Bilingual identity map

Freeze identity before choosing a target form

A translated spelling should not become a new person. Stable IDs preserve the connection between source and target forms even when an editorial decision changes. Enter one ID per identity and keep source form, target proposal and review state separate. A missing target is unresolved, not permission to copy a random name into the slot. This screen does not machine-translate names.

Collisions can appear after adaptation

Distinct source names may collapse into one target nickname or transliteration. The tool flags equal normalized target strings. That is a narrow comparison, not a full cultural review. Read the target edition’s actual dialogue forms as well as the cast list. A collision may be intentional, but its intended function needs to survive the change of language.

Record the function of a clue

A name can carry initials, a pun, a private nickname or a graphic clue. The clue-token field records an explicit source token and proposed target token. A mismatch or missing token is a prompt for redesign. Equal tokens do not prove a translated puzzle is fair: the surrounding evidence may still be inaccessible. Keep the clue’s narrative job in a separate editorial note.

Fictional example: the theater program

A localized theater program uses a stage name that unintentionally matches the concealed civilian identity. The original audience learned that connection much later. The translator revises the program credit and records a new path for the eventual recognition. Editing only the dialogue would leave the premature reveal visible in the artwork.

Compare costs rather than seeking one universal strategy

Retaining a form preserves a recognizable identity but may lose a semantic clue. Transliteration can preserve an intended sound while changing initials. A functional adaptation may restore a joke or deduction while requiring a new prop. List those consequences openly. The right decision depends on which function matters in this story, not on a blanket rule that every name must be translated or retained.

Hand off a decision, not just a string

After choosing a target, review credit text, signs, handwritten notes, subtitles, recordings and glossary entries. Preserve older spellings in the project history instead of silently replacing the evidence trail. A review status is supplied editorial metadata; the tool cannot certify that a native-language editor has approved it. Only complete, reviewed editions should be treated as publicly equivalent localized pages.

Measure a target-edition surface

The target-surface worksheet takes character ID, label, inner width, font size and local font family. It measures the target form belonging to that stable ID, keeping the proof attached to the translation decision. An empty proof worksheet is unreviewed, never passed. Save earlier and current mappings as separate private projects to restore either version without replacing character IDs. A fit result remains a typography observation, not acceptance of the translation or its clue function.

Treat the map as a production decision

A bilingual identity map has two responsibilities. It keeps one person linked across editions, and it records what a naming decision changes in the story. A list of translated strings can satisfy the first responsibility while failing the second. Initials, private address, a written signature and a spoken cue may all depend on a form that does not survive unchanged.

Start with stable IDs, not matching display strings. If the target spelling changes, the person should not become a new record. If two people share a name, they should not be merged merely because the strings match. The workbench exposes equal normalized target forms as a collision to review. It does not decide whether the resemblance is intentional or whether the result is natural in the target language.

Write the main function beside each difficult name. Is it a normal reference, a joke, an identity clue, a family marker or a production credit? One name can have several functions, but they should not be collapsed into the vague instruction “preserve the meaning.” Decide which function the scene requires and which losses can be explained or repaired elsewhere.

Reproduce an actual change request

Enter a previous mapping and a new mapping under the same ID. In the asset table, list four affected surfaces: a prop, a dialogue line, a credit and an audio cue. Give each an asset ID, the person’s ID, a supported asset type and its current target form. If the current form differs from the new target, the report creates an edit item. The revision panel also records the old-to-new mapping.

For example, suppose an earlier transliteration appears on a drawn badge, while the approved form has changed. Editing the script is not enough. The badge needs an art change; the spoken cue may need a pronunciation discussion; the credit may need a new proof. The report does not edit those files or claim to have listened to the audio. It makes the work visible so the responsible collaborators can perform it.

An unknown person ID is a conflict, not a request to guess whom the asset belongs to. Correct the mapping or the manifest. A missing target remains unresolved. Those failures help prevent a polished production handoff from silently linking a prop to the wrong character.

Original fictional scene: the program in the foyer

The translator noticed the problem on the theater program, not in the dialogue. The stage credit used the same target form that the later chapter would reveal as the performer’s civilian name.

In the source version, the two forms remained distinct. The audience saw a performer before they learned who had paid for the production. In the localized proof, the answer was printed beside the ticket price.

The editor suggested leaving the program untranslated. The designer pointed to the rest of the page: the program was meant to be read, and the letter concealed in its fold depended on its ordinary appearance.

“Then we cannot simply hide the page,” the translator said.

They tried retaining the stage name, then a closer sound-based spelling, then an approved public form that did not collapse into the civilian identity. Each solved one problem and moved another. One weakened a later initials clue. Another lengthened the credit enough to crowd the fold where the letter would be visible.

The team chose a public form and rewrote the later clue around a previously established private salutation. The designer revised the program. The voice team received a new pronunciation note. The editor checked the chapter where the audience was supposed to make the connection.

In the final scene, a character opened the program to find the letter. They learned what the story intended them to learn, not what the translation had accidentally printed early.

This is an original fictional production example. The supplied target choices are not asserted to be preferred real translations. Its point is that a naming revision can require a new clue path and several asset changes, not just a replacement entry in a glossary.

Compare three strategies through their consequences

Retention preserves a visible source form and may help continuity across editions. It may also leave an intended semantic association unexplained. A sound-oriented adaptation can support pronunciation while changing initials and graphic patterns. Functional adaptation can preserve a joke or deduction but may require new wording, props or performance cues. None is a universal default for every project.

Compare them using the same scene facts. What must the reader know before the letter is found? Which characters can recognize the salutation? Where is the clue first established? If one strategy preserves a pun but exposes an identity, that is not a minor stylistic loss. It changes the order of knowledge on which the plot depends.

The clue-token table is deliberately narrow. It compares the declared source and target tokens and flags missing or unequal values. Equal tokens do not prove that the surrounding puzzle is solvable. Different tokens do not mean a good adaptation is impossible. Record the deliberate redesign and test the target scene rather than forcing a superficial match.

Build a target-language review packet

Give a translator or editor the identity map, approved readings, the function of each difficult name, and the passages where the name matters. Include what has already been rejected and why. A string without context invites the reviewer to solve the wrong problem: naturalness alone, for example, when the scene actually depends on a constrained initials pattern.

Keep creative associations separate from factual claims. An invented name can have an author-assigned meaning within a world, but that does not establish a real-language etymology. A sourced reading may be attested in one context without validating every possible combination or era. A bilingual handoff should preserve these limitations rather than remove them to look decisive.

Review status is user-supplied metadata. Entering reviewed does not create a reviewer or authenticate an approval. In a real production record, retain who checked the decision, which context they saw, and what remains unresolved. This website does not provide a hidden native-language certification service.

Prevent drift after the first volume

One accepted mapping should feed the dialogue, cast page, credits and production notes. Keep previous versions so a returning contributor can understand an older script without reviving an obsolete spelling by accident. When an author revises the source name, reopen the dependencies rather than treating the target form as permanently independent.

The same issue applies to recaps and glossaries. A later identity connection should not appear in an early-volume cast page simply because the master map contains it. This mapping tool is a production document. Use the observer and chapter workflows to prepare reader-facing extracts with the correct knowledge scope.

Do not count an English page and a Japanese page as two independent tools. They are complete language editions of the same user task only when both their interface and their substantive guidance work. A translated heading above an English explanation is not a finished localization. The project keeps language readiness separate from route count for precisely that reason.

Finish with a scene-level regression

After accepting a map revision, reread the target introduction, the clue scene and the reveal without the source text beside them. Can a target-language reader identify the same people and make only the intended inferences? Check the prop image, subtitle and audio handoff as well as the narrative paragraphs. A successful string export is not evidence that these assets have already been revised.

Record losses openly. If a source pun cannot be preserved without harming an identity reveal, say which function you prioritized and how the other information is handled. That is an editorial decision a collaborator can assess. Claiming that nothing was lost would make the handoff less useful, not more faithful.

Counterfactual: preserving the letters while losing the person

If the translator kept the source initials at all costs, the program could preserve its visual clue while making two target identities collapse into one. If the translator instead changed the initials without revising the prop and reveal, the eventual solution would appear unsupported. Compare both revisions against the stable identity map: preserve distinct people first, then record which prop, line or earlier introduction must carry the clue's function. This is a choice about information the reader can use, not a demand that every edition share identical letters.

Is a target collision always an error?

Not necessarily. Two public forms may intentionally match, but the target edition must support that intention. Record whether the overlap is a clue, a social convention or an unwanted collapse. The tool preserves stable IDs and flags the shared form without merging the people. Test the actual dialogue references; a glossary distinction alone may not be available where the reader needs it.

Why list audio when the change is only a spelling?

The team needs to decide whether the approved pronunciation changed too. It may not have, in which case the audio note can close with that explicit reason. It may affect recording cues, subtitles or spoken credits. An asset list makes the decision visible rather than assuming either that every spelling revision requires a retake or that none can matter to performance.

Can equal initials certify a localized puzzle?

No. Matching clue tokens checks one declared condition. The target passage must still teach the rule, preserve access to evidence and avoid disclosing a hidden identity early. A different token can also be part of a good functional adaptation when the clue is deliberately redesigned. Keep the reasoning and the revised asset list with the mapping, rather than demanding surface equality at any narrative cost.

When should a locale be linked as a finished alternate?

Only when its actual interface, substantive guidance, examples and review state support that status. A translated title above unfinished source-language material is not a complete edition. The development site can offer language navigation while withholding public indexing and language-alternate metadata. Tool counts should still use the underlying user task, not add another tool for each localized URL.