To translate a historic masonry condition survey correctly, treat each photograph, drawing symbol, condition code and repair note as a record of a particular part of a particular building at a particular time. Readers searching for translation of masonry crack maps, stone conservation condition reports, mortar-joint descriptions, salt efflorescence, damp evidence or historic-building restoration records need an accurate distinction between what was observed, what was suspected, what was proposed and what was actually completed. A fluent sentence that turns an observation into a diagnosis, or a proposal into finished restoration, changes the conservation record.
A facade survey may identify friable mortar at grid C4, white surface deposits below a parapet, stepped cracks near an opening and previous cement-rich pointing. The evidence might come from a visual walk-through rather than laboratory salt analysis or structural monitoring. A translator must preserve the drawing reference, elevation, area, width if measured, date, observer and uncertainty. No glossary can compensate for losing the difference between a surface deposit and a confirmed cause; the real goal is that an architect, conservator, site contractor or historian reading the translation sees the same condition evidence.
Quick guide: locate the original drawing and survey date; preserve grid coordinates and numbered photographs; translate ‘crack’, ‘open joint’, ‘spalling’, ‘delamination’, ‘efflorescence’ and ‘staining’ as distinct observations; retain actual measurements and units; separate suspected mechanisms from verified findings; distinguish a recommendation, a specification, a work instruction and completed intervention; avoid declaring structural safety, heritage authenticity or successful repair from a translation alone.
The visible problem: one wall, several competing stories
A single stone wall can have a historic construction phase, several repair campaigns, a recent leak, and new survey observations. One photograph might show white crystals. A conservation surveyor may write ‘possible soluble salts’, a moisture specialist may later collect samples, and a contractor may record a trial cleaning patch. If a translation collapses those three events into ‘salt problem treated’, it loses evidence, date, agency and uncertainty. Future readers might believe the source proves diagnosis and completed remedial work when neither was recorded.
The building’s orientation matters too. ‘North elevation’, ‘west return’, ‘bay 6’, ‘lower plinth’, ‘intrados’, ‘extrados’, and ‘room-side face’ are spatial locators, not embellishments. Moving a record from north to south changes the object being discussed. Conservation plans commonly use condition legends and photographs, so translation is an alignment task between words, codes and geometry. A good reader must be able to move from the translated sentence to the exact source location.
Historic masonry varies greatly: brick, limestone, sandstone, granite, terracotta, rubble masonry and mixed construction behave differently. Even when two languages have similar everyday words for stone and cement, specialist distinctions can change meaning. The translator should record the term selected by the surveyor and ask for technical clarification where material identification is uncertain rather than reclassify the structure from a photograph.
First principles: record the wall without rewriting its history
Survey evidence is not a diagnosis
The note ‘damp patch observed’ describes appearance. It does not by itself identify rising damp, roof leakage, condensation or a damaged pipe. Translating observation as cause replaces the investigator’s uncertainty with a new claim. Keep hypotheses attributed to whoever made them.
Field-reading discipline: Each translated line should carry an object, place, time, status and source of evidence. A conservator should be able to locate the original condition on a drawing and distinguish material description from explanation. If either part is absent, add the source locator or preserve an explicit uncertainty note before releasing the translation.
Condition and treatment are different categories
A condition map shows present evidence: cracks, staining, voids, loss or deposits. A repair schedule may propose repointing, consolidation or monitoring. A work-completion register states what happened. Put these three categories in different translated columns and never promote one to another.
Field-reading discipline: Each translated line should carry an object, place, time, status and source of evidence. A conservator should be able to locate the original condition on a drawing and distinguish material description from explanation. If either part is absent, add the source locator or preserve an explicit uncertainty note before releasing the translation.
Historic fabric has a timeline
Original construction, later alteration, earlier patch repair and recent intervention need separate dates and voices. ‘Existing lime mortar’ does not necessarily mean ‘original lime mortar’. Avoid turning old into original or making a modern replacement seem historic without documentary support.
Field-reading discipline: Each translated line should carry an object, place, time, status and source of evidence. A conservator should be able to locate the original condition on a drawing and distinguish material description from explanation. If either part is absent, add the source locator or preserve an explicit uncertainty note before releasing the translation.
Position is a primary identifier
Elevation, bay, level, grid square, photograph number and drawing revision join an observation to a physical location. ‘J4 / east wall / photograph 18’ is not an attractive caption; it is how the statement can be checked. Preserve orientation and drawing numbering exactly.
Field-reading discipline: Each translated line should carry an object, place, time, status and source of evidence. A conservator should be able to locate the original condition on a drawing and distinguish material description from explanation. If either part is absent, add the source locator or preserve an explicit uncertainty note before releasing the translation.
Measured cracks are more specific than adjectives
A report may record a width in millimetres, a monitored displacement, or merely ‘hairline’. Do not replace an unmeasured observation with a measurement or infer structural movement from one snapshot. The translator needs to keep method, unit, comparison date and limitations together.
Field-reading discipline: Each translated line should carry an object, place, time, status and source of evidence. A conservator should be able to locate the original condition on a drawing and distinguish material description from explanation. If either part is absent, add the source locator or preserve an explicit uncertainty note before releasing the translation.
Efflorescence is a description before a cause
A visible powdery white deposit may be described as efflorescence, but material identification or moisture mechanism may remain unverified. Never translate ‘possible salts’ as a proven chemical analysis. If the source says ‘suspected’, that qualifier is part of the evidence.
Field-reading discipline: Each translated line should carry an object, place, time, status and source of evidence. A conservator should be able to locate the original condition on a drawing and distinguish material description from explanation. If either part is absent, add the source locator or preserve an explicit uncertainty note before releasing the translation.
Spalling, flaking and scaling are not interchangeable
Different forms of surface material loss can be described in specialist condition legends. Check the project’s definitions rather than assume everyday synonyms. A mapped symbol must retain the same category in all translated captions, tables and photographs.
Field-reading discipline: Each translated line should carry an object, place, time, status and source of evidence. A conservator should be able to locate the original condition on a drawing and distinguish material description from explanation. If either part is absent, add the source locator or preserve an explicit uncertainty note before releasing the translation.
Mortar terms need context
Pointing describes visible mortar joints; repointing is intervention to renew mortar in joints; rendering is a coating; grouting may fill voids. A conservation note about ‘incompatible pointing’ cannot become ‘the wall was rendered’ merely because both involve mortar.
Field-reading discipline: Each translated line should carry an object, place, time, status and source of evidence. A conservator should be able to locate the original condition on a drawing and distinguish material description from explanation. If either part is absent, add the source locator or preserve an explicit uncertainty note before releasing the translation.
A historic building is not necessarily defective
Conservation surveys record wear, alteration, patina and conditions that may or may not require intervention. Rewording every observation as a fault can bias a decision toward unnecessary work. Preserve the reporting style and significance language.
Field-reading discipline: Each translated line should carry an object, place, time, status and source of evidence. A conservator should be able to locate the original condition on a drawing and distinguish material description from explanation. If either part is absent, add the source locator or preserve an explicit uncertainty note before releasing the translation.
Protective actions may be temporary
A temporary scaffold, tarpaulin, area closure or tell-tale gauge might be installed for monitoring. It is not a final repair or a declaration that a crack is stable. Keep action type, date, duration and intended follow-up visible.
Field-reading discipline: Each translated line should carry an object, place, time, status and source of evidence. A conservator should be able to locate the original condition on a drawing and distinguish material description from explanation. If either part is absent, add the source locator or preserve an explicit uncertainty note before releasing the translation.
The twelve-stage method for multilingual conservation records
Step 1: Freeze the survey version
Record document title, drawing revision, author, survey date and photographic set before making language changes.
Example and decision: Two versions of a facade map may use the same code C4 for different conditions; mixing revisions can relocate damage. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 2: Draw the source hierarchy
Separate overview plan, elevation, condition map, legend, itemized schedule, photograph caption, analysis report and repair specification.
Example and decision: A phrase in the contractor’s method statement has a different force from a visual observation in the first survey. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 3: Compile an exact location key
Keep facade orientation, bay numbers, datum, floor level, grid, room number and photo IDs in a comparison table.
Example and decision: ‘East return B7’ is a locator, while ‘east-facing’ can describe orientation; choose the source’s actual spatial role. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 4: Preserve condition-code definitions
Translate each coded condition once using the project’s legend, maintaining codes unchanged where possible.
Example and decision: If F means ‘flaking’ in the legend, do not silently expand it to ‘failed repair’ in captions. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 5: Separate evidence from interpretation
Mark source sentences as observed, measured, suspected, laboratory-confirmed, historically inferred, recommended or completed.
Example and decision: ‘Possible salt-related deterioration’ is not the same as ‘salt damage demonstrated by testing’. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 6: Lock measurements and measurement method
Retain crack width, reference plane, moisture meter reading, survey date and any caveat about device limitations.
Example and decision: A surface conductivity number cannot be relabelled absolute moisture content. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 7: Map materials and building phases
Distinguish brick or stone type, bedding mortar, pointing, render, earlier replacement and new intervention.
Example and decision: ‘Later cementitious pointing’ should not imply all mortar in the building is Portland cement. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 8: Translate treatment vocabulary precisely
Differentiate monitoring, cleaning trial, repointing, desalination, consolidation, pinning and replacement as the source does.
Example and decision: A note recommending a test patch does not authorize full-scale removal of historic material. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 9: Check status verbs
Use proposed, specified, approved, instructed, commenced, paused, completed and inspected according to the record.
Example and decision: The contractor’s quotation must not become an executed repair log. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 10: Cross-reference photographs
Check image number, arrow, crop and direction against each translated caption, including historical and current views.
Example and decision: A caption describing the wrong elevation makes even perfectly translated condition terms unusable. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 11: Review high-consequence claims
Query descriptions of instability, public access, risk, temporary support, removal of original fabric and environmental hazards.
Example and decision: Language review cannot replace professional engineering or conservation assessment; refer to the authorized reviewer. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Step 12: Record uncertainties and publish a traceable version
Issue a bilingual termbase, translation change log, source-document revision and list of outstanding queries.
Example and decision: A later restoration team should be able to see what the original survey said and which terms required expert resolution. Write the evidence pointer beside the target-language sentence. For release, ask whether the translation changes the location, severity, time, intervention status or confidence level. One change is enough to require revision.
Twenty worked examples of condition and conservation translation
The following are constructed educational cases. They illustrate editorial decisions, not diagnoses of an actual building, approval for building work, or a substitute for conservation expertise.
Worked case 1: Powdery white deposit
Original evidence: Survey: ‘white powder-like deposit; suspected efflorescence; sample not taken’, lower course of north wall.
Tempting but wrong: Translate ‘confirmed salt crystallisation caused by ground moisture’.
Supported translation: Use ‘white powdery deposit, possibly efflorescence; no sample collected’.
Reasoning: The appearance, possible interpretation and missing test result must remain separate. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 2: Hairline crack at an arch
Original evidence: Map marks a hairline crack at the arch springing, photograph P12, no width recorded.
Tempting but wrong: Write ‘structural crack measuring 2 mm’.
Supported translation: Preserve ‘hairline crack’ and photo P12; do not invent width or structural classification.
Reasoning: A visual descriptive term cannot become quantitative evidence. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 3: Crack gauge monitoring
Original evidence: An engineer’s note reads ‘tell-tale installed; baseline photographed; no follow-up reading yet’.
Tempting but wrong: Write ‘crack proven stable’.
Supported translation: Translate installation and baseline only; follow-up remains outstanding.
Reasoning: Monitoring apparatus is not itself a longitudinal conclusion. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 4: Spalled brick face
Original evidence: Legend code SP indicates loss of outer brick face over 0.3 square metres.
Tempting but wrong: Write ‘surface cleaned’.
Supported translation: Record spalling, marked area and location; cleaning is not recorded.
Reasoning: The condition is material loss, not a maintenance action. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 5: Open mortar joints
Original evidence: Condition sheet states ‘open pointing joints at sill; extent approximate’.
Tempting but wrong: Translate ‘stone blocks have separated’.
Supported translation: Use ‘open mortar joints beneath sill; extent approximate’.
Reasoning: A joint condition must not become movement of the masonry units. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 6: Earlier cement patch
Original evidence: Caption reads ‘hard grey pointing, possibly twentieth-century repair’.
Tempting but wrong: Write ‘original Roman cement’.
Supported translation: Preserve visible material description, possible age and uncertainty.
Reasoning: Appearance alone does not establish exact binder chemistry or construction phase. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 7: Missing cornice fragment
Original evidence: Elevation notes ‘fragment absent by 1984 photograph’.
Tempting but wrong: Write ‘fragment removed in 1984’.
Supported translation: Translate ‘already absent in a photograph dated 1984’.
Reasoning: A historical photo establishes a terminus of absence, not the event’s date or cause. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 8: Repointing recommended
Original evidence: Schedule says ‘consider trial repointing after mortar analysis’.
Tempting but wrong: Write ‘repointing completed’.
Supported translation: Keep tentative proposal and required investigation.
Reasoning: An option pending research must not be recast as work performed. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 9: Moisture stain
Original evidence: Survey records darker masonry beneath gutter; cause undetermined.
Tempting but wrong: Translate ‘the gutter leaks into the wall’.
Supported translation: Use ‘dark staining below gutter; source of moisture not established’.
Reasoning: Location near an element can inform a hypothesis but does not prove causation. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 10: Salt laboratory result
Original evidence: Lab report identifies a measured soluble-ion profile for specimen S-17.
Tempting but wrong: Generalize ‘wall is saline’.
Supported translation: Report exact specimen, test and measured result without extrapolating to the whole facade.
Reasoning: Sample-specific data cannot automatically describe the entire structure. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 11: Photograph reversed
Original evidence: A mirrored display image causes the translator to label the left buttress ‘west’.
Tempting but wrong: Follow the visual left-right position.
Supported translation: Refer to drawing north arrow and photograph orientation, then confirm compass terms.
Reasoning: Image presentation can mislead spatial terms; true location comes from survey metadata. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 12: Consolidation trial
Original evidence: Contractor record says ‘trial applied to panel T2; evaluation pending’.
Tempting but wrong: Write ‘facade consolidated successfully’.
Supported translation: State ‘trial treatment applied at panel T2; evaluation pending’.
Reasoning: Scope and outcome are both limited. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 13: Replacement stone
Original evidence: Specification proposes replacement of a severely eroded stone, subject to approval.
Tempting but wrong: Translate ‘historic stone replaced’.
Supported translation: Keep proposal and pending authorization, with stone location.
Reasoning: Plans do not prove work or historical attribution. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 14: Render detachment
Original evidence: A condition map records ‘hollow-sounding render’ at grid D6.
Tempting but wrong: Write ‘wall structure hollow’.
Supported translation: Use ‘hollow-sounding render at D6’ with survey method if stated.
Reasoning: Surface finish and structural core are different objects. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 15: Freeze–thaw hypothesis
Original evidence: Report uses ‘possibly associated with freeze–thaw action’.
Tempting but wrong: Convert to ‘freeze damage confirmed’.
Supported translation: Preserve possible mechanism, source attribution and lack of confirmation.
Reasoning: Historic buildings often have several plausible deterioration mechanisms. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 16: Cleaning test patch
Original evidence: Photo caption says ‘trial cleaned; adjacent fabric untreated’.
Tempting but wrong: Call entire facade cleaned.
Supported translation: Translate exact trial area and untreated comparison zone.
Reasoning: The intervention’s spatial boundary is essential to later evaluation. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 17: Patina and soiling
Original evidence: Statement distinguishes natural patina from localized black crust.
Tempting but wrong: Use the same word ‘dirt’ for both.
Supported translation: Translate patina and black crust as separate source terms.
Reasoning: A conservation decision may value one surface feature while treating another as condition evidence. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 18: Urgent review notation
Original evidence: A survey box says ‘refer cracking at stair bay to structural engineer’.
Tempting but wrong: Translate ‘unsafe for public use’.
Supported translation: Render the referral instruction without inventing an access prohibition.
Reasoning: Professional review request is not a finding of immediate danger. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 19: Drawing revision
Original evidence: An updated elevation shifts photo P21 from bay 8 to bay 9.
Tempting but wrong: Use the old map’s bay reference because the photograph seems the same.
Supported translation: Match each caption to the correct drawing version and document the revision.
Reasoning: A revision is source history, not an inconvenience to be erased. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Worked case 20: Work completion certificate
Original evidence: Completion note lists repointing at west parapet; no work at east gable.
Tempting but wrong: Say ‘masonry restored throughout’.
Supported translation: Translate only the completed west parapet scope; preserve absent east-gable work.
Reasoning: Global claims of restoration erase localized work limits and leave false histories. The translator should compare the source text to the drawing or photograph to which it refers. Before release, check what a new reader might now believe about condition, cause, area or work status. The target should never create a stronger claim than the survey actually makes.
Failure modes: how conservation meaning is quietly lost
Observation becomes diagnosis
What happens: Translate ‘surface staining’ as ‘rising damp due to defective foundations’.
Correction: Separate condition description and possible mechanisms; preserve uncertainty and author. Read the sentence again with the location and date appended. If the claim is broader, stronger, newer or more certain than the source, it is not faithful. Record the correction in the change log so another translator does not repeat the same error.
Proposed work becomes completed work
What happens: A repair specification becomes a narrative of successful restoration.
Correction: Translate status verbs and retain order dates, approval states and completion certificates. Read the sentence again with the location and date appended. If the claim is broader, stronger, newer or more certain than the source, it is not faithful. Record the correction in the change log so another translator does not repeat the same error.
Grid and orientation migrate
What happens: North elevation G3 becomes south elevation or room G3.
Correction: Cross-check north arrow, reference key, image caption and drawing revision before final. Read the sentence again with the location and date appended. If the claim is broader, stronger, newer or more certain than the source, it is not faithful. Record the correction in the change log so another translator does not repeat the same error.
Material is renamed
What happens: Lime mortar, cementitious repair and render are all called ‘cement’.
Correction: Use the source material term and retain uncertainty where tests have not identified composition. Read the sentence again with the location and date appended. If the claim is broader, stronger, newer or more certain than the source, it is not faithful. Record the correction in the change log so another translator does not repeat the same error.
Extent is exaggerated
What happens: Localized spalling at two panels becomes deteriorated facade throughout.
Correction: Maintain the affected area, mapped points and whether the scope was measured or estimated. Read the sentence again with the location and date appended. If the claim is broader, stronger, newer or more certain than the source, it is not faithful. Record the correction in the change log so another translator does not repeat the same error.
Crack status is invented
What happens: One visual crack observation becomes ‘progressive movement’.
Correction: Only monitoring across dates supports a time-trend statement; preserve single-observation limits. Read the sentence again with the location and date appended. If the claim is broader, stronger, newer or more certain than the source, it is not faithful. Record the correction in the change log so another translator does not repeat the same error.
Heritage value is flattened
What happens: Patina, staining and deterioration become one universal ‘damage’ code.
Correction: Keep the conservation vocabulary and significance distinctions from the project’s legend. Read the sentence again with the location and date appended. If the claim is broader, stronger, newer or more certain than the source, it is not faithful. Record the correction in the change log so another translator does not repeat the same error.
New work is called original
What happens: A twentieth-century patch is translated as original stonework.
Correction: Use dated evidence and construction-phase labels exactly; do not infer age from colour. Read the sentence again with the location and date appended. If the claim is broader, stronger, newer or more certain than the source, it is not faithful. Record the correction in the change log so another translator does not repeat the same error.
Temporary support is called repair
What happens: A shoring arrangement or protective covering becomes an intervention finished.
Correction: Translate protective, temporary, monitoring and permanent status separately. Read the sentence again with the location and date appended. If the claim is broader, stronger, newer or more certain than the source, it is not faithful. Record the correction in the change log so another translator does not repeat the same error.
Technical approval is manufactured
What happens: The translator’s polished report implies a building is structurally sound or cleared for use.
Correction: Limit output to translated evidence and refer engineering decisions to authorized professionals. Read the sentence again with the location and date appended. If the claim is broader, stronger, newer or more certain than the source, it is not faithful. Record the correction in the change log so another translator does not repeat the same error.
Practice and transfer: exercises for translators and students
Exercise 1: Make a condition key
Draw a fictional wall with four bays and write separate codes for open joints, staining, spalling and loss of material. Translate the legend, then check that every symbol still points to the same observation.
Assessment criterion: A successful response keeps measured versus estimated information separate, retains survey coordinates, and avoids adding repair or structural conclusions. Learners should be able to explain why one apparently fluent sentence was rejected and point to the field evidence that supports the revised wording.
Exercise 2: Distinguish observed from inferred
Write ten sample notes containing words such as observed, possibly, suggests, measured, confirmed, recommended and completed. Sort by evidence status before translating the sentences.
Assessment criterion: A successful response keeps measured versus estimated information separate, retains survey coordinates, and avoids adding repair or structural conclusions. Learners should be able to explain why one apparently fluent sentence was rejected and point to the field evidence that supports the revised wording.
Exercise 3: Follow a repair timeline
Create a three-date training record: survey, proposed works, partial repair. Translate into a table without presenting any proposed action as completed.
Assessment criterion: A successful response keeps measured versus estimated information separate, retains survey coordinates, and avoids adding repair or structural conclusions. Learners should be able to explain why one apparently fluent sentence was rejected and point to the field evidence that supports the revised wording.
Exercise 4: Build a photo-to-grid matrix
Assign photographs to north and west elevations, including an awkward return. Ask a peer to find each point using the target-language captions alone.
Assessment criterion: A successful response keeps measured versus estimated information separate, retains survey coordinates, and avoids adding repair or structural conclusions. Learners should be able to explain why one apparently fluent sentence was rejected and point to the field evidence that supports the revised wording.
Exercise 5: Train technical terms
Use pairs such as pointing/repointing, spalling/flaking, patina/soiling, staining/efflorescence, crack/open joint. Define distinctions in your own words before choosing target terms.
Assessment criterion: A successful response keeps measured versus estimated information separate, retains survey coordinates, and avoids adding repair or structural conclusions. Learners should be able to explain why one apparently fluent sentence was rejected and point to the field evidence that supports the revised wording.
Exercise 6: Audit measurements
Create a training table with 0.2 mm, 2 mm, 0.2 m and an unmeasured hairline crack. Use a separate check to catch unit and scale errors.
Assessment criterion: A successful response keeps measured versus estimated information separate, retains survey coordinates, and avoids adding repair or structural conclusions. Learners should be able to explain why one apparently fluent sentence was rejected and point to the field evidence that supports the revised wording.
Exercise 7: Preserve alternative explanations
Describe a damp patch near a downpipe without declaring a cause. Write two possible hypotheses in clearly attributed language and keep the actual observation separate.
Assessment criterion: A successful response keeps measured versus estimated information separate, retains survey coordinates, and avoids adding repair or structural conclusions. Learners should be able to explain why one apparently fluent sentence was rejected and point to the field evidence that supports the revised wording.
Exercise 8: Write a field-ready handover
Include source image index, condition legend, drawing revision, translated report, unresolved queries and responsible reviewers. Test whether a conservator can trace every target claim.
Assessment criterion: A successful response keeps measured versus estimated information separate, retains survey coordinates, and avoids adding repair or structural conclusions. Learners should be able to explain why one apparently fluent sentence was rejected and point to the field evidence that supports the revised wording.
Frequently asked questions about translating historic masonry reports
How do I translate a masonry condition survey?
Begin with the source legend, elevation drawings and photo index, then create a bilingual key for codes and location labels. Translate the observations before descriptions of causes or treatment. Review every instance of proposed, completed, possible and confirmed.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
Is efflorescence the same as damp?
No. Efflorescence refers to a visible deposit associated with soluble salts, whereas damp describes moisture-related condition. A survey may observe both, one, or neither. Translate the specific evidence and preserve uncertainty about causes and composition.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
Can I translate a crack as structural damage?
Only when the source makes that determination. A crack may be observed visually without an engineering classification. Preserve its location, description and measurement, if any, and refer structural conclusions to a competent reviewer.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
What is the difference between pointing and repointing?
Pointing refers to the mortar finish in joints; repointing involves renewing or repairing mortar joints. The precise use depends on source context. Do not transform an observation about existing pointing into a description of completed repointing.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
Does white staining prove rising damp?
No. A surface appearance alone may have several explanations. Translate the observation and any named hypothesis or test result as separate claims; do not supply a cause from memory.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
Why are photograph numbers important?
They connect the written record to a documented view of the wall. Lost or swapped identifiers can move the observation from one bay to another, making a translated report unreliable even if its wording sounds correct.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
Should I modernize all old masonry vocabulary?
Not indiscriminately. A historical document may use dated material names that are themselves evidence. Preserve the original term or provide a clearly marked explanation rather than silently replacing it with a modern presumed equivalent.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
How do I translate restoration versus repair?
Respect the source’s conservation terminology and scope. Repair, conservation, preservation, restoration and reconstruction are not universal synonyms. The organization’s definitions and specific project documents should guide the target term.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
Can a proposed treatment be called work completed?
No. Recommendations, method statements, approvals, works orders, site records and final sign-offs are different stages. Match the translation’s tense and status to the recorded stage.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
What if a photograph contradicts the text?
Do not silently amend the report. Record a query identifying the photograph and text reference. The authoritative project reviewer can resolve whether the picture, caption or drawing number is incorrect.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
Is machine translation suitable for building conservation reports?
It can assist with drafts and terminology extraction, but condition codes, historic materials, units and negation require specialist review. Automated text is particularly vulnerable when source tables and figures are detached from context.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
What should a bilingual conservation handover include?
Include the source version, original drawings, photograph index, location key, terminology decisions, translation, uncertainty register and reviewer sign-off if required by the project. A language deliverable must not falsely claim structural or conservation approval.
The final review question is whether the translated text would lead a reader to locate a different defect, infer a different cause, or believe different repair work had happened. If it would, preserve the source boundaries and revise the wording.
Reference material and contextual learning routes
Relevant expert background includes Historic England on masonry buildings and moisture-related conditions, the U.S. National Park Service Preservation Briefs collection, and the Preservation Brief on repointing mortar joints. These sources inform the differences among appearance, material, intervention and diagnosis; they are not a project-specific engineering or conservation assessment.
Develop the broader translation method through Translate Meaning, Not Just Words, Context, Tone and Register, Idioms, Culture and Non-Literal Language and Terminology and Quality Checks. For precise definitions, use the Vocabulary Learning Hub; to study negation, sentence structure and careful attribution, continue to How English Works.
Historic-building translation is successful when a reader can move from a translated phrase to the right wall, image, observation and event. A record of fragile mortar should remain a record of fragile mortar. A suspected cause should remain suspected. A recommended trial should remain a recommendation. Fidelity protects both the building’s actual history and the judgments that future professionals must make.