Dover Castle and Royal Power in twelfth-century Kent

245 Dover Castle and Royal Power in Twelfth- Century Kent richard eales Dover castle may well have been, in strictly military terms, the ‘key to England’ in the civil war of 1216-1217, as claimed by Hubert de Burgh, who held the castle for King John and his son in that conflict, and reported by the chronicler Matthew Paris: at least many modern historians seem to agree.1 Hubert had every reason to maintain such a claim subsequently, as he was in effect burnishing his own record as a military commander, and perhaps justifying the vast expenditure lavished once more on the castle under his own custody in the 1220s. Understandably, the phrase has been much quoted, not always in context, by the authors of guidebooks and textbooks alike. Yet what value it has for the longer-term history of Dover, let alone English royal castles or castles in general, remains debatable. The needs of coastal defence, even if assumed to be real, fluctuated over time and were combined with other aims, especially in contemplating an expensive and prolonged building campaign like Henry II’s at Dover a generation before the famous sieges of 1216-17. What Henry aimed to achieve, or thought he had achieved, by pouring such resources into the construction of Dover, and above all its great tower, in the 1180s, emerges as a complex issue in recent research. Modern writers have greatly expanded the range of possible influences to be considered, and many of these approaches are complementary, even if a single agreed conclusion is unlikely. The aim in this paper is to set Dover in its regional context of Kent and south-east England in the twelfth century. Dover Castle and castle studies The specific study of a major monument like Dover castle creates areas of debate, approached from different viewpoints by historians and archaeologists, with their own agenda of enquiry. Site-specific research is evidently of prime importance, but even at Dover it has tended to be intermittent, pursued as opportunity arose. Thus in the 1960s the end of military residence in the castle and a major reassessment of the documentation for the publication of the first two volumes of The King’s Works in 1963 was followed by two important excavations outside the inner bailey, conducted respectively by Stuart Rigold and Martin Biddle.2 Only recently, through the English Heritage project initiated in 2009, has there been much more detailed survey work on the Henry II inner bailey walls, the rebuilt ranges within them, and above all on the great tower itself. The conclusion to RICHARD EALES 246 be drawn from this is how much is still unknown about the development of the site as a whole, and how difficult it will be to fill in many of the gaps. That is despite the fact that Dover is a major royal castle, constructed almost entirely, so far as its surviving fabric is concerned, in the Angevin ‘post-pipe roll’ period, with chancery and account materials to supply more information for the thirteenth century.3 This contrasts with the other royal great towers in Kent at Canterbury and Rochester, built earlier and with no such documentation. Yet the very scale of the later rebuilding means that it is almost impossible to say what was within the presumed Iron Age enclosure at Dover before Henry II’s builders set to work in about 1168, presumably to remedy its shortcomings. Biddle could reach only tentative, though entirely plausible, conclusions about early earthworks around the Roman pharos and St Mary’s church south of the inner bailey, the possible, though still speculative, original core of the castle.4 More surprising, there remain many unknowns about the staging of Henry II’s own campaigns of 1168-73 and 1179-1189. The construction of the keep in the 1180s, evidenced by the weight of expenditure and a handful of specific references, including those to the presumed designer Maurice the Engineer, is secure.5 Much more doubt though attaches to the progress of the walled enclosures: the inner bailey and the early phases of the outer walls to the east. Rigold’s discovery of the footings of slightly earlier structures just south of the inner bailey supplied more information but also more questions to which there are no obvious answers. Were these the products of Henry II’s first works of 1168-1173 or pre-existing features? There are consequent problems if almost all the rest of the surviving Henrician work has to be accommodated within the later phase of 1179-1189, which must already include the keep. Was there a single outline plan from the first or a radically evolving one? Was it conceivable, at any stage, that Dover could have remained as a keepless castle or can it be assumed that the inner bailey was intended from the first as the setting for an exceptionally large great tower and was constructed with it?6 The central importance of the work initiated in the 1180s remains, especially as the keep alone accounted for more than 60 per cent of Henry II’s total expenditure on the site, but much else is less clear. Further interpretation of site-specific research, even for a comparatively welldocumented site like Dover, therefore demands wider comparisons and larger contexts. If the needs of coastal defence, as in 1215-17, and to a lesser extent in the 1260s, cannot just be read back into the twelfth century, then how significant were they? What exactly was Dover the ‘key’ to in the reign of Henry II, or his predecessors back to William I? Here debates about national political history mesh in with debates in castle studies over the last generation. Among those who have contributed most to the study of Dover, and Angevin castle policy as a whole, Allen Brown never really doubted its primary military purpose. He applied this analysis equally to the keep itself, though curiously arguing that its designer, as well as presumably the king and his advisers, failed to incorporate ‘new principles of military architecture’ already evident elsewhere. Accordingly, despite the huge resources expended on it, the Dover great tower ‘like a conventional battleship in the atomic age, was in fact obsolete almost as soon as it was built’.7 Such an analysis presents obvious problems when it comes to making sense of royal policy, though it should be said that Brown never suggested that castles only ever served DOVER CASTLE AND ROYAL POWER IN TWELFTH-CENTURY KENT 247 military purposes. Like many others he was propounding a model of multiple functions for castles, but with an assumed priority given to defence, especially in their design, which he certainly had no hesitation in applying to Dover. The shift which followed after the 1970s was increasingly towards a similarly multi-functional but less weighted scheme of interpretation, in which the always-admitted roles of castles as residences, centres of government and estate administration, or as symbols of wealth and power, were just as important as their military functions. Charles Coulson, challenging the ‘militarist’ interpretation of castles in a pioneering article, drew attention to a letter of 1247 in which Henry III ordered his constable of Dover, about to receive a diplomatic visitor from France, ‘to show the castle off in eloquent style, so that the magnificence of the castle shall be fully apparent to him and that he shall see no defects in it’.8 Yet even Coulson, writing in 1979, was prepared to contrast this with events directly after the fall of Normandy in 1204, seeing military needs then as relatively more significant, even before French invasion actually materialized and led to the siege of 1216-17.9 In recent years more radically revised interpretations have been argued, implicitly or explicitly challenging the presumption of any military rationale for most English castles and largely severing the study of castellated buildings and their social importance from the history of war.10 Such conclusions have indeed long been advanced for the later middle ages, but usually on a basis that was compatible with the traditional interpretation. The assumption tended to be that these were ‘fortified’ houses rather than castles, or that at some date, variously placed in the fourteenth or fifteenth century, castles began to ‘decline’ and, in most cases at least, to lose their earlier roles of directly providing defence and security behind their walls. Reading back these ideas into the pre-1300 period constitutes a more fundamental reappraisal, in effect challenging those who take a contrary view, including political historians studying specific episodes of conflict, to restate their case. Such a debate should be welcomed, but one potential problem with it is the danger of juxtaposing current social or architectural history with outdated analyses of politics or warfare. A narrative formed simply by linking together battles and sieges has limited value, but so does a reaction which merely points out that such events were rare or untypical, for individual sites or in general. Instead it is necessary to employ broader and more inclusive concepts of power in medieval society. To proclaim status and wealth by architectural or other display might in itself imply consequent threats of force to back them up, if only as a last resort; these categories of action were not easily segregated in practice or even in law. Almost by definition, such assertions of hierarchy or authority if successful should not lead to actual violence. But there is little more basis here to postulate the peaceful nature of castles as a general rule than there was to associate all castles with ‘feudal anarchy’ under older interpretations: in reality almost all depends on the specifics of time and place. Some medieval societies were more peaceful than others. Wider debates have thus enriched the range of ideas which can be brought to bear on a particular example like Dover, but the key problem remains that of identifying the actual circumstances which led Henry II to rebuild it. Much the same can be said of the ‘debate-within-the-debate’ on the role of great towers in castles of this period. Recent writers have again become increasingly confident in assigning largely residential and ceremonial functions to these grand RICHARD EALES 248 and expensive buildings, so leaving Allen Brown’s ‘battleship’ not only obsolete but scarcely ever afloat.11 But this shift of categories in itself hardly begins to answer the specific question of what return the king expected to get for his huge outlay on the tower at Dover Dover Castle in the context of Kent A potential intermediate line of enquiry into castles lies between site-specific research and wider national or international generalisations. That is the regional context: in the case of Dover, Norman and Angevin Kent or South-East England. Examination of this context, both before and after Henry II’s reign, raises issues about similarities and differences between Dover and the other major royal castles in the county, about the influences exerted by castles of other lords, and more generally about the recurrent specific problems of governing the region. Regional studies of castles, in East Anglia and the South-West, as well as in parts of Wales, the Midlands and northern England have featured in recent literature and have addressed these kinds of problems, but have not so far been attempted for the South-East before the late medieval period.12 Much writing about Kent naturally emphasizes the strategic position of its setting in south-east England; a contact point with the Continent whether by armed invasion or peaceful influence. Such assumptions can be read back to Julius Caesar or St Augustine, back from Napoleon or Hitler. F.W. Maitland, dismissing the idea that distinctive traditions in later medieval Kent could be explained just by survival from earlier customs, wrote that the county ‘was no remote fastness … it is the garden of England, of all English counties that which is most exposed to foreign influences’.13 In relation to the Norman period a narrative can be constructed by linking individual crises from before 1066 onwards: the naval role of Sandwich in the Godwin civil war of 1051-52 and in earlier eleventh-century regime changes, William of Normandy’s decision to move east after the battle of Hastings, taking and fortifying Dover, which then had to be held against an attempted revolt by Count Eustace of Boulogne the following year. Stephen, having traversed Kent to launch his bid for the throne in 1135, left Dover and Leeds castles in the hands of his potential rival Earl Robert of Gloucester, who had been entrusted with them by Henry I. But on Robert’s defection in 1138, he was compelled to besiege and capture them to secure the county once more as a crucial base and a link to his lordship of Boulogne.14 Handled uncritically, this leads back to Dover as ‘the key to England’, not just in the special circumstances of 1216 but throughout the middle ages and beyond, as a fact of geography. Thus in David Cathcart-King’s Castellarium Anglicanum of 1983 Kent is discussed under the tell-tale heading, redolent of the 1940s, of ‘The Invasion Coast’. Continuity is certainly assumed so far as castles are concerned between the Norman period and the invasion fears of the Hundred Years War from the 1330s onwards, as well as later periods. Puzzlingly, given the almost universal view that castles in war had only a limited and relative capacity to block the passage of larger hostile forces, a verdict seemingly borne out by the events of 1216 or the successful French raid on Dover in 1295, he also argued that Dover, DOVER CASTLE AND ROYAL POWER IN TWELFTH-CENTURY KENT 249 uniquely among the castles of the area, ‘stood on what is arguably the best invasion route of all, and effectively blocked it’.15 More recent scholars have been rightly sceptical of this kind of geographical and military determinism. The narrative can be read in different ways, rather than just by highlighting selected episodes of threatened though rarely realized invasion. Thus relations between English rulers and the counts of Flanders and of Boulogne were complex and fluctuating in the twelfth century, including the elaborate treaties of 1101 and 1110 between Henry I and Robert of Flanders, revived and revised in Henry II’s treaty with Thierry of Flanders in 1163.16 All these agreements, it should be noted, were promulgated in Dover, whatever the limitations of the castle in that period. Anglo-Flemish relations are especially crucial for any assessment of Continental links passing through Kent and Dover, in part because this was never the natural route for kings and lords crossing to Normandy in the twelfth century, Portsmouth being the commonest embarkation point.17 The conclusion is that this was a history not of consistent strategic needs but of constantly shifting priorities, which alone can explain the commitment of large resources to Dover castle at a particular date like the 1180s, or in the aftermath of the civil war at the end of John’s reign, the 1220s. But the history of twelfth-century Kent, or of Dover castle, cannot be based only on an analysis of external influences on the region, even a more sophisticated analysis. There is also the question of its internal politics. The real answer to the ‘Invasion Coast’ view of medieval Kent is to recognise that the history of the region was characterized by interactions between such external influences, including increasingly that of London on the west and north of the county, and its complex internal structure. This process conditioned all kinds of political decisions in the twelfth century, from relations between kings and archbishops, the privileges of towns including the Cinque Ports, local loyalties in general as highlighted in Stephen’s reign, to the rebuilding of Dover castle. But though the outcomes were complex, some main features of Norman Kent can be simply summarized.18 First, there was the high proportion of the county’s lands held in ecclesiastical lordship, approaching 50 per cent by value in Domesday and subsequent estimates, a figure exceeded only in Worcestershire. This made Kent unpropitious for baronial powerbuilding on a large scale, and in particular created problems for those exercising comital or pseudo-comital authority in the county: Odo of Bayeux under William I, William of Ypres under Stephen, Hubert de Burgh in Henry III’s minority, and so on. Second, after Odo’s fall in 1082, made permanent by his revolt against William Rufus in 1088, a pattern of lesser baronial estates, mostly now held directly of the Crown, was established across Kent by the early twelfth century. It would be too much to call these local baronies, as many of their lords also had significant holdings outside Kent, but they were small or medium sized in scale. Only one baronial family of national magnate rank held a Kentish lordship all through the period, the Clares in Tonbridge.19 Many of these baronies owed castle-guard service at Rochester or Dover in the twelfth century, like the eight baronies owing a total of 116 fees to Dover, copied in several variant forms into the Red Book of the Exchequer and other thirteenth-century registers.20 There was also considerable tenurial intermixing; a majority of Odo’s Domesday tenants held lands from the archbishopric too; one reason why so few of them followed him into revolt and RICHARD EALES 250 dispossession in 1088.21 This created a legacy of uncertainty over multiple claims to overlordship in some cases. Third, there were in Kent an unusually large number of significant, and significantly privileged, towns: Canterbury and Rochester, the two major ports of Sandwich and Dover, the smaller ones at Hythe and Romney. All these coastal towns, later members of the formalised Cinque Ports confederation, owed some sort of individual ship service to the king by William I’s reign, and all possessed early mints. Recent research has tended to increase estimates for the likely population sizes of all the Kent towns in this period, giving Kent an unusually high ratio of urban to rural population in the twelfth century, though of course these categories are always debatable.22 Other distinctive local features could be added, such as the Kentish customs later allowed as an exception in common law, the diverse topography of the county with its large areas of woodland and coastal marshes, and the scale of fishing and coastal trade. From the king’s point of view, however, the issue was governance, and it can be suggested that there were broadly two different approaches which could be adopted in varying degrees. The first was for rulers to co-operate with and rely on the officials of the archbishop’s lordship, if not so much in routine administration and revenue collection, then certainly in the operation of local courts, as witness the preponderance of Kentish hundreds in ecclesiastical hands, and in more informal decision making. All through the period, whenever the sources permit, royal sheriffs and other officials can be seen doing exactly this. Even late in Stephen’s reign, when his relations with Archbishop Theobald (Abp 1138-61) were notoriously strained, the sheriff of Kent Ralph Picot, who retained office until 1160 under Henry II and had strong local connections, was brokering such working agreements, a process no doubt contributing to his own durability in office.23 The apogee of this approach in castle policy was Henry I’s grant of Rochester castle as a perpetual custody to Archbishop William of Corbeil (Abp 1123-36) in 1127, with permission to erect a turris there.24 The outcome was the construction of the massive Rochester keep over the succeeding decade or so, probably by the archbishop’s death in 1136 and presumably at his expense; at least there is nothing in the 1130 Pipe Roll or other sources to indicate royal subsidy, though the possibility of undocumented arrangements cannot be ruled out. This added a link to the chain of royal great towers: London, Canterbury, Rochester and Dover, which was to be completed by the 1180s and was built in that order, the date of the Canterbury tower, though still debated, lying almost certainly within the range 1090 to 1120.25 But there was an alternative royal system, of falling back on a pattern of secular control in Kent, relying on the direct use of the king’s officials, and more informally on patronage and influence over secular lords, as practised in many other parts of England. The most obvious reason to adopt this approach was the breakdown of the king’s working relationship with a particular archbishop, as with Henry II and Thomas Becket in the 1160s, though the precedents are clear in Anselm’s two exiles between 1097 and 1106, and the most drastic example of all was the Interdict of 1208-1214 in John’s reign. In these extreme cases the outcome was temporary confiscation of ecclesiastical lands and rights of lordship: if Stephen exercised some restraint over such action between 1148 and 1153 then Henry II was not so patient.26 Temporary extensions of royal authority could also arise periodically and less violently through archiepiscopal vacancies. But even in more DOVER CASTLE AND ROYAL POWER IN TWELFTH-CENTURY KENT 251 normal conditions, the royal castles, Rochester and Dover as well as Canterbury, which was the sheriff’s main base, were centres from which political control could be managed, with castle-guard services one point of contact and reinforcement of local lordship, in parallel with court attendances. Even the 1127 perpetual grant of Rochester, which transferred the castle-guard services to the archbishop, fell short of outright alienation, as it was made saving all allegiance owed to the king.27 Eventually the betrayal of Rochester to the rebels in 1215 by Reginald of Cornhill, who was both royal sheriff and the archbishop’s constable, ended a final attempt to maintain local peace by sharing power along these lines, initiating civil war and the siege of the castle in October to November 1215. Rochester was apparently a royal base in Henry II’s reign as in Stephen’s; one account even claiming that Becket’s murderers in December 1170 attempted to summon the garrison of the castle to their aid.28 The overall pattern of castles inherited from the Norman settlement in Kent is also compatible with this analysis. Odo of Bayeux himself built no castles on the demesne manors of his huge Kent lordship, but like later earls based his power on custody of royal castles, one reason why he was so vulnerable to any loss of royal favour, as were William of Ypres and Hubert de Burgh after him. In contrast, many of his tenants or their successors built lesser castles on their own lands, as at Folkestone, Coldred and in a band across west Kent south of Rochester: at Stockbury, Binbury, Thurnham, Leeds and probably others. A smaller number of castles was held of the archbishop’s lordship, notably at Saltwood, Eynesford and Stowting. In all these cases early dates are only probabilities, given the lack of documentary confirmation; it is well-known that a large proportion of archaeologically evident Norman castle sites across England lack any pre-1200 written history.29 Those which have documented later histories tended naturally to be those which were rebuilt more elaborately in stone, like Leeds, Allington and Sutton Valence. Tonbridge, though eventually acknowledged to be held from the archbishopric in a compromise agreement, was long in dispute and functioned effectively as an independent centre of power under its Clare lords.30 The distribution of castles reflects the balance of power in Norman Kent: royal authority variable because sometimes heavily delegated, great ecclesiastical estates, and a network of smaller lordships rarely in the hands of major magnates. Henry II’s rule in Kent A review of Henry II’s government of Kent through his reign therefore contributes to explaining his investment in castle-building in the region, culminating in the quite exceptional reconstruction of Dover and its great tower in the 1180s. It begins with the question of what Henry inherited from Stephen in 1154 and the restoration of royal power, so far as it was required in Kent.31 The prevalent view of Kent in Stephen’s reign used to be that it remained essentially peaceful and untroubled through the ‘anarchy’ because it consistently supported the king, and that such disturbance as did occur was mostly caused by Stephen’s importation of Flemish soldiers under the command of William of Ypres to bolster his campaigns elsewhere. The author has argued that this is clearly an oversimplification, which does not make sufficient allowance for the ecclesiastical, and Canterbury-centred, RICHARD EALES 252 bias of many sources. Local loyalties were always more complicated than this, and relative peace also owed much to ‘a balance of local forces … kept in being by compromises and alliances’ between leading figures, as already instanced in the royal sheriff’s close relationship with Archbishop Theobald and his household. There are indications that even William of Ypres participated in such arrangements too and was much more than a predatory alien in the county.32 The support of the Kentish lords for Stephen was also not quite as consistent as was assumed by earlier historians. At least three of the Kent barons: William de Crevequer, Hugh of Dover and William Patrick, two owing service in Dover castle and the third at Rochester, had gone over to Normandy and witnessed charters for the newly-installed Duke Henry by 1151, two years before his successful invasion of England and succession agreement with his rival.33 When Henry conducted his final meetings with Stephen early in 1154 to discuss the implementation of that treaty, at Canterbury and Dover, he was not necessarily entering a hostile zone, despite the story later told by Gervase of Canterbury that he returned to London because of a plot to kill him on the part of Stephen’s Flemish followers. The count of Flanders also joined the meeting at Dover, and was there again for a further meeting with Stephen later in the year.34 More convincing chronicle evidence tells us that William of Ypres, and presumably many of his followers, left Kent in 1155, soon after Henry II’s succession.35 But, as in many other regions, Henry took some time to assert his authority, first consolidating relations with Archbishop Theobald and appointing his archdeacon Thomas Becket as the new royal chancellor. Only in 1157 did Henry discontinue payments to William of Ypres from the Kent farm and compensate Faramus of Boulogne, another Fleming who had held the custody of Dover and its castle, for the loss of his position in Kent. Faramus, though, remained in England and apparently in royal favour, retaining his lands in East Anglia, so Henry’s decision was specifically one about the rule of the South-East. Apart from the Flemish withdrawal, there was actually much continuity of administration. Stephen’s sheriff retained office to 1160, and county farms were relatively stable in the post-1157 pipe rolls.36 As king, Henry naturally wished to maintain good working relations with Flanders itself, which he managed to do until the late 1160s. After the death of Stephen’s surviving son William in 1159, Henry was instrumental in arranging the important marriage of the old king’s daughter to Matthew younger son of Thierry count of Flanders, so confirming him as count of Boulogne in 1160, the formal treaty with Count Thierry following in 1163.37 During the central period of Henry’s reign, between the onset of the Becket dispute in 1163 and the aftermath of the 1173-74 revolt, this general stability and continuity in Kent came under threat. Both its elements: the balancing of overlapping interests within the county and the diplomatic resolution of problems with outside Continental powers, were called into question. In the event though, the previous system of control was eventually restored by the king along more-or-less the same lines when circumstances permitted. The importance of local and Kent problems in the early stages of the Becket conflict is perhaps often underestimated, because they were so soon overtaken by wider ecclesiastical issues. It is clear that Henry would have expected Becket (Abp 1162-70) to assume the same role as previous archbishops as a collaborator DOVER CASTLE AND ROYAL POWER IN TWELFTH-CENTURY KENT 253 in governing the region, and indeed to be much more sensitive to the king’s aims than Theobald had been. In these circumstances the flurry of demands issued by Becket in 1163 relating to long-running local disputes must have seemed like a deliberate rebuff. They included the requests for homage from Earl Roger de Clare for Tonbridge castle, and homage from William de Ros, holder of one of the Kent baronies appurtenant to Dover castle, for fees in Maidstone and the return of two fees in Saltwood, crucially including the castle there, as well as a reiteration of the archbishop’s rights at Rochester.38 These were not jurisdictional niceties but potentially significant moves in terms of local power. W.L. Warren saw them as typical of a series of ‘grand gestures’ of defiance by which Becket initiated his quarrel with the king.39 Anne Duggan, using the same evidence but giving more weight to the account of Ralph of Diceto, drew an almost exactly opposite conclusion: that Henry had already abandoned Becket and denied him royal favour by the beginning of 1163, encouraging all subsequent attacks on him, so that the archbishop was almost entirely the innocent party and in local terms was simply demanding what was owed to him.40 Study of the Kent tenurial claims themselves, and the tangled allegiances attached to them, makes this cut-and-dried assessment of legality unconvincing. In the Ros case Henry got a judgement in his favour, though Diceto claimed that the judges were biased. The division of the various Saltwood fees and the status of the castles there and in Tonbridge are also far from clear in the surviving evidence, making claims about them negotiable and ambivalent. It is more plausible to follow Frank Barlow in the wider conclusion that this was a drift towards ruptured relations fuelled by provocations on both sides, rather than a matter of right and wrong.41 The effects on the royal government of Kent and its castles were however potentially wide-ranging. While it can be argued that for Henry’s wider plans the Becket exile was little more than a distraction right up until its denouement in 1170, this was hardly the case in Kent, where the archiepiscopal and monastic lands of Canterbury were already seized by the king after 1165 and put into the hands of unsavoury agents like Ranulf de Broc. These local disruptions coincided with worsening relations with Flanders, as Matthew count of Boulogne demanded compensation in England for his wife’s inheritance claims. By 1167 he was rumoured to be planning actual attacks on Kent, though Gervase of Canterbury’s later account of his preparations, involving the use of 600 ships, is hard to credit.42 It came to nothing, probably because Henry II was too successful in defending Normandy against the allied threat from Louis VII of France. Given the entanglement of Anglo-Flemish negotiations and attempts to resolve the Becket dispute, it was no surprise that when opportunity presented itself the new count Philip of Flanders and his brother threw themselves into the Young King’s revolt against Henry in 1173. According to Roger of Howden, Count Philip was promised Kent with the castles of Dover and Rochester as his reward, but though Flemish forces fought unsuccessfully in East Anglia in 1173, no serious attack on, or revolt in, Kent actually materialized.43 Henry’s penance in Canterbury cathedral in 1174, to which he attributed the almost-simultaneous capture of the king of Scots near Alnwick, must qualify as the most significant narrative event in Kent during the whole crisis. There was also administrative continuity during and after the revolt. The royal sheriff Gervase of Cornhill retained office all through these events RICHARD EALES 254 from 1168 to 1176, as the ecclesiastical confiscations were unwound and a new archbishop Richard prior of Dover, a man with impeccable local experience and acceptable to the Canterbury monks as a former member of their community, was elected with royal approval in 1173.44 What really happened over these years in the 1160s and 1170s, even setting aside the temporary confiscations, was a strengthening of royal secular control in Kent. To an extent this simply mirrored the national trends of Angevin ‘reform’, with legal assizes, revived judicial eyres, the Cartae Baronum of 1166, the 1170 Inquest of Sheriffs and so on, but the South-East was one of the regions most specifically affected. In some respects the process was opportunistic. One holder of a barony owing service to Dover, Manasser Arsic, died in 1171 or 1172, and his lands came into the king’s hands in wardship. Two others, William of Avranches and Daniel Crevequer, died later in the 1170s with similar consequences. William Patrick, whose lordship owed service at Rochester, was unusual in actually joining the revolt or in giving grounds for suspicion; his lands were in the sheriff’s hands in 1173-75.45 Also in 1175, Saltwood castle was ‘destroyed’ while in royal hands, perhaps anticipating a future need to return it to the archbishopric, and Allington castle was apparently treated the same way.46 But the most revealing case is that of the barony of Chilham, whose lord Hugh of Dover, sheriff of Kent 1160-68, died in about 1170. Though his heir, his nephew John, was allowed to succeed, and no evidence survives of royal disfavour towards him, Henry retained Chilham castle and rebuilt it at a cost of about £428 recorded in the pipe rolls between 1171 and 1174, constructing the small polygonal great tower which still survives in a damaged and altered form just over five miles from Canterbury. Not surprisingly, after this investment, it remained in royal hands whatever the legal justification may have been, until it was eventually returned to John of Dover’s granddaughter and heiress on her marriage to King John’s illegitimate son Richard.47 In effect these individual actions formed part of a royal castle policy in Kent between 1168 and 1175, which included recorded expenditure, as well as the £428 at Chilham, of £102 at Canterbury, £132 at Rochester and almost £500 at Dover. Few specific details are given about the Dover works in the pipe roll accounts though it is clear that masonry was constructed as well as earthworks, the names of several ‘viewers’ (assessors) and the mason Ralph being mentioned.48 Given the wider pattern, it is hard to see the expenditure at Dover before 1173 as solely a military response to renewed threats of Flemish invasion by Philip of Flanders or Matthew of Boulogne. It also contributed to strengthening and displaying royal power in Kent during a period in which the king could not count on an archbishop of Canterbury as a reliable partner. It was designed to address internal as well as external problems in south-east England. So far as pipe roll evidence is concerned there was almost no expenditure on castles in Kent between 1154 and 1167, but there was little after 1175 too, except for the extraordinary sums committed to Dover: the second campaign there, including the great tower, costing almost £6,000 in 1180-88. This was something new and in clear contrast to 1168-1173: spending on Dover alone and on an unprecedented scale. More normal and modest work at Canterbury resumed only in 1191 after the Dover operations had come to a temporary close.49 It might be argued from the sheer cost of Dover castle in the 1180s, as well as DOVER CASTLE AND ROYAL POWER IN TWELFTH-CENTURY KENT 255 the impressiveness of the outcome, that it was rebuilt to reflect national rather than local priorities. But all of the motives which have been attributed to Henry II in constructing it also testify to importance of the local and regional setting. In the debate over military and non-military motives for castle-building, Dover great tower, and even the new defences around it, are rarely now interpreted in straightforwardly military terms. Whatever Henry II’s mounting Continental problems, and it remains debatable whether he was in serious difficulties until the very end of his life, there is little reason to propose that outside threats to South- East England in the 1180s necessitated such massive expenditure on Dover for defensive reasons. Any danger then was less pressing than in the 1170s, much less so than in John’s reign. Henry II’s castle was certainly strongly defensible, but it could have been made so at less dramatic cost. This leaves it open to argue that the great tower was demonstrative building, constructed to symbolize and provide a very visible setting for displays of royal authority, to celebrate Henry’s successful rule rather than as means to fight for it. This conclusion could be reached for Dover as an individual case, on the basis of where and when it was built, though naturally welcome to those who interpret great towers in general in this way.50 A case has also been made for seeing Dover keep as a specific architectural response of secular magnificence to the Becket cult, then attracting high status visitors to Kent, and the rebuilding of Canterbury cathedral after the 1174 fire, well advanced at the time when the castle works started.51 This may also be more compelling if placed in the context of local status and cooperation outlined here. The functional use of keeps of this period is still debatable, as documentation is almost entirely lacking and interpretations typically rely on imaginative reconstructions from surviving fabric, inevitably some more persuasive than others. Royal itineraries give some indication of when kings actually visited their castles and how often, though it is not always easy to identify consistent purpose from responses to political exigencies. It has already been observed that twelfthcentury kings had many opportunities to observe the shortcomings of Dover castle for themselves, before Henry II decided to take action on the grandest scale late in his reign. Gervase of Canterbury gives a rare specific example of the use made of another new royal castle in Kent when he described the subprior of Canterbury Christ Church going out to visit the king at Chilham in 1187, during the bitter dispute between the monks and Archbishop Baldwin. It seems Henry was conducting negotiations at his own base outside the city before entering it, the castle serving as the setting for an episode of political management.52 This is interesting, though far from a basis for arguing that he had built Chilham with just such a purpose in mind fifteen years before. In general it is better to allow for multiple uses and shifting priorities rather than trying to define a single purpose for the existence of individual castles, let alone the buildings or rooms within them. Part at least of the explanation for the extraordinary phenomenon of what was built at Dover in the 1180s must lie in the inheritance of earlier royal policies towards Kent and its castles, at Canterbury, Rochester and other sites as well as Dover. Most recently, this had led to a resumption of castle building across the county within the decade 1165-1175. But it seems that the 1180s also brought new motives and requirements, as later did events after 1200. RICHARD EALES 256 Pragmatic adaptability can be seen in almost every aspect of the reigns of the Angevin kings, and certainly in their castle policies. endnotes 1 H.R. Luard (ed.), Matthaei Parisiensis, Chronica Majora, Rolls Series, 57, vol. III, 1876, 28. For a summary of Hubert de Burgh’s career, with full references, see F.J. West, ‘Burgh, Hubert de, earl of Kent (c.1170-1243)’, Oxford Dictionary of National Biography, Oxford, 2004. 2 S.E. Rigold, ‘Excavations at Dover Castle 1964-1966’, Journal of the Archaeological Association, third series, XXX, 1967, 87-121; M. Biddle, ‘The earthworks around St Mary in Castro’, Report of the summer meeting of the Royal Archaeological Institute at Canterbury in 1969, Archaeological Journal, 126, 1969, 264-65. See also A.M. Cook, D.C. Mynard and S.E. Rigold, ‘Excavations at Dover Castle, principally in the inner bailey’, Journal of the British Archaeological Association, third series, XXXII, 1969, 54-104. The full summary of the documentary sources by Brown and Colvin, still invaluable despite growing disagreements over some of its conclusions, is in R.A. Brown, H.M. Colvin and A.J. Taylor, The History of the King’s Works, vol. II, The Middle Ages, London, 1963, 629-41. 3 The 2009 English Heritage project focused on the great tower included extensive surveys of the fabric as well as full contextual studies. It was edited for publication in 2011 and again in 2016, but has yet to appear. It was preceded by a preliminary report by C. Phillpotts, Dover Castle Great Tower: Revised Documentary Research Report, English Heritage (Properties Presentation Department), 2008. Some other more recent work has been published online. P. Dixon, The Great Tower at Dover, Philip Dixon Associates, 2009, includes a wider discussion of ‘the great tower during the 12th century, with emphasis on the design function of these buildings’. See also two titles in the English Heritage Research Department Report Series (available online or in paper copy): A. Brodie, Arthur’s Hall, Dover Castle, Kent, 40, 2011; A. Brodie and G. Higgott, Inner Bailey, Dover Castle, Kent, 41, 2011. This is by no means a full list of continuing research on aspects of Dover castle. As always, recent English Heritage guidebooks contain useful summaries: J. Coad, Dover Castle, 2007; S. Brindle, Dover Castle, 2012, and subsequent editions. 4 Biddle, ‘The earthworks …’, 264-65. 5 Brown and Colvin, in The History of the King’s Works, vol. II, 630-32, and 746 on the earlier employment of ‘Maurice’ at Newcastle in the 1170s. D.F. Renn, ‘The Avranches traverse at Dover castle’, Archaeologia Cantiana, lxxxiv, 1969, 79-92, is a valuable fuller discussion with an annotated table of recorded pipe roll expenditure on Dover at 90-92. 6 J. Coad, English Heritage Book of Dover Castle, London, 1995, summarises a range of views on Henry II’s work at Dover before the initiation of recent research, with tentative suggestions about the order of construction of the keep and inner bailey at 23-24. Renn, ‘The Avranches traverse’, 88-91, proposes the prior construction of the keep with more confidence. But on the evidence then given, none of these attempts to attribute more precise dates to the building history can be regarded as certain. 7 Brown and Colvin, in The History of the King’s Works, vol. I, 72-75, quotation at 75. It may be reasonable to attribute this to Allen Brown in view of his extensive work on Dover, and similar opinions expressed elsewhere. See his full and influential guidebook: Dover Castle, Department of the Environment, 1966, and subsequent editions. 8 C. Coulson, ‘Structural symbolism in medieval castle architecture’, Journal of the British Archaeological Association, CXXXII, 1979, 73-90, quotation and translation at 75. 9 Ibid., 75. For his more recent views, see C. Coulson, ‘Peaceable power in English castles’, Anglo-Norman Studies, 23, 2001, 69-95, with discussion of Dover at 69-75. Compare John Goodall’s views on Dover in ‘The key of England’ and ‘In the powerhouse of Kent’, Country Life, 193, 1999, 44-47 and 110-13, with the military narrative which informs his analysis of (some of) the architecture in ‘Dover castle and the great siege of 1216’, Chateau Gaillard, XIX, 2000, 91-102. 10 Among the large and growing list of ‘revisionist’ works, there are useful bibliographies and guides to debate in C. Coulson, Castles in Medieval Society: Fortresses in England, France and Ireland in the Central Middle Ages, Oxford, 2003; O.H. Creighton, Castles and Landscapes, London, DOVER CASTLE AND ROYAL POWER IN TWELFTH-CENTURY KENT 257 2002: R. Liddiard, Castles in Context: Power, Symbolism and Lordship, 1066 to 1500, Bollington, 2005; J. Goodall, The English Castle, London, 2011. 11 See especially the surveys and individual studies of Philip Dixon and Pamela Marshall. Their general conclusions are stated in P. Dixon, ‘The influence of the White Tower on the great towers of the twelfth century’, in E. Impey (ed.), The White Tower, New Haven and London, 2008, 243-75; P. Marshall, ‘The ceremonial function of the donjon in the twelfth century’, Chateau Gaillard, XX, 2002, 140-51; P. Marshall, ‘The great tower as residence’, in G. Meirion-Jones, E. Impey and M. Jones (eds), The Seigneurial Residence in Western Europe AD c.800-1500, BAR International Series, Oxford, 2002, 27-44. Other views are still possible, of course, but the shift in interpretation is clear. See, for example, the range of opinions in J.A. Davies, A. Riley, J-M. Levesque and C. Lapiche (eds), Castles and the Anglo-Norman World, Oxford, 2016. 12 R. Higham, ‘Public and private defence in the medieval South-West: town, castle and fort’, in R. Higham (ed.), Security and Defence in South-West England before 1800, Exeter, 1987, 27-49, was in some ways pioneering in asking ‘regional’ questions rather than providing a gazetteer of sites. The regional landscape approach is exemplified in R. Liddiard, Landscapes of Lordship: Norman Castles and the Countryside in Medieval Norfolk, 1066 -1200, BAR British Series, 309, 2000 and in O.H. Creighton, Castles and Landscapes. C. Coulson, ‘On crenellating, in Kent and beyond: a retrospection’, The Castle Studies Group Journal, 21, 2007-8, 189-202, deals with Kent in the later middle ages. 13 F. Pollock and F.W. Maitland, The History of English Law, 2nd edn, Cambridge, 1898, vol. I, 187. 14 On these events, see R. Eales, ‘Local loyalties in Norman England: Kent in Stephen’s reign’, Anglo-Norman Studies, VIII, 1985, 88-108; and on the Boulogne connection, H.J. Tanner, Families, Friends and Allies: Boulogne and Politics in Northern France and England c.879-1160, Leiden and Boston, 2004, 181-243. 15 D.J. Cathcart King, Castellarium Anglicanum, New York, 1983, vol. I, l-lv, quotation at li. 16 The texts of these treaties are conveniently printed together in P. Chaplais (ed.), Diplomatic Documents Preserved in the Public Record Office, vol. I, 1101-1272, London, 1964, 1-14 (nos 1-4). See also E. Oksanen, Flanders and the Anglo-Norman World 1066-1216, Cambridge, 2012, on the long-running diplomatic engagement behind these agreements. 17 J. Le Patourel, The Norman Empire, Oxford, 1976, 163-78, brought together a discussion of such Anglo-Norman ‘transfretations’. Oksanen, ibid., 158, points out that it was only in the final years of his reign that for Henry II ‘Dover to Wissant and back was among the king’s favoured Channel crossings’. For further discussion, see D. Gilmour, ‘Bekesbourne and the King’s esnecca 1110-1445’, Archaeologia Cantiana, cxxxii, 2012, 315-25. Ecclesiastics, on the other hand, usually preferred the shorter passage via Kent. 18 F or the background to this, see R. Eales, ‘An Introduction to the Kent Domesday’, in The Kent Domesday, London, 1992, 1-49; C. Flight, The Survey of Kent: Documents relating to the survey of the county conducted in 1086, BAR British Series, 506, Oxford, 2010. 19 Eales, ‘Kent Domesday’, 45-49; ‘Local loyalties’, 94-98. On the tenants of the archbishopric, D.C. Douglas (ed.), The Domesday Monachorum of Christ Church Canterbury, London, 1944; F.R.H. Du Boulay, The Lordship of Canterbury, London, 1966. On Tonbridge, J. Ward, ‘The lowy of Tonbridge and the lands of the Clare family in Kent, 1066-1217’, Archaeologia Cantiana, xcvi, 1980, 119-31. 20 H. Hall (ed.), The Red Book of the Exchequer, Rolls Series, 99, vol. II, 1897, 613-18, 706-12, 717-22. On the wider background, J.S. Moore, ‘Anglo-Norman garrisons’, Anglo-Norman Studies, XXII, 1999, 205-59. 21 S ee note 19 above. F. Barlow, William Rufus, London, 1983, 53-98, and W.M. Aird, Robert Curthose Duke of Normandy (c.1050-1134), Woodbridge, 2008, 99-117, are the fullest accounts of the 1088 revolt. 22 Eales, ‘Kent Domesday’, 27-34. D. Keene, ‘The South-East of England’, in D.M. Palliser (ed.), The Cambridge Urban History of Britain, vol. I, 600-1540, Cambridge, 2000, 545-82, covers wider issues. 23 E ales, ‘Local loyalties’, 102-105. RICHARD EALES 258 24 C. Johnson and H.A. Cronne (eds), Regesta Regum Anglo-Normannorum 1066-1154, vol. II, Regesta Henrici Primi 1100-1135, Oxford, 1956, 203 (calendar), 356 (text). Henry subsequently confirmed the archbishop’s right to exact castle-guard service at Rochester, ibid., 230. On the implications of the Rochester grant, see I.W. Rowlands, ‘King John, Stephen Langton and Rochester castle, 1213-15’, in C. Harper-Bill, C. Holdsworth and J.L. Nelson (eds), Studies in Medieval History Presented to R. Allen Brown, Woodbridge, 1989, 267-79. 25 D.F. Renn, ‘Canterbury castle in the early middle ages’, in P. Bennett, S.S. Frere and S. Stow (eds), The Archaeology of Canterbury, vol. I, Excavations at Canterbury Castle, Maidstone, 1982, 70-88, broadly accepted by more recent writers. The order of construction in Kent (Canterbury, Rochester, Dover) is precisely the opposite of later medieval priorities in maintaining these sites. 26 John of Salisbury commented on the fact that Stephen did not victimise Theobald’s ‘friends or fideles’, M. Chibnall (ed.), The Historia Pontificalis of John of Salisbury, London, 1956, 42. He and other members of the Becket circle later came to adopt a very different view of Henry II. On the earlier precedents, see B. Golding, ‘Tribulationes Ecclesiae Christi: the disruption caused at Canterbury by royal confiscations in the time of St Anselm’, in R. Foreville (ed.), Les Mutations Socio-Culturelles au Tournant des XIe-XIIe Siecles, Paris, 1984. 27 As pointed out by Rowlands, ‘King John’, 268-69. 28 J.C. Robertson (ed.), Materials for the History of Thomas Becket, Rolls Series, 67, vol. III, 1877, 131. This account of William Fitzstephen brackets Rochester with the castles of Canterbury, Dover, Saltwood and Bletchingley (Surrey), a possession of the Clare family. 29 Cathcart King, Castellarium Anglicanum, vol. I, 227-42, provides a map and outline account of these sites in Kent. More could be done with the evidence. 30 S ee note 19 above. Du Boulay, Lordship of Canterbury, 85-87, surveys the history of relations down to the agreement between Archbishop Boniface of Savoy and Earl Richard de Clare in 1259. 31 S ee E. Amt, The Accession of Henry II in England: Royal Government Restored 1149-1159, Woodbridge, 1993; G.J. White, Restoration and Reform 1153-1165: Recovery from Civil War in England, Cambridge, 2000. 32 Eales, ‘Local loyalties’, quotation at 108, and on William of Ypres 100-102. See also Amt, Accession of Henry II, 82-93, chapter on ‘The Anglo-Flemish community’; Tanner, Family, Friends and Allies; Oksanen, Flanders and the Anglo-Norman World. 33 E ales, ‘Local loyalties’, 106. 34 W. Stubbs (ed.), The Historical Works of Gervase of Canterbury, Rolls Series, 73, vol. I, 1879, 158. D. Crouch, The Reign of King Stephen 1135-1154, London, 2000, 278-89. 35 Notably William of Newburgh, Robert of Torigni and Ralph Diceto, as well as Gervase of Canterbury: references in Amt, Accession of Henry II, 90. 36 Amt, Accession of Henry II, 84-87, 91. 37 E ales, ‘Local loyalties’, 107-108, Tanner, Families, Friends and Allies, 200-203. Oksanen, Flanders and the Anglo-Norman World, 33-34, 100-102, argues that the marriage led to a ‘brief crisis’ in relations between Henry II and Thierry Count of Flanders in 1160, largely because of Thierry’s own strained relations with his son Matthew, though it was soon resolved. 38 Gervase of Canterbury, vol. I, 174, vol. II, 1880, 391 (Actus Pontificum Cantuarensis Ecclesiae); Herbert of Bosham, Materials for the History of Thomas Becket, vol. III, 251-52; represent the Canterbury tradition of these matters. F. Barlow, Thomas Becket, 1986, 83-84, 98, gives full references to other sources. 39 W .L. Warren, Henry II, 453-59, quotation at 455. Warren does describe these issues as ‘peripheral’ at 459, but that is a judgement of selective hindsight. 40 A . Duggan, Thomas Becket, 33-37; W. Stubbs (ed.), The Historical Works of Master Ralph de Diceto, Rolls Series, 68, vol. I, 1876, 308-12. Thus at 35, Duggan says ‘there is no doubt that Henry was not justified in this action’, in retaining Saltwood, since it ‘belonged to the archbishopric’. This was admittedly the strongest of Becket’s claims, despite some confusion over the division of the fees there. 41 Barlow, Thomas Becket, 83-84, 88-89. On some of the complications which led to compromises over Tonbridge, see Du Boulay, Lordship of Canterbury, 85-87, 295-96, 334, and on Saltwood, 367. 42 Gervase of Canterbury, vol. I, 203. DOVER CASTLE AND ROYAL POWER IN TWELFTH-CENTURY KENT 259 43 W. Stubbs (ed.), Gesta Regis Henrici Secundi, Rolls Series, 49, vol. I, 1867, 43-45. Warren’s account of the 1173-74 revolt in Henry II, 117-36, is now succeeded by that of M. Strickland, Henry the Young King 1155-1183, New Haven and London, 2016, 151-205, which is fuller and more alert to the motives of all the participants. At 194, Strickland argues that there was, at least briefly, a real threat of a Flemish invasion of Kent, at the time of Henry II’s famous Canterbury penance for the murder of Thomas Becket in July 1174. But the case is boosted by the hagiographer William of Canterbury’s natural desire to magnify dangers from which the region was delivered by the martyr (such as the implausible claim that ‘King Henry ordered the people of Kent to remove their belongings beyond the Medway for fear of the invading forces’) and evidence otherwise is thin. The only force to leave Flanders at this time was directed instead to the North-East, landing near Durham. 44 Rowlands, ‘King John’, 270, comments on the power of the Cornhill ‘dynasty’, though this has to be seen in the context of other local interests. 45 T he basic information is in I.J. Sanders, English Baronies: a Study of their Origin and Descent 1086-1327, Oxford, 1960, 36 (Cogges/Arsic), 45 (Folkestone/ Avranches), 31 (Chatham/Crevequer), 135 (Patricksbourne/Patrick). See also the notes on the descent of individual Kent baronies in Flight, The Survey of Kent, 245-63. The Kent returns to the 1166 royal inquiry into knights’ fees have now been re-edited and annotated in N. Stacey (ed.), Cartae Baronum, Pipe Roll Society New Series 62, 2019, 5-13. 46 The Great Roll of the Pipe for the Twenty-first Year of the Reign of King Henry the Second, Pipe Roll Society, 22, 1897, 109, 112. 47 Sanders, English Baronies, 111. Brown and Colvin, in The History of the King’s Works, vol. II, 613. 48 The tabulated figures are in R. Allen Brown, ‘Royal castle-building in England 1154-1216’, in Castles, Conquest and Charters: Collected Papers, Woodbridge, 1989, 55-57. Renn, ‘The Avranches traverse’, 86-89, gives more details on the Dover entries. 49 R . Allen Brown, ibid., 56, 58-59. 50 See note 11 above, and many of the chapters in Impey (ed.), The White Tower, on the general issues. Strickland, Henry the Young King, 301, suggested that ‘repairs’ at some castles in 1183 reflected Henry II’s renewed fears of insurrection in England as his sons again revolted against him in France, but it is hard to make such a case for Dover, given the intensive building there all through the decade. 51 S ee note 9 above for these views as advanced by Charles Coulson and John Goodall. It has been suggested, speculatively but plausibly, that Henry II’s hastily arranged reception at Dover of Louis VII, when the French king made a pilgrimage to Canterbury to seek the recovery of his sick son (Philip Augustus, who was to succeed him in 1180) may have precipitated his decision to rebuild Dover Castle. 52 Gervase of Canterbury, vol. I, 353.