How do you translate ballot-marking instructions, spoiled-ballot replacement notices, assisted-voting procedures and secret-ballot warnings without changing what a voter is permitted to do? This practical translation guide explains the difference between selecting, marking, reviewing and casting a vote; between correcting a mark and surrendering a spoiled ballot; and between helping a person understand instructions and influencing the person’s choice. It is written for translators, editors, election-information teams, accessibility reviewers, civic educators and readers checking multilingual voting materials.
The hidden translation problem is that a short instruction can carry several separate permissions, prohibitions, stages and consequences. “Mark one” is not the same as “vote for up to one.” “Review your ballot” is not “submit your ballot.” A voter who receives a replacement ballot has not necessarily cast two votes; a ballot that is spoiled before casting must not automatically be described as an invalid vote during counting. A fluent translation can still misdirect a voter when it changes the action, the maximum number of selections, who may offer help, or when a choice becomes final.
This worldwide reference gives a repeatable method for translating voting instructions, ballot secrecy, ballot corrections and voting accessibility language; it includes source-to-meaning examples, diagnostic checks, failure modes, practical workshops, and editorial handoff templates. Rules differ across election systems: examples here are teaching simulations, not universal legal instructions. Always use the competent electoral authority’s applicable wording, certified ballot design, and authorized language review where the material will be issued to actual voters.
- Start with the decision: identify what the voter may, must or must not do at the exact stage of voting.
- Protect cardinality: preserve “one,” “up to,” “at least,” ranked orders and contest limits.
- Preserve status: separate issued, marked, reviewed, spoiled, surrendered, replaced, cast, declined and counted.
- Keep the choice private: do not turn assistance or verification into a demand to reveal a selection.
- Finish with testing: compare the intended voter action, not merely the dictionary words, and obtain jurisdiction-specific approval.
For the wider system of cross-language meaning transfer, begin with Master Art of Translation | The Complete System for Moving Meaning Between Languages. For the distinct preceding step—check-in, identification and issuance of the ballot—see the polling-day check-in and ballot-issuance chapter. This chapter starts after lawful issuance and focuses on the meaning of the instructions that lead toward casting.
1. What the reader is really trying to solve
A search for “how to translate ballot instructions” often conceals a more operational need: a translated sentence must cause the same lawful action as its source. The translator must know whether a mark is provisional, whether several selections are permitted, whether a screen’s Continue button merely navigates, and whether a request for a replacement requires the old paper to be returned first. The most reliable unit of translation is therefore not the word or even the sentence. It is the instruction plus its decision state.
A reader searching for “spoiled ballot meaning,” “spoilt ballot translation,” “overvote and undervote meaning,” “assisted voting language,” or “secret ballot instructions” may encounter apparently interchangeable expressions used differently by different systems. Some spellings vary, such as spoiled and spoilt. Some legal classifications also vary. Avoid converting a procedural convention in one election into a universal rule. The translator’s job is to reveal where a governing rule must be looked up rather than hiding uncertainty inside a confident phrase.
The boundary of this chapter matters. Earlier articles handle the electoral register, voter eligibility, issuing the ballot and official ballot reconciliation. Later election operations may include tabulation and certification. Here we isolate the human interaction between an issued ballot and its lawful casting, because this is where a single mistranslated command can redirect a real-world action.
2. The hidden problem: one expression, several different actions
The expression “fill in” can describe physically shading an oval, supplying missing information, or completing a form. On a paper ballot, “fill in the oval completely” is a graphic marking instruction. On a touchscreen, “select the circle” might mean a tappable interface element rather than ink. A translator who chooses the most familiar equivalent without seeing the device may produce a grammatical instruction that cannot be carried out. Request the actual approved ballot, screen or illustrated guide before settling the verb.
The difference between “mark” and “cast” is also decisive. A person can make marks in a private voting compartment without yet delivering a ballot into the system. Depending on the process, a person may be able to correct, replace or review choices before final casting. After casting, the available remedies may be completely different. The translation must preserve that threshold with special care: do not make a review screen sound like a final confirmation, and do not make a final-submission control sound like a reversible preview.
Language introduces risk through negation, quantities and scope. “Choose no more than two” excludes three but may allow zero or one. “Choose two” can sound like an exact requirement. “For this contest” restricts the instruction to one portion of the ballot; omitting that phrase can make the rule appear to cover the entire document. “Unless directed by an election official” is a conditional exception that must not be expanded into unlimited official discretion. These changes appear tiny in text but large in conduct.
Status expressions form a second layer. An incorrectly marked paper handed back for replacement before casting might be labelled spoiled or spoilt under an authority’s procedure. An unissued damaged form is a different inventory problem. An overvote may be a choice pattern detected during processing, not a synonym for a surrendered ballot. An unmarked ballot need not be legally equivalent to declining a ballot. The translator should write these contrasts into the termbase before creating a public-facing sentence.
3. The mechanism: a six-field meaning ledger
Actor
Who acts: voter, chosen assistant, trained official, machine, adjudicator or observer? Keep the actor visible when passive voice makes authorization ambiguous. A translation can sound natural while losing this field. Record it explicitly in the project notes, and ask for documentary evidence if the source does not resolve it. The completed ledger lets a reviewer compare operational equivalence rather than relying on a subjective impression of fluency.
Object
What receives the action: whole ballot, one contest, one option, a ranking field, a write-in field, a secrecy sleeve, a replacement request or a display control? A translation can sound natural while losing this field. Record it explicitly in the project notes, and ask for documentary evidence if the source does not resolve it. The completed ledger lets a reviewer compare operational equivalence rather than relying on a subjective impression of fluency.
Action
What is being done: select, mark, review, correct, surrender, replace, cast, scan, deposit, verify, explain, assist or count? Do not conflate neighboring verbs. A translation can sound natural while losing this field. Record it explicitly in the project notes, and ask for documentary evidence if the source does not resolve it. The completed ledger lets a reviewer compare operational equivalence rather than relying on a subjective impression of fluency.
Constraint
What limits the action: maximum choices, exact ranking order, time, place, stage, required mark, consent, confidentiality or an eligibility condition? A translation can sound natural while losing this field. Record it explicitly in the project notes, and ask for documentary evidence if the source does not resolve it. The completed ledger lets a reviewer compare operational equivalence rather than relying on a subjective impression of fluency.
State change
What becomes different afterward: unmarked to marked, marked to spoiled, issued to surrendered, draft to final cast, anonymous to potentially exposed? A translation can sound natural while losing this field. Record it explicitly in the project notes, and ask for documentary evidence if the source does not resolve it. The completed ledger lets a reviewer compare operational equivalence rather than relying on a subjective impression of fluency.
Proof
What evidence lets a reviewer test the translation: sample ballot, approved instructions, screenshot, device behavior, legal text, multilingual glossary, user test or reviewer sign-off? A translation can sound natural while losing this field. Record it explicitly in the project notes, and ask for documentary evidence if the source does not resolve it. The completed ledger lets a reviewer compare operational equivalence rather than relying on a subjective impression of fluency.
4. A repeatable translation method: from source file to voter-safe instruction
1. Collect the controlling artefacts
Obtain the approved source text, ballot artwork, interface screenshots, procedural manual, terminology approvals and intended audience. A list of isolated sentences is not enough when a button’s placement changes its meaning. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
2. Separate voter text from official-only text
A line directed to an election worker must not become a voter-facing command. Track labels such as ‘for staff use,’ form instructions, observer notices and explanatory captions. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
3. Annotate the workflow state
Mark where issuance ends, marking begins, review occurs, a spoiled ballot can be requested, and casting becomes final. If this path is unknown, escalate rather than invent a workflow. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
4. Lock the selection grammar
Extract exact numbers, maxima, minima, conjunctions, exceptions and conditions. Test zero, one, the maximum and one beyond the maximum to see whether translated instructions still have the same reach. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
5. Build a controlled glossary
Keep select, mark, cast, spoil, surrender, replace, assist, observe, disclose, overvote and undervote distinct; add approved local equivalents and usage contexts, not one generic translation for every occurrence. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
6. Draft for functional equivalence
Translate what the reader must understand and do. Preserve candidate names, official titles, party designations and legally fixed labels according to the source authority’s naming policy. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
7. Read against the visual design
Check alignment of text and boxes, reading direction, line wraps, number placement, icons, color assumptions, button placement, tactile alternatives and audio prompts. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
8. Perform a negative test
For each instruction, ask what a reader might do if they misunderstood it. Try to produce the most plausible wrong action, then see whether the translation accidentally permits it. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
9. Perform a recovery test
Test what happens when a voter makes an error, changes their mind before casting, needs assistance, or cannot interpret a screen. The correction pathway must remain available where the authoritative procedure provides one. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
10. Protect the secrecy boundary
Verify that assistance guidance does not require a voter to announce or show a choice to an unauthorized person. Never add unofficial requirements for proof of preference. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
11. Review with qualified humans
Use subject-matter and language reviewers who know the actual voting process. A bilingual text that reads smoothly is not enough if its procedural consequence is wrong. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
12. Test with representative users
Use plain-language, readability, accessibility and observed task testing. Ask users to demonstrate what they would do, not merely whether they liked the wording. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
13. Version and approve
Record source version, reviewer, date, approved terminology, unresolved risks, affected ballot format, and replacement procedure. Publish only the approved artifacts. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
14. Monitor the released artefact
After layout or software changes, compare actual visible text, audio text and print. A late line break, software label substitution or cropped number can undo a correct translation. The operator should keep a short evidence note showing the original constraint, the translated constraint and the result of a simple user-action test. If the two actions diverge, treat the issue as a meaning defect even when the grammar is elegant.
5. Four classifications that must not be collapsed
Spoiled or spoilt ballot
Often an issued ballot returned or otherwise handled under an approved procedure because it has been marked incorrectly or is unusable. Whether and how replacement occurs is system-specific. Never infer that its selections were cast or counted. A reviewer should record the term in a controlled glossary together with the actor, stage, and jurisdictional source. The crucial diagnostic question is whether a reader would take the same next step after seeing the translated expression. If not, the glossary entry has become a false friend.
Invalid or rejected ballot
A status normally determined under counting or adjudication rules. It can describe a ballot or portion of a ballot that does not satisfy applicable requirements, but it is not a free synonym for an incorrectly marked form awaiting replacement. A reviewer should record the term in a controlled glossary together with the actor, stage, and jurisdictional source. The crucial diagnostic question is whether a reader would take the same next step after seeing the translated expression. If not, the glossary entry has become a false friend.
Overvote
A condition involving more selections than allowed in a contest under the applicable rules. Translation should preserve the unit ‘contest’ and should not imply that every separate contest on the same ballot is necessarily rejected. A reviewer should record the term in a controlled glossary together with the actor, stage, and jurisdictional source. The crucial diagnostic question is whether a reader would take the same next step after seeing the translated expression. If not, the glossary entry has become a false friend.
Undervote or blank contest
Fewer choices than the permitted maximum, or none in a contest, under a system’s definitions. A choice not to mark every allowed option may be intentional. Do not translate this as a mistake without evidence. A reviewer should record the term in a controlled glossary together with the actor, stage, and jurisdictional source. The crucial diagnostic question is whether a reader would take the same next step after seeing the translated expression. If not, the glossary entry has become a false friend.
Surrendered ballot
A paper handed back to officials; the act of surrender does not by itself tell a reviewer why it was returned, whether it was marked, or how it should be accounted for. A reviewer should record the term in a controlled glossary together with the actor, stage, and jurisdictional source. The crucial diagnostic question is whether a reader would take the same next step after seeing the translated expression. If not, the glossary entry has become a false friend.
Cast ballot
A ballot submitted into the official voting process at the designated final step. The implementation differs between physical boxes, scanners, voting devices and other authorized methods. A reviewer should record the term in a controlled glossary together with the actor, stage, and jurisdictional source. The crucial diagnostic question is whether a reader would take the same next step after seeing the translated expression. If not, the glossary entry has become a false friend.
Assisted ballot marking
Help provided under authorized rules, potentially including reading instructions or physically marking a ballot at the voter’s direction. Translation must retain consent, autonomy and secrecy conditions. A reviewer should record the term in a controlled glossary together with the actor, stage, and jurisdictional source. The crucial diagnostic question is whether a reader would take the same next step after seeing the translated expression. If not, the glossary entry has become a false friend.
Ballot secrecy
Protection of the association between the voter and their selections. A secrecy sleeve, private compartment, helper confidentiality, and prohibitions on disclosure serve different parts of that goal. A reviewer should record the term in a controlled glossary together with the actor, stage, and jurisdictional source. The crucial diagnostic question is whether a reader would take the same next step after seeing the translated expression. If not, the glossary entry has become a false friend.
6. Worked diagnostics: eighty-four practical translation situations
Every case below is a fictional instructional example. The ‘source message’ is an invented English training line, not an official form. The ‘unsafe version’ describes a possible meaning error; the ‘safer meaning’ is a semantic restatement in English, not an authorized translation for any specific language. Real ballot wording and the legality of the proposed actions must be confirmed with the responsible authority. The purpose is to train distinctions that can survive transfer into a chosen target language.
1. Marking medium and graphic target
The controlling principle for this family of instructions is straightforward: A voter must know which physical or digital object receives the mark. The central translation test is equally practical: Show a sample ballot or screen and point to the exact target of the instruction. The consequences are not only stylistic. A correct verb directed to the wrong visual object can create an unreadable mark or an unintended click. Translators should read the instruction as a real interaction in paper ovals, boxes, touch interfaces, accessible audio controls and illustrated instructions, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 01: Fill the oval
Illustrative source: “Fill the oval beside one name.” Unsafe interpretation: “Draw a circle around the name.” Safer meaning: “Shade the printed oval beside the chosen name.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The target is the printed oval, not the typography of the candidate name. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Ask a tester to demonstrate the ink motion on an empty sample. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for marking medium and graphic target is: Show a sample ballot or screen and point to the exact target of the instruction. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A correct verb directed to the wrong visual object can create an unreadable mark or an unintended click. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ovals, boxes, touch interfaces, accessible audio controls and illustrated instructions. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 02: Mark a box
Illustrative source: “Put a cross inside the box.” Unsafe interpretation: “Cross out the entire row.” Safer meaning: “Place the specified cross inside the option’s box.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Crossing out a candidate has a different visual effect from placing a cross in a box. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Inspect whether icons and examples preserve the same center point. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for marking medium and graphic target is: Show a sample ballot or screen and point to the exact target of the instruction. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A correct verb directed to the wrong visual object can create an unreadable mark or an unintended click. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ovals, boxes, touch interfaces, accessible audio controls and illustrated instructions. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 03: Tap a choice
Illustrative source: “Tap the button beside your choice.” Unsafe interpretation: “Tap the written name on any part of the screen.” Safer meaning: “Activate the on-screen control associated with your choice.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The touch target may be smaller than the text and may not encompass the full row. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Use the actual interface to confirm what activates. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for marking medium and graphic target is: Show a sample ballot or screen and point to the exact target of the instruction. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A correct verb directed to the wrong visual object can create an unreadable mark or an unintended click. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ovals, boxes, touch interfaces, accessible audio controls and illustrated instructions. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 04: Use a marking pen
Illustrative source: “Use the pen supplied at the booth.” Unsafe interpretation: “Use any writing instrument available.” Safer meaning: “Use the supplied marking instrument, as directed for this ballot.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Ink, pencil, stylus and machine interaction are not universally interchangeable. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check the approved equipment guidance rather than inventing substitutions. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for marking medium and graphic target is: Show a sample ballot or screen and point to the exact target of the instruction. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A correct verb directed to the wrong visual object can create an unreadable mark or an unintended click. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ovals, boxes, touch interfaces, accessible audio controls and illustrated instructions. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 05: Complete the arrow
Illustrative source: “Complete the broken arrow next to your selection.” Unsafe interpretation: “Draw a freehand arrow toward the option.” Safer meaning: “Join the printed parts of the arrow in the specified manner.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The action is bound to a particular optical design, not decorative drawing. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Compare the example illustration with the translated verb. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for marking medium and graphic target is: Show a sample ballot or screen and point to the exact target of the instruction. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A correct verb directed to the wrong visual object can create an unreadable mark or an unintended click. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ovals, boxes, touch interfaces, accessible audio controls and illustrated instructions. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 06: Do not sign the ballot
Illustrative source: “Do not write your name on this ballot.” Unsafe interpretation: “Sign only where there is room.” Safer meaning: “Do not add your identity to the marked voting paper.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Unrequested identification can conflict with secrecy and marking constraints. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test whether the warning is legible near optional write-in fields. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for marking medium and graphic target is: Show a sample ballot or screen and point to the exact target of the instruction. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A correct verb directed to the wrong visual object can create an unreadable mark or an unintended click. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ovals, boxes, touch interfaces, accessible audio controls and illustrated instructions. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck marking medium and graphic target after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
2. Selection limits and counting words
The controlling principle for this family of instructions is straightforward: The permitted number of selections must remain numerically and logically identical. The central translation test is equally practical: Present a choice of zero, one, the maximum and one more than the maximum; ask what is allowed. The consequences are not only stylistic. Mistranslating a quantity changes the permission granted to voters. Translators should read the instruction as a real interaction in multi-member elections, committee ballots, referenda with several questions and preference surveys, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 07: Vote for one
Illustrative source: “Vote for one candidate in this contest.” Unsafe interpretation: “Vote for at least one candidate.” Safer meaning: “Choose one candidate for this specific contest.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Exactly-one wording and minimum-one wording are not automatically equivalent. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Try zero and two selections against the source rule. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for selection limits and counting words is: Present a choice of zero, one, the maximum and one more than the maximum; ask what is allowed. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Mistranslating a quantity changes the permission granted to voters. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into multi-member elections, committee ballots, referenda with several questions and preference surveys. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 08: Vote for up to two
Illustrative source: “Vote for up to two candidates.” Unsafe interpretation: “You must choose two candidates.” Safer meaning: “You may select no more than two candidates.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. An upper limit does not in itself require the voter to use both selections. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test whether a reader thinks one selection is forbidden. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for selection limits and counting words is: Present a choice of zero, one, the maximum and one more than the maximum; ask what is allowed. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Mistranslating a quantity changes the permission granted to voters. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into multi-member elections, committee ballots, referenda with several questions and preference surveys. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 09: Vote for no more than three
Illustrative source: “Vote for no more than three.” Unsafe interpretation: “Choose three or more.” Safer meaning: “Choose at most three options.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The comparative direction of the limit is crucial. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check the Arabic or other target-language comparison carefully with a native reviewer. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for selection limits and counting words is: Present a choice of zero, one, the maximum and one more than the maximum; ask what is allowed. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Mistranslating a quantity changes the permission granted to voters. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into multi-member elections, committee ballots, referenda with several questions and preference surveys. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 10: Select one in each column
Illustrative source: “Choose one candidate in each column.” Unsafe interpretation: “Choose one across the whole page.” Safer meaning: “Apply the one-choice restriction separately to every column.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The scope is local to each column rather than a total for the ballot. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Mark two different columns and verify the inferred permission. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for selection limits and counting words is: Present a choice of zero, one, the maximum and one more than the maximum; ask what is allowed. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Mistranslating a quantity changes the permission granted to voters. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into multi-member elections, committee ballots, referenda with several questions and preference surveys. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 11: Up to five across two pages
Illustrative source: “Select up to five names across both pages.” Unsafe interpretation: “Choose five names on each page.” Safer meaning: “The total across both pages must not exceed five.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Page boundaries are not the units of the maximum. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test a four-plus-two pattern with the reviewer. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for selection limits and counting words is: Present a choice of zero, one, the maximum and one more than the maximum; ask what is allowed. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Mistranslating a quantity changes the permission granted to voters. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into multi-member elections, committee ballots, referenda with several questions and preference surveys. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 12: One proposal at a time
Illustrative source: “Answer each proposal separately.” Unsafe interpretation: “Choose one answer for the entire booklet.” Safer meaning: “Give one answer to each individual proposal.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The singular word ‘answer’ can hide multiple repeated actions. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Track how many separate questions are present. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for selection limits and counting words is: Present a choice of zero, one, the maximum and one more than the maximum; ask what is allowed. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Mistranslating a quantity changes the permission granted to voters. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into multi-member elections, committee ballots, referenda with several questions and preference surveys. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck selection limits and counting words after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
3. Ranking, numbering and preference order
The controlling principle for this family of instructions is straightforward: Ranked choices express ordered preference rather than a count of independent approvals. The central translation test is equally practical: Ask the tester to show what 1, 2 and 3 mean and whether the same number may be reused. The consequences are not only stylistic. Wrong ranking verbs can reverse voter preference or flatten an ordered ballot into an approval vote. Translators should read the instruction as a real interaction in preferential ballots, student representative elections and organizational ranking forms, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 13: First preference
Illustrative source: “Write 1 beside your first preference.” Unsafe interpretation: “Write 1 beside the last choice.” Safer meaning: “Use 1 to identify the preferred first-ranked option.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The numeral represents order rather than the number of marks to make. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Explain which option is first without using the word ‘best’ as a legal substitute. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for ranking, numbering and preference order is: Ask the tester to show what 1, 2 and 3 mean and whether the same number may be reused. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Wrong ranking verbs can reverse voter preference or flatten an ordered ballot into an approval vote. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into preferential ballots, student representative elections and organizational ranking forms. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 14: Second preference
Illustrative source: “Write 2 beside your second preference.” Unsafe interpretation: “Place two marks beside one candidate.” Safer meaning: “Use the numeral 2 to identify the next ranked preference.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Two as a rank is not two physical selections. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test a display containing three candidates and sequential numerals. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for ranking, numbering and preference order is: Ask the tester to show what 1, 2 and 3 mean and whether the same number may be reused. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Wrong ranking verbs can reverse voter preference or flatten an ordered ballot into an approval vote. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into preferential ballots, student representative elections and organizational ranking forms. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 15: Continue ranking
Illustrative source: “Continue numbering candidates in order.” Unsafe interpretation: “Put the same number by every acceptable candidate.” Safer meaning: “Assign subsequent ranking numbers in sequence.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Sequential numbering carries an ordering relationship. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check an example with a skipped numeral. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for ranking, numbering and preference order is: Ask the tester to show what 1, 2 and 3 mean and whether the same number may be reused. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Wrong ranking verbs can reverse voter preference or flatten an ordered ballot into an approval vote. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into preferential ballots, student representative elections and organizational ranking forms. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 16: Rank only as many as permitted
Illustrative source: “Rank no more than four candidates.” Unsafe interpretation: “You must rank every listed candidate.” Safer meaning: “Rank up to the stated maximum of four, if the rules allow optional ranks.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The rule combines ranking with an upper bound; do not add a completion requirement. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test both four and fewer-ranked examples. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for ranking, numbering and preference order is: Ask the tester to show what 1, 2 and 3 mean and whether the same number may be reused. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Wrong ranking verbs can reverse voter preference or flatten an ordered ballot into an approval vote. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into preferential ballots, student representative elections and organizational ranking forms. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 17: Do not repeat a rank
Illustrative source: “Use each ranking number only once.” Unsafe interpretation: “Use each number on every page.” Safer meaning: “Do not assign the same rank twice within the covered contest.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The prohibition applies to duplicate rank values. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Give a test ballot with two first preferences and ask what is wrong. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for ranking, numbering and preference order is: Ask the tester to show what 1, 2 and 3 mean and whether the same number may be reused. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Wrong ranking verbs can reverse voter preference or flatten an ordered ballot into an approval vote. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into preferential ballots, student representative elections and organizational ranking forms. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 18: Tie is not a rank
Illustrative source: “Do not mark two candidates as first choice.” Unsafe interpretation: “Choose both candidates as first.” Safer meaning: “Do not assign first rank to more than one candidate.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A translation that suggests joint first rank changes the ballot model. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Use an intentionally tied example to test comprehension. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for ranking, numbering and preference order is: Ask the tester to show what 1, 2 and 3 mean and whether the same number may be reused. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Wrong ranking verbs can reverse voter preference or flatten an ordered ballot into an approval vote. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into preferential ballots, student representative elections and organizational ranking forms. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck ranking, numbering and preference order after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
4. Referendum options and question polarity
The controlling principle for this family of instructions is straightforward: The vote relates to the precise proposition, including any negation in the question. The central translation test is equally practical: Ask the reviewer to explain what a Yes answer actually approves and what a No answer opposes. The consequences are not only stylistic. A reversed or simplified proposition can invert the political choice while sounding neutral. Translators should read the instruction as a real interaction in referenda, amendments, member resolutions and proposition ballots, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 19: Yes to the proposal
Illustrative source: “A Yes vote supports the proposal as written.” Unsafe interpretation: “Yes means you reject the proposal.” Safer meaning: “Yes expresses support for the particular proposition shown.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The answer does not describe personal approval of every possible related policy. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Read the question and answer label as one unit. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for referendum options and question polarity is: Ask the reviewer to explain what a Yes answer actually approves and what a No answer opposes. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A reversed or simplified proposition can invert the political choice while sounding neutral. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into referenda, amendments, member resolutions and proposition ballots. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 20: No to the proposal
Illustrative source: “A No vote opposes the proposal as written.” Unsafe interpretation: “No means you support the opposite party.” Safer meaning: “No opposes the presented proposition; it does not identify a party.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Avoid importing a party frame into a proposal question. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Back-translate both answer branches into concrete consequences. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for referendum options and question polarity is: Ask the reviewer to explain what a Yes answer actually approves and what a No answer opposes. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A reversed or simplified proposition can invert the political choice while sounding neutral. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into referenda, amendments, member resolutions and proposition ballots. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 21: Double negative
Illustrative source: “Should the ban on late applications be removed?” Unsafe interpretation: “A Yes vote supports keeping the ban.” Safer meaning: “The proposition asks whether an existing prohibition should be removed.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Negation applies to the removal action, not to the entire voting process. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Make an explicit decision table for Yes and No. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for referendum options and question polarity is: Ask the reviewer to explain what a Yes answer actually approves and what a No answer opposes. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A reversed or simplified proposition can invert the political choice while sounding neutral. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into referenda, amendments, member resolutions and proposition ballots. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 22: Approval wording
Illustrative source: “Do you approve the proposed amendment?” Unsafe interpretation: “Do you believe the amendment is popular?” Safer meaning: “The question asks for approval of the specified amendment.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Approval is an act of endorsement, not a prediction or general preference. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Verify the grammatical object of approve. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for referendum options and question polarity is: Ask the reviewer to explain what a Yes answer actually approves and what a No answer opposes. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A reversed or simplified proposition can invert the political choice while sounding neutral. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into referenda, amendments, member resolutions and proposition ballots. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 23: Multiple questions
Illustrative source: “Answer Question A and Question B independently.” Unsafe interpretation: “One Yes covers both questions.” Safer meaning: “Each question requires its own answer if the voter chooses to answer it.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Repeated answer words cannot be merged across proposals. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Use a two-question mock ballot. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for referendum options and question polarity is: Ask the reviewer to explain what a Yes answer actually approves and what a No answer opposes. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A reversed or simplified proposition can invert the political choice while sounding neutral. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into referenda, amendments, member resolutions and proposition ballots. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 24: Explanatory summary
Illustrative source: “The summary is explanatory; the official proposition controls.” Unsafe interpretation: “The summary replaces the legal question.” Safer meaning: “Treat the summary as an aid, while preserving the official question’s authority.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Where both are printed, they have different legal roles. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Label the summary consistently in the target version. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for referendum options and question polarity is: Ask the reviewer to explain what a Yes answer actually approves and what a No answer opposes. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A reversed or simplified proposition can invert the political choice while sounding neutral. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into referenda, amendments, member resolutions and proposition ballots. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck referendum options and question polarity after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
5. Write-in names and identity
The controlling principle for this family of instructions is straightforward: A write-in field identifies a person or option; it is not an invitation to translate a proper name freely. The central translation test is equally practical: Check whether the identity, official spelling and field usage remain verifiable. The consequences are not only stylistic. Automatic transliteration or grammatical normalization can alter the target identity. Translators should read the instruction as a real interaction in write-in lines, candidate lists, party names and official identity references, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 25: Write a candidate name
Illustrative source: “Write the name of a qualified write-in candidate.” Unsafe interpretation: “Write the translator’s preferred spelling.” Safer meaning: “Enter the candidate name according to the applicable official instructions.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Eligibility and required spelling may depend on the election rules. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check the approved candidate-name policy. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for write-in names and identity is: Check whether the identity, official spelling and field usage remain verifiable. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Automatic transliteration or grammatical normalization can alter the target identity. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into write-in lines, candidate lists, party names and official identity references. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 26: Use the write-in line
Illustrative source: “Use the line provided for write-in votes.” Unsafe interpretation: “Write in the margin beside any name.” Safer meaning: “Use the designated write-in field if this ballot permits it.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The physical location of text affects whether the choice can be read. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Verify field position after text expansion. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for write-in names and identity is: Check whether the identity, official spelling and field usage remain verifiable. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Automatic transliteration or grammatical normalization can alter the target identity. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into write-in lines, candidate lists, party names and official identity references. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 27: Activate a write-in box
Illustrative source: “Mark the write-in selection area when required.” Unsafe interpretation: “Writing a name always completes the selection.” Safer meaning: “Follow both the write-in-name and selection-mark instructions where specified.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Some ballot designs require more than text entry, but not all do. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test the complete workflow for the particular ballot. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for write-in names and identity is: Check whether the identity, official spelling and field usage remain verifiable. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Automatic transliteration or grammatical normalization can alter the target identity. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into write-in lines, candidate lists, party names and official identity references. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 28: Preserve diacritics
Illustrative source: “Copy the approved spelling of the candidate name.” Unsafe interpretation: “Remove unfamiliar accents to simplify the ballot.” Safer meaning: “Use the authorized candidate form without unapproved name changes.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Diacritics may distinguish identities and official records. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Request authoritative written guidance before normalization. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for write-in names and identity is: Check whether the identity, official spelling and field usage remain verifiable. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Automatic transliteration or grammatical normalization can alter the target identity. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into write-in lines, candidate lists, party names and official identity references. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 29: Non-Latin scripts
Illustrative source: “Use approved bilingual presentation for names.” Unsafe interpretation: “Replace every name with an unverified phonetic guess.” Safer meaning: “Follow the official script and transliteration policy.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Writing-system conversion and translation are different services. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Compare against the authority’s published candidate list. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for write-in names and identity is: Check whether the identity, official spelling and field usage remain verifiable. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Automatic transliteration or grammatical normalization can alter the target identity. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into write-in lines, candidate lists, party names and official identity references. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 30: Do not infer a name
Illustrative source: “Leave an unknown write-in name as entered for review.” Unsafe interpretation: “Correct the voter’s apparent spelling automatically.” Safer meaning: “Do not guess the intended person from an uncertain entry.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Interpretation of voter intent is not routine translator copyediting. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Escalate unclear recognition to the authorized process. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for write-in names and identity is: Check whether the identity, official spelling and field usage remain verifiable. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Automatic transliteration or grammatical normalization can alter the target identity. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into write-in lines, candidate lists, party names and official identity references. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck write-in names and identity after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
6. Corrections before final casting
The controlling principle for this family of instructions is straightforward: The reader must know whether an error can be corrected in place or requires a defined recovery path. The central translation test is equally practical: Ask the tester what they would do with an accidental mark before the vote is submitted. The consequences are not only stylistic. An unauthorized suggestion to erase, scratch or overwrite may make a recoverable situation worse. Translators should read the instruction as a real interaction in paper corrections, touchscreen changes, ballot reprints and device review screens, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 31: Wrong oval
Illustrative source: “If you marked the wrong oval, ask an official about the replacement procedure.” Unsafe interpretation: “Erase and redraw any way you like.” Safer meaning: “Seek the specified correction or replacement procedure before casting.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Whether erasure is permitted cannot be inferred from ordinary stationery practice. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Have a reviewer follow the actual written correction route. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for corrections before final casting is: Ask the tester what they would do with an accidental mark before the vote is submitted. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. An unauthorized suggestion to erase, scratch or overwrite may make a recoverable situation worse. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper corrections, touchscreen changes, ballot reprints and device review screens. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 32: Changed mind on screen
Illustrative source: “You may revise selections before the final cast step if the system allows it.” Unsafe interpretation: “The first tap is final.” Safer meaning: “Before final submission, follow the device’s provided change-selection controls.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The reversible stage depends on the device and approved configuration. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Inspect the real navigation and confirmation sequence. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for corrections before final casting is: Ask the tester what they would do with an accidental mark before the vote is submitted. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. An unauthorized suggestion to erase, scratch or overwrite may make a recoverable situation worse. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper corrections, touchscreen changes, ballot reprints and device review screens. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 33: Stray mark
Illustrative source: “Report a stray mark using the official assistance procedure.” Unsafe interpretation: “Add several marks to hide it.” Safer meaning: “Use the approved process for a stray marking error.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Stray marks can have technical or legal effects depending on the form. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Do not make up a universal threshold of harmlessness. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for corrections before final casting is: Ask the tester what they would do with an accidental mark before the vote is submitted. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. An unauthorized suggestion to erase, scratch or overwrite may make a recoverable situation worse. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper corrections, touchscreen changes, ballot reprints and device review screens. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 34: Folded or torn form
Illustrative source: “If the ballot is damaged, request instructions before casting.” Unsafe interpretation: “Tape the paper and submit normally.” Safer meaning: “Ask for the specified remedy when the ballot is torn or unusable.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Damage and marking error are distinct causes that may share or differ in remedies. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Verify how the authority classifies unusable materials. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for corrections before final casting is: Ask the tester what they would do with an accidental mark before the vote is submitted. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. An unauthorized suggestion to erase, scratch or overwrite may make a recoverable situation worse. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper corrections, touchscreen changes, ballot reprints and device review screens. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 35: Incorrect ranking
Illustrative source: “If you entered the wrong preference number, seek the permitted correction method.” Unsafe interpretation: “Overwrite every numeral without asking.” Safer meaning: “Use the officially authorized correction or replacement step.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A handwritten overwrite may leave ambiguous rankings. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check an error example with the official ballot design. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for corrections before final casting is: Ask the tester what they would do with an accidental mark before the vote is submitted. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. An unauthorized suggestion to erase, scratch or overwrite may make a recoverable situation worse. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper corrections, touchscreen changes, ballot reprints and device review screens. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 36: Machine cancellation
Illustrative source: “Choose Cancel only if you intend to abandon that stage.” Unsafe interpretation: “Cancel casts the current choice.” Safer meaning: “A Cancel control abandons the stated action rather than submitting a vote.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A single button label can affect status and recovery. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Inspect where the device returns after cancellation. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for corrections before final casting is: Ask the tester what they would do with an accidental mark before the vote is submitted. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. An unauthorized suggestion to erase, scratch or overwrite may make a recoverable situation worse. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper corrections, touchscreen changes, ballot reprints and device review screens. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck corrections before final casting after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
7. Spoiled, surrendered, replaced and accounted-for
The controlling principle for this family of instructions is straightforward: The status of a paper must be anchored to the exact election stage and accountable procedure. The central translation test is equally practical: Distinguish what happens to the original paper from what happens to any replacement. The consequences are not only stylistic. Confusing returned paper with counted votes can trigger erroneous records and public misunderstanding. Translators should read the instruction as a real interaction in voter notices, replacement forms, reconciliation explanations and official logs, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 37: Request a replacement
Illustrative source: “Ask about a replacement ballot if you make an error.” Unsafe interpretation: “Take another ballot without telling anyone.” Safer meaning: “Follow the authorized request procedure for an erroneous ballot.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The source is a request for official assistance, not unlimited self-service issuance. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check whether surrender is required by the actual procedure. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for spoiled, surrendered, replaced and accounted-for is: Distinguish what happens to the original paper from what happens to any replacement. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing returned paper with counted votes can trigger erroneous records and public misunderstanding. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into voter notices, replacement forms, reconciliation explanations and official logs. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 38: Surrender the original
Illustrative source: “Return the original ballot when the procedure requires it.” Unsafe interpretation: “Keep both ballots and cast either.” Safer meaning: “Return the designated original as required before receiving a replacement.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Keeping two live papers may defeat ballot accountability. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Ask the official reviewer about inventory records. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for spoiled, surrendered, replaced and accounted-for is: Distinguish what happens to the original paper from what happens to any replacement. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing returned paper with counted votes can trigger erroneous records and public misunderstanding. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into voter notices, replacement forms, reconciliation explanations and official logs. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 39: Spoiled is not cast
Illustrative source: “A spoiled ballot returned before casting is handled separately.” Unsafe interpretation: “Your spoiled ballot was cast as a vote.” Safer meaning: “Describe the returned unusable paper without presenting it as an accepted cast vote.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The same physical object can change status without becoming a counted vote. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check the stage where casting actually occurs. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for spoiled, surrendered, replaced and accounted-for is: Distinguish what happens to the original paper from what happens to any replacement. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing returned paper with counted votes can trigger erroneous records and public misunderstanding. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into voter notices, replacement forms, reconciliation explanations and official logs. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 40: Unused damaged stock
Illustrative source: “Set aside ballots damaged before issuance.” Unsafe interpretation: “Record damaged stock as a voter’s spoiled ballot.” Safer meaning: “Classify unissued defective forms separately from voter-returned papers.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The actor and stage distinguish printing defects from voter mistakes. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check stock control and issue counts. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for spoiled, surrendered, replaced and accounted-for is: Distinguish what happens to the original paper from what happens to any replacement. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing returned paper with counted votes can trigger erroneous records and public misunderstanding. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into voter notices, replacement forms, reconciliation explanations and official logs. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 41: Replacement does not mean extra vote
Illustrative source: “Issuing a permitted replacement does not add a second valid vote.” Unsafe interpretation: “Each replacement is another vote for the voter.” Safer meaning: “Explain replacement as substitution under the approved process.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Translation should not falsely imply duplicate participation. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Review how issuance and return are reconciled. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for spoiled, surrendered, replaced and accounted-for is: Distinguish what happens to the original paper from what happens to any replacement. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing returned paper with counted votes can trigger erroneous records and public misunderstanding. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into voter notices, replacement forms, reconciliation explanations and official logs. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 42: Declined versus spoiled
Illustrative source: “A voter may decline a ballot under applicable rules.” Unsafe interpretation: “Every returned unmarked ballot is spoiled.” Safer meaning: “Keep an intentionally declined ballot distinct from one returned for a marking mistake.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Declining, abstaining, abandoning and spoiling are not universal synonyms. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Consult the jurisdiction-specific decline procedure. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for spoiled, surrendered, replaced and accounted-for is: Distinguish what happens to the original paper from what happens to any replacement. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing returned paper with counted votes can trigger erroneous records and public misunderstanding. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into voter notices, replacement forms, reconciliation explanations and official logs. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck spoiled, surrendered, replaced and accounted-for after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
8. Review, submit, scan and cast
The controlling principle for this family of instructions is straightforward: Different interface verbs represent different levels of finality. The central translation test is equally practical: Ask users to identify the last point at which they can still change a selection. The consequences are not only stylistic. Confusing preview and submission can cause unwanted finality or failed completion. Translators should read the instruction as a real interaction in paper ballot boxes, optical scanners, digital ballots and audio voting systems, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 43: Review selections
Illustrative source: “Review your choices before casting.” Unsafe interpretation: “Your vote is already cast when the review screen appears.” Safer meaning: “Inspect selected options before the final cast action.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Review means evaluation, not necessarily final submission. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Map the next available buttons. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for review, submit, scan and cast is: Ask users to identify the last point at which they can still change a selection. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing preview and submission can cause unwanted finality or failed completion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ballot boxes, optical scanners, digital ballots and audio voting systems. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 44: Press Cast
Illustrative source: “Select Cast Ballot to submit your choices.” Unsafe interpretation: “Cast means go back to edit.” Safer meaning: “Use the final casting control only when ready to submit.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The action may be irreversible; the label must be explicit. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test a deliberate-choice moment on a training device. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for review, submit, scan and cast is: Ask users to identify the last point at which they can still change a selection. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing preview and submission can cause unwanted finality or failed completion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ballot boxes, optical scanners, digital ballots and audio voting systems. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 45: Insert into scanner
Illustrative source: “Insert the marked ballot as instructed.” Unsafe interpretation: “Insert any paper into any slot.” Safer meaning: “Follow the designated scanner intake instructions for the marked ballot.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A physical intake may lead to validation or additional prompts. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check what constitutes successful casting on that system. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for review, submit, scan and cast is: Ask users to identify the last point at which they can still change a selection. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing preview and submission can cause unwanted finality or failed completion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ballot boxes, optical scanners, digital ballots and audio voting systems. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 46: Deposit in ballot box
Illustrative source: “Place the completed ballot into the designated ballot box.” Unsafe interpretation: “Leave the paper on the table.” Safer meaning: “Deposit the ballot at the authorized receiving point.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A ballot box is not a general collection surface. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Trace the official route from booth to deposit. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for review, submit, scan and cast is: Ask users to identify the last point at which they can still change a selection. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing preview and submission can cause unwanted finality or failed completion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ballot boxes, optical scanners, digital ballots and audio voting systems. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 47: Successful submission message
Illustrative source: “Your ballot has been cast.” Unsafe interpretation: “You are ready to cast later.” Safer meaning: “The system indicates the final vote was submitted.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. An acknowledgement and a readiness prompt cannot be used interchangeably. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Compare actual device messages. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for review, submit, scan and cast is: Ask users to identify the last point at which they can still change a selection. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing preview and submission can cause unwanted finality or failed completion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ballot boxes, optical scanners, digital ballots and audio voting systems. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 48: Return to contest
Illustrative source: “Back returns to the previous contest screen.” Unsafe interpretation: “Back discards the whole election permanently.” Safer meaning: “Use Back to navigate as the approved interface specifies.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A navigation label may be reversible but must be device tested. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test Back from both selection and review screens. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for review, submit, scan and cast is: Ask users to identify the last point at which they can still change a selection. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Confusing preview and submission can cause unwanted finality or failed completion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into paper ballot boxes, optical scanners, digital ballots and audio voting systems. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck review, submit, scan and cast after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
9. Secrecy, privacy and pressure
The controlling principle for this family of instructions is straightforward: The instruction must protect the link between the voter and their private selection. The central translation test is equally practical: Ask who can see, hear, record or demand disclosure of a preference under the stated text. The consequences are not only stylistic. A poorly scoped confidentiality sentence can expose choices or normalize coercion. Translators should read the instruction as a real interaction in secrecy sleeves, privacy booths, audio devices, assistance protocols and observation rules, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 49: Vote privately
Illustrative source: “Mark your ballot in privacy.” Unsafe interpretation: “Show your marked ballot to the next voter.” Safer meaning: “Complete the selection without revealing it to unauthorized people.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Privacy is an operational condition, not merely a polite tone. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Inspect sightlines implied by illustrated instructions. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for secrecy, privacy and pressure is: Ask who can see, hear, record or demand disclosure of a preference under the stated text. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A poorly scoped confidentiality sentence can expose choices or normalize coercion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into secrecy sleeves, privacy booths, audio devices, assistance protocols and observation rules. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 50: Do not display choices
Illustrative source: “Keep your marked selections concealed.” Unsafe interpretation: “Hold the marked side up for verification.” Safer meaning: “Avoid displaying voting selections where unauthorized persons can view them.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Public display can compromise secrecy even without an identifier printed on the ballot. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check whether a sample image demonstrates the wrong orientation. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for secrecy, privacy and pressure is: Ask who can see, hear, record or demand disclosure of a preference under the stated text. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A poorly scoped confidentiality sentence can expose choices or normalize coercion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into secrecy sleeves, privacy booths, audio devices, assistance protocols and observation rules. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 51: No photo instruction
Illustrative source: “Do not photograph the marked ballot where prohibited.” Unsafe interpretation: “Post a picture to prove completion.” Safer meaning: “State the applicable rule without inventing a worldwide photo ban.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Photography rules differ; the translation must not overgeneralize. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Obtain legal approval of the actual prohibition. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for secrecy, privacy and pressure is: Ask who can see, hear, record or demand disclosure of a preference under the stated text. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A poorly scoped confidentiality sentence can expose choices or normalize coercion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into secrecy sleeves, privacy booths, audio devices, assistance protocols and observation rules. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 52: Do not announce choice
Illustrative source: “No one may require you to announce how you voted under the stated rules.” Unsafe interpretation: “Tell the official which candidate you supported.” Safer meaning: “Preserve the relevant protection against compelled disclosure.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The audience of the disclosure is central to the meaning. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check pronouns, agent and modal force. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for secrecy, privacy and pressure is: Ask who can see, hear, record or demand disclosure of a preference under the stated text. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A poorly scoped confidentiality sentence can expose choices or normalize coercion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into secrecy sleeves, privacy booths, audio devices, assistance protocols and observation rules. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 53: Secrecy sleeve
Illustrative source: “Place the marked ballot in the secrecy sleeve as instructed.” Unsafe interpretation: “Use the sleeve to identify the voter.” Safer meaning: “Use the sleeve to conceal selections during transfer where required.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A sleeve is a privacy device, not necessarily a voter identification container. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Inspect which side remains visible. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for secrecy, privacy and pressure is: Ask who can see, hear, record or demand disclosure of a preference under the stated text. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A poorly scoped confidentiality sentence can expose choices or normalize coercion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into secrecy sleeves, privacy booths, audio devices, assistance protocols and observation rules. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 54: Coercive helper
Illustrative source: “An assistant must not substitute their preference for yours.” Unsafe interpretation: “The assistant chooses what is best for you.” Safer meaning: “The voter’s choices remain controlling when help is authorized.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Assistance is not permission to direct or change a vote. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Ask user testers who makes the final decision. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for secrecy, privacy and pressure is: Ask who can see, hear, record or demand disclosure of a preference under the stated text. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A poorly scoped confidentiality sentence can expose choices or normalize coercion. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into secrecy sleeves, privacy booths, audio devices, assistance protocols and observation rules. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck secrecy, privacy and pressure after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
10. Assistance, permission and autonomy
The controlling principle for this family of instructions is straightforward: Eligibility for help, chosen helper and authority to mark must remain distinct. The central translation test is equally practical: Identify who requests help, who chooses the helper and what that helper may do. The consequences are not only stylistic. Translating assistance as an official takeover can erase the voter’s independent choice. Translators should read the instruction as a real interaction in accessible voting notices, oral assistance, reader services and assisted marking, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 55: Ask for assistance
Illustrative source: “If you need help, ask an election official about available assistance.” Unsafe interpretation: “Every voter must be helped by an official.” Safer meaning: “Help is available on request under the applicable rules.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. An offer of support is not an obligation to disclose private circumstances. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Observe how the request is made without exposing the selection. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for assistance, permission and autonomy is: Identify who requests help, who chooses the helper and what that helper may do. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Translating assistance as an official takeover can erase the voter’s independent choice. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into accessible voting notices, oral assistance, reader services and assisted marking. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 56: Voter-chosen assistant
Illustrative source: “You may select an eligible assistant where the rules allow.” Unsafe interpretation: “Only an official can select your helper.” Safer meaning: “The voter may choose a helper subject to the stated conditions.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The choice of assistant is a separate decision from the voting preference. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check exceptions or disqualifications in the governing procedure. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for assistance, permission and autonomy is: Identify who requests help, who chooses the helper and what that helper may do. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Translating assistance as an official takeover can erase the voter’s independent choice. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into accessible voting notices, oral assistance, reader services and assisted marking. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 57: Read but do not decide
Illustrative source: “An assistant may read instructions but must not choose for the voter.” Unsafe interpretation: “Reading aloud authorizes the assistant to mark any choice.” Safer meaning: “Reading support must preserve the voter’s direction.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Information access and decision authority are different permissions. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Ask the reviewer to paraphrase the assistant’s exact scope. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for assistance, permission and autonomy is: Identify who requests help, who chooses the helper and what that helper may do. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Translating assistance as an official takeover can erase the voter’s independent choice. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into accessible voting notices, oral assistance, reader services and assisted marking. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 58: Physical marking on instruction
Illustrative source: “The helper marks only the choice directed by the voter.” Unsafe interpretation: “The helper may infer the choice from conversation.” Safer meaning: “The assistant follows the person’s expressed selection if authorized to mark.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Inference is not a substitute for direction. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test with a deliberately ambiguous request. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for assistance, permission and autonomy is: Identify who requests help, who chooses the helper and what that helper may do. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Translating assistance as an official takeover can erase the voter’s independent choice. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into accessible voting notices, oral assistance, reader services and assisted marking. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 59: Confidential helper
Illustrative source: “A permitted assistant must protect the voter’s preference.” Unsafe interpretation: “Tell others afterward how the voter voted.” Safer meaning: “Keep the choice confidential under the relevant rules.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Secrecy obligations may extend beyond the polling space. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Ask whether the instruction includes post-assistance disclosure. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for assistance, permission and autonomy is: Identify who requests help, who chooses the helper and what that helper may do. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Translating assistance as an official takeover can erase the voter’s independent choice. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into accessible voting notices, oral assistance, reader services and assisted marking. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 60: Declining help
Illustrative source: “You may refuse offered assistance if the procedure allows independent voting.” Unsafe interpretation: “Anyone who needs accessibility support must accept help.” Safer meaning: “Present offered assistance as optional where that is the rule.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The word may is a permission rather than a duty. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test that the translation leaves room for independent voting. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for assistance, permission and autonomy is: Identify who requests help, who chooses the helper and what that helper may do. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Translating assistance as an official takeover can erase the voter’s independent choice. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into accessible voting notices, oral assistance, reader services and assisted marking. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck assistance, permission and autonomy after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
11. Accessibility channels and equivalent instructions
The controlling principle for this family of instructions is straightforward: The same decision must be available through different access modalities without changing rights. The central translation test is equally practical: Compare the visible, spoken, tactile and alternate-format message for equivalent constraints. The consequences are not only stylistic. A silent audio omission or a misleading symbol can make accessible voting less independent. Translators should read the instruction as a real interaction in audio ballots, large print, braille, tactile guides, switches and screen-reader interfaces, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 61: Audio next button
Illustrative source: “Use Next to move to the following contest.” Unsafe interpretation: “Next records the current vote permanently.” Safer meaning: “The spoken Next command advances to the next step as configured.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Navigation and casting remain separate in audio interaction. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Listen through the whole sequence. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for accessibility channels and equivalent instructions is: Compare the visible, spoken, tactile and alternate-format message for equivalent constraints. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A silent audio omission or a misleading symbol can make accessible voting less independent. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into audio ballots, large print, braille, tactile guides, switches and screen-reader interfaces. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 62: Repeat instruction
Illustrative source: “Press Repeat to hear this instruction again.” Unsafe interpretation: “Repeat makes a second selection.” Safer meaning: “Repeat replays information without creating an extra vote.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The repeated content must match the current screen’s rule. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test after the user changes a selection. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for accessibility channels and equivalent instructions is: Compare the visible, spoken, tactile and alternate-format message for equivalent constraints. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A silent audio omission or a misleading symbol can make accessible voting less independent. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into audio ballots, large print, braille, tactile guides, switches and screen-reader interfaces. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 63: Tactile marking guide
Illustrative source: “Use the tactile guide aligned with the ballot.” Unsafe interpretation: “Any hole represents any candidate.” Safer meaning: “The guide’s openings correspond to specific printed options.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Physical alignment carries semantic information beyond text. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Inspect the exact order of options. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for accessibility channels and equivalent instructions is: Compare the visible, spoken, tactile and alternate-format message for equivalent constraints. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A silent audio omission or a misleading symbol can make accessible voting less independent. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into audio ballots, large print, braille, tactile guides, switches and screen-reader interfaces. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 64: Large-print equivalence
Illustrative source: “Large-print instructions preserve the same marking rule.” Unsafe interpretation: “Large print means a different selection limit.” Safer meaning: “Format enlargement must not change the voting constraint.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Line wrapping must keep quantities next to the governed action. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Compare print formats side by side. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for accessibility channels and equivalent instructions is: Compare the visible, spoken, tactile and alternate-format message for equivalent constraints. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A silent audio omission or a misleading symbol can make accessible voting less independent. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into audio ballots, large print, braille, tactile guides, switches and screen-reader interfaces. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 65: Screen reader label
Illustrative source: “The focused control announces the candidate and selection state.” Unsafe interpretation: “Every button is read as Button.” Safer meaning: “Accessible labeling must convey the same choice and state.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Generic controls can hide which candidate a user is selecting. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test keyboard focus and spoken feedback. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for accessibility channels and equivalent instructions is: Compare the visible, spoken, tactile and alternate-format message for equivalent constraints. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A silent audio omission or a misleading symbol can make accessible voting less independent. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into audio ballots, large print, braille, tactile guides, switches and screen-reader interfaces. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 66: Language in audio mode
Illustrative source: “Choose a language for spoken instructions.” Unsafe interpretation: “Changing language changes the legal vote.” Safer meaning: “The language setting changes presentation, not the ballot’s underlying options.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A language switch must preserve selection state when designed to do so. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test round trips between language modes. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for accessibility channels and equivalent instructions is: Compare the visible, spoken, tactile and alternate-format message for equivalent constraints. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A silent audio omission or a misleading symbol can make accessible voting less independent. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into audio ballots, large print, braille, tactile guides, switches and screen-reader interfaces. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck accessibility channels and equivalent instructions after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
12. Multiple languages, scripts and layout
The controlling principle for this family of instructions is straightforward: Translation must remain aligned with the source option, number and interface element. The central translation test is equally practical: Compare paired languages while looking at the real ballot layout and reading order. The consequences are not only stylistic. A technically accurate text may point to the wrong candidate or omit a crucial modifier when laid out. Translators should read the instruction as a real interaction in bilingual paper ballots, right-to-left scripts, vertical text, icons and mobile layouts, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 67: Paired names
Illustrative source: “Each translation appears next to its source option.” Unsafe interpretation: “Put translated labels under a neighboring candidate.” Safer meaning: “Keep each target-language label associated with the correct option.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Vertical and horizontal alignment determine what readers believe they are choosing. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Trace the selection row with a ruler or screen overlay. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for multiple languages, scripts and layout is: Compare paired languages while looking at the real ballot layout and reading order. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A technically accurate text may point to the wrong candidate or omit a crucial modifier when laid out. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into bilingual paper ballots, right-to-left scripts, vertical text, icons and mobile layouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 68: Right-to-left order
Illustrative source: “Maintain the intended ordering of names and numbers.” Unsafe interpretation: “Mirror the candidates automatically.” Safer meaning: “Apply the authority-approved right-to-left layout without altering ballot order.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Layout direction and election order are separate design decisions. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Compare numeric ranks and list sequences. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for multiple languages, scripts and layout is: Compare paired languages while looking at the real ballot layout and reading order. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A technically accurate text may point to the wrong candidate or omit a crucial modifier when laid out. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into bilingual paper ballots, right-to-left scripts, vertical text, icons and mobile layouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 69: Mixed numerals
Illustrative source: “Choose up to 2 candidates.” Unsafe interpretation: “Change the number to 3 for a target script.” Safer meaning: “Preserve the numeric limit using approved numeral presentation.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Numeral shape may localize; quantitative value may not. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Run a two-versus-three regression test. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for multiple languages, scripts and layout is: Compare paired languages while looking at the real ballot layout and reading order. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A technically accurate text may point to the wrong candidate or omit a crucial modifier when laid out. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into bilingual paper ballots, right-to-left scripts, vertical text, icons and mobile layouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 70: Wrapped negative
Illustrative source: “Do not select more than one.” Unsafe interpretation: “Place ‘not’ on a distant line or remove it.” Safer meaning: “Keep negation visibly attached to the selection limit.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A line break can separate a small word that controls legality. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Review the narrowest screen and smallest print format. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for multiple languages, scripts and layout is: Compare paired languages while looking at the real ballot layout and reading order. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A technically accurate text may point to the wrong candidate or omit a crucial modifier when laid out. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into bilingual paper ballots, right-to-left scripts, vertical text, icons and mobile layouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 71: Icon plus text
Illustrative source: “The arrow icon indicates the Next control.” Unsafe interpretation: “A decorative arrow means a vote has been cast.” Safer meaning: “Ensure icons and translated labels describe the same action.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Icons should not silently contradict accompanying text. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Ask a user to act from the icon alone. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for multiple languages, scripts and layout is: Compare paired languages while looking at the real ballot layout and reading order. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A technically accurate text may point to the wrong candidate or omit a crucial modifier when laid out. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into bilingual paper ballots, right-to-left scripts, vertical text, icons and mobile layouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 72: Language change mid-process
Illustrative source: “Change display language without intentionally resetting choices.” Unsafe interpretation: “Switching language discards every selection.” Safer meaning: “Do not claim a reset or preservation behavior without testing the device.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The software’s true state transition must be documented. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Switch languages on a test ballot after making a selection. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for multiple languages, scripts and layout is: Compare paired languages while looking at the real ballot layout and reading order. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. A technically accurate text may point to the wrong candidate or omit a crucial modifier when laid out. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into bilingual paper ballots, right-to-left scripts, vertical text, icons and mobile layouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck multiple languages, scripts and layout after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
13. Exception conditions and branching
The controlling principle for this family of instructions is straightforward: Conditional words determine which route is open when something unusual occurs. The central translation test is equally practical: Draw an if-then diagram from the translated sentence and compare with the source. The consequences are not only stylistic. Moving an exception or weakening a modal verb can authorize the wrong behavior. Translators should read the instruction as a real interaction in equipment failures, exceptions, accessibility alternatives and correction notices, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 73: If the scanner returns the ballot
Illustrative source: “If the scanner returns your paper, follow the displayed instruction.” Unsafe interpretation: “A returned paper means your vote was successfully counted.” Safer meaning: “The return triggers a specific follow-up step.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A physical return may represent an error or another defined status. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Observe the machine’s exact error screen. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for exception conditions and branching is: Draw an if-then diagram from the translated sentence and compare with the source. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Moving an exception or weakening a modal verb can authorize the wrong behavior. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into equipment failures, exceptions, accessibility alternatives and correction notices. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 74: Unless an official instructs otherwise
Illustrative source: “Use this path unless the authorized official directs the approved alternative.” Unsafe interpretation: “Anyone nearby may choose another path.” Safer meaning: “The exception is limited to a specified authorized role.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. The actor attached to an exception is central to its scope. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test who qualifies as the source of the instruction. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for exception conditions and branching is: Draw an if-then diagram from the translated sentence and compare with the source. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Moving an exception or weakening a modal verb can authorize the wrong behavior. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into equipment failures, exceptions, accessibility alternatives and correction notices. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 75: Only after review
Illustrative source: “Cast only after reviewing the displayed selections.” Unsafe interpretation: “Review your vote after it has already been cast.” Safer meaning: “Review is a prerequisite to the stated final action.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Only after is directional and cannot be reversed. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Write the temporal order with arrows. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for exception conditions and branching is: Draw an if-then diagram from the translated sentence and compare with the source. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Moving an exception or weakening a modal verb can authorize the wrong behavior. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into equipment failures, exceptions, accessibility alternatives and correction notices. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 76: If you cannot read the text
Illustrative source: “If you cannot read the instructions, ask about available access options.” Unsafe interpretation: “You must disclose a diagnosis before receiving information.” Safer meaning: “Offer language or accessibility help through the approved route.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. An access need does not automatically authorize invasive questioning. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check what information is actually required. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for exception conditions and branching is: Draw an if-then diagram from the translated sentence and compare with the source. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Moving an exception or weakening a modal verb can authorize the wrong behavior. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into equipment failures, exceptions, accessibility alternatives and correction notices. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 77: Before leaving
Illustrative source: “Check that casting is complete before leaving the voting area.” Unsafe interpretation: “You can complete the casting process at any later place.” Safer meaning: “The instruction concerns completion at the authorized point.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A time-and-place boundary may determine whether the vote enters the system. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Observe the actual completion indicator. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for exception conditions and branching is: Draw an if-then diagram from the translated sentence and compare with the source. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Moving an exception or weakening a modal verb can authorize the wrong behavior. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into equipment failures, exceptions, accessibility alternatives and correction notices. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 78: If no selection appears
Illustrative source: “If your intended selection is not shown, use the available review controls.” Unsafe interpretation: “Cast anyway and assume the system remembers.” Safer meaning: “Resolve a missing displayed choice through the authorized controls before casting.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. A missing selection is a state mismatch, not proof of a choice. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test whether Back or Edit is available. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for exception conditions and branching is: Draw an if-then diagram from the translated sentence and compare with the source. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Moving an exception or weakening a modal verb can authorize the wrong behavior. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into equipment failures, exceptions, accessibility alternatives and correction notices. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck exception conditions and branching after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
14. Mistakes in official explanations and diagnostics
The controlling principle for this family of instructions is straightforward: Explanatory prose must not invent procedural authority or contradict the actual ballot. The central translation test is equally practical: Find any claim in the explanatory note that changes a right, limit, remedy or status. The consequences are not only stylistic. Helpful-sounding additions are especially dangerous when they are outside approved instructions. Translators should read the instruction as a real interaction in FAQ pages, explainer videos, posters, machine help text and translated handouts, rather than as a detached sentence in a bilingual glossary.
Before editing, build a small state map showing the input, the permitted action and the next state. Compare this map with the equipment or approved paper that a voter will actually see. If the source material omits a crucial condition, raise a query to the issuer; translators cannot legitimately supply new electoral rules by guessing. The following six exercises teach how to locate that missing field and keep the repaired wording honest.
Case 79: All votes count
Illustrative source: “Every correctly cast vote is processed under the applicable rules.” Unsafe interpretation: “Any mark whatsoever counts as a valid vote.” Safer meaning: “Explain the process without promising validity for every possible mark.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Counting depends on the governing rules and the submitted choices. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Require a legal/editorial signoff on universal claims. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for mistakes in official explanations and diagnostics is: Find any claim in the explanatory note that changes a right, limit, remedy or status. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Helpful-sounding additions are especially dangerous when they are outside approved instructions. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into FAQ pages, explainer videos, posters, machine help text and translated handouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 80: You can change later
Illustrative source: “Review your choices before the final casting action.” Unsafe interpretation: “You may always change a cast ballot afterward.” Safer meaning: “Do not promise revisions after the system’s finality threshold.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Before and after are not interchangeable time scopes. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Locate the exact final action. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for mistakes in official explanations and diagnostics is: Find any claim in the explanatory note that changes a right, limit, remedy or status. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Helpful-sounding additions are especially dangerous when they are outside approved instructions. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into FAQ pages, explainer videos, posters, machine help text and translated handouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 81: Any helper can vote for you
Illustrative source: “Eligible assistance may be available.” Unsafe interpretation: “Anyone may cast the ballot on your behalf.” Safer meaning: “Describe only the authorized assistance route.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Personal choice and delegated assistance must remain distinct. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Check permitted assistants and their limits. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for mistakes in official explanations and diagnostics is: Find any claim in the explanatory note that changes a right, limit, remedy or status. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Helpful-sounding additions are especially dangerous when they are outside approved instructions. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into FAQ pages, explainer videos, posters, machine help text and translated handouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 82: Ballot rejected for silence
Illustrative source: “Unmarked contests may have a defined treatment.” Unsafe interpretation: “Leaving one contest blank necessarily invalidates every contest.” Safer meaning: “Describe blank selections without inventing whole-ballot rejection.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Undervotes and unmarked choices are not automatically fraud. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Test with one intentionally blank contest. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for mistakes in official explanations and diagnostics is: Find any claim in the explanatory note that changes a right, limit, remedy or status. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Helpful-sounding additions are especially dangerous when they are outside approved instructions. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into FAQ pages, explainer videos, posters, machine help text and translated handouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 83: Translation is official
Illustrative source: “This translation is for information unless expressly approved.” Unsafe interpretation: “The translator’s draft overrides the certified ballot.” Safer meaning: “State the document’s approved or explanatory status accurately.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Authority labels are operational claims, not promotional badges. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Verify document approval before publication. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for mistakes in official explanations and diagnostics is: Find any claim in the explanatory note that changes a right, limit, remedy or status. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Helpful-sounding additions are especially dangerous when they are outside approved instructions. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into FAQ pages, explainer videos, posters, machine help text and translated handouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Case 84: Every system same
Illustrative source: “Procedures differ across electoral authorities.” Unsafe interpretation: “The same correction method applies in every country.” Safer meaning: “State the limited scope of any illustrated rule.” These are short English-language semantic comparisons; they are not a translation into another language, and they do not replace the controlling ballot text.
The important contrast is not that one version sounds more professional. Even widespread practices can have exceptions. A translator who copies a familiar but unsuitable verb may change the physical gesture, the number of selections, the actor’s authority or the moment of finality. Begin by underlining the operative verb and its object. Then identify the smallest condition that would make the meaning different. In this case, that boundary is the detail to protect when transferring the message to another language.
A reviewer should now perform this case-specific diagnostic: Identify the controlling authority for the actual material. Record what the reader actually does, not merely what the reader says the instruction means. If two competent readers follow different routes, return to the source drawing, official manual or device behavior and check whether ambiguity already exists upstream. The correct editorial response may be a source clarification, an approved illustration, a supplementary accessibility cue or an escalated legal query—not automatically a longer translated sentence.
The broader quality gate for mistakes in official explanations and diagnostics is: Find any claim in the explanatory note that changes a right, limit, remedy or status. An unsafe interpretation is a real defect when it leads someone to perform a different action under the same evidence. Helpful-sounding additions are especially dangerous when they are outside approved instructions. A safe revision preserves the same choice opportunity, constraint, timing and confidentiality; it does not import a rule from a different type of ballot. Preserve the context in the termbase so that a future translator does not reuse this sentence in a system with another workflow.
For transfer practice, take this scenario into FAQ pages, explainer videos, posters, machine help text and translated handouts. Redraw the user path with three boxes: what the reader sees, what the reader is allowed to do, and what state the ballot reaches. Run a second test in which the reader makes a mistake, needs to repeat a step or asks for clarification. If the path or remedy changes across languages, stop release, note the mismatch in the issue log and revise only after the appropriate source owner has resolved the question. This is how a small lexical difference becomes a verifiable translation decision.
Family review. This section’s examples should be tested together, because errors can interact. A translation that protects the right verb but loses a selection limit is still unsafe. A layout that retains the words but places them beside the wrong option also fails. Recheck mistakes in official explanations and diagnostics after any changes to the design, software or governing procedure, and require a fresh approval when the workflow changes.
7. Failure modes to inspect before publication
Failure 1: Verb substitution
Selecting an option is translated as casting the whole ballot. The user thinks a reversible stage is final or a final action is merely navigation. Check the actual state transitions. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 2: Quantifier drift
Up to becomes exactly, at most becomes at least, or a number is copied incorrectly. Test a zero, partial and over-limit selection. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 3: Contest-wide scope error
A limit applying to one question is made to apply to the entire document. Mark the range of the restriction in the source and target. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 4: Negation loss
A short negative particle is omitted or detached by line wrapping. Use negative-case testing, not proofread-only review. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 5: Wrong conditional branch
If, unless and only after become interchangeable. Convert the text into a decision tree and check every branch. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 6: Actor confusion
An instruction for trained staff is presented to the voter. Role labels and permissions must survive translation. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 7: Premature finality
The word submit appears on the preview screen or the final cast control is called Next. Device tests reveal whether finality moved. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 8: Unapproved correction advice
The translator instructs the voter to erase, scratch or amend a paper without governing support. Replace with the verified procedure. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 9: Status collapse
A spoiled or surrendered paper becomes a rejected or counted vote. Track whether the paper was ever cast. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 10: Inferred authority
An example or unofficial guide is labeled as binding. Approval status must be visible and evidenced. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 11: Privacy leakage
The new instruction tells the person to show or announce their selections. Confirm who is permitted to know and why. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 12: Assistance takeover
The translator converts an offer of help into permission for another person to decide. Preserve voter choice and helper restrictions. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 13: Name substitution
A proper name is translated or transliterated inconsistently. Check the certified option list. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 14: Screen-reader mismatch
Visible text and audio label have different limits or option names. Compare modalities action by action. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 15: Icon contradiction
An icon suggests confirmation while the translated text says navigation. Test without relying on familiar icon assumptions. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 16: Truncated quantity
A narrow mobile screen wraps or crops the word ‘not’ or the digit of a limit. Inspect device breakpoints and print layouts. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 17: Mixed-script order change
Right-to-left rendering rearranges names or numerals without authority. Compare the underlying option identities. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 18: Glossary false consistency
One fixed translation for ballot or mark is applied in a context where a different operational sense is required. Use context-tagged terminology. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 19: False universalism
A specific election law is described as global practice. State the evidence’s jurisdiction and require local authorization. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 20: Unsupported certainty
An ambiguous source is guessed into precise advice. Record the open question and withhold the affected text until the owner resolves it. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 21: Automated translation hallucination
Machine output invents a correction or helper procedure. Independent human technical review must catch added requirements. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 22: Translation-only signoff
A fluent language reviewer approves without a ballot-process specialist. The release needs both semantic and operational checking. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 23: Outdated specimen
The translated sample matches an old ballot but not the current artwork. Version locks and checksums catch mismatched assets. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
Failure 24: Missing fallback
A device error message explains what failed but not how to seek authorized help. Test the recovery interaction, not just normal use. Reconstruct the source instruction as actor, object, permitted action, limiting condition and resulting state. Then ask whether the translated sentence would cause exactly that transition. Record an observed test, a source citation and a named reviewer. If a discrepancy affects rights, secrecy, permissible selections or final submission, escalate it as critical rather than treating it as a minor stylistic preference.
8. Translation laboratories for editors, reviewers and advanced learners
Workshop 1: A two-page selection-limit test
Create a mock ballot with two pages and five total allowed selections. Ask the reader to mark two on the first page and four on the second. The intended diagnosis is a total-limit violation, not a per-page exception. Repeat with a right-to-left layout and check where the number appears. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 2: A ranked-choice comprehension test
Show three options and ask the reader to express first, second and third preference. Present a deliberately ambiguous translation that tells the reader to ‘mark three favorites.’ The reviewer must explain why three approvals do not preserve ranking. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 3: A finality threshold test
Create an interactive mockup with Choose, Review, Back and Cast. Ask a tester when a selection may still be changed. Swap Review and Cast in the target language to demonstrate how a single label makes the action unsafe. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 4: A spoiled-ballot recovery test
Give a voter an intentionally erroneous marked sample. Supply a translation that says only ‘discard this.’ The reader should identify the missing authorized procedure for handling the old paper and requesting a replacement. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 5: An assistance autonomy test
A helper is asked to read instructions to a voter. Insert the mistranslation ‘the helper selects the preferred candidate.’ Correct it by separating reading assistance from the voter’s decision and any authorized physical marking. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 6: A secrecy sleeve transfer test
Present two illustrations showing a ballot inside a privacy cover in opposite orientations. Test whether the localized caption points to the correct orientation and does not expose the marked face. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 7: An audio-screen mismatch test
The visual text says ‘vote for up to two’ and audio says ‘choose two.’ The reviewer must demonstrate that the difference affects a voter choosing only one option. Correct the audio text and retest. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 8: A referendum polarity test
Use an invented question asking whether to remove a ban. Have a tester explain the meaning of Yes and No without using party labels. Any reversal of the ban’s status exposes a scope or negation problem. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 9: A candidate-name consistency test
Provide a certified multilingual list of four invented names. Compare ballot artwork, audio prompts, large print and translated instructions. The task is matching identities, not creating elegant new spellings. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 10: A device-error recovery test
Display a fictional notice ‘ballot returned: see official.’ Ask whether the translated text incorrectly says ‘vote counted.’ The correct test separates the mechanical event from the authorized follow-up. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 11: A multilingual line-wrap audit
Place a negation and a numeral in a narrow column and enlarge font size. Inspect whether the visual grouping survives. Source and target words may be right while the rendered instruction is functionally wrong. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 12: A jurisdiction-transfer test
Place two fictional authorities side by side, one using a replacement ballot and another using a different approved correction route. The translator must preserve each procedure separately, not apply a single global instruction. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 13: A helper confidentiality test
Create a training scenario in which an authorized assistant knows a voter’s preference. Ask reviewers to rewrite the helper instruction so it protects confidentiality after the assistance is completed. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 14: A blank-contest test
Mark one contest, deliberately leave another blank and compare two translations. Reject the version claiming that every blank contest necessarily invalidates the entire ballot unless the actual rules expressly require that outcome. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 15: A bilingual option-alignment test
Print six candidate rows with labels in two languages. Move one translation down a line and ask testers to identify which candidate the mark selects. Correct layout, not just wording. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
Workshop 16: A pre-release decision-table test
For ten instructions, record actor, object, condition, action, next state, secrecy implication and evidence. Require a reviewer to sign each row. Missing fields become tracked release blockers rather than unspoken assumptions. Repeat the test with a reader who has not seen the English source. Ask the reader to perform the action on a simulated ballot, describe their expectation of what happens next, and identify whom they would ask if the action failed. Compare the observed path with the approved intended path. Mark each discrepancy as terminology, grammar, layout, accessibility, workflow, privacy, or source ambiguity; then revise and run the same task again. This keeps the exercise tied to behavior instead of self-reported confidence.
9. Quality-assurance release gates
Gate 1: Source governance
There is one identified controlling version of the ballot, manual or UI state. No translator has invented an unverified legal rule. The evidence should identify a specific artifact and an observed or documented decision. A statement such as ‘translated correctly’ is insufficient unless a later auditor can reproduce what was checked. Any unresolved error affecting a permissible action, privacy or final vote submission blocks publication of the affected material.
Gate 2: Terminology governance
Mark, select, review, cast, spoil, surrender, replace, assist, and count have documented context-specific meanings. The evidence should identify a specific artifact and an observed or documented decision. A statement such as ‘translated correctly’ is insufficient unless a later auditor can reproduce what was checked. Any unresolved error affecting a permissible action, privacy or final vote submission blocks publication of the affected material.
Gate 3: Quantity preservation
All maxima, minima, exact requirements, rankings and contest-specific limits have been stress-tested. The evidence should identify a specific artifact and an observed or documented decision. A statement such as ‘translated correctly’ is insufficient unless a later auditor can reproduce what was checked. Any unresolved error affecting a permissible action, privacy or final vote submission blocks publication of the affected material.
Gate 4: Stage preservation
Review, correction, replacement and final casting steps remain separate where the official process separates them. The evidence should identify a specific artifact and an observed or documented decision. A statement such as ‘translated correctly’ is insufficient unless a later auditor can reproduce what was checked. Any unresolved error affecting a permissible action, privacy or final vote submission blocks publication of the affected material.
Gate 5: Names and labels
Option identities, candidate spellings, ballot question numbers, answer labels and official names align. The evidence should identify a specific artifact and an observed or documented decision. A statement such as ‘translated correctly’ is insufficient unless a later auditor can reproduce what was checked. Any unresolved error affecting a permissible action, privacy or final vote submission blocks publication of the affected material.
Gate 6: Layout and accessibility
Every supported print, digital, audio and tactile presentation has been checked for functional equivalence. The evidence should identify a specific artifact and an observed or documented decision. A statement such as ‘translated correctly’ is insufficient unless a later auditor can reproduce what was checked. Any unresolved error affecting a permissible action, privacy or final vote submission blocks publication of the affected material.
Gate 7: Secrecy and consent
No translation unnecessarily requires disclosure of a selection or expands helper authority. The evidence should identify a specific artifact and an observed or documented decision. A statement such as ‘translated correctly’ is insufficient unless a later auditor can reproduce what was checked. Any unresolved error affecting a permissible action, privacy or final vote submission blocks publication of the affected material.
Gate 8: Human review
A trained language reviewer and a process owner have approved high-risk passages independently. The evidence should identify a specific artifact and an observed or documented decision. A statement such as ‘translated correctly’ is insufficient unless a later auditor can reproduce what was checked. Any unresolved error affecting a permissible action, privacy or final vote submission blocks publication of the affected material.
Gate 9: Observed users
Representative testers have completed practical tasks without relying on the source language. The evidence should identify a specific artifact and an observed or documented decision. A statement such as ‘translated correctly’ is insufficient unless a later auditor can reproduce what was checked. Any unresolved error affecting a permissible action, privacy or final vote submission blocks publication of the affected material.
Gate 10: Release evidence
The release contains version/date, approved wording, screenshots, issue log, reviewer names and rollback route. The evidence should identify a specific artifact and an observed or documented decision. A statement such as ‘translated correctly’ is insufficient unless a later auditor can reproduce what was checked. Any unresolved error affecting a permissible action, privacy or final vote submission blocks publication of the affected material.
10. What authoritative sources can and cannot establish
The Election Assistance Commission’s glossary of election terminology explains the importance of consistent terms but also warns that procedure and terminology differ among jurisdictions. This guide uses such reference material for distinctions and questions to ask. It does not treat the glossary as a universal election code.
The EAC guidance on designing polling-place materials describes clear language, visual hierarchy, paired language presentation, instructions and accessible navigation. The April 2026 voter-education design toolkit provides a practical model for preparing accessible public information. The October 2026 language-access resources include election-terminology glossaries and translated educational materials. These are examples of contemporary public-sector resources, not binding requirements for every voting system.
The ACE Electoral Knowledge Network discussion of spoiled ballots makes the distinction between controlled returned ballots, replacement processes and ballot accountability concrete. Its procedural discussion supplies an excellent diagnostic model, but the exact handling rule must still be established for the election in question. For assisted voting and secrecy, see ACE’s explanation of assisted voters at voting stations.
The published ISO 11669:2024 translation-project guidance supports needs analysis, risk assessment, agreed specifications and translation workflow planning. ISO 5060:2024 concerns evaluation of translation output, including analytic assessment of error types. Neither text should be misrepresented as prescribing electoral law or certified voting-equipment behavior. The editorial method in this article applies their general discipline to a high-consequence translation environment.
Source check: November 11, 2026. Election materials and guidance can be updated. For actual voting instructions, verify the current authority, exact approved specimen, software version and applicable rules as of the election, not merely the publication date of a reference article.
11. Frequently asked questions
FAQ 1: What is ballot-marking translation?
It is the task of preserving instructions about how a person records a choice on a paper or digital ballot, including format, numbers, sequence, permissions and the visual or tactile target. Its success is tested by whether the reader performs the same approved action, not by word-for-word similarity. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 2: What does a spoiled or spoilt ballot mean?
In many systems it refers to a ballot rendered unusable or returned under an official procedure, often due to an error before casting. The precise classification and replacement rules vary. It should not casually be called a counted invalid vote. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 3: Are spoiled and invalid ballot interchangeable?
No. A spoiled ballot returned before casting and an invalid or rejected ballot assessed under counting rules may have different actors, timing and consequences. Some local laws use terms differently; confirm the official source. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 4: Does replacing a spoiled ballot give someone another vote?
A properly authorized replacement process is designed to substitute a usable ballot while maintaining records, not to grant an additional valid vote. The exact issue and surrender procedures must come from the responsible authority. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 5: Is a blank contest an error?
Not necessarily. Some voters intentionally leave a contest unanswered. The treatment depends on the relevant voting system. Do not rewrite a blank choice as misconduct or automatic whole-ballot invalidity. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 6: What is an overvote?
Typically it means that a contest has more selections than the allowed number. The legal effect and correction procedure depend on the election rules and equipment. Translation must retain the relevant limit and contest scope. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 7: What is an undervote?
A term often used when fewer choices are selected than the allowed maximum, including a blank single-choice contest. A translator should not present the term as a moral judgment or universal error. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 8: What does vote for up to two mean?
It imposes an upper limit of two selections; it is not equivalent to ‘you must select two.’ If the governing ballot adds other requirements, translate those separately rather than inferring them from the limit. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 9: What is the difference between selecting and casting?
Selecting chooses an option; casting completes the system’s defined submission step. There may be a review or correction opportunity between them. Verify the actual workflow, especially for digital interfaces. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 10: Can a translator change a ballot question to make it easier?
A translator may propose a clarified wording for official review, but cannot change the legal proposition, scope, negation or option identities independently. Controlled question wording requires explicit authorization. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 11: How do you translate a ranked-choice ballot?
Preserve the order conveyed by first, second and subsequent ranks, and the exact permissions about skipped or repeated numbers. Do not describe sequential ranking as approving several favorites. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 12: How do you translate a write-in field?
Preserve the authorized field label and instructions while using the authority’s approved candidate naming and script policy. Translators should not guess which person a misspelled or ambiguous entry means. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 13: What should a voter do after a mistaken mark?
The translation should point to the actual authority-approved correction or replacement process. It must not invent a universal instruction to erase, cross out, tape, discard or overwrite a ballot. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 14: What does it mean to keep a vote secret?
The voter’s selection should not be exposed to unauthorized people. Physical booth design, sleeves, audio privacy and helper confidentiality can contribute to secrecy, but particular rules vary. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 15: Does assisted voting allow a helper to choose?
Assistance should preserve the voter’s own choice. Where helpers may read, explain or physically mark a ballot, the authorized scope, required consent and confidentiality obligations must be checked. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 16: Does an accessibility translation need a different selection limit?
No. Accessibility presentation should preserve the same underlying choices and constraints. Differences in modality do not authorize changed numerical permissions. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 17: Is machine translation acceptable for election information?
It may support drafting or comparison, but high-consequence public instructions need qualified human linguistic and procedural validation. Automated output can invent rules, misread modifiers or change the finality of controls. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 18: Should candidate names be translated?
Names and designations should follow the relevant authority’s official publication and transliteration policy. Avoid independent ‘improvements’ that can obscure identity or create inconsistent options. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 19: Can one glossary be reused globally?
A shared conceptual glossary helps, but many terms need context and jurisdiction tags. A single unqualified equivalent for ballot, mark, cast or invalid can mislead in another procedure. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 20: What if the source English is ambiguous?
Escalate the ambiguity to the text owner, record alternatives, and pause release of affected high-risk content. Translation should not disguise a source defect through confident guesswork. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 21: How should visual instructions be tested?
Use the real design or approved prototype, observe people carrying out the intended action, and test misreadings. Compare arrows, boxes, line breaks, number placement and candidate alignment across formats. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 22: What makes a translation error critical?
A defect is critical when it can change a legal choice, permissible number of selections, access to a remedy, secrecy, voter autonomy or final submission. Severity must follow potential consequence, not grammatical elegance. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 23: Where should a translated notice go for approval?
Through the responsible election organization’s document-control and specialist-review process. Keep source version, translated artifact, issue log, accessibility tests and sign-off evidence together. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 24: Can the examples here be printed as official voting instructions?
No. They are fictional semantic diagnostics for learning. Official materials require local approval, legal review and alignment with the certified ballot and election system. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 25: How does this guide connect to the rest of translation?
It demonstrates a general method: recover the underlying action, preserve conditions and authority, test the result with representative readers, and document sources. The same discipline applies to safety labels, consent documents and procedural notices. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 26: Why not simply translate every word literally?
Literalism can break functional equivalence when verbs, grammar, writing systems and visual layouts encode action differently. The goal is to preserve the authorized decision, not mimic a particular word order. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 27: What is the fastest first check for a short voting instruction?
Underline the verb, its object, any number or negation, and the point at which the action becomes final. Those four fields frequently expose the highest-consequence mistranslations. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 28: What is the best second check?
Ask a representative reader to carry out the instruction on a mock ballot without seeing the source language. Observe behavior and compare it to the intended action, then investigate any divergence. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 29: What does a good termbase entry include?
The term, sense, controlled equivalent, actor, object, stage, usage example, source authority, prohibited confusions, reviewer and date. Context prevents a formally consistent term from becoming systematically wrong. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 30: How is this different from voter registration translation?
Registration concerns eligibility, the roll and assigned access before a ballot is issued. This chapter concerns meaning and privacy during marking, correction and casting after issuance. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 31: How is this different from tally and certification translation?
Tally and certification concern how votes and results are accounted for after casting. This article focuses on the voter’s authorized actions and the meanings of instructions before and during casting. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
FAQ 32: When should a translated instruction be withdrawn?
When it demonstrably alters an authorized action or no longer matches the approved ballot or device. Follow the owner’s change-control and replacement procedure, document affected artifacts, and do not rely on silent online edits to fix distributed print materials. For a real election, the controlling legal language and voter-facing approved materials always take precedence over a general educational example.
12. Transfer the method beyond elections
The six-field meaning ledger also applies to medical-device controls, consent interfaces, financial authorizations, classroom assessment instructions, health-and-safety labels and software confirmation screens. In each domain, a person may select an option before formally submitting it, may correct an error only in a defined window, or may need help without surrendering decision authority. Treat words such as approve, confirm, cancel, revoke, replace and finalize as state transitions, not decorative verbs.
Consider an online consent form: checking a box, reviewing disclosures and submitting consent are separate stages. The same distinction prevents a review button from becoming a binding acceptance. Consider a medication instruction: a number governs the exact action and cannot be softened into an approximate suggestion. Consider an accessible application: changing language or using audio mode should not silently alter the underlying options. Translation quality becomes observable when these constraints survive each transfer.
For vocabulary-building exercises on scope, negation, precise verbs and contrasting terms, use the protected Vocabulary Mastery Learning Hub. For the grammar and meaning mechanisms behind instruction wording, the protected How English Works | Translation and Meaning Across Languages offers a separate route. These owners supply language foundations; this chapter remains a specialist translation diagnostic rather than rewriting their lessons.
13. Next routes and editorial checklist
- Translation architecture: Master Art of Translation | The Complete System for Moving Meaning Between Languages — return to the complete meaning-transfer map.
- Previous reserved-tens child: Polling-Day Check-In, Voter Identification and Ballot-Issuance Fidelity — understand the step before marking.
- Terminology and precision: Vocabulary Mastery Learning Hub — strengthen distinctions between close words.
- Grammar and semantic structure: How English Works | Translation and Meaning Across Languages — study how conditions, scope and intent are encoded.
- Practical final gate: can a reader independently perform the same approved action in every language and supported format without losing secrecy, choice or a lawful recovery path?
Before any high-consequence translation is released, ask: Have we identified the actor? Have we preserved the action and object? Are every number, negation and condition unchanged? Does the translated instruction retain the same stage of finality? Have we verified the actual paper or interface? Does any sentence invite disclosure of a private selection or authorize a helper to decide? Has a specialist reviewed the text and has a real task test demonstrated equivalent behavior? A translation that passes these tests has moved beyond fluent wording toward accountable meaning.
The essential principle is simple: translate the decision the reader must make, protect the boundaries around it, and test the action that follows. Words are the visible surface. The work of translation is to keep the underlying choice intact.
Reference chapter EDKSG-TRANS-MASTER-WORLD-2220. Published as a new independent world-facing guide; fictional exercises are not official election directions. Sources checked November 11, 2026.