8. Romney Marsh in the Early Middle Ages Nicholas Brooks This paper was originally printed in Rowley, R. 7.(editor) The Evolution of Marshland Landscapes (Oxford University Department for External Studies, 1981), 74-94 and is here reprinted in a revised form by permission of that department. 17,300 acres of land; they also had extensive properties Introduction in Walland and Denge Marshes and around their The evolution of Romney Marsh has been such a fertile manors at Appledore and the Isle of Oxney. Other ground for antiquarian conjecture and controversy over Kentish monastic houses, such as St. Augustine's, the last century and a half that armchair historians Canterbury, and Bilsington Priory, were also major might have learnt to leave well alone. But in the last lords in the Marsh (Smith 1943; Neilson 1928). The generation major advances in our understanding of the charters and estate-records of these houses, published technical processes by which the marsh has been formed and unpublished, provide an enormous corpus of have at last provided a more secure framework into material concerning the development of drainage and which the historical evidence needs to be fitted. The the many forms of land-use in Romney Marsh from the work of W. V. Lewis in the 1930s and more recently of mid-twelfth century. Supplemented from the thirteenth J. Eddison has transformed our knowledge of the shingle century by the records of royal statute and central beaches that comprise the Dungeness headland and of government and by the archives of the Cinque Ports, the way that changes in sea-level have contributed to the there is material here to keep several research students formation of the headland (Lewis 1932, 1937; Lewis and busy for many years. As yet the surface has only been Balchin 1940; Eddison 1983a). The second crucial scraped (Dugdale 1772; Holloway 1849; Teichman contribution was made in 1968 when R. D. Green Derville 1936; Smith 1943). Detailed and well published for the Soil Survey of Great Britain his final documented studies of the development of the marsh and full report on the area to seaward of the Royal landscape from the mid-thirteenth century or earlier are Military Canal. For the first time the complexity of the certainly both possible and urgently needed. marsh soils was revealed, defined and mapped; their Romney Marsh is also surprisingly well-documented stratigraphical relationship was established, and many in the early Middle Ages. The archives of Christ Church, of the technical problems of the variations in land-levels Canterbury, and to a lesser extent of St. Augustine's within the marsh and of the relics of former creeks and have preserved a series of charters concerning lands in water courses were solved (Green 1968). Then in 1980 the Marsh from the late seventh century onwards. Many Professor Cunliffe used these major contributions, of these charters were studied by the great Kentish together with scattered nuggets of archaeological, topographer, Ward, in a series of articles published in geol.3gical and historical information, to propound a Archaeologia Cantiana between 1931 and 1952 (Ward bold interpretation of the geomorphological changes 1931a, 1931b, 1933a, 193313, 1936, 1940, 1952). When that have occurred since neolithic times. As his synthesis we recall that Ward was working without any of the claimed to be no more than a 'preliminary' model, guides and handbooks that ease the path of the modern which required testing and refinement (Cunliffe 1980a, scholar through the difficult field of Anglo-Saxon especially p. 47), it may be a propitious time to set out charters (Stenton 1955; Whitelock 1955, 1979; Sawyer some of the historical evidence that can help to deepen 1968; Brooks 1973), it is astonishing how far his and to delimit our knowledge of the development of the enthusiasm and his eye for detail enabled him to seize landscape of the marsh. upon the significance of the topographical and economic Of all the marshland areas of Britain Romney Marsh information in the charters. Inevitably, however, there offers the best prospect for the historian to work were mistakes; moreover Ward never attempted any alongside the geologist and the archaeologist because it synthesis of the whole body of the charter evidence in the is uniquely well documented. In the Middle Ages the light of the evidence of place-names and of the archbishops and the cathedral community of Canter- Domesday survey and related documents. Now that bury were the dominant landowners in Romney Marsh Green and Cunliffe have confirmed some of Ward's proper, that is the marshland to the north-east of the findings and challenged others, it is time for the evidence Rhee Wall, where they held no less than 7,140 out of the to be examined afresh. The Early Middle Ages 9 1 .... I\B ............. ............ .1 T .... C .. .......................... ................ ./ ............ :/ / ........ ... 'f ... ....... [. ..... I -.. ....... ' W W A LrL A N D : .... \'.""........... "A'." " "' .. I ...... I I C.. ..... .\ ,. ................. I ........ I 0 KMS 5 .... (........ ....... 111111 ....... ....... t /...... ........,... .. ...... .* -r.. ......-. ........ C / .. / I- DecaLcif ied Marsh Calcareous Marsh A Fig. 8.1 Romney Marsh and the Dungeness headland. Fundamental to any understanding of Romney Calcareous (New) Marsh did not all occur at one time: Marsh in the early Middle Ages is Green's distinction that of the Hythe estuary is attributed to the Middle- or between the 'Calcareous' or New Marshland, which has Late-Saxon period, that of Walland and Denge Marshes been subjected to inundation by the sea within historic to the late Middle Ages and early modern period (Green times, and the 'Decalcified' or Old Marshland from 1968, 30-44; Cunliffe 1980a, 47-52). Another signifi- which the calcium has largely leached away after cant proposal was indicated in Green's mapping of the centuries of natural drainage (Fig. 8.1). Green and alluvial beds of two supposed former courses of the Cunliffe have argued from good evidence that some or northern branch of the river Rother: one finding its way all of the old marshland already existed as land available to the sea at West Hythe, the other at Romney. Green for colonisation between the first century BC and the identifies these two branches with the two courses of the first century AD (Green 1968, 18, 27; Cunliffe 1980a, river Limen recorded in pre-conquest charters (but 43-4). With a change in environmental conditions, this compare Eddison 198313, 54-6). Their meandering old marshland itself came under threat in late Roman courses establish that both are natural watercourses; but times. For Green detected in the decalcified marsh an it should be observed that the Romney branch cuts elaborate system of tidal creeks which seems to be related through the Old Decalcified Marshland, whilst the to the routes taken by the rivers Brede, Tillingham and northern branch runs through New Calcareous the two arms of the Limen (Rother) towards a wide Marshes. It has still to be determined whether this estuary which made its way to the sea past the Roman difference has any implications for their relative fort at Lympne (Cunliffe 1980a, Fig. 19). chronology. It is also argued that the reclamation of the areas of Mention must also be made of the extensive layer of peat, deposited in the second half of the second Hope) (but see Tatton-Brown 1988). We may therefore millennium BC, which extends over much of the western suspect that Fig. 8.2 reflects in part boundaries in half of the marsh, most thickly just to the south of existence since the early Middle Ages. Certainly the Appledore (Green 1968, 14-15). It is possible that a shape of the parishes seems to reveal some fundamental medieval rise in the level of the sea, together with the distinctions in the settlement pattern. To the north a desire to improve the marsh pastures, led to more and series of villages is situated on the much eroded cliff more elaborate drainage-schemes. Drainage in turn overlooking the marsh, and their parish boundaries would have caused the peat to contract and the level of extend south so that an area of marshland is included in the land to drop. As a result of one or more of these each of their territories - Appledore, Kenardington, factors, sea- and river-walls had to be built even higher. Warehorne, Orlestone, Ruckinge, Bilsington, Bonning- The familiar vicious circle of peat-marsh management ton, Hurst and Lympne. The parishes that lie wholly in provides one possible explanation for the growing the marsh divide clearly into two types: north-east of the concern shown in the medieval records with the ancient Rhee wall, in Romney Marsh proper, despite and customary obligations of marsh-landowners to build innumerable minor irregularities, the parishes form up the banks of sea-defences and of water-courses. This consolidated blocks, very approximately circular in lex marisci is already mentioned in charters from the early shape with the parish church in a central position twelfth century (Holloway 1849,669. It may have been (Burmarsh, Eastbridge, Newchurch, Blackmanstone, that the subsidence of the land caused by effective Orgarswick, St. Mary-in-the-Marsh, Hope, Snave); but drainage made this part of the marsh so vulnerable, south-west of the Rhee wall the parishes are especially when it was combined with changes in the predominantly of long, narrow form with straight barrier beaches and possibly with a gradual rise in sea- boundaries which follow