Seminar II · Translating the visual language

✦ A methods seminar · original sources required

Literal equivalence is a research hypothesis

A translator encounters Wappenrolle, armorial, wapenregister, and herbarz and puts “register of arms” into each English record. The resulting table is easy to read, but it may have made unlike documents appear equivalent. Each term can concern a collection of arms while carrying a different relationship to compilation, recording, and institutional authority. We have to establish those relationships before deciding how much the English label can safely compress. Fluency is useful only if the distinctions survive it.

A translator’s responsibility is to preserve the meaning of the source, including its institutional relationships, not to find the nearest familiar English word. When a medieval term has no exact equivalent, retain the original word at first occurrence, define its scope, and explain why the convenient English analogue falls short.

Consider Bürgerwappen. Translated carelessly, it could become ‘city coat of arms’, but that is more likely Stadtwappen. Bürgerwappen may denote the arms of non-noble citizens or families. A mistaken translation changes the bearer of the coat, turning a personal identity into a municipal one. This is not stylistic: it is a substantive historical error.

A comparable problem appears in the Dutch language. Gemeentewapen is municipal arms; familiewapen is family arms. Wapendiploma appears in public-body granting procedures, but not every ‘certificate of arms’ in a private commercial setting carries equivalent legal force. Institutional vocabulary must travel with its procedural context.

Legal language cannot be translated by analogy alone

The Flemish 1998 decree distinguishes erkenning (recognition) and verlening (grant). An English-language summary that calls both ‘registration’ erases a legal distinction. Similarly, a German association’s Eintragung (entry) in its own voluntary register must not be described as an English sovereign’s grant.

A translation apparatus should define the speech act of the original document: is the author describing, approving, recording, advising, ordering, proclaiming or asserting? A government act can grant an emblem; a catalogue may merely report that a document exists. Confusing these actions is among the most consequential translation failures in comparative heraldry.

Legal terminology is historical. The same word can acquire different effects following administrative reforms, constitutional changes or the repeal of older statutes. Date the applicable text and note whether the linked page is a consolidated later version or a reproduction of the original act.

Working with vernacular descriptions of shields

Spanish partido and cortado can describe vertical and horizontal partitions; cuartel may refer to a quarter or a compartment. Translate the geometry being described, not only the dictionary form of each noun. The Sotillo de la Ribera notice is useful precisely because it specifies six shield sections: the literal ‘six quarters’ requires the reader to understand that the Spanish term is used more flexibly in that context.

A controlled parser can help test the geometry of a proposed translation. If the instructions specify six compartments, we can inspect whether the rendering has six. But the program operates on the instructions we supplied; it has no independent access to the meaning of the Spanish source. It can therefore draw a misinterpreted shield quite successfully. Checking the result against source-language usage and comparable historical witnesses remains part of the translation, even when the software reports no error.

The distinction becomes even clearer with Japanese 家紋 (kamon). One may explain that it is a family emblem in a comparative course, but forcing the entire Japanese system into English medieval blazon grammar imports a set of social and legal assumptions the source never made. It is legitimate to compare distinct symbolic systems. It is not legitimate to call that comparison a literal translation.

What to publish, and what not to publish

The project should maintain at least four statuses:

1. Original editorial synopsis — new English prose explaining the nature and implications of a source, linked to that source. This is what our current atlas generally provides.
2. Translated brief excerpt — a limited source-language extract plus a properly attributed English rendering, with the location and reuse rights stated.
3. Public-domain/permission-based critical translation — full translated text with edition, folios, editorial notes and publication rights verified for the actual source expression.
4. Unverified working translation — a private research aid awaiting language and domain review, not confidently published as a final scholarly edition.

Contemporary legislation can be freely accessible on a government website without implying that every photograph, editorial explanation or modern scholarly publication nearby is free to redistribute. Digitization does not itself extinguish a modern author’s copyright. Separate the underlying historic text, the modern transcription, the translation, the facsimile and the photograph.

A translation record

Every substantial translation should record original title, named author if known, original date, edition or version, holding institution, shelfmark and folio or section, source language, script, source URL, translation date, translator or method, editorial interventions, unresolved ambiguities, reviewer, rights basis, and the version history of both source and translation.

This may look pedantic. It is also what allows another researcher to correct an error without reconstructing the entire project from the beginning.

From manuscript hand to translated argument

The transcription desk now gives this seminar a folio-level exercise. Its French witness is Getty’s Hours of Simon de Varie, Ms. 7, folio 2, with a reproduced CC0 full-leaf image. Its Portuguese witness is the Torre do Tombo’s Livro do Armeiro-Mor, book 19, folio 87, located in the archive’s presentation on PDF page 23. The latter remains a source link because image reuse has not been verified. Read the paired workshop before beginning the edition.

Do not begin with English. Begin with the form of the witness and a policy for representing it. A diplomatic transcription describes a chosen level of the written form. A normalized reading records editorial changes intended to make the language legible. A translation conveys the established reading’s meaning. Keeping them separate permits a reviewer to locate disagreement: an uncertain letter, a proposed expansion, and an English idiom are different decisions.

Prepare at least two apparatus entries. Each should identify a locus, an observed form, a proposed reading, a review status, and the supporting comparison or reason. A statement that a form was checked should name the check. It is not equivalent to specialist approval. Likewise, uncertainty that survives into the normalized reading should survive into the translation where it affects meaning.

For the French leaf, distinguish your inspection of the motto scrolls from the museum’s modern reference readings and its report of partly overpainted border arms. A dated manuscript does not date every visible intervention. For the Portuguese leaf, distinguish the modern archival caption from the actual hand. If access to the image fails, a caption-only study is possible, provided its conclusions concern that caption rather than an uninspected original inscription.

The file format introduces another editorial question. The desk’s portable JSON can restore the separate fields and apparatus; its Markdown export provides a readable report. Neither replaces the source image or confers scholarly approval. Preserve versions when a changed reading affects a translation. State what changed, why it changed, and which claims depend on the earlier version.

Edition assignment: Submit one short located inscription with all three text layers and a 900–1,200-word editorial report. Include a witness inventory, notation policy, two reasoned alternatives, source and image rights, reviewer status, and a concrete unresolved check. Then compare the mediation of the French and Portuguese dossiers in 250 words. The workshop’s rubric allocates credit to witness control, editorial policy, apparatus and translation, and proportionate interpretation.

The institutional sources are Getty’s folio 2 record, its IIIF manifest, and Torre do Tombo’s Portuguese presentation, page 23, consulted 9 October 2026. This project supplies original teaching analysis; an independently reviewed manuscript transcription remains a separate scholarly task.

Collation before interpretation

The comparison bench extends the edition into a study of two located readings. Its default French pair consists of folios 2 and 2v from one manuscript, rather than two independent manuscript copies. Read the collation workshop and state what that relationship permits you to argue. A repeated inscription within a book and a separately transmitted copy answer different questions.

Begin by aligning actual locations, not convenient modern strings. A lower scroll, an upper scroll, and a collar are distinct occurrences. If you supply a doubtful form from a parallel occurrence, retain the doubtful observation and the dependence of your proposed reading. The parallel cannot then be counted as independent confirmation of the completed text.

Compare the diplomatic and normalized layers separately. The exact-text preview highlights case, punctuation, accents, and whitespace without normalizing them. A line break you entered can therefore produce a difference. A normalization policy can remove it. Explain which operation belongs to the witness and which belongs to your edition before drawing a philological conclusion.

Translation adds another source of apparent agreement. Two English renderings may coincide because the translators chose the same idiom; different renderings may express the same established French wording. Write an apparatus note connecting each interpretive choice to the relevant source-language passage. The amount of highlighted English text cannot measure fidelity.

Collation assignment: Supply three located comparison entries and a 600-word methodological argument. Include one difference affected by editorial policy, one agreement with a precisely limited implication, and one proposed check capable of changing a reading. Export the comparison with both source records and the original single-witness apparatus. Name the version used for the argument; imported drafts are editable copies, not automatic updates of the originals.

The second French opening is Getty’s folio 2v, accession 85.ML.27.2v. Its image has separately documented Public Domain Mark / PD-Art provenance, described in the workshop. The Portuguese dossier can support a comparison of source mediation, but its unrelated wording must not be treated as a variant copy of the French text.

Comparative language exercises

Take three records from the source atlas: a German DWR entry, the Dutch noble-register explanation, and a Spanish municipal arms decision. Prepare a one-page concordance of the verbs they use for recording versus granting. Distinguish institution, private status, and the legal effect stated by the source itself.

Next, use the National Diet Library’s Japanese guide to collect the terms 家紋, 紋章 and 文様. Find contexts in which these terms should not all be translated ‘coat of arms’. Clearly mark which judgments are your own interpretations and which are derived from native-language scholarship.

Graduate assessment: Produce a 500-word bilingual critical translation note using a short, rights-permitted excerpt with an exact source locator. Include two plausible English alternatives for one difficult term and explain why you chose the final rendering. If the material’s rights are uncertain, submit a private note or a new English synopsis without reproducing the original passage.

Key witnesses

• HEROLD, Die Deutsche Wappenrolle
• Hoge Raad van Adel, Adellijke heraldiek
• BOE, Sotillo de la Ribera decision
• National Diet Library, Japanese heraldic research guide