William the Conqueror is probably the best known soldier and general of the eleventh century. The conquest of England in 1066 was not only a major historical event, it was also one which has stuck in the minds of at least the English-speaking world. William was a minor when his father died in 1035, and the struggle to impose himself upon Normandy was long and bitter. It was only with the help of his overlord, Henry I of France (1031–60) that the greatest rebellion against him was defeated at the battle of Val-ès-Dunes in 1047 of which we know almost nothing. However, the rebel leader, Guy of Burgundy, took refuge in the castle of Brionne where he held out for three years. Thereafter, although William’s position improved, the propensity for rebellion remained. In the wake of his capture of Tours in 1044 Geoffrey Martel, count of Anjou (1040–60), turned his attention to Maine, where the major city of Le Mans was captured in 1051. After the count of Maine’s widow, her son Herbert and daughter Margaret had fled to the Norman court, Geoffrey seized both Domfront, a fief held of the count of Maine by the Bellême family, and the Norman town of Alençon, offering as an inducement to their soldiers a licence to ravage in the Norman lands. William failed to take Domfront by coup de main and built four castles, probably earthwork and wood structures, to blockade it while maintaining an active posture which enabled him to rally his troops against an effort to relieve it by Geoffrey, whose forces retired intact and watchful. William now faced a difficult situation for their presence prevented him from ravaging. However, William had apparently kept a close eye on Alençon in the meantime, and, when he realised that its defences were weak, suddenly seized it, dealing so harshly with its garrison that Domfront decided to come to terms. The campaign certainly illustrates William’s generalship, with its tight control over events. It indicates how the castle and its supply dominated war yet not at the expense of mobility which was the key factor in William’s victory. It should also be added that Geoffrey was a good general, but here he was at the very edge of his authority, so his power was attenuated and his ability to bring it to bear without enormous effort limited. William’s own stabs against Maine failed for much the same reasons, until after Geoffrey’s death in 1063 when, taking advantage of the internal conflict then rending the house of Anjou, he advanced against Le Mans with fire and sword as described by William of Poitiers.
In the years 1051–2 there occurred a major shift in alliances in northern France. The Norman dukes had long been close allies of the Capetian royal house. William’s father, Robert I, had sustained Henry against the revolt of 1031 and in return the king had supported his son as we have seen. But the Capetians had also long been friendly with the house of Anjou, who had been their allies against the grave threat posed by the counts of Blois-Champagne, most recently accepting their conquest of Tours in 1044 from the Blésois.6 When these two allies quarrelled over Maine, King Henry supported the Angevins, posing a grave threat to William whose régime was still far from secure after his recent minority. In 1053 William of Arques, a great lord of upper Normandy with many allies, rebelled and his castle of Arques, newly built and well-fortified, was the focus of events. William’s men at Rouen, his principes militiae, tried unsuccessfully to interfere with the preparation of Arques, but when William arrived he built a counter-castle and settled down to a siege. King Henry led an army into Normandy, ravaging as he went, but was ambushed and, although he got supplies into Arques, his force was so weakened that the castle fell soon after his withdrawal. In the following year Henry tried again with two armies, one under Odo, his brother, striking into Eastern Normandy and the other under his own command, supported by the Angevins, advancing via Evreux. The duke adopted the classic tactic of shadowing his enemy, and one of his detachments fell upon French ravagers at Mortemer causing such loss that both French armies withdrew. The same tactics of shadowing the French, preventing them from spreading out to forage, were employed in 1057 and this time William fell upon the French and Angevin army as the tide cut it in two crossing the Dives at Varaville, causing very heavy losses. It was at this battle that, according to Wace, archers played a notable role. There is much to admire in William’s generalship in all these campaigns. He was a master of the contemporary techniques of war and succeeded in impressing his vassals and preserving their loyalty. Perhaps even more important is to notice the scale of effort which he managed to sustain despite his internal difficulties. He, and indeed his opponents, mounted major campaigns interspersed with sieges and lesser affairs over a period of very nearly ten years. This obviously says a great deal about the economic efficiency of the manorial economy, but it also says a great deal about the ability to organise, recruit and sustain armies. It is a theme not much discussed by modern historians of the period, but it was of course a vital skill in the circumstances of the crusade.
Even William’s admiring biographer, William of Poitiers, admits that he evaded battle whenever possible. Indeed, Varaville was the only occasion before Hastings when he engaged on any scale in the open field and it was then only in the most favourable circumstances. The qualification ‘on any scale’ is important, for there were many occasions during these years when there were fights, but they were of a limited kind which could only have limited results. In 1053 and 1054 King Henry simply absorbed minor defeats. William’s was not a technique without battles – rather he committed himself to a style of war which avoided heavy losses and conserved his forces, preferring the tactics we have noted above. In this he showed wisdom, for battle on any scale could be very expensive and was terribly hazardous. The battle of Cassel on 22 February 1071 was fairly widely noted by contemporaries. In 1070 Baldwin VI of Flanders died and the succession of his fifteen year old son, Arnulf III, who was supported by his mother Richilde, was contested by the dead count’s brother Robert I the Frisian, father of Robert II of Flanders who went on the First Crusade. Robert rallied support especially in northern Flanders and struck suddenly at Cassel where Arnulf’s army was concentrated; in its ranks was Eustace II count of Boulogne, a major vassal in Flanders and in England and father of three participants in the First Crusade, Eustace III of Boulogne, Godfrey de Bouillon and Baldwin. Arnulf was supported by his overlord Philip of France, whose aunt Adela had married Baldwin V of Flanders (1035–67), amongst whose forces was a contingent of ten knights from Normandy led by William FitzOsborn, a small force whose size indicates that the Conqueror, who had married Baldwin V’s daughter Mathilda, was very much more concerned with affairs in England. Robert seems to have advanced quickly toward Cassel, evidently seeing battle as offering a quick decision and needing to force it before the superior strength of his enemies could gather. We do not know for certain who held Cassel at the start of the battle, the details of which are largely lost to us. One source suggests that Robert lured the allies into an ambush by a feint, but beyond this there is confusion. What interests us is the extraordinary outcome of this battle. Arnulf III was killed and so was William FitzOsborn; Richilde was captured by Robert’s men, and Robert the Frisian was captured by Eustace II of Boulogne. Within a month the king of France had concentrated a much larger force at Montreuil and was ready to resume the war, but he was forced to recognise Robert who was freed in exchange for Richilde and was elevated to the county through the support of Eustace II. Baldwin of Hainault, the other surviving son of Baldwin VI, later unsuccessfully contested the county of Flanders, but was to die on crusade with Robert’s son, Robert II, in 1098. Robert the Frisian had had little option but to seek battle, for most of his support was in the poorer part of Flanders and his rival had powerful allies. The immediate outcome of his strategy was poor reward for his bravery, although in the long run the death of Arnulf opened the way for a favourable political solution. Over a century later the risks were just as great. In September 1198 Richard I of England (1189–99) fell upon the army of King Philip of France (1180–1223) as it tried to relieve Courcelles, inflicting a severe defeat during which the bridge at Gisors broke throwing the French king into the water where he ‘had to drink of the river’. Richard reported these events in a letter to the bishop of Durham which has a confessional, almost apologetic note, reflecting the hazards of resorting to battle: ‘In doing this we risked not only our own life but the kingdom itself, against the advice of all our councillors’. Such sober reflection from one of the greatest of all medieval generals explains why major battle was only to be undertaken in the most favourable circumstances, as William showed at Varaville, or for the highest stakes, as in the Hastings campaign.
Because of its spectacular and decisive results, Hastings is perhaps the most celebrated of all medieval battles. Certain aspects of the Hastings campaign need to be emphasised, however, because they illuminate the nature of war in the late eleventh century. In the first place the scale of the undertaking, requiring the collection and construction of a fleet, was enormous. The devoted biographer of William tells us that when his hero announced his intention of conquering England as news came through of the death of Edward and the usurpation of Harold, many advised him that such an undertaking was beyond the strength of the Normans and some seem to have refused to take part or promised, then reneged. Indeed it was a huge undertaking. William was obliged to consult with his magnates in a series of conferences at Lillebonne, Bonneville-sur-Touques and Caen at which they agreed to unprecedentedly heavy contributions to the army and, also apparently, to the provision of ships such as the sixty raised by William FitzOsborn. It seems likely that William established the number of troops which each lord owed him according to the extent of his lands, and then concluded agreements over and above such figures for the special circumstances of the great expedition. According to Wace, William FitzOsborn exhorted them to provide at least double their obligations and this caused anxiety amongst the magnates lest the increased contribution be seen as a precedent, leading the duke to assure them individually that this would not be so. Indeed, in one sense the critics of the expedition were proved correct, for William had to seek resources outside Normandy. The presence of Flemish, French and Breton troops in the host at Hastings, and afterwards amongst the new aristocracy of England, is too well known to need discussion here. The importance of Eustace II of Boulogne in the Bayeux tapestry testifies to this, and we know of the presence of soldiers from Poitiers. The Carmen de Hastingae Proelio suggests the presence of South Italian Normans. This indicates the range of his recruiting effort. Wace gives some hint of the diversity of the Conqueror’s arrangements when he speaks of soldiers coming to him in groups and singly. ‘Many wished for the duke’s lands should he conquer England. Some requested pay and allowances and gifts. Often it was necessary to distribute these, to those who could not afford to wait.’ Overall some 14,000 men including sailors were mobilised, of whom something like 8,000 were effectives, including 3,000 cavalry. Amongst the 5,000 foot were a lot of archers who appear, from the Tapestry, to have been lightly armed, and a sizable corps of what William of Poitiers calls pedites loricati, heavily armed footsoldiers. In the battle the duke would find it convenient to divide his force into divisions of Normans, Bretons and French. This vast assemblage must have stripped Normandy of troops, but such exposure was possible because two inveterate enemies had died in 1060, Henry I of France and Geoffrey Martel of Anjou. The regency of France was in the hands of William’s father-in-law Baldwin V of Flanders. This huge force had to be concentrated near Dives-sur-Mer where the fleet gathered in the summer of 1066, and it had to be supplied, for William of Poitiers tells us that William would not allow the troops to plunder and so arose what he describes as an extraordinary situation: despite the presence of squadrons of knights, farmers could get on with their business and travellers come and go without fear, an interesting comment on contemporary chivalry!
This concentration of forces at Dives of some 14,000 men and 2,000–3,000 warhorses presented a formidable problem of supply. The task of feeding and watering them, it has been suggested, demanded 9,000 cartloads of grain, straw, wine and firewood along with eight tons of iron for horseshoes alone. They generated 700,000 gallons of urine and five million pounds of horseshit during their stay and this had to be removed. In addition there must have been many draught animals and indeed the Bayeux tapestry shows us military supplies being moved on specialised vehicles. Warhorses were very valuable and supporting sizable numbers of them was a grave problem. Recent research indicates that the breeding of specialised strains of horses was a great burden, requiring enclosed parks to isolate mares and suitable stallions in well-found studfarms. In addition, it must be recognised that in western Europe there were few ranges where horses could graze and that these animals were stall-fed with grain and hay. They thus competed with men for grain while for the provision of hay, meadows needed to be developed. This explains the contrast between the west where the development of bigger and heavier animals was a necessary consequence of this costly regime, and the east where the availability of ranges in Asia Minor and the Euphrates plain, as in North Africa, fostered the development of a lighter breed, though the progress of this distinction was limited in the eleventh Century. Supporting such animals was a major drain on the peasant surplus at the best of times. In conditions of war, feeding horses presented terrible problems. In August and September of 1914 von Kluck’s First Army, which marched on the right of the German attack under the famous ‘Schlieffen Plan’, had 84,000 horses consuming two million pounds of fodder per day, or twenty-four pounds of grain and hay each. Although they were advancing in a most favourable season the cavalry were tired by the time they crossed the French frontier and in poor condition by the start of the Battle of the Marne on 6 September. The lot of the draught animals was worse and the guns were badly delayed. Difficult as the conditions of 1914 must have been, they were infinitely better for the survival of animals than in the eleventh century. William’s concentration at Dives took place at a most favourable time of year and his subsequent deployment enjoyed good fortune. However, the crusaders faced much more difficult conditions and the state of the horses rapidly became a major preoccupation for the army, as we shall see. Once into the Anatolian steppe, animals were very vulnerable, and it seems unlikely that any western Europeans animals survived the journey.
Contemporaries were deeply impressed by the fleet which William gathered, and which is so graphically illustrated in the Tapestry. Its actual size was not definitely known to contemporaries. The ship list of William the Conqueror suggests that the Norman lords should have produced some 776 ships, and Wace recollects being told that the fleet which sailed numbered 700 less four, though he had also found the figure of 3,000 written down. It is not necessarily the case that the Norman lords produced their quotas and figures as low as 400–500 have been suggested, but most writers believe that a total of between 700 and 1,000 concentrated at Dives where the army was gathering. William of Poitiers tells us that the duke ordered that ships be constructed, but it is unlikely that a huge fleet could have been built in the period between the death of the Confessor and the landing at Pevensey on 28 September. The evidence suggests that the duke acquired existing ships, in particular hiring them along with mercenaries from Flanders. The greater number of them were merchantmen suitable for the transport of horses and supplies as well as men, though a number of longships and skiffs were undoubtedly included. The emphasis on shipbuilding in William of Poitiers and the Tapestry probably owes much to the excitement generated by this activity. But evidently William was pressed to find enough ships, for the Tapestry appears to show unseasoned wood being cut for shipbuilding. It seems unlikely that William had special transports made for his horses, such as those used by the Byzantines, for the Tapestry does not show anything resembling them and the written sources do not give any indication of such exotic vessels. By early September the concentration of forces at Dives seems to have been complete and the fleet sailed on a westerly wind for St Valéry where it waited fifteen days until a gentle southerly took it to England. By any standards this was a remarkable logistical and organisational achievement. It is important to recognise that while exceptional, it was not unique.
King Harold of England knew of the intentions and preparations of the duke of Normandy; indeed William of Poitiers records the reception given to an English spy. By May Harold set in train his own preparations, hastened by the raids of his dissident brother Tosti on southern England. His fleet was apparently slow to mobilise, but he may well have attempted a spoiling attack on William’s forces across the Channel, while on land his troops stood ‘everywhere along by the sea’ for the English had an efficient military system. This Anglo-Saxon fyrd was centred on the retainers of the king and the great thegns and perhaps some mercenaries, supplemented by shire levies whose localities provided them with support. The peculiarity of the Anglo-Saxon military tradition was the failure to develop any effective cavalry. Although the élite of the army rode to battle there is every evidence that they fought on foot. Thus, although they could move quickly across country, they lacked battlefield mobility, the key factor in the coming war. Then on 8 September the Anglo-Saxon fleet and army broke up, the former going to London with losses, because, as the Chronicle tells us: ‘the provisions of the people were gone’. It is easy to contrast this logistic disaster unfavourably with the triumph across the Channel. However, to maintain an army and a fleet as long as this was a major achievement, especially as considerable forces stayed in the north to guard against the threat of attack from Tosti and Harald Hardrada. Moreover, when Harold heard of the Norse attack on York, he was able to gather his army and strike very quickly, which suggests that not all had dispersed. Probably the extent of his demobilisation has been exaggerated and the best troops remained with him. Furthermore, the English fleet took to the sea quickly to cut off the Normans after they landed on 28 September. On 12 September the Norman fleet left its concentration area in and around Dives and sailed east to St Valéry, just as Harold heard of the landing of Harald Hardrada at York with a fleet of 300–500 ships reinforced by Tosti; they defeated earls Edwin and Morcar at Fulford Bridge on 20 September with a great slaughter on both sides and took possession of York. By 24 September Harold, after a whirlwind march, was at Tadcaster. On 25 September he marched his troops through York and surprised and slaughtered the Danish army at Stamford Bridge. Hearing of the Norman landing at Pevensey of 28 September, he turned his army south and after spending 5–11 October raising more troops in London, marched out to confront William whose spies warned him of the coming of the Anglo-Saxon army on 13 October. The next day the battle took place and Harold was killed. The organisational effort made by both sides in this summer of 1066 was remarkable and it points to the abilities of commanders. It was paralleled elsewhere in Europe at this time. The Norman conquest of South Italy and Sicily reached its climax in the years 1071 and 1072 when the major cities of Bari and Palermo fell. Bari was the last major bastion of Byzantine power in Italy and its powerful fortifications were deservedly feared. When Robert Guiscard began the siege on 5 August 1068 he knew he was starting a major undertaking and that blockade by sea was vital. In 1060–1 the Normans had demonstrated their willingness to take to ships with a series of raids on Messina which culminated in its seizure by a force which included 700–1,000 cavalry whose mounts had to be ferried across to Sicily. This successful lodgement opened the way for a conquest made easier by divisions amongst the three major Muslim Emirates. Bari was a much greater operation in the course of which a land blockade was established and complemented with a sea blockade, during which the Norman ships were linked together to form a barrier to penetration into the port. A Byzantine relief force did break in, however, in 1069, and a sea and land diversion against Brindisi was heavily defeated. However, the Normans enjoyed aid from Pisa whose fleet brought troops and crossbowmen for land as well as sea operations. The defeat of a major Byzantine fleet in 1070 opened the way for negotiations which culminated in a negotiated surrender of the city in April 1071. This long operation was then followed by the siege of Palermo begun in August 1071, to which the Hauteville brothers, Robert and Roger, brought a force of fifty-eight vessels. On land they built siege machines and on sea a blockade was established which was not totally successful for a North African fleet broke through to provision the city. However, in the end hunger brought the city to a negotiated surrender on 10 January 1072.
These remarkable operations in the south were paralleled as feats of organisation by the German expeditions to Italy. Documentation on the military organisation of the German kings is sparse, but the Indiculus Loricatorum is a list of the reinforcements called for by Otto II (973–83) after his defeat in 982 at Cortone. A total of 2,090 mounted men were called to service on the basis of what appears to have been established servitia debita which formed the recruiting base of the imperial army. On the marches of Germany a regular levy, the census, was imposed upon the Slavs in order to maintain the garrisons and military forces of their conquerors. In 1026 Conrad II (1024–39) undertook the expedition to Italy which led to his imperial coronation. It is not generally seen as a major military action but Italy was unfriendly. After the crowning in Milan, Conrad ravaged the lands of hostile Pavia, though he was unable to take the city. He had to put down a revolt in Ravenna before proceeding to Rome. The imperial coronation was brilliant, but afterwards a German and a Roman quarrelled over a hide and severe fighting broke out involving the entire German army. The ‘Investiture Conflict’ was a German civil war involving bloody battles in a land where the castle was emerging as an important factor. During its course Henry IV led several major expeditions to Italy including the siege of Rome in 1083 in which Godfrey participated when siege machinery, including rams, was constructed. The regularity and scale of the Italian expeditions of the German emperors made a profound impact on the emergence of the German knightly class, the ministeriales. In the twelfth century the codes which governed their conduct were elaborated, particularly with regard to their duties on the ‘complicated and onerous imperial ventures into Italy’, with both heavy fines for failure to comply and fitting out allowances payable from their lord. In 1154 the archbishop of Cologne required that all holding land worth five marks should go, and they were given ten marks for equipment together with supplies, horses and pay of one mark per month once over the Alps. In 1161 the archbishop sent 500 men at a cost of 10,000 marks.
The organisation of war was the primary concern of government, but even at its best it remained, by our standards, simple. In essence those who held land of the king owed service in one way or another and this obligation co-existed with an older Germanic tradition that all free men had a duty to serve the king in moments of emergency. We have noted the establishment of quotas in Germany and the same process was at work in Normandy, although it should be stressed that ‘feudalism’ was emergent in the late eleventh century and that as yet there was only ‘a tangle of incipient feudal customs, partly built up from below’.39 In any case, powerful rulers had sources other than nascent feudal obligation for the raising of great armies. It is now clear that paid troops had always played a major role, as they did, for example, under William Rufus. The distinctions between mercenary, endowed knight and household knight are not clear – those serving from obligation beyond some fixed period might well be paid, and there was a strong tendency to argue about how far obligations went. The aristocracy and the knightly class certainly provided a large pool of skilled manpower trained in war from which soldiers could be recruited. Moreover, it was upon the royal household, their wealth and their leading followers, that the Norman kings relied to raise armies. These professional groupings of household followers around the king – paid and aspirant, or endowed and paid and hoping for better – were what the king relied on for the core of his army and its command. In time of war such a body could expand and serve as the command force of a great army. Through them the sinews of war were channelled, for in the end it was money which made victory. Although such bodies, such military households, can only be documented from the early twelfth century, it is unlikely that they were invented – rather they must have evolved over a period of time. In 1101 Henry I negotiated an arrangement with Robert II of Flanders whereby the latter swore to be his man and to provide 1,000 knights in return for a fee. William Rufus almost certainly made the same arrangement when he met Robert in 1093. It is interesting that the treaty specified that each knight was to be provided with three horses. It seems likely that this kind of organisation was the secret of Rufus’s reputation for raising and paying armies. A medieval army was a composite of forces around a core of loyal leaders whom we can regard as generals. They were not merely military men; they also formed an administrative corps for the vital task of handling and paying out money. Clearly both William Rufus and Henry I needed such a body if they were prepared to take on large Flemish forces. Of course we cannot describe such organisation with any certainty outside the Anglo-Norman sphere, and clearly for Suger such capacity was a matter of wonder. What is of interest is that such capacity had already come into being amongst the Normans on the eve of the First Crusade; they were a major element in the army of conquest which Urban II called into being in 1095. This organisational development indicates the degree to which war in the late eleventh century was not a matter of instinct, of ‘kick and rush’, but of guile and organisation, in short of generalship. This explains the rarity not of battle but of battle on a large scale. They understood the context in which they were making war. To attack your enemy’s economic base, isolate his castles, starve his population, these were surer methods and more applicable to the usually limited objectives for which men fought. However, there were occasions when the stakes were so high that all had to be risked on the throw of battle, and on these occasions the men who directed things sought to ensure that their chances of victory were as great as possible in what was the most risky of all undertakings.