Killing or Clemency? Ransom, Chivalry and Changing Attitudes to Defeated Opponents in Britain and Northern France, 7th-12th centuries .
by Matthew J. Strickland
On 25 September, 1066, the forces of King Harold II of England fell upon the unsuspecting Norwegian army of Harald Hadraada at Stamford Bridge in Yorkshire. In the fierce battle which ensued, the English lost many of their best warriors, but both Hardraada and his ally Tosti Godwineson, Harold’s own brother, were slain and the Norwegians virtually annihilated. According to the Anglo-Saxon Chronicle, the English king gave quarter to the Norwegian reserve force under Olaf, Hardraada’s son, and the earl of Orkney, who had not been present at the main battle, but of 300 ships which had sailed into the Humber earlier that month, only 24 were needed to carry away the survivors.We hear of no prisoners, no ransom.
Less than sixty years later, in 1119, another king of England, Henry I, but now also duke of Normandy, met an invading French army under Louis VI at Brémule in the Norman Vexin. The battle was a resounding victory for the Anglo-Normans, yet of the 900 or so knights engaged, only three were killed. The Anglo-Norman chronicler Orderic Vitalis, writing at the monastery of St Evroult in southern Normandy and one of our finest sources for the nature of contemporary warfare, offered his own explanation for this striking lack of casualties:
They were all clad in mail and spared each other on both sides, out of fear of God and fellowship in arms (notitia contubernii); they were more concerned to capture than kill the fugitives. As Christian soldiers, they did not thirst for the blood of their brothers, but rejoiced in a just victory given by God for the good of Holy Church and the peace of the faithful.
We must treat Orderic’s interpretation with caution, for it was influenced not only by his monastic vocation, but also his desire to portray Henry I’s wars as fully in accordance with Augustinian concepts of the just war; Henry’s soldiers fight not only a war of defence but fight with right intent, devoid of hatred. In this context, mention of ransom is studiously avoided, though we know from many other references by Orderic himself that the ransoming of knightly captives was in practice widespread. Moreover, while killing tended to be remarkably constrained in several major battles fought within the Anglo-Norman regnum in a context of civil war, the very low casualties at Brémule were more exceptional compared to other Franco-Norman engagements, where considerably higher casualties might occur in far smaller skirmishes.Nevertheless, the contrast with the battle of Stamford Bridge highlights one of the fundamental distinctions between Anglo-Scandinavian and Franco-Norman conduct in warfare that is the subject of my discussion here, namely the treatment of enemy warriors and developing concepts of ransom.
The Norman conquest of Anglo-Saxon England from 1066 and the subsequent penetration into the Celtic lands of Wales, Scotland and Ireland has been characterised by Robert Bartlett in his The Making of Europe as part of a wider aristocratic diaspora from the Frankish heartlands to the peripheries of Europe, achieved in large part by the superior military technology of castles and cavalry. While this model is largely valid for the Celtic lands,I have argued elsewhere that it is more problematic for late Anglo-Saxon England, which possessed sophisticated military institutions, including a well-organized army, a network of fortifications or burghs and a powerful fleet. Indeed, though the Anglo-Saxons fought predominantly on foot and not as cavalry, there were many similarities between the warrior aristocracies of Normandy and late Anglo-Saxon England. As the Bayeux Tapestry makes clear, their arms and equipment were virtually identical, while the military households of Anglo-Scandinavian lords had much in common with their Norman and Frankish counterparts. Both aristocracies, moreover, shared many of the essential martial values; while not identical, the ethos of the Song of Roland is close indeed to that of the great Anglo-Saxon poem The Battle of Maldon , which commemorates the heroic last-stand of Ealdorman Byrhtnoth of Essex and his men against the Vikings in 991. This is unsurprising, for core virtues – courage, loyalty to one’s lord and comrades, generosity, and a jealously guarded sense of honour and reputation – have had an almost universal validity among warrior elites. Yet although such virtues operated within the war-band (comitatus, familia, mesnie), between a lord and his military following or between comrades in arms, they were not necessarily extended to their opponents. Indeed, where notions of conduct might fundamentally differ between warrior aristocracies was in the treatment of the enemy and attitudes towards prisoners. In the British Isles before 1066, the general fate of those defeated in battle or taken in war was either death or enslavement. The Norman Conquest, however, was to mark the importation into England of a differing military ethos, which placed an increasing stress on ransom and the sparing of knightly captives, and which eschewed the enslavement of prisoners of war as a token of barbarism. Nevertheless, Anglo-Norman advances first into Wales then Ireland and intermittent warfare with Scotland brought them into direct and sustained conflict with peoples whose methods of warfare still involved the killing or enslavement of captives until well into the twelfth century. As a result, a combination of military factors and cultural preconceptions which cast the Celtic peoples as uncivilised barbarians, ensured that Anglo-Norman conduct towards the Welsh and Irish was marked by a ruthlessness and cruelty rarely exhibited in theatres of war within England or France. Given the scope of the subject, what follows can only be a broad overview, but one which, however impressionistic, attempts to map a fundamental shift in the nature of behaviour in war.
Attitudes to captives in early Anglo-Saxon warfare are nowhere better encapsulated than in an incident related by Bede in his Ecclesiastical History, completed in 731, concerning a Northumbrian thegn called Imma, who had been left for dead following a great battle on the river Trent between King Ecgfrith of Northumbria and Aethelred of Mercia in 679. Taken prisoner by the Mercians, Imma initially tried to pass himself off as a ceorl or simple peasant, who had only been bringing food to the army, the implication being that, as a warrior, his life was at far greater risk. Given away by his noble demeanour, however, he was told by his captor, a Mercian gesith (nobleman), that he ought to die in revenge for several of the gesith’s kinsmen who had been slain in the battle. Nevertheless, before he had discovered Imma’s noble identity, the gesith had promised him his life and he now kept his word; but while his life was spared, the Mercian did not ransom him back to the Northumbrians, but sold him as a slave to a Frisian merchant.The fact that Imma was a noble, an Angle – rather than a Celt – and a fellow Christian made no difference to his treatment.
Here we see the essence of war in the Anglo-Saxon heptarchy – the killing of noble opponents precisely because of their nobility and value as warriors, and the enslavement, not ransom, of those prisoners who were taken alive. There is great emphasis on vengeance for slain kin, powerfully echoed in Anglo-Saxon epic literature such as Beowulf, which is replete with bitterly pursued blood feuds. Beowulf himself comforts King Hrothgar, grieving over the death of his beloved thegn Aeschere, with the words, ‘Do not be sorrowful, wise man. It is better for anyone that he avenge his friend than mourn greatly’. Crucially, the value of a man of rank was assessed less at the price of his freedom, but at the sum paid in compensation for his slaying,and the payment of such a ‘man-price’ or wergild could buy off further feud. In the same battle in which Imma was involved, the death of King Ecgfrith’s brother, Aelfwine, ‘gave every indication of causing fiercer strife and more lasting hatred between the two warlike kings and peoples’, until with the aid of Archbishop Theobald, peace was established and the Mercians paid Aelfwine’s wergild to the Northumbrian king.Other mechanisms existed to regulate and limit the extent of war such as truces, the exchange or extraction of hostages, or the purchasing of peace by payment of tribute. Yet as in the Merovingian Gaul of Gregory of Tours, wars of territorial expansion and dynastic consolidation were still marked by the ruthless extirpation of defeated kings and royal kindred. At the great battle of Winwaed in 655, fought between Penda of Mercia and Oswiu, king of Northumbria, nearly all of the thirty duces fighting for Penda were slain,while in seventh-century England a confessional dimension may have added to the bitterness of warfare between pagan and Christian opponents. Thus, for example, after his defeat at the hands of the pagan king Penda in 641, the body of Oswald of Northumbria was dismembered and his head and arms placed on stakes, while following his invasion of the Isle of Wight in 686, King Cadwalla of Wessex executed the captured brothers of its heathen king, Arwald, pausing only to allow them to be first baptized. The same emphasis on slaying and enslavement is found in warfare between the Anglo-Saxons and the Celts, and between Celtic enemies. King Ecgfrith of Northumbria was slain with the greater part of his army at the hands of the Picts at Nechtansmere in 685, while the Pictish stone at Aberlemno, which probably commemorates this event, depicts the fallen king, with a raven pecking at his eyes. The later tenth-century carved stone at Forres in north east Scotland, known as the Sueno Stone, provides a still more graphic illustration of the fate of captured warriors; beneath a battle scene are depicted rows of decapitated prisoners, their hands tied behind their backs. The collection of enemy heads was still a commonplace in Anglo-Scottish warfare at the turn of the year 1000. Thus in c.1003, following the rout of a Scots army outside Durham, Earl Uhtred of had the severed heads of the enemy washed and groomed by local women, paid for their gory task by the gift of a cow each, then placed around the city on stakes, while the city’s market place was similarly decorated with Scottish heads following the defeat of King Duncan’s force in 1039. In Celtic warfare, raiding too was habitually marked by the massacre of able-bodied men and the enslavement of women and children. When, for example, the viking town of Limerick was sacked by Mathgamain, king of Munster, in 967, the captives were rounded up and ‘every one of them that was fit for war was killed, and every one that was fit for a slave was enslaved’. The profits of war came not from ransom, but from slaves, livestock and booty. In Wales, Ireland and Scotland, such a pattern of warfare was to continue until well into the twelfth century.
It was the vikings who made both slave-raiding and the levying of tribute the hallmark of their operations in Britain and Frankia from the early ninth century, with powerful armies extorting huge sums from kingdoms or localities to buy off devastation. While some campaigns, such as those of the ‘the great army’ from 865, or the invasions of Svein Forkbeard, were waged for conquest, for most aggression was primarily the means for amassing of wealth. As the Norse messenger tells Byrhtnoth in The Battle of Maldon; “…it is for you to send treasure quickly in return for peace, and it will be better for you all that you buy off an attack with tribute, rather than that men so fierce in battle as we should give you battle. There is no need that we destroy each other, if you are rich enough for this ”. From the 990s, the England of Aethelred the Unready was bled white by vast and almost annual payments to viking fleets; rune stones throughout Scandinavia boasted of the geld the drengrs had received from Svein or Cnut, and reality of such vaunts has been revealed by the finds of great hoards of English pennies. The nature and impact of what we may perhaps term this form of ‘corporate ransom’ in England and Frankia is well known, but the vikings’ search for profit also led to a ransoming of individuals on a scale greater than can be seen in earlier warfare.Leading ecclesiastical and secular figures were targeted for the great sums that might be paid for their release. In 914, the Welsh bishop Cyfeliog of Archenfeld was redeemed from the vikings by Edward the Elder, king of Wessex, while in 939, the Norse leader Olafr Gothfrithsson seized Wulfrun, mother of the great Mercian nobleman Wufric Spott, for ransom. In the same period, the vikings of Limerick and Dublin captured and ransomed several Irish kings. It was the refusal of Archbishop Aelfheah of Canterbury to allow himself to be ransomed that led to his death in 1012 at the hands of the men of Thorkell the Tall. The more extensive Frankish sources reveal a similar picture. In 858, Abbot Louis of St Denis and his brother, Gauzlin, were ransomed for a sum so enormous that it required not only a general levy of the churches of Neustria but also additional donations from Charles the Bald and his nobility. On the secular side, the Chronicle of Fontanelle records that in 841, the Vikings ransomed sixty-eight prisoners for twenty-six pounds of silver, while according to Regino of Prum, in 881 Evesa, wife of Count Megenhard, ransomed her son Abehard the Saxon for a large sum. The practice was widespread enough for Charles the Bald’s Edict of Pitres of 864 to forbid anyone to ransom themselves by giving the Vikings arms, armour, or horses. Clearly the vikings were trading captives for high quality Frankish swords, hauberks and mounts, and the Emperor was keen to prohibit such superior weaponry from leaving the kingdom or being turned against the Franks. Even slaves could be ransomed; in the treaty of 866 made between Charles and the Northmen, they stipulated that any slaves captured by them who had escaped were either to be returned or ransomed for a price set by the Northmen themselves.
It is possible that interaction with vikings, as with Muslim raiders in the Mediterranean, may have stimulated and disseminated the practice of ransom within Frankia, or, as in the case of the treaty of 866, broadened the social basis of its operation to include the unfree; but it did not create such conventions. By the ninth century, the enslavement of captives, widely practiced in the Merovingian period, had become largely restricted to pagan opponents beyond the Frankish lands, while in 866 Pope Nicholas I could explicitly tell the newly converted Bulgars that the slaughter of captives and of women and children was a grave offence. Certainly battles fought in the civil wars between the grandsons of Charlemagne and their successors could be bloody. The Annals of Fulda noted ruefully of the battle of Fontenoy that ‘there was such great slaughter on both sides that no one can recall a greater loss among the Frankish people in the present age’, while the Frankish bishops were moved to issue a penitential edict for those involved and a three-day fast.
The grave unease felt about such internecine killing among the Franks themselves,and the desire of the victors to portray the battle as the just judgement of God goes far to explain why mention of ransom is studious avoided in the sources for Fontenoy and other battles fought within the context of civil war, just as Orderic’s stress on the right intent of combatants at Brémule in 1119 led him to suppress any reference to the capture of prisoners for financial gain. The same motives led chroniclers to emphasize the restraint and clemency shown by the victors. Hence Nithard, who was present at the battle of Fontenoy in 841, was at pains to stress that the victorious army of Charles the Bald and Louis the German did not press home their pursuit of Lothair’s forces, while at the hard-fought battle of Andernach in 876, the victorious Louis the Younger ‘took alive many of Charles’ leading men, whom in his humanity he ordered to be spared unharmed’.It would seem likely that these men had to pay for their ultimate release, as did those captured in more localised conflicts as, for example, in 850, when Lambert, count of Nantes and Nominoe of Brittany ravaged Maine and took noble Franks prisoner.In their dealing with the Northmen, moreover, the Franks seem to have had an expectation of ransom. Thus Abbo of St Germain records how when during the siege of Paris in 886-7, a group of twelve Franks were surrounded and called upon to lay down their arms, they did so, expecting subsequently to be able to buy their freedom. Instead, however, they were treacherously slain.Similarly, in 890 the Frankish garrison of St Lo surrendered to the Norsemen who promised that they would spare their lives and take only the booty, but on leaving the fortification they were all killed, with the bishop of Coutances among them.
That the Franks could be rudely disabused in their expectation to be spared for a price serves, by contrast, to illustrate the serious limitations on the vikings’ own operation of ransom. First, the majority of those captured by Vikings were not ransomed back to their own people, but sold abroad as slaves, one of the most staple commodities of Viking trade from Dublin to the lands of the Rus.So adept did the Vikings become at this that attacks on certain monasteries and religious centres were co-ordinated with major feast days in order to capture the maximum number of pilgrims, and the churches themselves spared in order to regularly milk their congregations for slaves. Second, the enslaving of captives was far from habitual. The inhabitants of numerous Frankish towns were put to the sword, or, as at Noyon in 859, were carried off only to be subsequently slain.Similarly, the Translation of Saint-Germain recorded how the Vikings retired with their captives after a raid, but then hanged 111 of them, while according to Annals of St Vaast, captives taken in the Tournai region in 880 were executed. During the great siege of Paris in 886-7, the vikings deliberately put to death Frankish captives before the eyes of the defenders, while c. 935 a Norse force operating in the region of Thérouanne, confronted by the approaching army of King Ralph, strangled all its prisoners and withdrew. Even the most noble might be killed in cold blood; King Edmund of East Anglia, martyred by the Danes in 869, was but the most famous of many such royal and noble victims. Churchmen were equally vulnerable. Between 858-9, the bishops of Bayeux, Beauvais and Noyon perished at the hands of vikings operating on the Seine, while in 870 the abbot and monks of Peterbrough were slain, with these being far from isolated examples.
Thirdly, within battle itself, viking forces showed little propensity to spare the lives of warrior opponents, whether out of respect for their prowess in arms or for their financial worth. The Anglo-Saxon Chronicle records a striking number of high-ranking men slain in battle between Anglo-Saxon and viking forces. At Maldon in 991, not only did Byrhtnoth and his household retainers perish at the hands of a viking force under Olaf Trygvasson, but the ealdorman’s severed head was carried away as a trophy. In 1001, the Viking army fought with the men of Hampshire near Dean, ‘and there Aethelweard, the king’s high reeve, was killed, and Leofric of Whitchurch and Leofwine, the king’s high reeve, and Wulfhere, the bishop’s thane, and Godwine of Worthy, Bishop Aelfsighe’s son, and eighty-one men killed in all’.At Thetford in 1004 ‘the flower of the East Anglian people was killed’ while at Ringmere in 1010 ‘the king’s son-in-law, Aethelstan, was killed there, and Oswig, and his son, and Wulfric, Leofwine’s son, and Eadwig, Aelfric’s brother, and many other good thanes and a countless number of people’.At the great battle of Ashingdon in 1016 between Edmund Ironside and Cnut, several ealdormen were killed and ‘all the nobility of England were there destroyed’.
The vikings, however, did not have a monopoly on such slaying, and in defeat themselves faced death. One Viking king and five jarls fell against the West Saxons at Ashdown in 871, and the lives of seven more jarls were claimed in further engagements that year.In 878, Ubba, the brother of Ivar the Boneless, was killed in Devon, along with 840 of his men. while in 893, the Anglo-Saxon forces besieging Chester slew all those vikings emerging from the beleagured town. A similar slaughter of enemy nobles occurred in engagements such as the battle of Tettenhall, fought during the ‘Reconquest’ of the Danelaw by the kings of Wessex, while following Aethelstan’s great victory at Brunanburgh in 937 over a coalition of a coalition of Celtic kings and the Hiberno-Norse, ‘five young kings lay on that field of battle, slain by the swords, and also seven of Olaf’s earls, and a countless host of seamen and Scots’, while the son of King Constantine of Scotland also lost his life there.The emphasis here is again on extirpating the enemy’s elite, not on attempting to capture them for ransom.
How does one account for such intensity of warfare, and a general absence of clemency? For the Anglo-Saxons and Franks, two important factors exacerbated the bitterness of hostilities against viking opponents. The first was a powerful religious dimension to warfare against an enemy who, until the eleventh century, was usually pagan. There is a strong flavour of holy war in Alfred the Great’s struggle from the 860s against ‘the great heathen army’ and its successors, while a similar stress has been noted in the Old High German Ludwigslied, celebrating the victory of Louis III over the Vikings at Saucourt. The Anglo-Saxon nobleman Aethelweard, producing a Latin version of the Anglo-Saxon Chronicle, noted that at the great victory of Edward the Elder at Tettenhall in 910, the viking leader Ivar ‘lost his sovereignty and hastened to the court of hell’, while at Maldon in 991 Ealdorman Brythnoth proclaims, ‘Heathen shall fall in battle!’Penitentials recognized the legitimacy of killing in such a context. Whereas homicide in a public war, even at the command of a legitimate ruler, might require penance for forty days and abstention from church, ‘if an invasion of pagans overruns the country, lays churches waste, ravages the land, and arouses Christian people to war, whoever slays someone shall be without grave fault, but let him nearly keep away from entering the church for seven, or fourteen, or forty days, and when purified in this way, let him come to church’. Not only that, but killing the heathen might win religious merit, and it seems little coincidence that from the late ninth century comes the first surviving benedictional for the blessing of weapons, the Ordinatio super militantes.
Secondly, wars against the vikings were not merely waged against pagans, but were defensive wars to safeguard the patria or homeland from an invader, and as such were unequivocally just wars, a rihtlic gefeoht. Aelfric, the great Anglo-Saxon homilist and scholar of the early eleventh century, commented on Isidore’s definition of a just war; ‘justum bellum is just war against the cruel seamen or against other nations who wish to destroy our homeland’.The doctrines of Augustine and other Church fathers here blended easily with an equally powerful secular tradition, perfectly expressed by the poem preserved in the Anglo-Saxon Chronicle on Aethelstan’s great victory at Brunaburgh in 937 : ‘Edward’s sons clove the shield wall, hewed the linden-wood shields with hammered swords, for it was natural for men of their lineage to defend their land, their treasure and their homes, in frequent battle against every foe’.
How far the vikings were motivated by religious sentiments in war is far less clear. Acceptance that the gruesome rite of the ‘blood-eagle’ is the stuff of legend and not history still leaves open the wider question of the existence of militant paganism.There is evidence that the sacrifice of prisoners of war, visible among some of the early Germanic peoples, continued among the pagan Saxons till as late as the seventh century, and it is possible that some of the killing of captives and execution of leaders by the vikings may have occurred in this context. What is more certain, however, is that the deliberate execution of captives was itself an act of psychological warfare, designed to instil terror and to break further opposition. Moreover, the killing in or after battle of as many members of an opponent’s warrior elite as possible struck directly at the military potential of a kingdom or of an invading force, just as the killing of political, military and even spiritual leaders could further weaken an enemy’s ability to co-ordinate resistance or aggression. As the Anglo-Saxon Chronicle noted, describing Ealdorman Aelfric’s treachery in 1003, ‘As the saying goes, “When the leader gives way, the whole army will be much hindered”’. It was this rationale that must in large part have lain behind the vikings’ prediliction for executing kings and great nobles, just as it doubtless informed the decision of Otto I to hang captured Magyar princes at Regensburg after his victory on the Lech in 955, thereby striking a major blow at the command structure of the invaders.
As in earlier insular warfare, concepts of feud and vengeance may also have been powerful factors exacerbating the extent of killing. Reprisals were rarely on the scale of Charlemagne’s execution of 4,500 Saxons at Verden in 788, in vengeance for the deaths of two missi, Adalgis and Gailo, four counts and over twenty nobles, slain in battle by the forces of Widukind. Nevertheless, when Alfred the Great had two shipwrecked vikings hanged at Winchester in 898, following a naval engagement on the south coast, it was probably in vengeance for men of his own household and others slain in this battle. The desire to avenge a fallen leader or comrades appears as a powerful motif both in The Battle of Maldon and its literary analogues such as the Old Norse Bjarkamal, and, although the Jomsvikinga Saga is in itself unreliable, the duty enjoined on members of this legendary military brotherhood to avenge fallen comrades may well have reflected similar sentiments among the close-knit military households of housecarls serving Anglo-Scandinavian kings and great lords. In the early 2nd century, Tacitus had famously noted the custom among the early Germanic tribes that if their lord fell in battle, his warriors were obliged to fight on to the death, for flight brought a lasting ignominy, although scholars have been divided as to the continuing impact of such an ideology in Anglo-Saxon England, particularly in relation to its influence on the poem The Battle of Maldon. It is, of course, seldom if ever possible to gauge the reasons and personal influences that lay behind individual acts of courage, yet is certain that some men did in reality choose to fight to the death rather than flee and were honoured for so doing. Moreover, whilst the shame and disgrace of flight were clearly powerful factors, discussions of the Anglo-Saxon heroic ethos have rarely considered the enemy’s treatment of the defeated. Here we may suggest that the desire to fight on was intimately linked to the customary fate of prisoners and the absence of honourable surrender, for, as Philippe Contamine has noted, ‘the history of courage is bound up with that of risks’. While by the eleventh or twelfth century, a knight could surrender to a fellow knight with a reasonable expectation that his life would be spared for ransom, the choice in earlier centuries for vanquished warriors was a stark one between flight or almost certain death. The absence of ransom, and its subsequent development and dissemination, must thus play an integral role in the history of courage.
By the eleventh century, warfare in Normandy and France presents a markedly different picture to the ruthless conduct we have examined above. In pitched battles within or between the territorial principalities, casualties among nobles and knights could still be high. Thus, for example, in Normandy, the engagements of Val-es-Dunes, 1047, Mortemer, 1054, and Varaville, 1057 were all hard-fought, while smaller scale skirmishing and castle-based warfare among warring nobles in the duchy claimed a steady toll in lives. But while warriors might be killed in open combat, there was now a marked emphasis on the ransoming of knightly opponents, and the deliberate execution or mutilation of prisoners was rare. When in the early years of the eleventh century, Geoffrey of Thouars cut the hands off the knights (caballarios) of Hugh de Lusignan defending the castle of Mouzeil, the act was clearly regarded as an outrage; significantly, Hugh’s retaliation had only been to refuse to ransom those of Geoffrey’s men he himself held captive. The mutilation of knights rarely occurred outside the punishment of rebellion,  and even here the political community resisted the corporal or capital punishment of its members by king or duke, holding this to be unchivalric. As Orderic has the count of Flanders inform Henry I in 1124, when the king had ordered the blinding of Geoffrey de Tourville, Odard of Le Pin and Luke of La Barre for supporting the rebellion of his nephew William Clito; ‘My lord king, you are doing something contrary to our custom in punishing by mutilation knights captured in war in the service of their lord’. Lords guilty of habitual cruelty such as the notorious Robert of Belleme, who tortured, mutilated or starved his prisoners instead of ransoming them, were the exception and were criticised by contemporaries for their excesses.
By contrast, texts such as William of Poitiers’ Gesta Guillelmi (written c. 1077) show an emphasis on taking opponents for ransom, on the release of prisoners and on mercy as a noble virtue. Hence although the battle of Mortemer in 1054 had been bitterly fought, Duke William released French knights taken prisoner there, while Poitiers even praises Guy of Ponthieu, who had seized Earl Harold Godwineson en route to see Duke William in 1064, because he had treated him with honour and chose not to exercise his right to torture, mistreat or ransom his prisoner.  The good treatment of prisoners, moreover, was already considered an integral princely virtue by the early eleventh century. William the Great of Aquitaine was praised by his encomiast Ademar of Chabannes for refusing to execute or mutilate his political opponents,  and likewise Norman chroniclers were keen to stress the clemency of their early dukes. Thus according to Dudo of St Quentin, following a bloody rout of Theobald of Chartres men outside Rouen, c. 962, Duke Richard I (942-996) supposedly ‘felt much pity at the death of so many. He ordered them to be buried. Those who were still alive he had carried gently to Rouen on litters, and healed. Besides that, he had the thickets and bogs searched, and found many dead and wounded, and he attended to them with the same dutifulness’. William of Jumieges following this account closely, added that once the wounded men had recovered, Richard sent them back to Theobald, with the implication that no ransom was demanded.
The reality of conduct in later tenth-century Normandy, however, was probably more complex than this idealised image suggests. For although the Scandinavians who took control of Normandy from c. 911 came to assimilate Frankish codes of conduct as well as their methods of warfare, the process was complicated by Normandy’s sustained contact with Norse fleets until the early eleventh century. In 1013, during his attacks on Aethelred’s England, Svein Forkbeard of Denmark concluded a treaty with Duke Richard II whereby in return for assistance and medical aid, the Danes would sell their booty in Normandy;  such booty may well have included slaves, auctioned in the markets of Rouen. Not for nothing did Richer regard the previous duke of Normandy as merely a ‘pyratorum dux’. Norse forces, some of which were still pagan, were also operating as mercenaries within France itself, and their conduct in war appears just as ruthless as that of their forbears. Ademar of Chabannes, for example, describes how c. 1010-13 the knights of William the Great of Aquitaine were defeated by a mixed force of pagans and Christian Norse led by Olaf Haraldson (later St Olaf), when they charged into concealed ditches and many were captured. William was forced to withdraw from the battle ‘for fear of those men who had fallen first and were now held captive (thirty were among the noblest), lest they might be slaughtered by the Norsemen’. The duke later succeeded in ransoming these men for a great sum of money. Little wonder that Ademar couched Duke William’s struggle against Olaf’s army in terms of a holy war. Put within the context of Bartlett’s ‘centre-periphery’ model, here we see the forms of behaviour in war associated with peoples of the periphery, namely slaving and the slaying of prisoners, still operating within the ‘core’ of the Frankish heartlands at the opening of the eleventh century. Ironically, though the Normans were to introduce chivalry into Britain, the duchy of Normandy itself may have been among the last parts of Frankia to fully assimilate such chivalric conduct.
Nevertheless, the activities of the Norse only serves to highlight the contrast with the increasing stress on ransom and clemency visible in warfare within and between the territorial principalities of France by the year 1000. Two crucial factors go far towards explaining this shift of emphasis, namely the changing nature of warfare and the changing nature of the enemy. Let us first look at the nature of warfare. For much of the tenth and eleventh centuries, warfare in northern France was confined to comparatively restricted conflicts between the rival dynasties of Normandy, Anjou, Blois-Chartres, Flanders and the rulers of the Ile de France, or to the small-scale local skirmishing between rival noble kindreds within these principalities themselves. For the Franks in what had formerly been Neustria, the great annual hostings of Carolingian armies beyond the regnum had ceased, and with them the opportunities of booty and tribute on a grand scale, while the enslaving of Christian prisoners of war had all but ceased. It has been disputed how far the ‘feudal transformation’ of c.1000 witnessed the replacement of existing forms of violence within a framework of public order by a newer, coercive ‘private violence’ of lords and castellans. Like violence towards local peasantry, purveyance and the depredations of royal and seigneurial households, the ransoming of warriors within in Frankia, as we have suggested, was not in itself new. Nevertheless, the fragmentation of centralized authority and the rise of banal lordship created a political environment in which rival castellans needed to finance both castle-building and growing numbers of milites required to garrison them. This could be done by the levying of arbitrary taxes and labour services upon on the local populace and churches, such as the tenserie or angaria so bewailed in many ecclesiastical sources; or by the kind of brigandage, spoliation and extortion of ransom from the peasantry witnessed so graphically by the peace oaths, such at that of Beauvais (1023), enjoined upon castellans and their knights by the peace councils of the late tenth and early eleventh centuries. But if the operation of such predatory lordship witnessed an intensification of violence towards the laboratores and oratores, the small-scale castle-based warfare between castellans may well have stimulated the operation and dissemination of ransom among the bellatores themselves. For captured milites were a better financial prospect than captured peasants. Among a list, drawn up c.1000 of the many grievances he had suffered at the hands of his lord, Duke William of Aquitaine, Hugh of Lusignan noted that he could have made 4000 solidi from the knights he had captured from Geoffrey of Thouars, but that Duke William had taken them from him – an early reference to a lord’s prerogative right to a vassal’s captives. The ransoming of captured knights helped war pay for itself.
The process is well illustrated by Orderic Vitalis’ account of William the Conqueror’s prolonged siege of Sainte Suzanne, held against him by Hubert, vicomte of Maine, from 1083-5, where despite establishing a siege castle in the valley of the Beugy, the Normans could make no headway. The constant skirmishing, however, attracted experienced knights from Aquitaine, Burgundy and other parts of France to Hubert’s service in search of booty and ransoms, thereby further strengthening his position and prolonging hostilities. As Orderic noted wealthy Norman and English lords were frequently captured ‘and Hubert the vicomte, Robert the Burgundian, whose niece he had married, and their other supporters made an honourable fortune out of the ransoms of these men’. Ransom could thus serve to mitigate the extent of killing, yet paradoxically it could equally act as a powerful dynamic in the escalation of warfare.
The Normans’ failure at Sainte Suzanne highlights another important consideration. While siege warfare had long been an integral aspect of early medieval warfare, the proliferation of castles around the year 1000, coupled with the growing superiority of the art of defence over that of attack, made wars of rapid conquest increasingly difficult, as seen, for example by hostilities in the Touraine under Fulk Nerra, or the campaigns of William the Conquer and William Rufus in the Vexin or in Maine. In terms of conduct, this had important effects. The impregnability of major fortresses added further incentive to sparing noble opponents, whose ransom might include the surrender of key castles or cities. In 1044, for example, Geoffrey Martel, count of Anjou, captured Theobald III of Blois-Chartres and many of his nobles at the battle of Nouy, and exacted in return for their release not only the great castles of Chinon and Langeais, but the strategically vital city of Tours. When Geoffrey of Mayenne was captured by William Talvas, probably in the 1040s, he was only released on the condition that Geoffrey’s vassal William son of Giroie destroy his own castle of Montaigu ‘which was a threat to his power’. With the multiplying of ‘private’, seigneurial castles in the decades around 1000, such bargains became far easier and more widespread than had been possible with earlier ‘public’ fortifications. Conversely, however, the desire to speed the capitulation of castles or ransom might lead to the maltreatment of captives. Hence, for example, it was probably no coincidence that Count William VI of Aquitaine, captured by Geoffrey at the battle of Mont Couer in 1033, died shortly after his release, having surrendered several key fortresses, while the troubled reign of King Stephen (1135-1154) furnishes several instances of lords being maltreated or threatened with hanging in order to gain possession of key castles.
Nevertheless, the protracted nature of many sieges gave opportunities for parley and negotiation, and the spread of castles helped to disseminate and develop the ‘customs of war’ regulating respite, truce and surrender which formed an important aspect of chivalric conduct. Thus besiegers might grant a beleaguered garrison permission to seek aid from their lord, with the condition that if such aid was not forthcoming, they could surrender with their lives and their honour safe. A remarkable picture of the kind of conduct that might occur in such circumstances is furnished by Orderic’s account of the siege of Le Mans, whose near-impregnable citadel was held by a picked Norman garrison against Count Helias of Maine in 1100. William Rufus’ campaigns of conquest in Maine in the 1090s had seen bitter fighting and much devastation of the countryside, while the garrison itself had earlier set fire to the city of Le Mans. Nevertheless, daily the two sides held parleys and threatened each other, but jokes were often mixed with the threats. They gave Count Helias the privilege whenever he wished of putting on a white tunic and in this way having safe passage to the defenders of the citadel. Since he trusted the good faith of the men he knew to be both valiant and honourable, he often visited the enemy wearing the distinctive white garment and never feared to remain alone for long conferences with them. Besieged and besiegers alike passed their time in jocular abuse and played many tricks on each other in a far from malevolent spirit, so that the men of those parts will speak of them in days to come with wonder and delight.
On the death of Rufus, the garrison sent messengers to his brothers Robert Curthose, duke of Normandy and Henry, king of England, but since both were unable to send aid, they gave the garrison permission to surrender. They obtained not only honourable, but even lucrative terms, and when the garrison finally yielded the citadel to Helias, the count escorted them out of the city with their horses, arms and possessions ‘not as vanquished foes but as faithful friends’.
This brings us to our second major consideration, the change in the nature of the enemy. For warfare within France itself in the later tenth and eleventh centuries now generally lacked the religious dimension which lent so much bitterness to wars against the heathen, be they vikings, Magyars or Saracens. The raids of Olaf Haraldson against Brittany and Aquitaine in the early decades of the eleventh century were among the last launched by pagan Scandinavians, and Olaf himself was baptised in Rouen between 1013 and 1014. The decision of Count Helias of Maine in 1096 to regard his war of defence against William Rufus as a crusade, and to bear the sign of the cross on his armour and equipment, was very much an exception, and it was only with the thirteenth century that political crusades within Europe became widespread. Opponents in warfare in northern France of the eleventh and twelfth century were thus not foreign, pagan invaders, but warriors sharing the same faith, culture, language and perceptions of conduct. The question of the origins of knighthood is controversial. Some, such as Karl Leyser and Janet Nelson, have argued for the existence of a shared brotherhood in arms between conmilitiones as early as the ninth century, while others such as Jean Flori would see the process largely the as creation of the twelfth century. Certainly, however, the evidence of writers such as William of Poitiers and Orderic Vitalis suggests that by the mid-eleventh century at the latest, the heavily armed and well-mounted warriors operating within the territorial principalities were conscious of common membership of a military elite, which encompassed the lowliest milites gregarii to the greatest of lords in a shared profession of arms. Poitiers’ Gesta Guillelmi seeks to portray Duke William himself as the flower of the militia of France, and a revealing episode relates how when aiding the king of France at the siege of Mouliherne against Geoffrey Martel of Anjou, c. 1049, the duke with only four companions attacked a larger group of enemy knights. He unhorsed one but took care not to run him through, then succeeded in taking seven others captive. Here Poitiers, who himself had been a knight, is keen to celebrate not the killing of the enemy but the duke’s youthful feat of arms against opposing milites ‘glorying in their horses and arms’. The emphasis on a brotherhood of arms becomes still more explicit in the pages of Orderic, whose striking account of the good-humoured relations between besieged and besiegers at Le Mans in 1100 we have already noted. Similarly, Orderic recounts how when in 1098, during his attempted conquest of Maine, William Rufus entered the castle of Ballon, held by the hated and feared Robert of Belleme, the Angevin and Manceaux prisoners there called out to the king for him to free them. The king immediately ordered them to be released on parole, and when his councillors objected that they might escape, he replied, ‘Far be it from me to believe that a knight would break his sworn word. If he did so, he would be despised forever as an outlaw’. Here the emphasis is on Rufus’ clemency and franchise not so much as a king, but as a fellow knight.
But what of conduct in the localised but bitter conflicts which occurred in Normandy during periods of ducal weakness in the 1030s and 1090s between rival families such as Tosny, Grandmesnil, Clere and Beaumont? Was the general trend towards clemency and ransom, and developing notions of chivalry retarded by the nature of war between opposing aristocrats exploiting, or reacting to, a vacuum in centralised power? There is debate as to whether such hostilities were actual bloodfeuds, or merely the kind of expansionist warfare practised by the territorial princes but in miniature. Certainly, murders, mutilations and revenge killings formed a prominent feature of such aristocratic conflicts. Yet significantly, these rarely occurred within the context of open warfare, but instead were assassinations, poisonings, night attacks or ambushes on men all unprepared for combat; thus in the turbulent minority of Duke William, William Giroie was mutilated at a wedding feast by William de Belleme; Count Gilbert of Brionne was assassinated during a parley whilst among his retinue; William of Montgommery cut the throat of Osbern, steward to the young duke William, in the ducal bechamber, and in turn was slain by Osbern’s steward, Barnon de Glos, again by stealth at night; and so on. By contrast, though localised conflicts could be bloody, open warfare could be announced by formal challenges many days in advance, and often was less about killing than the seizure of arms, horses and prisoners for ransom. This is made clear by a clause in the Norman Consuetudines et Justicie, drawn up in 1091 but purporting to record key ducal rights in the time of the Conqueror: ‘No one was allowed to capture a man in war, to ransom him, nor take money from a war or a fight, nor take away arms nor a horse’. Here the duke was aiming to prohibit the fundamental mechanisms of localized warfare. Though opponents could be killed in these encounters, there was an underlying expectation that vanquished knights who were captured or who surrendered would be ransomed and not slain out of hand. Thus when in the war of the Breteuil succession, Reginald de Grancy personally slew all the members of a defeated garrison on surrender, Orderic could remark, ‘This was the chief act that made him an object of universal hatred’. Moreover, even in the context of such ‘private war’ killing could be a source of regret. Orderic, for example, records how Gilbert de Laigle was surprised without his armour at Moulins la Marche and pursued by a group of enemy knights ‘wishing to take him alive’, but was accidentally killed by a lance thrown by one of his pursuers during the chase ‘to the great grief even of those who had done the deed’. Geoffrey, count of Mortagne, ‘considering that his men had committed a serious crime and had sown the seeds of terrible troubles for his land by murdering such a warlike baron, made peace with his nephew, Gilbert de Laigle, and gave him his daughter Juliana in marriage’. 
Indeed, it was such ties of kinship and marriage which existed between many of the Anglo-Norman aristocracy that accounted for much of the restraint from killing seen in battles waged within the context of civil war. At Tinchebrai, fought in 1106 between Henry I and his brother Robert, there were very few noble casualties, while at Bourgthéroulde in 1124, the archers of Henry I’s familia regis were ordered only to shoot the horses of their rebel opponents, who were all swiftly unhorsed and captured without any fatalities. A similar ploy of shooting the horses, but not the riders of the enemy cavalry may well have occurred at Brémule, where Henry I’s army faced a mixed force of Norman rebels and Louis VI’s French, and where, as we have seen, casualties were negligible. The second battle of Lincoln in 1217 certainly witnessed a deliberate limitation on the lethal potential of the crossbow, for here the royalist commander Fawkes de Bréauté ordered his crossbowmen only slay the warhorses of their French and baronial opponents, while the royalist forces hung back from pursuit to give friends and relatives a chance to escape. Once again, the battle was virtually bloodless, and the death of one of the leading French nobles, the count of Perche, was a cause of consternation and regret on both sides. The stress on magnanimity towards a defeated knightly opponent, moreover, was beginning to be reflected in courtly vernacular literature, and is nowhere better expressed that in Chretien de Troyes’ Conte du Graal. When Perceval’s tutor in arms, Gornemant of Gohort, girds him with the sword-belt of knighthood, he then instructs him; ‘Young man, remember that if you are ever compelled to go into combat with any knight, I want to beg one thing of you; if you gain the upper hand and he is no longer able to defend himself or hold out against you, you must grant him mercy rather than killing him outright’.