Castles, lordship and settlement in Norman England and Wales: O.H. Creighton examines the many and varied reasons behind the siting of Norman castles, and considers their decisive effect on the cultural landscape of Britain.
As well as being defensible strongholds and elite private residences, most castles were also the hubs of estates. The castle was also a conspicuous emblem of royal authority or seigneurial status. Such qualities are reflected not only in their physical remains, but also in their portrayal within literature and art. While one of the words most commonly identified with castles is `keep', the term is virtually unknown in medieval documentation where the term donjon was generally used. This derives ultimately from the Latin dominium (`lordship'). This word reflects the qualities of these conspicuous buildings of the medieval period not simply as the ultimate place of refuge within a fortified complex, but as a proclamation of lordly ambition.
Castles are notoriously difficult to define, not least because perceptions of what they constituted altered markedly during the period of castle-building, which in England and Wales ranged broadly from the mid-eleventh to the fifteenth centuries. What is clear, however, is that the medieval landscape was dotted with castles of all shapes, sizes and status, built by a wide range of individuals from kings, bishops and major magnates, through to petty manorial lords and sometimes their tenants. Well over 1,000 sites of all dates and types are known from across England and Wales. These cluster most thickly in border regions. In counties such as Herefordshire the density of castles exceeds one site per ten square miles, reflecting a proliferation of castle-building among the ranks of tenants and subtenants in response to insecurity in the late eleventh and early twelfth centuries. Other counties, too, could have surprising numbers of private fortifications: Essex and Northamptonshire both have in excess of thirty castles. Most of these sites survive not as impressive ruins, but as the grassed-over remains of slumped earthworks formerly supporting timber structures. These earth-and-timber castles should not be taken as representing crude forms of fortification, as recent excavations, notably at the site of Hen Domen (Powys), have shown that sites with timber buildings could be equally as defensible and impressive as their counterparts built in stone.
The distribution of medieval fortifications makes clear that castle-building was carried out not as part of any grand national strategy, but as a result of a multitude of individual decisions. While the crown may, in a broad sense, have been able to encourage or discourage private castle-building in given places and regions at certain times, and to actively discourage it in other contexts, there is no evidence of a master-plan of castle-building designed for national defence.
Another misconception is that most castles were built on hilltops. While photogenic gems such as Beeston Castle (Cheshire), Corfe Castle (Dorset) and Peveril (Derbyshire) have stunning settings that grace the pages of coffee-table books, in fact these sites are rather atypical. The domestic, administrative and economic functions of castles ensured that relatively low-lying positions with better access to communications and resources were more likely to be selected, and it is notable how many such castles are overlooked by higher ground. Most castles were not located at inaccessible isolated positions within the medieval landscape, but closely associated with settlements and other features reflecting their status as centres of lordship and consumption, such as fish ponds, deer parks and mills, or ecclesiastical sites signifying seigneurial patronage.
The great castle of the de Warenne family at Lewes (Sussex) was one of numerous Norman baronial capita `twinned' with a nearby monastic house--the Cluniac priory of St Pancras being established at the foot of the fortress in 1077. Such a foundation could reap spiritual rewards for a lord who could gain comfort from the commemorative practices offered by the religious community for their benefactor. The juxtaposition of Norman castle and monastery, meanwhile, acted as a monument to the founding family and a symbol of an essentially colonial settlement. Castles might also be tied into networks of ecclesiastical patronage by the foundation of castle chapels, which might be served by members of monastic establishments of which the castle lord was patron, and by the association between castles and parish churches.
A close spatial relationship between sites of lordship and parish churches is relatively common. It was not unknown for a church to lie within a castle bailey itself, standing at the junction between the zone of lordship and the community beyond, and emphasising the seigneurial influence over ecclesiastical provision. The tiny Herefordshire parish church of Kilpeck is one such building. This structure stood directly between a Norman castle and a small appended settlement which, having withered away following the abandonment of the seigneurial focus, now survives as a series of earthworks within an embanked and ditched enclosure.
The most profound and enduring impact of the medieval castle on its surroundings was, arguably, its effect on patterns of human settlement. This was most marked in the Norman period as the power holders in a new order sought to make their mark on society and landscape. Countless British settlements, from cities to villages and hamlets, came into being as a direct result of, or in some way bear the imprint of, Norman castles and the policies of their authorities. The evidence of historic landscapes and townscapes throughout England underlines not only that Norman castles acted as catalysts for settlement change--their lords energetically engaged in the creation, planning or re-shaping of communities--but also the psychological impact these sites had on contemporary populations.
In the years after 1066 the towns of Saxon England were subjected to a whirlwind programme of castle-building as William the Conqueror sought to suppress the populations of regional centres, creating a network of fortified power-bases which dominated the pattern of communications and government nationwide. These castles disrupted urban road networks, infringed on ecclesiastical property, displaced housing, and doubtless overawed civilian populations. In Wales, by contrast, which was essentially non-urbanised at the time of the Conquest, administratively significant castles were not placed within towns, although some were associated with dependent boroughs, underlining a powerful link between Norman colonial expansion and urbanisation.
The urban castles of England were clearly statements of power to the indigenous population. Nowhere was this more necessary than in London, where William of Poitiers commented that the early Norman fortifications raised in 1067, were:
... completed in the city as a defence against the inconstancy of the numerous and hostile inhabitants.
One of these fortifications later developed into the site known as the Tower of London. The Norman donjon started c. 1078 was a looming feature over the eleventh-century cityscape: built with its most impressive side facing in towards, rather than out from, the city, and an unmistakable feature to river traffic entering the capital from continental Europe. Norman London, however, ,contained three castles, the other two being Baynard's Castle, in the south-west of the city, and, immediately to the north, the short-lived site of Montfichet. Unprecedented elsewhere in Britain, the raising of three contemporary fortifications in a single city demonstrated the need to keep an independently minded population under subjection but also reflected a balance of seigneurial and royal interests. While the Tower remained a royal fortress, the other two castles were under separate private control, both being named after their twelfth-century lords.
William's suppression of the south-western revolt of 1067-68 resulted in the construction of a large ringwork castle in the city of Exeter; the building of other castles in the region followed. In the initial phase of the great northern campaign of 1068, the Norman chronicler Orderic describes the construction of royal castles at Nottingham, Warwick and York, and fortifications at Cambridge, Huntingdon and Lincoln were raised on the King's return journey to the south. Castles were founded in many other urban centres including Chester and Stafford in the wake of associated campaigns, and in total more than forty such sites were built.
Imposed within extant communities, urban castles also fulfilled important administrative, economic and legal functions, and many of these centres emerged as county towns. While the technologies of Norman castle-building were innovative--their structures representing the instruments and symbols of a new alien order--the strong association between castles and important towns perpetuated a royal interest in urban affairs and an urban role in royal government already established in the pre-Conquest period.
Most urban castles were built immediately within the line of existing urban defences. The royal fortresses at Exeter and Winchester, for instance, were set within the corners of city walls of Roman origin; at Wallingford and Wareham mottes and baileys were placed at angles within the ramparts of Late Saxon burhs. Sitting like cuckoos in the nests of other birds, these castles were positioned pragmatically to reuse extant defensive features while allowing ready access to wider hinterlands and keeping civilian populations in check. The unusual position of Colchester Castle, built from the mid-1070s, near the centre of the medieval town is explained by the fact that here the Norman builders made use of an earlier high-status site. The donjon was located on the podium of the former Roman temple of Claudius, also the site of a Late Saxon villa regalis. Prominently. incorporating re-used Roman tile, this immense building--in terms of ground plan the largest Norman donjon in England--was one of a number of Norman castles that made propaganda of the past through continuity of occupation on a site imbued with high-status connotations in the minds of the population. At Chepstow too, Roman tile and brick was deliberately re-used, here to form a conspicuous string course on the exterior of the Norman donjon that, with its echoes of imperial authority, faced into the unconquered lands of South Wales.
In many other cases, evidence of masonry features or earthworks allows us to appreciate that the structures of urban castles embodied the architecture of authority. The immense Norman motte inserted into the Saxon community of Thetford, Norfolk, for instance, still has a base diameter of c.100 metres and is elevated c.20 metres above the surrounding settlement, yet it represents only the vestiges of the original Norman earthwork that would have been topped with an imposing superstructure. Masonry structures such as the Norman donjons at Norwich and Rochester incorporated ostentatious Romanesque detail and, as at Durham, were `twinned' with nearby cathedrals, while the gatehouse at Exeter had elaborate gallery-type openings in an upper storey apparently built for reasons of display rather than to enhance defensibility. One of the earliest surviving fragments of a masonry castle in Britain, this remarkable structure blends Anglo-Saxon architectural elements with Romanesque, suggesting the use of a native workforce accustomed to church-building. In all these cases, the metaphysical impact of castles on contemporary communities was clearly as important as their physical presence.
Vivid testimony of the disruption caused by the construction of these early Norman urban castles is provided in Domesday Book, which in several places retrospectively recorded the impact of castle-building on taxable assets. At Cambridge twenty-seven houses were said to have been destroyed on account of the new castle. At Gloucester the figure was sixteen, although these statistics pale into insignificance compared to Lincoln, where in 1086, 166 of the 970 houses recorded in 1066 were levelled because of castle-building. Here, the precincts of the castle (built in 1068) and Bishop Remigius' new cathedral transplanted from Dorchester on Thames in the early 1070s occupied the entire former Roman upper city--an area in excess of 150,000 square metres--forming a walled focus of Norman military and ecclesiastical authority. At York, the damming of the Foss to create an immense pool flanking one of two royal castles resulted in the inundation of much agricultural land and two newly built mills, and overall one of the city's `seven shires' was said to have been laid waste on account of the castles.
Archaeology has added substantially to our understanding of the physical impacts of Norman castle-building on towns. At Oxford and Winchester, evidence of Late Saxon properties and streets cleared and buried beneath defensive earthworks has been excavated. But by far the most impressive evidence relates to Norwich, where Domesday Book recorded that some ninety-eight houses were demolished in occupatione castelli. Here, excavations on the site of Castle Mall in the late 1980s and 1990s--one of the largest urban digs in Europe--uncovered a wealth of evidence including buildings, rubbish pits, industrial activity and a church and cemetery, indicating the clearance of part of a flourishing community to make way for the royal castle. The area of crown land known as the Castle Fee, defined by a ditch demarcating the jurisdiction of the royal castle from the borough, occupied some nine hectares, forming the administrative and commercial focus of a radically re-shaped town.
The evidence from Norwich indicates that ecclesiastical property was not immune to the effects of castle-building. Such infringements were particularly common in the face of extreme military circumstances, as highlighted in the Gesta Stephani, whose pages are filled with accounts of sieges during the civil war of Stephen's reign (1135-54). The erection of castles within abbey precincts at Malmesbury and Reading, for instance, was met with outrage, while abject horror was expressed at the desecration of the graveyard of Hereford cathedral, where in 1140 a siege castle was built, resulting in the unceremonious exhumation of decomposed bodies. The tensions that could emerge between the castle authorities and other urban stakeholders are well documented at Worcester. In 1069, Urse d'Abitot, Sheriff of Worcestershire, was admonished by Archbishop Ealdred of York for the fact that the bailey of the newly built castle cut off part of the cathedral cemetery. It was not until 1216, when King John was buried inside the cathedral church, that the monks' land was eventually returned.
Norman castle-building was sometimes also associated with changes in the ethnic composition of townscapes. The topography of medieval Nottingham was distinguished by a `French borough' which grew up outside the castle gate in the late eleventh century, its legal distinctions from the `English borough' being preserved through different traditions of inheritance and the possession of separate sheriffs and bailiffs until the fifteenth century. Similar settlements for French immigrants attracted by favourable urban privileges were established in close association with castles at Hereford, Norwich and Shrewsbury, forming new focal points within towns that marginalised the communities of old English boroughs. A number of royal castles were also associated with newly established Jewish communities, the protection of members of which was a duty of constables and sheriffs. The massacre of Jews at the site of Clifford's Tower, York in 1190 occurred as the city's Jewish community sought sanctuary from persecution in the timber tower of a royal castle, and it seems no coincidence that Jewish enclaves at Bristol, Lincoln and Winchester lay close to urban royal fortresses.
With the exception of those temporary and entirely martial castles built in towns during the turbulences of Stephen's reign, urban castle-building was essentially a phenomenon of the late-eleventh century. Later members of the medieval urban elite in England and Wales did not possess defended private residences within towns. The small number of castles or fortified dwellings built by members of the wealthy urban classes tended to be on rural sites, as with Stokesay Castle, Shropshire, adapted from an existing manorial centre by the wealthy merchant Nicholas de Ludlow in 1291. In contrast, many Scottish and Irish towns did contain tower-houses, forming part of a far wider medieval European tradition of private urban fortification most pronounced in parts of the Mediterranean region. The Tuscan town of San Gimignano, for instance, featured more than seventy tower-houses (some up to 50 metres high) built by the patrician families as symbols of their wealth and status. Higher levels of political centrality, as manifested by the enduring royal presence in major towns, would seem to be an important factor explaining the dearth of comparable urban fortifications in English and Welsh towns, and conditions of relative security.
A slightly later and more prolonged phenomenon was the growth of nascent boroughs in association with royal and baronial castles. Settlements such as Ludlow, Pembroke and Richmond are among the numerous market towns that sprang up as appendages to lordly sites. So marked was the connection between castle foundation and the urbanisation of medieval Britain that an estimated three-quarters of all towns in England founded between 1066 and 1150 adjoined castles, and an even greater proportion in Wales. As estate centres and high-status residences, castles attracted immense and varied human concourse, including guests and their retinues, a miscellany of officials, and military personnel including those owing `castle guard' duties.
The magnetic pull of a castle could create a ready-made market, attracting craftspeople and traders, and many castle boroughs started life as informal nucleations at the castle gate. Beyond the defensive perimeters of many castles the vestiges of medieval market places can still be found, often with market crosses. The grant of a borough charter might formalise the urban status of such communities, the town-plan evolving to assume a more economically inspired layout through the planning of a regulated street network, burgage plots and defences. To a great extent, the relationship between castle and borough in the Norman period was symbiotic--the seigneurial nucleus provided commercial opportunities and the settlement a source of labour, services and income accrued through rents and tolls. In the countryside too, the policies of castle lords were expressed in the re-casting of settlements. This was especially true in Yorkshire--where in the wake of William's brutal `harrying' of 1069-70, ambitious new Norman landlords re-organised devastated estates and planned villages in regular form--and in South Wales, where nucleated villages were imported, along with a new manorial structure and lordship.
The Exon Domesday provides insight into the assertive economic policies of castle lords in the late eleventh century in a reference to the manor of Launceston in Cornwall, in the hands of William the Conqueror's half brother, Robert, Count of Mortain. Here, a nearby market was taken away from the control of the canons of St Stephen's-by Launceston and transplanted to the Count's castle of Dunheved, which became the economic focus of the locality and around which grew up a medieval town, displacing the earlier settlement and adopting its name. Launceston is one of a small yet distinctive group of towns whose street patterns suggest they initially grew up within the confines of large outer baileys or enclosures appended to castles. Perhaps the clearest example is the Wiltshire town of Devizes, where a remarkably intact semi-circular street plan includes a curving market place established at the foot of the Bishop of Salisbury's castle. Derived from the Norman-French le devises (`the divisions'), the place-name reflects the foundation of Bishop Roger's new castle and borough between the medieval hundreds of Potterne and Cannings. Many of these towns were de novo foundations, their establishment requiring the creation of new parishes, although other boroughs were attached to existing villages. Norman nomenclature is reflected in the names of new castle-towns such as Mountsorrel, Leicestershire (`sorrel-coloured hill'), Egremont, Cumbria (`sharp-pointed hill') and Montacute, Somerset (`pointed hill'), all of which feature topographical observations about the castle site.
In the eleventh and twelfth centuries, the establishment of castleborough units was a crucial mechanism through which the Norman Conquest was made permanent, although the late twelfth and thirteenth centuries were to gradually see the `new town' phenomenon lose its `castle' component. Edward I's programme of castle-building in North Wales, however, ensured a final flourish for the phenomenon of castle-town foundation. The great royal castles at Conwy, Flint and Rhuddlan were all accompanied by towns for English settlers, designed in regular grid-plan form and provided with urban defences integrated with and complementing those of the castles. But the centrepiece of Edward's strategy and capital of the principality was Caernarfon, where work on the castle and new town started simultaneously in the summer of 1283, when existing houses were demolished to make way for the defences of the borough and twenty men were employed to take away the dismantled timbers. That Edward's fortress incorporated the motte of an earlier Norman castle indicates that here castle-building actually signified a re-conquest of territory. With its polygonal towers and conspicuous banded masonry in deliberate emulation of the Theodosian walls of Constantinople, the design of Caernarfon Castle drew heavily on imperial imagery as a statement of English colonial settlement.
Not all castle-towns flourished, and a large number survive only as the bare earthworks of a deserted site where the population drifted away, or else where the borough was stillborn and failed to develop as a viable commercial site. The Welsh Marches has a concentration of such sites, including examples at Caus, Kilpeck and Richard's Castle. The most dramatic of these, Caus was a castle of the Fitz Corbets, named after the family's homeland in Caux, France. Founded high on Long Mountain within an abandoned iron-age hillfort, the castle's presence inspired the growth of a seigneurial borough that was also crammed into the massive prehistoric defences, constricting its growth and ensuring its eventual fate. Perhaps the most celebrated example of a deserted castle-town is Old Sarum. Located on a bare chalk hilltop dominated by the Bishop of Salisbury's castle, the town was progressively deserted from c.1219 in favour of the bishop's new plantation of Salisbury, on an un-enclosed site in the water meadows below. The inconvenience of the borough's position adjacent to the castle was highlighted in a remarkable poem of Henry d'Avranches, a court poet of the time of Henry III:
A fortress stood upon the hill, exposed only to the winds, which were strong enough to shake its summit. Little water was to be found; but chalk in abundance. The winds howled, but no nightingale ever sang. The chalk soil was bad enough, but the shortage of water worse. The former dazzled the eyes, and the latter provoked thirst. The silence of birds was a loss still worse than the violence of the wind. The one deprived us of pleasure, and the other destroyed our very dwellings.
The poet went on to compare the juxtaposition of fortress and cathedral, which both lay within the reconditioned defences of a former hillfort, to having the Ark of the Covenant in the Temple of Baal. The town's abandonment thus demonstrates that the initial advantages of a fortified borough site closely associated with a castle could be overtaken in the face of social and economic realities. Several examples of castles associated with deserted villages are also known--one of these being More in Shropshire, where the earthworks of a rural community that nucleated within the confines of a large outer bailey attached to a Norman earth-and-timber castle are readily apparent on the ground or from the air.
The impact Norman castles have made on the development of Britain's landscapes and townscapes thus reflects many dimensions of medieval power and lordship. The intrusion of urban castles into Saxon towns is the very embodiment of the `Norman yoke' school of thought, whereas the foundation and nurturing of other seigneurial settlements in association with fortified sites demonstrates the role of Norman lordship in urbanisation and economic growth. All these perspectives demand that we look beyond the defences of these most characteristic, but often misunderstood, features of the medieval world to appreciate their place within, and contribution to, the making of the British landscape.
FOR FURTHER READING
O.H. Creighton, Castles and Landscapes (Continuum, 2002); O.H. Creighton and R.A.H. Higham, Medieval Castles (Shire, 2003); R.A.H. Higham and RA. Barker, Timber Castles (Batsford, 1992); M. Johnson, Beyond the Castle Gate, Medieval to Renaissance (Routledge, 2002); R. Liddiard, "Landscapes of Lordship": Norman Castles and the Countryside in Medieval Norfolk, 1066-1200 (Oxford, BAR No. 309, 2000); K.D. Lilley, Urban Life in the Middle Ages: 1000 to 1450 (Palgrave Macmillan, 2002); T. Rowley, The English Heritage Book of Norman England (Batsford and English Heritage, 1997).
Oliver Creighton is a Lecturer in Archaeology at the University of Exeter.
|Printer friendly Cite/link Email Feedback|
|Date:||Apr 1, 2003|
|Previous Article:||Files on the Abdication Crisis of 1936 have been released by the Public Record Office after the overruling of a 100-year secrecy order. (News).|
|Next Article:||Christian-Muslim understanding in the later Middle Ages: Richard Fletcher asks to what extent medieval Christians and Muslims sought to move beyond...|