In 1066, as everybody knows, the Normans invaded England. That most engaging of all medieval sources, the Bayeux Tapestry, shows them landing their horses at Pevensey in Sussex and racing to occupy nearby Hastings, from where they would shortly set out to fight the most famous battle in English history.


Before that, they paused to have an elaborate sit-down meal – barbecued chicken is on the menu – and attend to their own protection. “This man,” says the caption of an important-looking Norman holding a pennant, “orders a castle to be dug at Hastings,” and to his right we see a group of men, armed with picks and shovels, setting to work.

The Normans’ decision to erect a castle at the very moment of their arrival might not strike us as particularly remarkable. After all, medieval warfare revolved around the building and besieging of fortresses, and the English landscape of today is liberally studded with their remains. But at the time of the invasion in late September 1066, the Normans’ action was startlingly novel: prior to that point, castles had been virtually unknown in England.

The exceptions comprised a handful constructed a few years earlier by the French friends of King Edward the Confessor. “The foreigners had built a castle in Herefordshire,” says the Anglo-Saxon Chronicle for 1051, “and had inflicted every possible injury and insult upon the king’s men in those parts”.

The fact that the chronicler was reporting a new phenomenon is conveyed not only by his palpable outrage at the Frenchmen’s behaviour, but also by his need to borrow their word for the offending object: this is the first recorded use of ‘castle’ in English.

More like this

The Conquest that followed 15 years later ensured it would not be the last, because the castle was the primary instrument by which the Normans stamped their authority on England. From having almost no castles in the period before 1066, the country was quickly crowded with them. According to one conservative modern estimate, based on the number of surviving earthworks, at least 500, and possibly closer to 1,000, had been constructed by the end of the 11th century – barely two generations since the Normans’ initial landing.

Of course, England had not been without defences before 1066. The pre-Conquest landscape was studded with, among other things, Iron Age hillforts, Roman legionary forts, and the fortified towns built by the Anglo-Saxons themselves, known as boroughs or burhs. But all of these structures differed from what followed in that they were large enclosures designed to protect sizeable communities including, in some cases, non-military personnel. Castles, by contrast, were comparatively small affairs, designed to be defended by a limited number of fighting men. They had originated in France around the turn of the first millennium as a result of the collapse of royal and provincial authority, when power ultimately devolved to those who had the means to build their own private fortifications and fill them with mounted warriors.

As well as being smaller in area, castles were also taller. Some of the earliest French examples were great stone towers, such as the soaring donjon at Loches on the river Loire, built by the buccaneering Fulk Nerra, count of Anjou, around AD 1000, and still impressive 1,000 years later.

But the crucial thing about castles was that they could be created without the need for such colossal investment. It was quite possible to obtain the same advantage of height quickly and on a fraction of the budget by throwing up a great mound of earth and topping it with a tower of wood. As every schoolchild knows, such mounds were known from the first as ‘mottes’.

The point about size and speed is reinforced by the Normans’ behaviour in England immediately after their arrival. At Pevensey they created a castle by adapting a Roman fort, and at Hastings by customising an Iron Age hillfort, in each case hiving off a smaller section of the much larger original.

After their victory at Hastings, as they set about crushing the remaining English resistance, the Normans continued to follow this method of construction. They added new fortifications to the ancient defences at Dover, and almost certainly created the castle at Wallingford by destroying a corner of the Anglo-Saxon borough.

When, late in 1066, the citizens of London at last submitted to William the Conqueror, his first thought was to plant a castle in the south-eastern angle of the city – the site that would soon become home to the Tower.

The White Tower, one of the oldest parts of the Tower of London complex, was built by William the Conqueror. (Getty Images)
The White Tower, one of the oldest parts of the Tower of London complex, was built by William the Conqueror. (Getty Images)

Rising in revolt

In the years that followed, the castle-building campaign intensified. The Normans, wept the Anglo-Saxon Chronicle for 1067, “built castles far and wide throughout the land, oppressing the unhappy people, and things went ever from bad to worse”.

Part of the reason for this intensification was the repeated attempts by the English to throw off the rule of their conquerors. The south-west of England rose in revolt at the start of 1068, apparently led by the surviving remnants of the Godwin family, while in the summer of the same year there were similar risings in the Midlands and northern England.

William crushed them all, marching in with his army and planting castles in major towns and cities. Exeter, Nottingham, Warwick, York, Lincoln, Cambridge and Huntingdon all received new royal fortresses at this time, and further examples were added in the years that followed: Chester and Stafford in 1069–70, Ely in 1071 and Durham in 1072.

The northernmost outpost of Norman power was established in 1080 by the Conqueror’s son Robert, who planted a “new castle” upon the river Tyne, while William himself marked the western limit of his authority during an expedition to Wales the following year, founding a new fortress in an old Roman fort called Cardiff.

The foundation of castles, however, was far from being an exclusively royal affair. William may have raised armies to quell major rebellions, but for the rest of the time he relied on other Normans to keep order in his new kingdom.

In the two decades after 1066 the king rewarded his closest followers with extensive grants of land in England, and the first act of any sensible incoming lord was invariably to construct a castle. In some instances it appears that these were planted on top of existing English seigneurial residences, to emphasise a continuity of lordship.

But in most cases such continuity was lacking because the process of conquest had caused the country’s existing tenurial map to be torn up. Sussex, for example, was sliced up into half-a-dozen new lordships, known locally as rapes, which paid no heed to earlier patterns of ownership. New lordships required new castles, and the rapes were named in each case after the fortresses that sprung up at Chichester, Hastings, Bramber, Arundel, Lewes and Pevensey.

The reorganisation of Sussex into continental-style, castle-centred lordships seems to have been a decision determined by cold military logic. The county had been the Normans’ beachhead, and also the former Godwin heartland. The rapes run north-south, and their castles are all located near the coast, as if to keep the route between London and Normandy secure.

In recent decades, however, the scholarly trend has been to emphasise that castles had other roles beyond the military. The fact that they were often sited to command road and river routes, for example, meant that their owners were also well placed to control trade, and could both protect and exploit mercantile traffic. We are reminded, too, that part of the reason for building a castle could be symbolic. A great fortress, towering above everything else for miles around, provided a constant physical reminder of its owner’s power – a permanent assertion of his right to rule.

During the Conqueror’s reign, this was most obviously true in the case of the three great stone towers the king himself is known to have created at Chepstow, Colchester and (most famously) London. In each case these giant buildings, the like of which England had not seen since the time of the Romans, have strong Roman resonances and were partially constructed using the stone from nearby Roman ruins; not for nothing did 20th-century scholars christen the style ‘Romanesque’.

Indeed, in the case of Colchester it is difficult to suggest a reason for the construction of so massive a building – beyond a desire to be associated with the town’s imperial past. There are no reports of rebellions or military action in Essex during William’s reign, but the great tower he created in Colchester was erected on the ruins of the town’s Roman temple. The Conqueror’s sycophantic biographer, William of Poitiers, draws frequent comparisons between his royal master and Julius Caesar. To judge from buildings such as Chepstow, Colchester and the Tower of London, it was a comparison that the king himself was keen to cultivate.

The towers that William the Conqueror built during his reign had “strong Roman resonances and were partially constructed using the stone from nearby Roman ruins”. (Getty Images)
The towers that William the Conqueror built during his reign had “strong Roman resonances and were partially constructed using the stone from nearby Roman ruins,” writes Morris. (Getty Images)

At the same time, we need to guard against hyper-correction. In recent years, it seems to me, the revisionist arguments about Norman castles have been pushed too far, to the extent that some historians now come close to arguing that they had almost no military function at all.

Take, for example, the castle that William the Conqueror caused to be built at Exeter in 1068. Its original gatehouse still survives, and has been judged defensively weak because it was originally entered at ground level. This may be so, but it takes a considerable leap to conclude from this, as one historian has done, that the whole castle was “militarily ineffectual”.

Much of the site has now vanished, but it occupied an area of around 185 metres by 185 metres (600 by 600 feet); Domesday Book suggests that 48 houses were destroyed in order to make room for it. It was built on the highest point in the town, and was separated by a deep ditch and rampart.

Exeter had fallen to William in 1068 after a bitter three-week siege that saw heavy casualties on both sides – and during which, if we believe the later chronicler William of Malmesbury, one of the English defenders signalled his defiance by dropping his trousers and farting in the king’s general direction. It beggars belief to suppose that the Conqueror, having taken the city at such cost, would have commissioned a building that had no military capability, and was concerned only with the projection of what has been called ‘peaceable power’.

The notion that castles had little military purpose also requires us to ignore the testimony of contemporary chroniclers. William of Poitiers repeatedly describes the castles his master besieged on the continent before 1066 using terms such as “very strong” or “virtually impregnable”. Such descriptions are borne out by the fact that it took the duke months, and in some cases years, to take them.

Yet some scholars are curiously reluctant to allow that castles built after the Conquest served a similar military purpose. The Conqueror’s great stone tower at Chepstow, for instance, has been plausibly reinterpreted in recent years as an audience chamber where the king or his representatives could receive and overawe the native rulers of Wales.

But the fact is that Chepstow Castle was still a formidable building, situated high on a cliff above the river Wye, and defended at each end by ditches cut deep into the rock. True, it does not bristle with arrowloops, turrets and machicolations – but then no castles did in that period, because the technology of attack was primitive in comparison with what came later. Without the great stone-throwing machines known as trebuchets, there was not much an enemy at the gates could do, beyond mounting a blockade and trying to starve a garrison into submission.

In these circumstances, a well-situated and well-stocked castle could be militarily decisive. In 1069 the people of Northumbria overran Durham, massacring its Norman garrison, which tried and failed to hold out in the hall of the local bishop. But when the Northumbrians attempted to take the town again in 1080 they failed, because they were unable to take its new castle.

Subduing the English

One of the remarkable things about the Norman conquest was how quickly the rift between the English and the Normans was healed. Within a generation or two, it is possible to point to castles that did owe more to ideas of peaceful living than military deterrence. But in the years immediately after 1066, filled as they were with bloody rebellion and even bloodier repression – when a few thousand Normans lived among a population of two million English in the daily fear of violent death – in these circumstances castles have to be regarded first and foremost as military installations, introduced to subdue an unwilling population.

Unfashionable though it may be among castle scholars, there is every reason to listen to the testimony of the half-English, half-Norman historian Orderic Vitalis, born in Shropshire within a decade of 1066, who attributed the success of the Conquest to one factor above all others.

“The fortifications that the Normans called castles,” he explains, “were scarcely known in the English provinces, and so the English – in spite of their courage and love of fighting – could put up only a weak resistance to their enemies.”

William’s castles 1066–87

From the moment his army landed on English soil, the Conqueror embarked on a remarkable programme of castle-building...


Established by the Conqueror’s friend William fitz Osbern soon after 1066, Chepstow was acquired by the king in 1075, after which construction is reckoned to have started on its Great Tower.


William built his first castle in England here, the point of the Normans’ disembarkation, to protect his army while they prepared to engage Harold Godwinson.


After his victory at Hastings, William reportedly spent eight days at Dover, an Iron Age hillfort, “adding the fortifications it lacked”. Afterwards it was entrusted to his half-brother Odo of Bayeux.


This was established shortly before Christmas 1066, “as a defence against the inconstancy of the numerous and hostile inhabitants” (wrote William of Poitiers). Work on the White Tower started in the 1070s and continued until the early 12th century.

Old Sarum

Planted in the middle of an Iron Age hillfort, Old Sarum was probably begun before 1070, when the Conqueror went there to dismiss his army after the Harrying of the North.


This most famous of English castles was created a short distance from an existing royal hunting lodge, probably before the council held at Windsor in 1070.


On his return from Scotland in 1072, William stopped to plant a castle in Durham where, three years earlier, his troops had been massacred by the Northumbrians.


William built not one but two castles in York: the first (Clifford’s Tower) was constructed in the summer of 1068, the second (Baile Hill) early the following year.


Norwich was begun before 1075; that year Ralph Guader, the rebellious earl of East Anglia, was besieged here for three months.


A gigantic building, with close affinities to the Tower of London, Colchester illustrates William’s desire to be compared to the Romans before him.

Marc Morris is a historian and broadcaster specialising in the Middle Ages. He is the author of The Norman Conquest (Hutchinson, 2012).

To listen to our podcast interview with Marc on the story and legacy of the Norman Conquest, click here.


This article appeared in BBC History Magazine's The Story of the Normans bookazine