The Norman Conquest of England, 1066–87

Edward the Confessor, last king of the ancient Wessex line, died on 5 January 1066. Next day he was buried in the new abbey he had built at Westminster and Harold was crowned in his stead, scenes graphically depicted on the Bayeux Tapestry. Earl over Wessex, brother-in-law to the Confessor, Harold had dominated England since the death of his father, Earl Godwin, in 1053. The Confessor had no children and had in his last feeble days designated Harold as his successor. The great men gathered at Westminster sanctioned the choice and both archbishops probably officiated at the coronation. Tall, strong, clever and courageous, Harold had long planned this take-over. He was mighty in arms: in 1063 in a campaign by land and sea he had crushed Gruffudd ap Llywelyn, ‘king over all the Welsh’, and laid his head before the Confessor. He was also a clever politician: in 1065 he had condoned, if not encouraged, the northern rebellion against his brother Tostig, earl in Northumbria since 1055, and had thereby removed a rival, conciliated the north and made alliance with the only family whose wealth remotely approached his own, that of Edwin, earl in Mercia, and his brother Morcar, who succeeded Tostig as earl in the north.

Having gained the throne, Harold could be both optimistic and anxious. He had succeeded to a state of great power, or at least of potential power. Yet his title to the throne was open to challenge, notably from Duke William of Normandy, and Harold himself had seen at first hand the cavalry and the castles which made the Norman military machine superior to the Anglo-Saxon.

During the tenth century, the Confessor’s forebears had moved out from their Wessex base and brought under their rule the areas in East Anglia, the midlands and the north seized and settled by the Danes after 865, for the first time unifying England under a single king. The subsequent Danish conquest under King Cnut (1016–35) had shaken up the English aristocracy, bringing new families to the fore (like the Godwins), but had not replaced it with Scandinavians. On the death of Cnut’s son Harthacanut in 1042, leaving no direct heirs, the old Wessex dynasty had been peacefully restored with Edward the Confessor returning from exile in Normandy. The Anglo-Saxon state had three pillars: a pervasive sense of Englishness which held king and people together, something already discussed (above, p. 3); a kingship high in status and strong in administrative structures; and a church, gentry and nobility integrated within the king’s government of the realm.

The ideal of the king as a great warrior was a constant throughout the medieval period. Even the Venerable Bede wrote with admiration of how King Aethelrith of Northumbria, ‘ambitious for glory’, had conquered more territories from the Britons than any other ruler. Yet Anglo-Saxon kingship also glittered as Christian and civilian. The ordo for the 973 coronation of Edgar, shot through with Carolingian influences, formed the basis for all later ceremonies. The king was crowned and anointed with holy oil just as the prophet Samuel had anointed King Saul and King David. He was thus chosen by God, imbued with the blessings of the Holy Spirit, and elevated above all his subjects. Hence it was crowned and enthroned that the Confessor was depicted on his seal. The oath the king took at his coronation stressed his duty to maintain peace and dispense justice. In the late Anglo-Saxon period he was taking increasing steps to do so. In the past the king had not so much punished crime as regulated the compensation payments its victims or their families were to receive from the perpetrators, the role of the courts being to oversee such settlements. Now, alongside this system, the king started to isolate a group of major crimes, namely homicide, robbery, serious theft, rape, arson and treason, which he, and only he, had the right and duty to try and punish, the punishment often being death by hanging. Such offences constituted breaches of the king’s ‘peace’, which by 1066 probably extended throughout the country. This itself was related to the oath of loyalty to the king taken by all adult males, which included a pledge not to be thieves or receivers of thieves. Peasants probably swore this oath when they entered the tithings (as they were later called), the associations of ten adult males whose members stood surety for each other’s good behaviour. In later practice (and quite probably before 1066) the tithing’s failure to arrest a delinquent member resulted in an amercement, a monetary penalty imposed by the king for an offence.

The old English kings had also developed an administrative structure which enabled them to run this peacekeeping system and also to exact their revenues, always the twin purposes of local government. They thought of England as a series of counties, making their announcements to the officials and men of Northamptonshire, Oxfordshire, and so on. The counties were the essential building blocks of the kingdom, and indeed survived unchanged in shape until 1974. By the year 1000 there were thirty-two of them. Some, like Kent and Essex, were ancient kingdoms, others in the midlands were probably created by the Wessex kings as they expanded their rule in the tenth century. Each county was subdivided into hundreds, of which there were over 600 in the thirteenth century. Both counties and hundreds had courts, the former meeting, at least according to legislation, twice a year, the latter once every four weeks. These courts were central to maintaining the peace and settling civil disputes over property. If they were very much ‘popular’ courts run by the leaders of local society doubtless in very different ways, they were also presided over, as we shall see, by royally appointed officials.

The king’s revenues came in part from his lands, which were scattered through most of the country, the proceeds in cash or kind being worth some £6,000 per annum in 1066, according to the figures in Domesday Book. These sums were either paid into the treasury at Winchester or given to the king directly (and if in kind, consumed) when he was in residence in the locality. More remarkable was England’s land tax, a tax with Carolingian parallels but now unique in Europe. It depended on a territorial division called the hide or carucate, which varied considerably in size but was often around 120 acres. Each hide had to contribute so many shillings to the tax (the rate could be varied), with the king expecting a lump sum from each hundred depending on the number of hides within it. If the figures from the Anglo-Saxon Chronicle can be believed (a debated point), early in the eleventh century the geld, as this tax was called, had raised vast sums to pay off the Danes, the £72,000 taken by King Cnut in 1018 being the largest. Equally impressive is the possibility that regular annual gelds between 1012 and 1051 raised up to £14,000 a year to support the king’s fleet and army. Underpinning this taxation and the economy more generally was the superb coinage, universal, exclusive, of high quality, and controlled centrally by the king (see above, p. 30).

To run this system the king appointed local officials of whom the most important were the sheriff and the earl. It was often to them as the governors of the shire that the king addressed his orders and pronouncements. The sheriff was the official who did the day-to-day work of collecting the king’s revenues and executing his orders. If there was no earl in the county, as often there was not, the sheriff was directly answerable to the king. The number of earls varied, as did their powers and the counties over which they presided. Often an earl would head a group of shires, and with the sheriffs in some sense as his deputies, command their military forces and preside over their courts. Both earl and sheriff were present in the Herefordshire county court during an important case in Cnut’s time.

There were various ways in which kings sought to proclaim their will and enforce their decisions in the localities, thus solving the problem of government at a distance. The royal court was both the centre and the cementer of the realm. As it travelled the country the local nobility came in to discuss their affairs with ministers, king and queen, the latter (her role is discussed later) often being at the heart of court ceremonial. In fact the royal itinerary, though imperfectly known, was still largely confined to the midlands and the south. But if the king did not visit all his realm, he brought the realm to him in great assemblies or ‘witans’ where legislation was promulgated and important decisions taken. The king could also send his envoys into the localities. Present at the meeting of the Herefordshire county court referred to above was Tofi the Proud, there ‘on the king’s business’. As his ultimate ‘muscle’, there was a body of troops attached to the royal household and supported by the allotment of land and by pay. Some of these were thegns, others of Scandinavian origin were described as ‘housecarls’, a name sometimes given to the body as a whole. Apart from forming the nucleus of armies, these men watched over the king’s interests in the shires and if necessary put down trouble, for example punishing the non-payment of taxation in brutal fashion.

In granting favours and issuing orders kings also utilized written communication. Indeed they were doing so with increasing efficiency, gradually embodying grants not in the elaborate and wordy ‘diploma’ but in a far more concise and standard form document, namely the ‘writ’ (brevis), essentially a letter authenticated by the king’s seal being attached to it. Where diplomas and writs made concessions to ecclesiastical institutions they were often written by the beneficiaries themselves, but there is considerable evidence that the king relied on his own clerks, however few in number, to write his orders to local officials. In the last years of his reign the Confessor certainly had a chancellor, the official who can be seen later keeping the king’s seal and presiding over his writing office. The train was already starting for its thirteenth-century destination where the chancery controlled the whole government of England through issuing thousands of documents a year, all derived in form from the sealed writ of the Anglo-Saxons.

None of this structure of government, of course, was sustainable without military force. At the heart of the king’s armies were the troops of the royal household just described. They were supplemented through a system (set out most fully in Domesday Book for Berkshire) in which every five hides of land had to provide one warrior for the army or one sailor for the fleet, a great lord thus supplying a contingent in accordance with the hidage of his estate. The Old English kings could certainly raise armies of very substantial size, as the ability to fight three battles in 1066 shows. There was, however, one key feature about these forces. They seem to have been composed of foot soldiers, not combined foot and cavalry. At Hastings the English fought exclusively on foot, just as they had at the battle of Maldon in 991. In 1051 the reason for an English defeat in Wales, according to the twelfth-century chronicler John of Worcester, was because they had been ordered to fight on horseback ‘contrary to their custom’. There was also one other key feature of the English military system. England had walled towns but, as the chronicler Orderic Vitalis pointed out, lacked the kind of fortifications ‘called castles by the French’. There is little to suggest Orderic was wrong in this judgement. Excavations have revealed only one pre-Conquest castle in the sense of a significantly fortified residence. This is at Goltho in Lincolnshire, and there the pre-1066 earthworks were later supplemented by a much stronger Norman motte-and-bailey castle. One qualification for being a thegn, according to an early eleventh-century document, was to have an enclosure with a gate-house, but there is nothing in the historical narrative (about the crisis of 1051–2 for example) to suggest that such residences had much military value. The importance of these features of the English military system, which certainly set it apart from that developing on the continent, the events of 1066 were to show.

If there was a question mark over English military practices, there was none over the way the church, aristocracy and gentry were part and parcel of royal England. The church indeed had helped to unify England in the first place. By 1066 there were twelve English bishoprics, Durham being subject to the metropolitan authority of the archbishopric of York and the rest to the authority of the archbishopric of Canterbury. There were also some forty-five monastic houses (including eight nunneries), the richest, like the episcopal sees, being very wealthy. It was clearly vital for kings to control an institution which held some 26 per cent of England’s land, and they certainly did so, appointing the bishops and abbots and ensuring that church land owed military service as part of the five-hide system. That the monasteries existed at all was thanks to King Edgar (959–75), for he had sponsored the reforms which saw their foundation or re-foundation. They prayed daily for the king and queen and were mostly situated within the monarchy’s midlands and southern bases. The bishop was often addressed with the sheriff and the earl in the king’s pronouncements and attended the county court, ecclesiastical pleas (see below, p. 101) being heard there or in the court of the hundred.

The thegns were the county gentry of Anglo-Saxon England, vital for the governing of the shires. They gave judgements in the county courts; they were addressed along with royal officials in the king’s writs – ‘all my thegns of Somerset’, ‘all my thegns of Berkshire’; and it was they who served in royal armies under the five-hide system. There were perhaps four or five thousand thegns spread through the counties. In theory the minimum qualification to achieve the status included having the residence with gate and enclosure mentioned above together with five hides of land, the equivalent of a medium-sized manor. Some thegns were specially attached to the king, ‘king’s thegns’; others were in the service of great lords.

Lordship too was central to the workings of government and society. The king relied on great lords to bring contingents to the army, give counsel at witans, and hold office as earls. Domesday Book shows that there were ninety or so lords on the eve of the Conquest with wealth substantially larger than the ordinary county thegns and above them there were three great families who combined their own lands with the tenure of earldoms, namely the families of Seward, of the brothers Edwin and Morcar, and above all of Harold himself. There was always a danger, of course, of such men becoming over-mighty, but in general lordly power before 1066 seems to have been less disruptive of the state than in some areas of the continent. This was due both to the absence of castles and to the limited competence of lords in the areas of justice and law and order. Lords before and after 1066 were certainly developing manorial courts but the latter’s purpose was essentially to discipline the peasant workforce and ensure the smooth running of the manor. At a higher level, although at least 100 hundreds in 1066 seem to have been dominated by great lords, lay and ecclesiastical, the king never appears to have formally relinquished control of their courts. Even if he had, provided he reserved the royal monopoly over the serious crimes mentioned above it would only have meant conceding jurisdiction over cases of minor crime and disorder. It was certainly only that which was involved in the grants the king did make of ‘sake and soke, toll, team and infangenthief’, the most important right here being ‘infangenthief’ (often later seen attached to manorial courts) which involved having a gallows on which to hang petty thieves taken red-handed on the lord’s property.

The structures of Anglo-Saxon monarchy might be strong, yet individual kings weak, as a result of personal inability and political circumstance. Weakness in a king could also have an effect on structure. The Confessor reduced the number of hides on which some hundreds paid geld and (in 1051) abolished the highly unpopular annual army geld altogether. It is difficult to think that the Confessor was other than dangerously threatened by Harold’s family whose lands in 1065 were £2,000 a year more valuable than his own. Harold, as earl throughout the counties of Wessex, had been encroaching on royal lands and absorbing them into those of his family. He had also recruited numerous sworn followers. Yet these problems vanished with Harold’s accession. The over-mighty earl became the king. In combining his family lands with those of the crown he was far more powerful than his predecessor. Apart from a doubt over its military structure, Harold’s difficulties lay not with the state he inherited but with his political position. His alliance with Edwin and Morcar was fragile, despite his marriage to their sister. His brother Tostig in exile plotted revenge. Nothing could give Harold the prestige of the ancient House of Wessex. His own father, Godwin, who had made the family fortunes through service to King Cnut, came from a Sussex family of merely local importance. Harold’s succession had depended on brushing aside a clear representative of the Wessex dynasty, Edgar Atheling, who at some stage had perhaps received the title ‘Atheling’, meaning throne-worthy, from the Confessor himself. A great-grandson of King Aethelred (971–1016), and grandson of the heroic Edmund Ironside, Edgar was a teenager born in exile, and kept landless in England, doubtless through Harold’sinfluence. Another claimant, William, duke of Normandy, was a very different proposition.

* * *

The basis of Duke William’s claim was that he had been made heir to the throne by the Confessor himself back in 1051. This was probably true enough and was part of the latter’s abortive attempt to throw off the Godwins. Edward naturally looked to Normandy. His mother, Emma, was the daughter of Duke Richard I, the Conqueror’s great-grandfather. Edward himself had spent the period of Danish rule as an exile at the Norman court and it was with Duke William’s support that he had returned in 1042 to secure the English throne. The Normans also alleged that Harold had taken an oath to accept William’s succession. Again true enough, though probably the oath had been forced from him in 1064 or 1065 as the price of release after he had fallen into Duke William’s hands during an ill-fated diplomatic mission.

The precise merits of William’s claim, however, were beside the point. The first part of the Bayeux Tapestry opens and closes with a magnificent picture of Edward the Confessor, huge and imposing, crowned and enthroned. Against him William, holding court as duke of Normandy, bare-headed like everyone else, seems small and insignificant. William had every incentive to reach out for the wealth of England and the manna of a crown. He certainly had the resources to do so. In 911 the Frankish Carolingian king, Charles the Simple, had granted Rouen and its surrounding area to the Viking leader Rollo and his followers. This was the origin of Normandy. In the next century Rollo’s descendants, frequently using the title Duke, established their rule within Normandy’s historic frontiers. Although Viking settlement had probably been quite extensive, the newcomers ultimately lost their connection with Scandinavia and became essentially French in language, politics and social structure.

For the power of the Norman nobility the years between c. 1025 and c. 1050 were critical. The flaccid rule of Duke Robert (1027–35) was followed by the long minority of his son William, the future Conqueror. Only seven or eight on his accession, he was the fruit of an unconsecrated union between Robert and the daughter of a Falaise tanner. The great nobles, many probably descended from long-established families, gained control of both ducal lands and local office, notably that of the vicomte, the duke’s chief local agent. They also increased the number and geographical spread of the men who owed them allegiance, in the process rewarding followers with ‘benefices’ or ‘fees’ in land, and asserting lordship over ‘allodial land’ which families had previously held freely from no lord at all. The nobility likewise seized monastic lands and challenged what had previously been a ducal monopoly by themselves founding numerous monasteries, which protected the souls of the patron family and proclaimed its status. This was a period of violence and instability which spawned nobles – ambitious, aggrieved, unrewarded – who sought outlets beyond the duchy’s frontiers. Thus Normans began to arrive as fighters in southern Italy where they ultimately established their own kingdom. It was the same dynamism which established the Normans in England.

Put like this the Normans appear as a uniquely expansionist, martial and chosen race, and that is how they appear in the marvelling pages of one of the greatest of all medieval chroniclers, Orderic Vitalis (1075–c. 1141). He was born in England of a French father and English mother but became a monk at one of those ‘noble’ monasteries in Normandy, St-Evroult. In fact, however, Norman militancy and expansion were inseparable from a much wider aristocratic diaspora which was Frankish rather than specifically Norman. The nobles of Anjou were just as aggressive as those in Normandy and knights from many parts of France took part in the conquest of southern Italy, as indeed they did of England. Also common to many parts of Europe, it has been suggested, were changes in family structures which help to explain this aristocratic conduct. These involved large kin groups, which had shared property widely on the death of one of their members, narrowing into lineages in which the patrimonial land passed intact to the eldest son. The validity of this hypothesis has been much debated. The evidence is treacherous and may, on one view, reveal less a general trend than a favouring of the direct male line in particular situations. In Normandy, however, the development of lineages does seem to be reflected in the appearance, from the 1040s, of toponymic surnames like Beaumont and Montgomery, the result, it can be argued, of a single line retaining and thus becoming attached to its core properties. If land was concentrated in fewer hands than before it would explain how the nobility gained resources to build castles, found monasteries and extend its lordship over men. It would also help to explain the pressure to expand and conquer, for this solved the problem of younger sons, the custombecoming – perhaps the result of the conquest of England – for acquisitions to go to them, leaving the patrimony intact to descend by primogeniture.

Whether or not this was an exceptionally violent age throughout much of Europe has also been debated. On one view, the impression that it was is simply the result of better documentation. There was certainly nothing new about the military preoccupations of the aristocracy. Many of the ingredients of knighthood can be found in the ninth century. Yet the fact that the Normans did indeed concentrate on and define their status through the practice of arms remains of overwhelming importance. With the rest of the Franks, they favoured above all heavy cavalry, like the contingent pictured on the Bayeux Tapestry charging into Brittany to attack Dinan. A ‘knight’ might simply be a cavalry soldier equipped by a lord. But because the armour and the horse could only be afforded by an elite, the term was also used increasingly as an honorific title to a rank, with initiation ceremonies like that depicted on the tapestry where William ‘gives arms’ to Harold. In one vital respect, moreover, the nobility in this period did gain an entirely new means of wielding power, ultimately violent power. This was the castle. The encastellation of Europe between the tenth and the thirteenth centuries was indeed ‘a development of quite fundamental military and political importance’ (Robert Bartlett). In Normandy, although the evidence is fragmentary, the aristocratic elite seem to have been building castles from the 1030s. Just how second nature it had become to construct such buildings is illustrated by the castle thrown up at Hastings as soon as the Normans arrived in England, another scene depicted in the tapestry.

The Norman nobles were central to the Conquest, but only because William had brought them under control, reviving ducal authority in Normandy after its collapse during his minority. In 1047, in alliance with Henry I, king of France (1031–60), he won a great victory over his domestic foes at the battle of Val-ès-Dunes, and then, in the next decade, turned and beat off the assaults of King Henry and his ally Count Geoffrey of Anjou. William maintained a tight hold over appointments to bishoprics and presided over a series of reforming councils which enhanced the authority of bishops within their dioceses. He also won the favour of the pope and obtained a banner signifying the latter’s approval for the conquest of England. To counter the spread of aristocratic power, he went some way towards establishing a tenurial hierarchy in which nobles, lay and ecclesiastical, did him homage and held their land from him in return for military service, if unspecified in extent. William also prevented the castellans from effectively destroying the public structures of local government inherited from the Carolingians by taking over for themselves the maintenance of law and order. He asserted the right to control private castles and he built castles of his own; the one at Caen, between the two great monasteries founded by himself and his wife, still seems to hinge the city.

By the 1060s the duchy was dominated, under the duke, by a small and tested group of around a dozen nobles, most from long-established families and as a group far more unified than their equivalents in England. Where Harold had quarrelled disastrously with Tostig, William’s half-brothers, Robert, count of Mortain, and Odo, bishop of Bayeux, were among his leading lieutenants. Individually the great men of Normandy were very different, Count Robert stolid and reliable, Odo voluble and boastful. It was Odo who commissioned the tapestry, still at Bayeux, which tells the story of the Conquest. But they were all, as William of Malmesbury put it, ‘inured to war’. The duke was absolutely equipped to be their leader. Bursting with confidence, he could make and take jokes, and accept and reject advice. Above all, he too was a warrior. ‘It was a sight at once both delightful and terrible to see him managing his horse, girt with sword, his shield gleaming, his helmet and his lance alike menacing,’ wrote his chaplain, William of Poitiers. Stout (a very different physical type from Harold), but always leading from the front, William was a brilliant and ruthless exponent of the lightning cavalry expeditions, the siege operations and the burning and pillaging which were the business of war. And it was above all through warfare, external warfare, that William canalized the energies of his men. Here he was helped by events across his frontiers, which freed him from external threat. In 1060 both King Henry of France and Count Geoffrey of Anjou died. Anjou entered a period of weak rule and civil war. The new king of France was a minor, in the custody of Baldwin, count of Flanders, whose daughter William had married, thereby securing an important ally to the north. In 1062 William invaded Maine and wrested it from the count of Anjou. In 1064 or 1065 he led a punitive expedition into Brittany. It was this drive, expertise and organization which was now to be felt in England.

* * *

With promise of great rewards, William raised an army, in part mercenary, from Normandy and other parts of France. And then he was lucky. Hostile winds, so William of Poitiers states, highly unusual in the Channel at that time of year, kept him from crossing while Harold’s army guarded the coast, and then shifted to release him when the army had departed to deal with a crisis in the north. Tostig had thrown in his lot with the king of Norway, Harold Hardrada, who nourished his own claims to the throne, and in September they sailed up the Ouse to York and mauled the forces of Edwin and Morcar at the battle of Fulford Gate. By this time Harold was already hurrying north and on 25 September he surprised and killed both Tostig and Hardrada at Stamford Bridge. Two days later, on the evening of Wednesday, 27 September, the Norman army embarked at St-Valéry-sur-Somme in Ponthieu (whose count was William’s vassal), and next morning landed at Pevensey in England.

William also made his own luck. His military administration had kept his forces in being during the long wait. Some of Harold’s, by contrast, had departed through lack of supplies even before the invasion of Hardrada. William’s strategy once ashore was equally clever. He refused to be drawn inland, away from his lines of communication. Instead he made Hastings his base, built a castle there, and ravaged the surrounding countryside, much of it Harold’s ancestral land. ‘Here a house is burnt’ ran the caption on the Bayeux Tapestry above a picture of a pathetic woman with child in hand fleeing as her home is fired. William’s tactic was to provoke Harold into coming to him and Harold duly obliged. Perhaps he felt his political position would not brook delay; perhaps he was over-confident and sought to repeat his success at Stamford Bridge. Either way, instead of harrying William’s army and blockading the Channel with his fleet, he hastened south to do battle without Edwin and Morcar – without half his army, according to the chronicler John of Worcester (writing between 1124 and 1140).

That battle was joined at nine in the morning of 14 October, six miles north-west of Hastings. Harold had formed up his army on foot in a solid shield wall occupying the plateau, about a quarter of a mile broad and half a mile deep, overlooking a little valley. William had thus to drive his cavalry and infantry up a steepish hill merely to reach the English lines. Soon a contingent of Breton foot in William’s army gave way, and the rumour spread that Duke William was dead. This was the climax of the battle and William was its equal. The tapestry shows him raising his helmet to reveal his face, a necessarily static picture whereas the actual event was fast-moving, William galloping down his lines to inspire his men. Throughout the battle he had far more command and control than Harold who, wedged in his shield wall, was invisible within a few feet of his standard. There is here a wider point. The English, as we have seen, had no cavalry. It was a fatal deficiency. Once the Bretons had given way and panic had spread through the Norman ranks, the moment had come for a cavalry charge down the hill to sweep all before it; as it was, the English charged on foot and were cut off by Norman cavalry. So little indeed did the Normans fear such foot charges that they twice feigned retreat in order to repeat the cutting-off tactic. So Hastings became a killing match and the Normans had more effective means of killing: horsemen, and also archers. The lack of English archers is one of the puzzles of the battle. Only one English bowman is shown on the tapestry, the consequence perhaps of the haste with which the army had been assembled. By contrast the shields and bodies of the English bristle with Norman arrows. As the day wore on, and the English numbers dwindled, the Norman cavalry established themselves on the plateau and broke into the shield wall. They killed Harold’s brothers Gyrth and Leofwine, then Harold’s immediate bodyguard, and then at last, as the shadows of the autumn evening lengthened, they cornered Harold himself, already wounded by an arrow in the head, and struck him down.

Only part of the English forces had been engaged at Hastings, and in London Archbishops Stigand and Ealdred, Earls Edwin and Morcar and the townsmen rejected William’s demands for submission and nominated Edgar Atheling, who was with them, as king. Early in December, William began a long circular march round the city burning and pillaging as he went. Edgar was as powerless to resist him as he had been earlier to resist Harold. When William reached Wallingford, Stigand came to him and swore allegiance; he was followed at Berk-hampstead by Archbishop Ealdred, Bishop Wulfstan of Worcester, and Edgar himself, as well as by Edwin and Morcar, with citizens of London and many others. On Christmas Day, 1066 William was crowned king in Westminster Abbey.

William’s triumph was due to more than luck and superior tactics. It also reflected significant differences between the Norman and English polities, as we have seen. In England the Normans certainly did not meet a state inferior to their own. Quite the reverse. The duke’s role in the maintenance of law and order in the duchy, though significant, was less pervasive than that of the king in England. The duke had no chancellor, no writs, no seal, and no geld tax. He minted a greatly inferior coinage. Yet William conquered, as Cnut had done earlier in the century. The state which the English created but could not defend: there is some truth in that aphorism. In terms of armour, martial values and military effectiveness there was little, in some ways, to choose between the English and the Normans, which was why Hastings was such a long battle. The five-hide system, on paper at least, was a more clearly defined and established method of raising armies than any in place in Normandy. Where the Normans had the edge was not in the way the armies were recruited, but in what they produced. In a conflict like Hastings the simple truth was that a force which could fight on horse as well as foot was superior to one which could only do the latter. Given the amount of training mounted warfare takes, English society was, in that respect, less militarized than Norman. On both sides of his seal as king of England, the Confessor showed himself enthroned. William, on the seal he made once he became king, was enthroned on one side but on the other appeared as a mounted knight. Nothing sums up better this vital contrast between the Norman and the English polities.

There was another significant difference, as we have seen. Orderic Vitalis did not merely mention the absence of English castles. He also affirmed that this was why ‘the English, in spite of their courage and love of fighting, could only put up a weak defence against their enemies’. Fundamentally, when it came to defences the English had fallen between two stools. The defences of royal and magnate residences were too small to allow any effective resistance to the Normans. Those of the towns were too large to do so. According to a document of the early tenth century (‘The Burghal Hidage’), 27,000 men were needed just to defend the boroughs of Wessex, numbers impossibly large in a time of dislocation like 1066 when so many men were already called up in the armies. And there was here a double bind. Just as the absence of castles enabled the Normans to seize England, so it was by building huge numbers, over 500 by the year 1100 according to one plausible estimate, that they secured their rule. The castles of the Conquest, built both by king and lords, took a variety of forms. In London the great stone keep begun by the Conqueror, now called the White Tower, is ninety feet high, has walls over thirteen feet thick, and still looks a mighty building even against the present London skyline. Elsewhere structures were simpler. Some were just ditches and banked enclosures topped with palisades (called by historians ‘ring works’). Others, indeed the great majority, were of the motte-and-bailey type, as at Berkhampstead. The motte was a great earthen mound protected by a ditch and topped with a stockade and a wooden keep, the adjoining bailey an enclosure protected by earthen ramparts, topped again with a stockade, and surrounded by an outer ditch. Such structures could be built speedily and defended by small numbers of knights, yet they were quite sufficient to dominate their strategic environment; that, not the protection of the local population, was of course their purpose. Utterly novel, standing high above town and village, road and river, the castle as both symbol and sanction lay at the heart of the Conquest.

Cavalry and castles were integral to the intensely competitive military and political environment in France, where small principalities – Anjou, Maine, Brittany, Normandy, the French kingdom – were engaged in constant fast-moving warfare across great plains and open frontiers. England was very different. Since Cnut’s accession in 1016 it had suffered neither invasion nor civil war. Harold had triumphed in Wales, but this was not cavalry territory. There had been no need to develop either cavalry or castles. The English state had been too successful for its own good.

* * *

William’s success in 1066 cannot be explained entirely in military terms. Following his initial victories the English had accepted him as king. After all, he proclaimed himself the Confessor’s designated heir and took the same coronation oaths as his Anglo-Saxon predecessors. When he returned to Normandy in March 1067, he left as regents his brother Bishop Odo and his friend since childhood, William fitz Osbern, but he also retained many English sheriffs and recognized and appointed English earls. In the next few years William, faced with a series of rebellions, was utterly to destroy this Anglo-Norman co-operation.

There was immediate discontent at the taxation and the castles which were a concomitant of Norman rule. The Anglo-Saxon Chronicle commented at once on the breach of William’s coronation oath to rule his people well. Worst of all, as Orderic Vitalis later put it, was ‘the anger at the loss of patrimonies and the death of kinsmen and fellow countrymen’. From the start, as a writ to the abbot of Bury St Edmunds shows, William demanded the lands of all who ‘stood in battle against me and were slain’, including the lands of Harold and his brothers. From William’s point of view, these dispossessions were justified on the grounds that Harold was a perjured usurper (in Domesday Book he was not even given the title of king). They were also utterly necessary both to grip the kingdom and, as with taxation, reward those who had fought for his victory and in the case of the Norman monasteries had prayed for it. Yet the dispossessions sent shock waves through English society. They destroyed Harold’s house which had dominated England for a generation and disinherited the families of the other fallen. Thoroughly alarmed, large numbers of English landholders flocked to William or his deputies hoping to secure or recover their lands. Some were lucky, like Azo the steward of the Confessor who found William at Windsor, but many had to offer money. ‘The people bought their lands,’ grumbled the Anglo-Saxon chronicler.

Trouble began before William returned to England in December 1067. In the new year he had to march into the West Country, where he besieged and took Exeter and built a castle in the town. Then he drove on into Cornwall where he ultimately established his half-brother Robert, count of Mortain. The coronation of William’s wife Matilda in May 1068 was the last of the great Anglo-Norman occasions. Earl Edwin found himself threatened locally by Roger of Montgomery’s earldom in Shropshire and challenged at court by ‘greedy Normans’, as Orderic put it, who opposed his prospective marriage to William’s sister. Between 1068 and 1070 he was joined in rebellion by Earls Morcar, Waltheof and Gospatric, whom the Conqueror himself had placed over Northumbria. Most dangerous of all, Edgar Atheling, with succour and support from King Malcolm of Scotland, now revived his claims to the throne. In his first response, in 1068, William marched north, building castles as he went at Warwick (where Edwin and Morcar surrendered), on the rock at Nottingham, and at York. Early in 1069, however, the Atheling attacked York and was chosen king by his supporters. So William went north again and gave York a second castle. It was not enough. In the autumn of 1069 a great fleet sent by Sven Esthrison, king of Denmark (1047–76), a nephew of King Cnut, entered the Humber, and joined up with the Atheling. On 21 September they defeated and captured William’s sheriff of York and seized the city. Meanwhile Exeter and Shrewsbury (the seat of Roger of Montgomery’s new earldom) were also attacked.

This was the crisis of William’s reign and he knew it. He ignored a rebellion which threatened his conquest of Maine, and acted in England with a combination of energy, brutality and conciliation. Leaving Bishop Geoffrey of Coutances to deal with the trouble in the west, he again marched north and on Christmas Day, 1069 wore his crown, especially dispatched from Winchester, in the ruins of York Minster, a symbolic riposte to the pretensions of the Atheling. He then marched to the Tees, ravaging the country as he went. The Danes were bought off, the Atheling retired to Scotland, and Gospatric and Waltheof admitted defeat, retaining their earldoms. But William was not finished. He led his troops on an extraordinary winter march across the Pennines, fell upon the Shrewsbury rebels, built castles at Chester and Stafford, ravaged the surrounding areas, and was back at Winchester in time for Easter (April 1070). By this time his forces in Yorkshire had reduced much of it to a wasteland.

Historians have sometimes been sceptical about the extent of ‘the harrying of the north’ but the evidence is terribly powerful and consistent. William ‘went northwards with all his army that he could collect and utterly ravaged and laid waste that shire [Yorkshire],’ wrote the Anglo-Saxon chronicler. Sixteen years after the event Domesday Book still recorded 33 per cent of Yorkshire ‘waste’ and another 16 per cent as virtually without resources. The sheer scale of the waste, amounting to 80 per cent of that recorded in Domesday, can scarcely be explained by clerks simply writing off areas for which they had no information, as has been suggested. Columns of refugees, young and old, women and children, fleeing the famine caused by ‘the devastation’, reached as far south as Evesham, where they died of weakness even as they ate the food provided by the abbot. Simeon of Durham, writing in the early twelfth century but with local knowledge, likewise spoke of the great famine; the exodus of refugees; the decaying corpses and ‘the land deprived of anyone to cultivate it, reduced for nine years to an extensive solitude…. There was no village inhabited between York and Durham.’ The Conqueror’s knights were masters of the ravage and they rose to the appalling challenge. Small parties of them, moving rapidly from village to village, could easily have accomplished the destruction in the months between Christmas and Easter. They were helped by the winter season. The corn for both eating and sowing was in the barns. By setting them ablaze the food for two years was effectively destroyed. The last word is with Orderic Vitalis, who was born in 1075 and grew up in Shropshire till he was ten:

My narrative has frequently had occasion to praise William, but for this act which condemned the innocent and the guilty alike to die by slow starvation I cannot commend him. For when I think of helpless children, young men in the prime of life, and hoary grey-beards perishing alike of hunger I am so moved to pity that I would rather lament the griefs and sufferings of the wretched people than make a vain attempt to flatter the perpetrator of such infamy.

William’s northern campaign almost ended English resistance, but not quite. Edwin and Morcar had been kept out of the upheavals of 1069–70, but in 1071 they escaped from court. Edwin was soon trapped and killed; but Morcar fled to the Isle of Ely where he joined up with an adventurous Lincolnshire thegn, Hereward (Hereward the Wake of later legend). King William invested the isle and in October 1071 Morcar surrendered, ending his life as a prisoner. Hereward escaped. His further exploits inspired poets but did not threaten Norman rule. The Conquest itself was over. In 1074 the Atheling recognized as much; he left Scotland and became a pensioner at William’s court. Lacking a real sense of dynasty and destiny, his later role was as a captain of Norman and Scottish armies, not as a candidate for the throne. William of Malmesbury provides a last glimpse of him in the 1120s, aged and obscure, living quietly in the country.

* * *

The Normans had come to exploit the peasantry, not replace them. The overwhelming bulk of the English population thus remained in place after the Conquest, if battened down by more exacting lords. The latter, often absentees, strove to get their income in cash rather than in kind, a desire for money which accelerated the end of slavery. Reorganization of manors also led to a substantial decline in the numbers of sokemen and free peasants (see above p. 52). Peasants laboured on the new castles and fled or starved to death when a Norman army burnt its way through the countryside. Towns too suffered. Much of York was wrecked in the great rebellions of 1069–70 while elsewhere houses were pulled down to make way for the new castles and cathedrals. There was some immigration. At York, 145 properties once held by Anglo-Scandinavians were taken over by Frenchmen. There were French quarters too at Norwich and Northampton. But probably the great bulk of the town populations remained English.

If peasants and townsmen remained in place, a huge swathe of English landowners was dispossessed, including virtually all the aristocracy. It was that, more than anything else, which secured the Conquest so absolutely. This disappropriation had begun after Hastings and increased in pace with every rebellion. The result can be seen in Domesday Book which surveyed England both in 1066 and 1086. Only four Englishmen, Edward of Salisbury, Gospatric son of Arnkell, Thorkell of Warwick and Colswein of Lincoln, remained as major landholders. Gone were Harold’s family and those of the other English earls, gone were all ninety or so of the lords who had possessed land worth £40 a year or more, and gone too, at least in the record, were perhaps fifteen or twenty thousand smaller English landholders. Many of the latter were very small fry indeed, which shows how devastatingly low the blades of disappropriation had been set. Others, perhaps four or five thousand of them, were thegns, who, as we have seen, formed the country gentry of Anglo-Saxon England. In many counties page after page of Domesday Book, manor after manor, reveals hardly a single Anglo-Saxon lord. ‘Henry de Ferrers holds Kingston [in Berkshire]. Ralph holds it from him. Stanchil held it in the time of King Edward’ runs a typical entry, where the 1066 lord Stanchil has been replaced by the great Norman baron Henry de Ferrers, who has in turn granted the manor to his knight, Ralph de Bacquepuis. Just occasionally the survival of wills and charters shows what all this meant for a particular English family. For more than a generation prior to 1066 a spread of lands in Lincolnshire had supported Ulf, son of Tope, and his kinsmen. ‘They acted in concert, endowed their favourite monastery (Peterborough), transacted business together, attested each other’s charters and lent each other a helping hand in troubled times’ (Robin Fleming). Within half a decade of the Conquest this family network was no more. Ulf, son of Tope, had departed for the Holy Land, never to return, and the family’s lands were split between five Norman lords. Where families did survive it was often in much reduced circumstances. In Cambridgeshire, Almaer, lord of Bourn, was reduced from an estate of twenty-two hides to one of fewer than four.

William’s Conquest had thus turned out very differently from that of Cnut, who, as William of Malmesbury shrewdly noted, restored their lands unimpaired to the conquered. Cnut had rewarded his followers with money rather than with land. The Conqueror chose land, perhaps inevitably given Norman expectations, and thus set off the vicious cycle of rebellion and deprivation which ended with the elimination of so many English landholders. Malmesbury was a monk, writing in the 1120s, half Norman and half English. He believed that the sins of the English explained the Conquest. Yet he could still write about it with great bitterness. Hastings was ‘a fatal day’, ‘a melancholy havoc of our dear country’, making it ‘the residence of foreigners and the property of strangers’.

The Conquest was, therefore, devastating, but large numbers of Englishmen did survive at levels above the peasantry. Almaer of Bourn was not alone. When a great Norman baron, a Henry de Ferrers or Ilbert de Lacy, swept into an area to take possession of the estates granted him by the Conqueror, he was met by dozens of Englishmen promising faithful service and seeking to obtain or retain land. Some were lucky and, like Aelfwine, tenant of the Ferrers at Brailsford in Derbyshire, were predecessors of major gentry families of the twelfth and thirteenth centuries. Indeed around ten of Henry de Ferrers’s twenty-five tenants in Derbyshire had English-sounding names and the proportion was even higher on the Lacy estates in Yorkshire. Further evidence of English survival is provided by the lists of king’s thegns at the end of every county section of Domesday Book. Most held just a few hides of land, but they were also probably local officials – huntsmen, foresters, sheriff’s bailiffs – and thus significant. Domesday Book may also conceal a whole class of English sub-tenants and officials because it rarely reveals who were the lessees or bailiffs of the manors held by the Norman lords. Thaxted, the most valuable Essex manor of Richard de Clare, was leased to an Englishman. He may well have been one of many. For running the hundred, the basic unit of local administration, the English remained vitally important. Indeed they provided nearly half the jurors drawn from the Cambridgeshire hundreds who gave evidence to the Domesday commissioners. Much later history – the bid made by Norman kings for English support, the survival of the English language – becomes understandable against that background. At the level of the county and the hundred, unless there was to be constant disturbance, Normans and English had to work together. Gradually a new nationality and local society formed to replace the old.

In that formation a significant role was played by women. After the Conquest the Normans had no place for the male kin of the killed and dispossessed. Women, or at least women of a certain status and place in the life cycle, were a different matter. Marriage to the widow or daughter of a thegn might help secure possession of his lands. It was to escape such a fate, or worse, that English women after the Conquest fled to monasteries. But many such marriages did take place, like that between Robert d’Oilly and Ealdgyth, daughter of Wigod of Walling-ford. Of course, the whole purpose of such matches was to divert property away from English kin. But Ealdgyth and the rest cannot have suddenly disowned their Englishness. They passed it on to their children and thus took a first step in bridging the divide of the Conquest.

* * *

The church, with some struggle, by and large kept its lands intact before and after the Conquest, holding some 26 per cent of Domesday England. Beyond these possessions, however, virtually all the land of England had come into the king’s hands, and Domesday Book shows what he did with it, illuminating both the resources of the crown and the structure of the new aristocracy. The first and most striking fact is the amount of land William kept for himself: in all some £12,600 worth, 17 per cent of Domesday England. These royal lands, giving a presence in nearly every shire, were double in value those held by the Confessor. Whereas before 1066 the leading nobles had roughly 16 per cent more land than the king, afterwards it was exactly the reverse. William was thus far more powerful than the Confessor, though that would have been true for Harold too had he remained on the throne.

Apart from his own possessions and those of the church, William had redistributed the great bulk of the land of England to his followers from Normandy and elsewhere in France. The most important land-holders in each county were listed individually at the start of each county section of Domesday Book and came to be called tenants-in-chief of the crown, a ‘baron’ being simply a major tenant-in-chief. In the twelfth century there were between 150 and 200 of the latter. Domesday shows that these tenants-in-chief had themselves granted to their own tenants, in varying amounts, roughly 45 per cent of their land in terms of value. We have seen how Henry de Ferrers gave Kingston in Berkshire to Ralph de Bacquepuis. On the Clare estates, 40 per cent of the land in terms of value was held by around fifty tenants. Similar ‘sub-enfeoffments’ were made by the new Norman bishops and abbots on their church estates. The reasons for such grants were various. One was to win friends and influence people, hence some of the tenants werevery great men, barons themselves or major tenants of other barons. Another reason was to reward followers, so that many tenants of the Clares and the Ferrers were, like Ralph de Bacquepuis (Bacquepuis near Evreux), their tenants or neighbours in Normandy.

The structures of power which resulted from these landed endowments can be analysed in different ways. Where tenants-in-chief had granted manors to tenants, it was generally the latter who enjoyed the revenues. Yet tenants-in-chief still expected service and support from their tenants and continued to have rights over their lands, as will be seen. Thus in assessing their power it is not irrelevant to add the value of tenanted land to that they kept in their own hands, that is ‘in demesne’. Below is an analysis made along these lines, based on the pioneering work of W. J. Corbett published in the 1920s:

Tenants-in-chief, 1086: value of lands in demesne and tenanted

A	£650–£3,240	10
B	£400–£650	10
C	£200–£400	24
D	£100–£200	36
E	£15–£100	200+

Calculated like this the concentration of wealth at the top of the scale is remarkable, for the barons from class A commanded between a quarter and a fifth of Domesday England. Since they were all close associates of the Conqueror, and included his brothers, this explains much of the hold he established over the country.

A rather different picture, however, emerges if one makes a survey of the landed wealth of tenants as well as tenants-in-chief, and does it in terms simply of the land held in demesne. This both hugely increases the numbers of people surveyed and reduces the separation between the great barons and the rest. An analysis along these lines by J. J. N. Palmer shows that 34 per cent of secular wealth (apart from the lands of the king) was in the hands of 940 landholders with lands valued at between £5 and £45 a year. Another 11 per cent was possessed by 1,720 landholders with lands worth between £1 and £5 a year. Below this there were some 3,470 men with lands valued at less than a £1 a year. All told, the analysis reveals over 6,200 landholders. If a proportion of these were English it was a comparatively small one. Having made some estimate of the numbers in Essex, Norfolk and Suffolk (counties excluded from his analysis), Palmer puts the total of Norman land holders (Norman here meaning loosely everyone from France) at over 8,000.

These findings are immensely significant because they explain the grip which the Normans gained on England and the power of local society in later centuries against both great lords and the king. Even if some of these lesser Normans had other sources of income (for example as professional knights) and were largely absentees, the great majority were almost certainly resident on their estates in England. These were the men who were the first to speak the language and intermarry with the native English. That half the personnel of the Cambridgeshire hundred juries were Norman, some holders of substantial properties, others with only tiny holdings, is as significant as the fact that half were English. Nothing shows more clearly how involved the Normans had become from the start in running local affairs in England. The more substantial families with lands worth £5 and upwards came to form the cream of a new gentry class, the county society of the knights coming to replace the county society of the thegns. Their establishment in England was often reflected in place names. Thus the Bacquepuis’s Kingston became Kingston Bagpuize, showing very clearly how this family from the Evreçin had taken root in Berkshire.

In terms of their physical configuration, the lands of individual barons, both demense and tenanted, were often spread over many shires while having a core in one or more particular regions (see below, pp. 404–5). Since such estates were usually built up from the possessions of many dispossessed Englishmen, they were as a whole new creations. A common pattern was for a baron to receive a large proportion of the lands of a few great English nobles together with the properties of numerous smaller fry. Of the Clare lands in Surrey, 40 per cent in terms of value derived from three major ‘ancestors’ (as previous owners of significance were called), while the other 60 per cent came from seventeen men, only three of whom contributed more than one property. Many of the large antecessorial grants took place soon after the Conquest, while the lands of the lesser individuals were swept up later in disappropriations following the rebellions. In all this William showed a keen strategic eye. Outside the Welsh borders, he never gave all the lands in a shire to a single baron, but he was quite prepared to create smaller blocs of power in order to hold down particular areas. In the north, Henry de Ferrers gained nearly all the land within Derbyshire’s Appletree wapentake, while Ilbert de Lacy received nearly all that in Skyrack wapentake in Yorkshire. (A wapentake was the equivalent of a hundred.)In Sussex, William accepted the ancient divisions into rapes, but completely ignored earlier tenurial patterns when he concentrated the land within them under single lords. There was a similar concentration of Clare holdings around their great castle at Tonbridge in Kent.

Politics and family history were always reshaping the contours of great estates, if at a slower pace than in the Conquest period. In the long run, what was more significant was the transformation in the terms on which land itself was held. After the Conquest, the king and nobility, it can be argued, came to form a new kind of tenurial hierarchy. This hierarchy, together with castles and cavalry, lies at the heart of ‘feudalism’ as defined by many English historians, who thus eschew the much broader definitions, including in effect the whole of medieval society, adopted by continental colleagues. The introduction of feudalism to England is a complex and controversial subject; it may be helpful to describe the structures as they can be seen in place after 1066 before going back to consider how far they were comparable with what had existed in Anglo-Saxon England and how far they were significantly different.

Domesday Book makes clear, for example in its entry for Shropshire, that the great lay and ecclesiastical landholders listed at the start of each county section ‘held’ their land from the king. Such men came to be called ‘tenants-in-chief’. In 1086 they had probably all performed an act of homage in which they knelt before William, placed their hands between his and swore allegiance for the land they held from him. The land was usuually called a feodum, that is a ‘fee’ or ‘fief’: hence ‘feudalism’. The ceremony of homage was solemn and consequential. Breach of the oath was treason and involved the forfeiture of the fief. In return for its tenure, the Conqueror required each tenant to provide contingents of knights either for the royal army when it was called out or to guard royal castles, the individual quotas (on the evidence of a survey in 1166) ranging from fifty or sixty knights down to a handful. William’s imposition of such service is described in the twelfth century by Orderic Vitalis and the chroniclers of Abingdon and Ely. If the survey of 1166 does indeed reflect the quotas he determined, as is not unlikely, then the tenants-in-chief as a whole owed some 5,000 knights. Although tenure of this kind came to be called tenure by ‘knight service’, actual military service represented only one part of its value. There were also, probably from the start, what historians have sometimes called ‘the feudal incidents’. When a tenant-in-chief died his heir had to make a money payment to the king to inherit the fee, a payment later described as a ‘relief’ when considered reasonable in size and a ‘fine’ when arbitrary and exorbitant. If the heir was a minor, which was often the case given the prevalence of early death, the wardship of the lands and the revenues from them passed to the king. The king also contolled the marriages of the widows of tenants-in-chief; and he controlled too those of heirs and heiresses when they were his wards. Although it was quickly accepted that the fees were hereditary, the king retained power to influence their descent, especially where there was no direct male heir, and he used this power to raise up those who were in favour and pull down those who were not. Ultimately, in default of heirs the fee would ‘escheat’, that is return to the king, although this was comparatively rare. What feudalism gave the king was thus military service, money, sources of patronage (in the marriages of widows and – best of all – of heiresses), and also social and political control. There was in addition the ecclesiastical equivalent of wardships, since after the death of bishops and abbots the king received the revenues from the lands they had held in chief until a successor was appointed.

Henry I’s coronation charter of 1100 suggests that the feudal structure just described was in place and that relations between the tenants-in-chief and their own tenants had developed on similar lines. The former expected their own tenants to supply quotas of knights which would help them provide the service owed to the king as well as military support at other times, for example rebellion. Very broadly tenants of a single property might owe a single knight, and so on in increasing numbers. A baron, just like the king, through inheritance payments, wardships and the bestowal of marriages gained sources of revenue, patronage and social control. His ability to discipline his tenants and to forge them into a real community was, potentially at least, facilitated by the ‘honourial court’ as it is sometimes called by historians. This was a court held by the baron and attended by his tenants by knight service, a court which had jurisdiction, amongst other things, over disputes concerning the tenure of the fees and the services due from them. The ‘fee’ or ‘honour’ (the words at this level were used inter-changeably) which the baronial tenant-in-chief held from the king and which would pass to his heir was thus composed of his tenants and their lands as well as the lands he had kept in hand, ‘in demesne’.

The feudal structures in England after the Conquest were not imported intact from Normandy. There both the duke and individual lords had been extending their lordship over allodial lands, but the process was far from complete and the rights and service were probably often undefined. Nor did the structures have any general parallel in England before 1066. England was an ‘old’ country with diverse forms of lordship and land tenure. There certainly were tenants holding land in forms comparable with those found after the Conquest (for example, some of the lessees of the bishop of Worcester). Lords also tried to assert lordship over family land, much as they did in Normandy. But Anglo-Saxon wills and charters, together with Domesday Book which throws much light on conditions before the Conquest, suggest there was still a great deal of allodial land held from no one at all. Those with such land might well have sworn fealty to a great lord or in a special way have been in the service of the king, but they were not their tenants. The circumstances of the Conquest meant a complete new start, bringing all secular land into the hands of the king. He gave it out again to men he made his tenants, and likewise made tenants of the bishops and abbots. That was how all land came to be held from the king. The king’s position, unique in Europe, at the head of a tenurial hierarchy with all its attendant rights and revenues, stemmed from these unique events.

Of course, a new form of tenurial hierarchy is still perfectly compatible with kings and lords before 1066 having enjoyed in some respects equivalent powers. When it came to the raising of armies, the contrast between the two systems certainly seems of little moment. The core of royal armies was always provided by household forces, the thegns and housecarls before 1066 and the household knights thereafter. Beyond that, although in neither case is there real evidence for the size of the forces produced in actual practice, the system in which lords owed so many warriors according to the hidage of their estates seems just as good as one in which they owed a number of knights (unrelated to hidage) as determined by the Conqueror.

In other ways too kings and lords before 1066 exercised powers comparable with those of their post-Conquest successors. Land was certainly forfeited for breach of faith. Indeed it was forfeited to the king for any failure to turn up when summoned to the army. Both king and lords had large numbers of followers in their special allegiance, and on the death of such men they could demand a ‘heriot’ from their families, a death duty rather than the later feudal payment to inherit, but probably just as valuable.

Yet despite these similarities, the change brought by the Conquest was still momentous. The fact was that kings and lords before 1066 wielded much less power over their men than their post-Conquest successors. Since there was often no tenurial content to lordship, they lacked the same ability to manipulate the descent of land and take possession of it when heirs were under-age; there are no pre-1066 references to wardships. By the same token they also had less power over the marriages of women, and therefore significantly fewer resources of patronage, something discussed more fully in a moment. The king’s exploitation of his new feudal rights was absolutely central to the workings of politics and society in the century and a half after the Conquest. It could be a major cause of friction between the king and his baronage as the concessions in the early clauses of both the Coronation Charter of 1100 and Magna Carta in 1215 show. At the level of the baron and his tenants the honour and its court were also new. If lords before 1066 sometimes had rights of soke or jurisdiction over extended areas it is difficult to see how this was equivalent, let alone the origin (as has sometimes been suggested), of later honourial jurisdiction because the key to the latter was the kind of feudal tenure which simply did not exist in any general way before the Conquest. There has been debate about the importance of the honours after 1066. They were never self-contained and autonomous institutions since from the start there were tenants who held from several honours or were tenants-in-chief themselves. Honours also had very different histories and came in various shapes and sizes. Yet when all these qualifications have been made, in the century after the Conquest and for many years thereafter honours formed an essential element in magnate power. (For further discussion, see below, pp. 404–7.)

Associated with the introduction of feudalism were wider changes in the structure of the family. England too saw the transition from the extended to the linear family, which, as we have noted, has been detected taking place rather earlier elsewhere in Europe. Neither the speed nor the extent of the change should be exaggerated. The numbers of kin amongst whom property was usually divided seems to have been narrowing before the Conquest, with a bias in favour of direct male heirs at the expense of brothers, nephews, sisters, widows and others. After 1066, eldest sons did not in fact get everything. Fathers could still make provision for younger offspring from acquisitions and sometimes from patrimonial lands, and they also continued to provide for widows and daughters. None the less, the Normans, with stronger notions of patrimony and primogeniture than those current in pre-Conquest England, arguably shut the door more firmly on the wider kin, and created a greater expectation that the key properties, including the principal castle, would go to the eldest son. A change in practice is reflected in the absence of toponymic surnames before 1066 and their development thereafter. Likewise there is a remarkable alteration in the nature of wills: those before 1066 went into great detail, allotting land to various members of the kin, while those after 1100 say nothing at all about land and concentrate on distributing the movable property. There was indeed no need to say anything because, unless previous provision had been made, the descent of the patrimony by the rules of primogeniture was simply accepted.

It is difficult to generalize about how these changes in lordship and family structure affected the position of noblewomen, partly because the evidence bearing on the descent of property through women is tenuous, partly because politics, status and life cycle could make female experiences far more diverse than those of men. Anglo-Saxon wills show widows and daughters receiving property, and widows also disposing of it, apparently as they wished. In a case in the Herefordshire county court a woman (presumably widowed) actually announced that she would grant nothing to her son. Yet after the Conquest too women could hold property. The Coronation Charter of Henry I in 1100 laid down that widows were to receive both their dower and ‘marriage portion’. The former was land assigned by the husband on marriage for his wife’s support after his death. (In the law book Glanvill of c. 1189 it was specified as a third of the husband’s estates unless a smaller amount was stipulated.) The latter was land given with the bride by her own family. The Charter also shows that women could inherit land. Indeed in the law of the twelfth and thirteenth centuries there was nothing to prevent a widow alienating her inheritance, so in theory the Herefordshire episode could have been replicated after 1066. If the development of primogeniture after the Conquest meant that daughters were less likely to get a share of the inheritance, they were by the same token more likely to scoop the lot if they had no brothers.

If all this suggests that women, before and after 1066, enjoyed independent power as landholders, the impression may be misleading. In the Herefordshire case, the woman is not named and the whole transaction was probably masterminded by its chief beneficiary, the great thegn Thurkill the White. After the Conquest, when widows alienated land, they usually did do so with the consent of their heirs. In any case few twelfth-century widows remained so for very long, pressure to marry again being intense. In all marriages, in twelfth- and thirteenth-century law the husband had total control over his wife’s property.

The Conquest did, however, bring changes especially when it came to the control exercised by king and lords over marriage. Women before 1066 were certainly forced into marriage, as the charter issued by King Cnut (perhaps at his coronation) shows. But after 1066 the new tenurial rights of lordship gave the king and lords tighter controls in this area, as the greater precision of Henry’s Charter compared with Cnut’s demonstrates. As it was, the attempts at regulation in the former were unavailing. Henry did not keep the promise not to force widows of tenants-in-chief to marry. He did not relinquish the right to find husbands for the heiresses of deceased tenants-in-chief, promising only to arrange such marriages after having taken counsel from his barons. Since such heiresses would be wards in the king’s hands until they married and only then able to inherit, he was well placed to stand his ground. He also had every incentive to do so because the gift of a wealthy wife was by far the easiest way to enrich both his close family and rising ministers. There were also plenty of heiresses about. Fifty-four of the 189 honours in existence in 1166 passed into the female line at least once after 1086. The king’s ability to conjure up heiresses was also enhanced by the relative fluidity of female (and male) inheritance rights. A woman could, for example, be made more eligible at the expense of an unfavoured brother by giving her a large marriage portion. Later the opportunities for manipulation became all the greater when the practice developed (perhaps influenced by a decree of Henry 1 around 1130) that in default of sons the inheritance was to be divided equally amongst all the daughters, instead of the eldest, like an eldest son, getting everything. This both increased the pot of patronage available to the king and allowed him, despite the supposed equality of division, to chop up the respective portions according to the pecking order of the prospective husbands. As for widows, a second husband in the law as defined later in the century would control his wife’s dower from her first marriage, together with her inheritance and marriage portion, as long as she lived. The dower would then pass back to the heir of the first husband, the inheritance and marriage portion to the heir of the wife – one and the same person if the first marriage hadproduced offspring but not otherwise. However, if the second marriage produced a child, the husband could continue to keep the marriage portion and inheritance for the length of his own life, before they passed back to any child of the first marriage. Complex! The interpretation and implementation of such rules gave plenty of room for royal interference. The resulting disputes were the stuff of politics.

* * *

King William thus introduced a new ‘feudal’ aristocracy to England, yet he established it within the framework of the Anglo-Saxon state. Here there were deep continuities before and after 1066, and also significant changes.

For William his coronation, which followed closely the old Anglo-Saxon order of service, had immense significance. From a duke he was now a king and he made sure no one forgot it. Surrounded by his bishops, barons and knights he wore his crown three times each year, at Christmas, Easter and Whitsun, rites based on German imperial practice and probably introduced into England in the 1050s but now made much grander, judging from the awe-struck description in the Anglo-Saxon Chronicle. Likewise William continued the Laudes Regiae, the great ceremonial hymns invoking God’s aid for the king and the whole Christian community: ‘Victory and long life to the most serene William, crowned by God, great and pacific king,’ chanted the choirs. ‘Behold I see God,’ cried one of William’s jesters, like so many jesters getting at a truth. Such rituals were a constant reminder that the king was the Lord’s anointed, ruling with his blessing and protection. They were of immeasurable importance in securing service and stilling revolt.

The basic structures of Anglo-Saxon government, superior in many ways to those in Normandy, William gratefully took over: the counties and hundreds with their courts; the sheriffs; the pervasive royal peace with its specially reserved royal pleas; the geld; the coinage; the chancellor and the sealed writ. The nerve centre of William’s government, the royal household, was similar to the Confessor’s. Here were the chamberlains receiving, storing and spending the money for his day-to-day expenses, here some of the household knights (others might be away trouble-shooting on special missions), here his chancellor heading the clerks of the chapel who sang the daily services and wrote documents, here the kitchen staff providing food and drink, and the grooms looking after the dogs, horses and stables. Around the household the wider court was gathered, with its councillors, some of whom were great barons, lay and ecclesiastical, others men of much lower status; and around the court William regularly assembled the nobility in great councils, similar to the pre-1066 witan, to witness his wearing of the crown and discuss affairs of the realm, for example the making of Domesday Book.

Yet Norman rule was also significantly different from that of the Confessor and his Wessex predecessors. In the first place, after 1072 William was largely an absentee. Of the 170 months remaining of his reign he spent around 130 in France, returning to England on only four occasions. This was no passing phase. Absentee kings continued to spend at best half their time in England until the loss of Normandy in 1204. This was not simply because the continent was home; rather, as William of Malmesbury shrewdly appreciated, Normandy with its long open frontiers was far less secure than England and needed constant minding. In the purely personal sense these kings must, therefore, have been less ‘hands-on’ and interventionist than their Anglo-Saxon predecessors, though perhaps not less than Cnut who was also king of Denmark. But this absenteeism solidified rather than sapped royal government since it engendered structures both to maintain peace and extract money in the king’s absence, money which was above all needed across the Channel. Malmesbury indeed compared England and Normandy to two sisters joined at the waist, the healthy one supporting the other, who is terminally diseased.

By far the most important of William’s regents was Bishop Odo, who had authority to act independently in the king’s name. At other times, Archbishop Lanfranc, or a small group of magnates headed by Lanfranc, received the king’s orders, but William also wrote direct to sheriffs and other local officials when he was abroad. Many such orders were reactions to complaints and petitions. Already the trek across the Channel to the real centre of power had begun. ‘Do this so that I quickly hear the truth of the matter by your letter,’ one writ to Lanfranc concluded. However far away he might be, England was always under that stern exacting oversight.

As the exchange between William and Lanfranc suggests, the exigencies of the cross-Channel state significantly increased the role of written orders in government. William took over Regenbald, the Confessor’s chancellor, and with him the sealed writ, which soon came to be written in Latin rather than English. William issued large numbers of writs in favour of abbeys and bishoprics confirming their rights and properties. Here he was frequently being reactive, the initiative coming from the beneficiary who not surprisingly (as before 1066) often wrote the document. The chancellor’s task was just to check the draft by the beneficiary, supply standard writ formulas and affix the king’s seal to the finished copy. But William was also proactive. Writs were issued to summon armies, and give orders to local officials. Had more of these survived in the original, we would probably detect the hands of a small group of clerks at court writing for the king, although in fact only one has been identified.

One particularly important use of the writ was to command local courts to hear law cases. Although employed for that purpose before 1066, such orders multiplied thereafter as the Conquest produced numerous disputes over the tenure of land. William also sent Lanfranc, Odo of Bayeux and the bishop of Coutances into the county courts to hear such pleas. If all this enhanced the role of the king in local affairs, so did changes in the structures of government. One innovation, derived apparently from Normandy, lay in the way William began to subject parts of England (the New Forest for a start) to a type of royal forest law which was later to be a major source of both income and unpopularity. Another change related to the earls because, as compared with 1065, the number of counties subject to their authority was greatly reduced. William set up earls in Kent and the Welsh borders, who guarded the frontiers very much as the counts did the frontiers of Normandy. But these frontier earls were the only ones William had. There would be no over-mighty officials in his England, any more, in all probability, than there would have been in Harold’s.

This limited use of earldoms meant that William established a direct relationship with the sheriff, who was no longer the earl’s deputy and was more powerful as a result. Even more important was the way the king’s castles, planted in the county towns, became the military and administrative bases for royal government in the shires, housing the sheriff’s office, the mint, and the county court over which the sheriff presided. Secure in the castle and keeping troops there to enforce his will outside it, the sheriff could afford to be far more unpopular, and thus more ruthless in the king’s interests (and his own), than before. ‘Art thou called Urse? Have thou God’s curse,’ Archbishop Ealdred thundered at Urse d’Abetot, sheriff of Worcestershire. The sheriffs were powerful but William was determined to control them. He ordered the regents to summon them together and forbid seizures of property from the church. He moved them frequently from one sheriffdom to another and more often appointed ‘new men’, owing everything to him, than he did great barons; he was doubtless aware of how the equivalent office in Normandy, that of the ‘viscount’, had fallen into the hands of magnates, some of whom had been notably disloyal in the great crisis of 1046–7. William was not going to have a repeat performance in England.

The Anglo-Saxon coinage was far superior to the Norman and William naturally continued it, but he standardized the weight of the coin and increased the annual money payments made by the mints. The symbolic importance of the coinage remained. Circulating everywhere, seen by everyone, it was the most visual demonstration possible of the unity of the realm under the king. Although the evidence is fragmentary, William probably levied the geld annually, usually at 2s. a hide (producing perhaps £2,500) but occasionally at higher rates. The 6s. a hide in 1084 provoked cries of protest from the Anglo-Saxon chronicler. If William never equalled the sums raised by geld early in the century, assuming the figures can be believed, that was because he levied it at lower rates (that in 1018 was possibly 20s.), and exempted the demesne manors of his barons (though not the peasant lands within them) from payment. Yet these early gelds had in effect been harried out of the kingdom by invading armies. The more regular army geld levied between 1012 and 1051 had itself proved unsustainable. Up to a point William may actually have revived a tax which was falling into decay. He certainly needed the money to pay the mercenaries who had helped make and sustain the Conquest, but his noble followers he had rewarded with land. That being the case, it made sense to reduce the tax liability of his barons on their main reward, especially when the exemptions could be reversed, as some of them were after 1087. How right William was to choose land rather than money. His followers, unlike those of Cnut, had come to stay.

Much of this makes kingship seem more powerful and pervasive after the Conquest, especially when one adds in the new ‘feudal’ package of powers. But this is only part of the story. The baronial fees or honours introduced new structures of magnate power, cutting into and across the counties and hundreds of Anglo-Saxon England, destroying the monopoly of their courts and raising the question of whether the allegiance of under-tenants belonged to the baron or the king. Before 1066 the only private court had been that of the manor. Afterwards, while the upheavals of the Conquest doubtless brought business into the courts of shire and hundred, it is difficult to think that they were not weakened by the loss of land pleas to the honourial court, as also of ecclesiastical pleas to the courts of the church. At the centre of honourial power, moreover, was something completely new: the private castle. If the royal castle enhanced the king’s local position, so did the private castle that of the baron. Of the 500 or so castles in England around 1100 perhaps two-thirds were in private hands. The castle was thus not confined to a small elite as it had been in Normandy. It was common to all barons and many major under-tenants. What had begun as an instrument of conquest, continued as an instrument of lordship. In the castle, the lord held his honourial court, feasted his friends and followers, and generally displayed his wealth and status – hence the elaborately decorated entrance to the great stone keep at Castle Rising in Norfolk built by Earl William II d’Albini after 1138. There was nothing in any of this necessarily threatening to the king. William I had probably ordered the building of castles by his followers in order to secure the country. He and his successors depended on strong loyalist barons. But if barons were ever to be disloyal, the Conquest has certainly placed new weapons in their hands with which to resist the crown.

* * *

In 1066 William had wanted to postpone his coronation until his wife, Matilda, then holding the fort back in Normandy, could be crowned with him. In the event he went ahead without her, but Matilda’s eventual coronation in May 1068 was still a magnificent affair. Queens, therefore, were important and, as the ceremony of crowning and anointing showed, they held a formal office. Matilda had her own set of regalia, which may have included such symbols of authority as the orb and sceptre. Certainly Henry I’s queen, Edith Matilda, appears with both on her seal. All this had probably been equally true before 1066. If the framework of queenship altered after the Conquest, its basic structures did not. Matilda’s ceremony in 1068 was similar to those performed for Emma, queen of both Aethelred and Cnut, and Edith, queen of the Confessor. All these queens gained lustre through the growing cult of the Virgin Mary, who was depicted as a mother and increasingly as a crowned queen, the Queen of Heaven. She appears as both in a Winchester prayer book of Emma’s time.

Matilda’s status did not derive solely from her marriage and her office. It also, in common with other queens, came from her own family. As her tombstone proudly proclaimed, she was the daughter of a count of Flanders and the granddaughter of a French king (Robert the Pious). She was of far higher status than William himself, hence his anxiety to be crowned with her in 1066, and the manner in which a form of the Laudes Regiae, introduced for her coronation, applauded king and queen in almost parallel terms. Cnut, trying to establish a new dynasty, had felt the same need of linkage to his wife: the coronation ordo of 1017, after his marriage to Emma, widow of the previous king and daughter of a duke of Normandy, described her as consors imperii, a sharer in his rule.

Anglo-Saxon precedent and practice also meant that as queen Matilda enjoyed her own resources. Some of these probably came, as they certainly did under Edith, from a share in the financial offers made to the king for favours, a payment later called ‘queen’s gold’. More important was land. Before 1066 certain individual estates were traditionally used to provide for the queen, although they were neither drawn on exclusively nor uniformly. Since the queen as widow seems to have kept the lands she held as queen, a new clutch had to be found for her successor. But the fact that queens did have lands which they could expect to keep after the king’s death gave them their own income and a measure of independence. The Conqueror himself allowed Edith to keep her extensive possessions until her death in 1075, and did not then pass them to Matilda. But he found Matilda other lands worth around £500 a year, most of them coming from the estate of the fallen thegn Beorhtric. From these Matilda was able to endow monasteries and, like Emma and Edith, support her own household, rewarding her chamberlains, for example, with grants of land. This household was separate from that of the king when Matilda and William were apart, but merged more or less with his when they were together.

At their coronation queens, unlike kings, took no oath of office. Was there none the less a conception of what they should so? Before and after 1066, they sometimes appear in charge of the royal treasure but perhaps this was more the result of particular crises, or of activity as the king’s deputy, than of any role they were generally allotted. Queens certainly were expected to feature prominently in court ceremonial. Indeed they could mastermind it, at least if we can believe the Life of Edward the Confessor which has Edith, who commissioned the work, both encouraging her unworldly husband to put on royal finery and arraying him in it. Whether Matilda did the same for the Conqueror may be doubted but she was often at court, judging from the large number of royal documents she witnessed, something which itself reflects her status. The coronation ceremony indicated that queens should be fruitful, the prayers linking them with the biblical women, Sarah, Rebecca and Rachel, who produced the line of David. The influence that queens established over their sons (sometimes through controlling their upbringing) could indeed lead to important political roles both as queens and queen-mothers. According to the coronation prayers, the queen was also to imitate the biblical example of Esther and persuade the king to act with mercy. Intercession was not, of course, an exclusively queenly prerogative, but that of the queen did have a special moral force, as well as the unique advantages provided by proximity to the king not merely at court but in the bedchamber. Nor was there really any clear line between pleading from a sense of mercy and pleading from a sense of politics.

All of this, of course, amounted to the queen influencing the men who pulled the levers of power, while not pulling them herself. And the men did not have to listen. When Matilda interceded with William for Robert, her eldest son, she got nowhere. When, testimony both to her spirit and resources, she dispatched money to him despite William’s prohibition, he ordered one of her servants to be blinded. The queen was not in any sense a joint ruler of the kingdom. Indeed, neither the consors imperii of the 1017 coronation nor the queen’s place in the 1068 Laudes were generally adopted. Emma, in the first depiction of an English queen, might stand beneath the Virgin opposite Cnut, but he was crowned, she was merely veiled. There had to be a king. There was no permanent need for a queen. William did not remarry after Matilda’s death in 1083. His successor, William Rufus, did not marry at all. Scotland had no queen between 1130–31 and 1186. Nor was queenly status itself unalterable. It depended on life cycle, as the king’s did not. If her husband’s death did not ‘de-queen’ her, it certainly diluted her status, particularly if the new king married and created another queen.

For all these limitations, the queen was unquestionably in a position of potential influence and, in effect, of power. What she actually achieved depended on a whole range of variables including political circumstance, personal ambition, and the particular relationship with the king. In one way, queens after 1066 did have potentially more scope than before. Cnut had governed several realms, yet there is (perhaps surprisingly) virtually no evidence that he ever made Emma his regent. William was different. He made Matilda regent on several occasions both in Normandy and in England. In England she was sought out by those with grievances and presided over important law cases in the counties. She was the first in a series of formidable post-Conquest queens given opportunities by the exigencies of the cross-Channel state.

* * *

The church was central to royal power both before and after the Conquest, and William was no less determined to control and exploit it than his predecessors had been before 1066. Kings also swore at their coronations to protect the church, which at the very least meant maintaining its properties. It also implied they should be supporters of reform. The struggle for reform envelops the whole period covered by this book and constantly impinges on the course of politics. The reform movement had begun outside the papacy but in the pontificate of Leo IX (1048–54) was taken over by it. The ultimate aim was to ensure that the church worked with enthusiasm and devotion for the spiritual welfare of its flock. Monks were to devote themselves to the round of divine service and pray for the salvation of the faithful. Priests and bishops were, as Pope Gregory VII (1073–85) put it, to be ‘pastors of souls’. For this to be achieved, certain abuses had to be eradicated. One was clerical marriage, which snared priests in the world and threatened to make their offices hereditary. Another was pluralism, the holding of more than one benefice, behind which there was often simony, the buying and selling of ecclesiastical office. If the church was to be rid of these abuses, a necessary condition was clear lines of authority. Bishops needed to be able to rule their dioceses and archbishops (metropolitans) their provinces. Above all, the ultimate authority of the pope in matters of doctrine, law and discipline needed to be recognized in theory and exercised in practice. To that end the great mass of canon law – passages from the Bible, the pronouncements of popes and the decrees of councils – was edited and arranged so as to lay bare the basis of papal power.

For the future of royal power the implications of reform were explosive in England as elsewhere. Was the pope to exercise a direct authority over churchmen? Was the church to be freed from control by the king? To break the royal hold over ecclesiastical appointments, in particular, seemed essential to reformers, given that such appointments implied the king had some kind of spiritual authority, and resulted in totally unsuitable royal clerks gaining high office. How could they be instruments of reform? But such a fissure seemed utterly unreasonable to the king, given the wide lands held by bishops and abbots, for which, apart from anything else, they owed military service. Until the last years of William’s reign, however, there was little to presage any fundamental conflict, thanks in large measure to William’s relationship with the man who replaced the discredited Stigand as archbishop of Canterbury in 1070: Lanfranc.

From Pavia, born around 1010, Lanfranc had studied arts in the Italian Schools, and had gained an easy mastery of its fundamental method: the deployment of evidence in support of argument. He had gone to Normandy to teach and then, around 1042, had entered the impecunious infant monastery at Bec. There, in novel fashion, he used the methodology of the arts to study the Bible, and made his school internationally famous in the process. Having defused an early quarrel with Duke William through a joke (‘I would go into exile more quickly if you gave me a better horse’), in 1063 Lanfranc became abbot of the new ducal monastery of St-Étienne at Caen, a resounding vote of confidence. From there the move to Canterbury was a natural one. Lanfranc was far more than a dry and cloistered academic. He had a good head for business and had masterminded building operations at Bec. He tempered sternness with humility, and had a brisk, humane common sense: ‘Whatever death overtakes the just man his soul will be in peace,’ he declared, quoting scripture. Here was the man who could act as archbishop and also as regent. Lanfranc believed wholeheartedly in reform yet he also thought that ‘the practice of Christianity’ could only be established through good kings. Hence he entreated God to grant William a long life, for William was a good king. Indeed with the archbishop of Rouen, he had presided since 1049 over reforming councils in Normandy.

‘In order to confirm his power in the kingdom he had acquired,’ as the chronicler John of Worcester put it, William replaced the English bishops and abbots. By the end of the reign there were only three important native prelates. Yet in establishing his power, reform of the church was almost as important as the change in personnel, although that was far from the only motive for it. William complained that church laws before 1066 contravened the ‘precepts of the holy canons’, and the English themselves, if William of Malmesbury is at all representative, came to believe him. Thus the Normans could regard the Conquest as a divinely sanctioned mission (the pope after all had given his approval) and the English could accept it as a punishment for their sins. The propaganda had at least some basis in fact. True, Wulfstan of Worcester, the only English bishop remaining in 1087, was a model of erudition, eloquence, practical piety and unostentatious austerities. English bishops attended the great papal reforming council in 1049 and avoided the schismatic Stigand. Yet there had been no reforming synods in England before 1066, and no parallel to Lanfranc’s school at Bec and the duchy’s other vibrant monasteries. The pre-Conquest church under the other-worldly Confessor and the all too worldly Stigand was very different from that after 1066, driven forward by William and Lanfranc.

Not all Lanfranc’s work was successful or well judged. His attempt to assert Canterbury’s primacy over the whole of Britain and more particularly over the archbishopric of York was motivated by the passionate concern with Canterbury’s rights. But it lacked precedent, sucked his successors into a quagmire, and actually weakened rather than strengthened church discipline and the unity of the kingdom. Indeed it became impossible in later centuries, thanks to disputes over status, for the two archbishops to appear in each other’s presence. Lanfranc, however, was sensitive to English conditions. He came to revere some of the Anglo-Saxon saints and accepted the uniquely English institution (found at Canterbury itself) of cathedral clergy organized as monks rather than as chapters of canons. But there was still much to do. Between 1070 and 1076 five councils were held which promulgated statutes for the reform of the church; the first two were presided over by papal legates. A sensibly cautious start was made in eliminating clerical marriage – a more radical approach in Normandy led to the stoning of the archbishop of Rouen. Much attention was given to increasing the authority of the bishop within his diocese and improving its administration. Several cathedrals were moved to more populous centres: Dorchester on Thames to Lincoln (1072), Selsey to Chichester (1075), Sherborne to Salisbury (1078) and Elmham ultimately to Norwich (1094). Dioceses seem to have been formally divided into territorial archdeaconries and then subdivided (sometimes using old Anglo-Saxon minister divisions) into deaneries. Archdeacons may have existed before 1066 but they were now more able, along with the rural deans, to ‘scrutinise the character of [the local clerics] and their competence as priests’, as Lanfranc put it, describing his own activities.

The parish priesthood was the lowest rung of the church, yet the most important. Here significant changes were taking place, both for good and ill. Characteristic of Anglo-Saxon England had been large territorial areas served by priests based in minsters. Occasionally, as at Farnham and Chertsey in Surrey, these survived, in whole or in part, to form later parishes. But for the most part, before and after 1066 such jurisdictions were being undermined as individual lords founded new churches, an activity itself related to the breaking-up of great estates and the formation of manors and nucleated villages. The pace may well have quickened after the Conquest for William of Malmesbury speaks of churches rising in every village. Thus the new Norman lords of manors marked their arrival in the country and God’s sanction for it. By 1086 in Surrey 60 to 70 per cent of later parish churches had already begun their life; the actual formation of the new parish boundaries was largely complete here and elsewhere by 1200, a process as central to the shaping of England as it is hard to trace.

Another way William and his followers thanked God for their victory and indirectly tied kingdom and duchy together was by giving land in England to their family monasteries back home. Around thirty continental houses, most of them in Normandy, received land in England. Fécamp’s endowment was worth £200 a year. But the Normans also thanked God and proclaimed their arrival by founding monasteries in England, around thirty-four of them by 1087, just over half being daughters of continental houses. Battle, William’s own foundation, was placed symbolically on the very site of his victory. Other houses like the Warenne foundation at Lewes in Sussex, a daughter house of Cluny, marched side by side with the baronial castle. The Normans also began to endow pre-1066 English houses, indeed by the end of the century such grants were on a far greater scale than those made to the continental monasteries. The major English houses, of course, acquired continental abbots. Some of the latter, like Thurstan at Glastonbury, quarrelled violently with their English monks, but all of them (and the same was true of the Norman bishops) had one decisive advantage over their English predecessors: they were far more successful, as the contrasting fortunes of Abingdon under English and Norman abbots shows, in recovering and retaining property, hence the church’s success in broadly maintaining its estates over the Conquest period. In general, William of Malmesbury was right to think that monastic life flourished under the Normans. Lanfranc himself drew up a series of constitutions to govern the life of the monks at Christ Church, Canterbury and these were adopted by at least twelve other monasteries. Many of the new abbots were men of exemplary sense and piety, and they attracted recruits. The monks at Gloucester rose from ten to 100 under Serlo, abbot from 1072 to 1104. The Normans were also great builders. The massive, cold stone columns of their new abbeys and cathedrals seem almost with the aid of God and man to be treading down Norman rule into English soil.

Lanfranc’s reforms depended absolutely on the support of the king, and none more so than one of the most important of these. Churchmen before 1066 had been aware that there was a category of offences committed by the laity but related to ‘the rule of souls’ (blasphemy and adultery, for example) which should come within the jurisdiction of the church; so should disputes over wills and burials. But the bishop or his deputy had none the less heard such cases in secular tribunals, usually in the hundred court. To Lanfranc this seemed a scandalous breach of canon law, and William agreed. ‘I order and by my royal authority command’ ran the consequent royal ordinance, which banned the practice, laying down that henceforth such cases were to be heard in a place decided by the bishop. Indeed, if necessary they were to be brought there ‘by the force and justice of the king and the sheriff’. There could be no clearer indication of William’s support for the church, even though the competence of his own courts was diminished. Although in practice there was no sudden or clear-cut break with the past, the ordinance facilitated the development of separate ecclesiastical courts, over which, for most routine cases, the archdeacon and rural dean presided.

Under Lanfranc and William, therefore, the English church, as archbishop Anselm later put it, was like a plough drawn by two well-matched oxen. If the pope was formally the driver, he followed where they led. The pope had been useful in getting rid of some English prelates and sanctioning through his legate some early reforms. William and Lanfranc were meticulous in showing proper respect, but essentially it was ‘we’ll call you, don’t call us’. The relationship changed with Gregory VII (1073–85) who was determined to make the papal headship of the church a reality. As his quarrel with the Emperor Henry IV deepened and the latter set up his own anti-pope, Clement III, so William’s attitude became increasingly ambivalent. Gregory complained that William was preventing Lanfranc and other bishops visiting the Holy See. But Lanfranc’s loyalties were with William. He explained that while England (‘our island’) had not abandoned Gregory for Clement it might decide to do so. After Gregory’s death in 1085 William and Lanfranc hesitated to recognize his successor. This, however, was only the beginning of the story. In the long term William and Lanfranc’s pick-and-choose attitude to the papacy was to prove unsustainable.

* * *

After the suppression of Edwin and Morcar in 1071, English attacks on individual Normans continued, something which William sought to counter by introducing the ‘murder’ fine. It had to be paid by the hundred or the village in which a murder took place if it could not apprehend the murderer or prove the victim was English. But any general resistance was at an end. The rebellion of 1075 illustrated as much. It was joined rather weakly by Waltheof, the last surviving English earl, who had followed Gospatric in Northumberland and also apparently presided over Northamptonshire and Huntingdonshire. The leaders, however, were two young continental nobles, one at least irritated by the way the king’s sheriffs were challenging his local power. Since equivalent tensions had sometimes surfaced before 1066, normal politics had been resumed. The revolt was quickly put down. Waltheof was executed, Earl Ralph of Norfolk fled to his estates in Brittany and Earl Roger of Hereford (son of the ultra-loyalist William fitz Osbern, as Lanfranc never ceased to remind him) was imprisoned for life. There he flung the fine robes sent him by the Conqueror on the fire, a furious and futile gesture which sums up the rebellion. The Conqueror had demonstrated his mastery and he was to do so again even more dramatically in 1082. Faced by the vaulting ambition of his half-brother, Bishop Odo, for whom Kent, the regency and the Bayeux Tapestry seemed not enough, William hurried across the Channel and had him tried and imprisoned.

William juggled his problems in England with those across the Channel where, as we have seen, he spent most of his time. In Normandy the brief respite which had permitted the Conquest had ended and he was on the defensive. He faced a count of Flanders, Robert the Frisian, who was a sworn enemy, a count of Anjou, Fulk Rechin, who aimed to recover Maine, and a king of France, Philip I (1060–1108), who had gained hold of the French Vexin in 1077 and so was able to prowl along the Norman frontier. All of them were ready to exploit William’s quarrels with his son Robert who, approaching twenty-five in 1077, coveted a good deal more power than his sceptical father would accord him. There was also one final challenge to William’s rule in England: an invasion planned by King Cnut of Denmark in alliance with the king of Norway and the count of Flanders. William levied a heavy geld and then in 1085 came over with a large paid army. But it was not needed. The Danes quarrelled among themselves and eventually in July 1086 Cnut was murdered. For England this was a decisive moment. The kings of Denmark had made their last bid for the throne.

‘Having found out for a fact… that his enemies could not carry out their expedition,’ as the Anglo-Saxon chronicler put it, William decided to survey his winnings. At his Christmas court of 1085 he ‘had much thought and very deep discussion with his council about this country – how it was occupied or with what sort of people’. The result was the great survey of England embodied in the two volumes of Domesday Book. (The name, signifying the final and definitive nature of the testimony, was in use by 1179.) Domesday mentions 13,418 places and contains 2 million words. ‘Not one ox, cow or pig was left out,’ grumbled the Anglo-Saxon chronicler. To carry out the survey England was divided into at least seven circuits each with its own commissioners. They had some existing material to help them, for example lists of geld liabilities and of dues from royal manors, but the great bulk of the information probably derived from written returns about their properties presented by the tenants-in-chief. This material was then co-ordinated with the evidence presented by local juries at sessions of the hundred courts. The survey’s rapid completion, quite probably by August 1086, testified to the strength of pre-1066 governmental structures and William’s ability to exploit them.

Studies of Domesday are full of friendly controversy. One ingenious suggestion (by David Roffe) is that William commissioned the survey but not the book, the latter being the brainchild in the 1090s of William Rufus’s chief minister, Ranulf Flambard. But a Worcester chronicler of the early twelfth century states that the Conqueror ordered that everything be written in a book, and the returns would certainly have been useless unless edited. As for the purpose of the whole exercise, one view is that William wanted information to enable a reassessment of the geld. But there is no indication that the geld was generally reassessed, nor does Domesday Book itself highlight the information which would have been relevant. If the Conqueror wished to increase his revenue from the geld, it could be done much more simply by altering the rate at which it was levied, and reducing the number of exemptions. The form of the book suggests the principal aim was rather different: the king wanted information about his own properties and those of his tenants. Domesday arranges its information county by county. At the start of each section there is a survey of the county town and a statement of the customs of the shire. Then follow the estates of the landholders one after another, listed hundred by hundred and then manor by manor. The first landholder is always the king. The survey thus provided William with detailed information about his own lands, including assessments of their potential value. His harshness as a landlord is referred to explicitly by the Anglo-Saxon chronicler and he was now ideally placed to demand more from his reeves and lessees. Any efficient landholder taking over a new estate would have it surveyed; for William, Domesday Book was such a survey on a gigantic scale. William was equally determined to know about the estates of his tenants. The list of their names at the start of each county section made it easy to find the page where the entry for each began. In 1085 William had parked out his army, brought to meet the Danish threat, on the lands of his barons, hence perhaps his immediate desire for more information about their possessions. But the broader background was the gigantic turnover in landholding which had taken place since the Conquest. For all William’s attempts to control it through written orders and special officials, some barons had just helped themselves, and there was no record of the final results. Domesday provided just that. Now, when William wanted to seize estates after a tenant’s forfeiture or death, he knew what to take. If he kept those estates in his own hands during an ecclesiastical vacancy, or through wardship or escheat, then he could exploit them as effectively as any land of his own. Domesday Book was therefore very much about exploiting the king’s feudal rights and revenues.

Why then did the magnates themselves co-operate in the making of the survey? The answer is that Domesday Book provided them with something approaching a written title to their lands, if not a definitive one. This was why it recorded the name of the Anglo-Saxon holder of the land in 1066. That was completely irrelevant to the exploitation of the estate, but it was very relevant indeed if (as was so often the case) land was claimed by a Norman lord as the successor of an English ‘ancestor’. Of course what precisely an ‘ancestor’ had held often gave rise to disputes, and the Domesday commissioners were far too busy to determine all of them. Nevertheless the whole process provided an opportunity to ventilate claims which the king issued writs ordering local courts to settle.

In the generation after its construction, and probably for longer, Domesday was central to the exploitation of the king’s lands and his feudal rights and revenues. Almost at once, moreover, William responded to something else the survey had been designed to reveal: the names of the under-tenants enfeoffed by the tenants-in-chief. Would these men be loyal to the king or simply to their overlords? William, with characteristic precision, provided an answer. He could not demand homage from the under-tenants because they did not hold land from him. But he could demand an oath of fealty. In August 1086 he summoned to Salisbury ‘all the landholding men of any account that were over all England whosoever men they were’ (my italics) and made them swear just such an oath. Domesday Book thus revealed the dual polity which had emerged out of the unitary Anglo-Saxon state. On the one hand there were the pre-1066 counties and hundreds giving the king a direct relationship with all his subjects, particularly through maintenance of his peace. On the other, within this old framework, were the new structures of feudalism.

After the Oath of Salisbury, William returned to Normandy and within a year lay buried in his great abbey at Caen. ‘I was brought up in arms from childhood,’ he groaned on his deathbed, according to Orderic Vitalis. This martial, stern, demanding, jovial, pious, intelligent and farseeing man had transformed the face of Europe.