A displayed ornament, a recognized bearing and a proven right to use it are distinct claims. Ask which one each source can actually establish.
The medium makes claims possible
Heraldry is often taught as a set of abstract colored shapes. Yet much of our evidence comes from objects whose materials resist certain kinds of representation. An impression in plain wax preserves form but not the field’s original color. A stained-glass window turns painted tinctures into light. A monumental carving gives a shield physical depth and allows a sculptor to add attendant figures around it. An armorial book fixes the arrangement within a page and may impose a compiler’s scheme on otherwise independent achievements.
These aren’t incidental constraints. They affect what a historian can infer. The object does not merely transmit an identity; it filters and sometimes transforms it.
A seal is not a grant
The presence of an armorial design on a seal is evidence that the seal’s owner, maker or user chose to associate the design with a given authentication context. It may be highly important evidence of period usage. It is not automatically a surviving grant of arms. A seal can also preserve a crest, supporter-like figures or inscriptions omitted from another witness.
The Bibliothèque nationale de France’s manuscript catalogue describes a shield with helmet, gold-and-red mantling, a wing-shaped crest and a motto in a manuscript associated with an armorial treatise. Its catalogue wording is particularly useful because it separates the observed painted features from the cautious attribution to the author of a particular text. ‘Could be’ is not the same as ‘is’.
An unpublished claim is not made reliable by typing the name of a castle beneath a photograph. We need the object’s repository, shelfmark or inventory number, dating evidence, and preferably the original cataloguing record.
A photograph has a date too
A photograph of a medieval glass panel is not itself a medieval source. It is a recent witness to a potentially older object. Restoration, relocation, damaged portions, overpainting and modern display affect what we see. A catalogue might correctly identify a work’s medieval origin while the portions containing heraldic decoration were materially restored in the nineteenth century.
Thus a careful object entry records both date of original object and date and status of its present documentation. A photograph licensed for reuse is yet another separate question. The site’s object studies already preserve individually documented image rights; that policy should apply to every future heraldic photograph.
The useful comparative unit is a claim
Rather than build a database record that simply says ‘has supporters: yes’, consider a claim-based model. A contemporary English grant might prove that a specific armiger received two named supporters on a given date. A German seal might show two creatures next to the arms but contain no evidence of a separately inherited supporter entitlement. Recording both as the same Boolean value creates the illusion of comparability where it is precisely comparison that requires work.
A richer record would contain the element, visual observation, source witness, claim type, jurisdiction, period, degree of verification, and possible alternatives. The conclusion ‘depicted as supporters’ could coexist with the qualification ‘no grant or registration located’.
Negative evidence and historical change
If supporters disappear from a later engraving, three explanations are immediately possible: a historical change in rights or conventions, a change in artistic presentation, or a practical reproduction choice. More are possible. Conversely, the appearance of a new collar might reflect a real honour received during the bearer’s lifetime. Only dated evidence lets us discriminate among such alternatives.
This is where heraldry becomes an exercise in documentary historical method. The shield is useful precisely because it gives us a relatively stable point of comparison. Around it we can observe change that may signal office, allegiance, privilege, taste or a different reproduction medium.
Capstone: publish a critical miniature catalogue
Select three independent witnesses to one armorial identity, preferably a seal, a manuscript or grant, and a later physical reproduction. Produce an 800–1,200-word catalogue entry. Transcribe the earliest surviving wording; inventory each external ornament; separate firsthand observation from editorial reconstruction; and identify any legal or genealogical proposition that remains unproven. Supply an image-rights statement, a source locator, and one plausible competing interpretation.
Then review your catalogue as an adversarial reader. What is its strongest claim? What is its weakest? Which assertion would you remove if the original source were no longer accessible? A successful study is not the one with the most ornaments. It is the one whose claims survive examination.
These English-language syntheses cite institutional and historical sources but have not been independently reviewed by all relevant regional specialists. Current guidance, historical practice and enforceable law must be distinguished.