Seals and their Context in the Middle Ages

In searching seal catalogues with this in mind, a broad pattern emerges. With few exceptions sigillum does not appear on the principal seals of sovereigns before the sixteenth century; it normally appears, however, on the seals of queens and other members of royal families, just as on the seals of nobles and lesser named seal-owners. Nearly all deputed royal seals call themselves sigillum, not only in defining their area of authority, as the seal of the exchequer or of a court of justice, but also where we might expect more exact duplication of the principal seal’s form, as on the French kings’ seals of absence or the seal of Edward I as ruler of Scotland. The principal seals of bishops – their seals of dignity – normally start the legend with sigillum, but at least in France, England and Scotland there was a tendency from the mid-twelfth to the mid-thirteenth century to omit this and to give just the name and title.

Most interesting, however, is that sigillum does not appear on seals before the late eleventh century; apart from the seal of Fulk, to be discussed, the seals of the archbishop of Bourges in 1089 and of the bishop of Beauvais between 1089 and 1095 seem to be the earliest examples from France.⁵ The only exceptions to this are from England. Certainly, on the only known seal of an English king before Edward the Confessor, the lead bulla of Cenwulf, king of Mercia 796–821, the word sigillum does not appear,⁶ nor on the earliest of all known English seals, the late-seventh-century gold matrix of Baldehildis,⁷ but we see it on all the few other English seals believed to date from before the Norman Conquest.

Ethilwald, bishop of southern East Anglia, 845–70:

+SIG’EĐILVVALDI:EP⁸

Aelfric, late 10th or early 11th cent.:

+SIGILLVMÆLFRICI⁹

Godwin, minister, late 10th or early 11th cent.:

+SIGILLVMEGODǷINIMINISTR’¹⁰

Godgytha, nun, late 10th or early 11th cent. (engraved on handle of Godwin’s seal):

+SIGILLVMGODGYĐEMONA CHE D’O DATE

Wulfric, first half of 11th cent. (Schøyen): +SIGILLVMǷVLFRICI¹¹

Others, that long continued in use, we know only from later impressions. The best known is the seal that Wilton Abbey used down to its dissolution in 1539; this was the personal seal of its early saint, Edith, who died in 984. Its legend reads:

+SIGILL’EADGYĐEREGAL’ADELPHE

She was the daughter of King Edgar and the sister (adelpha), correctly half-sister, of two kings, Edward the Martyr (975–78) and Aethelraed the Unready (978–1016). In an important article in 1980, T. A. Heslop argued convincingly that six other English religious bodies were using seals in the twelfth century and later that had been engraved at latest in the eleventh century and most likely before the Norman Conquest. Of one, from Glastonbury Abbey, we have only a fragment of the legend, but on all the other five we can see that the legend began with a cross and the word sigillum. These seals were from Exeter, Sherborne, Athelney and the cathedral priories at Canterbury and Durham.¹²

Until the late eleventh century, then, it seems that it was a peculiarity of English seals to start the legend with sigillum. What did the word mean in Anglo-Saxon England? The word is a diminutive of signum, which can mean a sign or token. This is exactly how signum is used on our three earliest seals of queens of France, the gold signet rings of Arnegundis (died 573), Radegundis (died 587) and Berthilda (who married Dagobert I in about 635); each is engraved with the queen’s name and a monogram reading ‘REGINE SIGNUM’ (with variants).¹³ These seals may well have been used to authenticate documents; early royal charters from France were authenticated by the monarch’s seal attached en placard. However, we know of no English sealed documents earlier than charters of Edward the Confessor and earlier seals were probably used to give authority to someone sent on the sigillant’s business; they authenticated the messenger rather than a written message.¹⁴ What authenticated a formal document was not a seal but the persons testifying to its import, in a list normally at the end of the text. Before each name was a cross that the charter called the signum crucis or, simply, signum.

Even as late as 1095–97 we have a royal writ in this form, though further strengthened with the pendent seal that was by now the normal means of authentication.¹⁵ On some late examples the crosses appear to be autographs, actually drawn on the document by the persons named. Earlier, the crosses were normally drawn in uniform style by the writer of the document. In either case, however, we may be reasonably sure that authentication did not consist solely in writing the name and drawing the cross, the signum; each signatory will have strengthened the confirmation by making the sign of the cross, which is what the signum recorded. In most original charters that survive from the years 939–57 the sentence that names each witness not only opens with a large cross, but has a small cross interlineated above the reference to the sign of the cross, very often called not signum crucis but sigillum crucis;¹⁶ we can only guess what this may have implied in the ceremonial of attesting the charter.

It is against this background that we should see the word sigillum on early English seals. The sigillum is not the seal itself, but the small cross at the start of the legend;¹⁷ it was this that was the authentication, the sigillum crucis of the seal’s owner, and we may well suppose that the sign of the cross was made over it every time it was brought into service.¹⁸ It is in this sense that we should understand the word sigillum on all seals before the Norman Conquest of 1066, the seal of Edward the Confessor among them. Oddly, this is borne out by certainly one, and very likely the other, of the two earliest Anglo-Saxon seals that, we have seen, are the only ones that do not have sigillum in their legend. That on the lead bulla of Cenwulf, king of Mercia, reads:

obverse + COENVVLFI REGIS

reverse + MERCIORUM

The king’s name and title are not in the nominative case but in the genitive: it is the cross ‘of Cenwulf king of the Mercians’.¹⁹ The word sigillum does not appear – perhaps would never have appeared at this date – but is to be understood. The cross of King Cenwulf is what the seal is all about. Our other seal without sigillum may be an exact parallel. Its legend is simply:

Although here there is room for some doubt, this too may be read as a genitive: the cross ‘of Baldehild’.²⁰

One question that arises is the parallel with contemporary English coins. It has been generally accepted that in the early Middle Ages there was a close link between the designs of seals and of coins, in England and elsewhere, though this has never been seriously investigated. Certainly most of our Anglo-Saxon seals look very like contemporary English pennies; there too a cross appears at the start of the wording around the edge, at first occasionally on coins of Offa of Mercia (757–96), then increasingly until by the ninth century it was a normal part of the design. We might reasonably suppose that the cross on the seal was simply taken over from the design of the coin. However, the words of Stuart Rigold are of interest here; he was a scholar whose interests covered both coins and seals and speaking of these seals in a published lecture some 40 years ago he wrote:

‘All have been, justly, compared with coin-types and, quite unjustly, regarded as isolated designs derived from existing coins. All positive evidence, throughout history, points in the opposite way, to wit, that the seal-type has the priority and the coin-type is derivative.’²¹

Thus, he would see the cross on the coins as coming from contemporary seals, not the other way round. Chronologically this is acceptable; the seals of Cenwulf and Ethilwald are a generation or so later than the pennies of Offa, but the seal of Baldehild is considerably earlier and we should anyway remember that the number of coins struck will have been infinitely greater than the number of seals engraved and that this will be reflected in their survival. Logically it makes sense for the cross on the seal to precede the cross on the coin: if the coin is based on the seal the cross will carry across from the seal its guarantee of authenticity, whereas otherwise there would be no particular reason to put a cross at the top of the coin.

It is not suggested that sigillum was never understood as seal in Anglo-Saxon England; it clearly could have this meaning. The seven seals that closed the book in the Vulgate’s translation of the Revelation are sigilla,²² and in his glossary Aelfric gives insegel, which unquestionably meant seal, as the English translation of sigillum or bulla.²³ The word was ambiguous and Jane Roberts has recently shown how we see this ambiguity at work in the Old English translation of Felix’s account of Guthlac of Crowland.²⁴ This is an early-tenth-century text that we know only from a late-eleventh-century version that has undergone some revision, and we may wonder whether in late-Anglo-Saxon England seal was being seen more and more as the primary meaning of sigillum – a process that its appearance on seals will have furthered. Certainly the few occurrences of sigillum in Domesday Book are unequivocal; to the Norman clerks who were its authors the word meant either a sealed document or, probably, the seal that a messenger carried to authenticate his message.²⁵ However, it is in the context of English diplomatic of the ninth and tenth centuries that sigillum first appeared on seals, and in this context it is unambiguously the cross at the start of the legend, not the seal itself.

If we accept that the word sigillum on seals originated in England and in Anglo-Saxon diplomatic we still have questions to answer. The word’s appearance in continental seal legends follows a pattern that can be easily defined, but the appearance of a cross at the start of the legend does not; we find it in continental seal legends long before we find the word sigillum, but inconsistently, and its incidence would probably repay investigation.²⁶ Another question is how and why did sigillum cross the Channel to become common on seal legends throughout western Europe? Alongside this we may place another, older, question: how did the two-faced pendent seal, which originated in the court of Edward the Confessor, come to be adopted in other countries? It seems as if, unexpectedly and for reasons that are obscure, mid-eleventh-century English usage had a strong influence on Continental seals. In our present state of knowledge neither question can be answered, but in the case of the word sigillum we have what may be just a glimmering of the process of transmission. In January 1843 a matrix was found ‘dans le canal à Amiens’; carved in ivory it was on one side the seal of Fulco, archdeacon, on the other the seal of Fulco, bishop, presumably reflecting the career of its single owner. This matrix has long disappeared, but it is known through plaster impressions. Most remarkable is that in form and lettering it is at one with the seals assigned to Anglo-Saxon England – and the legend on each side begins with a cross and the word sigillum. Fulco has been assumed to be bishop of Amiens and there were indeed two of this name at an appropriate period: one held office from 993 to 1030, the other from 1036 to 1058.²⁷ This, however, rests only on the slender evidence of the find-spot. Beauvais is not far away, and it could as well have belonged to Fulco de Dammartin who, we have seen, was bishop there from 1089 to 1095 and whose known seal is among the earliest outside England to bear the word sigillum. Venturing still further into speculation we may suppose that he first, as archdeacon, had a seal in wholly English style and had a similar seal carved on the back when he became bishop, soon replacing this, however, with a more conventional bishop’s seal – it shows him standing and giving a benediction – into which he introduced the cross and the word sigillum from his earlier seal.²⁸ This will have done no more than record the sign of the cross that would be made when the seal was used, probably in France as in England. We may speculate even a little further. Could the omission of sigillum from so many bishops’ seals between the mid-twelfth century and the mid-thirteenth reflect changing usage? – it was left off because it was no longer the custom to make the sign of the cross over every transaction and reinstated a century later when the word signified nothing more than the seal itself.

This is guesswork that further research may confirm or reject. It seems reasonably certain, though that the word sigillum in a seal’s legend first appeared in England in or before the ninth century, referring not to the seal itself but to the cross at the start of the legend. It appeared on seals on the Continent from the late eleventh century and before long became a normal part of the legend, meaning now not the cross – though that was still there – but the seal itself.

1

This is a seal

P. D. A. Harvey