2

THE CONQUEST OF NORMANDY, 1200–1204

The treaty of Le Goulet should have ushered in a period of relative stability between the kingdoms of England and France. For Philip it granted space to sort out his problems with the Papacy over his bigamous marriage to Agnès of Méran; for John it meant time to consolidate his inheritance of the English crown. Both countries were in a position to benefit from the increase in trade that peace would surely bring to them; this was particularly important to John who would have wished to recoup much of the huge sum of 20,000 marks that he had agreed to pay his overlord, Philip, by terms of the treaty, in return for formal recognition of his continental fiefs. John visited these fiefs on a comprehensive tour between June and August of 1200. From Dieppe in Normandy to St Sever in Gascony, he was accompanied by his army in a formidable display of power, designed to impress upon the many rebellious factions within his empire that he possessed the means and determination to force his will upon his widespread dominions. As a further incentive to cooperation, John took hostages as guarantees of good behaviour.

Whatever success this ‘triumphal progress’¹⁰¹ might have achieved in its tacit aims of intimidation, it was utterly nullified by one of John’s greatest political blunders: his crass insensitivity to Hugh le Brun and the Lusignans. This episode reveals how John’s acute political awareness was, as was so often the case, squandered by his hopeless inability to manage people or inspire them to place their confidence in him as their lord. Discarding his first wife, Isabella of Gloucester, with an ease that must have been the envy of Philip of France, John determined to marry another Isabella, the twelve-year-old daughter and heiress of Count Ademar of Angoulême, claimant to the country of La Marche. He wed her in late August and whisked her back to England. This swiftness of events was in a large part prompted by the uncomfortable fact that Isabella had been betrothed to John’s vassal, Hugh le Brun, Count of La Marche and head of the most influential baronial family in lower Poitou. John’s action was one of a series of offences he committed against the Lusignans (he would argue these were in response to slights against him), but the elopement with Isabella proved the most instrumental in the decisive struggle that was to follow. His motives for marrying Isabella had little to do with lust – although this cannot be entirely discountenanced, as nudging contemporaries pruriently suggested – and even less to do with romance, as some chroniclers and historians have also suggested; instead they were generated by a keen grasp of geo-politics. John wished to prevent Hugh’s marriage to Isabella as the concomitant territorial control would not only have created a physical barrier to his communications between Poitiers and Bordeaux along the network of Roman roads, it would have also dangerously empowered the already refractory Angoulême lords. There was also the major consideration that the existing power structures in Aquitaine might be overthrown: John was fearful that if Hugh combined possession of the counties of Angoulême and La Marche with the lordship of Lusignan, then any shift in his allegiance away from the English crown to Philip of France would cut Aquitaine in half and make it virtually impossible to hold.¹⁰² The affair might have ended favourably for John had he suitably recompensed Hugh for his loss of face; instead he fixed on a vindictive course which added insult to injury: in place of making amends he offered trial by single combat (by champions of course); worse, in the spring of 1201, he invaded the county of La Marche, seized it in his new wife’s name and attacked the Norman county of Eu, which belonged to Hugh’s brother, Ralph. Daniel Power has written of the importance of this affair: ‘John’s military inadequacies alone do not explain’ his unfolding position; ‘equally significant were his deteriorating relations with the Lusignan family since 1200.’¹⁰³

John had feared a strengthened Lusignan family agitating under the aegis of the Capetians; now, by his own actions, he had forced a weakened but embittered family decisively into the French camp. Hugh and Ralph appealed to King Philip against John’s shabby treatment of them. Philip, as overlord to the Duke of Aquitaine, heard this appeal with some discomfort. His relations with the Papacy over his matrimonial problems were still at a sensitive stage, despite the lifting of a papal interdict on France, and John’s display of power on inheriting his continental lands in 1200 may well have served its purpose in earning Philip’s cautious respect. Furthermore, John had added to his war chest in 1201 when he summoned his troops to Portsmouth. There the expeditionary force was equipped for a campaign that John had arguably never intended to embark upon. Instead, his purpose in gathering this army was to appropriate the soldier’s campaign funds, being the money that they had brought with them to cover their expenses while serving the host: as one chronicler succinctly puts it, ‘He took from some of them the money they would have spent in his service and let them return home.’¹⁰⁴ The money thereby collected was used to employ 200 mercenaries in Normandy: one half under the command of William Marshal, the other under Roger de Lacy – two generals who were to play such a large part in the coming wars. But Philip’s own position was improved by a settlement of differences with Rome and by the death of Count Theobald of Champagne. This latter event meant a sudden windfall for the French crown: the Count’s heir was a minor, which allowed Philip, as overlord, temporarily to add the substantial resources of this great fief to those of the royal demesne.¹⁰⁵

The two kings entered into a period of seemingly successful negotiations, culminating in John’s stay in Paris at the end of June, where he was lavishly entertained by Philip. The treaty of Le Goulet was confirmed and John promised Philip to have the Lusignan matter settled in his court as Duke of Aquitaine. John’s troops, however, continued to harass Lusignan forts and to lay waste French territory in the Touraine. The Lusignans pressed Philip to act on their behalf, and by 1202 he was prepared to do so. John’s foot-dragging over fulfilling his promise of the previous June had incited Philip to demand, somewhat ambitiously, the surrender of the most important castles in Normandy – Falaise, Arques and Château Gaillard – as security for his word. John continued to ‘make his excuses’, as Gervase of Canterbury correctly identified them,¹⁰⁶ and by April, Philip, as feudal overlord, could not be seen to be waiting any longer. Nor did he wish to do so. When, despite having pledged as security the two small satellite castles of Château Gaillard (Tillières and Boutavant), John failed to heed a summons to the French royal court to respond to the charges of injustice raised against him, he was judged, in absentia, as a contumacious vassal and condemned to forfeit his lands held of Philip: his fiefs of Aquitaine, Poitou and Anjou were to revert to the French crown.

Mirebeau and Anjou

The fragile agreement of Le Goulet had already been violated by both Kings: John had renewed his support for his nephew Otto of Brunswick’s claim to the imperial throne; Philip had engaged his daughter Marie to Arthur of Brittany, thereby re-establishing ties with John’s enemy. But Philip’s dramatic judgement meant all-out war. Hostilities were immediately opened by French attacks on Boutavant and Tillières, both of which were taken and razed to the ground.¹⁰⁷ John set up his campaign headquarters at Pont l’Arche in the Seine Valley where he must have rued his change of fortune. He had lost valuable allies to the Fourth Crusade, upon which the Counts of Flanders, Blois and Perche had embarked; and the Counts of Toulouse and Boulogne were transferring their allegiances to Philip.

The French King swept through the north-eastern frontier, meeting with little resistance until he reached Gournay.¹⁰⁸ Philip’s strategy had been to isolate Gournay by first taking the castles in the Forest of Lions. With this achieved Gournay had only its formidable defences to rely upon. The castle was situated in marshland and was protected by three curtain walls, wide and deep moats and the river Epte. It lay under the control of a loyal Angevin officer by the name of Brandin and both he and the garrison were offered considerable rewards by John if they maintained a successful defence. Philip, however, whose military achievements lay primarily in his skill as a castle-breaker, rose to the challenge with great ingenuity. Seeing that the castle was all but impregnable to anything but a lengthy siege, he turned, as he so often did, to his engineers. He instructed them to break the dam wall of a large weir that lay farther up the river. The result was an inundation the sheer power and volume of which so compromised Gournay’s defences the garrison was compelled to surrender. William the Breton claimed the whole area looked like a sea. Philip rebuilt the defences and by mid-July had moved to Arques, which lies on the Varenne river and which protected the vital port of Dieppe (that Richard I had given to Walter, Archbishop of Rouen, in exchange for Andely and the construction there of Château Gaillard). John hoped to raise the siege here by cutting off French supplies: he intended to do this himself by land while his ships from the Cinque Ports did likewise at sea.¹⁰⁹

In the southern theatre of war Arthur led his Bretons and Poitevin allies up the Loire valley into the strategic nerve centre of the Angevin Empire. As a rival claimant to the throne of England, the teenaged Arthur (he was born in 1187), who had been brought up in the French court with Philip’s son, Prince Louis, was an obvious weapon in Philip’s armoury. The French King had knighted him and accepted his homage for Brittany, Anjou, Poitou, Maine and Touraine – on the provision that Arthur could seize them. Philip intended to keep Normandy for himself. Philip had furnished him with money and 200 elite knights and sent him to Poitou where his forces were augmented by the Lusignans, Savary de Mauléon and other barons, including feudal contingents from Berry and Bourges. In all, Arthur may have been at the head of 1000 men when he marched on the Castle of Mirebeau at the end of July. This castle, lying between Angers and Poitiers, was at that moment offering hospitality to his grandmother, and John’s mother, the redoubtable Eleanor of Aquitaine, now approaching her eighties but still a major player on the political scene. Despite Philip’s characteristic advice to proceed cautiously, the proud and headstrong Arthur was not inclined to miss this opportunity of bagging such a great prize. Eleanor managed to despatch an urgent letter to her son, begging him for his immediate assistance. John, already moving south, met her courier near Le Mans on 30 July. Prompted by William des Roches, the castellan of Chinon whom John had made Seneschal of Anjou following his alienation from Philip, he marched with truly remarkable speed to Mirebeau, which his troops reached within 48 hours, having covered a distance of some 80 miles.

The sources do not agree on the events at Mirebeau, but a detailed composite picture can be drawn up. The anonymous narrator of Béthune, the most complete and reliable of the sources, relates that the town of Mirebeau surrendered but the castle, to which the garrison had withdrawn, remained defiant. Arthur requested Eleanor to leave the castle; she expressed her indignant surprise at the affrontery of his actions. Arthur’s force billeted in the town and settled down for a siege, unaware of John’s rapid approach. Early in the morning of 1 August, the English king’s army under the lead of William des Roches, burst upon the besiegers. The startled look-outs sent up the cry of ‘To arms! To arms!’ Ralph of Coggleshall claims that all the town’s gates except one had been secured; the Anonymous confirms that the Poitevins had been unable to close this gate. It was presumably through this poorly defended entrance that William and his troops stormed into the town. Once in, they fought to open the other gates. The element of surprise was total and had been used to the fullest advantage by the royalist forces. Geoffrey de Lusignan, we are told, did not stir himself from his breakfast dish of pigeons; if this were true, he must have mistakenly considered his defences secure enough to deter a precipitous assault. Others were not so confident. Hugh le Brun and his brother Ralph mounted their horses and rushed to the gates where they were met by the sight of William des Roches’ men breaking through. Royalist troops cascaded along the streets of Mirebeau, converging on the town centre. In the ensuing combat, des Roches is recorded as having three horses killed beneath him. We are led to believe, somewhat improbably, that even John entered the thick of the mêlée that erupted throughout the town. The Poitevins were completely routed. No one of any consequence escaped. William the Breton, forever making excuses for the defeats suffered by the French and their allies, claims that John’s soldiers had made a cowardly and, by implication, unchivalrous night attack that offered Arthur’s forces no chance of a spirited resistance. Relying on the effects of the day’s labours and drink to put the besiegers into a deep sleep, William depicts the royalist troops creeping furtively into the town and overcoming their opponents who were still in their beds (as if this somehow places the French troops in a better light). Roger of Wendover’s account differs slightly again and should not be entirely dismissed: his Flores Historiarum becomes contemporaneous around this time. In his version, the besiegers left the town ‘in pompous array’ to meet the oncoming Angevin troops. Both sides drew up in battle order and engaged with each other. The royalists gained the upper hand and Arthur’s force withdrew hastily to the town; but they were pursued so closely by the royalist cavalry they were unable to close the gates behind them (other references remark on an unsecured gate) and both sets of belligerents entered the town together. Although none of these sources mention it, it is possible that at the crucial stage of the battle some of Eleanor’s garrison in the castle sallied forth to aid the relief army, thereby catching the besiegers in the middle of two hostile onslaughts. Whatever the exact details, John’s decisive response to the situation at Mirebeau had earned him a great victory.¹¹⁰

John’s swift reaction to the threat faced by his mother reveals his ability to act rapidly in a crisis. His forced march to raise the siege offers a good example of the need for a military commander to act quickly and decisively. However, this in itself was not enough: the efficacy and use of such rapid movement was equally important. John might easily have rushed headlong into an ambush; we might suppose that William des Roches provided him with good intelligence of the enemy’s disposition. Philip Augustus had once been caught by acting over-zealously in a military situation and inadvertently hurled himself into a dangerous ambush – although William the Breton unconvincingly claims that this was both understandable and excusable given Philip’s unrestrained martial vigour and eagerness for the fray.¹¹¹ John would also have recalled how Richard failed to lift the siege of Aumâle in 1196: his attempt to surprise the French camp after a forced march floundered because of the well-entrenched and thoroughly prepared defences of the besiegers.

Although the Omanian school of thought on medieval warfare has been discounted, we can see how its thinking may have developed when we encounter such instances of spontaneous reactions by medieval generals. The idea that soldiers, and knights in particular, had only to get a whiff of the enemy to charge headlong into battle is exaggerated; indeed, medieval commanders placed great emphasis on battle avoidance. Chroniclers, especially those favourable to the subject of their attentions, liked to stress the resoluteness of the commander who responded immediately and boldly to any danger. The real skill lay in knowing when to act quickly and when not to act at all: some military actions were undertaken with the express purpose of provoking the enemy into taking steps that were to their ultimate disadvantage (a major strategy of the 1215–17 war in England). John’s response to Mirebeau was appropriate and vindicated by the hugely successful outcome. He was well aware of the benefits that speed could bestow upon a commander. His father, Henry II, said while campaigning in France: ‘Many castles, farms and cities lie exposed to us which we can easily overrun by a forced march.’ His brother Richard, the epitome of energetic generalship, characteristically commented: ‘To those who are well prepared, delay has always been and always will be dangerous.’ John would have also remembered the great effectiveness of Philip’s speed in raising the siege of the important Castle of Vaudreuil in 1194. William the Breton was astonished by this remarkable forced march:

I am amazed

That he [Philip] could, like a giant, complete an eight day march in three;

And who could not be astounded that this king, with his troops,

Fully armed, travelling as if with wings rather than feet,

Could make so many days’ march in so short a time?

What runner or pilgrim with winged feet,

Having fulfilled a vow and wishing to return home,

Can boast of having ever similarly covered

One hundred and fifty miles in three days?

It is interesting to compare the rate of this march (150 miles in three days) with John’s (80 miles in two days); unsurprisingly, both impressed contemporary observers.¹¹²

John had done well. He appreciated the urgency to raise the siege before it succeeded through storming, mining, bombardment or the arrival of reinforcements; Mirebeau would be hard to win back if the castle fell into enemy hands. As John was to find to his cost later, the loss of an important stronghold could prompt a rapid realignment of alliances and defections to the side deemed to be gaining the upper hand in the contest. John was justly elated by his victory – he did not have too many – and expressed his joy in a letter sent to England telling of his ‘happy success’ in seizing over 200 prisoners. Among the captives were Arthur himself (seized by William de Braose), Geoffrey de Lusignan, Hugh le Brun, Andrew de Chauvigny, Raymond de Thouars and Savary de Mauléon. No wonder he crowed that ‘he had got the lot’; Warren assesses that ‘not until Crécy were English arms to gain so resounding a success.’¹¹³ If John had deliberately provoked the Lusignans into revolt through his marriage to Isabella, it had seemingly worked out brilliantly. The shock-waves hit King Philip at Arques, where his siege machines had been pounding the town’s defences for over a fortnight. He immediately raised camp, abandoned the investiture, and force-marched his troops to Tours, but arrived too late to salvage anything from this heavy defeat. With nothing to be done but to assimilate the new political and military situation into a new strategy, he returned to Paris. William Marshal pursued the French army, hoping to inflict telling damage during its retreat; but the French kept their discipline and withdrew in good order and did not expose any weaknesses that could be exploited by the harassment of the Angevin soldiers. This pursuit was limited, being curtailed by effective and well-executed counter-measures put into operation by Philip. As he withdrew, he left the Norman borders in flames, sparing neither churches nor monasteries. It was one of only two major victories over Philip.¹¹⁴

At one blow John had become master of events. Arthur of Brittany and Geoffrey de Lusignan were incarcerated in the mighty fortress of Falaise; Hugh le Brun was thrown into the donjon of Caen Castle. Many of the other prisoners were shipped to Corfe Castle in England, where a dramatic postscript to events occurred. Amongst these prisoners was the romantic adventurer Savary de Mauléon, who led an attempted prison break-out. Having apparently made four guards drunk, he took possession of the keep, which then had to be invested by English troops. Tellingly, 22 of the prisoners starved to death rather than surrender, an indication of the harshness of the conditions of their captivity. Savary, through the mediation of the Archbishop of Canterbury, Hubert Walter (who was sympathetic to Arthur’s cause), came to terms with John: his wife and mother were amongst the hostages he had to pledge for his future loyal service.¹¹⁵ The one remaining possible focus of revolt, Viscount Guy de Limoges, also fell into John’s hands by September. Not only had the opposition leadership been removed; with their capture came the enormous military assets of many of their castles. But John failed miserably to capitalise on his extraordinary good fortune. With no lessons learned, once again his egregious mishandling and poisonous mistrust of his most powerful subjects had disastrous consequences. If indeed John could not resist kicking a man when he was down, nor could he retain his balance when putting the boot in; the result was usually John ending up on the floor.

It was, ultimately, John’s treatment of his most illustrious captive, Arthur of Brittany, that caused him to squander the aces in his hand. William des Roches, whose invaluable assistance at Mirebeau was given on the understanding that he would have a say in Arthur’s fate, went unheeded when he pleaded for Arthur’s release. John had, instead, moved Arthur from Falaise to Rouen, capital of Normandy. Dark rumours soon spread concerning Arthur’s fate and his suspected murder; speculation abounded in much the same way as would in the similar case of Richard III and the princes in the tower, when another English monarch stood accused of slaying his nephews, but again without conclusive evidence. A story of Ralph of Coggeshall relates how John had attempted to blind and castrate Arthur (to prevent any heirs laying claim to the throne), only to be prevented by the king’s chamberlain, Hubert de Burgh. It has been suggested that Arthur actually died from shock after being castrated. By Easter of 1203 it was widely believed that John had done away with Arthur. Two contemporary writers offer circumstantial evidence for Arthur’s murder. The Annals of Margam claim that John, drunk one night after dinner, killed his nephew with his own hands and threw his body, attached to a heavy stone, into the Seine. Though dramatic, this is a serious account: Margam Abbey had for its patron William de Braose, who was with John and party to events at this time, which may have been a reason for de Braose’s downfall. William the Breton paints an even blacker picture of John, who coldly murders Arthur after taking him out alone in a boat at night, plunging his sword into his stomach, and then rowing three miles in darkness on the Seine before dumping the corpse overboard. However, it is highly unlikely that John was possessed of enough courage to risk being alone with Arthur, or even capable of rowing three miles. But the agreement on the use of the Seine for the purpose of disposing Arthur’s body does lend some verisimilitude to this aspect of the accounts. Later assertions that Arthur had died of self-pity or in trying to escape by swimming across the Seine have done little to alleviate suspicions. Whatever the real story, it is highly probable that John did have Arthur slain; certainly, this is what the Bretons believed.¹¹⁶

It was John’s treatment of young Arthur that prompted William des Roches’s defection to Philip Augustus; he took Viscount Aimery de Thouars with him to the French side. The Bretons responded to the rumours of Arthur’s murder by going on the warpath; they took Angers while William and his rebel allies took control over most of the Loire’s counties, blocking the movements of John’s agents between Le Mans and Chinon. The citadel at Tours managed to hold out for John until 1204: abandoned by des Roches’s man, Hamelin de Roorta, Tours was placed under the command of Brandin, one of John’s mercenary captains; it was subsequently lost and partially regained. Loches and Chinon, under the mercenary captain Gerard d’Athée and Hubert de Burgh respectively, resisted until 1205. Vital as these strongholds were, they were isolated pockets of resistance in expanding enemy territory. John had even to send a mercenary force under Peter des Préaux to rescue his queen from Chinon. Chinon was a great administrative centre (David Carpenter has suggested that the loss of records was very significant) and an Angevin treasury; its fall was a deep psychological blow for John. This grim situation was entirely of his own doing. William des Roches was the most vacillating of vassals but, as Powicke observes, his actions were those of intelligent ambition. Had John kept him loyal he would have retained this vital region, which would have been ‘the most important guarantee against the loss of Normandy’.¹¹⁷

His desertion marked a trend. The great majority of the Poitevin nobility had relations or friends taken prisoner by John at Mirebeau, and his cruel treatment of his captives lost him any residual goodwill. John’s talent for alienating those he needed had come to the fore. Fearful of being bogged down in Anjou, John sent out peace feelers to Philip; unsurprisingly, Philip rejected these. English sources blamed the disaster that was to follow on the mass desertions of the continental baronage, but with more skilled and energetic military leadership, John may have minimised even this great wave of treachery that was now unleashed; instead, he succumbed to fatalistic lethargy, allowing des Roches and his allies further successes in Anjou. The defections gathered momentum. Hugh de Gournay went over to Philip and delivered up John’s castle of Montfort, having let in French troops at night.¹¹⁸ Count Robert of Alençon triggered further tergiversations with his change of allegiance to Philip; by 1203, John was bereft of important allies north of the Loire. But it was the military blows that struck the hardest. John’s castle at Vaudreuil had been stocked up with provisions and strengthened against the French King’s obvious designs on Normandy. Philip prepared for a lengthy siege of this strategic fort but, to everyone’s amazement, its English commanders, Robert Fitzwalter and Saer de Quincy, capitulated without a fight. Vaudreuil served two major purposes: an offensive one as a launching pad for expeditions into the Vexin; and a defensive one as an important satellite protection to Château Gaillard, itself the chief bulwark for Rouen and hence Normandy. With Vaudreuil and now also Conches in his hands, and the counties of the Loire increasingly under allied control, Philip was able to focus on the great but formidable prize of Normandy.

The marches of Normandy were the most heavily fortified areas of France. John was relying on their defences to absorb anything that Philip could throw at them. Normandy was of vital importance to both England and France. The French crown wished for access to the northern seaboard and for control of the mouth of the Seine, not least because Paris, the seat of the French government, lay up river and was vulnerable to attack. Normandy offered a different prospect to Philip than the regions of the Loire. Although there had been substantial ‘continental drift’ between England and Normandy, much Norman sentiment remained pro-English: many barons, including William Marshal, held territory both there and in England; the kingdom and the duchy were also bound by strong trading and commercial interests. The network of castle defences; the resources in men and money; the ease with which reinforcements could be sent unhindered across the Channel – all pointed to the belief that an energetic resistance would ensure Normandy’s safety.¹¹⁹But Philip had probed deep into the eastern regions of the duchy, and John was left placing his hopes on the garrison at Château Gaillard. Normandy’s fate hung on its defence.

The Siege of Château Gaillard

The siege of Château Gaillard was the great set-piece of the struggle for Normandy. The castle itself was justly renowned by contemporaries as a marvellous feat of engineering. Sited on a dramatic crag overlooking the confluence of the rivers Gambon and, crucially, the Seine, this castle was the prized personal project of Richard the Lionheart. He had endured an interdiction on Normandy to build his ‘beautiful castle on the rock’, his ‘saucy castle’ as it is often called (’bold’ or even ‘hardy’ castle may be more accurate translations). It was constructed in one startling phase between 1196 and 1198, a remarkably short time. Richard spent more than £11,500 on it, more than on all his castles in England during his entire reign, an amount that came to more than twice Normandy’s total annual revenues. The castle was the heart of its own complex defence system that included the island fort of Île d’Andely, the new walled town of Petit-Andely, the fortifications of the original town of Andely, a stockade and outlying forts. It represented the apogee of castle-building and techniques, and it was with good reason that the writers of the age considered it impregnable. It was here that one of the most dramatic sieges – arguably the most dramatic siege – of the entire Middle Ages was carried out between the end of September, 1203 and March 1204.¹²⁰

The place of Château Gaillard in history has been immortalised by William the Breton’s hugely detailed, and therefore invaluable, accounts. He was an eyewitness to most of the events that occurred there and this is reflected in the space he devotes to it; other chroniclers, in comparison, barely mention the siege in more than passing.¹²¹ Despite the epic qualities of the siege and its enormous consequence for the course of English history, its details have been largely overlooked by writers outside France, the last major published accounts of any note in English being those by Kate Norgate in 1887 and 1902, with my contributions on one aspect of the siege from the last few years.¹²² What follows is based closely, but by no means exclusively, on William the Breton’s description of events, especially as laid down in his Philippidos (a source which requires some careful handling and which, due to its abstruse nature, has not been used by historians as much as William’s more straightforward chronicle).

Having descended down the Loire by boat, taking Saumur and Loudon along the way, King Philip of France returned to Normandy to invest Château Gaillard. He began by isolating it. With Vaudreuil already in his hands, in August he besieged another satellite fort of Radpont. This resisted for three weeks before surrendering in late September; a substantial garrison of 20 knights, 100 sergeants and 30 crossbowmen was taken prisoner. Philip then rested and reinforced his men before fully investing the defence complex at Château Gaillard. The flow of events was going his way. His soldiers had raised John’s siege of Alençon and regarrisoned Tours; William des Roches and Philip’s mercenary captain Cadoc had taken possession of Angers; by April 1203 Le Mans was in Philip’s hands; his allies were making advances throughout Anjou. Pushed out of Anjou and Touraine, John resorted to ravaging Brittany, possibly to draw enemy forces away from the pincer formation they had developed. But he achieved nothing. His army and resources were needed on two fronts, in the south-west and the north-east; it was in the north-east that Philip’s grip was hurting most.

The French army marched up to Petit-Andely and pitched camp on the bend of the river opposite the town. Philip, ever cautious and realistic, knew his objective was both daring and ambitious, and prepared for a lengthy siege. The castle was in the charge of Roger de Lacy, the Constable of Chester. Roger of Wendover describes this seasoned and ruthless soldier as ‘noble and warlike’ (both Richard and John rated him highly as a commander) and this was not just the view of the English chroniclers; the heroic defence that he and his garrison were to make become widely recognised. Roger was an Englishman with no landed interests in, or attachments to, Normandy; he stood to benefit only from the English king’s fortunes and was therefore completely loyal to John. Indeed, in 1199 de Lacy had hesitated in his support of John’s accession to the throne and had consequently fallen under the king’s suspicion; here was his chance to prove himself in his lord’s eyes. His first defensive measure was to destroy the bridge joining the left bank to the island, thereby denying the French both crossing and easy communication to the facing banks. Philip therefore had need to construct his own bridge, but this process was hampered by the defensive stockade across the Seine, which obstructed the transport of materials necessary for the bridge’s construction. Philip’s initial objective was to remove this obstacle.

For this task, he relied on what proved to be the first of a series of extraordinarily brave acts performed during the siege. While Philip’s siege machines and artillery gave covering fire to keep down the heads of the defenders (and hopefully at the same time inflict some useful damage on the defences), a group of young Frenchmen, possibly led by Galbert de Mantes, swam under a hail of fire to the stockade and hewed at it with axes. Given the river’s strong currents, the weight of the axes and the exertions of swimming and axe-wielding, it is likely that the fatalities suffered by this group were not only due to the barrage of missiles raining down on them from the castle, but were also due to drowning. Eventually they hacked a gap wide enough to allow large boats through. The passage thus created, though distinctly uncomfortable for those who had to pass through it under the eyes of the castle, was essential not only for the construction of the new bridge but also for the supply of the camp, including livestock and provisions, needed to sustain a long siege. It was Philip’s intention to starve the garrison out.

Philip immediately set about organising the building of the bridge. For this he assembled a number of broad, flat-bottomed barges and ferries that worked the Seine transporting commercial goods and livestock along and across it. These he had bound together to form a pontoon bridge. Strengthened by stakes, the bridge was able to support the erection of two tall, strong towers on four particularly large boats. These towers served the dual purpose of defending the bridge and directing arrows on the enemy walls. The bridge was a substantial piece in itself, and an early indication of the scale of engineering works necessarily brought to bear on the Angevin defences. With the bridge now complete, Philip led the greater part of his forces over the water and tightened the siege of the town. William the Breton, in a suitably overblown reference, likens Philip to Xerxes, the Persian King who famously crossed the Hellespont by similar means in the fifth century. The French pitched another camp and attacked the island fortress from both sides. Having fully bottled up the besieged, and having captured Château Gaillard’s satellite forts, the French were free to roam the Vexin at will, foraging and plundering the region’s fields and dwelling places to the extent that feeding his huge army provided no overwhelming logistical problems for Philip. The French King had clearly understood the need to neutralise not only the direct military threat posed by the satellite garrisons but also their capacity to hamper the essential task of foraging and provisioning to meet the insatiable demands of such a numerous besieging force. Philip, ever the master of poliorcetics, paid as much attention to the no-less important but prosaic details as to the grander, more militaristic ones. No doubt the booty from these operations kept the troops happy too and helped maintain morale: well fed and with opportunities to line their pockets, the French soldiers would have enjoyed higher spirits than their hemmed-in English counterparts; but they were lulled into a false sense of security.

John, never far from Rouen, was roused from his ineffectual meanderings into one of his periodical bursts of intelligent and focused activity. In the early phases of the Normandy campaign, he treated the flood of ill-tidings about the military situation with seeming equanimity. Roger of Wendover reports that John’s reaction to being told that the King of France had entered his territory, taken many of his castles and dragged off their commanders to prison ignominiously bound to the tails of horses, was to reply nonchalantly: ‘Let him be; whatever he takes from me now I will one day recover.’¹²³ Now, however, at the very end of August while staying in Rouen, he proposed a daring plan to raise the siege at Château Gaillard. No doubt hoping to emulate his spectacular success at Mirebeau, John devised a scheme involving a night attack on the French lines. In this plan hatched by the King and William Marshal, two English divisions were to make their way by land and river to Andelys. In the combined operation the land force would fall upon the camp while over 70 transport vessels, laden with provisions pillaged from the Channel Islands and protected by a flotilla of small war ships, would re-supply the garrison on the Isle and break the pontoon bridge. Secondary sources have too readily accepted the numbers William the Breton provides for this relief force. He puts the figures of the Angevin land troops at 300 knights, 3000 mounted sergeants and 4000 infantry, plus a mercenary force under the infamous routier captain Lupescar; the naval element is said to have consisted of pirates under their commander, Alan, and 3000 Fleming soldiers. A total of perhaps 11,000 men is highly improbable: one modern estimate calculates that the French King had only around 2300–2600 paid troops on the Norman marches in 1203–4, and of these only 250 were knights, the rest being made up of infantry and crossbowmen.¹²⁴ William inflated the figures to create a sense of great odds pitted against the French. Nevertheless, it is clear that this was a major operation and one designed at least to match the French in military capacity. On receiving John’s orders, William Marshal immediately put the plan into motion.

Whether through over-confidence and poor guard detailing by the French or through the skill of William Marshal, the English land force approached undetected to the main camp on the left bank peninsula shortly before dawn. But the plan went wrong from the start: the fleet was late for its part in the operation. John’s strategy had been a sound one, and clearly caught the French off guard, but perhaps it had been conceived too hastily. He and his advisers had overlooked the crucial factor of when the Seine could be best navigated upstream that night; consequently, the flotilla was delayed in its journey against the tide. The Marshal, probably unaware of the flotilla’s problems, would not in any case have wished to hold back his attack: if he waited he would lose the advantage of the cover of the night and in the meantime increase the risk of being discovered by the French, a certainty as daylight approached. The better and greater part of the French army had crossed to the other bank; Philip had left only a smaller contingent at the camp. Attached to the siege camp were the non-combatants: ‘merchants, ribaldi and scroungers, and all those who march to follow the army camps in order to sell all sorts of things’, writes William. These lay drunk and in a deep sleep in the fields outside the camp (which was almost certainly fortified). At the appointed hour, the Marshal launched his onslaught. His men rushed on the camp. The element of surprise was total. The camp erupted in terrified panic as the alarm went up. All who could, fled to the pontoon bridge in a frenzied effort to escape to safety on the other side and away from the slashing and hewing of English swords and axes. Over 200 were cut down. Such was the weight of numbers on the bridge it partially collapsed: it could not sustain the panicked rush of so many. Already one English objective had nearly been achieved.

William des Barres, the great French general whom William the Breton labels ‘the flower of chivalry’, and who was everything William wished his King to be, halted this precipitous flight. He admonished the fleeing troops and organised his own company of soldiers for a counter-attack. He had torches lit to deny the English the cover of darkness and ordered rapid makeshift repairs to the bridge. When temporary repairs had been put into effect, des Barres’ banners led the French back across the bridge and headlong into a bloody engagement with the English. The English force could not hold the bank or camp against this onslaught and were soon routed; they suffered many casualties, dead and wounded, and a large number were taken prisoner. William Marshal made good his own escape. It is unsurprising that the Marshal’s biographer, who offers copious details of the great warrior’s military career, omits this serious defeat in his writings.

Just as the French were recovering from the exhausting encounter and celebrating their victory, they were faced with the second wave of the English attack: the flotilla had arrived. But it was late. The misreading of tides ensured that the counter-attack had lost its synchronicity. As dawn broke, the English ships could be seen making their way up-river to the isle. Once more the cry of ‘To arms!’ went up. French troops lined both banks and the pontoon bridge; crossbowmen were ordered to take up positions on the bridge’s towers. The bridge was afforded the greatest protection; it was the priority of attackers and defenders alike; on it stood des Barres, Simon de Montfort (of Albigensian Crusade fame) and the elite of the French troops. Alan and his sailors displayed extraordinary courage and determination in pressing on into the midst of the enemy that were arrayed on three sides, especially as the Marshal’s forces were in retreat. His flotilla had been discovered sooner than he had counted on and the Marshal’s assault had obviously alerted the French to the presence of danger. As his ships drew near the bridge they were met with a shower of arrows, javelins, stones and other missiles. Their position in the middle of the river afforded some protection and they doggedly held their course. They reached the bridge with a crash of timbers and a downpour of arrows, crossbow bolts, sling-shots, stones, logs, pieces of iron, boiling pitch and tar let loose by the French lining the bridge and manning the towers. Persevering bravely against this barrage, the English began to attack the fabric of the bridge and to strike at the cables, stakes and boats which held together the beams of the bridge. A deadly hand-to-hand mêlée ensued. William the Breton relates the violence of the combat almost gratuitously, but he captures the horror of medieval combat. Blows by the sword inflict terrible injuries with deadly efficiency: eyes, hands, feet and ears are lost by many victims. Throats are cut. Stones crush skulls. Axes shatter knee caps and clubs spatter brains. One man is engulfed in boiling tar and another sees his intestines hanging from his stomach. It is only when a huge beam of oak is toppled on two boats, crashing into them and holing them that the deadly encounter took a decisive turn. The boats manoeuvred about and employing their oars, retreated down the river, still incurring heavy losses. They were pursued by some young French sailors under the command of Jean le Noir and Galbert de Mantes, the latter proving himself to be one of the heroes of the siege. These caught up with two English vessels and captured them with the complements of crew, soldiers and booty. The rest of the attack flotilla fled to Rouen. John had failed to lift the siege of Château Gaillard.

The hearts of the garrison trapped in the island fortress must have sunk as they watched the disaster unfold before their eyes and as the flotilla withdrew into the distance. The English offensive had concentrated French minds and shaken them from their complacency; reminded of their own vulnerability, they intensified their action against the island. The besieged had placed some hope in a wooden palisade before the fort as an extra line of defence. It was here the French now turned their attention. Once again Galbert was instrumental in events. Some incendiary devices were prepared and secured in water-tight containers. These were attached by ropes to the waists of a group of men under Galbert’s lead.¹²⁵ This group swam unobserved (it is not clear whether what follows occurred in the day or at night) to the far, eastern side of the island, which was not so closely guarded: the defenders believed that the castle above them extended security to this part of the wall and had accordingly intensified their manpower on those points most under direct pressure. When in range, the French swimmers hurled their incendiaries at the palisade, which immediately caught fire. The breeze whipped up the flames and carried them into the fort itself. Soon all was engulfed in the conflagration and the whole fort went up in smoke. The shouts of triumph from the French looking on from the river banks were matched by the cries of horror from the isle. As the garrison attempted to seek safety from the blaze, its already critical plight was worsened by the barrage of stones and arrows that continued to assail them from the siege towers of the French. A few managed to escape in boats, but many were overcome by fumes as they cowered in the fort’s vaults. The palisade had proved to be not their saviour, but their nemesis. The French took to their boats and crossed with ease to the island, seized the survivors and made themselves master of the place. King Philip’s men had won him two spectacular early victories: the repulsion of the English relief force and the taking of the Îsle d’Andely, a vital step to taking the castle itself. For John in Rouen, the news was grim indeed.

As William the Breton, who had such a keen military eye, said: ‘The island fortress having been taken, it was easy to take the town.’ This was something the townspeople of Petit-Andely had anticipated: witnessing the fall of the fort, they fled headlong en masseinto the opened gates of Château Gaillard, placing their trust in the castle’s walls now that their town’s main defence and garrison had been lost. For many it was to be a fatal mistake; for many more the short path to the castle led them to unimaginable and unforeseen horrors. To prevent the empty town and fort from being retaken and in order to press the siege of the castle more closely, Philip filled them with his own people. These comprised not only the garrison needed to guard the place and to assist in the siege, but also new inhabitants – settlers – who, through the spoils of war, now took possession of the fugitive’s homes (Edward III did the same thing at Calais in 1347 after winning an eleven-month siege there, and crusaders cast out the defeated townspeople of Carcassonne in 1213). The two forces to whom this task fell were both mercenary elites: one company under a certain Walter; the other under the famous captain Cadoc, Philip’s counterpart to Mercadier, Richard the Lionheart’s mercenary captain, and now John’s. Philip placed great trust in Cadoc; so much so that his company received the lucrative sum of 1000 French pounds daily. With the town and fort occupied, Philip’s next task was familiar to a commander who had taken a strongpoint by force and intended to keep it rather than destroy it: the repair of the damage done (especially to the bridge) and a general refortification and reprovisioning.

It was at this point, with the castle securely invested, that Philip left to personally conduct the siege of Radpont, which fell in less than three weeks. John, still smarting from his recent defeat, did nothing for this important satellite castle, but instead withdrew to Falaise and Mortain, possibly with the intention of recruiting fresh troops. Radpont’s garrison had made a few spirited sorties, probably prompted by dwindling supplies, but to no avail. Philip wanted to lead his army onto Rouen, but this was not practicable in military terms while the castle of Château Gaillard continued to cast its formidable shadow over French ambitions, tying down vast resources of the French kingdom. With its soldiers occupying the banks, the island fort and the town, the castle’s garrison was easily contained; but an investiture of Rouen, with its reputation of impregnability, would be a lengthy, and hence hazardous process. Château Gaillard had to be taken first.

Patiently and realistically settling down for a long siege, Philip appreciated that bombardment and storming were unlikely to produce results on their own. And so he resolved to starve the besieged out. He set in motion a series of monumental engineering works. Huge trenches of circumvallation and contravallation were excavated the length of the siege force’s perimeters. These were designed to prevent sorties from the castle and to defend the French siege camp from any further outside relief attempts. One leading historian of the fall of Normandy has cast doubt on these trenches being excavated, but overlooks clear contemporary evidence for this.¹²⁶ Seven wooden forts (brestaches) protected both trench lines and were placed equidistant from each other; each had its own moat and drawbridge and was of considerable strength. All were filled with soldiers. Within the greatly fortified camp the French army, soldiers and camp followers alike, prepared for winter by building wooden and thatched huts to replace their tents. They had created their own small town. Here they were to live – and some die – until the spring. The harsh winter that followed bore witness to perhaps the most tragic episode in all the decades of Angevin–Capetian conflict.

The Useless Mouths

At some unspecified time early on during the siege, the castle’s commander, Roger de Lacy, the Constable of Chester, regretted having opened his gates to the inhabitants of the captured town, Petit–Andely, whose number had been swollen by an influx from the surrounding region, all of whom had sought refuge in the castle from the French forces.¹²⁷ He realised that with so many mouths to feed his garrison, which had little hope of fresh supplies, would be unable to sustain the siege for long. By itself, however, the garrison had sufficient stores to sustain it for a year. On one occasion, William the Breton informs us that many thousands of non-combatants had taken shelter within the castle’s precincts, but this is clearly an inflated number; the exact figure is hard to ascertain: an upper estimate of 2200 and a lower one of 1400 can be calculated. In order to preserve the stores that would be quickly exhausted by these ‘useless mouths’¹²⁸ de Lacy evicted some 500 of the oldest and weakest amongst them. The French took pity on the feeble group that emerged from the castle gates and allowed them to pass safely through their lines. A few days later a similar number were turned out from the castle and, being equally incapable of aggression, they were also permitted to cross the besieger’s lines.

King Philip was absent from the siege at this stage, away campaigning and organising help for his Breton allies who were being troubled by John in an attempt to draw the French away from Château Gaillard. When he heard of this lenient treatment of the refugees, he immediately sent orders forbidding the safe passage of any further non-combatants expelled from the castle, regardless of their age or wealth; and he also ordered that any further groups were to be driven back to the fortress. He was keen to ensure that as many people as possible whittled down the garrison’s supplies. De Lacy resolved to send out the remaining townspeople, judging his victuals as sufficient for twelve months if only the garrison and most able men remained in the fortress; all the rest were cast out. The number of this last group of men, woman and children was possibly as high as 1200, but was probably nearer 400; it comprised the weakest and most vulnerable of the refugees. William wrote that de Lacy knew he was sending them to certain death.

What followed was a full-scale tragedy to match any of the worst horrors of the Middle Ages. When this ragged and disorderly band spilled from the castle they believed that they were going to rejoin their families and fellow townsfolk in safety; but instead they were to be subjected to warfare in its most vicious form. They were not met by the opening lines of the besieging forces, but by a hail of arrows and javelins. The French had followed their new orders. The refugees flew back to the castle only to find the gates locked and bolted against them. The guard responded to their pitiful entreaties to allow them back into safety with the words: ‘I do not know you; go and search for shelter elsewhere: it is forbidden to open the gates to you.’ With this the soldiers on the battlements hurled down stones and fired arrows onto the terrified masses huddled below. Racked with fear and panicking from the incomprehensible violence of their erstwhile protectors, the wretched crowd withdrew from the foot of the castle walls, and moved out of reach of the missiles and into no man’s land between the besiegers and the besieged. There on the rocky slopes beneath the castle, they were to remain for three winter months.

Finding what precious little cover they could on the steep, barren and rocky gradients around the fortress, the people tried to protect themselves from both the elements and the artillery barrages by harbouring in the shallow fissures and clefts in the rock face. These offered little shelter from the wet and cold of winter. William was astonished by the harshness of those within the fortress: it was not surprising that the enemy should prevent the group passing through their own lines; but that they should be condemned to ‘a wretched and miserable existence’ by the inhumanity of their own friends and relations was truly shocking to him. These poor people had sought refuge for themselves and their goods in the strongest castle in Christendom, only to be ejected and abandoned by their own side. All they had with which to sustain themselves were wild herbs, rarely to be found in winter, and the waters of the nearby river.

William catalogues the horrors that these people were driven to by deprivation and starvation. A chicken that had come out of the castle was fought over by the strongest of them and consumed with its feathers, bones, eggs and excrement. A baby that had just been delivered was snatched from its mother by ravenous men who ripped up the infant and devoured the parts. The starving wretches feasted on the dogs that de Lacy had kicked out of the castle. He did not want these animals to consume even the scraps of the garrison’s food, and he was probably moved by a belated compassion for the plight of the people whom he had defended at the start of the siege. According to William these dogs were skinned with bare hands, although use was undoubtedly made of any arrows that had fallen into no man’s land. When these dogs had been eaten, so too were their skins. William wrote that all feelings of shame were suppressed in the fight for survival. For three long, bitter cold months the desperate souls were tormented by hunger, existing in a nether-world where many ‘neither lived nor died; being unable to hold onto life, they could not quite lose it’. In fact, at least half perished from starvation and hunger. But for the waters of the Seine, all would have died.

King Philip returned to the siege in February 1204 and witnessed the horrible results of his earlier orders. When the emaciated survivors saw the resplendent sight of the French monarch, given to corpulence and regally attired, crossing the bridge to the island fort, they called out for mercy and for a release from their slow death. William, always ready to praise Philip, rejoices in the king’s clemency when he commands that the wretches be released and fed. William witnessed amongst the ragged band a man still resolutely clutching onto a dog tail. When told to throw the tail away the man refused: ‘I shall only part with this tail that has kept me alive for so long when I am full on bread.’ There was, however, further tragedy in the ending to this episode, which adds veracity to William’s eye-witness account: deprived for so long of proper food and nourishment, most of the survivors gorged themselves on their first real meal with fatal consequences. Some historians have sensed hyperbole in all this. However, even in medieval times, it was known by some people, the writer Wofram von Eschenbach incuded, that it was dangerous for starved people ‘to gorge on empty stomachs’. At Belrepeire, his hero Parzival, having raised the siege there, fed its famished inhabitants while being aware of this danger, ‘so he gave them enough food and no more’. Numerous attested accounts from history reveal an authenticity to William’s writing at Château Gaillard.¹²⁹ With this pitiful chapter of the siege concluded, the more familiar business of war was resumed.

Historians have had relatively little to say on these events, yet they lay bare the commonplace reality of warfare rather than the glamorous but rare spectacle of battle: the non-combatant was as much at risk as the knight, and frequently perhaps even more so. William’s graphic account may give the impression of being artificial and sensationalist, but it actually bears close similarities with other chroniclers’ harrowing descriptions of different sieges. The situation at Château Gaillard was repeated twice during the Hundred Years War. At the siege of Calais in 1346–7 Edward III allowed the 1700 poor sent out by the garrison to pass through his lines, even giving them food and money. But then he sealed the investment and left 500 to die from appalling hardships in no-man’s-land. Those that were left behind suffered agonies from hunger. A message for help from the besieged was intercepted by the English: it read that dogs, rats and cats were being eaten, and nothing remained but to eat each other. Henry V’s siege of Rouen during 1418–19 was almost an exact repetition of Château Gaillard, similarly lasting six months over Christmas. The besieged were soon short of food. The eyewitness chronicler John Page informs us:

They ate up dogs, they ate up cats,

They ate up mice, horses and rats.

Here a rat cost 30 pence, a mouse sixpence; those in the town were fortunate enough to add vegetable peelings and dock roots to their meagre fare. Those expelled from the town and denied access through the besieging lines of the English endured the winter starving in the ditch surrounding the walls. Some 5000 were said to have died of hunger or from exposure in the winter downpours. Starving children, their parents dead, begged for food; an infant was seen trying to suckle from its dead mother; there were ten or twelve dead to everyone alive; and there were rumours of cannibalism.¹³⁰ The heart-rending scenes depicted by William the Breton at Château Gaillard were repeated throughout the Middle Ages on various scales.

The motivation for Philip’s response to the outcasts at the siege was clear enough: he was using hunger as a weapon. All the sources attest to the presumed impregnability of the fortress, and Philip believed that it could only be taken by blockade and starvation. As William, a seasoned observer of war wrote, ‘it is cruel hunger alone that conquers the invincible.’ The chronicler and French royal biographer Rigord declared that Philip’s intention was to take the castle by ‘hunger and want…in order to spare the blood of men’. But this is either naïve or disingenuous, attempting it as it does to portray a compassionate king: the unmitigated suffering of the non-combatants involved little actual blood-letting. Such a policy of starvation took its toll at Château Gaillard. The English chroniclers believed it to be the principal reason for the fall of the castle: Ralph of Coggeshall wrote of the garrison’s shortage of victuals; Roger of Wendover thought that Roger de Lacy made a gallant but doomed sortie so that he might die by arms rather than of hunger.¹³¹The role of hunger as a significant military weapon will be seen again later when we investigate the siege of Rochester in 1215; here we shall limit ourselves to focusing on why it was that the non-combatants were made to suffer so terribly.

From Philip’s point of view, the situation was straightforward. First, by forbidding the outcasts to make the short walk to safety he hoped to oblige the garrison to take them back into the castle and so exhaust the garrison’s stores many times faster than would otherwise have been the case. This was especially desirable as we may assume that these people brought foodstuffs with them into the fortress, thereby unwittingly providing more victuals for the soldiers there and increasing their ability to hold out even longer. Starving the garrison out by the burden of extra numbers would facilitate an early surrender at relatively little cost in blood and treasure for the French. Secondly, showing mercy meant revealing a potentially fatal weakness. If King Philip let ‘the useless mouths’ through his lines then other towns and castles would also expel their non-combatants whenever Philip might besiege them; in this way their stores would last longer and the siege would be protracted, probably without success for the besieging forces. Philip Augustus did not earn his reputation as a castle-breaker by displaying any such weakness. Finally, the fate of those outside the walls reminded those within the castle of what awaited them should the blockade run its full course: the victims offered the besiegers an opportunity to apply some psychological pressure.

Roger de Lacy’s motivations on the English side are more complex and, as this episode has been largely neglected by historians, never fully examined. Why should such a renowned and respected warrior (one chronicler describes him as ‘vir magnificus et bellicosus’) resort to such cruel tactics against his own people? For a medieval commander to be successful required a degree of ruthlessness that affronts modern sensibilities. Obviously, de Lacy wanted to preserve his food supplies long enough to maintain a lengthy resistance; and such was the strength of the castle under his command he could hope to hold out as long as his stores did. Those that were not of use to the castle’s defence were therefore thrown out. To have accepted them back in would have been as much a sign of weakness as any leniency that Philip might have displayed; it would also have utterly defeated the military objective of defending the fortress – and according to the anonymous writer of Béthune, de Lacy had vowed that he would only come out of the castle dragged by his heels.¹³² De Lacy has been blamed for his ‘terrible blunder’ in allowing the townsfolk to enter the castle in the first place, only to force them out later. But this overlooks the basic feudal obligation of a lord to protect his vassals. A more practical concern was that if a lord failed in his duty, his people might actively seek another lord they deemed more capable of protecting them: the thirteenth-century Schwabenspiegel asserts ‘We should serve our lords for they protect us; if not, justice does not oblige us to serve them.’¹³³ Furthermore, at the time when he accepted the refugees into the castle, it was more than probable that a fortress as important as Château Gaillard would be quickly relieved. We have already seen how King John had sent out an expeditionary force to raise the siege, and it was only after its failure that de Lacy sent out the non-combatants.

Historians have made little of a letter sent by King John to de Lacy, and have not made any possible connection between it and de Lacy’s expulsion of the non-combatants. This might be of something of an oversight as the letter clearly implies that no more help would be forthcoming, and that de Lacy must look to his own devices:

We thank you for your good and faithful service, and desire that, as much as in you lies, you will persevere in the fidelity and homage which you owe to us; that you may receive a worthy reward of praise from God and from ourself, and from all who know your faithfulness. If however – which God forbid! – you should find yourself in such straits that you can hold out no longer, then do whatsoever our trusty and well-beloved Peter des Préaux, William of Mortimer, and Hugh of Wells our clerk, shall bid you in our name.

The letter, though undated, seems to have been written in early November in nearby Rouen, capital of Normandy.¹³⁴ It is possible that de Lacy received this letter before removing the townsfolk from the fortress, and perhaps only took this extreme measure when he knew the garrison had been left to fend for itself. He did have the option of surrender at this point, and the letter appears to accept such a possibility. But the letter is ambiguous on this score: John does, after all, express his wish for de Lacy to ‘persevere’. The charge of holding a lord’s castle was a heavy one, especially when the castle was as important as Château Gaillard and the lord was King John. Medieval chronicles bear witness to the fate of the commanders who were too ready to capitulate. The Duke of Norfolk wrote in 1453: ‘it has been seen in many realms and many lordships that for the loss of towns or castle without siege, the captains that have lost them have been dead and beheaded, and their goods lost.’¹³⁵

William the Breton and some modern French historians have condemned de Lacy for his ‘cruel and pitiless decision’ in expelling the refugees, but to do so ignores the exigencies of war and the situation in which the castellan found himself. If such large numbers were kept within the castle precincts – unfed, but at least with the benefit of shelter from the winter – they may have risen up against the garrison who denied them food and overwhelmed them by sheer weight of numbers.¹³⁶ They may also have been driven away from the foot of the castle walls to minimise the chances of the outcasts effectively doing the work of the enemy through covering a surprise night attack by scaling the walls. Lowly ribaldi, the non-professional element in armies, had achieved something similar when they scaled the walls of Tours in 1189. It is interesting to note that only after their three months of deprivation did Philip judge this group as ‘unable to harm anyone’; but William the Breton had drawn the same conclusion when they were first ejected, describing them as ‘inutile bello’. At least by sending the people out de Lacy could hope that there was a chance the French would allow them to pass through as they had done with the first two groups.

Some French historians have equally been too quick to follow William the Breton in praising Philip’s show of mercy when, on his return to the siege, he was so moved by the plight of the outcasts he relented on his earlier instructions and ordered that they be freed and fed. William makes the most of this humanitarian gesture of the French king, who was ‘always responsive to supplicants, because he was born to have compassion for unfortunates and to always spare them’; he was ‘moved by the lamentations … of those who had already suffered too much’. Philip Augustus was moved probably as much by military considerations as by charitable ones: he feared ‘the not improbable outbreak of a pestilence which might easily spread to his own entrenchments’.¹³⁷ This was an ever present fear at siege camps, where numbers could be decimated by the spread of disease. In the siege camp at Avignon in 1226, ‘There arose from the corpses of the men and horses which were dying in all directions, a number of large, black flies, which made their way inside the tents, pavilions and awnings, and affected the provisions and drink, and being unable to drive them away from their cups and plates, caused sudden death amongst them.’¹³⁸ It was during this siege that Roger of Wendover suggests that Philip’s own son, Louis VIII, died from the dysentery that had spread through the besieging camp. More important to the French King than the alleviation of the suffering of the non-combatants was the need to keep his forces at full strength to press home the siege.

The End of the Siege

Philip may have simply wanted the refugees out of the way for an assault on the castle. Throughout the winter the castle had been bombarded and assaulted, but with few tangible results. Spring was imminent and with it could be expected a renewed military campaign by John. John had slunk back to England in early December, having dismantled a number of castles which he feared Philip might make use of on his advance to Rouen. These were not the actions of an inspiring military leader. Shortly after Christmas the King of England was at Oxford where he organised the granting of war subsidies from both the laity and the clergy. This was the beginning of a concerted effort to raise war finances: tallages were levied on towns; privileges and concessions went on the market; goods in ports were taxed at a fifteenth. Scutage was also collected: a number of barons were fearful of losing their lands in Normandy and were therefore prepared to pay ‘shield-money’ in lieu of providing an actual military presence on campaign.¹³⁹Philip had probably heard of these preparations. Further delay incurred mounting dangers and left unfinished business in other territories, especially in Poitou; Philip could not afford to have the best part of his army indefinitely tied down at one siege, no matter how important. John’s inactivity over the last three months was not going to last. By the end of February 1204, Philip was making ready to storm the castle.

Another phase of major engineering works now took place. Philip concentrated his forces around the narrow stretch of land that reached out to the castle rock. This was the only accessible approach to the castle, a fact reflected in the intricate construction of the castle itself; its design was focused to counter any attack from this direction. By massive military landscaping, the whole area before the castle was flattened and widened, an enormous task which involved the breaking-up and removal of rocks. The purpose was two-fold: to provide a quicker and safer approach to the castle’s outer defences; and, most importantly, to facilitate the movement of siege machines – ballistae, mangonels, trebuchets – closer to the castle walls. Protected runs allowed soldiers to fill the ditch before the salient tower with various materials, thereby creating firm ground for the belfries to threaten the walls; but it remained highly dangerous work. It was clear to the defenders what the French had in mind, and they directed their fire on these manoeuvres, inflicting heavy casualties. Nevertheless, the task of filling the moat was continued, its importance and dangers understood by those taking it. At the siege of Acre, directed by Richard the Lionheart and King Philip during the Third Crusade just over a decade earlier, a story circulated of a mortally wounded woman asking for her body to be thrown into the town’s moat to fill it and to help with the attack.¹⁴⁰ The crusading spirit was not so prevalent at Château Gaillard; but the exercise and its purpose remained the same, ensuring the loss of many lives in its accomplishment. The defenders were equipped with their own mangonels and ballistae, which they used with varying effect. The protected path to the ditch had enabled at least one very large belfry to be drawn up to the walls. From this huge wooden tower, crossbowmen and archers aimed their bolts and arrows down on the garrison with deadly results, easing the plight of the French working to fill the moat.

Events suddenly speeded up. Before the moat was half-full, the French descended ladders into the moat and then set them up on the other side. The ladders, however, were too short to be of much use and the soldiers had to resort to using their daggers and swords to assist them in clambering up the rock to the base of the walls. These men were terribly exposed to the missiles falling down on them from above; covering fire from the belfry together with the shields strapped to their backs and held above their heads afforded them some protection. In the siege tower the crossbowmen and archers of Périgas Blondel demonstrated why they were so highly regarded, and so well rewarded, by King Philip. The air whistled with sounds of bolts, arrows and slingshots from both sides. Philip, exhorting his men onwards from the front ranks by the edge of the ditch, exposed himself to the dangers of the fierce combat and was struck several times (if we are to believe William the Breton), but his shield and armour protected him from serious injury. However, it is unlikely that Philip had forgotten Richard I’s fate five years earlier and probably orchestrated the assault from a reasonable distance.

At the foot of the salient tower which formed the apex of the two angled forward walls, the French used picks to dig out a hole at the wall’s base large enough to provide them with shelter from the stones and arrows cascading down on them. From here Philip’s miners went to work. The contemporary chronicler Robert of Auxerre identifies the importance of miners to the French king: they operated like moles and followed Philip everywhere.¹⁴¹ These miners are a prime consideration in Philip’s reputation as a castle-breaker; they had gained much experience from services to their employer, especially at successful sieges such as Le Mans in 1189 and Boves in 1185. As these sappers mined under the foundations of the tower they supported the roof over their heads by means of sturdy timber pieces, cut from tree trunks. They excavated the mine in relative safety; the defenders had not yet organised an effective counter-mine to break in on the French tunnel and halt their progress. The miners piled up some tinder, lit it and hastily withdrew from the mine as the timbers burned. They got out just before the props went and the roof caved in. With its foundations eaten away and all the underpinning gone, half of the tower came crashing down into the moat, filling the air with dust. Not only did this leave a breach through which the French could storm; the masonry rubble from the tower also helped to fill the moat further and made a passage across it easier. Faced with the immediate prospect of storming, Roger de Lacy ordered his men to torch the buildings of this outer ward of the castle, not wishing to leave the enemy any useful materials or cover and also probably with the intention of buying some time for a retreat. This hastily done, the garrison manning this untenable sector withdrew quickly across the drawbridge to the castle’s middle bailey. Before the flames had died down or the smoke had cleared, the French, led by Cadoc, poured through the breach to take the outer ward. On the ruined remains of the tower he raised his banner.

The middle bailey was guarded by a ditch 30 foot wide and a strong wall; it was no less formidable than the first defences. The French were keen to keep up the momentum that they had just created and did not allow themselves to be daunted by this broad moat and staunch rampart. They immediately launched themselves on a follow-up attack. A group of sergeants scouted around the edge of the ditch in the standard practice of searching for a relative weak point in the defences. They found one. Almost inevitably, given his misfortunes in war, it had been created by King John. A year earlier he had a building added to the middle ward. William the Breton describes the building’s upper storey as consisting of a chapel and the lower one a latrine,’against religion’, he complains, appealing to spiritual outrage. At a shallower part of the ditch Peter Bogis and some comrades scrambled unseen down and across the moat until they were directly beneath the chapel window. Bogis, standing on the shoulders of a companion named Ralph, with considerable agility was able to reach the window and pull himself up. Remarkably the window had been left unbarred and Bogis climbed into the chapel. An alternative version puts forward the idea that Bogis, presumably of slight build, actually clambered up the latrine chute and opened up the window from the inside. Either account should not be dismissed: during the First World War, one of the most powerful fortresses in the world was taken in a very similar manner.¹⁴²

From the window Bogis lowered a rope to his men and pulled them up. Once inside, the band of Frenchmen began to break down the door to the interior of the ward. The noise of their efforts alerted the defenders who feared they were about to be subjected to a mass break-in by the enemy. As before, in their panic they decided to abandon this ward also, and so they fired it (although it is possible that they may have intended only to burn the building the French were in). Then they hurried back to the strongest part of the castle – the inner bailey which housed the keep. As the French troops outside the castle watched this drama, they feared that Bogis and his group had been engulfed by flames. They had in fact taken shelter in a vault used as an armoury and survived the blaze. To the undoubted cheers of their onlookers, Bogis and his men emerged from the smoke to cut the ropes of the drawbridge and lowered it to their fellow soldiers. The French troops rushed across.

The inner bailey was enclosed by a 500ft wall eight feet thick. It was less a wall than a series of seventeen D-shaped towers, or convex buttresses, providing tremendous strength and the capability for withering flanking fire. The keep itself, surrounded by its own ditch 50 feet across, was the same thickness as the wall. De Lacy was left with less than 180 men to defend it, a sufficient number under his experienced leadership, although fatigue would have been a further factor with which to contend. However, like the middle bailey, this part also suffered from a major design flaw. The gate-house, though well positioned at an awkward point for attackers, had no drawbridge; instead the ditch was spanned by a bridge hewn out of the rock itself. Thus the weakest part of the inner ward – its entrance – could not be protected by the wide ditch. That noted, any approach across this bridge might expect to encounter the deadliest barrage to repel it. The most intense action of the whole siege now led to its climax.

Philip sent forward a siege machine known as a cat. The cat, catus, also known as a sow, scrofa, was a reinforced roof on walls – in effect, a huge, mobile shield – under which soldiers and engineers could approach a stronghold’s walls to damage its fabric or to undermine it. The French mining operation met with effective resistance. Roger de Lacy now ordered a counter-mine to be excavated; this opened up on the French mine where the French miners were attacked and forced to retreat, their work only partially completed. By early March an enormous petraria, called ‘Chadaluba’, was brought to bear on the walls, discharging its artillery of large stone blocks against the ramparts. The wall, weakened both by the effects of Philip’s miners and, ironically, by de Lacy’s counter-mine, collapsed when struck by the petraria for the third time. The besiegers, seeing the creation of this breach, hurled themselves forward. They scrambled up and over the tumbled masonry and stormed the breach in force.

Despite the hopelessness of the situation, de Lacy and his men fought on. Even now they disdained what would have been the most honourable of surrenders. However, overwhelmed by sheer weight of numbers, they were soon under physical restraint and led away in chains. They were lucky not to have been slaughtered on the spot: the laws of war placed the fate of victims of a successful storming in the hands of the victor. According to the anonymous chronicler of Béthune, de Lacy met with a relatively lenient, if ignominious treatment. Each day throughout the long siege, when called upon to yield up the castle, de Lacy defiantly replied that he would never surrender the fortress, even if he were dragged out by his feet. His heroic defence did indeed end in this dishonourable manner.¹⁴³ He and his garrison were shackled in irons. King Philip had finally taken Château Gaillard.

Both the Anonymous of Béthune and Roger of Wendover provide alternative endings to the siege. Both agree that the English were, at the last, starved out (Philip’s early strategy). The Anonymous states that the garrison capitulated after exhausting their food supply, which included their horses. Wendover claims that de Lacy and his men, ‘preferring to die in battle to being starved’, mounted their warhorses, rode bravely out of the castle in a daring sortie, and killed many of the enemy before they were eventually overcome, and then only with great difficulty. These two accounts offer plausible scenarios; but it must be considered that they cover the siege in only a few lines, and the details concern themselves with only the siege’s finale.¹⁴⁴ William the Breton, despite his faults, was an eyewitness to these dramatic events.

In the warm afterglow of victory following the brief outburst of triumphalism meted out to de Lacy, Philip afforded a genuinely magnanimous treatment to the enemy commander who was detained prisoner on parole in France. De Lacy was one of the few men King John placed consistent faith in: he contributed £1000 toward paying his ransom to secure his release from France; on his release he made him sheriff of Yorkshire and Cumberland, entrusting him to help defend the volatile north from Scottish incursions.¹⁴⁵John had recognised the great, if ultimately futile, service de Lacy had rendered the English crown by his lengthy defence of Château Gaillard; it was the king’s own fault not to have profited from de Lacy’s defence. By the time of the siege’s conclusion at the end of the first week in March, de Lacy had been left with some auxiliaries, 120 men-at-arms and 36 knights. Only four defending knights were killed during the siege.

For six months, Philip’s army, estimated at between 2300 and 2500 strong on the Normandy frontier at this time, was preoccupied with this siege; but John had failed to secure any advantages from this situation. Powicke ponders whether the garrison would have performed even better in a less ‘complicated’ castle; he questions whether the castle’s defences were too restrictive; if its ‘elaborate arrangements’ were ‘mutually injurious’; or, indeed, if it was too scientifically advanced for its defences to be fully understood and exploited by its garrison.¹⁴⁶ These are valid questions; however, de Lacy was an experienced commander and, over the course of the siege, would have studied, analysed and exhausted all of the castle’s defensive potential. What is curious is that such a justly renowned castle, rightly considered as the apogee of castle-building, should have, in John’s chapel and the bridge of rock, such serious design flaws in its innermost defences. Whatever the possibilities might have been, the English and French forces had played their part in perhaps the most epic siege of the entire Middle Ages. Its consequences were soon apparent to all.

The Collapse of Normandy

John heard the news of Château Gaillard’s fall on the day he was making his hunting arrangements for his return to the duchy. But John was never to return to Normandy as its duke. Philip immediately established his grip on the region by repairing the castle’s ruined defences so that they were, in the words of William the Breton, ‘more solid than they had ever been’. Given Philip’s eye for detail and his mastery of siege warfare, this is wholly credible. This was the castle that symbolised control of Normandy. The implications of its loss were grasped straightaway by the duchy’s inhabitants; within three months it was totally subjected to French rule. Philip, ever the cool customer, did not let success go to his head. His gaze was set on Rouen, the capital of Normandy, and which lay farther down the Seine.¹⁴⁷

Rouen prepared for the siege: resources and supplies poured in while the defences were made ready. The city’s strong walls and triple ditch made it a formidably strong objective (as Henry V discovered during his six-month-long investiture of it between 1418–19). Its Angevin loyalties were reinforced by those who had fled to the city before the Capetian advances. All the time negotiations prompted by the Papacy continued apace, but these were little more than diplomatic formalities and they achieved nothing. It was during these talks that William Marshal, in an act of realpolitik, paid homage to Philip for his Norman lands while still serving his English lord.

Contrary to what might have been expected, Philip did not make straight for the city but instead directed his army to the west: a strategy which was to reap great reward at minimal cost. John’s preparations for war, initiated at Oxford just after Christmas, were now tardily coming to fruition. If Rouen held out against a besieging French force then Philip could expect to face another relief expedition from England, this time larger than the one he had confronted at Château Gaillard and while deeper in hostile territory. Rather, he determined on the piecemeal territorial conquest of Normandy, leaving its capital isolated and increasingly surrounded, and hence reducing its strategical value. On 7 May he took Argentan and the satellite castles of his next objective, Caen. Intelligence gathered from Norman deserters revealed Falaise to be deeply demoralised, a condition not unsurprising giving Château Gaillard’s recent loss. The stronghold of Falaise withstood the French forces for only a week, despite the extensive defence excavations John had recently carried out there; Lupescar, its commander, bowed to the tide of events and went over to Philip, taking his mercenaries with him. It is possible that Lupescar’s actions were in part prompted by financial incentives offered by the townspeople of Falaise who did not wish him to resist the French king: they feared the material and personal damages that a siege would inflict on them, a telling effect of Philip’s ruthlessness at Château Gaillard. With Falaise gone, Caen soon followed without a fight; with it went Bayeux, Lisieux, Coutances, Barfleur, Cherbourg and lesser places. In the first week of May, Breton forces under Guy de Thouars, burning with revenge for Arthur’s murder, had broken through Normandy’s southwestern defences and destroyed Mont Saint-Michel, a task which had to be completed quickly as the tides that transformed the rock into an island gave them only limited opportunities to do so. They then moved on to sack Avranches, linked up with French forces and joined Philip at Caen. Here Philip divided his army into two. Guy de Thouars, the Count of Boulogne (erstwhile ally of John), William des Barres, a large body of French knights and the mercenaries who had changed sides at Falaise completed the subjugation of south-western Normandy where castle after castle fell in rapid succession, a sure sign of military and political momentum. Philip, meanwhile, now turned to Rouen.

Philip must be given full credit for his strategy. The successful sweep south west and link-up with the Bretons left, as Philip intended, the ducal capital exposed and friendless. The political and psychological effects of this were as important as the explicitly military ones. Prince Louis attempted a similar approach during his invasion of England in 1216–17 when he was confronted by Dover Castle: he temporarily abandoned the difficult siege of this castle to first subdue lesser ones around it. John, too, tried this for London in 1216. In eastern Normandy all that was left of John’s defences were Arques, Verneuil and Rouen itself; these decided to act in unison during the crisis. They saw clearly that the writing was on the wall in one form or another. The duchy’s records of government had already been shipped to England before the end of May; and the indecent haste with which Falaise and Caen had capitulated had done nothing to stiffen the resolve of the remaining Angevin pockets. They felt the acute vulnerability that Philip had designed for them.

At Rouen we witness another example of why Philip Augustus had such a reputation for the subjugation of strongholds. Force and logistical expertise were not the only ingredients of his poliorcetic success. Before the gates of Rouen, he played on the defendant’s fears, self-interest and common sense to come to an agreeable conclusion. Faced with the might of Philip’s conquering army, the leading figures of the capital’s community considered, as Falaise did, not only the mortal dangers they faced but also the financial threats that a siege would entail. Rouen’s civic and trading riches rested in large measure on its privileges and rights; if John could no longer guarantee these then Philip might instead; but he would only be predisposed to do so if the city submitted quietly without forcing him to expend French blood and treasure. Philip, ever cautious and calculating, had been alert to such sensibilities throughout the Normandy campaign and had confirmed the liberties of the places now under his control. This had happened at Falaise, the fate of which was so influential on decision making in Rouen. If resistance failed, the capital’s jealously guarded and coveted privileges could be revoked and instead granted to rivals.

Rouen’s military commander, John’s ‘trusty and well-beloved’ Peter des Préux, assessed the long-term hopelessness of the situation. He knew that only direct intervention from England could save the situation and harboured doubts that King John would provide the deus ex machina. Peter, who had enjoyed a close friendship with Richard the Lionheart, knew John’s character and had less faith in him than his predecessor; John, conversely, had relied heavily on Peter and his brothers over the years in keeping Normandy secure. Rouen could look to its own formidable defences for a while if it so wished, but the balance of gains and losses made this option an unattractive one. Des Préux and the castellans of Arques and Verneuil made a truce with Philip on 1 June, in which all three agreed to surrender if help was not forthcoming within 30 days. They wrote to John in supplication and could not have been overly surprised by the tone of John Softsword’s reply: they were not to expect any help from him and they were to act as they saw fit. This was not an exhortation to arms that would steel his subjects’ nerves to counter Philip’s threats, for the French King had added psychological intimidation to the incentives for surrender. Philip had explained to the citizens, governors and castellans of the towns and castles in his path that they should accept him as their lord now that John had abandoned them (not least by fleeing to England at Christmas). Wendover writes: ‘he begged them as a friend to receive him as their lord since they had no other; but he declared with an oath that if they did not do so willingly, and dared to vie against him, he would subdue them as enemies and hang them all on the gibbet or flay them alive.’¹⁴⁸ After Château Gaillard, few doubted this threat. Hostages had been offered to Philip and their safety, along with that of the city, was put before faith in John. On receiving John’s reply, de Préaux did not even wait for the truce to expire; on 24 June the strong fortresses of Arques, Verneuil and the capital of Normandy, considered invincible by many, surrendered to the French king. John had already lost Maine and Anjou. Now, too, Normandy was lost.

This was a momentous occasion in English history. Nicholas Vincent has recently summed up its significance: ‘it fatally tarnished the military and political reputation of King John and set the Plantagenet kings on the road to harsher taxation, failed attempts to recover the duchy and, in the aftermath of this failure, the breakdown in relations between kings and barons that resulted in Magna Carta. King John’s road from Normandy to Runnymede was a straight one.’ David Carpenter has judged that ‘The Capetian Conquest of Normandy was a turning point in European history. It made the Capetian kings dominant in western Europe, and ended the cross-Channel state.’¹⁴⁹ The French were in command of the northern seaboard. Paris, at times within just a day’s ride of English forces in the twelfth century, was now safe. And England lay under threat of invasion.