Part II
Chapter 4
By 1060 the political position of Duke William had been stabilized as a result of fourteen years of continuous war. He had moved out of his perilous minority, and freed himself from dependency upon the king of France. He had withstood a combined assault from Paris and from Anjou; and the deaths of Count Geoffrey and King Henry had removed from his path his two most formidable opponents in Gaul. Never before, during his reign, had Normandy been so secure from attack, and its ruler was now offered the opportunity yet further to strengthen his position during the six years which were to elapse before he undertook the invasion of England. He was in his early thirties, and it is small wonder that what he had accomplished had won for him an ever-increasing prestige. This appears in all the comments of contemporaries. Of itself, therefore, it merits some attention since it was to make a real contribution to what he was later to achieve.
A warrior age salutes a warrior, and in the young William it found a warrior to salute. Tall in stature and of great physical strength, his personal exploits in battle (particularly in the campaigns of 1051 and 1052) had attracted notice, and more sober contemporaries might already have detected in him a commander of considerable capacity. Yet these attributes (which were shared by many of his contemporaries) do not suffice by themselves to account for the special admiration with which in 1060 he was beginning to be surrounded. An explanation has therefore to be sought in his more individual qualities. Brutal himself, it was not merely through successful brutality that he had been able to elicit support from so many of the ruthless men who might have been expected to withstand him. Doubtless, fortune had sometimes favoured him, and certainly Norman chroniclers, writing after 1066, were liable to offer him fulsome adulation. But when all deductions have been made, his courage in adversity had in fact been outstanding, and it commands admiration. There must, moreover, have been a wonderful tenacity of purpose in this young man, who threatened by murder in infancy and menaced by treachery in adolescence, had none the less saved himself by long years of war conducted against great odds. His success in his struggle between 1046 and 1060 must in the last resort be adjudged as in some measure a triumph of character.
The value of the prestige he had acquired was to be abundantly displayed in the future. But by itself it will not explain the power of the Norman duchy which confronted England in the third quarter of the eleventh century. If in 1066 the Norman duke in the Norman duchy could aspire to become, by force of arms, one of the dominant rulers of Europe, this was due in the first instance to the political structure of pre-Conquest Normandy. In particular it depended upon the results of two movements which had already by 1066 been brought to fruition in the province, namely the rise of a very remarkable aristocracy, and a most notable revival in the Norman church. The one supplied the sinews of Norman strength, the other gave a special direction to Norman policy. Both these movements had started before Duke William II began his effective rule, but both gathered new impetus under his direction; and he was to be signally successful in co-ordinating them to his own advantage. To these developments – aristocratic, ecclesiastical, and ducal – attention must now be turned, for they were to mould the future, and to provide not only the force but also much of the character of the Norman impact upon England. The greatest period of Norman achievement did not begin until they had taken place, or before they had been fused together by a great constructive statesman to provide the overmastering energy of a province which in 1066 was unique in Christendom.
Of these developments the rise of the Norman feudal aristocracy must, both for its causes and its consequences, challenge immediate attention. For it is by no means easy to ascertain with any precision why it took place. General theories of lordship and vassalage will clearly not of themselves suffice to account for the rise of an exceptional aristocracy in an exceptional province, and if actuality is to be given to the inquiry, reference must be made to the particular families which then arose. An endeavour must be made to determine how they acquired such power, and how their doing so affected their relations with the duke. Such an investigation must inevitably on occasion be concerned with points of detail, but it can be justified by the magnitude of the issues which were involved. Even so (as must be confessed) the evidence is somewhat intractable. The statements of later chroniclers respecting the ancestors of their patrons are notoriously suspect, and in the case of Normandy a special problem is posed in connexion with the pedigrees which, late in the twelfth century, Robert of Torigny added to the chronicle of William of Jumièges.1 These famous genealogies have been widely used by modern scholars, but it is usually hazardous to rely on them unless they can be corroborated by independent testimony. In short, any account of the rise of that Norman secular aristocracy which was under the duke to effect the Norman conquest, and to give a new nobility to England, must not only be related to individual families: it must also be firmly based upon the testimony of the Norman charters.
Four such families may here therefore be selected as illustrating this development which was to be of such importance to Europe. The first of these is the house of Tosny.2 Its earliest undoubted member is a certain Ralph de Tosny, and this man (who may be styled Ralph II) or perhaps another Ralph who was possibly his father (and may thus be styled Ralph I) was the original grantee of Tosny,3 which had belonged to the see of Rouen in the time of Hugh, archbishop of Rouen, from 942 to 990. Ralph II, about whom there is definite information, was in 1013 or 1014 entrusted by Duke Richard II with the defence of Tillières,4 but just before this, or shortly after, he is to be found in Italy, and when and where he died is not known. He was succeeded as lord of Tosny by his son Roger I who, likewise, had a varied career, both within and outside Normandy. At some time he went adventuring to Spain,5 and he is known to have married a woman named Godehildis, who later (after Roger's death) became the wife of Richard, count of Évreux. Roger himself perished in one of the feuds of the minority, for in, or shortly after, 1040 he was slain in private war by Roger of Beaumont. This disaster, however, did not prevent Roger's son by Godehildis – Ralph III – from succeeding to Tosny, and he too played a prominent part in Norman politics. Ralph III was active against the king of France in the campaign of 1054, and later he fought at Hastings.6 Nevertheless, it was always in Normandy that his chief interests lay, and before his death, which occurred before 24 March1102, he had become, within the duchy, a benefactor of many religious houses, including Saint-Évroul, La Croix-Saint-Leuffroi, Le Bec-Hellouin, and Jumièges.7
Tosny is the earliest family to be discovered in Normandy wherein a territorial appellation is found to be descendable in the manner of a surname, and the succession which has here been displayed is thus of exceptional interest. Similarly, the marriage alliances formed by members of this family are symptomatic of its growing power. Not only did the widow of Roger I marry a count of Évreux, but his sister linked Normandy to Maine, by allying herself with Guy de Laval, and one of his daughters married William fitz Osbern, steward of Normandy, and later earl of Hereford.8 Nor was the Old English aristocracy itself to be unaffected by the Tosny fortunes, for Ralph IV of Tosny (the son of the man who fought at Hastings) married a daughter of Waltheof, son of Earl Siward of Northumbria.9 A better illustration of the expanding influence of a rising Norman family in the eleventh century could hardly be obtained.
The rise of Tosny was, however, not to be unchallenged even in central Normandy, for some twenty miles to the west of it lay Beaumont on the Risle which was to give its name to an equally illustrious and bitterly hostile house.10 The first member of this family who possessed Beaumont was Humphrey ‘of Vieilles’ who was probably the son of one ‘Thorold of Pont-Audemer’, and more doubtfully grandson of a certain ‘Torf’.11 This Humphrey was a supporter of Duke Robert I in whose company he is frequently to be found, and before 1035 he founded at Préaux near Pont-Audemer two monasteries: Saint-Pierre for men, and Saint-Léger for women.12 He died before 1047 and was succeeded by his son Roger, soon to be styled ‘of Beaumont’, who first brought the family to greatness. For a long time, however, its fortunes were precarious, since during the minority of Duke William a fierce struggle took place between Beaumont and Tosny in which not only Roger II of Tosny but also Robert the brother of Roger of Beaumont perished.13 Nevertheless, Roger of Beaumont himself prospered. Leaving Vieilles, he established himself on the neighbouring hill of Beaumont, where he built a famous castle, and there he remained in power throughout the Conqueror's lifetime.14 He did not take part in the campaign at Hastings, but was represented at that battle by his eldest son, Robert.15 His own interests, in fact, remained in Normandy, though in 1086 he is recorded as possessing some estates in Dorset and Gloucestershire.16 But his two sons, Robert and Henry, became great landowners in England and in due course respectively earls of Leicestershire and Warwick. Thus it was that a man who had been active in Norman politics in the time of Duke Robert I survived until after the time of Domesday Book, and having established the fortunes of a great Norman family, left behind him two sons who were to acquire English earldoms. The rise of Beaumont can in truth be described as both rapid and spectacular.
Few Norman families of the eleventh century were more powerful than those of Tosny and Beaumont, but the same period also witnessed the rise of many lesser houses, and of these the first family of Vernon may be taken as an example. When, some time between 1032 and 1035, Duke Robert I gave land to Saint-Wandrille at Sierville, some ten miles north of Rouen, he did so with the consent of a certain ‘Hugh of Vernon’,17 and other documents show that the family of Hugh had already become possessed of other estates in this district,18 for in 1053 William ‘of Vernon’, together with his father Hugh, who had by now become a monk, gave to Holy Trinity, Rouen, land at Martainville within five miles of the city.19 It is probable, moreover, that the full lordship of Vernon passed to this family at some time between these two dates. Early in his reign Duke William had given Vernon to his cousin Guy of Burgundy, and Guy's disgrace and forfeiture after 104720 may well have provided the opportunity for the rise of the new family. A charter for Saint-Père of Chartres,21 which was passed before 1061, and probably before 1053, shows that at that time the family had then obtained full lordship of Vernon, together with its castle;22 and William of Vernon retained this lordship until after the Norman conquest, his last recorded act being in 1077 when he made a grant to the monastery of Le Bec.23 It is rare indeed that the origin of a Norman territorial family of the second class can be illustrated with this particularity from the independent charters of four religious houses.
Finally, there may be taken the instance of the family of Montfort-sur-Risle.24 The first known ancestor of this house is Thurstan of Bastembourg who appears in a ducal charter of 1027 as possessed of land at Pont-Authou, and who, perhaps, also subscribed two charters for Saint-Wandrille which passed at the ducal court at about the same time.25 This Thurstan had a daughter, with whom the notorious Geré of Échauffour fell in love at first sight while dining with her father,26 and also two sons – William Bertram, who can perhaps be seen in charters for Le Mont-Saint-Michel, and Hugh I of Montfort (some five miles from Pont-Authou) who perished in private war with Walchelin of Ferrières during the anarchy.27 It was, however, the son of Hugh I, namely Hugh II of Montfort, who finally established the fortunes of the house. He was one of the leaders of the Norman forces at Mortemer, and between 1060 and 1066 attested ducal charters for Bayeux and Caen.28 He fought at Hastings, and so influential had he become by that time that in 1067 he was left in England to assist in the government of the kingdom during William's absence and placed in charge of the important castle of Dover.29 In due course he was to become a great landowner in England, and in Normandy to add to hishereditas of Montfort the distinct honour of Coquainvillers.30
The four families, whose origins have here been briefly illustrated, may safely be regarded as typical of the Norman aristocracy which came to supply the greater part of Norman strength in the time of William the Conqueror. It will be noted, moreover, that in all these cases the family only acquired the lands from which it took its feudal name during the earlier half of the eleventh century. At a later date the family of Tosny might claim descent from an uncle of Rolf,31 but its association with Tosny only started with Ralph II or Ralph I, and its feudal greatness only began with Ralph III. In like fashion the lords of Beaumont might subsequently cite remote ancestors, but the authentic history of their house really starts with Humphrey of Vieilles, and their full power was only attained under a man who was to survive the Conqueror. Again, the first family of Vernon has been shown to have acquired its domainal lands between 1035 and 1053, whilst the house of Montfort-sur-Risle began about the same time with Thurstan of Bastembourg, and only achieved greatness with a man who is recorded in Domesday Book. The history of these families therefore points inexorably to an important conclusion. The Norman aristocracy which surrounded Duke William, and which he led to the conquest of England, was in his time of comparatively recent growth.
The manner in which this new aristocracy acquired its lands can only be sparsely demonstrated in the available documents, for the surviving texts are not sufficiently numerous to reveal in any comprehensive fashion what was the distribution of land in Normandy before the establishment of the great feudal honours. The pedigrees of Robert of Torigny suggest, however, that the advancement of the kindred of the Duchess Gunnor, widow of Richard I, may have been a factor in the rise of many Norman houses during the latter years of the reign of his son,32 and there is charter evidence which indicates that some at least of these families acquired lands which had formerly belonged to the ducal dynasty. Thus among the possessions of the Duchess Judith, first wife of Duke Richard II, was a large block of territory in the Lieuvin.33 After her death most of this went to the abbey of Bernay,34 but some of the manors do not seem to have been thus disposed, and of these Ferrières-Saint-Hilaire and Chambrais which had both belonged to the Duchess Judith became the special endowment of one of the new Norman houses. Walchelin of Ferrières was clearly established at that place before his death about 1040, and Chambrais (which is the modern Broglie) probably came into the possession of this family at about the same time, since Chambrais, situated but three miles from Ferrières, was at a later date held by the lords of Ferrières in demesne, and afterwards became the head of their Norman barony.35
A clearer example of the acquisition by a feudal family of lands, which had earlier been part of the ducal demesne, can be seen in the descent of the possessions of Count Rodulf, half-brother to Duke Richard I.36 Among the lands held by this man were estates situated on the Risle near Saint Philibert; estates on the Eure, including Cocherel, Jouy, and, it would seem, Pacy; lands dependent on Breteuil; and lands centred on Ivry. Many of these lands, particularly those on the Eure, were inextricably intermingled with the earliest demesne of the Norman dukes and must have come to Rodulf through his stepfather or his half-brother. Their subsequent devolution is thus of particular interest. Part of the Ivry lands went to the count's eldest son Hugh, bishop of Bayeux, whilst the barony of Saint Philibert passed through the count's second son, John, bishop of Avranches, to that cathedral church. But the larger part of Rodulf's possessions, including the honour of Pacy and the distinct honour of Breteuil, descended through the count's daughter, Emma, to her husband Osbern, the steward of Duke Robert I, and one of the guardians of the infant William. And Osbern was typical of the new aristocracy which at this time was rising to power. Few of his wide estates had been held by his father Herfast, whose meagre hereditas passed eventually to the monastery of Saint Père of Chartres.37 Osbern's own extensive lands were acquired by him between 1020 and 1040. Carved out of the original demesne of the ducal dynasty, they passed after 1040 to Osbern's great son, William fitz Osbern, the future earl of Hereford, and became the endowment of one of the greatest feudal honours of Normandy.
If the new magnates thus enriched themselves with lands which had previously been in lay possession, they also effected a considerable spoliation of the Church. It was not for nothing that an ecclesiastical synod before 1046 denounced prelates in the province who gave lands to laymen.38 When Ralph II of Tosny went to Apulia he was already known by the name of his chief Norman possession which had from an early date belonged to the cathedral church of Rouen, and if this particular alienation may perhaps be explained by the kinship between Ralph and Archbishop Hugh, none the less about the same time other archiepiscopal estates centred on Douvrend passed likewise into lay hands.39 Similarly, Robert, bishop of Coutances, was accused of having bestowed cathedral prebends on his relatives to be held by them as lay fiefs.40 It is probable, however, that the older monasteries of Normandy were the chief victims. As late as 1025, for instance, the abbey of Bernay was in possession of Vieilles, Beaumont, and Beaumontel which had been given to the monks by the Duchess Judith,41 but shortly afterwards Humphrey ‘de Vetulis’ obtained Vieilles, and Beaumont was ceded to him before 1035.42 The best example of this process might, however, be taken from the history of the family of Montgomery which seems to have acquired many of its earliest possessions from monastic lands. Roger I, the first ascertainable member of that house, is reported to have abstracted lands in the possession of Bernay,43 and between 1028 and 1032 he seems to have acquired Vimoutiers,44 which in 1025 had been held by the monks of Jumièges.45 Again, a charter of Duke Richard II displays Troarn with its dependencies as belonging to the abbey of Fécamp, and in 1025 this abbey also held Airan and Almenèches. Yet Troarn and Airan appear to have been possessed by Roger I of Montgomery, while Almenèches is known to have belonged at some time to his son Roger II.46 Three of the oldest monastic houses of Normandy clearly had reason to regret the rise of Montgomery in the second quarter of the eleventh century.
Such transactions may safely be regarded as representative, for our knowledge of them depends upon the chance survival of texts from a remote age, and on the possibility of making certain identifications of the places mentioned in them. The spoliation of the Norman church by the new Norman aristocracy must have been considerable, and the numerous new religious houses, which in the latter part of the eleventh century were founded by these magnates, were often endowed by them with lands which had fairly recently been taken from the older ducal foundations. It should, however, be noted that these ecclesiastical alienations would tend to figure with undue prominence in the documents, and the transference of lay lands must also have been extensive, even though for the most part it was unrecorded. Only because the monastery of Saint-Taurin of Évreux was interested in the property is anything known of the manner in which Meules, which was part of the ducal demesne in the time of Duke Richard I, passed into the hands of the family of Gilbert of Brionne, the count, to supply at last a territorial name for the first Norman sheriff of Devon.47
The large-scale transference of landed property which created the new Norman aristocracy involved a tenurial revolution which coloured the whole history of the duchy during the Norman reign of Duke William II. For while the beginnings of this social movement may be referred to the first half of the eleventh century (though not earlier) its end had not been fully accomplished even on the eve of the Norman conquest. The disorders in the duchy between 1035 and 1060 provided plentiful opportunity for men of the new aristocracy to win estates and power by means of the sword, and each crisis within that period involved the fate of families which were later to rise to dominance in Normandy and England. The disturbances of the minority directly concerned the fortunes of Tosny, Beaumont, Montgomery, Ferrières, and Montfort, whilst the campaigns of 1047 and 1051 not only brought misfortune to many magnates in Lower Normandy but also influenced the rise of men as far to the east as William of Vernon. But probably the greatest changes in this period occurred after the forfeiture of William, count of Arques, in 1053. The count's possessions had stretched westward across the Seine,48 and in that region Beaumont and Montfort were ready to take advantage of his disgrace, whilst within the Pays de Talou itself the upheaval must have been even wider spread. Thus, while Bolbec, some twenty miles from Le Havre, seems to have been the original home of the Giffards,49 they were enabled about this time to establish themselves at Longueville in the heart of the Pays de Talou, and this was to remain the head of their Norman honour.50 Similarly, the establishment of Warenne at Bellencombre certainly took place during these years, and as a result of these events.
The rise of this aristocracy inevitably presented a problem of special urgency to Duke William, the more especially as several of these families not only obtained extensive possessions but also acquired official positions within the duchy. Emphasis has already been laid on the importance of the vicomte in bridging the transition between Carolingian Neustria and ducal Normandy, and in providing an instrument of administration for the descendants of Rolf. It was therefore a matter of great moment when during the earlier half of the eleventh century many of the chief Norman vicomtés themselves passed into the hereditary possession of some of the most important of the feudal families. Nigel of Saint Sauveur, vicomte of the Cotentin, who rebelled in 1047, was either the son or possibly the grandson of the first man known to have held this office, and he was himself a lord of great power. He was to regain his vicomté after Val-ès-Dunes, and to survive until long after the Norman conquest.51 Equally notable were the hereditaryvicomtes of the Avranchin. Richard, later vicomte of Avranches, who was established by 1046,52 and who was perhaps also lord of Creully,53 was the son of Thurstan Goz, likewise a vicomte54 and he was to continue in that office at least as late as November 1074.55 The Bessin, too, produced an outstanding dynasty of vicomtes. At the beginning of Duke William's reign the vicomte of the Bessin was Rannulf, who was the son of a vicomte named Anschitil.56 He married a daughter of Duke Richard III57 and was among the defeated rebels at Val-ès-Dunes. None the less, the office continued in the family, for he was succeeded by another Rannulf (II) who was established at Avranches before the Norman conquest, and who survived until after April 1089.58 Moreover, this second Rannulf married Maud, daughter of Richard, vicomte of the Avranchin, thus linking together two powerful vice-comital dynasties which were later in turn to determine the succession of the earldom of Chester.59
These descents have far more than merely a genealogical interest. They reflect the rise of great feudal families whose possession of the vice-comital office was to influence the growth of Normandy and the fate of England. Moreover, whilst the greatest vice-comital dynasties were to be found in Lower Normandy – in the Cotentin, the Avranchin, and the Bessin – the same process could be watched elsewhere in the duchy. In 1054, for example, one Rainald was vicomte of Arques.60 His possessions passed in due course to Goscelin, son of Hedo, vicomte of Rouen,61 and Goscelin's daughter married a certain Godfrey who in turn became vicomte of Arques.62 The feudal connexion thus exhibited between the vicomtés of Arques and Rouen in these years is not unremarkable since it affected the social structure of Upper Normandy at a critical period. But farther to the west, in the centre of the duchy, a more striking manifestation of the same process was taking place. Between July 1031 and July 1032 there was included in a charter for Saint-Wandrille the sign of ‘Roger, vicomte of the Hiémois’, and this was none other than Roger I of Montgomery.63 His father is unknown, and his career is obscure, but he was to be succeeded by his famous son, Roger II, who was to bring the family to its full power. Already in 1051, as has been seen, Roger II of Montgomery was distinguishing himself round Domfront, and at about the same time he made his brilliant marriage with Mabel, heiress to many of the lands of Bellême.64 From that time forward his career was bound up with that of William the Conqueror, but he was still proud to be vicomte of the Hiémois even after (about 1075) he had become the first earl of Shrewsbury.65
The establishment of the great vice-comital families was fraught with danger to the ducal power. But it also provided Duke William with a great opportunity which in the event he was not slow to seize. The vicomte had been, and was to remain, the chief agent of ducal administration, so that while the acquisition of vicomtés in hereditary possession might in a special way challenge the ducal authority, it might also be made to provide the means whereby prominent members of the new nobility could be made to act, like their predecessors in this office, as deputies of the count of Rouen. As will be seen,66 Duke William was enabled, even in the changed conditions, continuously to employ his vicomtes as agents of his administration, but his success in so doing none the less depended directly upon his solution of the more general problem which confronted him in respect of the aristocracy. For the vice-comital families were but a section of that aristocracy, and their rise was but part of the same process which at the same time led to the advancement of other families of equal or greater power. The progress of Tosny and Beaumont, for example, was not essentially different from that of the hereditary vicomtes of the Cotentin and the Bessin, and the possession of the vicomté of the Hiémois was as much a result as a cause of the enrichment of Montgomery. The crucial task for Duke William was thus in this matter the same throughout. It was to maintain his own position within a social structure which throughout his reign was being progressively modified by the rapid acquisition of landed wealth by members of a new nobility.
For the establishment in Normandy between 1030 and 1060 of so many of the families which were later to dominate the feudal province entailed also the advancement of their dependants, and contributed to the formation of a new political organization of the duchy based upon lordship and vassalage. The chief tenants of the greater Norman lords in England after the Conquest often bore territorial names which reveal their families as near neighbours of their lords in Normandy;67 and it is impossible to avoid the conclusion that they owed their rise to power to an earlier connexion with the great feudal houses from whom they were later to hold their widely scattered English estates. Sometimes indeed positive evidence of this can be found in pre-Conquest Norman documents. The connexion between Pantulf and Montgomery, for instance, which was to be so strikingly exhibited in Shropshire at the time of Domesday Book may thus, for instance, be confidently referred back to the time of Roger I of Montgomery, who between 1027 and 1035 issued for the abbey of Jumièges a charter which is subscribed with the sign of ‘William Pantulf’.68
The best example of the early growth of Norman vassalage may, however, be found in the dependency of Clères upon Tosny which likewise was to continue in England after the Conquest. At the end of the eleventh century, Gilbert of Clères, the son of Roger I of Clères, gave land at La Puthenaye to the abbey of Conches with the assent of Ralph (III) of Tosny ‘to whose fief it belongs’.69 Similarly, the same Ralph of Tosny, between 1071 and 1083, confirmed to the abbey of Croix-Saint-Leuffroi all the possessions of Gilbert's son Ralph who had become a monk.70 Moreover, shortly before the Conquest, Gilbert's father – Roger I of Clères – made a grant to Saint-Ouen of Rouen with the assent of his lord Ralph (III) of Tosny,71 and at about the same time he gave land to the abbey of Conches for the soul of his former overlord Roger (I) of Tosny, and with the assent of Ralph (III) of Tosny who was then the chief lord of the fief.72 It is seldom, indeed, that early Norman vassalage can be illustrated with this particularity, and in this case the connexion between the two families can be noticed at an even more remote date. Among the numerous acts of violence characteristic of Duke William's minority two are in this respect especially heinous. Roger I of Tosny was killed by Roger of Beaumont, and shortly afterwards Robert of Beaumont (Roger's brother) was assassinated by Roger I of Clères.73 In the light of subsequent evidence it is impossible not to regard this latter act as the revenge of a vassal for the murder of his overlord. The dependence of Clères upon Tosny (which in fact was to continue until the thirteenth century) can thus be traced back to the beginning of the reign of Duke William.
Such early connexions between Norman families are challenging, but it would be wrong to deduce from them a conclusion that before the Norman conquest the structure of Norman society had as yet been made to conform to an ordered feudal plan. Norman charters of the period 1035–1066 reveal unmistakably in the duchy a society which was based upon vassalage, but with equal certainty they display a social structure in which feudal arrangements had not as yet been reduced to a uniform pattern capable of precise definition in the interests of the duke as feudal overlord. Dependent tenure is, however, to be met with on every side, and gifts of land seem normally to have required the consent of a superior. When a certain Urso, about 1055, gave land to Holy Trinity, Rouen, he asserted the previous consent of his lord (who was dead), and also the approval of his lord's wife and sons;74 and when, about the same time, Ansfred, son of Osbern the vicomte, granted his hereditary possessions to the same monastery, he did so with the consent of ‘my lords Emma, the wife of Osbern the steward, and her sons William and Osbern’.75 Such vassalage could in fact be very widely illustrated, but it would seem impossible to define it in terms characteristic of later feudal society; and the drafters of Norman charters at this time made little attempt to do so. Right down to the Norman conquest the most common word for a dependent estate in Normandy was the older term beneficium. About 1050 Rodulf I of Warenne, for instance, assigned to Holy Trinity, Rouen, land which had previously pertained to the ‘old benefice’ of a certain Roger,76 and when Guidmund gave land in Normandy to Saint-Père of Chartres he did so with deference ‘to my lord William the count from whom I hold it as a benefice’.77 In like manner when Guazo in 1060 founded the priory of Croth in the Évreçin, his endowments were made with the consent of his lord Hugh Bardo, ‘part of whose benefice they were’.78
The terminology of the Norman charters of this time is characteristic of an age in which feudal obligations have not yet been fully defined and few of these deeds use any of the terms of feudal status in anything like the precise sense they subsequently acquired. The word baro, for instance, is rare in Norman documents of this period, and much more frequent is the vaguer term fidelis (liegeman). Thus Gilbert, son of Erchembald the vicomte, is the fidelis of Osbern the steward;79 Oylard is the fidelis of the Countess Lesceline of Eu;80 and Roger, son of Humphrey, is the fidelis of Duke William.81 Most significant of all is the vague connotation in these texts of the word miles, which later acquired the meaning of ‘knight’. The father of William of Warenne can, for instance, be described as a miles;82 so also can the founder of the priory of Croth; and so also can a multitude of lesser men. It would indeed be very rash even to suppose that at this time the word miles inevitably designated a mounted warrior. Many of these men undoubtedly fought on horseback at Val-ès-Dunes and Mortemer, but it would be difficult to assert what was the precise status of those seven equites given to the abbey of Saint-Ouen by Osbern of Écquetot before 1066,83 or that of Atselin equitis mei whose land Duke William himself, about 1050, confirmed to the monastery of Saint-Wandrille.84
There thus seems little warranty for believing that anything resembling tenure by knight-service, in the later sense of the term, was uniformly established, or carefully defined, in pre-Conquest Normandy. Such definition as was attained in the duchy during the Conqueror's reign is recorded for the most part in Norman charters issued after the conquest of England. Nor is there anything in the testimony to suggest that before the Conquest, Normandy had been made generally familiar with the ‘feudal incidents’ which half a century later were so carefully to be discussed after the coronation of King Henry I of England. In all the texts that have been examined there would appear to be only one mention before 1066 of the payment of a ‘relief’, and this occurs in a charter which was issued only a very few years before the Norman conquest.85
The evidence reveals within the duchy between 1035 and 1066 a society in a state of flux, and it must in this sense qualify any such generalization as that ‘Norman society in 1066 is a feudal society, and one of the most fully developed feudal societies in Europe’.86 Dependent tenure, of various kinds and in different degrees, was widespread through Normandy, but there seems little reason to believe that feudal law and feudal obligations were generally accepted and enforced in the duchy at this time in the manner characteristic of a later age. The feudal order which the Conqueror was able to impose upon England before his death was much more developed, and much more centralized, than anything that can be discovered in Normandy before 1066, and it was his success in his kingdom which enabled him to attain in his duchy a part of those feudal objectives which might otherwise have eluded him. If Normandy gave feudal principles to England, England under the Conqueror profoundly influenced Normandy in the matter of feudal practice. Before 1066 older notions of vassalage were still widespread in Normandy, and still imperfectly co-ordinated in a regular feudal scheme. It was therefore a vital problem for Duke William whether he could so link his own interests with those of the aristocracy which surrounded him between 1047 and 1066 as ultimately to dominate the developing social order to which they belonged. And upon the outcome of this endeavour was to depend much of the future history of two countries, for if it had failed, the Norman conquest of England would have been impossible.
The character of Duke William's success in this matter can only be appraised in relation to the difficulties he faced and the manner in which he overcame them. During the earlier years of his reign the duke's position in the new social order might seem to have depended less on the enunciation of legal theory than upon the creation of such personal, and particular, loyalties as the young ruler could elicit from the magnates who surrounded him. In the midst of a social transformation controlled by great families eager to extend their newly won power, it was essential for the duke judiciously to favour the fortunes of those upon whom he could rely. This he seems to have done from an early date as a matter of definite policy. William fitz Osbern and Roger II of Montgomery, themselves still ‘young warriors’, had already in 1051 been selected by the duke for his special confidence,87 and both these men consistently enjoyed their master's support in their advance to power, and co-operated with him closely, both in Normandy and England. To reward such men could, however, before 1066, only be done by impoverishing the ducal demesne, or at the expense either of their fellow magnates in Normandy or of the Norman church, so that the policy was always fraught with the danger of open conflict.
It may, therefore, be regarded as a sign of the growing strength of the duke that in 1055 or 105688 he was able very summarily to disinherit William Warlenc, count of Mortain, and to establish in that comté his own half-brother Robert, who later guarded Norman England from the Sussex rape of Pevensey, and became one of the richest English landowners. Often, however, the advancement of the duke's supporters came about with greater hazard. One of his first acts appears to have been to grant the comté of Arques to his uncle, William, on the understanding that the new count, ‘having obtained this benefice would in return remain faithful in all things to the duke’. The confidence was woefully misplaced as the great revolt of 1052–1054 was to prove, but the count's defeat, as has been seen, enriched several other families, and the duke was able to use this situation in its turn to his advantage. It was, for instance, by William's direct intervention that the count's forfeiture was made to entail the decline of Mortemer and the rise of Warenne.
The early history of Warenne89 may indeed serve as an excellent demonstration of the duke's feudal policy during these critical years, and at the same time illustrate how, as a result of that policy, a particular family might be made to rise from small beginnings to great power. At the opening of the reign, Warenne was of little account. A certain Rodulf of Warenne can be discerned holding some estates near Rouen,90 and he apparently survived until 1074.91 He had, moreover, two sons, Rodulf the elder and William.92 As a younger son this William can have inherited little of the meagre hereditas of his father.93 Yet it was none the less this man who established the greatness of the house. In the campaign of 1052–1054 he distinguished himself by his special loyalty to the duke, and though still a young man (tiro legitimus), he was, after the fall of the count of Arques, singled out by the duke for special favour. It is specially recorded that when, after the critical campaign, Roger of Mortemer forfeited a large part of his Norman lands, the castle of Mortemer itself was bestowed upon William of Warenne.94 Nor was this all; for the extensive lands later held by Warenne in the neighbourhood of Mortemer must have been acquired at the same time and as a result of the same forfeiture. Such at least must have been the origin of the Warenne possessions round Bellencombre (some fifteen miles from Mortemer), and those situated in the region of Dieppe some eighteen miles to the north of Bellencombre. It was a notable advance, but even the acquisition of these lands was not of itself sufficient to raise Warenne to the front rank of the Norman aristocracy. Only after 1066 was this to be achieved, and then too as a consequence of continuing service as a specially trusted adherent of the Conqueror.
To establish relationships such as these, which might assure him faithful service from important men in his duchy was a cardinal feature of the policy of Duke William during his Norman reign, and upon his success in this matter depended to a large extent the position he was eventually to occupy in the feudal order that was soon to be established. The suppression of successive revolts in these years did not merely enable the duke to survive; it enabled him in each case to reward his friends out of the lands of his defeated enemies, and very gradually to mould the developing feudalism of Normandy according to a pattern that might subserve his own interests. The process, which must be sharply contrasted with the more sudden introduction of military feudalism into England, was in Normandy a slow one, and only after a considerable time did it make possible the practical application in the duchy of those feudal principles which the Normans were so soon to introduce into England. In his kingdom, the Conqueror, by interpreting Norman feudal custom in a sense advantageous to himself, was enabled to assume, as if by legal right, a position in the feudal order to which he had long aspired, but which he never wholly attained in Normandy before the Norman conquest.
It would therefore be misleading to describe the aristocratic structure of pre-Conquest Normandy in the light of the conditions prevailing in England, or indeed in Normandy, at the close of the Conqueror's reign. In particular, there would seem no reason to suppose that before 1066 the greater Norman magnates had been made generally to regard their position as depending upon ducal grant, or as conditional on their performing military service for the duke with a specified number of trained and equipped knights. The principle of the ‘service owed’, the servitium debitum, precisely defined and rigidly enforced, was one of the cardinal features of feudal organization in the Anglo-Norman kingdom between 1070 and 1087.95 But it is very doubtful how generally it had been applied in the duchy before the Norman conquest of England. The magnates who were most directly concerned had, as has been seen, established many military tenures on their own lands, but, as must be supposed, they had done this in their own interest, being compelled in a fiercely competitive society to take all possible measures to safeguard the possessions they had so recently won. The feuds during the minority in which so many members of these families perished made it necessary for them to attract to their support as many dependents as possible, and the persistence of private war as a recognized institution in Normandy certainly encouraged sub-infeudation far in excess of any demands that the duke might have felt himself able to make.
Perhaps, indeed, there may here be discerned a factor of feudal development which has received less attention than it deserves and which is particularly applicable to the situation here being discussed. It is generally assumed that the imposition of a fixedservitium debitum was in the interests, and for the benefit, of the ruler, and such was undoubtedly the case in England after the Norman conquest. But in pre-Conquest Normandy the situation may have been different. The fixed payment of the ‘relief’ was welcomed at a later date by the feudal tenants of the king as protecting them against arbitrary and excessive demands. A fixed servitium debitum may on occasion have been likewise welcomed as conferring a similar benefit in a changing society wherein the increasing authority of the duke might have appeared to menace an aristocracy already entrenched in power. In other words, the imposition in such cases may sometimes have occurred by mutual consent, and have approached very closely to a free bargain. At all events, it was not until after the conquest of England that William was able to attempt a uniform regulation of Norman vassalage in his own interests.
None the less, it deserves full emphasis that the imposition of servitia debita in Normandy began to be widely made during the reign of Duke William II. Indeed there is testimony to show the manner in which the burden was imposed in his time, and sometimes on estates which had not been so burdened when they were originally acquired. Thus in 1072 the abbey of Saint-Évroul owed the service of two knights,96 and since this was not a ducal foundation it is difficult to see how a conditional service to the duke would have been attached to the original grant of these lands, and in any case the servitium debitum of this abbey could not have come into existence before its re-establishment about 1050. There is, again, reason to suppose that no defined service attached either to the Breteuil or the Ivry baronies when these were in the possession of Count Rodulf early in the eleventh century, and the later service of five knights owed to the duke by the count of Meulan cannot have been established before the acquisition of Beaumont-le-Roger, with its dependencies, by the family between 1026 and 1035, and may very possibly be of still later origin.97 Finally, the service of five knights owed in the twelfth century to the duke by Hugh of Mortemer must derive from the partial restoration of the Mortemer lands after their forfeiture in 1054.98 It is rarely possible to supply such specific illustrations from the surviving testimony. But it may be added as very probable that before his rebellion in 1047 Grimoald had been made to perform knight-service for his lands at Plessis,99 and that between 1060 and 1066 John, bishop of Avranches, rendered a service of five knights in respect of the honour of Saint-Philibert.100
The results of such arrangements were so substantially to increase the military strength of Normandy that it is important to realize how gradually, and with what difficulty, they were achieved in the time of Duke William. At the time of his accession the process can scarcely have begun. The Norman exodus to Italy in the time of his father may be diversely explained, but it is inconceivable that a ruler able to exact a servitium debitum regularly from his magnates would, like Duke Robert I, have allowed so many of them to depart with their followers to distant lands.101 Moreover, the disorder which afflicted Normandy between 1037 and 1047, followed by the continuous war which lasted until 1054, must have prohibited any more general application of ducal rights in this matter. The point, indeed, deserves some emphasis. The period which elapsed between the accession of Duke Robert (1028) and the battle of Mortemer (1054) was precisely that in which the new aristocracy established itself in Normandy. But during these years there can have been little opportunity for formal assertion of the military demands of the duke over his magnates. If the special position of the duke in the military organization of Normandy, and his legal claims on the service of his magnates, were more generally recognized in 1066 than they had been three decades earlier, this must be attributed to particular arrangements which Duke William had from time to time been able to make with individuals, and more especially to his rule over his duchy during the twelve years which preceded the Norman conquest of England.
By the beginning of 1066 the process was well advanced. It seems certain that before the Norman conquest of England, contractual military service had been imposed on all the older Norman monasteries, and on most, if not all, of the Norman bishops. It had been imposed also upon many, if not on most, of the lay magnates. This for the duke was a practical success of a high order. On the other hand, it is by no means certain that even in 1066 the Norman secular aristocracy had been brought uniformly to accept the principle that they held their lands conditionally upon service, or that at that time they had generally, or in a formal manner, recognized the duke as head of a unified feudal order dependent upon the regular discharge of specified military obligations. Only after William had successfully imposed such legal doctrine on his followers in a newly conquered kingdom was he able at last to apply it generally, and as a matter of principle, to his Norman duchy.
Feudal organization developed gradually in Normandy in connexion with the tumultuous rise of a new aristocracy: it was not imposed rapidly, as in England, by the administrative policy of a prince. All the more noteworthy, therefore, was William's success in bringing this movement so much under his control during his own hazardous reign as duke; for the men who at this time first arise to greatness in Normandy were themselves remarkable. Their vigour was astonishing, and if the superabundant virility which was apparent in their private lives was in part responsible for bringing them to dominance in their own province, none the less many of them early learnt that political sagacity which won for them the panegyric of William of Poitiers.102 Yet, stained as they were with many of the worst vices of a violent age, and as yet unorganized in any rigid feudal scheme, they were in very truth unamenable to control. Consequently, it is wholly noteworthy that they found in William a leader able to dominate them by his personality, and capable also of directing their immense energy into the paths of constructive statesmanship. Only thus were these men enabled to claim the future for their inheritance, so that much of the history of Normandy and of England was to be a record of their acts. Not the least of the achievements of William the Conqueror was that before 1066 he made to subserve his purpose the ambitions and the divergent interests of the most remarkable secular aristocracy produced in eleventh-century Europe.
1 Ed. Marx, pp. 320–329.
2 Complete Peerage (G. H. White), vol. XII, part I, pp. 753 et sqq.
3 Eure, cant. Gaillon.
4 Will. Poit., p. 64; Pfister, Robert le Pieux, p. 213.
5 On a possible literary inference deriving from Roger's Spanish exploits, see Douglas, ‘Song of Roland and the Norman Conquest of England’ (French Studies, vol. XIV (1960), pp. 110, 111).
6 Will. Poit., p. 197.
7 Ord. Vit., vol. II, p. 404; vol. IV, p. 183.
8 Cal. Doc. France, no. 1171; Le Prévost, Eure, vol. I, p. 415.
9 Ord. Vit., vol. IV, p. 198; Vita et Passio Waldevi (ed. Michel), p. 126.
10 Complete Peerage, vol. VII, pp. 521–523.
11 Ord. Vit., vol. II, p. 14; vol. III, p. 339; Robert of Torigny, interp. Will. Jum., p. 324.
12 Complete Peerage, vol. VII, p. 521, note ‘c’.
13 Ord. Vit., vol. II, p. 370; vol. III, p. 426.
14 He was a constant witness to ducal charters before 1066. See R.A.D.N., passim.
15 Will. Poit., p. 197.
16 D.B., vol. I, fols. 80, 168.
17 Lot, Saint-Wandrille, p. 54, no. 14.
18 Cart. S. Trin. Roth., no. LXX; p. 460, no. LXXVIII.
19 Ibid., p. 441, no. XXXVII.
20 Ord. Vit., vol. III, p. 230; vol. IV, p. 335.
21 Cart. S. Père Chartres, p. 178.
22 Cart. S. Trin. Roth., no. LXX.
23 Neustria Pia, p. 442; Porée, L'abbaye du Bec, vol. I, p. 373.
24 Douglas, Domesday Monachorum, pp. 65–70.
25 Chartres de Jumièges, vol. I, p. 41, no. XII; Lot, Saint-Wandrille, nos. 9, 14.
26 Ord. Vit., interp. Will. Jum., p. 163.
27 Cal. Doc. France, nos. 703, 704; Will. Jum., p. 116.
28 Will. Poit., p. 73; R.A.D.N., no. 219; Regesta, vol. I, no. 4.
29 Will. Poit., p. 267. See also below, pp. 207, 208.
30 Cart. S. Ymer-en-Auge (ed. Breard), no. I; Cal. Doc. France, no. 357.
31 Ord. Vit., interp. Will. Jum., p. 157.
32 Will. Jum., pp. 320–329; cf. G. H. White, in Genealogist, New Series, vol. XXXVII, p. 57.
33 R.A.D.N., no. 11.
34 Ibid., no. 35.
35 Will. Jum., p. 116; Ord. Vit., vol. I, p. 180.
36 Douglas, Eng. Hist. Rev., vol. LIX (1944), pp. 62–64; vol. LXI (1946), p. 131.
37 Cart. S. Père Chartres, vol. I, p. 108.
38 Bessin, Concilia, p. 42, cl. X.
39 Valin, Duc de Normandie, Preuves, no. 1; Stapleton, in Archaeological Journal, vol. III, pp. 6–7.
40 Gall. Christ., vol. XI; Instrumenta, col. 218.
41 R.A.D.N., nos. 11, 35.
42 Gall. Christ., vol. XI; Instrumenta, col. 199; Robert of Torigny, De Immuatatione Ordinis Monachorum (Mon. Ang., vol. VI, p. 1063).
43 Robert of Torigny, loc. cit.
44 Chartes de Jumièges, vol. I, no. XII.
45 Ibid., no. XIII, but the editor's comments need correction.
46 R.A.D.N., no. 34; Sauvage, L'abbaye de Troarn, Preuves, no. I; Chartes de Jumièges, no. XII.
47 Douglas, Rise of Normandy, p. 19.
48 Chartes de Jumièges, no. XX.
49 G. H. White, in Genealogist, New Series, vol. XXXVII, p. 59. ‘Osbern of Bolbec’ seems none the less a somewhat nebulous figure.
50 Stapleton, Rot. Scacc. Norm., vol. I, p. civ.
51 On all that concerns the family of St Sauveur, and the vicomté of the Cotentin in this period, see L. Delisle, Saint-Sauveur (1867).
52 R.A.D.N., no. 110.
53 L. Musset, Actes Inédits de XIe Siècle, pp. 8, 9. (This is Bull. Soc. Antiq. Norm., vol. LII, 1954.)
54 His reputed father was ‘Ansfrid the Dane’, and his career is illustrated in many charters between 1015 and 1040 (R.A.D.N., passim).
55 Cart. Bayeux, vol. I, nos. 2, 3; Cal. Doc. France, no. 1211.
56 R.A.D.N., no. 111. The career of Anschitil is illustrated in many other charters of an early date, particularly in those of Saint-Wandrille (e.g. Lot, op. cit., nos. 13, 14).
57 Gall. Christ., vol. XI; Instrumenta, col. 70; Robert of Torigny (ed. Delisle), vol. I, p. 34.
58 Cart. Bayeux, no. IV.
59 It seems to have been generally assumed that there were only two Rannulfs, vicomtes of the Bessin, at this time, namely Rannulf ‘Meschin’ who became earl of Chester in 1120, and Rannulf his father who married Maud, daughter of Richard, vicomte of the Avranchin. Chronology, however, makes it imperative to distribute them into three: namely (i) Rannulf, son of Anschitil, who fought in 1047 at Val-ès-Dunes; (ii) Rannulf, presumably his son, who occurs in or before 1066; and (iii) Rannulf ‘Meschin’.
60 Chevreux et Vernier, Archives de Normandie, plate IX. For date, see Haskins, op. cit., p. 258.
61 Martène and Durand, Thesaurus Anecdotorum, vol. I, col. 167; R.A.D.N., no. 72; Valin, op. cit., Preuves, no. I. He is usually described as having been vicomte of Arques, but this can hardly have been the case since he appears in charters alongside of Rainald, vicomte of Arques (Chevreux et Vernier, op. cit., plate IX), and also alongside of Godfrey, Rainald's successor as vicomte of Arques (Cart. S. Trin. Roth., no. I). That Goscelin was vicomte not of Arques but of Rouen is indicated by no. IX of the same cartulary.
62 Ibid.
63 Lot, Saint-Wandrille, no. 13.
64 Above, pp. 90, 91.
65 Below, pp. 294, 295.
66 Below, pp. 141–144.
67 Below, pp. 270, 271.
68 D.B., vol. I, fols. 257, 257b; Chartes de Jumièges, no. XIII.
69 Gall. Christ., vol. XI; Instrumenta, col. 132; Delisle-Berger, Rec. Actes – Henry II, vol. I, p. 553.
70 Lebeurier, Notice – sur Croix-Saint-Leuffroi, p. 46, no. III.
71 Le Prévost, Eure, vol. III, p. 467.
72 Gall. Christ., vol. XI; Instrumenta, col. 132. Round, Cal. Doc. France, no. 625. Ord. Vit., vol. II, p. 403; vol. V, p. 180.
73 Ord. Vit., vol. I, p. 180; vol. II, pp. 40, 41.
74 Cart. S. Trin. Roth., no. X.
75 Ibid., no. XLIX.
76 Ibid., no. XXX.
77 Cart. S. Père Chartres, vol. I, p. 145.
78 Le Prévost, Eure, vol. I, p. 570.
79 Cart. S. Trin. Roth., no. VI (R.A.D.N., no. 96).
80 Ibid., no. LXIX.
81 R.A.D.N., no. 128.
82 Le Prévost, Eure, vol. III, p. 324.
83 Ibid., vol. II, p. 38.
84 R.A.D.N., no. 109.
85 Gall. Christ., vol. IX; Instrumenta, col. 132; Le Prévost, Eure, vol. III, p. 467.
86 Haskins, Norman Institutions, p. 5.
87 Will. Poit., p. 38.
88 William ‘Werlenc’ was still witnessing charters as count about this time (R.A.D.N., nos. 161, 162).
89 L. C. Loyd, ‘The Origin of the Family of Warenne’ (Yorks. Arch. Soc. Journal, vol. XXXI (1933). pp. 97–113).
90 Cart. S. Trin. Roth., nos. XXVII, XXIX, XL, XLI; Man. Ang., vol. VI, p. 1101.
91 Cart. S. Trin. Roth., no. XXXV.
92 Ibid., no. XXXI.
93 Loyd, op. cit., pp. 106–110.
94 Ord. Vit., vol. III, p. 237.
95 Below, pp. 281–283.
96 Red Book of Exchequer (ed. Hall), p. 626.
97 Ibid. Cf. Ord. Vit., vol. III, pp. 263, 336.
98 Ibid.
99 Haskins, op. cit., pp. 16, 17, and the authorities there cited. The inference may, however, have been pushed somewhat too far.
100 Ibid. Cf. Le Prévost, Eure, vol. III, p. 183.
101 Stenton, Anglo-Saxon England, p. 551.
102 Will. Poit., p. 149.