In June 1204, King Philip Augustus of France completed his subjugation of Normandy and overthrew the so-called ‘Anglo-Norman realm’, following the flight overseas of the duke of Normandy, King John of England, the previous year. All landowners in the duchy were ordered to do homage to the king of France by Easter (10 April) 1205 or else suffer confiscation of their Norman lands. Several months after the deadline, on Sunday 13 November 1205, a great gathering of surviving members of the Norman aristocracy at Rouen issued a proclamation concerning the traditional rights of the ruler of Normandy over the clergy. The document recording this declaration survives in the Archives Nationales in Paris and forms the subject of this article, especially its fine and unusual collection of seals and its curious method of sealing.2
This study first considers the history of seals in Normandy before 1205, before briefly outlining the context for the declaration on the Norman clergy. It then describes the seals themselves in detail and considers their place within the history of Norman seals. It thereby sets out the history of Norman aristocratic seals in the Anglo-Norman and early Capetian periods and shows their diplomatic and political significance at the time of the transition from Angevin to Capetian rule.
Early Norman seals have received substantial historical attention since Germain Demay’s Inventaire des Sceaux de la Normandie, and specific collections have been examined and sometimes catalogued, such as the fine collection of nearly 200 seals from the muniments of the Cistercian abbey of La Noë near Évreux, now in the Bibliothèque Nationale de France, most of which belonged to very minor local landowners.3 Nevertheless, some other collections remain neglected, notably the magnificent series for the much more famous abbey of Savigny, which survive mainly in the Archives Nationales and the Bibliothèque Muncipale in Rouen.4 The subject still lacks a comprehensive overview: Norman seals have not been subjected to the sort of quantitative analysis that Brigitte Bedos-Rezak and Nicolas Civel have executed for Picardy and the Île-de-France, for instance.5 Moreover, an important recent consideration of Norman seals as a group by Maïté Billoré remains unpublished: it contains a short thematic and statistical analysis as well an inventory and reproductions of about eight hundred twelfth- and thirteenth-century Norman seals from the national collections in Paris and the departmental archives of Seine-Maritime, Calvados, and Eure.6 Norman seals have more often been considered either in a broader French context,7 or within an Anglo-Norman framework8 – with good reason, since many Norman sigillants were also landowners in England. Indeed, much of what has been written about the design and function of British seals holds good for Norman seals as well.
There is some slight circumstantial evidence for seals in Normandy before 1066, when seals were becoming popular in some other parts of Northern France;9 but it was in the wake of the Norman Conquest of England that the seal appears to have been brought across the sea to the duchy as a means of authentication or confirmation, no doubt influenced by the Conqueror’s rapid adoption of the Anglo-Saxon sealed writ.10 Sealing soon became a popular form of validating acts: Emily Zack Tabuteau lists sealing clauses in acts before 1107 not only of the kings of England and dukes of Normandy but also of counts, viscounts, archbishops, bishops, and men of humbler status.11 Although some of these references may not be authentic and a few should certainly be redated to a later period,12 it is clear that the seal was gaining in popularity in Normandy by the early twelfth century, much as it was more generally across Northern France. Nevertheless, given the diversity of diplomatic forms such as pancartes, chartes-notices, and diplomas, many acts remained unsealed,13 and it was only after 1150 that the seal became an essential feature of Norman charters. In 1151, an act of Gilbert de Say in favour of the abbey of Aunay contained both a seal and signa, as did a slightly later act of Geoffrey III de Mandeville, earl of Essex.14 Yet in 1153, William II de Roumare, earl of Lincoln, issued a writ-charter addressed to the archbishop of Rouen that had ex parte witnesses and a seal but no signa.15 Thereafter the sealed charter rapidly displaced older diplomatic forms, and signa vanished as modes of assurance before the end of the reign of Henry II (duke 1149/50–1189).16 The lateness of this development bears comparison with other parts of France, where similarly the late twelfth century, not earlier, was the main period when seals began to be used.17
From then on the sealing clause became a standard feature of Norman diplomatic, although not every act bearing a seal mentioned that fact in the text. As in England and elsewhere in Northern France, most Norman acts in the twelfth and early thirteenth centuries were sealed sur double queue, usually with a plica or lower edge folded over for strength, and with the seal attached either by parchment (or occasionally leather) tags or by silk or woollen cords. Aristocratic acts were occasionally sealed sur simple queue, without a fold and with the seal impressed upon a tongue of parchment protruding from the main body of the act. Multiple sealing, rare at first,18 became more common in the second half of the twelfth century, normally with a hierarchy of importance working from left to right.19
The majority of sealed charters must have been produced by their ecclesiastical beneficiaries, but there are signs of aristocratic ‘chanceries’ in the twelfth century,20 and a good number of contracts between lay parties from the late twelfth century onwards. As might be expected, most seals bore the name of the person issuing the act, but acts were frequently sealed by confirmatory authorities such as bishops or lords, whether or not their act of sealing was mentioned in the text. A statement in an act of Robert and Ralph de Gauville for the abbey of Aumale (c. 1165) is typical: ‘since we did not have our own seal, we confirmed this grant with the seal of Robert the constable [of Aumale] and placed it sealed upon the altar’.21 Just as minor landowners increasingly issued their own acts in the late twelfth century, so, too, they increasingly sealed their own acts. In 1196 Robert de Rétonval confirmed the gifts of his father and grandfather to the abbey of Foucarmont, and sealed the act himself; significantly, however, his act claimed that his grandfather’s grant had been recorded only in a sealed charter of his lord, Count John of Eu (d. 1170).22 Even so, seals were not necessarily used by the person whose name they bore: in an act of 1175, the three sons and wife of William Pantulf of Samesle stated that they had confirmed William’s gifts by affixing his seal to their act; the seal is lost but the act bears a single sealing slit.23 In addition, third-person notifications continued to be drafted and sealed on occasion until the end of the twelfth century.24
One popular use of seals from the mid-twelfth century onwards was for bipartite chirographs. Fine early originals survive for the abbeys of Aumale and Almenêches from c. 1150 and 1157 respectively,25 and they grew in popularity thereafter.26 Similar phrasing to English final concords was adopted for chirographs drawn up at the Norman exchequer by King John’s reign, and while some final concords appear to have been unsealed,27 there does not appear to have been a concerted move towards using unsealed tripartite concords as in England. On the contrary, in 1212, a final concord that had been drawn up at assizes in Normandy in 1199 for the division of a cross-Channel inheritance was rejected in the king’s court at Westminster because it bore a seal, explicitly because this was contrary to English practice for final concords.28 More often, the parties either sealed both halves of the chirograph, or sealed the half in their own name, or sealed each other’s half.29
Hence, after a hesitant start over the previous century, seals acquired a standardised purpose and function in Normandy under the Angevin dukes, and by the time of the fall of the Angevin régime in 1204, almost all Norman acts were sealed: if not as the sole means of authentication, then as the most popular means of strengthening an act’s validity, both for ecclesiastical and for lay transactions. While exact sealing practices are rarely apparent, and the early Norman custumals give no hints – in contrast, for instance, to the remarkably informative statements in Philippe de Beaumanoir’s Coutumes de Beauvaisis (1283)30 – charters occasionally give some indication of sealing conventions in Angevin and early Capetian Normandy. Around 1200 a clerk recorded that his father had resigned the patronage of the church of Feugerolles-sur-Orne to the priory of Le Plessis-Grimoult, but had died before he could seal the act of resignation: this shows that there was a clear delay between the father’s oral ceremony and the son’s sealing of the act.31 For male landowners, acquiring a seal was as much a feature of coming of age as the knighting ceremony was for knights, as can be seen in the record of a dispute waged in royal assizes in 1225 over the fief of Montgaroult near Argentan. In support of his claim to two thirds of the fief, the abbot of Saint-André-en-Gouffern produced several charters for the abbey from the predecessors of the rival claimant, Gervase de Joué, as well as a confirmatory charter which had been issued by Gervase himself with his father and elder brother and which bore their seals. Gervase denied that he had confirmed the grant or sealed the charter: he retorted that if his father and brother had indeed issued this deed, it was at a time when he had not yet acquired a seal or come of age, and so the charter would have been made contrary to the uses and customs of Normandy.32 However, an inquest found that Gervase had already reached the age of twenty-one when the charter was made and sealed with three seals, and that he was one of the sigillants.33 The inquest’s verdict was correct, for the charter of Gervase, his father and brother survives today in the Archives du Calvados; regrettably, it has lost its seals.34 After the verdict, a new record of the case was sealed by arbiters who negotiated a compromise between the abbot and Gervase de Joué.
Other acts indicate how grantors dealt with the loss of a seal. An aristocratic widow, Joanna du Breuil, notified Bishop Jordan of Lisieux (1201–18) that she had made gifts to the abbey of Saint-André-en-Gouffern after her husband’s death and confirmed them with her charter, but she had lost her seal, as the bishop apparently already knew; she therefore implored him to protect her gifts against her sons.35 On the borders of Normandy and Maine, a knight called Hamelin de l’Écluse granted customary rights to the monks of Savigny, and sealed the act with a new seal since he had lost his old one ‘during the war’, presumably one of the conflicts that had pitted his lord, Juhel de Mayenne, against King John in 1199 and 1202–03.36 When sigillants altered their seals for some reason, it could undermine the authenticity of their earlier grants. The simplest solution, adopted by William de Semilly in 1236, was to issue a fresh act with the new seal.37 A more complex solution was adopted by William de Husson in 1237. In 1223, he and his mother had granted grain renders in the diocese of Bayeux to Savigny, but soon after he had broken up his seal and replaced it with a new one. Now, in 1237, he wished to enlarge the grant to the abbey, and so he issued a new act, to which he transferred the old seal: he then added his new seal and publicly avowed that both seals were his, and for good measure he arranged for the rural dean of the Val de Mortain in the diocese of Avanches (where his chief estates lay) to seal the act as well.38
In contrast to their standardised function, twelfth- and thirteenth-century Norman seals varied enormously in their appearance, but these, too, were generally similar to those of seals in England or other parts of France. Their size, shape and designs were heavily influenced by the social status and function of the seal’s owner, but no exact correlation between these features can be established. Unlike in English royal acts, or indeed French royal acts by the end of the reign of Philip Augustus, the colour of the wax used for Norman aristocratic seals does not appear to have been significant.
For the first few decades after the spread of lay seals, the most popular design for landowners were equestrian seals: these varied enormously in size and quality, and many were issued by very minor ‘gentry’ figures as well as magnates. Most equestrian seals were military in design, but a well-known exception for a family whose interests straddled Normandy, France and (eventually) England, were the series of seals issued by the lords of Montfort-l’Amaury, counts of Évreux and afterwards earls of Leicester: although these depicted a knight on the obverse, they bore a huntsman and hound on the reverse, possibly an allusion to the dynasty’s ancestral office as the grierius of the great royal Forest of Yvelines.39
Also popular were symbolic designs, many of which may be described as ‘proto-heraldic’ or ‘quasi-heraldic’. A particularly fine, large and early example is the seal of Geoffrey de Clinton for Savigny in 1158, which depicts a griffin attacking a lion: this is a reminder that the remit of Norman seals transcended the borders of the duchy, since Geoffrey’s father had been Henry I of England’s chamberlain in England.40 Around the same date, we see ‘canting’ symbols being used, such as the wolf passant used by the Franco-Norman magnate William Louvel of Ivry (c. 1162),41 the eagle displayed on the seal of Richer II, lord of l’Aigle (d. 1176),42 or, at a rather humbler level, the ash tree of William du Fresne, a minor landowner near Conches, in 1174.43 From the early thirteenth century, non-heraldic symbols were mainly used on smaller seals, particularly of minor landowners and commoners, who presumably did not always have recognised coats of arms.
As David Crouch and others have discussed, equestrian seals depicting true heraldic devices on shields and horse-trappings go back to well-known examples from the 1140s, especially the chevrons of the Clare kin-group44 and the famous ‘Warenne group of chequered shields’ of Count Waleran II of Meulan and his maternal uncle Count Ralph I of Vermandois, shared by a kin-group whose lands were spread across England, Normandy, and Francia.45 Norman seals with a shield as the main device survive from the end of the twelfth century. In 1196, the cession of the Norman frontier fortress of Vernon to Philip Augustus bore two seals: Richard I de Vernon’s equestrian seal, which depicted his arms – a saltire – on the shield and horse-trappings, and the seal of his son and heir Richard II de Vernon, which simply bore a shield with the same arms differenced by a label of thirteen points.46 The seal of William, brother of the count of Sées or Alençon (c. 1200) bore almost identical arms to his distant cousin and namesake William, count of Ponthieu.47
The seals of the higher clergy in Angevin and early Capetian Normandy were usually of similar form and design to elsewhere in the Anglo-French world, namely vesica-shaped and with conventional prelatial figures. Noblewomen’s seals, too, were usually vesica-shaped and depicted them in very standard fashion.48 One of the earliest is the seal of Matilda, countess of Évreux, which survives in the Archives de Yonne in Burgundy on an act for the abbey of Pontigny;49 but by the beginning of the thirteenth century, much humbler women were issuing seals with this standard noblewoman design: two designs survive for the minor Évrecin landowner Basilia de Glisolles, for instance.50 There were some more distinctive female seals, however, such as the seal of the Anglo-Norman heiress Agnes de Beaumont, widow of Richard du Hommet, constable of Normandy, which appears on an act concerning her inheritance in Oxfordshire.51
Burgess seals had become common in Normandy by the end of the twelfth century. Very extensive collections from Rouen include the seal of Matthew le Gros, mayor of Rouen in 1199–1200, which depicted an intaglio with a bust identified as the goddess Minerva, while his sons used a joint seal showing the Roman Empress Julia and her son Emperor Caracalla; the seal of Geoffrey, son of the viscountess of Rouen, a burgess but shown hunting on horseback;52 and the seal of the commune of Rouen itself, closely modelled on the lion seal occasionally used by the Plantagenets as dukes of Normandy.53 If the merchants of Rouen had special reasons to aspire to such grandeur, the burgess of smaller towns also adopted seals by 1200, such as Ralph l’Abbé, a prominent burgess of Sées and Argentan and a leading administrator in the duchy from Richard I to Philip Augustus, and his son Herbert, a clerk who was a candidate for the contested episcopal see of Sées in 1201–3.54
A final feature of Norman seal design to note was the introduction of the counterseal, beginning under Henry II and becoming common in the early thirteenth century.55 As elsewhere in the Anglo-French world, these could include Classical intaglios, symbols and heraldry, or miniature equestrian figures.
It will be apparent that Norman seals had few if any unique features in their design and use, and if English influences after the Norman Conquest initially created contrasts with seals in other parts of northern France, from the reign of Henry II onwards they developed very much in tandem with both England and northern France. Landowners with property on both sides of the Channel, or elsewhere in France, used the same seals in Normandy and other regions. While the seals of the higher aristocracy were distinctive, the similarity of seals of three different gentry benefactors of the abbey of La Noë suggests that their matrices were manufactured by a single workshop in the nearby city of Évreux.56 Indeed, the very lack of variation of seals in this period across numerous kingdoms and provinces merits further discussion: even though customary law and judicial practice were hardening along provincial lines, the use of seals appears remarkably similar across political and legal boundaries. Meanwhile, the chief value of the Norman seals may simply lie in their number: thousands survive from the twelfth and first half of the thirteenth century.57
By 1204, then, the seal was the main established method of validating charters and letters in Normandy: its variety of designs went hand in hand with its common purpose and function. However, the overthrow of the ducal régime in 1204 ended the direct political connection with England, at least for the vast majority of landowning families. It had an immediate impact upon the political complexion of the duchy, and arguably also a long-term effect upon the diplomatic practices of the Norman aristocracy, as Northern French charters abandoned the witness-list and dating clauses became ubiquitous.
The French victory forced many of the most prominent barons to retreat to England, including eleven earls and counts and leading barons such as the lords of Briouze, Gournay, Montbray (Mowbray) and Tosny. At a stroke, the duchy lost much of the élite that had dominated the region for the previous couple of centuries. However, this disaster for much of the Anglo-Norman ruling class served some people well: it brought to the fore those Norman magnates who remained in the duchy; it allowed the rapid rise of Norman families who served the new régime; and it introduced a number of courtiers, soldiers, administrators and clerks from the Île-de-France.58 Under the kings of England, Normandy had become accustomed to the ruler’s frequent absences, and the early Capetian kings also rarely ventured further north and west than the duchy’s south-east corner, preferring to control Normandy through bailiffs and castellans – mostly from the Île-de-France – and through periodic taking of pledges.59 Despite the surveillance of the Capetian régime, gatherings such as the twice-yearly exchequer meetings suggest that the Norman aristocracy enjoyed considerable autonomy and could act together as a single community upon occasion. Apart from the surrender of Rouen itself, in which seven barons and knights and thirty-three burgesses treated with the king of France and his leading courtiers,60 the declaration concerning the Norman clergy in 1205 appears to have been the first significant demonstration of common action by the survivors of the Capetian conquest of Normandy.
The exact reason for the declaration is not recorded, although references in the text suggest that it may have been provoked by a dispute over the judicial rights of the bishop of Lisieux: it is also clear that the jurors themselves proved jealous of lay rights.61 Philip Augustus’ relationship with the Norman Church after 1204 was testy, and his regulation of the Norman Church formed part of a broader move to define and tighten royal rights over the Church across the Capetian regnum.62 The declaration of 1205 concerned a wide range of issues as they had supposedly obtained in the time of Henry II and Richard I – King John’s reign being treated as a legal nullity.63 The barons swore that the rights of the clergy had been very restricted in matters such as advowsons, excommunication of royal bailiffs and clerks, the jurisdiction of ecclesiastical courts, both across the duchy and in specified regions, debts and usurers’ goods, and criminous clerks. It was not, however, intended to be a definitive statement, for the act expressly stated that important Norman barons were absent: these would have included the counts of Eu and Alençon and the lords of Harcourt, Le Neubourg, Creully and l’Aigle, amongst others, and, indeed, fourteen of the twenty-two named barons were from eastern Normandy, chiefly the region of Rouen itself. Admitting the fallibility of human memory and the problem of absences, the barons stated that a further meeting would be required.64 Philip Augustus would regulate the property and legal rights of the Norman clergy on several more occasions during his reign.65