by John France
One such conflict pitted Ascelin Goël against William of Breteuil. It began, says Ordericus, when Ascelin’s brother William offended against a lady and was adjudged guilty by William as overlord. Ascelin took his vengeance in 1089 by betraying the important castle of Ivry, which he held for William, to Robert Curthose, who extracted a large sum for its return. Both families prepared for war, the Goëls in 1091 enlisting Richard of Montfort and some of the household troops of Phillip I of France, with whose aid they defeated William in a skirmish and captured him. This phase of the affair ended with a peace by which William of Breteuil restored the castellancy of Ivry to Ascelin, gave him his daughter in marriage and made various payments. The peace, says Ordericus, was a relief to all, but it was short lived. William of Breteuil seized the abbey of St Mary as a fortress against the Goëls, but was ejected when they attacked and burnt it. However, the considerable resources of the house of Breteuil were greatly increased ‘with the help of ransoms of captives and plunder taken from the country people’. Thus enriched, William persuaded Gurthose and Phillip I of France to help him, and in 1092 together they attacked Brévol castle. It was strong, but Robert of Bellême also joined them for he hated the Goëls. His expertise in siege equipment forced Ascelin to a peace which restored Ivry to the Breteuil. However, Robert of Bellême evidently was left out of the peace negotiations and as the siege force broke up he attacked and nearly captured, the castle at St Céneri held by another ally of Curthose, Robert Giroie. The Conqueror’s youngest son Henry who had taken Domfront from Robert of Bellême now decided to join the struggle against him, during which St Céneri fell to Bellême.47
This convoluted affair, involving some of the greatest men of Normandy and even the king of France, was merely one of many such episodes in the years of Curthose which were unusual in a duchy hitherto strongly ruled. The war was fought for a limited objective – really it was about possession of Ivry. There was no wish to destroy the enemy à la Clausewitz, though poor Robert Giroie suffered badly enough, and so it was punctuated by peaces. There was one major skirmish and the fact that at the taking of St Mary’s abbey Ordericus mentions the death of a single knight suggests its trivial scale. It was not a war of sieges, although there was one major siege at Brévol, rather it was a war which turned on possession of strongpoints, a war of position. The most significant fact about it we learn only in passing as Ordericus comments on the sufferings of ‘the country people’, for in such a war the knights deliberately attacked and ravaged the lands of their enemies. In an age of poor logistics and major problems over the preservation of food, ravaging was a military necessity to keep an army in the field. This had always been true. Vegetius commented that ‘The main and principal point in war is to secure plenty of provisions and to destroy the enemy by famine’.48 William of Poitiers described the manner in which his hero, William the Conqueror, made war: ‘He sowed terror in the land by his frequent and lengthy invasions; he devastated vineyards, fields and estates; he seized neighbouring strongpoints and where advisable put garrisons in them; in short he incessantly inflicted innumerable calamities upon the land’.49 Ravaging was inevitable in a world where all activities were so directly linked to the peasant surplus. War, though restricted in its extent and usually in its ambitions, often involved destroying an opponent’s economic base and this meant appalling suffering for the people who lived on his land. The savagery of the crusade must be seen in this context. But of course it was also the consequence of castles. Assault was expensive and difficult. It took the combined efforts of the king of France, the duke of Normandy and the Breteuil family to take Brévol. Ravaging undermined the economic base and the morale of the enemy. When Bouchard of Montmorency defied a judgement by King Philip in favour of the abbey of St-Denis, Prince Louis initially ravaged his lands and strongpoints, and only later attacked his castle. In the end, losses forced Bouchard to come to a settlement.50 The same process, of course, shook the allegiance of one’s enemy’s vassals. In 1055 the Conqueror began the construction of a castle at Ambrières in the lordship of Geoffrey of Mayenne who recognised this as a prelude to the devastation of his lands and appealed to his overlord, the count of Anjou. When this latter was unable to prevent the construction, Geoffrey did homage to William.51
It was a much cheaper strategy, indeed one which paid for itself. It is worth noting too that important men, including even the king himself, were quite ready to profit from such affairs which enabled them to pay their retinues and give them an opportunity to plunder, for this remained an important attribute of leadership.52 In the affair of Brévol this was probably at least as important to Philip I as any notion of weakening the duke of Normandy. This kind of war, so common across the face of France, so brutal to the mass of the population, explains the anxiety of the church to divert or control the energies of the military caste. But, although there was little glory in it, warriors continued to practise what was ‘the normal business of war’, which, it should be remembered, in the context of a brutal age was not so very different from the coercive practices of civilian government. In 1051 Edward the Confessor ordered Godwin to harass Dover as a judicial punishment.53 In the early tenth century St Odo of Cluny had written his celebrated Life of St Gerald in which he praised Gerald for forbidding followers to take peasant goods casually – this was evidently common behaviour then and things had not much changed on the eve of the First Crusade when the first knightly act of Ralph, son of Albert of Cravent, was to despoil a monk.54 It is important to see the knight in this context – of a war of position where ravaging and skirmishes turned on possession of strongpoints. To attack them by siege was always possible, but could be expensive. Ravaging could undermine both the economic base of the castle and its psychological underpinning, the loyalty of the vassals who manned it, though it could be a protracted affair. Nor should the castle simply be seen as a passive defence point, for its building, as we can see by the example of Ambrières, could be an offensive act, a first stage in bringing fire and sword to an area with a view to its subjugation. It was a style of war which necessitated the mobility of the knight and at the same time demanded that he turn his hand to many tasks, especially when it came to a siege. Even Robert of Bellême turned out to have engineering skills. Nor is the knight the only figure in such skirmishes – footsoldiers figure, usually suffering heavier losses, and archers as well. It was a paradox, but a form of war which turned on possession of strongpoints also placed a heavy emphasis on mobility. And in the set pieces of this conflict, the sieges, the knightly cavalry had their role. After Robert Curthose had rebelled against his father, Duke William raised a great force and cornered the rebels at the castle of Gerberoi where they had received help from the king of France and his allies. After some skirmishing Robert suddenly led out his forces and defeated the royal army. William was engaged and wounded by his son and his horse was shot from under him. One source suggests that an Englishman, Toki son of Wigod, tried to bring him a fresh one but was shot by a crossbow, while another says that Curthose recognised his father’s voice and sent him away on his own horse.55 The incident nicely illustrates the value of cavalry even at a formal siege and explains why horsemen are so often seen besieging cities in manuscript illuminations. The garrison could use mobility to strike unexpectedly at their besiegers and the attackers needed to patrol and guard against this and be ready to take advantage of any weakness which might appear in the defence. At Gerberoi, Robert was cornered and probably desperate when he launched his charge which evidently involved horse and foot, but the royal army was caught in disarray and defeated. This was a very successful sally but a lesser objective such as the burning of siege equipment would equally make such mobility desirable.
This warfare of position with its concomitant of brutal ravaging was not confined to the Anglo-Norman world. About 1086 Godfrey’s possession of the castle of Bouillon was challenged by Albert of Namur, count of Verdun, who had earlier shown his hostility, as we have noted. He allied with Theoderic bis
hop of Verdun to besiege Bouillon. Godfrey was able to raise the siege after a costly battle outside and he countered by releasing one of his prisoners, Henry of Grandpré, on condition that he devastate the county of Verdun. The count and the bishop then attacked Stenay, where Godfrey erected a castle, with such persistence that despite much fighting Godfrey was forced to seek help from his elder brother, Eustace of Boulogne, who was accompanied by the younger, Baldwin. The affair was brought to an end by the arbitration of Henry of Liège which was probably a device to save the faces of the count and the bishop, for Godfrey seems to have lost no lands by the peace which followed. Shortly after, in 1087, Godfrey was invested by Henry IV with the duchy of Lower Lorraine.56 Raymond of St Gilles, who would later lead the largest army on the First Crusade, had to wage a struggle for at least thirteen years against Robert count of Auvergne to assert his claim to the Rouergue which by the time of his success in 1079 was a devastated zone.57 Even the Papal Legate on the crusade, Adhémar bishop of Le Puy, seems to have been a man of military experience. He was probably of the house of the counts of Valentinois and is spoken of as an exceptional horseman. His election in the late 1070s was imposed by Gregory VII’s legate, Hugh of Die, and bitterly contested by the important local house of Fay-Chapteuil, lords of Polignac, who were only defeated by a fine use of military force and diplomacy. He was not secure in his see until the mid–1080s when he may have gone on pilgrimage to Jerusalem.58
The Normans established themselves in South Italy in the early eleventh century by serving as mercenaries of the Byzantines and the Lombard princes. William of Apulia is frank about the methods which enabled them to dominate the area. They lived as brigands serving any who would pay them and preying on anyone, even admitting known criminals into their ranks. Malaterra tells us that Roger of Sicily told him to report his early life as a brigand.59 Robert Guiscard arrived in South Italy as an impoverished younger son whose older brother Drogo was already established in the area and who finally conferred on him the job of guarding the valley of Crati in Calabria close to Cosenza, from whence he later moved his position to San Marco. Here he lived as a brigand paying his men by ravishing the land and building up a reputation as a ruthless leader. The pattern of conquest is made clear by the story of Peter, the Byzantine governor of Bisignano. Guiscard had concluded a truce with him, presumably the fruit of extortion. He asked to meet Peter who came with an escort only to be ambushed and ransomed for 20,000 ounces of gold.60 Devastation and brigandage were the foundations of the conquest of Calabria and, therefore, of the greatness of the house of Guiscard. His son, Bohemond, who would be one of the leading figures on the First Crusade, had a great military reputation and was adept in this kind of war. After the death of his father, Robert Guiscard, in 1085, and the subsequent collapse of their effort to conquer the Byzantine empire, Bohemond found himself landless, for the lands of his father had passed to his half-brother Roger Borsa. But Bohemond was able to enlist the help of his family’s rival, Jordan of Capua, and to attract young men into his service by his military reputation. He ravaged Otranto and Taranto which submitted to him by 1086, together with a number of lands and counties in Apulia; eventually he agreed to hold them of his brother. In 1087 he profited by an alliance with one of Roger’s rebellious vassals to gain a foothold in Calabria, from which he attacked and seized the important city of Cosenza which he later exchanged, with the weak duke Roger, for Bari. By the summer of 1089 this able soldier was confident enough to act as host at Bari to the visiting pope, Urban II.61 Southern Italy presented a different landscape from northern Europe, principally because it was dominated by cities, but the same techniques of ravaging and siege served the same purposes. It is worth noting how self-sustaining war could be. A successful captain could attract knights from far and wide, as Curthose did in Normandy, and Guiscard and Bohemond in the south. Even the king of France saw this as a means of offsetting the costs of his military entourage. The pattern of war imposed by the strength of fortresses and cities did not necessarily result in set-piece sieges; more often than not the technique of destroying food supplies and isolating strongpoints produced some kind of result. But, as we have noted, they were frequent enough and even at the level of localised war the need for the siege caused the development of poliorcetics.
At the siege of Brévol in Lent 1092 the allied forces of the duke of Normandy, the king of France and William of Breteuil needed the assistance of Robert of Bellême who was an expert in siege machinery. Ordericus’s description of the machines he built is not very clear. The mention of a machine which was rolled right up against the wall suggests an armoured roof or penthouse which would normally be used to protect troops undermining the wall. However, Ordericus says that it was used to ‘hurl great stones at the castle and its garrison’ which implies some sort of catapult. It is possible that more than one machine is being described, and certainly the assault was effective for the outer defenses of the castle, its wall and palissades, were driven in and many roofs collapsed on the defenders. It is notable that the local clergy were made to impress their flocks for the purposes of constructing these machines.62 But at the siege of Paris (885/6) the Danes had built a three tiered siege tower, the construction of which was halted by the killing of its builders in a raid. At the siege of Laon in 938 Louis IV employed a similar device and covered its forward movement by fire of archers. In 985 King Lothar assaulted Verdun with a siege-tower higher than the ramparts, which was propelled forward by ropes which were turned around stakes planted close to the walls and pulled by oxen out of arrow range. To counter this sophisticated device the garrison built a wooden tower at the point on the wall to which it was to be applied. In 988 Hugh Capet attacked Laon with a similar, through less elaborate, weapon.63 In 1087 the Pisans and Genoese mounted an expedition, which took upon itself much of the character of a crusade, against Mahdia under its Zirid ruler Tamin (1062–1108) who had encouraged attacks on Christian commerce. In the successful attack on the Islamic fortress of Pantelleria tall wooden towers were built lignis nimis altis facti sunt turrifices’ to dominate the walls.64 This kind of wheeled wooden tower seems to have been the most effective siege engine of the eleventh century. The vagueness of the sources sometimes makes it difficult to distinguish the siege tower proper from the mobile armoured roof or penthouse which could be wheeled up to the wall to provide shelter for attackers who could then ram or undermine it. Such engines were very important during the First Crusade.65 The siege tower was a complex machine which seems to have been difficult to build and deploy. Its purpose was to cover the mounting of ladders by fire-power, as well as to deliver troops onto the wall itself. Attackers brought ladders forward and sheltered themselves, and archers, behind mantlets of woven twigs or sheets of leather which provided cover and perhaps protection. But the other kind of machine found frequently in our sources, and mentioned at Brévol by Ordericus, were projectile throwers which are given various names: mangana, mangonella, petraria, ballista and all their variants. These are the names used in Roman texts and inherited by medieval writers. Unfortunately there is no uniformity of nomenclature between writers and as their descriptions are often very vague it is exceedingly difficult to understand quite what kind of machine is being referred to and how it worked. In the Roman world such weapons were driven by torsion, the effect of huge windings of hair and sinew. However, by the thirteenth century when our sources become much clearer and more precise this principle of propulsion had been completely superceded in favour of lever action often combined with counterbalances. For the period in question here there is much doubt. The ballista was a Roman weapon in which the two arms forming the bow were pulled back against torsion coils. In medieval usage it often means simply a crossbow, and sometimes this is made very clear by a reference like arcu baleari.66 But in the context of siege it can mean a machine which was a very large crossbow mounted on a frame used as a flat-trajectory, anti-personnel weapon. In his account of the siege of Paris, Abbo tells us that on one occasio
n a single ballista bolt killed two skilled workman, while on another Abbot Ebles of St-Germain killed seven Danes with a single shot, causing him to jest that his victims should be sent, like skewered meat, to the kitchen!67 But the other weapons, which we can generically call catapults, were very different. They seem to have thrown stones and other heavy objects, occasionally even bodies or heads, at the walls and defenders. A tentative conclusion based on a survey of our sources suggests that petraria and mangana refer to heavy weapons, while mangonella means lighter machines capable of throwing a projectile of only some five kilograms. In all cases the effective range was very limited – 50 to 75 metres being a maximum. The likelihood is that by the time of the First Crusade torsion had almost totally disappeared in favour of lever action. Certainly the huge windings of rope and sinew used by the Romans are no longer in evidence.68 The construction of such weapons was a difficult business. Ordericus regarded Robert of Bellême as unusual, and the reference by Abbo to very skilled men building the Danish machines used against Paris makes the same point. The siege of Pantellaria was conducted by sailors whose engineering skills were those needed to build machines. But the decision to use siege machinery of any complexity did not simply depend on finding skilled artisans. The commander had to have knowledge of the possibilities and to appreciate what was needed. The raw material, especially tall trees, had to be available. Above all, people had to be organised for the business of siege. This was not simply a matter of ordering soldiers about. Trenches and palisades as well as machines had to be made, and food had to be brought up. At Brévol, Ordericus mentions the impressing and organising of the local populace, and we can see the same phenomenon on the Bayeux tapestry as the Normans built their castle at Hastings.69 These were the skills of a commander, a vital part of the business of war. The larger the scale of action the more important such skills became. The common experience of those who went on the First Crusade was of a kind of war which turned on possession of strongpoints, which might be assaulted by siege or undermined by ravaging. At all times the castle and the walled city were dominating factors but there was a high premium on mobility and flexibility. The needs of siege also imposed a premium on capacity for organisation and the need to combine all the various arts of war. We have seen these factors at work in smallscale conflict, but it was not so very different in the major military undertakings of the age. However, this distinction is made purely for analysis. Louis VI’s biographer catalogues some major expeditions, but for the most part shows him besieging the castles of robber barons. In this he enjoyed the support of the future crusader, Stephen of Blois, who seems to have worked closely with the French monarchy in the years before the First Crusade. At some stage between 1081 and 1084 Stephen killed the notorious robber baron Count Bouchard of Corbeil, with a blow from his lance.70 Out of this grinding process emerged royal strength and prestige as exemplified by Louis’s appeal to the chivalry of France when Henry V invaded in 1124.71
