The Medieval English Landscape, 1000-1540 (38 page)

Earthwork and timber castles were relatively quick and inexpensive to build. Orderic Vitalis and William of Poitiers both mention eight days in connection with the Conqueror’s structures at York and Dover respectively but cannot have meant that the entire projects were completed within this time; most would probably have taken a few months in total.
²⁵
As such, they were ideal either for kings and major barons in the initial conquest-and-settlement phase of the occupation of the country (when the prime sites had not necessarily been identified) or for lesser lords with a limited amount to spend who felt obliged to defend themselves and their households as best they could: the apparent proliferation of mottes during the civil war of Stephen’s reign must have taken ‘castle-building’ below the ranks of the baronage. In a category of their own are ‘siege-castles’, designed to house troops engaged against an established castle nearby: some 31 have been identified in England and Wales, mostly between 180 and 280 metres distant from the enemy site. ‘Pampudding Hill’ and ‘The Rings’, ringworks built, respectively, by Henry I against Bridgnorth Castle (Shropshire) in 1102 and by Stephen against Corfe Castle (Dorset) in 1139, are good surviving examples; it is likely that these were put up in a matter of weeks, with rudimentary accommodation within them. However, as temporary structures without an administrative purpose, intended to house troops (often mercenaries without any local connection), it is questionable whether they should really be called ‘castles’, despite the common usage.
²⁶

We should certainly not think that timber buildings in castles were necessarily mean in appearance, or that they were always replaced in stone at the earliest opportunity. An open-backed timber-framed tower was built into the defences of the motte at Deddington (Oxfordshire) as late as the end of the twelfth century. The illustrations in the Bayeux Tapestry apparently showing decorated timber structures on the tops of mottes, and the description of one such building at Durham Castle, possibly in the 1140s, as ‘glittering with splendid beauty … each face … girded by a beautiful gallery’ suggest that there was already a concern for display which transcended purely military requirements. There continued to be a timber-building tradition within castles partly constructed in stone well into the fourteenth century. At York, this included the rebuilding in timber of the tower on top of the motte after the original had burned down in 1190 (only to be replaced by the present Clifford’s Tower, in stone, in 1245). A ‘great wooden tower’ on the motte at Shrewsbury, around which a wooden palisade had been built in the 1220s, survived until 1270.
²⁷

Despite the longevity of timber as a favoured material, it was of course vulnerable to fire and rot, so stone featured as an alternative almost from the outset: for example at the White Tower (the great square tower of London Castle), possibly begun in 1078, where imported Caen stone was used as the ashlar; at Colchester, where an even larger square tower, again incorporating Caen stone, seems to have been erected late in the Conqueror’s reign; and at Corfe, where the pre-1100 curtain walls of a motte-and-bailey castle on an elevated site utilized local stone as the most readily available material. Stone also featured in some late eleventh-century castles which lacked either a motte or a dominant great tower: in the curtain walls of baronial castles at Peak, Ludlow (Shropshire) and Richmond (Yorkshire) and in the gate towers of Ludlow, Richmond and William the Conqueror’s Exeter, the latter surviving largely intact.
²⁸

So in the century which followed the Norman Conquest, we have castles built with an emphasis on a ‘dominant feature’ – either motte-and-baileys or those with great stone towers – and we have others which relied for defence mainly on their enclosing ‘ringworks’ or stone curtain walls. Timber was the most widespread building material, but a minority were already incorporating stone. There is no neat progression through time here, but a series of local adaptations to what was readily available, affordable and met the requirements of the site. As we move on to consider the story of castles in the landscape through the rest of the middle ages, this remains the case: we cannot adhere too closely to a chronological approach. That said, castles did not develop in isolation from their political context, and in this connection the conclusion of the civil war of Stephen’s reign does have a claim to be of major importance. The Treaty of Winchester of November 1153, which brought an end to the
war by promising the throne to Henry Duke of Normandy (the future Henry II) on the death of the incumbent king, Stephen, included provision for the destruction of ‘newly built’ castles, and it is clear that efforts were made, even while Stephen was alive, to put this into effect. A statement by Robert de Torigni that in excess of 1,115 were to be ‘overthrown’ is incredible, but it does suggest that the number of active castles dropped dramatically around this time. Recent estimates have suggested that only about 30 brand-new castles were built during the reign but even if deliberate destruction was confined to some of these, far more would have been considered unusable against the improved siege artillery coming into use by the middle of the twelfth century and certainly available to the incoming king. Overwhelmingly, it was the earthwork-and-timber castles which dropped out of use, as – in Charles Coulson’s words – ‘obsolescence, physical decay, tenurial change and the cost of rebuilding in masonry contracted the social range of castle-ownership’.
²⁹
From the late twelfth to the early fifteenth century, the castle – as fortified residence from which authority was exercised over the surrounding area – was essentially a royal or baronial phenomenon where for reasons partly of status and partly of defensibility against the improved siege artillery of the day the more significant buildings had to be in stone. Lesser lords could not afford to keep up, and most contented themselves with moated manor houses from hereon.

The construction in stone of one great tower which stood out as superior to the rest of a castle – as those at London and Colchester were already doing by the 1080s – was to be an enduring tradition throughout the medieval period, although there would be differences in design and purpose according to time and place. Most of them have traditionally been given the appellation ‘keep’, a term which by the sixteenth century conveyed the notion of a defensible structure which could hold out if the rest of the complex fell. Thus, in 1541, two military surveyors recommended to Henry VIII that ‘within the cyrcuyte of the said castelle [Wark-on-Tweed, Northumberland] a strong towre or kepe [be] devysed and made for the savegarde of such mens lyves as were within the said castell when in extreme need shoulde chance’.
³⁰
However, the word was not in common use in the medieval period and the extent to which great towers (or
donjons
to use a medieval term) served a serious defensive purpose remains a matter for debate. The thirteenth-century
donjon
at Barnard Castle (County Durham) had comfortable residential accommodation in the lower portion but a guardroom above, which had direct access to the curtain wall
and could be shut off in the event of attack; this seems to have been intended both as residence and as ‘defence of last resort’ in a manner conventionally ascribed to medieval castle keeps.
³¹
But many of these towers seem to have been designed with other functions in mind. Those built for William the Conqueror at London (the White Tower) and at Chepstow (Monmouthshire) initially lacked any residential accommodation and seem to have been built primarily for formal occasions and as statements of authority to overawe the local populace; the Tower went on to be used mainly as a prison, as did the
donjon
at Lydford (Devon), begun in 1195.
³²
Other early structures generally known as ‘keeps’ also seem to have been used mainly for ceremonies and formal receptions; they might be seen as their lord’s ‘official residence’ and even incorporate some domestic living space, but most of the time he dwelt elsewhere within the castle. This seems to have been the case, for example, at Henry I’s great rectangular tower at Norwich, which had only a small amount of private accommodation within it; at Hedingham (Essex), where the great tower, built around 1140 for Aubrey de Vere apparently to mark his elevation to the earldom of Oxford, originally lacked any private chambers or kitchens (
Figure 33
); and at Castle Rising (Norfolk), where the highly decorated rectangular tower begun at about the same time for William d’Aubigny – proud to have married Henry I’s widow – remained unfinished for well over a century while the lords of the castle normally lived in the bailey below.
³²
There are echoes here of the great halls of early manor houses, already discussed in
Chapter 3
, which until the twelfth century seem normally to have been freestanding structures for entertainment and administration, separate from the living quarters elsewhere in the manorial enclosure.
³³

Figure 33: Hedingham Castle (Essex)
. The great tower which dominates the castle’s inner bailey was built in the 1140s, evidently not as defensible accommodation but for grand receptions celebrating its owner’s elevation to an earldom. It was well-appointed internally, with two rows of narrow windows (the middle pair shown here) lighting the Upper Hall.

By the late twelfth century, however, there are signs that these towers were being provided with more private rooms to serve as domestic accommodation. This has been identified, for example, where they were newly built for Henry II, as at Scarborough (Yorkshire) and Newcastle-upon-Tyne, and also where older structures were converted, such as the great tower at Norham, originally of the 1120s, where the first-floor reception area was divided up. Most great towers would retain a residential function from hereon – not necessarily for the lord of the castle – but the question of how far they were seen as defensible buildings remains. Several visually impressive examples fail to withstand close scrutiny from this point of view. At Dover, the great tower built for Henry II seems to have been intended above all to provide both power-statement and fitting accommodation for a king whose territories bestrode the English Channel, but a ground floor entrance which gave direct access to the storerooms left it vulnerable: it was the construction of inner and outer curtain walls at Dover, fully in place by the end of John’s reign, which seems to have been the principal defensive device. William d’Aubigny’s tower at Castle Rising has the appearance even to the most casual observer
today of a residence rather than a fortress: a neatly proportioned building of greater breadth than height, with many large windows and decorative blank arcading to impress the approaching visitor.

Further issues are raised by some of the great towers built from the mid-twelfth century to designs which deliberately offered alternatives to the traditional square or rectangular layout – such as the polygonal towers erected for Henry II in the 1170s at both Orford (Suffolk), a new site, and at Tickhill (Yorkshire), a former baronial castle, and also the circular
donjon
added to the castle at Conisbrough (Yorkshire) probably for the king’s half-brother in the 1180s. The corners of rectangular towers were susceptible to mining and created blind spots which defenders could not cover, but these rounded or polygonal keeps were scarcely an improvement from a defensive point of view: the supporting buttresses which projected from the external walls at Orford and Conisbrough, for example, were liable to restrict defenders’ field of vision and increase the number of vulnerable corners. The concern for external decoration which is apparent at these towers – at Conisbrough finely dressed ashlar covered the entire external walling while at Orford different
coloured stone was used – strongly suggests that a desire for ostentatious display was a principal motivation in their design.
³⁴

But should a military purpose to these great towers be ruled out altogether? In interpreting them, much depends on how much functionality is accorded to the elements of ‘fortification’. Thick walls, often splayed at the base to add stability, can be interpreted as giving added insurance against battery and mining, but also as a commitment to both opulence and insulation. Machicolation – stone parapets set on corbels which projected from the tops of some keeps, with holes in the floor through which items could be dropped – had some military purpose as a way of assaulting an enemy immediately below the wall, but are equally convincing as a decorative finish at the summit of the structure. Excessive height minimized the risk that attackers could reach the battlemented roofs and facilitated visual command far and wide, but also allowed the castle to be seen – and its lord feared or revered – over the same extensive area. A single well-protected doorway, normally at first-floor level, made a frontal assault difficult but also obliged visitors to arrive by a processional route and to look up to those waiting to greet them. And, once arrow-loops were introduced to castles towards the end of the twelfth century, a balance had to be struck between wide window-openings for sunlight and narrow openings for these loops, the strategic positioning of which sometimes lacks conviction from a strictly defensive point of view. A debate over how far ‘conspicuous defensive features … were rather to conform to the conventions of a militaristic society than for serious use’
³⁵
will continue, but it is surely wrong to see the different purposes as mutually exclusive. Orderic Vitalis was well aware that both considerations could come into play when he described the ‘very well-fortified castle of ashlar blocks’ at Montreuil in western Normandy, built by Henry I’s former justice Richard Basset who ‘swollen with the wealth of England had made a show of superiority to all his peers and fellow-countrymen by the magnificence of his building in the little fief he had inherited’; this castle successfully resisted two assaults by invading Angevin forces during 1136.
³⁶
The weaknesses in the design of the great tower at Rochester – lacking facilities from which to snipe at the enemy and with right-angled corners susceptible to undermining – eventually led to its garrison having to surrender to King John at the end of November 1215; but while this can be interpreted as a typical twelfth-century structure with decorated entrance and magnificent second-floor hall intended primarily for display, ceremonial receptions and a certain amount of accommodation, it played its part in the castle’s seven-week resistance to the king, with the defenders continuing to hold out behind a cross-wall even after the south-east corner had collapsed. It would almost certainly have been hailed as a ‘successful’ military building if the hoped-for French relieving army had turned up.
³⁷
In troubled times it was certainly safer to be inside a
tall stone
donjon
than in an poorly fortified outbuilding, and if the erection of such a structure afforded an opportunity to display wealth, power and aesthetic taste, while also giving some protection, the temptation was hard to resist. We have to see these great towers as serving a variety of purposes – ceremonial, residential, military, aesthetic and aggrandizing – for the rest of the medieval period, while also questioning in each particular case whether they can truly be regarded as independently defensible.