France: Armory, monarchy, and the great register

✦ A substantive research chapter

A registry is not the whole history of a practice

A French armorial image might survive in a seal, a manuscript, or a building. Each object places the design in a different activity: authentication, compilation, or architectural display. Medieval French armory circulated through courts, warfare, noble and urban society, and literature as well. To explain a particular use, we need its date, bearer, and source history. Those connections let us investigate local and royal interests without making one later register stand for the whole practice.

The surviving armorial is often an uneven evidentiary archive. One witness may identify a bearer in a roll, another may depict colours in a manuscript, and a third may preserve a seal with the same figure but none of the painted tinctures. Historians should ask whether a document describes an already-used device, records a transaction, or creates an administrative entry. Those are different evidentiary events.

French scholarship speaks of armoiries (armorial bearings), blason (the shield or armorial device, depending on context), and blasonnement (the formal act of describing arms). English-language readers sometimes treat blason and blazon as direct equivalents. They overlap, but translation should attend to the grammatical role the term plays in the actual sentence. When a cataloguer labels a volume blasons coloriés, it is describing painted armorial representations, not asserting that every painted page is a unique grant.

The documentary shock of 1696

The Bibliothèque nationale de France’s Gallica guide describes the Armorial général de France associated with an edict of 1696. The register’s scale is startling: more than 115,000 family entries and approximately 3,000 cities and communities, according to the BnF guide. The catalogued volumes are divided into different manuscript series, including textual descriptions and painted armorial images.

Why did the French monarchy assemble this vast corpus? This is not simply a story about protecting the reputation of an ancient noble caste. Registration was bound up with the fiscal and administrative logic of the late-seventeenth-century monarchy. A record generated under such circumstances can tell us a great deal about state practices, but one must resist treating each entry as a straightforward certification of noble status or of antiquity. Recorded does not mean ancient, and neither necessarily means noble.

That distinction changes how we interpret the extraordinary number of entries. Numbers in a government register are counts of the administrative population reached by that register, not counts of legally impeccable medieval armigers. A careful historian reconstructs the conditions of enrolment before making quantitative claims about the social distribution of arms.

Three manuscript series, three kinds of evidence

The BnF distinguishes textual register series from the illustrated blasons coloriés. A research project must specify which physical or digital witness was consulted. A painted shield is useful evidence of colour and composition as that witness rendered them, but may be later than the act of registration. A text entry may preserve the name and armorial description, but its spelling and wording may not match a modern edited index. An alphabetical table helps locate witnesses; it should not be silently substituted for them.

Consider a proposed dataset of heraldic charges from the Armorial général. If the sample contains only easily indexed coloured pages, it may disproportionately reflect what was painted and digitized rather than everything registered. If the study conflates a textual listing and a matching image as two independent arms, counts will be inflated. If it treats alternative spellings of a name as distinct persons, the error compounds. These are ordinary problems of source criticism, not something mysterious peculiar to heraldry.

A responsible dataset would contain at least an original-language name string, normalized lookup name, archival series, volume, folio or entry, observed blazon, transcription method, image source and confidence about reconciliation between witnesses. Its description would make clear how the source was sampled.

Authority and rupture

The French Revolution and subsequent regimes altered the symbolic and political setting of French arms. It is tempting to narrate this as a clean abolition followed by a complete restoration, but material usage and legal and institutional continuity do not necessarily move together. Buildings continue to bear carved shields even when the political status of their bearers changes. Civic emblems may be redesigned. Families may preserve armorial traditions irrespective of the reigning political theory.

At graduate level, the problem is not to pronounce one universal verdict on whether ‘French heraldry survives’. It is to ask who had authority in a defined period to grant, recognise or regulate a defined class of insignia; who used them in practice; what records survive; and what legal consequence any entry actually carried.

Read a French shield from its grammar

Imagine a teaching shield with a blue field, a gold chevron, and three silver stars arranged around it. A French description begins with the field, then identifies the chevron and the accompanying stars. This order gives the artist a construction sequence. The description is not a caption listing whichever objects happened to attract the writer’s attention first; the spatial relationships belong to the identity being described. A star on the chevron and a star beside it produce different arms even if an inventory counts exactly the same objects.

The distinction becomes especially useful with chargé and accompagné. A figure charged with another figure carries it on its own surface. A figure accompanied by others sits in a composition with them. In our invented shield, putting the three stars on the gold chevron would require a different description and invite a different contrast analysis. Silver stars against a blue field and silver stars against a gold chevron have different immediate backgrounds. Before announcing that a composition follows or breaks a tincture convention, establish which surface carries each figure.

French direction words likewise describe the shield from its bearer’s position. Dextre is the bearer’s right and appears on the viewer’s left in a frontal drawing; senestre is the reverse. A bande and a barre run in opposite diagonal directions. They cannot be exchanged merely because an artist thinks the composition balances better that way. The distinction survives a change of drawing style because orientation is information, rather than a decorative preference.

Now divide the teaching shield vertically. Parti describes that division; coupé divides horizontally. If a lion crosses the division and changes tincture with the background, the artist needs the relationship expressed by counterchanging. Describing two halves and then mentioning a lion without explaining its relationship to them leaves a consequential gap. Translation into English should preserve the operation, even when the French and English phrases have different lengths.

These examples are constructed exercises, rather than attributed historical arms. Their purpose is to make the language usable before the student encounters a manuscript whose spelling, punctuation, and abbreviations introduce additional difficulties. The Academy’s blazon lesson develops the general procedure. Here the task is to retain French distinctions rather than silently replacing them with the nearest English word and hoping the artist will work out the rest.

A fleur-de-lis can occupy several different roles

A stylized lily can be a charge in a shield, an element repeated across a surface, a badge-like decoration, or part of a border. The familiar form makes these occurrences easy to recognize, but recognition alone does not tell us whether they perform the same armorial job. A single lily carved into a furnishing does not provide the field, number, arrangement, or bearer needed to reconstruct a complete coat of arms. The motif is a useful clue; the composition supplies the identification.

Consider two invented blue shields. One carries three gold lilies in a defined arrangement; the other is strewn with gold lilies whose repetition continues towards the edges. The first requires the artist to retain the count. The second asks for a field treatment, commonly described as semé, whose visual grammar depends on repetition rather than a census of complete figures. An artist may draw different numbers of visible lilies in two realizations of a strewn field without necessarily changing the identity. Cropping such a field to three visible lilies does not turn its underlying specification into a three-charge coat.

The same distinction applies outside the shield. A textile covered with lilies can create a powerful association through repeated decoration while a small shield on that textile carries the more specific armorial statement. Their conjunction makes the garment legible at different scales. At a distance, colour and repeated motif organize the spectacle; close inspection can reveal the particular composition, lettering, or relationship to a wearer. A study that extracts only the small shield loses the work done by the larger surface.

This is a useful starting point for French dynastic heraldry because royal imagery circulated through objects and environments as well as formally bounded shields. The question is how a particular device identifies a ruler, a branch, a patron, or a claim in that setting. We can then investigate differences between branches through dated examples. A bordure, label, or changed tincture has to be connected with a particular bearer and period before it becomes an explanation of relationship. One cannot derive a complete succession system from the mere existence of a familiar lily.

Differences and combinations express relationships

A brisure changes an armorial composition to distinguish a related bearer or branch. Its value lies in the relationship it makes visible. For an invented family, suppose the earlier shield contains three stars and a later branch adds a red border. The second design preserves the recognizable core and changes the perimeter. The relationship is easy to see, but its historical explanation still needs the family’s documents: the same formal change could be associated with several different circumstances.

We should consequently separate two tasks. The first is descriptive: identify what changed and whether the changed feature belongs to the blazon or only to one drawing. The second is genealogical or institutional: explain why that change distinguishes these bearers. The first can often be answered through a comparison of the objects. The second requires names, dates, and the relevant relationship. A convincing account earns the connection instead of substituting a universal table of marks for the family’s history.

Marshalling introduces another operation. A quartered shield may combine arms associated with different inheritances or political claims, but its meaning depends on the composition’s particular circumstances. The four quarters are positions in a design, rather than four automatically verified ancestors. Repeated quarters can give a paired identity a balanced arrangement. An additional small shield can express another relationship. Counting fields and counting represented lineages are therefore different exercises.

Use the marshalling lesson to draw two alternative compositions for a fictional alliance: two shields placed side by side, and one quartered shield. Each arrangement can associate the parties, but it makes a different visual statement. Now attach dates to the hypothetical marriage, assumption of the combined display, and manufacture of an object showing it. The chronology prevents the artwork from doing impossible work as proof of every event at once.

An imperial armorial organizes a political system

The First Empire supplies a particularly useful comparison with the earlier general register. Henry Simon’s Armorial général de l’empire français, published in 1812, includes rulers, dignitaries, and classified cities, together with material on internal signs and external ornaments.¹ The book’s structure presents heraldic identity in relation to an imperial order. That is a contemporary printed witness to classification, rather than merely a modern collector’s assortment of attractive shields.

A classification can operate both inside and outside a shield. A reserved element within the composition associates a bearer with a category; an external ornament can communicate another part of the status. These layers need separate descriptions because they may vary independently. Removing the ornaments to create a simplified modern drawing can retain the shield while deleting the part of the original display that communicated its place in the system.

The Archives nationales preserves another layer: registers transcribing letters patent concerning conferral, confirmation, and transmission of titles and nobility.² Those acts must be compared with the separate armorial records. A person’s title, armorial description, and later succession may belong to related documents, but they are different administrative events. The printed classification helps us understand the visual vocabulary; the enrolled instrument helps establish what happened to the named bearer.

This relationship gives students a better historical question than whether Napoleon “brought heraldry back.” Which kinds of identification were being organized, through what acts, and with which internal and external signs? How did the display connect an individual distinction with the regime’s institutions? Once those questions are answered for specific records, broader conclusions about rupture and continuity have something solid underneath them.

Dupuch: the words constrain the drawing

The catalogue entry for the chevalier Dupuch, in register BB/29/1072, describes a blue field with a silver sword placed in the barre direction, gold mounting, accompanying stars, and a red lower area carrying the knights’ sign. It also records the livery relationship.³ Even this compact entry connects orientation, material colour, category, and use beyond the shield.

Treat the wording as a set of drawing instructions and record where it leaves questions. The sword’s diagonal is specified; its handle and blade cannot be recoloured as one undifferentiated object. The reference “of the first” must be resolved from the wording rather than from what a modern illustrator expects the stars to look like. A critical edition would compare the manuscript image, preserve the catalogue reading, and explain any proposed correction.

For a classroom exercise, change one feature at a time in a clearly fictional derivative: reverse the sword’s diagonal, move a star onto the blade, or remove the lower area. Compare the altered drawing with the original description. Each modification demonstrates why blazon is a specification of relations. The exercise becomes misleading if the derivatives are published as historical versions of Dupuch’s arms, so label the altered images as controlled examples and retain the actual record separately.

The livery statement also invites a productive extension. A shield’s palette can inform dress or other display without requiring every use to reproduce the entire achievement. Investigating that relationship needs evidence of the particular realization, but it begins with a real documentary link. The practical heraldic question is how an identity moves between media while preserving some features and adapting others.

Republican identification extends beyond the shield

The Presidency’s account of the Great Seal follows changes of design through successive regimes and explains the Republican composition established in 1848. It also records a difference between prescribed wording and the engraver’s placement of inscriptions.⁴ This is a national identification system whose history cannot be reduced to the inheritance of one armorial shield. Personification, lettering, attributes, and the seal’s institutional use all contribute to the result.

The comparison with blazon is instructive. A written specification constrains an object, but its realization can introduce choices or discrepancies. A historian needs both the normative text and the material witness when asking whether the object follows the prescription. Calling the engraving authoritative does not make every visible choice part of the original wording; quoting the decree does not make the completed object disappear. Their relationship is itself historical evidence.

A separate presidential account describes the lictorial fasces and related diplomatic and United Nations compositions, while distinguishing frequent use from formal official status.⁵ That distinction defeats the urge to assign one convenient modern picture to a box labelled “France’s coat of arms” and regard the problem as settled. National representation can involve several emblems, contexts, and decisions whose authority needs to be identified individually.

The wider lesson is that a republic can retain heraldic devices, create new public emblems, and use seals or other forms alongside them. The surviving French armorial tradition and the state’s contemporary identification practices overlap without becoming identical. A useful reference therefore studies family, civic, dynastic, and Republican displays through their actual mechanisms, rather than forcing all of them into one national shield.

A fifteenth-century leaf with a later history

The Armorial général is a large administrative corpus. A manuscript miniature supplies a different scale and kind of witness. Getty’s Simon de Varie Kneeling in Prayer, Hours of Simon de Varie, Ms. 7 (85.ML.27), folio 2, offers a located French leaf dated 1455 and assigned to Jean Fouquet in the collection record. Its leaf accession is 85.ML.27.2. The object study reproduces the full photograph under the CC0 declaration in Getty’s IIIF manifest.

Look at how the leaf organizes identity. A kneeling figure occupies the miniature; a woman supports a shield behind him; smaller armorial forms occupy the corners; lettered scrolls frame the scene above and below. Extracting one shield would discard much of that arrangement. A useful description identifies the relations among person, display, lettering, and devotional scene before translating the image into isolated heraldic components.

The museum identifies the kneeling figure and reports partial overpainting of the border arms. Those statements come from the modern catalogue. The photograph allows inspection of the present surface, but does not independently date every painted intervention. Keep the manuscript’s catalogue date and the history of a particular armorial form separate. A reconstruction of an earlier display needs its own documentary or technical basis.

The motto scrolls are a practical philological exercise. Use the transcription desk to record a short span of visible lettering, a normalized reading, and a working English translation. Explain supplied spaces and joined letter forms in the apparatus. Then compare the institutional reference reading without treating it as your own inspection of the hand.

This small edition also changes the questions you can ask of larger French registers. A normalized lookup name and a diplomatic name string are useful for different operations. An image’s present armorial form and an earlier recorded display need not be identical. A database must represent those relationships if its counts are to carry historical meaning. The general register’s scale makes disciplined witness control more necessary, not less.

Comparative exercise: Read the manuscript workshop and compare the French motto dossier with the Portuguese archival caption. Identify which claims arise from inspected lettering, modern identification, institutional dating, and reported material change. End with one proposed source that could settle a doubtful reading. Sources consulted 9 October 2026; specialist transcription and conservation review remain outstanding.

A second opening, a different evidentiary relationship

Getty’s Coat of Arms Held by a Woman and a Greyhound, Ms. 7, folio 2v, accession 85.ML.27.2v, supplies a second photographed opening. Inspect it beside folio 2 on the manuscript comparison bench. Record each occurrence by its actual position. A collar inscription and a scroll inscription remain different loci even if your normalized readings agree.

Because the openings belong to one book, this comparison investigates recurrence and arrangement within that manuscript. It cannot independently establish transmission through another copy. Nor does agreement in wording date the paint or prove that two occurrences were executed together. Keep those possible claims separate from what the present photographs show.

The collation workshop asks you to compare diplomatic readings, normalized readings, and translations in turn. Explain a difference that disappears through editorial intervention, then record an uncertainty whose proposed resolution depends on the parallel occurrence. The second image’s Public Domain Mark / PD-Art source record is retained separately from the first image’s Getty CC0 declaration.

Design laboratory: a French reading with a defensible result

Begin with the fictional chevron-and-stars shield. Write its French description, then ask another student to draw it without seeing your picture. Compare field, ordinary, count, arrangement, and immediate backgrounds before discussing style. If the second artist draws a rounder shield or sharper stars, determine whether the difference changes identity. If a star moves onto the chevron, identify the missing or misunderstood relationship in the words.

Next make two deliberately different versions: a strewn field with the same motif and a quartered composition combining two teaching coats. Explain which information each specification requires and why their counts mean different things. This exercise joins formal grammar with the interpretation problems encountered in registers: a repeated motif, an occurrence in a quarter, a bearer, and a manuscript entry are different counting units.

Finally choose a located historical record and replace the classroom description with a small documented study. Describe its composition, establish its institutional setting, and trace one realization in a different medium where evidence permits. The resulting account should explain an actual heraldic operation. Its bibliography supports that explanation; it should not become a substitute for the explanation itself.

Language laboratory

These equivalents are editorial guides, not a substitute for reading the French sentence in context.

Primary-source workshop

Follow the BnF’s genealogical portal into the armorial catalogue. Record the distinction among the description, the painted witness, and the index. Identify a locality for which the manuscripts can be traced and write a reproducible research note: what is the original item, where does it appear, how was the transcription verified, and what do the records not tell you about entitlement?

Graduate assignment: Design a sampling protocol comparing civic and familial entries in one provincial volume. Identify errors introduced by family-name normalization, delayed illustration, missing digital scans, and registration motivated by taxation. Propose a versioned relational schema rather than merely counting attractive painted shields.

Essential original-language sources

1. BnF / Gallica, Ressources généalogiques : Noblesse — institutional guide explaining scope and manuscript series.
2. BnF catalogue, Armorial général : Table des blasons coloriés — bibliographic finding aid, not the original register.

The primary next step is folio-level study. The general archival guide alone cannot support a claim about an individual armiger.