Translate | ISO 8601-2:2019 Uncertain Dates, Approximate Time, Recurring Intervals and Calendar Data
Translating uncertain dates is not the same as formatting a known appointment. If a historical archive says a letter was written around 1890, a bilingual catalogue must not display 1 January 1890 as though someone discovered the precise day. When you translate approximate dates, extended time intervals, missing components and recurring calendar entries for a database, ISO 8601-2:2019 offers an important reference point: temporal expressions can preserve uncertainty rather than disguising it. This practical guide answers how to translate those expressions into clear English and other languages while keeping machine-readable values, human-facing explanations and evidence status aligned.
A curator may have three dates for a photograph, an analyst may receive an event interval with no confirmed ending, and an app may schedule meetings every second Tuesday. Their problem is not simply which punctuation to use. Each reader needs to know whether the source states a certain instant, an estimated period, a set of candidates, an alternative choice or a repeating rule. A translator who turns those distinctions into one tidy calendar date can break searches, legal retention calculations, chronology and user trust. The worked examples below show how to detect that change, ask the correct question and publish a more faithful rendering.
At a glance: preserve uncertainty words; distinguish an open interval from a long but finite one; do not invent days for month-only evidence; keep recurrence patterns separate from recorded occurrences; document UTC and local time assumptions; tell readers which original data is certain, disputed or missing; and have a data owner approve every conversion that affects a system. This article is a specialised application of the existing translation system, not a substitute for the complete standard or a new language-learning hub.
Quick answer: the safe translation workflow
- Read the source record and determine the precise entity or event it is trying to describe.
- Separate facts, estimates, labels and internal identifiers before drafting prose.
- Translate for a specific reader while retaining what the source does not establish.
- Use a controlled terminology and data-field register, with the original evidence accessible.
- Test the target version in its real interface or document and have the data owner approve consequential changes.
Understanding the standard and the translation question
The ISO catalogue describes extensions to Gregorian calendar and 24-hour clock representations for interchange, including uncertainty, approximation, unspecified portions, intervals, grouped units, sets, choices, recurrences and arithmetic. ISO 8601-2:2019 has a published 2025 amendment; its successor is still a draft as of November 2026. This guide is conceptual translation education, not a claim to reproduce or validate the standard’s entire grammar.
This guide is written for archivists, translators, genealogists, software localizers, data stewards, librarians and anyone preparing cross-language calendars. Its narrow search intent is how to translate uncertain dates, approximate historical dates, open-ended date intervals, recurring events and partial temporal data without accidentally converting incomplete evidence into exact dates. It concentrates on the problem a reader has when translating that kind of record. It does not claim that the referenced ISO publication supplies a certified translation, an authoritative government record or a ready-made software system. The fictional examples below are original exercises designed to expose choices that ordinary word-for-word substitutions would hide.
Chapter 1. Begin with the difference between an event and a representation
A date expression is a statement about what is known, not automatically a complete record of when something occurred. A manuscript can be dated from a postmark, a diary entry, a cataloguer’s estimate or an inherited database value. The translator needs to preserve that provenance before selecting natural target-language words.
Consider a portrait described as ‘late nineteenth century’. An online form may demand a full timestamp, but the form’s limitation does not make the century estimate more exact. Store the source assertion separately from the software’s placeholder.
Ask who inferred the date and whether the label is for display, searching or computation. Those three purposes may justify separate fields and distinct translation treatments.
The practical test in this chapter is to compare the source assertion, the target wording and the record’s purpose side by side. A translator should write down which details survive unchanged and which are only presentation choices. If an editor requests a smoother sentence that erases a limitation, the response should show exactly what fact or inference would be added. Readers need a usable explanation, but usability is not permission to alter the underlying claim.
Chapter 2. Know what the current edition covers and what it does not
ISO 8601-2 extends the exchange-oriented representations of ISO 8601-1. Its public scope lists partial, approximate and uncertain dates, extended intervals, divisions of a year, sets and choices, grouped scales, recurrences and arithmetic. It remains concerned with Gregorian dates and the 24-hour clock.
A calendar belonging to another tradition cannot be silently converted using these rules without a documented calendar-conversion method. Nor should an editorial English approximation be presented as a validated ISO data string.
Check whether a legacy system implements the 2019 edition, the 2025 amendment or only common ISO 8601-1 timestamps. Version differences affect acceptance of expressions and should be tested against the actual parser.
The practical test in this chapter is to compare the source assertion, the target wording and the record’s purpose side by side. A translator should write down which details survive unchanged and which are only presentation choices. If an editor requests a smoother sentence that erases a limitation, the response should show exactly what fact or inference would be added. Readers need a usable explanation, but usability is not permission to alter the underlying claim.
Chapter 3. Separate exactness, precision and confidence
An exact value identifies a specific moment or period under a defined calendar; precision describes how finely it is recorded; confidence describes how well evidence supports it. A year-only entry can be accurate at year precision. A day-level guess can be precise in appearance and weak in evidence.
Suppose an invoice says ‘April 2028’. Recasting it as 2028-04-01 implies a selected day. A search index can sometimes use a proxy start date, but the public translation must still say ‘April 2028’ and label the proxy as a technical field.
Design the source-to-target record so the human date, approximate-status flag, confidence note and any computing proxy are not overwritten by one display formatter.
The practical test in this chapter is to compare the source assertion, the target wording and the record’s purpose side by side. A translator should write down which details survive unchanged and which are only presentation choices. If an editor requests a smoother sentence that erases a limitation, the response should show exactly what fact or inference would be added. Readers need a usable explanation, but usability is not permission to alter the underlying claim.
Chapter 4. Translate bounds without inventing endpoints
An interval can describe a period with two known limits, an unconfirmed beginning, an unconfirmed end, or endpoints whose inclusion is uncertain. Natural-language ‘before’, ‘after’, ‘through’ and ‘between’ are not interchangeable because the boundary matters.
A sale notice saying ‘valid through Friday’ may include Friday, whereas ‘expires before Friday’ may not. A record saying ‘after 2017’ does not reveal when the interval ends. Translating these as identical ranges creates materially different dates.
Ask the source owner for an inclusion convention whenever a boundary drives eligibility, deadlines, validity, archival retention or a statistical window.
The practical test in this chapter is to compare the source assertion, the target wording and the record’s purpose side by side. A translator should write down which details survive unchanged and which are only presentation choices. If an editor requests a smoother sentence that erases a limitation, the response should show exactly what fact or inference would be added. Readers need a usable explanation, but usability is not permission to alter the underlying claim.
Chapter 5. Distinguish a set of dates from a choice between dates
One document can describe multiple events as a set, while another may offer alternative candidate dates for one event. The difference determines whether all dates are true records or merely competing hypotheses. The same comma or slash cannot solve that semantic problem.
If a historical exhibition opened on two dates in two cities, those are two occurrences. If a faded inscription could read 1824 or 1829, those are competing interpretations. Flattening both into ‘1824, 1829’ hides the evidence type.
A robust translation uses explicit labels such as ‘on both dates’ and ‘either date is possible’, then preserves the source field’s original ordering and evidence.
The practical test in this chapter is to compare the source assertion, the target wording and the record’s purpose side by side. A translator should write down which details survive unchanged and which are only presentation choices. If an editor requests a smoother sentence that erases a limitation, the response should show exactly what fact or inference would be added. Readers need a usable explanation, but usability is not permission to alter the underlying claim.
Chapter 6. Retain recurrence logic when localizing events
A recurring schedule defines a rule, not just a list of occurrences. Holidays, maintenance periods and subscription dates may recur according to weekday, month, offset or exception rules; moving them between time zones can create different calendar days.
A conference recurring at 09:00 Singapore time is not the same event rule as a meeting recurring at 09:00 in every participant’s local time. Translating both as ‘every Monday at nine’ leaves the operational difference invisible.
Name the time zone, recurrence anchor and exception authority. Do not generate future dates from a rule until the target system has been tested against daylight saving transitions where applicable.
The practical test in this chapter is to compare the source assertion, the target wording and the record’s purpose side by side. A translator should write down which details survive unchanged and which are only presentation choices. If an editor requests a smoother sentence that erases a limitation, the response should show exactly what fact or inference would be added. Readers need a usable explanation, but usability is not permission to alter the underlying claim.
Chapter 7. Keep calendar words and numeric strings in different layers
Machine strings are for exchange; human-readable labels are for readers. A translator may localize month names, order of date components, approximate markers and explanatory notes, but changing identifiers inside data fields can damage validation and parsing.
For example, ’03/04/26′ can mean different dates in different regions. An interface should resolve the actual source interpretation rather than copy the ambiguous shape into another language. A qualified technical owner should authorize normalized values.
Treat user-facing strings, data values and validation rules as separate translation assets, with a sample of each in the review package.
The practical test in this chapter is to compare the source assertion, the target wording and the record’s purpose side by side. A translator should write down which details survive unchanged and which are only presentation choices. If an editor requests a smoother sentence that erases a limitation, the response should show exactly what fact or inference would be added. Readers need a usable explanation, but usability is not permission to alter the underlying claim.
Chapter 8. Handle historical time and uncertainty as editorial evidence
Archives often combine an inscription, a modern scholarly estimate and a later correction. An English display should distinguish ‘dated’, ‘attributed’, ‘probably’, ‘not before’ and ‘recorded in’ because they describe different sources of knowledge.
A museum label might say a painting ‘was created c. 1905’ while its acquisition register records 1938. Translating both as the date of creation collapses two events. The same error can spread into search snippets and catalogues.
Keep a provenance column with source type, evidence date, recording date and whether an estimate has been superseded. Public labels can be concise while internal records remain explicit.
The practical test in this chapter is to compare the source assertion, the target wording and the record’s purpose side by side. A translator should write down which details survive unchanged and which are only presentation choices. If an editor requests a smoother sentence that erases a limitation, the response should show exactly what fact or inference would be added. Readers need a usable explanation, but usability is not permission to alter the underlying claim.
Chapter 9. Build review around consequences rather than typography
A translation can be typographically elegant and semantically wrong. Prioritize tests where an uncertain date becomes exact, a future recurrence changes days, a deadline changes inclusion rules, or software silently converts a display string into a timestamp.
High-risk uses include legal notices, public-health records, financial periods, application deadlines and machine schedules. Lower-risk museum labels still deserve faithful uncertainty because later researchers may treat the wording as evidence.
Sign-off should include a language reviewer and the owner of the data model. When a field cannot represent the source truth, report the limitation instead of pretending the conversion was lossless.
The practical test in this chapter is to compare the source assertion, the target wording and the record’s purpose side by side. A translator should write down which details survive unchanged and which are only presentation choices. If an editor requests a smoother sentence that erases a limitation, the response should show exactly what fact or inference would be added. Readers need a usable explanation, but usability is not permission to alter the underlying claim.
Worked translation cases: original scenarios and reasoned repairs
The following 28 invented exercises represent distinct document and data situations. The example source phrases are not quotations from the ISO standard. They demonstrate a disciplined habit: identify the referent, make the false inference visible, repair the wording and name a test that would catch the error in actual work. The safer rendering is an English explanatory gloss; translation into another target language must preserve the same conditions and be reviewed by a proficient speaker.
Case 1. An estimated birth year
Source situation. Source record: ‘Born about 1887’, based on later census testimony. Misleading shortcut. ‘Born on 1 January 1887’ invents a day and a high-confidence birth certificate.
Better reader-facing rendering. “’Born around 1887 (estimated from census records).’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. The tempting shortcut is to make the target text look complete, but completeness of typography is not completeness of evidence.
What to verify. Verify the evidence category and whether 1887 is an indexed approximate year, not a registered date. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Keep the supporting source field available to the reviewer. If the record owner cannot confirm the missing fact, the published copy should say that it is not established rather than conceal the gap. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 2. An undated letter
Source situation. A manuscript has only ‘spring, perhaps 1912’ in pencil added by an archivist. Misleading shortcut. ‘March 1912’ assumes a calendar season and a specific month.
Better reader-facing rendering. “’Possibly spring 1912; later archival annotation.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. Here the difficulty is not a rare vocabulary item. It is the relationship between the words and the record they purport to describe.
What to verify. Check hemisphere and provenance; keep the annotation distinct from the writer’s own date. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
In an actual workflow, compare search results, user instructions and any exported record, because a correct sentence can still be attached to the wrong data value. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 3. A month-only invoice
Source situation. The source records ‘September 2030’ without a transaction day. Misleading shortcut. ‘2030-09-01’ suggests a first-of-month transaction.
Better reader-facing rendering. “’September 2030 (day not recorded).’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. This is a useful test of whether a translator can preserve a distinction while still giving a busy reader an immediately usable sentence.
What to verify. Inspect whether accounting software used a default day merely for indexing. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Ask the requester to confirm the business or research consequence of this wording. If it changes where someone goes, what they believe or what a program does, sign-off must be explicit. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 4. Before an acquisition
Source situation. An object is labeled ‘produced before 1750’. Misleading shortcut. ‘1750 production’ turns an upper bound into a specific year.
Better reader-facing rendering. “’Made before 1750; exact date unknown.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. A second reviewer who sees only the final sentence could reasonably infer more than the source says; that is why the qualification must remain visible.
What to verify. Ask whether the boundary excludes the whole calendar year or an exact instant. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
A faithful translation should remain accurate when copied into a report without its surrounding paragraph. Include enough identifying context and qualification for that realistic reuse. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 5. After a storm
Source situation. Field notes say ‘repaired after the 2016 flood’, with no completion record. Misleading shortcut. ‘Repaired in 2017’ replaces a relative bound with an unsupported date.
Better reader-facing rendering. “’Repaired sometime after the 2016 flood.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. The tempting shortcut is to make the target text look complete, but completeness of typography is not completeness of evidence.
What to verify. Locate the flood event identifier and avoid guessing the next year. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Keep the supporting source field available to the reviewer. If the record owner cannot confirm the missing fact, the published copy should say that it is not established rather than conceal the gap. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 6. Two plausible years
Source situation. A handwriting reading could be 1843 or 1848. Misleading shortcut. ‘1843–1848’ implies an interval instead of alternative candidates.
Better reader-facing rendering. “’The inscription may read 1843 or 1848.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. Here the difficulty is not a rare vocabulary item. It is the relationship between the words and the record they purport to describe.
What to verify. Confirm that only one year was intended and preserve both readings. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
In an actual workflow, compare search results, user instructions and any exported record, because a correct sentence can still be attached to the wrong data value. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 7. Two recorded visits
Source situation. A log states that inspectors attended on 6 May and 12 May. Misleading shortcut. ‘Visited between 6 and 12 May’ blurs two observations into a continuous period.
Better reader-facing rendering. “’Two visits took place, on 6 May and 12 May.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. This is a useful test of whether a translator can preserve a distinction while still giving a busy reader an immediately usable sentence.
What to verify. Retain separate events and their corresponding inspection findings. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Ask the requester to confirm the business or research consequence of this wording. If it changes where someone goes, what they believe or what a program does, sign-off must be explicit. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 8. An interval with unknown start
Source situation. A service contract says the arrangement was already active when records began in 2009. Misleading shortcut. ‘Started in 2009’ treats the first recorded observation as the launch date.
Better reader-facing rendering. “’Active by 2009; start date not established.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. A second reviewer who sees only the final sentence could reasonably infer more than the source says; that is why the qualification must remain visible.
What to verify. Compare contract and records, and mark the beginning as unknown. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
A faithful translation should remain accurate when copied into a report without its surrounding paragraph. Include enough identifying context and qualification for that realistic reuse. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 9. An interval with unknown end
Source situation. A historical trade route is recorded as used from 1670 onward, with no surviving end date. Misleading shortcut. ‘Used 1670–1700’ adds a neat but unsubstantiated closing year.
Better reader-facing rendering. “’Documented in use from 1670; end date unknown.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. The tempting shortcut is to make the target text look complete, but completeness of typography is not completeness of evidence.
What to verify. Distinguish no evidence of an end from evidence that use continued indefinitely. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Keep the supporting source field available to the reviewer. If the record owner cannot confirm the missing fact, the published copy should say that it is not established rather than conceal the gap. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 10. A report issued after midnight
Source situation. A message stamped 00:15 at a remote branch is merged into headquarters’ previous-day queue. Misleading shortcut. Translating only the calendar day conceals the offset and reverses order.
Better reader-facing rendering. “’00:15 local branch time on [date], corresponding to [verified UTC time].’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. Here the difficulty is not a rare vocabulary item. It is the relationship between the words and the record they purport to describe.
What to verify. Have the system owner calculate the offset from the actual location and date. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
In an actual workflow, compare search results, user instructions and any exported record, because a correct sentence can still be attached to the wrong data value. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 11. A recurring Monday meeting
Source situation. The source says the same meeting happens every Monday at 09:00 Singapore time. Misleading shortcut. ‘Every Monday at 09:00 local time’ moves the meeting for foreign readers.
Better reader-facing rendering. “’Every Monday, 09:00 Singapore time (UTC+08:00).’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. This is a useful test of whether a translator can preserve a distinction while still giving a busy reader an immediately usable sentence.
What to verify. Verify the service’s recurrence time zone and how translated invitations are generated. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Ask the requester to confirm the business or research consequence of this wording. If it changes where someone goes, what they believe or what a program does, sign-off must be explicit. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 12. A recurring first weekday
Source situation. The maintenance notice says work runs on the first business day of each month. Misleading shortcut. ‘On the first day of the month’ forgets weekends and holidays.
Better reader-facing rendering. “’On the first business day of each month, using the stated business calendar.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. A second reviewer who sees only the final sentence could reasonably infer more than the source says; that is why the qualification must remain visible.
What to verify. Confirm which jurisdiction’s holiday calendar determines a business day. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
A faithful translation should remain accurate when copied into a report without its surrounding paragraph. Include enough identifying context and qualification for that realistic reuse. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 13. A fiscal quarter
Source situation. The board minutes refer to Q1 of a fiscal year ending in June. Misleading shortcut. ‘January–March quarter’ assumes a calendar-year system.
Better reader-facing rendering. “’First quarter of the financial year ending in June.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. The tempting shortcut is to make the target text look complete, but completeness of typography is not completeness of evidence.
What to verify. Confirm the organization’s fiscal-year boundary and whether the first quarter includes July. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Keep the supporting source field available to the reviewer. If the record owner cannot confirm the missing fact, the published copy should say that it is not established rather than conceal the gap. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 14. A school term
Source situation. The archive says ‘Term 2, 2028’ without identifying the country. Misleading shortcut. Translating ‘Term 2’ as ‘April–June’ assumes the wrong school calendar.
Better reader-facing rendering. “’Second school term of 2028; dates depend on the jurisdiction.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. Here the difficulty is not a rare vocabulary item. It is the relationship between the words and the record they purport to describe.
What to verify. Ask the institution for the official term schedule before making a range. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
In an actual workflow, compare search results, user instructions and any exported record, because a correct sentence can still be attached to the wrong data value. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 15. A season in two hemispheres
Source situation. A seed trial says ‘autumn 2019’ but its location is in South America. Misleading shortcut. Automatically mapping autumn to September–November changes the season.
Better reader-facing rendering. “’Autumn 2019 in the trial location.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. This is a useful test of whether a translator can preserve a distinction while still giving a busy reader an immediately usable sentence.
What to verify. Check region and whether meteorological or astronomical seasons are used. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Ask the requester to confirm the business or research consequence of this wording. If it changes where someone goes, what they believe or what a program does, sign-off must be explicit. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 16. An inclusive offer
Source situation. The message says applications are accepted ‘up to and including 31 July’. Misleading shortcut. ‘Before 31 July’ excludes the final eligible day.
Better reader-facing rendering. “’Applications accepted through 31 July, including that date.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. A second reviewer who sees only the final sentence could reasonably infer more than the source says; that is why the qualification must remain visible.
What to verify. Confirm local cut-off time rather than assuming midnight means the end of the day. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
A faithful translation should remain accurate when copied into a report without its surrounding paragraph. Include enough identifying context and qualification for that realistic reuse. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 17. A deadline with time zone
Source situation. The instruction says ‘submit by 17:00 UTC’. Misleading shortcut. Converting to ‘5 p.m.’ without a zone makes the deadline location-dependent.
Better reader-facing rendering. “’Deadline: 17:00 UTC; local equivalent must be calculated for your location.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. The tempting shortcut is to make the target text look complete, but completeness of typography is not completeness of evidence.
What to verify. Validate any displayed local conversion and date rollover. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Keep the supporting source field available to the reviewer. If the record owner cannot confirm the missing fact, the published copy should say that it is not established rather than conceal the gap. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 18. An approximate decade
Source situation. A collection note says ‘early 1980s’. Misleading shortcut. Translating as ‘1980’ suppresses the multi-year uncertainty.
Better reader-facing rendering. “’Early 1980s, with no verified individual year.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. Here the difficulty is not a rare vocabulary item. It is the relationship between the words and the record they purport to describe.
What to verify. Ask whether cataloguing policy defines an early-decade range before computing one. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
In an actual workflow, compare search results, user instructions and any exported record, because a correct sentence can still be attached to the wrong data value. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 19. A missing century
Source situation. A damaged form retains the year ’32’ and no century. Misleading shortcut. ‘2032’ silently chooses the current century.
Better reader-facing rendering. “’Year ending in 32; century not known.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. This is a useful test of whether a translator can preserve a distinction while still giving a busy reader an immediately usable sentence.
What to verify. Preserve the illegible or omitted portion in the source transcription. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Ask the requester to confirm the business or research consequence of this wording. If it changes where someone goes, what they believe or what a program does, sign-off must be explicit. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 20. A probable event date
Source situation. The editor marks an expedition arrival as ‘probably 14 August’. Misleading shortcut. Omitting ‘probably’ makes the inferred date appear witnessed.
Better reader-facing rendering. “’Probably arrived on 14 August, according to the editor.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. A second reviewer who sees only the final sentence could reasonably infer more than the source says; that is why the qualification must remain visible.
What to verify. Keep attribution to the person making the inference. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
A faithful translation should remain accurate when copied into a report without its surrounding paragraph. Include enough identifying context and qualification for that realistic reuse. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 21. An interval from oral history
Source situation. Interviewee says ‘I lived there from roughly 1982 until about 1988’. Misleading shortcut. ‘Resident 1982–1988’ strips both estimates from the endpoints.
Better reader-facing rendering. “’Lived there approximately 1982 to 1988, according to recollection.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. The tempting shortcut is to make the target text look complete, but completeness of typography is not completeness of evidence.
What to verify. Retain that both bounds are recalled, not administrative records. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Keep the supporting source field available to the reviewer. If the record owner cannot confirm the missing fact, the published copy should say that it is not established rather than conceal the gap. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 22. A date in French
Source situation. French source ‘vers 1900’ is shown beside a catalogued day. Misleading shortcut. ‘On 1900’ is ungrammatical and discards ‘vers’ as uncertainty.
Better reader-facing rendering. “’Around 1900′ in display; preserve the source phrase in notes.” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. Here the difficulty is not a rare vocabulary item. It is the relationship between the words and the record they purport to describe.
What to verify. Confirm whether the original is an artistic dating convention or documentary estimate. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
In an actual workflow, compare search results, user instructions and any exported record, because a correct sentence can still be attached to the wrong data value. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 23. An interval in Spanish
Source situation. Spanish ‘hasta finales de 2020’ refers to the end portion of a year. Misleading shortcut. ‘Until 31 December 2020’ invents an exact boundary.
Better reader-facing rendering. “’Until late 2020.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. This is a useful test of whether a translator can preserve a distinction while still giving a busy reader an immediately usable sentence.
What to verify. Seek a project-level definition before treating this as a computational cut-off. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Ask the requester to confirm the business or research consequence of this wording. If it changes where someone goes, what they believe or what a program does, sign-off must be explicit. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 24. Overlapping candidate windows
Source situation. Two scholars propose 1902–1905 and 1904–1908 for a photograph. Misleading shortcut. Publishing ‘1902–1908’ as one proven interval disguises disagreement.
Better reader-facing rendering. “’Estimates differ: 1902–1905 or 1904–1908.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. A second reviewer who sees only the final sentence could reasonably infer more than the source says; that is why the qualification must remain visible.
What to verify. Save citation for each interpretation and flag the record as contested. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
A faithful translation should remain accurate when copied into a report without its surrounding paragraph. Include enough identifying context and qualification for that realistic reuse. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 25. A recurrence exception
Source situation. A reminder happens monthly except during the August closure. Misleading shortcut. ‘Monthly’ by itself triggers an August action in a new locale.
Better reader-facing rendering. “’Monthly, excluding August while the service is closed.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. The tempting shortcut is to make the target text look complete, but completeness of typography is not completeness of evidence.
What to verify. Test the target scheduler against the exclusion calendar. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Keep the supporting source field available to the reviewer. If the record owner cannot confirm the missing fact, the published copy should say that it is not established rather than conceal the gap. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 26. A year in a date picker
Source situation. The interface accepts only full calendar dates but a researcher has ‘circa 1920’. Misleading shortcut. Choosing any single day as the underlying truth damages later exports.
Better reader-facing rendering. “’Estimated year: around 1920′ in human text; store any search proxy separately.” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. Here the difficulty is not a rare vocabulary item. It is the relationship between the words and the record they purport to describe.
What to verify. Verify the database schema’s supported uncertainty fields before migration. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
In an actual workflow, compare search results, user instructions and any exported record, because a correct sentence can still be attached to the wrong data value. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 27. A planned versus actual date
Source situation. A transport project was planned for completion in 2027 but actually opened in 2029. Misleading shortcut. One field called ‘completion date 2027’ turns a target into an event.
Better reader-facing rendering. “’Planned completion: 2027; actual opening: 2029.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. This is a useful test of whether a translator can preserve a distinction while still giving a busy reader an immediately usable sentence.
What to verify. Confirm the type of event and evidence date in both languages. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
Ask the requester to confirm the business or research consequence of this wording. If it changes where someone goes, what they believe or what a program does, sign-off must be explicit. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
Case 28. A corrected catalogue entry
Source situation. An archivist revised a ship launch estimate after finding a newspaper dated 12 June. Misleading shortcut. Overwriting the prior approximate date with a new exact day without provenance destroys auditability.
Better reader-facing rendering. “’Updated launch date: 12 June [verified year]; previous estimate retained in revision history.’” The aim is not to make every target sentence longer, but to preserve the piece of information that determines what the text legitimately means. A second reviewer who sees only the final sentence could reasonably infer more than the source says; that is why the qualification must remain visible.
What to verify. Record source citation, correction authority and known year before public display. This check is not optional decoration: a second-language reader may act on the translated line without access to the original. The review should therefore test the assertion, its limits and its link to the underlying record rather than merely compare dictionary equivalents.
A faithful translation should remain accurate when copied into a report without its surrounding paragraph. Include enough identifying context and qualification for that realistic reuse. For a useful practice exercise, draft a deliberately overconfident one-line translation of this case, then repair it by restoring only the evidence-backed qualification. Explain to a colleague why the shorter version would have misled them.
A complete editorial and data review method
Stage one is intake. Record the source language, target language, audience, document owner, authoritative version, intended display surface and any source identifiers. Separate source quotations from translator commentary. State whether the task is to explain an uncertain record, migrate database content, localize a UI label or prepare a legally consequential notice. These are different deliverables, even if the initial sentence looks the same. A team that skips intake will spend time debating English style while important meaning is lost in an unexamined field or attachment.
Stage two is a controlled bilingual issue log. Create a row for the original expression, its proposed translation, its referent, what is known, what remains uncertain, an alternative rendering and the reason for acceptance. Include the reviewer and date of approval. The log should distinguish a data correction from a translation decision: changing the truth of the record requires an authorized source correction, while changing an explanation may be permitted within the translation brief. For repeat terminology, document the concept as well as the wording so that later translators do not mechanically copy an expression into a different context.
Stage three is testing. Open the translated item as a reader would: mobile page, printed extract, search result, map, schedule, spreadsheet or database export as appropriate. Ask whether the target text still points to the same thing, describes the same period, leaves the same uncertainty and supports the same authorized decisions. Challenge any symbol that looks technical and exact when the source was tentative. Challenge any comforting general term that hides distinct categories. A concise target version is welcome only when the original evidential boundary remains recoverable.
Stage four is publication control. The source owner confirms factual assertions and release scope; a qualified translator checks fluency, register and terminology; a domain reviewer checks consequential meaning; and an editor checks headings, explanations, accessibility and links. Record unresolved questions explicitly. If a source contains a contradiction, do not make up a seamless compromise to protect an attractive headline. Either secure an authoritative correction or tell readers the limitation in plain language.
A four-week practice plan for translators and reviewers
Week one: read the public scope of the cited standard, then build a terminology register from ten authentic but permitted sample records. For each record, identify which fields are labels and which fields establish identity or validity. Produce a short English gloss that preserves the original ambiguity and write one question that would resolve it.
Week two: have a colleague draft ten deliberately plausible but misleading paraphrases. Correct each one with the smallest necessary phrase, not a sprawling warning. Compare the reader’s likely inference before and after the repair. Note where a simple omitted adjective, qualifier or authority label changes the decision.
Week three: place the revised sentences into realistic presentation surfaces: a catalogue, a data table, a browser snippet, a public notice and a bilingual training handout. Check whether the surrounding interface changes what the reader assumes. Test exports and link targets with the same rigor as the main article.
Week four: commission a blind review from a fluent target-language reader and a domain specialist. Give them the target version first and ask what they believe it asserts. Then provide the source evidence and compare the interpretations. Summarize recurrent errors and revise the project’s termbase and review checklist.
Frequently asked questions
Does ISO 8601-2 mean every approximate date is machine-readable?
No. Implementations differ, and a human phrase does not automatically become a valid interchange string. Check the actual supported syntax and published edition. If your specific document goes beyond this example, preserve the uncertainty in the first published version and obtain a source-owner decision before replacing it with a more definitive statement.
Can I translate circa as exactly?
No. ‘Circa’, ‘around’ and similar markers communicate approximation and must not be discarded to make a date look tidier. If your specific document goes beyond this example, preserve the uncertainty in the first published version and obtain a source-owner decision before replacing it with a more definitive statement.
Does the 2019 edition remain relevant in 2026?
Yes. ISO’s catalogue lists the 2019 edition as published, with Amendment 1 issued in 2025; a later edition is still under development as of this article’s research. If your specific document goes beyond this example, preserve the uncertainty in the first published version and obtain a source-owner decision before replacing it with a more definitive statement.
Should I convert a year-only entry to the first of January?
Only as an explicitly labeled technical proxy, if the record owner authorizes it. Never describe the proxy as the actual event date. If your specific document goes beyond this example, preserve the uncertainty in the first published version and obtain a source-owner decision before replacing it with a more definitive statement.
Can a date range mean two candidate dates?
Not necessarily. A continuous interval, a set of events and alternative possibilities express different information. If your specific document goes beyond this example, preserve the uncertainty in the first published version and obtain a source-owner decision before replacing it with a more definitive statement.
What if the database cannot store uncertainty?
Keep the original expression and qualifier in a separate preserved field, document the technical limitation and request a schema change when the use requires exact fidelity. If your specific document goes beyond this example, preserve the uncertainty in the first published version and obtain a source-owner decision before replacing it with a more definitive statement.
Is a recurring schedule the same as recorded occurrences?
No. The rule for future dates and the historical list of events are separate kinds of data. If your specific document goes beyond this example, preserve the uncertainty in the first published version and obtain a source-owner decision before replacing it with a more definitive statement.
Who approves translated deadlines?
The operational owner and legal or compliance reviewer when the deadline controls eligibility, payments or rights. If your specific document goes beyond this example, preserve the uncertainty in the first published version and obtain a source-owner decision before replacing it with a more definitive statement.
Final principle: truthful clarity is more useful than false certainty
A trustworthy date translation does not claim knowledge that the source never possessed. Retaining an explicit unknown, an approximate bound or a recurrence rule may feel less elegant than a neat date, but it is more useful for researchers, applications and future readers.
Professional communication is not an exercise in making every translated phrase appear tidy. A reader needs to understand what the source establishes, why it matters, what remains unknown and where an answer could be checked. Precise translation follows the underlying evidence through changes of vocabulary, script, format and software. It does not manufacture completeness in the gaps left by the original. The most valuable sentence is often the one that gives a practical instruction without hiding the condition that makes it correct.
Verified reference and protected learning routes
Official catalogue, reviewed 5 November 2026: ISO 8601-2:2019 — public ISO publication description; related official or technical reference: supporting reference. This article offers original translation-training examples and does not reproduce or claim compliance with paid normative clauses. Readers who need to implement a formal ISO profile should consult the licensed standard and the relevant technical or professional authority.
The conceptual owner for this series remains The Master Art of Translation (EDKSG-TRANS-MASTER-WORLD-000). Term and meaning development belongs to the protected Vocabulary Learning Hub. Sentence-level clarity, reference and interpretation connect to How English Works: Translation and Meaning Across Languages. This narrow technical guide is a child application of those resources and does not create a competing translation hub.