How do you translate scientific names, chemical formulas and laboratory records correctly without changing the experiment, specimen, substance or data? Separate language from notation. Identify which elements are translatable prose, which are controlled scientific names, which are formulas, symbols, identifiers, units, sample codes, instrument outputs or method labels, and which terms have established discipline-specific equivalents. Scientific translation becomes reliable when the target preserves the same entities, measurements, relationships and evidential status as the source.
People searching for scientific translation, laboratory translation, chemical translation, scientific nomenclature translation, chemical formula translation, lab report translation, biotechnology translation, research data translation and scientific terminology translation are dealing with a mixed-language environment. A laboratory document can combine Latin binomials, gene symbols, protein names, chemical formulas, CAS numbers, concentrations, temperatures, sample identifiers, equations, units, instrument models, statistical values and free-text observations. Some of these should be translated; many must not be altered.
This guide builds a practical method for translating scientific records, laboratory reports, methods, specimen logs, research notes, analytical certificates, experimental tables, chemical and biological terminology, figure labels and supporting documentation. The aim is to preserve the same scientific object and the same evidence while making explanatory language natural and precise in the target language.
Scientific Translation Is Entity-and-Data Preservation
Scientific writing connects words to real entities and measurements. A species name identifies an organism. A chemical formula identifies a composition. A sample code points to a physical specimen. A concentration describes an experimental condition. A gene or protein symbol points to a recognised biological entity. A translation that changes any of these can create a different scientific record.
The translator therefore needs two modes at once: linguistic translation for prose and conservation for notation. The challenge is knowing which mode applies to each element.
Step 1: Identify the Scientific Domain
Start by identifying whether the document belongs to chemistry, biology, biotechnology, medicine, physics, materials science, environmental science, engineering, geology, food science, pharmacology or another field.
Domain controls terminology. “Culture” in microbiology, “stress” in mechanics, “field” in physics and “buffer” in computing or chemistry represent different concepts. General dictionaries are rarely enough for specialised scientific terms.
Step 2: Build a Protected-Notation List
Before translating, identify elements that should be preserved exactly: formulas, sample IDs, CAS numbers, gene symbols, protein symbols where applicable, accession numbers, catalogue numbers, instrument model names, equation variables, raw data fields and file names.
This protected list prevents language tools from normalising or translating technical strings. The translator can then focus attention on explanatory prose and controlled terminology.
Step 3: Distinguish Scientific Name From Common Name
An organism can have a scientific name and one or more common names. The scientific name may remain unchanged across languages, while common names often vary.
Do not replace a Latin binomial with a common target-language name unless the source itself uses the common name or the document requires both. Scientific names support precise cross-language identification.
Step 4: Preserve Binomial Nomenclature Structure
Genus and species names follow conventions of spelling and capitalisation. Do not translate them word by word. Preserve the accepted scientific form and italicisation where required by the document style.
When an abbreviated genus is used later, keep the abbreviation relationship clear. Do not expand or alter it unless the project style calls for first-use clarification.
Step 5: Handle Taxonomic Changes Carefully
Scientific classifications can change over time. A source may use an older accepted name, a synonym or a current name depending on publication date and field practice.
Do not silently modernise taxonomy unless the task includes scientific updating. If a newer name is important, add it only through an authorised editorial note or documented target convention.
Step 6: Preserve Chemical Formulas Exactly
Formulas such as H2O, NaCl, CO2 or more complex molecular formulas are notation. They should not be translated into target-language words unless the source also supplies the compound name and the task requires translating that name.
Protect subscripts, superscripts, charges, parentheses and stoichiometric numbers. A formatting error can change the substance or ion.
Step 7: Translate Chemical Names by Established Nomenclature
Chemical names may have systematic names, retained names, common names, trade names or laboratory shorthand. Identify which type the source uses.
Use established target-language nomenclature rather than literal word substitution. Where international nomenclature is standard, preserve that structure; where target-language conventions differ in spelling or word order, verify them with authoritative chemical references.
Step 8: Distinguish Substance Name From Formula
A document may use both “sodium chloride” and “NaCl.” The name is language; the formula is notation.
Translate the name using the accepted target term, but preserve the formula. Check that both still point to the same substance after translation.
Step 9: Preserve CAS and Other Registry Numbers
Registry numbers identify substances independently of language. Copy them exactly.
Do not infer or correct a registry number from the chemical name. If the source appears inconsistent, flag the issue. Silent correction can make the target record diverge from the original evidence.
Step 10: Preserve Concentration Expressions
Scientific records may use molarity, mass concentration, percentage, parts per million, dilution ratios or other concentration expressions.
Translate labels and explanatory prose, but preserve values, units and mathematical relationships. “5% w/v” is not interchangeable with “5% v/v.” The notation carries experimental meaning.
Step 11: Keep Unit Systems Stable
Scientific documents commonly use SI units, derived units and discipline-specific units. By default, preserve the source unit unless conversion is explicitly authorised.
If conversion is required, verify arithmetic, significant figures and uncertainty. Do not introduce false precision by converting a rounded source value into an excessively precise target value.
Step 12: Preserve Significant Figures
Numbers such as 2.0 and 2.000 may imply different measurement precision. Do not remove trailing zeros casually.
When translating tables, check that spreadsheet or publishing software has not reformatted values. Numerical presentation can encode the precision of the measurement.
Step 13: Preserve Uncertainty
Scientific data may include ± values, confidence intervals, error bars, detection limits or qualifiers such as approximately, below, above, less than and greater than.
Translate the prose accurately and preserve mathematical symbols. Do not turn “below detection limit” into zero, or “approximately” into an exact value.
Step 14: Keep Sample Identifiers Exact
Sample codes link text, instruments, storage locations and results. They are identities, not labels for stylistic translation.
Check case, hyphens, underscores and leading zeros. If a sample is “S-004B,” the target should not become “S-4B” or a translated equivalent.
Step 15: Preserve Batch and Lot Numbers
Batch, lot and production identifiers can be essential for traceability. Treat them like serial numbers.
Translate the field label if required, but not the identifier. Cross-check repeated values across certificates, lab sheets and result tables.
Step 16: Translate Specimen Descriptions Precisely
Biological and environmental records may distinguish blood, plasma, serum, tissue, swab, sediment, soil, water, air sample, culture or extract.
Use the correct scientific target term. Do not generalise “serum” to “blood” or “sediment” to “soil.” Specimen type affects method interpretation.
Step 17: Preserve Anatomical and Biological Terms by Concept
Scientific and medical anatomy contains standard target-language terminology. Research the accepted term rather than translating from ordinary-language roots.
Where Latin terminology is used internationally, determine whether the target document convention preserves Latin, uses a local equivalent or provides both.
Step 18: Gene Symbols and Gene Names Need Different Treatment
A gene symbol may be a controlled identifier, while the full gene name can be explanatory language. Do not translate the symbol.
For the full name, follow discipline conventions. Some target-language scientific communities retain the English full name; others translate descriptive components. Verify current usage in target-language literature.
Step 19: Protein Names, Symbols and Isoforms Need Traceability
Protein nomenclature can include names, symbols, isoforms and modification states. Preserve official symbols and numeric designations.
Do not create target abbreviations casually. A translated explanatory name should still map unambiguously to the same protein entity.
Step 20: Preserve Cell-Line Names
Cell lines are identifiers. Names such as laboratory codes or established cell-line names should remain exact.
Translate descriptive surrounding text but not the cell-line identity. Check hyphens and capitalisation carefully.
Step 21: Preserve Strain and Isolate Identifiers
Microbiology records may identify strain, isolate, clone or variant. These labels can contain letters, numbers and source references.
Do not translate them as ordinary adjectives. Preserve the identifier and translate the category label only.
Step 22: Distinguish Species, Strain and Sample
A species name identifies a taxon, a strain identifies a particular lineage or variant, and a sample code identifies the specific material tested.
Translation should not collapse these levels. Laboratory traceability depends on knowing whether a statement applies to the species generally, a strain or one sample.
Step 23: Preserve Enzyme and Assay Names
Enzymes, kits and assays may have official, generic or branded names. Check manufacturer materials and scientific literature.
If an assay name is a product name, preserve product identity. If it is a generic methodological description, translate using established laboratory terminology.
Step 24: Translate Method Names Carefully
Methods such as chromatography, spectrometry, PCR variants, titration, microscopy or sequencing have established scientific names.
Do not create novel target terminology from component words when the field already has a standard expression. Target-language journals and textbooks are useful comparable sources.
Step 25: Instrument Names and Model Numbers Are Separate
The generic instrument type may be translated; the manufacturer and model number should normally remain unchanged.
For example, translate the instrument category if needed, but preserve the exact model identifier. This supports reproducibility and equipment traceability.
Step 26: Preserve Calibration Information
Laboratory records may include calibration date, standard, lot, reference material and instrument status.
Translate labels and notes without changing the underlying traceability record. Dates, certificate numbers and reference-material identifiers deserve their own QA pass.
Step 27: Translate Standard Operating Procedure Language as Procedure
SOPs combine scientific terminology with technical instructions. Preserve sequence, conditions, temperatures, durations, volumes and safety language.
The Technical Instructions, Manuals and Procedures article provides the procedural layer. Laboratory translation adds specimen and data integrity.
Step 28: Preserve Reagent Preparation Steps
A reagent recipe may specify mass, volume, concentration, pH, sequence and final volume. Do not reorganise it casually.
“Add water to a final volume of 1 L” is not the same as “add 1 L of water.” Translation should preserve the preparation logic.
Step 29: Preserve Temperature Conditions
Laboratory temperatures can include room temperature, controlled ranges, refrigeration, freezing and high-temperature treatment.
Translate prose but preserve values and symbols. Check minus signs carefully; losing one minus sign can reverse the storage condition.
Step 30: Preserve Time Conditions
Incubation, reaction, centrifugation, drying and storage durations may be given in seconds, minutes, hours or days.
Keep unit and number together. “30 s” should not become “30 min.” A dedicated numerical pass is essential.
Step 31: Preserve Centrifugation Conditions
Laboratory protocols may report rpm or relative centrifugal force. These are not interchangeable values.
Translate the surrounding instruction but do not convert between measurement modes unless explicitly required and technically justified.
Step 32: Preserve pH and Other Scale Values
pH, absorbance, optical density and other scale values should remain associated with the correct measurement label.
Do not add units to dimensionless values or remove them where they belong. Scientific data labels must remain conceptually correct.
Step 33: Translate Qualitative Observations Faithfully
Lab notes may say cloudy, colourless, precipitate observed, faint band, slight odour, no visible growth or other observational descriptions.
Preserve degree and uncertainty. “Faint” should not become “clear,” and “no visible growth” should not become “no growth” if the detection method is visual.
Step 34: Separate Observation From Interpretation
“A white precipitate formed” is an observation. “The reaction was complete” may be an interpretation based on evidence.
Keep these levels distinct. Scientific records are stronger when the target preserves what was seen versus what was concluded.
Step 35: Preserve Detection Limits
Results can be below limit of detection, below limit of quantification, not detected or not tested. These are different statuses.
Use established analytical terminology. Do not translate “not detected” as zero unless the source explicitly does so.
Step 36: Preserve Positive, Negative and Inconclusive Results
Diagnostic or analytical reports may classify results. Translate according to the test’s intended meaning.
“Negative” can mean absence of the target under the assay conditions, not a general statement that nothing is wrong. Avoid explanatory overreach unless the source provides it.
Step 37: Preserve Replicate Structure
Experiments may use technical replicates, biological replicates or repeated measurements. Do not flatten replicate numbers into one average if the source records individual values.
Translate table labels so the reader can still identify replicate relationships.
Step 38: Preserve Control Groups and Control Samples
Positive control, negative control, blank, vehicle control, untreated control and reference standard can have specific roles.
Use consistent target terminology. Do not translate every control as one generic word if the source distinguishes them.
Step 39: Translate Statistical Terms by Discipline
Mean, median, standard deviation, standard error, confidence interval, variance, correlation and regression have established target-language equivalents.
The Academic Writing Without Changing the Claim article covers claim strength and statistical interpretation. In laboratory translation, the same terms must also align with tables and instrument outputs.
Step 40: Preserve Equations
Equations and variable symbols generally remain unchanged. Translate explanatory text and variable definitions.
Check subscripts, superscripts, Greek letters, operators and equation numbering. Publishing systems can introduce notation errors even when the translation is correct.
Step 41: Preserve Figure and Table Relationships
Figures, captions, legends, axis labels and table headings form one information system. Translate all human-readable components consistently.
Verify that labels still correspond to the correct series, sample or axis after layout. A translated legend attached to the wrong colour or symbol changes the data presentation.
Step 42: Preserve Colour and Symbol Keys
Scientific figures may encode groups through colours, shapes, line styles or markers. Translate the legend text but preserve the mapping.
Do not reorder legend entries casually if that makes readers associate the wrong description with a series.
Step 43: Protect Raw Data
Raw instrument output, data tables, chromatograms, sequences and machine-generated fields should not be linguistically normalised unless the project explicitly requires annotation.
Translate explanatory headings or notes while maintaining the source data as evidence.
Step 44: Distinguish Missing Data From Zero
Blank, NA, ND, not applicable, not determined, not detected and zero can represent different states.
Preserve the source convention or translate it through an approved data dictionary. Do not replace all non-numeric entries with one target abbreviation.
Step 45: Preserve Sequence Data
DNA, RNA or protein sequences are machine-readable scientific data. They should not be translated.
Protect spacing, case and line breaks according to the file format. Translate only surrounding annotations and labels.
Step 46: Preserve Accession Numbers
Database accession numbers and repository identifiers connect the document to external records. Copy them exactly.
If a hyperlink is present, verify that the target link still points to the same record after publishing.
Step 47: Translate Safety Language With Scientific Precision
Laboratory safety instructions may mention corrosive, toxic, flammable, oxidising, biohazardous or other hazards. Use established hazard terminology rather than everyday synonyms.
Where regulated labels or hazard statements apply, use qualified current references. Do not improvise safety wording from memory.
Step 48: Preserve Waste and Disposal Instructions
Laboratory records and SOPs may distinguish chemical waste, biological waste, sharps, contaminated materials and ordinary waste.
Translate the category and action according to the source procedure. Do not generalise disposal instructions in ways that could change handling requirements.
Step 49: Translate Quality-Control Records as Evidence
QC records can include acceptance criteria, control results, deviations, out-of-specification findings and corrective actions.
Preserve whether a criterion was met, failed, repeated or under investigation. Do not polish a deviation note into a successful result.
Step 50: Preserve Deviation and Nonconformance Status
Open, under investigation, closed, accepted with justification and rejected can be defined quality states.
Use stable target terms and keep case numbers, dates and responsible roles exact.
Step 51: Translate Certificates of Analysis Carefully
A certificate of analysis can contain product identity, batch number, tests, specifications, results, pass/fail status and authorised signatures.
Translate test names and headings consistently while preserving numeric results, limits, method references and batch identifiers exactly.
Step 52: Preserve Specification Versus Result
A table may contain a specification column and a result column. Swapping the translated headings can make a compliant batch appear noncompliant or vice versa.
Review table architecture after layout, not only individual words.
Step 53: Preserve Method References
Methods may cite standards, internal SOP numbers, pharmacopoeial methods or literature references.
Keep identifiers and edition information exact. Translate explanatory titles only according to the citation or project convention.
Step 54: Translate Laboratory Roles Precisely
Analyst, reviewer, principal investigator, technician, study director, quality unit and authorised signatory can have distinct responsibilities.
Use target terms appropriate to the scientific and organisational context. Do not collapse every role into “researcher.”
Step 55: Preserve Signature and Approval Fields
Signed, reviewed, approved, verified and released can represent different workflow actions.
Translate the action labels consistently and preserve names, dates and signatures as documentary elements.
A Scientific Data-Provenance Ledger
| Element | Example | Rule |
|---|---|---|
| Scientific entity | species, gene, compound | Use official name or identifier |
| Sample identity | S-004B, batch 22A | Preserve exactly |
| Measurement | 2.50 mg/L | Keep value, unit, precision |
| Condition | 4°C for 30 min | Preserve sequence and values |
| Observation | faint precipitate | Preserve degree |
| Status | ND, OOS, pass | Use controlled target term |
Worked Example: Formula and Name
Source: “A 0.9% sodium chloride solution (NaCl) was prepared.” Translate the compound name according to target scientific usage while preserving 0.9% and NaCl.
Do not translate NaCl as words inside the formula or change the percentage basis. The prose and notation should still refer to the same preparation.
Worked Example: Significant Figures
Source result: 1.20 mg/L. A target table automatically reformats it as 1.2 mg/L. Numerically they are equal, but the presentation may no longer communicate the same reported precision.
Check numeric formatting after export. Scientific translation includes data presentation.
Worked Example: Not Detected
Source: “Analyte X: ND; LOD 0.05 mg/L.” “ND” means the analyte was not detected under the method conditions; it does not prove a concentration of exactly zero.
Translate the legend or status according to the laboratory convention while preserving the detection limit.
Worked Example: Taxonomic Name and Common Name
A source gives a scientific species name followed by an English common name. The target may preserve the scientific name and translate the common name if a recognised target common name exists.
Do not replace the scientific name with the common name. They serve different identification functions.
Worked Example: Reagent Preparation
Source: “Dissolve 5.0 g of compound A and make up to 100 mL with solvent B.” This means the final solution volume is 100 mL, not that 100 mL of solvent is added after the solid.
Translate the preparation logic, not just the words. Laboratory procedures depend on physical interpretation.
Worked Example: Control Result
Source: “Positive control passed; negative control failed acceptance criteria.” A target that merely says “controls completed” removes essential quality information.
Preserve each control’s status and any consequence described in the source.
A Six-Pass Scientific Translation Review
- Entity pass: species, genes, proteins, chemicals, strains, cell lines and instruments.
- Identifier pass: sample IDs, registry numbers, batches, accessions and method numbers.
- Numerical pass: values, units, significant figures, ranges, uncertainty and symbols.
- Procedure pass: sequence, time, temperature, concentration, volume and conditions.
- Evidence pass: observations, interpretations, controls, status and conclusions.
- Layout pass: tables, figures, equations, legends, subscripts and superscripts.
Build a Scientific Termbase
Record discipline-specific terms, preferred target equivalents, definitions, symbols, abbreviations and source references. Link terms to official nomenclature or trusted target-language literature.
The Terminology, Glossaries and Quality Checks article provides the wider control system.
Use Target-Language Scientific Literature as Evidence
Original target-language research papers, textbooks and standards show how scientists naturally name concepts and describe methods.
The Dictionaries, Corpora and Parallel Texts article explains how to compare candidate terminology without trusting a single search result.
Keep Scientific Translation and Scientific Editing Separate
A translator may notice outdated nomenclature, apparent calculation errors or inconsistent units. Do not silently repair them unless authorised.
Flag source issues separately. Translation should remain traceable to the original record, especially in regulated, quality-controlled or evidential environments.
AI and Machine Translation in Scientific Work
AI can help explain terminology, extract protected tokens, compare repeated methods and generate draft prose. It can also help identify where notation and language are mixed.
But a model can change numbers, formulas, symbols, sample IDs or evidential strength. Protect structured scientific data from generation and verify all high-risk elements independently.
A Better AI Audit Prompt for Laboratory Records
Ask the system to extract every scientific entity, identifier, formula, value, unit, condition, sample code, method reference and result status from the source and target.
Then compare the inventories manually. This method makes scientific QA finite and explicit.
Practice Drill: Translate a Laboratory Worksheet
Create a fictional worksheet with five samples, one reagent preparation, two controls and a result table. Mark protected tokens before translating.
After translation, reconstruct the experiment from the target alone. If any sample, control or condition becomes ambiguous, revise the structure.
Transfer Drill: One Concept Across Scientific Genres
Take one chemical or biological concept and translate it in a journal abstract, SOP, certificate of analysis and educational explanation.
Keep the scientific identity stable while adapting register and detail. This develops the ability to separate concept preservation from genre adaptation.
Common Failure Modes
- Translating scientific names word by word.
- Changing formula formatting or charge.
- Altering a sample or accession number.
- Removing significant figures.
- Converting “not detected” into zero.
- Confusing specimen types.
- Translating gene symbols.
- Inventing target abbreviations.
- Changing rpm to relative centrifugal force without authorisation.
- Losing a minus sign in a temperature.
- Flattening positive and negative controls.
- Swapping specification and result columns.
- Modernising taxonomy without editorial approval.
- Changing a qualitative observation into an interpretation.
- Trusting AI output without a protected-token and data audit.
Frequently Asked Questions
Should scientific names be translated?
Scientific taxonomic names generally remain in their accepted scientific form. Common names can be translated where a recognised target-language equivalent exists.
Should chemical formulas be translated?
No. Formulas are notation and should remain exact. Translate the accompanying chemical name or explanation according to established scientific terminology.
Can units be converted?
Only under an explicit project policy. Preserve the source unit by default, and verify any conversion, uncertainty and significant figures carefully.
Why are sample IDs so important?
They connect the text to physical specimens, instruments, storage and results. A changed identifier can break the scientific chain of evidence.
What is the biggest risk in laboratory translation?
Changing structured scientific information—entity identity, value, unit, condition, sample, formula or result status—while producing otherwise fluent prose.
Can AI translate scientific records?
AI can assist with prose and terminology research, but formulas, identifiers, units, values, nomenclature and evidential status require independent scientific and linguistic verification.
Where This Article Sits in the Translation Architecture
This article owns the scientific-nomenclature-and-laboratory-record lane inside Master Art of Translation. It extends Academic Writing Without Changing the Claim into the physical record layer of scientific work: specimens, formulas, measurements and methods.
It also connects to Technical Instructions, Manuals and Procedures for operational steps and to Vocabulary Learning Hub for deep terminology knowledge.
The Principle to Keep
Scientific translation succeeds when the target describes the same scientific world. The organism is the same organism. The compound is the same compound. The sample is the same sample. The number has the same value and precision. The method runs under the same conditions. The evidence supports the same conclusion.
Translate the language around the data, not the data into something new. When identity, measurement and provenance survive intact, scientific meaning can cross languages safely.