How do you translate a handweaving draft correctly when the same sheet combines a threading sequence, a tie-up diagram, treadling, and a drawdown? Translating weaving instructions into English—or out of English—requires more than replacing textile vocabulary. Every number and symbol tells a weaver which warp end, shaft, treadle, or weft pick belongs where. One misplaced instruction can produce a completely different cloth.
This guide explains how to translate weaving drafts, loom threading charts, shaft numbering, treadle tie-ups, lift plans, rising-shed and sinking-shed conventions, weave structures, and WIF-related notes. These are the natural searches made by bilingual handweavers, teachers, conservators, technical editors, textile students, and pattern publishers trying to keep a woven design reproducible across languages.
The central rule is simple but demanding: preserve the behaviour of the cloth-making instructions, not just the visual resemblance of the sentences. A reader must be able to recover the original threading, tie-up, weaving order, fabric face and repeat boundaries from the translation. A translator who is unsure which convention applies should identify the uncertainty, not guess a convention from an attractive illustration.
This article handles loom-draft document translation, rather than sewing-machine operation, cross-stitch design, or restoration of historic textiles. All example instructions below are teaching examples, not manufacturer-specific operating procedures. Your loom type, the draft key and the designer’s instructions take priority over any demonstration numbers used here.
The distinction matters because an ordinary paragraph can be paraphrased while a technical weaving draft functions like a small program. The source expresses a sequence of decisions. A reliable translation is one whose decisions produce the same weaving structure, with the same qualifications and without silently changing a sinking-shed draft into a rising-shed draft.
The visible problem: a correct sentence can produce incorrect cloth
Imagine a reader faithfully following a translated six-end point-twill sequence. The words look familiar and each shaft number exists on the loom. Yet the translated edition omits the statement that the first pick is drawn at the bottom of the page, then supplies a rising-shed key for a sinking-shed source. The weaver may spend hours preparing a warp before discovering that the structure differs from the source. The failure is not careless vocabulary. It is loss of the operating assumptions under which the entire document works.
Translation quality therefore has two independent tests: can a language reader understand the explanation, and can a competent weaver derive the same structural outcome? An attractive translation can pass the first and fail the second. The rest of this guide makes that second test explicit, measurable and repeatable.
First principles: what each piece of a weaving draft actually controls
A draft is a coordinated system, not four unrelated pictures
A loom draft normally relates the threading of individual warp ends to shaft identities; the tie-up or lift plan relates controls to shaft movement; the treadling relates successive weft picks to controls; and the drawdown predicts the woven interlacement. A translator can reproduce the four names correctly yet disconnect one relationship. Ask whether the reader could reconstruct a small section of the cloth after reading the translated key. If the answer is uncertain, the translation is not complete.
Threading order is sequential evidence
A sequence such as 1–2–3–4–3–2 describes a journey through shaft positions, not six descriptive labels. Repeating the first or last shaft at a turning point changes the pattern. The beginning edge matters as much as the numerical set. Translation notes should identify whether the designer starts at the right-hand or left-hand end and whether any chart has been mirrored for page layout. A fluent language conversion cannot compensate for reversing the warp-end order.
A shaft number is an identifier, not a measurement
Shaft 1 is not always the same physical location on every loom. When a source uses a loom-specific front-to-back numbering convention, translate the convention explicitly. Do not replace the numeral with a presumed near or far position without checking the equipment. The most robust multilingual edition leaves the numeric identities unchanged and explains the orientation alongside them. A weaver should not have to infer which shaft is meant from a regional phrase.
Treadle labels and lever labels are different controls
A treadle is a foot-operated control in many floor-loom arrangements, while a table loom may use hand levers to obtain a shed. Some translations use one broad word for both. That can be acceptable only when the instruction is abstract and does not describe the physical action. If a diagram says ‘press treadle 3’, do not tell a table-loom user to press a nonexistent pedal; explain the pattern’s control mapping and direct the user to the actual loom manual.
Rising and sinking sheds require explicit treatment
An illustrated tie-up may mark shafts raised by the selected treadle in one system and shafts lowered in another. A weaver can weave a structurally related but inverted face if a translation silently assumes the wrong convention. The editor should extract the precise key: does a filled cell represent lift, sink, attachment, or a software display setting? Only after the key is identified should the translator choose terms such as ‘raise’, ‘lower’, ‘active’ or ‘connected’.
Treadling order has a start and a reading direction
Many North American published drafts place the first pick near the top of the treadling column. Other presentations use a bottom-origin system, particularly when the threading is displayed below the cloth. The Handweaving Academy describes these variants and their effect on visual orientation. Thus ‘read downward’ is not a universal assumption. A translated caption must preserve the source’s stated start marker and reading direction, not import one from the translator’s favourite publication style.
The drawdown is a testable prediction
A drawdown is not an optional decoration. It maps the warp-over and weft-over intersections implied by threading and shedding. If a translated tie-up label changes which shafts rise, a small hand-computed drawdown can reveal the inconsistency. One does not need to reproduce a full textile to test a few rows. The editor can compare an explicit four-end, four-pick sample before approving hundreds of repeated lines.
The face of the fabric is not always the front of the page
Terms such as face, reverse, obverse, right side and working side may refer to different viewing conventions. A photograph of the woven sample may show the opposite face from a loom-side diagram. Translating ‘front’ as if it meant the weaver’s physical position can invert instructions for a reversible fabric. Record whether the source is referring to a cloth surface, a loom direction, or the printed orientation of the draft.
Repeat brackets are structural, not ornamental
A notation like ‘(1 2 3 4) × 6’ describes six cycles of four threading positions, not an invitation to multiply shaft numbers by six. Border ends may lie outside the bracket. An added phrase such as ‘repeat to the edge’ can be false where a separate selvedge is specified. The translator should identify start index, end index, repeat count and exceptions, then verify the total warp-end calculation against the source.
Sett, width and counts are separate quantities
Ends per inch or centimetre, weaving width, total warp ends, and finished width after shrinkage cannot be used interchangeably. An apparently harmless change from ‘reed width’ to ‘final width’ makes a pattern misleading. Preserve which stage of production the measurement describes. Convert units only if the source licenses a conversion and the rounding cannot change the intended end count. When converting, keep the original figure for audit.
Float length describes construction, not an artistic preference
A float is the distance for which a warp or weft thread passes over multiple opposing threads without interlacement. Translations sometimes turn ‘maximum float’ into a decorative statement about shine or surface. That loses a practical property relevant to snagging and handling. If the source reports a float count, preserve the count and whether it describes warp or weft. A change in draft convention can alter that count even when a picture still looks plausible.
A historical transcription and a modern working draft have different jobs
An archive may preserve the orthography, omissions and conventions of an old weaving manual. A modernised draft may reinterpret it for present-day looms. Do not blend these two products invisibly. Label any editorial reconstruction separately from the faithful transcription; keep a source-image reference, line number or manuscript folio where possible. A reader should know which part is written evidence and which part is a contemporary explanation.
Colour order is independent of weave structure
Warp colour sequences and weft colour orders can be translated with correct colour words while the pattern repeat is still lost. The draft may use A and B as yarn identifiers, not literal colours. Keep those IDs stable across the materials list, colour drawdown and yarn notes. If the translator substitutes colour names for arbitrary IDs, readers may lose the ability to substitute yarn while preserving the intended structure.
Uncertainty must stay visible
Old drafts may have unreadable squares, missing treadle keys, or notes that use a local term with multiple possible meanings. A translator’s desire for fluency can transform ‘possibly’ into certainty. Maintain ‘unclear in source’ or ‘editorial reconstruction pending review’ rather than silently choosing an interpretation. Practical reproduction can proceed once a qualified reviewer has reconciled the gap against a scan, loom model, other pages and a tested drawdown.
A repeatable ten-stage weaving-draft translation workflow
1. Identify the document and its intended loom
Record the publication, designer, language, edition, loom type, number of shafts, number of treadles and whether the draft is a historical transcription or an instructional adaptation. This establishes the legitimate scope of change. If the document refers to a jack loom, a table loom, or a countermarch installation, the translator cannot treat a mechanism-specific phrase as universally applicable. Record source page and version for every ambiguous chart.
2. Make an invariant ledger
Before drafting target-language prose, list every quantity and ID that must survive: shaft numbers, treadle numbers, repeat counts, width stages, warp-end totals, picks per repeat, colour codes, directional arrows and the notation key. Compare the ledger after translation. This resembles checking variables in a mathematical proof: if a value changes, the editor must explain why. A corrected typo belongs in a signed editorial note, not an invisible translation change.
3. Extract the chart’s reading convention
Check the location of the threading, tie-up, treadling and drawdown. Locate any arrow or numbered first end. Determine whether the chart moves left-to-right, right-to-left, top-to-bottom or bottom-to-top. If the image lacks a key, record a question. Layout software and historical period influence convention, so a general textbook habit does not count as sufficient evidence for this specific source.
4. Create a controlled vocabulary sheet
Include warp, weft, end, pick, heddle, shaft, reed, dent, sleying, treadle, tie-up, liftplan, sinking shed, rising shed, drawdown, float, selvedge and repeat. Distinguish near synonyms that can direct different operations. Provide one preferred term per target-language function and record acceptable regional variants. Do not use the vocabulary sheet to override a source’s physical convention; it is a consistency tool, not a source of new instructions.
5. Translate constraints as constraints
Words such as ‘only’, ‘except’, ‘start’, ‘until’, ‘alternating’, ‘every second pick’, ‘do not repeat’ and ‘if using this loom’ are high-risk. Colour these terms in the editing copy and check them line by line. In particular, preserve conditional wording: a tie-up that applies to a rising-shed loom should not become a general command for all looms. The output must not turn a scenario into a universal rule.
6. Test a miniature cloth
Construct a four-end or eight-end example using the translated threading, tie-up and treadling. Compare the predicted drawdown with the source’s same intersections, noting which face is being displayed. An error in one repeated unit will be magnified by a full-width project. If the source deliberately leaves a border outside the repeat, test that junction separately. A tiny manual test is cheaper than a large corrected warp.
7. Reconcile totals and units
Calculate the number of warp ends implied by the repeats and edge ends. Check sett separately from widths and check whether centimetres or inches have been rounded. Keep the unit symbol close to its quantity, especially in bilingual tables. Do not ‘correct’ the total end count by shifting a repeat boundary. When figures disagree, flag the inconsistency and ask the source owner which figure controls.
8. Preserve machine-readable codes
If the draft is supplied as a WIF or other editable file, keep structural keywords, numeric fields and identifiers intact unless the format specifically permits localization. Human-facing comments can be translated, but machine keys should not be guessed at. A decorative translated label is not useful if the software can no longer parse the file. Maintain original file and revision alongside the translated document.
9. Use a bilingual expert review
A language editor should review readability and idiomatic accuracy, while an experienced weaver or technical reviewer should check the physical consequences. Ask both to mark exact passage and reason for correction. A translator may know a language but not the loom; a weaver may know a loom but not the nuance of a conditional sentence. Their responsibilities are complementary rather than interchangeable.
10. Release a change log and reader test
Publish the source edition, version date, scope, terminology note, and list of resolved ambiguities. Invite a test reader to explain back the first four picks without seeing the source text. Their explanation should agree with the source diagram. Record any correction as a versioned erratum so that a printed chart and a website copy do not silently diverge.
Twenty worked translation cases: source, tempting error, repair and proof
Case 1: Threading: point reversal
Source situation. A draft lists 1–2–3–4–3–2, then starts the next repeat at 1.
Tempting mistake. A fluent translation calls it ‘straight threading, one to four and back’ without the sequence. A student includes the peak shaft twice, producing an extra stripe.
Safer translation. Preserve the sequence and the repeat boundary explicitly: ‘Thread shafts 1, 2, 3, 4, 3, 2, then begin the next six-end repeat.’
Verification. Count six ends per repeat and examine the turning point. Treat the textual description as an aid to the diagram, never a substitute for its exact structure.
Case 2: Threading: direction from the edge
Source situation. The scanned caption says ‘from the right selvedge’.
Tempting mistake. The editor assumes left-to-right reading because the page is printed in a left-to-right language.
Safer translation. Keep ‘starting at the right selvedge’ and mark the first end on the translated chart. Distinguish the page’s text direction from the loom’s physical starting point.
Verification. Ask a weaver to number the first eight ends using the translated layout. Their sequence must match the source, including the first border end.
Case 3: Shaft numbered at the front
Source situation. The manual defines shaft 1 as nearest the weaver.
Tempting mistake. A regional translator rewrites ‘front shaft’ without stating how numbers map, although the target loom may number the other way.
Safer translation. Retain ‘shaft 1’ and add the source’s orientation note beside the chart. Do not redefine shaft numbering for another loom as if that were translation.
Verification. Check every subsequent ‘front’ or ‘rear’ reference against the same orientation note. A stable glossary prevents a hidden reversal later.
Case 4: Rising-shed tie-up
Source situation. A solid cell indicates a shaft that rises when a treadle is pressed.
Tempting mistake. The target caption says ‘the shaded shafts go down’, reversing the convention.
Safer translation. Translate the legend as ‘filled square = shaft lifted by this treadle’ and label the diagram ‘rising shed’ if the source does.
Verification. Compute one drawdown row to see whether the translated legend yields the same warp-on-top squares as the source chart.
Case 5: Sinking-shed variant
Source situation. Another publication marks the shafts that descend.
Tempting mistake. The translator copies a rising-shed key from a previous article into the new booklet.
Safer translation. Quote the specific source key: ‘filled cell indicates a shaft lowered in this draft’. Explain that a different source may use the opposite convention.
Verification. Cross-reference the original graphic and any manufacturer description before claiming physical equivalence of tie-ups.
Case 6: Treadling from the bottom
Source situation. An old layout places threading underneath the drawdown and the first pick at the bottom.
Tempting mistake. The translation’s generic introduction says ‘always start at the top’.
Safer translation. Replace the unsupported universal command with the source-specific instruction ‘start at the numbered bottom row and work upwards’.
Verification. Follow the first four weft picks and compare motif orientation. The Handweaving Academy documents why draft layout conventions matter.
Case 7: Sleying: dent versus end
Source situation. A note specifies two warp ends per reed dent.
Tempting mistake. The target language conflates each end with each dent, halving or doubling the sleying count.
Safer translation. Write ‘place two warp ends through each specified reed dent’ with a glossary distinction between end and dent.
Verification. Compare total ends and number of occupied dents; do not use a guessed reed conversion as evidence that the source is wrong.
Case 8: Warp ends versus yarn length
Source situation. A table has 240 ends and a warp length of 3 m.
Tempting mistake. The translator describes 240 metres of warp, changing a count into a length.
Safer translation. Keep count and length in separate table columns with units attached: ‘240 warp ends’ and ‘3 m warp length per end’.
Verification. Have the reviewer read the purchase calculation aloud. Quantity type must remain apparent even without the original table headings.
Case 9: Weft pick count
Source situation. The designer instructs 16 picks per motif, not 16 motifs.
Tempting mistake. An editor turns a pick into a repeated pattern block.
Safer translation. Use ‘weave sixteen successive weft picks to complete one motif unit’, then identify the treadling cycle that covers them.
Verification. Check a counted drawdown row by row and do not infer that one pick equals one treadle repeat.
Case 10: Alternate every second pick
Source situation. A note says use colour B on every second pick.
Tempting mistake. The translation says ‘repeat colour B twice’ because the phrase ‘second’ is read as a quantity.
Safer translation. Translate the alternation: ‘use colour B on picks 2, 4, 6 … of this sequence’ when consistent with the source.
Verification. Verify the first eight picks, observing whether the source begins numbering at one or uses an initial setup pick.
Case 11: Colour code A and B
Source situation. The materials chart calls yarns A and B while sample colours are blue and white.
Tempting mistake. The translator replaces all A and B labels with the photographed colours.
Safer translation. Keep yarn IDs as A and B; describe the example colours separately as optional palette choices, not structural identifiers.
Verification. Test a substitute palette. The same threading and colour-order instructions must remain intelligible without blue or white yarn.
Case 12: Selvedge outside repeat
Source situation. A two-end border is specified before an eight-end repeat.
Tempting mistake. The target line ‘repeat eight ends across the whole width’ absorbs the border into the interior count.
Safer translation. Write ‘work the two selvedge ends once; repeat the eight-end interior sequence as stated’ and preserve the other edge details.
Verification. Recompute the total ends. The border should appear once at the appropriate edge, not at every repeated unit.
Case 13: Float annotation
Source situation. The source warns that a certain variant creates a seven-end float.
Tempting mistake. A cheerful description turns this into ‘seven decorative threads’ and omits the performance warning.
Safer translation. Translate ‘seven-end warp float in this variant’ and preserve the caution or usage limitation the source gives.
Verification. Check whether the affected row belongs to the warp or weft. A reviewer should inspect structure rather than a photograph alone.
Case 14: Cloth face and reverse
Source situation. A photo shows the underside while the drawdown marks the face.
Tempting mistake. The translation calls both images ‘front view’.
Safer translation. Use explicit captions ‘face as represented by drawdown’ and ‘reverse-side photograph’ where supported.
Verification. Compare the pattern’s over-under points. Face references can change when a sample is turned over without any threading change.
Case 15: Shrinkage and finished width
Source situation. The document gives a weaving width and a finished measurement after wet finishing.
Tempting mistake. The translator merges them into a single ‘final loom width’.
Safer translation. Preserve ‘on-loom width’ and ‘after-finishing width’ with the corresponding values and method qualification.
Verification. Check whether the numeric difference is a measured sample result or a promised outcome. Do not generalise one material’s shrinkage.
Case 16: Historical damaged cell
Source situation. A manuscript’s tie-up includes one illegible square.
Tempting mistake. A digital edition fills it in and presents the result as a transcription.
Safer translation. Keep a transcription note ‘one tie-up cell illegible’ and distinguish any suggested modern reconstruction.
Verification. The reviewer should cite the scan location and compare adjacent repeats. A reconstruction is an editorial hypothesis, not source certainty.
Case 17: Lift plan and treadle substitution
Source situation. A weaver exports a lift plan from a computer-assisted loom file.
Tempting mistake. A translation says ‘press treadle 2’ although the file specifies a shaft-lift set, not a pedal.
Safer translation. Translate the plan as the selected shafts raised or lowered according to its key. Only adapt to treadles in a labelled separate instructional version.
Verification. A test should compare shed state pick by pick, because multiple mechanical controls can yield the same abstract interlacement.
Case 18: Plain weave as a structural description
Source situation. A draft shows alternate warp-over and weft-over cells.
Tempting mistake. A translation calls any alternating colour stripe ‘plain weave’.
Safer translation. Describe plain weave by interlacement, not by colour arrangement. Preserve separate stripe and yarn-colour instructions.
Verification. Check the drawdown’s checkerboard alternation independent of the colour palette and independent of a photograph’s lighting.
Case 19: Twill direction
Source situation. A 2/2 twill shows a diagonal in a labelled orientation.
Tempting mistake. The translated diagram is mirrored during layout but its caption still says the original diagonal direction.
Safer translation. Retain chart orientation or explicitly re-label the mirrored drawing; do not assume diagonal directions survive reflection.
Verification. Compare one repeat and the viewing face. A full visual audit can detect a change that text-only proofreading misses.
Case 20: Versioned corrected erratum
Source situation. The designer issues a corrected tie-up for edition two.
Tempting mistake. The translator combines first-edition notes with a second-edition diagram.
Safer translation. Identify the current edition and state which tie-up belongs to each revision. Add a clear correction note to any translation already released.
Verification. A user should be able to identify the correct file and date without silently mixing incompatible versions.
Failure modes that survive ordinary proofreading
Treating a schematic as ordinary illustration
A chart conveys operations. Reflowing its cells, altering row order or mirroring for typesetting can change the instructions even if the translated prose remains grammatical. Keep the chart and its legend under the same review owner. Ask the layout editor to confirm that the sequence of cells and their relationship to nearby labels is stable after export to PDF, web or print.
Repairing a source with unmarked assumptions
When an old draft seems wrong, translators may silently ‘improve’ it. The change can destroy historical evidence or conceal an editorial decision. Retain a faithful rendering and put any suggested repair in an explicitly labelled note. The source author, archive or competent textile reviewer should decide whether the correction is justified. Leave the original accessible for comparison.
Mixing units and measurements
Ends per inch is a density, while warp width and finished width are dimensions. If a text says twenty ends per inch across twelve inches, the numbers have a different job from a general yarn estimate or a count after edge treatment. Do not convert and round a density then recompute counts without marking the calculation. Readers need the original measurement frame as well as the translation.
Overgeneralising from software
A program’s default tie-up convention is not the rule for every weaving book. Digital tools may invert shed displays, transpose axes or use a colour scheme that has no relation to traditional ink marks. Include the software name, version and setting if they matter. A translator should not rewrite a historical key to agree with a contemporary editor’s display preference.
Making an accessible version without safeguarding the structure
Screen-reader descriptions and prose equivalents are valuable, but a sentence such as ‘several diagonal lines’ does not reproduce a threading or treadling list. Accessibility requires a structured textual representation of sequences, row labels and intersections. The editor can provide accessible tables or lists along with the image, then have a user verify the same first four picks.
Using literal verbs that imply the wrong mechanism
A term that literally means ‘push’ in one language may refer to moving a foot control, raising a lever or activating a shaft group. Choose verbs that match the source and clarify equipment specificity. Do not make a table loom behave as though it has floor treadles. Where a draft is mechanism-neutral, keep the wording at the abstract shed-operation level.
Losing the exception at a boundary
A short instruction such as ‘except at the border’ can carry more practical significance than a paragraph describing decorative texture. Verify qualifiers attached to selvedges, last picks, first-end setup, floating selvedges and finishing instructions. A translation that preserves 99% of the main repeat but loses one boundary exception may fail when a weaver attempts the full width.
Separating terminology control from outcome testing
An approved glossary is helpful but insufficient. A review can find every noun correct and still miss a reversed arrow or omitted condition. Pair terminology QA with a tiny drawdown test, numeric reconciliation and a reader demonstration. Maintain one issue log linking the source line, proposed wording, reason for change and final approval.
Four-week practice plan for translators and weaving students
Week 1: learn the representational grammar
Print a small plain-weave and twill draft from a trusted source. Identify threading, tie-up, treadling and drawdown without translating anything yet. Highlight the shaft numbering and reading-direction key. Explain to another person which chart part controls each decision, then write a concise bilingual glossary. Translate only captions and compare the first four warp ends and first four weft picks with the source. Record every term that seemed obvious but changed meaning after examination.
Week 2: reconstruct a compact sample
Choose a draft with no more than four shafts. Translate a defined eight-end and eight-pick section, then reconstruct the result from the target-language instructions alone. Compare both diagrams and mark the first difference, if any. Repeat with an instruction that contains an exception, such as a fixed border. Avoid judging translation quality from whether the cloth looks attractive; judge it from agreement with the source’s stated structure.
Week 3: practise competing conventions
Use two documented drafts that apply different reading directions or rising-versus-sinking conventions. Build an invariant ledger for each before drafting text. Then have a second editor explain the convention without seeing the original caption. Resolve any error through the source key rather than assumptions. This exercise builds habits that transfer to maps, scientific diagrams, music notation and other texts whose graphics carry operations.
Week 4: produce a publishable handoff
Deliver one short bilingual edition with the source citation, date, version, full symbol key, numeric ledger, translated instructions, visual equivalence check and a one-page errata record. Have a weaver follow a small sample or reason through its interlacement. Include a plain-language FAQ for non-specialists but keep every warning and operational condition. The project is finished only when another reader can use the translation without private explanations.
Frequently asked questions
Can I translate the words and leave the draft image unchanged?
Sometimes, but only if the image’s labels, legends, arrows and abbreviation key are comprehensible to target-language readers. A chart can contain its own language. Translate or explain those parts without altering the underlying rows and columns. Test the result with someone who has not seen the source.
What is the difference between threading and treadling?
Threading specifies how warp ends are assigned to shafts. Treadling specifies the sequence of controls used for successive weft picks. Mixing them changes which part of the loom a sentence describes. A translation should use distinct terms and preserve the chart’s physical relationship.
Should I turn a sinking-shed draft into a rising-shed draft?
Not as an invisible act of translation. That is a technical adaptation. If needed, release a separate labelled conversion reviewed for the intended loom, while retaining the faithful source representation. Explain how the two presentations relate and what test confirmed the conversion.
Is a WIF file itself a translation?
A WIF file represents weaving-related data in a digital format. Converting human-facing explanations around it into another language is translation; changing structural keys or data to suit a different loom can be adaptation. Keep the original file, the transformed file and the reason for every structural change distinct.
Why do some drawings seem upside down?
Drafts can employ different positions for threading and different expected starting points for treadling. A visual inversion may be a convention rather than a mistake. Check source arrows and keys before rotating or mirroring artwork; compare the first few picks to see what the original intends.
Must every regional textile term receive one literal equivalent?
No. A controlled glossary should preserve function and distinguish local terminology. A direct dictionary equivalent may be misleading if it names a different part on another loom. Choose terms that support the same action and include a note where the physical mechanism differs.
Can I substitute centimetres for inches?
Yes when suitable, but state the conversion method and keep enough precision to preserve counts and usability. Density measurements and dimensions require different treatment. Never silently recompute a warp-end total from a rounded density and publish it as if the source supplied it.
What if the scanned draft contains a suspected error?
Report the discrepancy with page or folio reference. Leave the faithful transcription intact, and provide a separately marked possible correction after technical review. Readers should see the difference between historical evidence and a new reconstruction.
How should I translate a drawdown for a screen reader?
Represent shaft assignments, tie-ups, picks and selected warp-over intersections in ordered accessible text or tables. A general description of the picture’s appearance is not enough. Verify that a user can determine the first repeat without access to the graphic alone.
What proves that a translated draft is correct?
No single check is sufficient. Combine bilingual proofreading, numeric reconciliation, convention verification, a small drawdown comparison, and a target-reader test. The best evidence is that the translation yields the same specified interlacement under the same explicitly stated assumptions.
What belongs in a published translation note?
Include source and edition, translator or editor attribution where appropriate, equipment assumptions, glossary variants, unconverted codes, adjustments, unresolved ambiguity and correction date. This allows teachers, weavers and archivists to understand precisely which text can be trusted for what purpose.
What should a beginner learn next?
Learn the distinction between descriptions, instructions and evidence. A reader who can preserve that distinction in a loom chart will recognise the same problem in scientific figures, craft patterns, bilingual timetables and technical manuals. The vocabulary and English learning hubs below provide transferable language practice.
Further reading, neighbouring translation jobs and English-learning routes
For detailed demonstrations of draft orientation and the relation between treadling and drawdown, see Handweaving Academy’s discussion of upside-down motifs. To compare real historic patterns and practical draft files, explore Handweaving.net’s archive. These are source and specialist context, not substitutes for an individual designer’s specific chart legend.
Keep neighbouring canonical tasks separate. For grids and symbol-to-material identity, continue to translating cross-stitch charts. For machine controls and needle settings rather than shaft-based handweaving, use the sewing-machine instruction guide. For another craft in which piece numbers and geometric instructions matter, see translating stained-glass patterns.
For students building the terms needed to read technical craft instructions, use the eduKateSG Vocabulary Learning Hub. For the wider relationship between words, representation, interpretation and reliable instructions, explore How English Works. Clear language helps; the final test is whether the same system of decisions survives translation.
Final working checklist
- Source edition, author and draft key identified
- Threading sequence and start edge preserved
- Shaft numbering and physical orientation documented
- Tie-up or lift-plan convention confirmed
- Treadling direction and first pick tested
- Drawdown compared on a small repeat
- All border exceptions and colour IDs retained
- Units, densities and warp-end totals reconciled
- Any proposed technical conversion separated from faithful translation
- Bilingual and weaver reviews recorded
- Unresolved ambiguities and errata made visible
A weaving translation succeeds when another reader can reconstruct the same pattern, understand the same limitations and locate every uncertain choice. That is a stronger standard than smooth wording, and it is attainable through small, well-designed checks before a large loom project begins.