CHAPTER ONE

HUMBLE ORIGINS

THE STORY BEGINS nine hundred and sixty years ago. A monk was sitting in Canterbury, writing his chronicle of the year’s events. It was the year 1051 – and what a year it had been! A great struggle had taken place in the kingdom between two powerful factions. On the one side stood the king, Edward the Confessor, with his friends and allies. On the other stood Earl Godwin and his sons, the most powerful noble family in England. The question they were debating, with armies and swords at the ready, was of the highest importance. Who was going to be king after Edward died?

The monk set down these events in detail. At one point, however, he departed from his main account to report an incident that had taken place in distant Herefordshire. Some members of the king’s party – Frenchmen, if you don’t mind – had been given lands in that county, and had been getting up to some outrageous things.

‘The foreigners,’ wrote the monk, ‘committed all kinds of insults and oppressions on the men in that region.’ But that wasn’t the worst of it. What really surprised the monk was the thing that these foreigners had built.

It was a great mound of earth, topped with a large wooden tower, surrounded by an enclosure of wooden palisades. It was so new and so different that the monk didn’t even have a word of his own to describe it. In the end he had to settle for the word the foreigners themselves used, and called it a castle.

We know all this, of course, because the monk’s chronicle has survived. It’s one version of the famous Anglo-Saxon Chronicle, now kept in the Bodleian Library in Oxford. Sadly, this manuscript doesn’t tell us anything else about the monk – not even his name – other than the fact that he lived in Canterbury. But is is the first surviving document, written in English, to use the word ‘castle’, and the earth-and-timber structure that the monk was describing was the first castle to be built in England.

No one is exactly sure where this castle was. Most historians think that the monk was talking about the mound of earth at Ewyas Harold on Herefordshire’s border with Wales. Two other castles were built in the county at the same time, in Hereford itself and at Richard’s Castle, while a third was constructed at Clavering in Essex. None of them is much to look at today. They are overgrown with trees and bushes, and their wooden towers and walls have long since vanished. If you didn’t know what you were looking for, you’d probably never guess they were there at all. Yet these few mounds of earth are the earliest castles in England. None of them were built by Englishmen – they were all built by the French friends of Edward the Confessor.

Although he came from a long line of English kings, Edward had grown up a stranger to England. When he was about ten, in the year 1013, the country was invaded and conquered by the King of Denmark. Edward’s father, King Ethelred the Unready, gathered up his family and fled across the English Channel to France, where he sought refuge at the court of his brother-in-law, the duke of Normandy. It was in Normandy, living the life of an exile, that Edward grew to manhood.

For a long time, it looked as though he would remain in France forever. His father and older brothers made several attempts to win back the kingdom they had lost, but to no avail: one by one, they all died trying. But disaster overtook the Danish royal house just as quickly as it had engulfed Edward’s family. In 1035, the Danish king, Cnut, died. By 1042, his two sons had followed him to the grave. The way was suddenly clear for Edward to reclaim his inheritance. In 1043, with the consent and support of the English aristocracy, Edward found himself back in England, being crowned king.

His fortunes had improved no end, but after his accession Edward still had one major problem. In his bid for the throne he had been supported by Earl Godwin, an Englishman who had collaborated with the Danes and was now the greatest aristocrat in England. After Edward’s coronation, the two men cemented their alliance when the king married Godwin’s daughter Edith. But Edward had grave doubts about his new father-in-law, and one very good reason to dislike him – the earl, it was rumoured, had been involved in the murder of the king’s brother. After a few years, therefore, of ruling with Godwin by his side, Edward decided it was time to take action. He invited to England some of his old friends from France, and began appointing them to positions of power. In 1050, he made his nephew, Ralph of Mantes, Earl of the East Midlands; shortly afterwards, he appointed his Norman friend Robert of Jumièges as Archbishop of Canterbury. The king’s intention, it seems, was to create a counterweight to Godwin. By 1051, surrounded by his Continental supporters, Edward seems to have felt that he was powerful enough to take on the earl and his family.

That year, a major row erupted between the two men. The official cause of the dispute was petty – some local trouble in Godwin’s town of Dover. The more likely cause for disagreement, however, was the question of the succession. Despite seven years of marriage, Edward and Edith had produced no children. Godwin couldn’t be certain – and neither, of course, can we – but it seemed that his son-in-law was deliberately resisting his daughter’s charms, and spitefully frustrating any chance that there would one day be a little Godwin sitting on the English throne.

In 1051, Godwin’s worst suspicions were confirmed. In the summer of that year – or so it was later claimed – Edward promised the throne of England to his cousin, an energetic young man called Duke William of Normandy. This, it seems, was the real trigger for Godwin’s defiance. It was now absolutely clear to the earl that he and his family were being cheated of their inheritance. In September, the row boiled over and threatened to come to blows. Robert of Jumièges, the archbishop, accused Godwin of plotting Edward’s death. The king’s other French friends started building their castle at Ewyas Harold in anticipation of the coming storm. Both sides were squaring up ready for a fight, amassing hundreds of troops in their own territories. It looked, to everyone’s despair, as though England was about to be plunged into a civil war.

But then, at the last minute, Godwin and his sons sensed it was a struggle that they could not win, and fled the country. Edward, finally, was free – master in his own kingdom after years of ruling in the earl’s shadow. He set the seal on his victory by confiscating the lands of the Godwin family and giving them to his French allies. Tellingly, he banished his queen to a nunnery, and later that autumn William of Normandy paid a visit to the English court.

Edward’s victory, however, was short-lived. The following year, the Godwins returned, invading the country and demanding the restitution of their lands. Confronted with superior numbers, the king had no choice but to give in. His French friends, realizing that this time their defeat was inevitable, chose to run. Some of them went west, to the castle at Ewyas Harold. The archbishop headed east, and set sail for the Continent. Our Canterbury monk, who clearly despised the Frenchmen, reported their departure with undisguised glee, and laid all the blame for the dispute at their door.

‘Archbishop Robert was declared an outlaw unconditionally, together with all Frenchmen,’ he wrote, ‘for they had been mainly responsible for the discord which had arisen between Earl Godwin and the king.’

So by 1052, everything was back to normal. The Godwins had been restored to power. Edward had taken back his queen. No one, if they were wise, was saying anything more about Duke William of Normandy. It was as if the events of 1051 had never happened. There were no more arguments, no more Frenchmen, and no more of their new-fangled fortifications – these so-called ‘castles’. Everything in England was back as it should be.

And so it might have remained, had not Edward made his famous promise in 1051. It was a promise that meant that when the king died fifteen years later, the French would be back. No one could have guessed it at the time, but that castle in Herefordshire was the first drop of rain before the deluge. Within a generation of its construction, England would be filled with hundreds and hundreds of castles, from sea to sea.

But let’s not race too far ahead. Instead, let’s dwell for a moment on the events of 1051, and what they tell us about castles. One thing emerges very clearly: the French definitely had them, and the English definitely didn’t. The Canterbury monk was quite outraged to discover that there were people building a castle in his backyard. Castles were a French invention and, as far as people in England were concerned, the French could keep them. By the same token, Edward the Confessor’s Continental friends had shown themselves to be enthusiastic and experienced castle-builders. At the first sign of trouble, they had quickly constructed a castle, and they must have built the other early castles in England at around the same time. Had this been France, where people had been building castles for generations, no one would have blinked an eyelid. Constructing a huge mound of earth was simply what you did in such circumstances. In France, when the going got tough, the tough built castles.

This difference in attitudes might seem, on the face of it, quite strange. After all, here were two societies, both governed by warrior aristocracies, both at roughly the same level of economic development, and separated from each other by only a narrow strip of water. Yet their feelings and opinions about fortification were apparently quite divergent. So how had this divide come about?

The simple answer is: because of the Vikings. The Vikings, we used to believe, were the bad boys of medieval Europe, looting and pillaging with fire and sword long after everyone else on the Continent had calmed down a bit and taken up farming. Nowadays, of course, we are taught to see them differently. Economic migrants rather than shameless pirates, traders as much as raiders: the Vikings, it turns out, were not such a bad bunch after all. But whether the indigenous peoples who lived in northern England at the close of the eighth century saw the Vikings in such a rosy light remains open to question. The monks on the island of Lindisfarne, who in 793 encountered the first batch of new arrivals, might well have disagreed. In the century that followed, the Norsemen swept all before them. One by one, the several kingdoms that made up ninth-century England collapsed in the face of the Viking onslaught. The ancient kingdoms of Northumbria and East Anglia, and even the mighty Midland kingdom of Mercia, all eventually succumbed. By the 870s, only one Anglo-Saxon kingdom, the kingdom of Wessex, remained.

Wessex, however, fought back. The resistance was led by King Alfred (871–99), who for this reason, as well as for his legendary lack of culinary skills, became an English national hero. The king and his descendants protected their people by instituting a sophisticated programme of defences, which they called burhs, or boroughs. These were nothing less than planned towns, strongly fortified so as to protect large communities within their walls. In many towns in southern England, the outline of a burh can be still be identified, and in each case the total area enclosed is very similar, suggesting that burhs were built to something approaching a standard model. By building them, Alfred and his successors were able to push forward their frontier with astonishing speed and success. By 927, they had all but reversed the effect of the invasions; that year the Viking capital of York fell, and the power of the Viking leaders was broken. Many Scandinavian settlers, of course, remained in the northern and eastern parts of the country, but they were now ruled by the kings of Wessex – or, as they had begun to style themselves, the kings of England.

Indeed, by driving the Vikings back, the kings of Wessex created a country that, in territorial terms, was recognizably similar to modern England. Where formerly there had been a handful of competing English tribes, there was now a single, united English state. As states went in the Middle Ages, it was a mighty one. The kings of England enjoyed powers on a scale unrivalled by any other European rulers at the time. They issued one type of coin throughout their realm, and manipulated the currency for their own profit. Their laws and their government likewise extended to all parts of their kingdom. Most importantly, they restricted the building of fortifications. Burhs were public defences, maintained and owned by the king. Building a private fortification – like a castle – was not permitted. When Alfred’s descendant Athelstan took the city of York in 927, his first action was to destroy the stronghold that the Viking leader had built there. If you were a reasonably prosperous landowner in tenth- or early eleventh-century England, the most you could get away with was a small fortified homestead, confusingly also known as a burh, but sometimes called aburhgeat. Archaeological excavation suggests that these amounted to a collection of domestic buildings surrounded by an earthwork and a wooden stockade. In England, serious fortification was the business of the king, and the king alone.

On the other side of the Channel, however, it was a different story. Here, too, the Vikings attacked in the ninth century, sailing their longboats up the Seine in 854 and burning Paris. But whereas in England the Viking attacks ultimately brought unity, in France the end result was political fragmentation. The formerly strong kingdom created in the late eighth century by the famous Charlemagne crumbled away during the rule of his heirs. In France there was no national epic in the making, no hero in the mould of Alfred to lead resistance against the invaders. Instead of building communal fortifications under the direction of the king, powerful men began to take the matter of defence into their own hands – to protect themselves, their families and their households. In 864, the then King of France, Charles the Bald, watching his kingdom disintegrate before his very eyes, attempted to reverse the process with a royal proclamation.

‘We will and expressly command,’ he said, ‘that whoever at this time has made castles and fortifications and enclosures without our permission shall have them demolished.’

This is the first recorded use of the word ‘castle’ in French – almost two hundred years before it occurs in English. It is also an indication that the spread of private fortification in France had reached the extent where it was irreversible; the French king might as well have ordered back the sea.

Of course, the Vikings weren’t the only cause for castle-building. Castles might be necessary for defence, but they were also very useful for enforcing one’s right to rule over others. As royal authority began to disintegrate in France, all aspects of government – law-making and law-enforcement, tax-collecting and control of the coinage – began to fall into private hands. French society, in a word, was becoming feudalized, and the symbol of a lord’s feudal authority was his castle.

Next comes an elegant twist in the tale. The Vikings who raided France, like the ones who raided England, decided to stay for good. However, whereas in England the power of Norsemen was eventually broken, in France they just kept getting stronger. In 911, the French king recognized the authority of the Viking ruler Rollo, who had colonized a large chunk of territory in the North-West of his kingdom. The region became known as the land of the Norsemen, or Normans. The province of Normandy had been born.

The remarkable thing about the Normans was how quickly they shook off their Viking past, and how readily they adopted the ways of their more sophisticated French neighbours. Within a couple of generations they had started speaking French, and had embraced the Christian religion. Their leaders started experimenting with French titles, like ‘count’, and later, ‘duke’. They also adopted French ideas about fortification and defence – the very ideas that their not-too-distant Viking ancestors had inspired as a result of their initial raids. By the eleventh century at the very latest, the Normans were following the fashion of French lords, and building castles.

What did these early French and Norman castles look like? Unfortunately, in the case of the very early ones – the kind against which Charles the Bald tried to legislate in the ninth century – we have no idea. The earliest surviving castles date from over a hundred years later, and are to be found along the River Loire. In the small town of Langeais, for example, not far from the city of Tours, are the remains of a stone tower, built around the year 1000. Stone castles, however, were highly exceptional at such an early date. It was far, far more common in the tenth and eleventh centuries for castles to be built from earth and timber. Like the stone castles of later periods, these castles came in all shapes and sizes, depending on the needs and resources of the owner. At the simplest end of the spectrum, they might consist of timber buildings encircled by a ditch and an earthen rampart. However, while there was no single design, by the eleventh century something approaching a standard procedure had evolved. Looking back from the early twelfth century, a French clergyman remembered:

The richest and noblest men… have a practice, in order to protect themselves from their enemies, and… to subdue those weaker, of raising… an earthen mound of the greatest possible height, cutting a wide ditch around it, fortifying its upper edge with square timbers tied together as in a wall, creating towers around it and building inside a house or citadel that dominates the whole structure.

It was these huge mounds of earth that ultimately distinguished the strongly fortified, private defences of French lords from the comparatively weakly defended homes of their Anglo-Saxon counterparts. Contemporary authors, writing in Latin, described these great mounds as aggeres, but the popular thing to call them was ‘mottes’ – which is curious, because the word ‘motte’ itself seems to be Celtic in origin. Mottes were almost always accompanied by a much larger enclosure, known as a bailey, created by digging a ditch and making an earthen rampart. The two elements, taken together, produce the common-or-garden early castle – the classic ‘motte and bailey’ design.

The earthworks of a motte-and-bailey castle (Tomen y Rhodwydd in Wales).

Today, of course, all the timber parts of such castles have long since rotted away, leaving only the earthworks, and even these have been considerably diminished by erosion over the centuries. Working out what these buildings looked like when they were first constructed therefore requires considerable detective work. In the absence of surviving timbers, we have to look for other evidence. Of course, we have some descriptions of castles, like the one above. We also have one or two bits of pictorial evidence, of which the Bayeux Tapestry is by far the best. The famous Tapestry, commissioned shortly after 1066, not only describes the Norman Conquest of England; it also deals with events in France leading up to the invasion. In the first section of the story, we find some of the earliest and most detailed images of wooden castles.

The Bayeux Tapestry – castles at Dol, Rennes and Dinan.

Altogether, there are four French castles on the Bayeux Tapestry – three in Brittany (at Dol, Rennes, and Dinan) and one in Normandy (Bayeux itself). As you can see, when the artist who directed the work thought of a castle, he clearly pictured a motte in his head. Castle experts have poured over these images for many long hours, trying to work out what the artist’s highly stylized depictions actually represent. All the mottes are clearly covered in buildings, and the appearance of two crafty Norman knights at Dinan, attempting to set fire to the castle with burning torches, strongly suggests that these buildings are made of wood. In the case of three of the castles (Dol, Dinan and Bayeux), access to the top of the motte from the ground is made possible by a ‘flying bridge’, which has a shallower angle than the steep side of the motte itself. At Rennes, visitors seem to have made their way to the top of the motte using steps cut into the side – taking care, of course, to avoid the animals grazing on the slope. It also appears that we might be looking at gatehouses, both at the top of a flying bridge (Dol) and also at the bottom (Dinan).

The Bayeux Tapestry – the castle at Bayeux.

It is, however, the towers on top of the mottes that have generated the most interest and speculation. Are they one, two, or more storeys tall? All of them certainly look very different. The one at Dol is particularly difficult to interpret – what are the curvy things hanging off the left-hand side of the tower? Are they shields belonging to the defenders, or flames licking the side of the building? Nobody can say for sure. The picture of Dinan, which has the most activity, shows a garrison of half a dozen knights defending the castle during a siege. It seems clear enough that the top of the motte is protected by a wooden fence or palisade around its edge, but what about the tower at the centre? Is it a solid building, or is it raised up on stilts to make it even higher? One of the knights on the motte, readying himself to throw his javelin, seems to pass his arm behind a post supporting the building above. Is that the main entrance to the tower, shrunk out of all proportion, squeezed in against the top border of the tapestry? And what are we to make of the fancy tower on top of the motte at Bayeux, complete with what appears to be a domed roof, stepped gables and round-arched windows, as well as an elaborate entrance, decorated with a carved animal head? Are these depictions realistic, or just something the tapestry artist invented? Had he actually visited the castle at Bayeux, or had he just heard second-hand that it was a very impressive building? As you can see, the Bayeux Tapestry, wonderfully rich source that it is, raises as many questions as it answers.

Another problem with the Tapestry is that it offers us no information about the baileys of the castles it depicts. A bailey was a large area that housed all the buildings necessary for a medieval household – not just the lord and his immediate family, but also their domestic servants and a number of soldiers. There had to be accommodation enough for all these people, as well as a chapel to cater for their religious needs, and buildings for the storage of grain and tools. Most importantly, the bailey had to have a hall, so that the lord could sit down with his whole household and dine in public, and so that he could receive and entertain guests in style.

So the bailey is quite straightforward – it is simply an enclosure for all the buildings needed by a small private community. But what was the purpose of the motte, with its wooden tower on top? From some contemporary descriptions of mottes, it is clear that their towers could provide additional accommodation for the lord. A famous description exists of the twelfth-century wooden tower at the now vanished castle of Ardres in northern France. The chronicler, one Lambert of Ardres, describes at great length a magnificent three-storey building, with storerooms and chambers piled on top of one another. It contained just about everything the lord of Ardres could wish for – not just a great chamber for him and his lady, but private rooms for their servants, as well as a chapel, a kitchen, and numerous cellars, larders and smaller rooms. In many cases, however, the surviving earthworks are too small to have accommodated such a huge structure, and must have supported something rather more humble.

Part of the reason for building a motte, of course, was defence. By raising a tower well above the ground, the castle-owner gained an obvious advantage over any attacker. There is also, however, an element of showing off. According to that twelfth-century description, French lords built mottes not just to protect themselves, but also ‘to subdue those weaker’. By building a great earthen mound, and topping it with a big wooden tower, you were making a statement. It was not so much, ‘I’m a bit concerned about my own safety,’ as ‘I’m in charge here, and don’t you forget it.’

We can also begin to understand why mottes were a popular option. Most obviously, they were cheap to build, and the building materials – earth and clay – were at hand and plentiful. It took some time of course, and you needed to persuade a lot of peasants with strong backs to do the digging, but it wasn’t nearly as demanding or expensive as building in stone. When stone is hard to come by and peasants are ten a penny, building a motte makes good sense.

All of this, however, is fairly obvious, and none of it gets us any closer to the heart of the question. Why France, and why the eleventh century? We can look at a motte and understand the motives that prompted a person to build it, but at the same time, the reasons seem to be universal and timeless – the desire to protect oneself and one’s family, while simultaneously lording it over everybody else. There’s no apparent reason why the Normans should have built mottes, but not the Romans, the Celts or the Vikings. Clearly someone somewhere in northern France must have had a brainwave one day around the turn of the first millennium, and the idea caught on fast.

One reason for the sudden adoption of motte-building might have been the advances that the French aristocracy were making in mounted warfare at the time. The turn of the first millennium was the period when we see the emergence of a class of men who would dominate European society for the next five centuries – knights. If strong-armed men in mail shirts were starting to charge around the place on horseback, a big mound of earth could be interpreted as a counter-cavalry measure.

Certainly, there is an important relationship between cavalry and castles. Castles, it has been observed, work a bit like aircraft carriers; they might be big and impressive, but without their moving parts – the aircraft, or the horses – they are not much use. Setting out at dawn, the cavalry garrison could ride out on daily patrols, making their presence felt and striking at their enemies, before returning to the safety of the castle in the evening. For the same reason, I tend to think of early castles as being like old-fashioned US cavalry forts. Surrounded by timber stockades, overlooked by watchtowers, and home to cavalry garrisons, such forts have much in common with motte-and-bailey castles.

Their use of cavalry was one of the great differences between the way the French and the English made war in the eleventh century. The English, of course, had horses, but they did not ride them into battle, preferring to dismount and fight on foot. The French aristocracy, on the other hand, galloped into battle, armed with swords and javelins, and were perhaps starting to experiment with lances. The cross-Channel difference in opinion was made very clear to Ralph of Mantes, one of Edward the Confessor’s castle-building chums. When he tried to train Englishmen from his earldom in the French art of cavalry warfare, and led them against the Welsh at Hereford, the result was a military disaster.

‘Before a spear was thrown,’ sighed the author of the Anglo-Saxon Chronicle, ‘the English fled, because they had been made to fight on horseback.’

Castles and cavalry, then, were the two major differences between the English and French approaches to warfare in the middle of the eleventh century. The gulf, however, may have opened quite quickly and recently, especially in the case of the Normans and the English. The Normans, after all, had been Viking settlers at the start of the tenth century, and as such they were originally accustomed to travelling by longship and fighting on foot. Only in the course of the tenth century can they have adapted to fighting on horseback. In the case of castles, the gap between the two peoples had appeared even more recently. It used to be thought that every self-respecting Norman lord had a little castle of his own to call home, but recent research has shown this was far from the being the case. It is, in fact, very difficult to find rock-solid evidence that the Normans were building castles to a motte-and-bailey design before 1066. When it comes to establishing dates, archaeology relies on identifying disturbances in the soil; this becomes difficult when the thing you are excavating (in this case, a motte) is made up entirely of soil that has been disturbed. Fortunately, however, the reputation of the Normans as castle-builders appears to be safe. At several mottes that have been excavated, pottery and other small finds have suggested a construction date somewhere in the first half of the eleventh century.

Archaeology, therefore, has emphasized the fact that the majority of Norman castles are likely to date from a period only a generation or so before 1066. This discovery tallies well with what we know of the history of Normandy around this time. From its creation in 911 down to the early years of the eleventh century, the story of Normandy had been one of unmitigated success. From 1026, however, the duchy experienced twenty years of almost perpetual crisis. In that year, the old Duke of Normandy, Richard II, died after a long and successful rule of thirty years, leaving behind two sons by his first wife. The elder of the two sons, Richard, succeeded his father as duke, but only one year elapsed before he also dropped dead – murdered, some would later claim, by his younger brother and successor, Robert. Whether or not Robert was indeed guilty of his brother’s death, his rule was an unsuccessful one, which saw the leading nobles of Normandy appropriating local offices and powers that properly belonged to the duke himself. In 1035, things went from bad to worse when Robert set off on a pilgrimage to the Holy Land and never returned. When news reached Normandy that he had died on the way back home, many must have despaired – their new duke was Robert’s only son, a boy of eight years old, and a bastard. His name was William.

Little William, as we all know well, would grow up to become the most famous of all Norman dukes. In 1035, however, few would have put money on him living past his ninth birthday. Normandy was soon plunged into a state of civil war, and all the evidence suggests that it was in this period, and during the rule of William’s father, that the number of motte-and-bailey castles in the duchy began to shoot up. Until this time, castles had only been built by the duke and his most powerful supporters. Now they were being built by anyone who could lay their hands on enough materials and manpower to do so.

Tackling these new castles was the principal challenge for Duke William. His career as a young man reads as the story of one siege after another. Controlling the duchy became a matter of destroying the castles of his enemies, and building new ones of his own. After a successful battle against his greatest opponents in 1047, a Norman chronicler observed that the balance of power had been tipped in William’s favour.

‘All those magnates who had renounced their fealty to the duke,’ wrote the chronicler, ‘now bent their stiff necks to him as their lord. And so, with castles everywhere destroyed, none afterwards dared to show a rebellious heart against him.’

From that point on, William went from strength to strength. By the time he was in his late thirties, he could reflect with a great deal of satisfaction on his success. The dark days of his boyhood were far behind him; he was now respected and feared not just in Normandy, but throughout all of northern France. At the same time, however, he had not taken his eyes off a far bigger prize – the one that had been held out to him in 1051, only to be immediately snatched away. In 1065, the throne of England was once again uppermost in William’s thoughts.

On the other side of the Channel, things had been reasonably quiet since the dramatic events of 1051–52. After their triumphant return to England, the Godwin family had manoeuvred themselves into positions of power. Old Earl Godwin himself had died in 1053, but he left several healthy sons to succeed him. The eldest, Harold, had inherited his father’s position as Earl of Wessex, and his younger brothers had become Earls of Northumbria, East Anglia and Kent. By 1065, the Godwin boys were easily the most powerful force in English politics.

However, the sad contrast between the Godwin clan and the royal family was plain for all to see. King Edward the Confessor, now in his sixties, was clearly not going to produce a son to succeed him, and his brothers, of course, had died decades ago. Attempts to find a suitable candidate for the English throne were becoming increasingly desperate. A few years beforehand, the great men of England had sent messengers to find the king’s long-lost nephew, Edward the Exile, who for half a century had lived in Hungary. They managed to find him and ship him home; but he died the moment he set foot on English soil, leaving only a young son, Edgar, in his place.

With a lack of obvious strong candidates, the wolves were beginning to growl and snarl around England, sensing easy prey. The King of Denmark was known to be interested. So, too, was the King of Norway. Most worryingly, the Duke of Normandy had apparently not forgotten King Edward’s rash promise of 1051. Back then, no one had been very concerned about this. Duke William was a young man, with a few easy victories behind him, but a ruler barely able to control his own territories, never mind seriously threaten England. Now, though, in 1065, the duke looked considerably more menacing. He was undisputed master of northern France, and an experienced general with a reputation for brutality and success.

In the event, however, when Edward finally gave up the ghost in January 1066, it was Harold Godwinson, the man on the spot, who unexpectedly seized the moment. How long he had been plotting this move we don’t know. Certainly, his nomination by the dying Edward the Confessor can’t have come as a surprise – it was simply a useful piece of last-minute propaganda. For many years now Harold had been the power behind the throne, and he seems to have decided he might as well just sit on it himself and deal with the consequences. Of course, this meant that he had to push young Edgar out of the way first, but no one of any importance seemed particularly bothered about that. The choice was between a strong, powerful and experienced man with a weak claim, and an inexperienced child with a better one. With England threatened by other, much less appealing overseas contenders, ready to wage war in pursuit of their ambitions, most people probably thought that backing Harold was the wise choice.

And so it proved, for most of 1066. Throughout the summer, Harold showed what a capable leader he was, summoning and holding together a great army in readiness for the invasions that everybody now expected. When the King of Norway landed in September, Harold marched straight up to Yorkshire and won a famous victory. The Norwegians had arrived in three hundred ships, but they sailed home in just twenty-four. The more poetical English soldiers were probably already composing songs to their new king’s greatness when messengers arrived from the south, bringing the news that William of Normandy had landed with an army of seven thousand men.

Landing his ships at Pevensey on the morning of 29 September, William’s first concern was to establish a beachhead, and he did this by building a castle. Pevensey was the site of an old Roman fort, and William and the Normans proved adept at customizing such ancient sites. There is also the intriguing possibility, suggested by a twelfth-century chronicler, that the Normans brought this castle with them. The fact that this is only mentioned in a later source casts some doubt upon its veracity, but there is nothing inherently implausible in the idea of a flat-pack fortress. The Bayeux Tapestry shows the elaborate lengths to which the Normans went in preparing their invasion fleet – transporting barrels of wine, armour, weapons, and the like. Landing in hostile territory, they didn’t necessarily want to go scurrying around looking for suitable timber, and waste time cutting it to shape. We have at least one example in later centuries of an invading army taking a wooden castle with them ready to assemble when they landed. It seems quite possible, therefore, that the first castle built in England by William the Conqueror was a prefab.

The castle at Pevensey, and the second castle that the duke began further along the coast at Hastings, can be used to explain in part why Harold rushed headlong into battle with William. The new English king, as recent events had shown, was by no means a bad general, yet he plunged his exhausted army straight into battle at Hastings without pausing for breath. Why was he so hasty and intemperate? Historians have tended to conclude that Harold was responding to William’s provocation. For his part, William knew that his only hope of success was to draw his opponent into battle as quickly as possible; above all, he needed a decisive victory. Landing in Sussex made this somewhat easier – the county was part of Harold’s own earldom. William was deploying a tried-and-tested technique of medieval warfare: attack your enemy in his own back yard. Terrorize his tenants, burn his crops, slaughter his sheep and cattle. To act in this brutal way exposes the weakness of your opponent’s lordship, and underlines his inability to protect his own people. Castle-building, of course, fits perfectly into this catalogue of terror. One need only recall the words of the Canterbury monk, for whom the construction of a castle was associated with ‘insults, injuries and oppressions’. Forcing Harold’s tenants to build castles and burning them alive in their houses (activities which are shown side by side on the Bayeux Tapestry) were all part of the same process of humiliating the king and provoking him to fight. And it was a tactic that proved highly effective.

The Battle of Hastings, contemporaries recognized, was a strange affair. One side – the English – just stood stock still, trusting to the ancient tactic of presenting a solid wall of shields to the enemy. The Normans, for their part, had little option but to try and break this wall, using archers to rain down arrows on to their enemies’ heads, and charging up the hill on horseback, throwing their spears at the English line. It went on all day, which shows that it was a very close-run thing, with both sides equally matched. Two mistakes, however, eventually cost Harold the battle, the crown and his life. First, the English line failed when some of the less-experienced recruits, seeing the Normans retreating, and thinking the day was theirs, broke ranks and charged down the hill in pursuit. It was, it turned out, a cunning Norman ruse. No sooner had the line broken than the Normans wheeled round and attacked their pursuers. The second mistake, as everyone knows, was Harold’s own. Late in the day, at precisely the wrong moment, he looked up.

Few battles ended as decisively as Hastings. Not only was Harold killed; all his brothers and a large number of major English landowners also perished. And yet, in spite of this catastrophic defeat, the remaining English leaders in London showed themselves in no rush to submit to William. Instead, they persuaded young Prince Edgar to wear the crown. William was obliged to continue pressing his candidacy with violence. After a short rest at Hastings, he headed east along the coast, burning and sacking the towns of Romney and Dover. The town of Dover was protected by an ancient fort on the top of the cliffs, which quickly submitted. At this point, one of our main sources for the duke’s career, his chaplain, William of Poitiers, says that, having taken possession of this fortress, William ‘spent eight days adding the fortifications that it lacked’. This has long been taken by some historians as an indication that, when the chips were down, it was possible to build a motte-and-bailey castle really quickly. You will notice, however, that the chaplain’s words are not very specific, and it takes a considerable leap of imagination to believe that what Dover ‘lacked’ was a motte – especially since there is no trace of one at the castle today. Nevertheless, the figure of eight days has in the past been eagerly seized upon, and seems to be supported by the comments of another chronicler on the building of a castle at York, which did have a motte.

The figure of eight days can be tested, to some extent, by measuring the size of an ‘average’ motte, and the amount of soil one man could shift in a day. A recent geophysical survey of the motte at Hamstead Marshal in Berkshire has revealed its volume to be 10,000 cubic metres – a weight of 22,000 tonnes. How much earth a man could move in a day is more speculative, but some idea can be gleaned from nineteenth-century military manuals. The regulations of the Victorian Army suggest that one soldier could dig fifteen cubic feet in an hour, or eighty cubic feet in a day (they evidently allowed for tiredness as the day wore on). By using these figures, therefore, we can say that to build an average-sized motte in eight days, we would need about five hundred men.

While this might at first seem a feasible recruitment target – especially if you had Normans with swords and whips to round up the diggers – it is doubtful whether such a large workforce could be effectively deployed on such a small site without the whole operation descending into chaos. Building a motte was not simply a matter of making a big pile of soil. If that were the case, the Normans’ earthworks would have been washed away by the first shower of rain, and would certainly not have made suitable foundations for the buildings that we know went on top. Where mottes have been excavated, archaeologists have found that they were constructed by using alternating layers of different material: a band of soil would be followed by a band of stone or shingle, followed by another band of soil, and so on. This is also reflected by the picture of the motte being built at Hastings on the Bayeux Tapestry, which shows several men raising a mound with different coloured bands. What we might first imagine to be an artist’s impression of height or depth turns out to be another very literal rendering of reality by the tapestry artist, who clearly understood the fundamentals of motte construction

The Bayeux Tapestry – men building a motte at Hastings.

To build a motte in eight days, therefore, would seem to be pushing it. It would take several weeks, probably running into months, if you didn’t want the whole thing to subside under the weight of the tower. A week might be enough to lay out and establish the site, but a full-scale motte-and-bailey castle would take a lot longer.

It seems, then, that Duke William probably only had time to carry out a few improvements to the existing defences of the burh at Dover, before heading off with his army eight days later. They marched through Kent, and set about laying waste to the land south of London in an effort to induce the remaining English leaders to submit. William crossed the Thames at Wallingford, where several leading Englishmen surrendered, and eventually stopped his army at Berkhamsted, where the final capitulation of the Londoners took place. If he stayed in Berkhamsted for any length of time (and it seems quite likely that he did), then the very large motte-and-bailey castle standing in the town today might have been begun by his men

The next significant date in William’s diary was, as far as we know, Christmas. On Christmas Day, 1066, in the new abbey church which Edward the Confessor had built at Westminster, the Duke of Normandy was crowned King of England.

After his coronation, William was faced with the dilemma common to many conquerors: how to rule his new subjects with fairness, and at the same time reward his victorious comrades-in-arms. Having claimed to be the legitimate successor of King Edward, he wanted to prove to the English that he would be a good king, willing and able to uphold the laws and customs of his predecessor. At the same time, however, he had an army of seven thousand men at his back, all recruited by the promise of rich pickings, and all now hungry for payment. In the early days of his reign, we see William trying to balance these contradictory expectations and demands. Certainly, many Normans grew rich at the expense of Englishmen. Plunder and booty – which the Continental chroniclers called ‘gifts’ – were shipped back to Normandy in large quantities.

Yet even as churches and monasteries were being pillaged, William was being lenient and generous in his dealings with the governing class of England. Of course, a lot of aristocrats, including Harold and his brothers, had perished at Hastings, but there was little anyone could do about that. To those who survived, however, William was quite charitable, allowing them (once they had sworn allegiance, naturally) to remain in possession of their existing lands and titles. When it came to governing his new subjects, the king exhibited the same sensitive streak. Letters drafted by his ministers continued to be written in English, and William was so keen to make a good impression that he even started learning the language himself. He seems to have believed that, given enough time, the English and the Normans could settle down and live happily side by side.

But William’s lenient approach did not endear him to the English. On the contrary, treating them with kid gloves actually provoked the opposite reaction. In the first five years of his reign, William faced a series of rebellions up and down the country. His response was to deal with them in much the same way as he had dealt with his opponents in Normandy. At the first sign of trouble, he marched his army into the affected region, put down the insurrection, and began to build a major new castle. These new royal foundations were, almost without exception, constructed in the larger towns and cities of England, where the population and the resistance were most concentrated. The king had already enforced his authority in London in the weeks immediately after his coronation, building a castle in the south-east corner of the city. When, early in 1068, the first rebellion broke out in the West Country, William wasted no time marching his troops down to Exeter and repeating the exercise. Likewise, when in the summer the two English earls who controlled the Midlands and the North cast off their allegiance, William pushed his way northwards, establishing castles at Warwick and Nottingham. When he reached York, he began the construction of the giant motte that still stands in the city centre(Clifford’s Tower). Returning south, the king planted three more new castles at Lincoln, Cambridge and Huntingdon, mopping up pockets of resistance as he went.

None of this, of course, was especially good for Anglo-Norman relations. When building these new castles, the king and his engineers showed little concern for the English inhabitants of the town or city in question. Nothing was allowed to stand in the way once the optimum site had been selected. At Cambridge, twenty-seven houses were razed to the ground to clear a space for the works to begin. In Lincoln, the number of dwellings destroyed was 166. But while William showed few or no scruples about building castles over people’s homes, he could at least claim to be acting out of strategic necessity. Outside the towns and cities, the king was still reluctant to indulge in any wide-scale disinheritance of Anglo-Saxon landowners.

A handful of his leading men had been rewarded with grants of land at this time, and they were busy asserting their own authority in similar fashion. In Sussex, for example, a number of Continental-style lordships, each organized around a castle, were created immediately after 1066. But how far castle-building extended in general is not known. Writing just one year after the Norman invasion, a monk at Worcester said that, when the king was away in Normandy, his regents ‘built castles far and wide throughout the land, oppressing the unhappy people’. How much this statement reflects the general situation, however, is open to question. One of the regents, William Fitz Osbern, had been made Earl of Hereford, and constructed several castles in the Severn valley region before 1070; our Worcester monk may have heard more horror stories about castles going up than most people. We should also perhaps allow for the fact he was clearly very depressed about the Conquest in general.

‘Things went ever from bad to worse,’ he said in his next sentence. ‘When God wills, may the end be good.’

* * *

What did transform the situation, however, was the great rebellion of 1069. It was a response, in part, to William’s castle-building programme of the previous year. The king’s new foundations were seen as a provocation – an invitation, even, for the English to rise up and smash them. When the men of Northumbria and Yorkshire rose early in the year, the lightly defended motte and bailey at York was an obvious and tempting target. William soon retook the castle and ordered the construction of another, but the city still fell for a second time in the summer. On this occasion the northerners came in greater numbers, aided in their rebellion by the arrival of a Danish army.

‘Forming an immense host, riding and marching in high spirits, they all resolutely advanced on York and stormed and destroyed the castle, seizing innumerable treasures therein, and slaying many hundreds of Frenchmen.’

For the third time in eighteen months, William was obliged to move his army into Yorkshire and retake its principal city. On this, his final attempt, defeating the rebels took considerable effort, and the Danes had to be paid to withdraw. By the time he rode triumphant through the smouldering ruins of York, the king himself was fuming.

Dealing with the rebellion of 1069 appears to have caused something inside William to snap. He had, after all, tried to be nice to the English, letting many of them keep their lands and promising to uphold their ancient laws and customs. Yet all they had done in return was repay his generosity with contempt, and force him to spend time, money and energy in putting down their insolence. What’s more, even now, after three years, they showed no signs whatsoever of giving up. So, since the softly-softly approach had evidently failed, William now allowed the more brutal side of his character to take over. After a sombre Christmas in York, he divided his army into small contingents and sent them out into the countryside of Yorkshire and Northumbria. Their mission was to burn crops, homes and livestock, in order to render the entire region incapable of supporting human life. Modern historians have dubbed this the ‘Harrying of the North’, but only a contemporary author can fully capture the horrific consequences of the king’s decision. One northern chronicler described it thus:

So great a famine prevailed that men, compelled by hunger, devoured human flesh, [and also] that of horses, dogs, and cats… [some] sold themselves to perpetual slavery, so that they might in that way preserve their wretched existence; others, while about to go into exile from their country, fell down in the middle of their journey and gave up the ghost. It was horrific to behold human corpses decaying in the houses, the streets, and on the roads, swarming with worms while they were consuming in corruption with an abominable stench… There was no village inhabited between York and Durham; they became lurking places to wild beasts and robbers, and were a great dread to travellers.

In retrospect, the Harrying was seen as the most savage and merciless act of William’s whole career. At the time, however, the king regarded it as just the beginning of a new direction in royal policy. If the English did not want him as their king, and were never going to give him their love or loyalty, why should he worry about respecting their laws or customs? This cold logic soon translated itself into action. Not only did William abandon his English lessons, and start spending much less time in England; he also decided there was no point in upholding the rights of Englishmen when there were loyal Normans who needed rewarding. In the year 1070, therefore, he deposed many native bishops and abbots, including the Archbishop of Canterbury, and replaced them with Continental newcomers. In the same year, the king permitted English monasteries to be plundered for cash.

The biggest change, however, was not felt in church cloisters, but in the countryside at large. In the wake of the English rebellions, William created huge new blocks of power for his most trusted followers, and charged them with holding down their new territories by whatever means they chose. Above all else, this meant building many hundreds of castles.

One of the main beneficiaries of William’s change of heart in 1070 was Roger of Montgomery. Roger was one of William’s oldest and closest friends: we first spot the pair of them together when William was in his late teens, and their friendship may have stretched back even earlier. Two major things underline the degree of trust between the two men. First, when William set sail for England in 1066, Roger was the man he left in charge of Normandy during his absence. Second, when Roger joined William in England shortly after the invasion, the king rewarded him with large grants of land. Roger was one of the individuals who profited from the early redistribution of property in Sussex, and in 1070 he received an even bigger prize. In the carve-up that followed the Harrying of the North, William made Roger Earl of Shrewsbury (or Shropshire).

This was a very large gift, and it catapulted Roger right to the top of English society. In the list of the top ten Normans in England after 1066, Roger ranks number three – below William himself and his brother, Odo, but above the king’s other brother, Robert. With great power, however, came great responsibility. As earl, Roger was expected to keep order in the region, and also to defend the English border with Wales. Shropshire, like Yorkshire, was one of the remotest and wildest parts of William’s new kingdom. In order to carry out the task appointed to him, Roger built several new castles. One of the most important of these, to judge from its name, was the one he called ‘Montgomery’, after his own home town of Montgommeri in Normandy. This castle, a perfect little motte and bailey, still survives, but for centuries it has been known by its Welsh name, simply meaning ‘the Old Mound’. It is called Hen Domen.

Hen Domen provides an interesting contrast with castles built by William the Conqueror at around the same time. Rather than being constructed in the middle of a town or city, Roger of Montgomery’s new castle was built in the open countryside. Despite its isolation, however, it was of crucial importance for Roger in controlling his earldom. He picked the site in order to command an ancient crossroads, and also to control the traffic across a major ford on the river Severn. Today, the castle is no less lonely than it was nine centuries ago. It squats between two formers’ fields, is overgrown by trees and bushes, and looks for all the world like nothing more than a woodland copse. But despite its apparent obscurity, Hen Domen has once again become very important. In fact, it is one of the most talked-about castle sites in Europe.

For a period of almost forty years, Hen Domen was the site of a massive archaeological dig. Every summer, from the early sixties to the late nineties, archaeologists gathered at the castle for weeks on end to try to uncover its secrets. With a total of over two years spent digging, this was the biggest and most sustained investigation of its kind ever undertaken. Thanks to the work done at Hen Domen, a great deal has been learned, not only about the nature of early castles, but about what life was like within their vanished wooden walls.

In itself, Hen Domen has good reason to be considered special. Although it is only a small- to medium-sized motte and bailey, the strength of the castle’s defences reflect both the high status of its builder and the dangerousness of its position on the border. As at the royal castle at Berkhamsted, built by either William or his brother Robert, we find multiple lines of defence. Three earthen ramparts ring the whole site, forming two deep ditches around the castle. Anyone approaching with hostile intent would have had to cross the first ditch, climb over a wooden fence with a fighting platform behind it, and then negotiate another, deeper ditch – all this before they reached the castle’s main walls, which stood twelve to fourteen feet high.

Of course, it is impossible to say exactly what stood above the ground by digging underneath it. Nevertheless, the excavations at Hen Domen permitted some reasonable estimates. They revealed two rows of post-holes, one set behind the other, which indicated that the walls must have been backed by a fighting platform, raised off the ground by the posts. In order to allow a man to pass underneath it, the platform must have been raised to a height of at least six or seven feet. Similarly, a man standing on top of the platform would need to be protected from attack, so we must assume that the wall rose at least another six or seven feet in front of him, bringing the total height of the wall up to the suggested height of twelve to fourteen feet.

This artist’s impression of Hen Domen, based on the archaeologists’ findings, shows how the castle might have appeared in the twelfth century.

In a similar fashion, the archaeologists were able to estimate the size of bailey buildings at Hen Domen. Certain post-holes were evidently home to very large timbers, and from the scale of these foundations the overall shape of the buildings can be guessed. At the foot of the motte, for example, the archaeologists uncovered the remains of a very large building. In all probability, this was the castle’s great hall. Judging by the massive size of its foundation ditch, the hall stood two storeys high, providing space downstairs for storage, and a main first-floor room where Roger and his household would have sat and dined. Behind the hall the team discovered evidence of a flying bridge of exactly the kind depicted on the Bayeux Tapestry. Again, it was the size of this structure that was striking. The foundations (and also, remarkably, a surviving timber that was found preserved in the ditch) indicate that the bridge must have been twelve feet wide; large enough to ride a horse up, if necessary. Finally, on the top of the motte, the diggers uncovered evidence for a great tower – or rather, several great towers, for it seems that the buildings on the motte were replaced several times over the years. Again, the scale of the foundations suggest that the greatest of these towers was at least two storeys tall.

How were these buildings actually constructed? The trees, as you might expect, were felled using axes and dragged to the site by animals in order for construction to begin. The trunks, however, were not cut to shape using saws, but by the more efficient process of splitting. Starting with a large oak tree, wooden or metal wedges were driven into the trunk along its length, using a wooden mallet or hammer. Eventually a crack would open and, with a little encouragement from crowbars, the tree would split in half. After this, the process could be repeated several times – the half could be split into quarters, the quarters split into eighths, and so on. In fact, if you had a good-sized oak tree, it was possible to get over a thousand square feet of planking from a single trunk. Once you had produced enough timber in this manner, you could start building with them right away – provided your boss wasn’t too concerned about the rough quality of the finish. If, however, he demanded smoother surfaces on his castle walls, these could be produced by working the split wood with an axe, and then dressing it with a smaller, subtler tool called a T-axe.

Other materials besides timber went into constructing an early castle. The walls of buildings could be built or reinforced with clay, as well as the well-known ‘wattle and daub’. When it came to roofing, slate tiles may have been used in some cases, but no such slates were ever uncovered at Hen Domen. Thatched roofs may also have existed, but using thatch obviously meant that there was a much greater danger from fire. Bearing both these things in mind, the archaeologists assumed that the roofs at Hen Domen would also have been made of timber, built either from planking or by using shingles. There was nothing low-status about any of these materials – especially wood. Roger of Montgomery was a very powerful man, and wood was his material of choice. Likewise, the castles built by William the Conqueror and his brothers were constructed in almost every case from earth and timber. The diggers at Hen Domen were slightly disappointed that none of the buildings there seems to have been very ornate – no carved timbers were uncovered. Roger’s castle, it seems, was not a fancy example like the one at Bayeux on the Bayeux Tapestry, with its dragon’s head over the doorway. Nevertheless, the size and number of the buildings was in itself revealing. It gradually became clear to the archaeologists at Hen Domen that they were not uncovering a small huddle of shabby-looking structures, but a site that was thickly planted with buildings, built on a scale that matched the fabulous descriptions of the chroniclers.

The only genuine disappointment for the archaeologists at Hen Domen was the limited number of ‘small finds’ they uncovered, and the fact that none of these items suggested a truly aristocratic lifestyle. There were no brooches or jewellery to compare with the finds at Threave (see Chapter Five); the most exciting find was half a wooden bucket. Of course, we can make certain allowances for the lack of luxury items. This was a castle, not a town or a battlefield; people were not necessarily dropping and losing things all the time. They must have had rubbish pits in which to throw away their unwanted or broken items, but these were never found: despite digging for forty years, the archaeologists only had time to excavate half the bailey. Who knows what treasures – or rubbish – might be concealed in the other half? Hen Domen has by no means given up all its secrets.

But even with all these excuses, the inescapable conclusion was that life at Hen Domen was not exactly luxurious. It was not a place where Roger of Montgomery turned up with his precious things: certainly no gold or jewels, and probably not even much money – only one coin was found on the site. In its early days at least, it was a garrison castle, manned entirely by knights and soldiers, whose standard of living was basic, not to say Spartan. Only two of the bailey buildings showed signs of being heated by fires and, to judge from the animal bones that were found, the diet of the occupants was quite simple. They typically ate beef, mutton and pork, and from time to time they got to dine on deer – a slightly classier dish. All this food, however, could be sourced locally; there was no indication that fancier foodstuffs ever found their way to the castle.

But this would not have been unusual. In the eleventh century, knighthood was still a long way from the fine living and pageantry of the late Middle Ages (see Chapter Four). In Roger of Montgomery’s day, it was not such an exclusive club; knights were numbered in thousands, not hundreds, and the poorer ones were not much better off than peasants who had done well for themselves. The men whom Roger sent to Hen Domen to guard the fringes of his earldom no doubt cursed the cold and criticized the cooking. But their experience was probably little different from that shared by Norman knights all over England.

* * *

Hen Domen was just one of Roger of Montgomery’s castles in his new lordship of Shropshire. He built several others, including the one that used to stand in Shrewsbury itself. But the region he had been given to govern was too big for one man to manage. So, just as William the Conqueror relied on Roger, the earl likewise delegated lands and authority to his supporters, and they in turn built castles of their own. The mottes at Clun, Maesbury and Kinnerley were all built by such men. One of the earl’s most powerful followers, Roger Corbet, decided to follow his boss’s example in an even more direct fashion. Caus Castle commemorates the region known as the Pays de Caux in Normandy – another example of a Norman knight, a long way from home, choosing to commemorate the old country when he came to name his castle. The effect of all this building by Roger and his tenants was that Shropshire was soon thickly planted with new fortresses. Today there are eighty-five surviving castle earthworks in the county, and an additional thirty-six in the former county of Montgomeryshire. The vast majority of these were established in the early years after the Conquest by Roger and his allies. Between them, they transformed the region into the most thickly castellated area of England.

It was, however, only in terms of overall numbers that Shropshire was exceptional. The pattern of castle building in the border region was replicated all over the country, with the greater Norman lords establishing castles, and their minions soon following suit. There was little about this process that was systematic, and very little supervision between one layer of authority and the next. William the Conqueror, for example, personally directed the business of building castles in the major towns and cities of England, but he had little control over what went on in Roger of Montgomery’s earldom of Shropshire. Having decided on a policy of total conquest, he had to place a lot of power in the hands of others. This meant, of course, that the way these men exercised that power was largely up to them – the king had no way of monitoring and supervising their activities. As a means of establishing Norman control over the English, William’s decision was remarkably successful. After 1075, there were no more rebellions in England; the last one took place in East Anglia that year, and the Anglo-Saxon chronicler put its failure down to the fact that the castles in the region were too strong. However, at the same time, a policy of handing large amounts of power to individuals was a double-edged sword. The king knew that, left unchecked, a laissez-faire approach to conquest and castle-building might one day make matters worse. He had, after all, spent most of his youth fighting his enemies in Normandy to deprive them of their castles.

So it was that, twenty years after he had landed at Pevensey beach, William made another momentous decision. The king decided it was time to take stock of his accomplishment, to draw a line under the process of conquest, and to remind everyone – Norman and Anglo-Saxon alike – exactly who was in charge. At Christmas 1085, he launched a great inquiry – a survey of his kingdom so expansive in its scope and so intrusive in its nature that men compared it to the last reckoning of God. They called it Domesday.

After the Conquest itself, the Domesday Book is William’s most famous achievement. As one of the most important documents in English history, it has attracted a lot of controversy over the years. Was it really a one-off original, or had the Anglo-Saxons been carrying out similar surveys for years? More importantly, what was the Domesday Book actually for? It has been suggested several times that it was a tax inquiry, but the arguments never quite convince. The best explanation, to my mind, is that Domesday was created for two reasons. In the first place, it was intended to serve as a reference work for William’s ministers; in order to conduct the business of government effectively, they needed an accurate record of who owned what. Domesday, however, was intended to do much more than this. The point of the exercise was that it was a legally binding document, like a charter or title deed. England had seen twenty years of chaotic land acquisition, but the survey set a seal on this process. It was no longer going to be possible, in theory at least, to grab land from someone else and claim it was yours by right of conquest. The Domesday Book set everything in stone. like God’s last judgement, the book’s verdict was final.

All this means that the Domesday Book is very useful for historians, since it provides rock-solid documentary proof for lots of things – including the early existence of castles. If a castle is mentioned in Domesday, we know that it must have been built before the survey was carried out in 1086. For example, if we turn to the county of Shropshire in Domesday, the first major landowner we find is (surprise, surprise) Roger of Montgomery. At the bottom his entry, we find the Latin sentence Ipse comes construxit castrum Muntgumeri vocatum (The earl himself [Roger] built the castle called Montgomery). Hen Domen, in other words, was built between 1070, when Roger was made earl, and 1086, when the Domesday scribe wrote that sentence.

When it comes to working out exactly how many castles the Normans had built, however, Domesday is a bit of let-down. Although it mentions castles from time to time, the book is a long way from being comprehensive. The king’s surveyors were much more interested in recording the number of manors, plough teams and peasants than they were in noting down where all the castles were. Certain castles, which we know from other evidence had been built before 1086 (such as Dover), are not mentioned in Domesday. Altogether, William’s great survey only provides us with evidence of fifty castles.

How, then, can we go about coming up with a total number? One option is to go looking for mentions of castles in all the other written evidence that survives from the eleventh century. Doing this pushes the total number up to just under one hundred. It is quite clear, however, from surviving numbers of earthworks, that there must have been considerably more than this. For castle scholars, therefore, the only solution has been to go out and count the sites on the ground – not as easy as it sounds, as some have been concealed or destroyed by later rebuilding. In recent decades, however, historians and archaeologists have between them come up with a total figure of around a thousand sites in England and Wales. Probably at least half of these castles were built before the year 1100, with the majority of them being built in the years immediately after the Conquest. This means that, even if we err strongly on the side of caution in our calculations, we have to conclude that around five hundred castles were built by the Normans in England during the reign of William the Conqueror.

When you arrive at a figure as big as this, it really makes you think about the scale of William’s achievement, and the invaluable role that castles played in the Norman Conquest. By 1086, the king’s policy of building castles himself and entrusting his great men with castle-building had proved spectacularly successful. Using five hundred castles, a force of seven thousand men had conquered and held down a country of almost two million people. Not since the days of Julius Caesar, a thousand years before, had such a feat been achieved; never again in the history of the British Isles would it be repeated.

Of course, William’s success was not due entirely to the fact that he and his followers built castles. We could also point to the king’s outstanding ability as a general, and remind ourselves that men like Roger of Montgomery were also zealous and experienced military leaders. Similarly, we should not forget that William and the Normans had more than their fair share of luck. The Battle of Hastings, after all, was almost too close to call – things would have been very different had it been William and not Harold who died that day. Perhaps most importantly, the country that William invaded, for all that it had been buffeted by misfortunes in the eleventh century, was still the strong centralized kingdom of England created by Alfred and his heirs. Taking over such a well-organized state was far easier than conquering a land where government was weak – as later generations of Normans in Wales and Ireland found to their cost.

Bearing all these qualifications in mind, have we been exaggerating the importance of castles? Recently, historians have begun to suggest as much, even to the extent of denying that castles were important at all. The technological differences between the Normans and the English, we are now informed, actually counted for very little in practice: knocking out the Anglo-Saxons in battle was the most important thing. Building huge mounds of earth was all very well but, when it came down to it, they were really symbols of lordship and not weapons of conquest. Personally, however, I wonder if we can really push castles out of the picture to this extent, or redefine them in such terms. Historians have, of course, the enviable advantage of hindsight. From a safe distance across the centuries, and using every available source, we imagine we can see the general picture better than contemporary chroniclers. Men who lived through such traumas are not only likely to be biased; their opinions are also fatally compromised by their provincial perspectives. I have already questioned the credentials of the Worcestershire monk who reported the events of 1067 earlier in this chapter.

But not all chroniclers were so confused and befuddled, or wrote with such enormous axes to grind. Our principal authority for the Norman Conquest is a monk called Orderic Vitalis. He too wrote with hindsight, composing his chronicle fifty years after the invasion, from the safety of his monastery at St Évroul in Normandy. Orderic himself, however, was only half-Norman. His father was a servant in the household of Roger of Montgomery, who travelled to England after 1066, and married an English girl. Originally, this Continental monk was a Shropshire lad; as he tells us in his history, he arrived in Normandy unable to speak French. Unlike his other contemporaries, therefore, Orderic was able to see things from both sides. He still, like all of us, had his prejudices and his bugbears, but his is the least biased contemporary opinion we have on the Norman Conquest. And for him there was absolutely no doubt as to why the Conquest was successful:

The fortifications which the Normans called castles were scarcely known in the English provinces, and so the English, in spite of their courage and love of fighting, could only put up a weak show of resistance.

For Orderic at least, the castle was the instrument with which the Normans had riveted their power into place.

When the Domesday Book was compiled, William the Conqueror was aged sixty or thereabouts. He had lived to grow old, and he had grown to be fat. Neither age nor girth, however, could persuade him to slacken the pace of his lifestyle, or to desist from the brutal kind of warfare that he had made his speciality. In 1087, he was at war with the King of France, and had recently captured and burned the French town of Mantes. As he rode through its smoking ruins, however, his victory was suddenly undone. His horse started and reared up in fright, driving the pommel of its saddle into the king’s ample stomach. It was a fatal injury. In great pain, William returned to his ducal capital of Rouen, to the priory of Saint-Gervais. It was there, at dawn on 9 September, that he died.

The news of William’s death sent shock waves throughout Normandy and England. When it reached Canterbury – where our story began – the author of the Anglo-Saxon Chronicle interrupted his record of the year’s events to record a detailed and impassioned obituary.

‘What can I say?’ he began. ‘If anyone desires to know what kind of man he was, or in what honour he was held… then shall we write of him as we have known him, who have ourselves seen him, and at one time dwelt in his court.’

The chronicler went on to describe the king in a balanced way, setting down both his good and evil deeds. William, he wrote, ‘was a man of great wisdom and power. Though stern beyond measure to those who opposed his will, he was kind to good men who loved God. We must not forget the good order he kept in the land, so that a man of substance could travel unmolested throughout the country with his bosom full of gold. No man dared slay another, no matter what evil the other might have done him.’

Among his blacker deeds, however, castle-building topped the list.

‘Assuredly in his time men suffered grievous oppression and manifold injuries,’ wrote the chronicler. ‘He caused castles to be built, which were a sore burden to the poor.’

So ends William’s story. But the story of earth-and-timber castles, which started well before William’s day, had a long way to go once the king was gone. Some motte and baileys, particularly those built along the Welsh border, continued to be inhabited and improved right down to the end of the thirteenth century. Hen Domen, for example, was not abandoned until the 1280s. When civil war erupted in the middle of the twelfth century, many new earth-and-timber castles were built from scratch, and hundreds of older ones were quickly repaired and refortified. Likewise, when the Normans later carried war into Ireland and Scotland, motte and baileys were still the weapon of choice.

However, in England after the Conquest, the trend was towards peace rather than war. Men who had built castles to secure their acquisitions in the years immediately after 1066 soon found there was no need to keep all of them in constant readiness and good repair. In many cases, they followed the example of Orderic Vitalis’s father, and settled down to marry a nice English girl. Later generations of Norman knights found there was little point in investing time, energy and money in repairing and renovating all the castles that their fathers and grandfathers had built. From the start of the twelfth century, the number of occupied sites began to fall. Abandoned and left to decay, in time their baileys grassed over, and their timbers rotted away.

With castles no longer needed as instruments of conquest and oppression, those which survived this process of thinning down were the ones that could adapt to play new peace-time roles. Many royal castles, for example, survived because they were necessary as prisons, as residences for sheriffs, and as treasuries for the king’s gold and silver. In most cases, however, the castles that survived were simply the ones their owners liked best, either because they were conveniently situated at the heart of their estates, or because they were well-placed for hunting, trade and travel. As they let some of their earlier castles fade into the landscape, and began to invest more and more of their resources in one or two favourite residences, later generations of Normans found they were able to invest in something a little more spectacular than earth and wood.

It was William the Conqueror, once again, who had led the way. In the weeks and months after his coronation, he had built a timber castle in the south-east corner of London. By the middle of the 1070s, however, the king had decided that his new capital required a more permanent and more grandiose royal residence – a building made of stone. It was a castle that took almost thirty years to build, and which William never lived to see completed. Its importance to future generations of castle-builders was correspondingly colossal. As the great stone building slowly inched its way skywards, it became known simply as the Tower. This, without question, was the shape of things to come.