How do you translate a geotechnical borehole log, Standard Penetration Test record or ground investigation report correctly? Keep the test interval, soil description, borehole identity, depth reference, sample recovery and recorded measurements intact. A dependable geotechnical translation does more than replace engineering terms with familiar words. It allows an engineer or student using the target-language report to understand which material was encountered, at what level, under which test conditions, and how much of the stated conclusion comes from observation rather than interpretation.
This guide explains the top ways to translate geotechnical site investigation records, borehole logs, SPT blow counts, soil and rock descriptions, groundwater observations, laboratory test sheets and engineering geology reports. It is intended for translators, engineers, construction students, infrastructure teams and international project coordinators. The practical goal is precise multilingual communication without altering the ground investigation evidence that later design, monitoring and review depend upon.
A geotechnical document can be dangerously easy to mistranslate because much of it looks repetitive: depth columns, symbols, abbreviations and numbers. Yet ‘below ground level’ and ‘above datum’ are different reference systems; a sample interval is not a borehole depth; and three SPT blow-count increments do not mean three independent N-values. You can make an entire report sound professional while changing the measured ground conditions.
The visible problem: translating the numbers is not enough
Consider a fictitious borehole BH-04 in which a field log records shallow fill over natural ground, a test at a certain interval and a note that groundwater was observed during drilling. An unsafe translation says the upper material is an engineered foundation fill, moves the water observation to the final stabilised groundwater level and changes a partial penetration record into an ordinary SPT N-value. Nothing in the target sentence may look awkward. Every critical distinction, however, has been erased.
An engineer reading the translated report might conclude that the soil has a different provenance, the ground is wetter at another level or the penetration resistance is more comparable than the field record permits. These are not literary disagreements. A wrong depth reference or number can change how a project team understands ground conditions. This is why high-stakes technical translation should always be compared with the original logs, issued testing procedures and an authorised geotechnical reviewer.
The same warning applies to apparently minor typography. A decimal point, minus sign, unit suffix, borehole suffix or soil-classification code may connect narrative text to drawings and databases. During layout, a tab can shift a row into the next depth interval. A translator who checks only spelling may miss the most consequential change in the document.
First principles: separate field observations, laboratory findings and engineering interpretation
A borehole log is a structured record of an investigation at a particular location. It generally identifies the borehole, drilling method, levels and depths, recovered samples, encountered materials, field tests and remarks. A soil description can include observed colour, particle character, plasticity, fabric, moisture condition and other descriptors according to the system used. A classification applies an explicit rule or standard; a design conclusion interprets the evidence for a specific engineering decision. Translation must not collapse these into a single assertion.
A field observation such as ‘brown sandy clay with gravel fragments’ differs from an engineering judgment such as ‘suitable founding stratum’. A translator may help clarify the former’s grammar, but cannot infer the latter. Equally, a laboratory result may use a specimen and sampling interval that differ from the depth at which a field test was made. All those associations must remain traceable in the target text.
A Standard Penetration Test, or SPT, ordinarily records separate penetration increments and blow counts, then reports a field N-value from the prescribed increments when the test is valid under its procedure. As a widely used example, the standard field N-value is the sum of the second and third 150 mm (6-inch) drive counts for a completed test. The first increment is a seating drive. A partial drive, refusal event or corrected resistance value needs its own notation and applicable method. A translator should follow the governing project standard and must not manufacture a conventional N from an irregular test.
Measurements need a datum. Depth below existing ground level, elevation above a project reference and length along a borehole are different quantities. The words ‘deep’, ‘high’ and ‘below’ do not carry the geometry by themselves. Translation should preserve the reference surface, units, sign convention and whether a value is observed, calculated or inferred.
Finally, the state of the investigation matters. The log might be a preliminary field record; a revised factual report; a laboratory appendix; or a design interpretation. Do not change ‘preliminary’ to ‘verified’ or turn ‘groundwater not established’ into a stable water-table measurement. Technical trust begins with honest status labels.
A repeatable method for accurate ground investigation translation
- Confirm the project and report version: site, contractor, investigation phase, borehole number, drawing revision and issue date.
- Identify the controlling reference: project specification, approved standard, log legend, soil and rock classification guide, and laboratory methods.
- Make a field inventory: location, elevation datum, borehole inclination, depth intervals, drilling technique, sample IDs, tests and groundwater notes.
- Preserve source numbers, units, decimal precision, intervals, symbols and original field codes before drafting prose.
- Separate observed material description from classification and engineering interpretation; do not infer bearing properties from adjectives.
- Map every test to its actual borehole, sample, depth and recorded drive sequence.
- Translate abbreviations with the original code retained where it functions as a recognised data key.
- Check all table rows, depth scales, hatching legends, notes, appendices and referenced drawings in the rendered target file.
- Have a qualified geotechnical reviewer validate field meaning and calculations; language review should then improve clarity without altering results.
- Record unresolved queries and issue a version-controlled translation with its source report reference.
This method is slower than opening a document and translating sentence by sentence, but it prevents large corrections later. If a translator cannot identify what a depth number is measured from, the responsible action is to query the reporting authority. An elegant translation with the wrong datum is less useful than a clearly marked unresolved question. If the source contains a contradiction between table and figure, preserve the evidence and request an authorised correction rather than editing one to match the other.
Treat every changed number as a separate intervention requiring verification. Even converting feet to metres can change reported precision and invite rounding issues. Where a client requests conversions, create a documented conversion table approved by the relevant engineer, leaving the source values recoverable. Translation and engineering calculation are related services, but they are not the same task.
Twenty-four worked examples: identify what a technical translation must preserve
These are fictional training scenarios rather than real soil investigations or design recommendations. Source fragments are annotated English glosses of imagined records; actual multilingual translation must be checked against the original file, project legend and governing test method. The examples demonstrate quality-control reasoning, not a certification of ground conditions.
Borehole identity
Hypothetical source entry: Project P-26; borehole BH-04A; log revision R2. Faulty target interpretation: The identifier is simplified to BH-04. Controlled target wording: Project P-26, borehole BH-04A, log revision R2.
The suffix may distinguish a re-drilled hole or a related investigation. Losing it can attach readings to the wrong location. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Compare borehole codes on the location plan and laboratory schedule. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Explain why two similar IDs should not be reconciled by an editor.
Depth below ground level
Hypothetical source entry: Material boundary at 3.20 m below existing ground level. Faulty target interpretation: The boundary is reported as 3.20 m above sea level. Controlled target wording: Boundary at a depth of 3.20 m below the recorded existing ground level.
Depth and elevation use different zero points. A target report must preserve the source’s reference plane. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Read the log’s datum legend and ground-level note. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Draft a target-language annotation for a depth column.
Ground elevation datum
Hypothetical source entry: Ground level RL 27.40 m relative to project datum. Faulty target interpretation: The 27.40 m value becomes borehole depth. Controlled target wording: Ground elevation RL 27.40 m in the stated project datum.
Reference level is not total drilled depth. A shared numeral does not prove shared physical meaning. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Compare RL labels with section drawings. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: List what information would be required for a datum conversion.
Stratum top and base
Hypothetical source entry: Layer begins 2.4 m and ends 5.1 m below ground. Faulty target interpretation: The target says the layer is 5.1 m thick. Controlled target wording: Layer recorded from 2.4 m to 5.1 m depth, thickness 2.7 m.
A depth to base is not a layer thickness. If thickness is introduced, it must be correctly derived and clearly marked. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Check interval endpoints and arithmetic independently. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Explain why the upper boundary must be subtracted from the lower.
Fill versus natural ground
Hypothetical source entry: Made ground over naturally deposited sand. Faulty target interpretation: The fill is translated as undisturbed native soil. Controlled target wording: Fill or made ground over natural sand, as described in the log.
Artificially placed and naturally deposited materials may require different investigation interpretations. Translation cannot erase provenance. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Keep the source classification basis and descriptive terms. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Write one sentence that distinguishes origin from grain size.
Sand versus sandy clay
Hypothetical source entry: Description: stiff sandy clay with occasional gravel. Faulty target interpretation: The main material is called loose sand. Controlled target wording: Stiff sandy clay containing occasional gravel particles.
The head noun identifies the reported dominant material; modifiers describe associated components. Changing their grammatical hierarchy changes the soil description. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Check the project’s classification legend. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Explain how ‘sandy clay’ differs from ‘clayey sand’ in source terminology.
Consistency versus density
Hypothetical source entry: Soil described as very stiff clay under local method. Faulty target interpretation: The term is replaced by very dense clay. Controlled target wording: Very stiff clay, using the report’s consistency terminology.
Consistency and relative density descriptors are not universally interchangeable; they may apply in different contexts or methods. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Refer to the reporting scheme before substituting adjectives. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Make a glossary entry including the applicable soil category.
Groundwater observation
Hypothetical source entry: Water encountered at 4.1 m during drilling; later level not measured. Faulty target interpretation: The report gives a stable water table at 4.1 m. Controlled target wording: Water was observed at 4.1 m during drilling; no subsequent stabilised level is reported.
A drilling observation can be affected by operation and timing and is not automatically a long-term groundwater level. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Preserve observation time and conditions. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Write a query asking whether monitoring observations exist.
Water strike versus piezometer reading
Hypothetical source entry: Water strike noted at 6.0 m; standpipe reading 3.8 m later. Faulty target interpretation: The two readings are merged as one level. Controlled target wording: A water strike was recorded at 6.0 m during drilling; a later standpipe reading was 3.8 m.
Observations from different methods and dates may differ legitimately. Their provenance is crucial. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Keep the method and time beside each figure. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Construct a two-row table rather than forcing a single number.
SPT seating increment
Hypothetical source entry: Field SPT blows 4/7/9 for the three 150 mm drives. Faulty target interpretation: The N-value is reported as 4+7+9=20. Controlled target wording: SPT field sequence 4/7/9; uncorrected N-value 16 for the completed test.
The first 150 mm is the seating drive in the cited standard method; N sums the second and third increments. Changing the computation alters the reported resistance. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Check the specification’s accepted method and original field notes. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Explain why the first count is still recorded but excluded from N.
SPT corrected value
Hypothetical source entry: Uncorrected N 18; corrected value not issued. Faulty target interpretation: The translation reports N60 = 18. Controlled target wording: Uncorrected field SPT N-value 18; no corrected value reported.
An energy-corrected value requires specified apparatus and correction inputs. The translator must not label raw N as N60. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Keep distinct columns for field and corrected values. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Draft a note that avoids implying corrections were completed.
Partial penetration
Hypothetical source entry: Fifty blows for 90 mm during a test increment. Faulty target interpretation: The value becomes a completed increment N of 50. Controlled target wording: Fifty blows produced 90 mm penetration in the stated increment; test condition as recorded.
Partial penetration cannot be represented as a conventional completed 150 mm increment without explanation. Refusal conventions depend on procedure and project specification. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Preserve actual penetration and blow count. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Explain why ’50/90 mm’ should not be silently rounded.
Sample recovery percentage
Hypothetical source entry: Recovered core length 0.72 m from a 1.00 m run. Faulty target interpretation: The report calls sample recovery 100%. Controlled target wording: Recovered length 0.72 m from a 1.00 m run; recovery 72%.
Recovery describes material retained over an investigation run, not merely whether a sample container exists. The source lengths enable review. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Check recorded run length, recovery and units. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Write a caption distinguishing recovered material from drilled interval.
Undisturbed sample designation
Hypothetical source entry: Tube sample U-04 submitted; sample disturbance to be assessed. Faulty target interpretation: The target claims the specimen is fully undisturbed. Controlled target wording: Tube sample U-04, labelled according to the project’s sampling procedure; disturbance assessment pending.
A sampling method name does not guarantee the physical condition of every recovered specimen. Do not convert a designation into a quality claim. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Retain source method and laboratory assessment status. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: State what evidence would be needed to describe sample quality.
Rock quality designation
Hypothetical source entry: RQD calculated for the specified core run. Faulty target interpretation: The acronym is translated as recovery percentage. Controlled target wording: RQD, or rock quality designation, reported for the specified core run.
RQD and total core recovery are different indicators. A similar percentage format does not make the measures equivalent. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Check definitions in the project’s core logging standard. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Provide separate headings for recovery and RQD.
Core run direction and depth
Hypothetical source entry: Core run 8.0–9.5 m with fracture observations. Faulty target interpretation: Fractures are assigned to the 9.5–11 m interval. Controlled target wording: Fracture observations from the 8.0–9.5 m core run.
Depth intervals determine which material observations belong together. A shifted row is a substantive error. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Compare log blocks against the graphical depth scale. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Recreate the interval using a simple numbered vertical column.
Soil colour and moisture
Hypothetical source entry: Dark grey slightly moist silty sand. Faulty target interpretation: The target describes dry black silt. Controlled target wording: Dark grey silty sand described as slightly moist.
Colour, particle description and moisture state are independent observed fields. Translating ‘silty’ as the head material changes the reported soil. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Preserve the hierarchy and moisture qualifier. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Explain why ‘slightly’ cannot be removed for style.
Laboratory water content
Hypothetical source entry: Water content 22% for sample T-06. Faulty target interpretation: The value is turned into the site’s humidity. Controlled target wording: Laboratory water content 22% for sample T-06, as reported.
A property measured on a soil specimen is not an atmospheric reading. The test method and specimen identity are part of the claim. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Find the sample schedule and method reference. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Write a glossary note separating humidity from soil water content.
Atterberg limits
Hypothetical source entry: Liquid limit and plastic limit separately listed in laboratory results. Faulty target interpretation: Both limits are merged into one plasticity score. Controlled target wording: Liquid limit and plastic limit retained as distinct laboratory properties.
These terms refer to different properties; a derived plasticity index would require the stated calculation. Do not invent it by shortening two columns. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Check each test result against sample ID and unit. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Describe when a derived index may be shown as a separate value.
Sieve distribution and grading
Hypothetical source entry: Passing percentage at each sieve size tabulated. Faulty target interpretation: The table is called percentage retained without recalculation. Controlled target wording: Particle-size distribution reported as percentages passing the listed sieves.
Passing and retained are complementary but not equivalent data columns. Rewording a header without transforming the values reverses interpretation. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Compare table heading, legends and plotted curves. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Calculate one hypothetical retained percentage solely as a check.
Borehole termination reason
Hypothetical source entry: Investigation terminated at planned depth; deeper ground not tested. Faulty target interpretation: The translated report claims bedrock encountered at termination. Controlled target wording: Drilling ended at the planned termination depth; underlying materials were not investigated.
An administrative stopping point is not proof of a geological boundary. This error can create false subsurface conclusions. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Retain termination reason and investigated extent. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Ask what additional evidence would be needed before stating bedrock depth.
Refusal versus hard layer
Hypothetical source entry: SPT test stopped under prescribed refusal condition. Faulty target interpretation: The whole borehole is described as stopped at bedrock. Controlled target wording: SPT ended under its documented refusal condition; borehole status reported separately.
Penetration refusal can occur for different reasons and does not automatically identify rock. A test stopping rule and a geological interpretation must be separated. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Check the actual drive, depth and material description. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Write a neutral report sentence without inferring a rock boundary.
Interpreted founding level
Hypothetical source entry: Engineering note: candidate founding level subject to further assessment. Faulty target interpretation: The target says the foundation design is approved. Controlled target wording: A possible founding level was identified for further engineering review.
An engineering proposal is not an issued design decision. Qualifications, assumptions and responsibility should survive translation. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Identify report status and sign-off authority. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Explain which later document might confirm the design decision.
Document revision and scope
Hypothetical source entry: Factual report R1 superseded by R2 with corrected logs. Faulty target interpretation: The translation cites R1 figures as current without note. Controlled target wording: Report revision R2 supersedes R1 for the stated corrected log sections.
Engineering information often circulates through controlled revisions. Translating old numbers without a status notice misleads future readers. This kind of error may survive ordinary proofreading because every word looks familiar to a non-specialist. The translation must be evaluated against the original field, test standard and log legend. Rewriting for clarity is valuable only when the reader still receives the same measured or qualified meaning, down to the correct borehole and interval.
Track source revision against every table and drawing. A practical reviewer also checks the rendered table and its associated plot: a misplaced line break or shifted column can alter the data relationship even when the translated words are individually correct. Record discrepancies as queries with the source location and expected value. Exercise: Prepare a change note showing which translation sections need revalidation.
A deeper lesson: the SPT example shows why technical literacy matters
Suppose an SPT field record shows blow counts 5, 8 and 10 for three successive completed 150 mm drives under the relevant method. The first five blows are the seating increment. The uncorrected field N-value is 8 + 10 = 18, not 23. If a translator treats every listed blow count as an independent test or adds all three values into one resistance, the output can look numerically precise while being technically false. The original sequence should remain available even when the report also states N.
Now consider a partial drive: a record states 50 blows for 90 mm penetration. It cannot simply be called N = 50 unless the reporting procedure expressly defines that representation for the situation. Preserve the actual resistance observation and the stopping condition. A qualified reviewer should decide whether any comparison with other tests is appropriate. The translator should not hide an irregular test inside a conventional-looking number.
This illustrates a broader rule: when technical terms have arithmetic consequences, translate the procedure as well as its label. A glossary entry for ‘standard penetration resistance’ should say what increments contribute under the project standard and what corrections or exceptions are reported separately. The glossary becomes a reasoning aid, not merely a dictionary.
Visual review: why a borehole log is more than a column of prose
A borehole log typically combines a vertical depth scale, coloured or hatched material descriptions, sampling markers, field test results and notes. The page layout itself contains relationships. Translation can increase word length, move notes outside their cells and break the alignment between a soil description and its depth interval. Therefore a target-language log should be checked at final page size by someone who can compare it side by side with the source.
Check wrapped headings carefully. A split caption might make ‘groundwater observation’ look like a permanent water-table measurement. A displaced number may appear to belong to the layer above it. The use of translated legends must remain unambiguous in black and white and in accessible digital copies. If a report uses a graphical symbol with a formal definition, translate the legend or add an authorised explanation without editing the underlying test classification.
Digital exports have their own risks. Spreadsheet decimal separators, Unicode symbols, negative signs and superscripts can change under language or locale settings. A comma may be a decimal separator in one context and a thousands separator in another. Preserve numeric data in machine-readable fields according to the agreed import specification, and make the presentation conventions explicit rather than allowing software to guess.
Quality control by field type
Identity pass: compare project number, borehole ID, sample ID, test ID, laboratory reference, drawing number and revision. Sort repeated references and check all occurrences. It is easy for a single character to distinguish two boreholes, so automated matching can help, but the list should still be reviewed against the source.
Measurement pass: examine depth ranges, elevations, units, blow counts, recovery lengths, percentages, water levels, material descriptions and values preceded by less-than signs. Do not convert measurements without an approved procedure. Mark any number that appears inconsistent; leave correction to the responsible author.
Meaning pass: check whether a statement records observation, classification, interpretation, proposed treatment or final approval. Look for negation, provisional language, sample quality, termination reason and data gaps. An honest unknown is more valuable than an invented certainty.
Presentation pass: review both the prose and the sheet: row heights, symbols, table alignment, linked drawing legends, headings and searchable text. Ask a reviewer to reconstruct a particular test and its associated sample solely from the target-language document. The practical success criterion is retrievability, not how effortlessly the document reads aloud.
A transferable practice workshop
Provide students with a fictional one-page borehole log, a location plan and a small laboratory table. Ask them to highlight immutable identifiers and measurement fields first. Next, they draft a reader-friendly translation of the explanatory notes. Finally, they create an audit trail containing each depth reference, test identifier and source field. No student may change a technical number without a separately labelled, authorised calculation.
Pair students for a reverse-lookup test. The second reader receives the translation and must identify the depth interval of a chosen sample, the original SPT blow sequence, the source groundwater observation and the report revision. If a detail cannot be recovered, classify the failure as identity, alignment, measurement, method, uncertainty or interpretation. Repair the category, not merely the sentence.
For transfer, give students an oceanographic CTD profile and an archaeological context register. The specific vocabularies change, but the invariant reasoning remains. Each field names a measurement or object, a position, a time or sequence and a method of recording. Good translators recognise that structure before selecting words. This is why technical literacy makes language work both more accurate and more efficient.
Frequently asked questions
Does SPT N include all three blow-count increments?
For a completed standard field test under the commonly used method, the first 150 mm drive is seating and N is the sum of the next two 150 mm blow counts. Always consult the governing specification for exceptions and correct reporting. The controlling project document and the responsible professional reviewer remain the final reference for engineering meaning.
Can depth below ground level be replaced by elevation?
Only through a separately checked conversion with the relevant surveyed ground level and datum. They are not interchangeable labels, and the translation should preserve the source representation. The controlling project document and the responsible professional reviewer remain the final reference for engineering meaning.
Is refusal proof of bedrock?
No. A test stopping condition does not independently establish geology. Record the observed penetration and stopping rule without adding an unsupported material identification. The controlling project document and the responsible professional reviewer remain the final reference for engineering meaning.
What is the difference between a field log and an interpreted report?
A field log records investigation conditions and observations. An interpreted report uses that and other evidence for engineering conclusions. Both are important, but a translation should preserve the distinct authority and certainty of each. The controlling project document and the responsible professional reviewer remain the final reference for engineering meaning.
Should soil descriptions be modernised into another classification standard?
Do not silently replace one scheme with another. Retain the source system and, where the project expressly requests mapping, document the basis, uncertainty and approval for any new classifications. The controlling project document and the responsible professional reviewer remain the final reference for engineering meaning.
What if borehole data appear inconsistent across drawings?
Record the discrepancy with report revisions and field identifiers. Seek clarification from the issuing engineer or data owner instead of choosing the value that seems more plausible. The controlling project document and the responsible professional reviewer remain the final reference for engineering meaning.
Can machine translation handle geotechnical reports?
It can help draft descriptive prose and flag inconsistent terminology, but it cannot certify test interpretations, units, depth references or design conclusions. Engineering review and final-layout checks remain essential. The controlling project document and the responsible professional reviewer remain the final reference for engineering meaning.
How do I make a translated borehole log accessible?
Use clear headings, explicit legends, labelled units and accessible digital text while preserving the original field positions and values. Test whether a reader can retrieve a particular sample, test and depth interval. The controlling project document and the responsible professional reviewer remain the final reference for engineering meaning.
Useful primary references
The Texas Department of Transportation field testing and sampling manual describes the individual SPT drive increments and uncorrected N-value under its specified procedure. The US Geological Survey GeoLog Locator illustrates the value of documented borehole identifiers, formats and retrieval metadata. These are educational references, not substitutes for a project’s issued engineering standards or site-specific technical review.
Continue into related translation and language routes
For general meaning preservation, read Translate Meaning, Not Just Words. For controlled definitions, use Terminology, Glossaries and Quality Checks. The Technical Instructions, Manuals and Procedures guide covers wider procedural translation without duplicating this article’s borehole and field-test emphasis.
Compare the oceanographic cruise and CTD guide for another system where measurement position and method govern meaning. Browse the series index, Vocabulary Learning Hub and How English Works: Translation and Meaning Across Languages for broader transfer.
The takeaway
A faithful geotechnical translation does not merely sound like engineering. It preserves the source investigation as an auditable body of observations. A reader should be able to find the original hole, interval, material, test, sample, datum, units, method and uncertainty behind every target-language statement. That discipline gives multilingual project teams a common basis for understanding the same ground rather than accidentally inventing different ground through different words.