is postponed.

Question: Did an action due “at the next meeting” acquire a different date? Model answer: Do not invent the new deadline without instruction. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 16: A project is closed for lack of funding.

Question: Does closed mean successfully delivered? Model answer: Retain cancellation or termination rationale. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 17: “Members agreed to consider” appears in a summary.

Question: Is the agreement about consideration, not implementation? Model answer: Keep the infinitive object attached to agreed. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 18: Two tasks have the same deadline but different owners.

Question: Have they been merged in the translation? Model answer: Preserve the separate rows and identifiers. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 19: A subcommittee recommends adoption to the main board.

Question: Can the subcommittee itself approve? Model answer: Keep the recommendation and approval stages distinct. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 20: A revised document is distributed after the meeting.

Question: Does distribution prove sign-off or ratification? Model answer: Report circulation separately from formal approval. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 21: A draft action log contains a blank owner field.

Question: Does the translated version supply an owner? Model answer: Keep it unassigned and flag the omission. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 22: A proposal is withdrawn before voting.

Question: Was it defeated by the meeting? Model answer: Say withdrawn, not rejected. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 23: A test passes but an independent audit is pending.

Question: Is the action verified? Model answer: Separate the test report from final acceptance. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 24: An anonymous comment is paraphrased as the group’s position.

Question: Is there evidence of collective support? Model answer: Return the statement to its attributed or anonymous discussion role. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 25: A formal vote passes with an amendment.

Question: Which version of the motion did the body adopt? Model answer: Use the final adopted wording and flag replaced text. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 26: A safety note records “No decision, pending inspection”.

Question: Does a fluent summary turn this into permission to proceed? Model answer: Preserve no decision and the pending inspection. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 27: The source says “for information only”.

Question: Does the target imply an instruction? Model answer: Keep informational force, especially in headings. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 28: An earlier meeting approved a pilot but current notes question it.

Question: Did the later discussion automatically revoke approval? Model answer: Locate a new formal decision before stating reversal. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 29: A status report says “awaiting external confirmation”.

Question: Has completion been guaranteed? Model answer: Describe the outstanding dependency. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 30: One attendee leaves before a vote.

Question: Is this person’s attendance conflated with voting participation? Model answer: Preserve item-specific presence when provided. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 31: An action table says “ongoing”.

Question: Does the translation imply progress against milestones? Model answer: Preserve open work without inventing performance results. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

Drill 32: A board agrees that someone will seek estimates.

Question: Is spending authorised? Model answer: Keep information gathering separate from expenditure approval. For assessment, identify the exact source verb, the actor, the approved object if any and the evidence needed to support a stronger claim. The aim is to practise detecting false authority even when every individual word looks ordinary and the target grammar is polished. A good answer explains not only which wording to use, but what unauthorised reader action the wrong wording could provoke.

13. A practical release checklist for translators, secretaries and reviewers

Start by comparing the number of agenda items and the item identifiers in both versions. Then compare every decision, every action, each assigned owner, each deadline and each explicit condition. An omitted “not” or “subject to” can matter more than a page of minor stylistic differences. Run the audit using a table, not solely by reading from top to bottom.

Review each apparent approval twice: once as a linguistic sentence and once as an operational permission. Ask whether the translated version allows implementation, purchase, distribution, publication, access or closure earlier than the source permits. If so, rephrase. Do the same for rejection: a deferral, referral or withdrawal should not be stated as a failed motion or closed issue unless the record supports it.

Compare all occurrences of names, roles and offices. A role can be an action owner in one item, presenter in another and approver in a third. The translation must not transform those relationships just because a single convenient term exists in the target language. For multilingual names and honorifics, consult the related chapter on personal-name and form-of-address fidelity where necessary, without repeating its separate full treatment.

Verify numerical details in a structured pass: meeting dates, vote totals, thresholds, money amounts, counts of pilot sites, version identifiers and document references. For cross-border events, use the preceding time-zone chapter to separate instant, offset and recurring local time. This article treats timing as a property of meeting decisions and action logs, not as a replacement for the detailed time-zone guidance.

When the source is incomplete, state the uncertainty and route a precise query. Silence in a record does not always mean no event occurred; it means the available source does not establish the event. Avoid converting an evidential absence into either certainty or a speculative accusation. The translator’s responsibility is to preserve documented information and seek clarification through authorised channels.

Finally, request a user test: give the target to someone who did not attend and ask them to list every decision, assigned task and unresolved issue. Compare their reading against the structured source ledger. If they infer extra approvals, missing conditions, a fabricated owner or false completion, revise before release. That test measures what the translation makes possible in practice.

14. How the method transfers without taking over other subject owners

In education and collaborative research, the method helps readers distinguish programme proposals from adopted curricula, assessment changes from pilots and assigned reading tasks from approved policy. Its purpose is to protect institutional actions, not to become a general vocabulary lesson. If the reader needs systematic word-learning and sentence construction rather than meeting-status analysis, the separate Vocabulary Learning Hub is the correct route.

In publishing and editing, the method prevents editorial recommendations from appearing as author consent or published corrections. Keep record statuses, version identifiers and sign-off evidence intact. Where the underlying challenge is English discourse structure, reporting verbs or summarisation rather than decisions and assigned tasks, the protected How English Works | Translation and Meaning Across Languages provides complementary language analysis.

In technology teams, a proposal, approved architecture, code merged, release published and security verification are different events. A meeting can authorise testing without authorising deployment. The meeting-record method detects those differences but does not replace domain-specific testing or release governance. In every field, ask what source record supplies the authority.

In arts and community programmes, approved minutes can document committee choices while comments and proposals remain expressions of perspective. A public-facing translation should avoid presenting a selective summary as a unanimously adopted position. Respect privacy, audience, context and the organisation’s chosen level of record detail. Do not create implied consent for people whose words or names appear in the original.

In high-stakes public communications, the prudent approach is to keep unresolved status explicit, verify the operative instruction against the competent authority and label any translation as informational until it acquires the approvals required by its context. Sound translation cannot cure a defective underlying meeting process; it can prevent a defective process from being disguised by confident prose.

15. Thirty frequently asked questions on translating minutes, decisions and action logs

FAQ 1: What is the difference between meeting notes and minutes?

Notes are often informal or provisional; minutes typically serve as a structured record under an organisation’s procedure. The title alone does not settle authority. Check version and approval status. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 2: Can a meeting summary be translated instead of full minutes?

Yes for an agreed informational purpose, but a summary is selective and must be labelled. It should not claim the authority or completeness of approved minutes. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 3: Does “discussed” mean “approved”?

No. It usually records consideration, not adoption. An approval requires separate evidence such as an explicit decision or valid procedure. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 4: Does “seconded” mean a motion passed?

No. Seconding is a procedural step in systems that use it; passage requires the outcome prescribed by the relevant rules. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 5: How do I translate “motion carried”?

First confirm the phrase is used for an adopted motion under the source procedure. Choose a target expression that conveys adoption, not physical movement. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 6: Can I translate “tabled” as postponed?

Not automatically. Formal usage differs between procedural traditions; consult the minutes and governing rules to identify whether the motion was introduced or delayed. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 7: What is a decision register?

A structured record of decisions, scope, authority, conditions and supporting source references. It differs from a task list, although both may be cross-referenced. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 8: What is an action item?

An identified follow-up task, normally with a deliverable and an owner when assigned. Its creation does not necessarily mean a related proposal has been approved. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 9: Does an action owner automatically approve the finished work?

No. The owner responsible for preparation may be different from the person or body with sign-off authority. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 10: What if a source minute has no owner?

Preserve that gap, note the ambiguity and ask an authorised source custodian. Do not invent a plausible person. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 11: What if the due date says “next meeting”?

Keep the relative event reference unless an authorised source establishes its date. If the next meeting changes, confirm how the deadline should be treated. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 12: Does completed mean verified?

Not necessarily. A task can be reported finished while acceptance or review remains open. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 13: Does closed mean the issue was successfully fixed?

Not always. An issue may be closed because it was cancelled, superseded, accepted or otherwise disposed of. Check the closure reason. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 14: Can I translate draft minutes as official minutes?

Not without preserving the draft status. The target should disclose that review or approval remains pending. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 15: Does approving previous minutes repeat all earlier decisions?

Generally it approves the record’s accuracy under the organisation’s procedure; do not infer a new substantive decision without evidence. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 16: What is the difference between a recommendation and a resolution?

A recommendation suggests a course for another actor to consider; a resolution records a decision under the relevant body’s procedure. The words are not equivalent. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 17: How should dissent appear?

As the source and the applicable minute procedure record it. Do not invent or erase dissent merely to make the account sound unified. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 18: What about abstentions?

Keep abstentions distinct from votes for, votes against and absence, matching the source categories exactly. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 19: Can machine translation handle meeting minutes?

It can assist with drafting, but high-impact status, actor, condition and scope claims need verification against the authorised source. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 20: Which international standards are relevant?

ISO 11669:2024 for translation-project guidance, ISO 17100:2015 for translation-service requirements and ISO 5060:2024 for translation-output evaluation. None independently decides what a meeting approved. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 21: Can a bilingual translation itself be binding?

That depends on the governing document and authority, not the translator’s fluency. Clarify which version, if any, is formally authoritative. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 22: How do I avoid changing a technical term?

Record the target term with a clause-level example, procedural meaning, forbidden confusions and review owner. Check each term against its local context. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 23: Why preserve apparently awkward negative phrases?

Negations such as “not approved” and “no wider rollout” can be the core of the decision boundary. Natural rewriting must not delete their force. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 24: How should unresolved questions be shown?

As open questions or translation queries linked to the source item. Do not make an answer up to improve flow. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 25: Can one meeting decision generate several action items?

Yes. Keep the one decision and separate deliverables traceable; do not collapse roles or tasks. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 26: Can an action exist without approval of a broader proposal?

Yes. The action may be to investigate, test or draft the proposal. It is a real task but not final adoption. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 27: How do I preserve a pilot-only approval?

Name the approved activity and limited scope together. Do not imply general implementation. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 28: When should a translator ask for clarification?

When plausible readings change authority, owner, deadline, object, condition, vote result or release permission. Frame a precise query with evidence. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 29: What is the most useful final review question?

Ask whether a new reader could take an action not permitted by the source. If yes, the translation needs correction. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

FAQ 30: Where should I continue studying?

For the full meaning-transfer framework use the translation root. For word learning use the Vocabulary Learning Hub; for English structures use How English Works. These are complementary owners, not duplicate articles. To check your own material, compare the exact source clause with the corresponding target clause and identify the actor, act, object, qualification and evidence. If one is missing, decide whether it was absent in the original or lost during translation; only the latter can be repaired without new source information.

16. High-authority references and carefully limited terminology sources

ISO 11669:2024 — Translation projects, general guidance: ISO overview. It supplies a framework for specifications, needs analysis, risk assessment, communication and workflow planning; the article applies those principles to the narrower risk of meeting-record meaning.

ISO 17100:2015 — Translation services, requirements: ISO overview. The ISO listing identifies requirements for translation-service processes and resources and notes that a revision is in development. It is not a meeting-governance code.

ISO 5060:2024 — Evaluation of translation output: ISO overview. It discusses human evaluation using error types, penalties and review competencies. Our diagnostic categories are illustrative, not an assertion of a certified standard process.

Multidimensional Quality Metrics: MQM Council. Accuracy, terminology, fluency and issue severity offer a useful way to classify errors. A wrong adoption status is principally a meaning and accuracy failure, even if the target sentence is graceful.

RFC 5545 — iCalendar: IETF / RFC Editor. The standard distinguishes VTODO statuses such as NEEDS-ACTION, IN-PROCESS, COMPLETED and CANCELLED. Those task states are useful contrasts but do not determine a meeting body’s decisions.

Meeting practice: a minute-taking and action-log guidance resource from NHS Elect illustrates the practical separation of meeting details, agenda summaries, agreed actions and ownership. Different organisations may follow different governing procedures; check the source institution rather than assuming a universal template.

Research and source review were conducted in November 2026. Where a link describes a standard or guidance edition, use the edition date shown by its issuing body rather than treating the date of this article as the publication date of that standard. This page does not reproduce proprietary standards, promise legal validity or claim certification.

17. Next reading routes

Return to the Master Art of Translation | The Complete System for Moving Meaning Between Languages to understand the overarching method of moving meaning across languages. The preceding time-zone, UTC offset and daylight-saving fidelity chapter explains time representation when an action due date or meeting instant is ambiguous. These are adjacent questions, not a substitute for the present chapter’s analysis of discussion, decision, assignment and verification.

For systematic vocabulary study rather than documentary authority, continue to the protected Vocabulary Learning Hub. For English grammar, discourse and meaning-transfer skills, continue to How English Works | Translation and Meaning Across Languages. No existing route or article needs rewriting for this new chapter to stand on its own.

Final operating principle: Preserve the exact boundary between what was spoken, what was proposed, what the authorised body decided, what someone was assigned to do and what was subsequently verified. That boundary is the meaning of a meeting record. When it survives translation, a reader in another language can understand not only the conversation but also the limits of the authority the conversation created.