Medieval settlement and society in the Broomhill area and excavations at Broomhill church

Introduction Broomhill lies towards the southern side of Walland Marsh in a position which has always been very vulnerable to flooding. During the last one thousand years the land here has been inundated on many occasions by water coming either directly from the sea to the south, or through the Wainway Channel to the north. A high sea wall has since been built to protect the marsh, and now stands some five metres above the level of the land behind it. A combined study using both documentary and archaeological evidence has sought to record some of the episodes of reclamation and flooding. The Broomhill area is particularly well covered by a series of documents from the late 12th and first few decades of the 13th centuries. During this period the Abbeys of Roberts- bridge and Battle were very active in purchasing land which had already been enclosed, and reclaiming further areas. In total more than 140 charters and analogous documents survive from this period. This very valuable documentary collection can be compared with, and may be supplemented by, the results of archaeological work. Survey and excavation work around the site of Broomhill church has revealed the plan of the building and identified physical remains of flooding in the area. Deep deposits of sediment have been found to overlie the final phase of the church. buildings, and these must extend to cover a considerable area around. The Topography of Broomhill The geography of the land in the vicinity of Broomhill has been transformed since the medieval period, both because of repeated floodings, and because of the very considerable erosion which has taken place along the coast to the south. The retreat of the coastline can be followed on maps from the late 16th century, and quite clearly erosion would have been taking place before then. In the mid-16th century, when thesea was making very considerable inroads, it was reported that a bank called Old Camber Head, on which a hermitage had stood, had been washed away. ' Before it had been finally removed by the sea, it had stood half a mile out from the shore, yet the same hermitage in the 13th century had been part ofthe mainland.2 Nearer Broomhill, at Wayes End even within the living memory of a 16th-century witness, erosion had caused the coast to retreat 200 metres or more inland.3 During the medieval period the most significant change in the coastline was the gradual erosion and destruction of the site of the town of Old Winchelsea, which lay on the coast west of Broomhill. The town had grown up around a port, which had been founded probably in the early l l th century, and was certainly in existence by the 1040s when Harthacanute granted or confirmed to the Abbey of Ftcamp, "two parts of the tolls in the port which is called Wincenesel" (Matthew 1962, 19-20). It was one of a number of harbours which developed along the south-west Kent and east Sussex coast at the time to serve an increased level of trade with France. The geography of the Winchelsea area is very problematic, since the town has been swept away so completely that even its position has been lost (Fig. 10.1). A number of charters, however, provide some evidence of the area. The Abbot ofBattle, in an undated charter, claimed any whales or porpoises which might be stranded within the bounds 0fBroomhil1.~ He described such animals as those whales and porpoises which (i) by entry of the port of Winchelsea went to the bounds of Kent and Lewes which are called Chene, or (ii) by entry of the said port went as far as Hamelesforde and Swanysmerke. In addition to the two channels implied there, a further charter, probably of the early 13th century, mentions another, "the great fleet which goes towards Rye".5 At this period then, Winchelsea stood towards the seaward end of a system of creeks or 'fleets' as they are called locally, which drained the southern part of Walland Marsh. The situation had altered by 1258 at the latest, when the River Rother had changed course and also debouched into the sea through the channel at Winchelsea (Eddison 1983, 54; Tatton-Brown 1988). A water course running between Appledore and Winch- elsea, which would have been the line of the Rother, is mentioned in a charter of Henry III.= Winchelsea, it seems, had grown up near a channel through the shingle by which some streams from the marsh reached the sea: its place-name indeed refers to a river (Mawer and Stenton 1930, 537-8). With the capture of the River Rother, the channel may have become a sizeable Medieval Settlement and Society in the Broomhill Area estuary. The location of Winchelsea may well have been comparable with that of New Romney on the other side of Romney Marsh, which was likewise sited near the entrance of an inlet and had been founded at about the same time (Tatton-Brown 1984, 26-7). Certainly such positions would have provided sheltered anchorages for shipping on what was otherwise an exposed and dangerous coast. At the eastern side of the entrance to the channel7 there was a hermitage which served as a lighthouse to guide ships up to the port of Old Winchelsea at night (Homan 1938, 219). It must have been in a more protected position than Old Winchelsea itself, for this was the building which survived in an increasingly attenuated position until the mid-16th century, when it was described as lying on a shingle bank called Old Camber Head, half a mile out to sea. It was then said to be in the parish of St. Thomas of Winchelsea but earlier, in the 13th century, it was in Broomhill, the lands of which extended as far as the 'great fleet'.8 Broomhill 113 then probably included the whole of the eastern bank of the Winchelsea inlet, and conversely this may indicate that the town and port of Winchelsea lay on the western side. Two major channels in the Broomhill area are implied by the charter of the Abbot of Battle laying claim to stranded whales and porpoises. The first of the fleets mentioned led up to Chene and a number of other medieval documents refer to the 'Water of Chene'.g This place-name survives in the modern Cheney Court and the channel should be sought in that area. It seems probable that the 'Water of Chene' was the same as the Wainway Channel, a substantial fleet which is shown as a prominent feature on early maps of the Marsh. Even in the 16th century it was sufficiently wide and deep for lighters to use to transport goods towards Lydd.Io The course of the second of the fleets in the Abbot's claim has now been largely lost through marine erosion, but its route can be identified with a channel shown on the 1596 Symondson map of Kent. The stream, Jury's ........ ........ ........ Spikespich Swanmere - L Fig. 10.1 A reconstruction of the Broomhill area in the early 13th century, based on documentary evidence. 114 Mark Gardiner Gut Sewer, did not empty formerly into the sea directly, but flowed westwards parallel with the coast behind a shingle bank, and reached the sea through the Winchelsea estuary. Small lengths of this fleet still survive, where it has not been lost through the cutting back of the coast, and it can be traced on the Soil Survey Map and indeed on the ground (Green 1968). Hamelesford, also referred to in a number of charters, must have been the site of a ford across the fleet and Swanmere an extensive area of marsh along its course (Historical Manuscripts Commission 1925, 72, no. 441). The Broomhill area was bounded on the northern side by the Wainway Channel and to the west it extended as far as the Winchelsea estuary and the great fleet. On the east, the parish of Broomhill straddled the county boundary and included an area of Kent called Pedling which lay in the lordship of Old Romney and it also was bounded by the Battle Abbey manor of Dengemarsh. In the south of Broomhill a number of 13th-century documents mention the tautological place-name Capenesse.I2 It must have been on the coast, and is probably to be placed in the vicinity of Jury's Gut. This is too far west to have been Dungeness by another name, rather the place-name suggests a second promontory on the coast nearer Old Winchelsea. Twelfth- and Thirteenth-Century Inning at Broomhill Its vulnerability to flooding did not make Broomhill immediately attractive to medieval settlement. It is probable that this area was initially held as appendant pasture land belonging to estates lying on the higher ground beyond the marsh in Sussex. Brooks (1981,78-9; 1988) has noted that some of the upland parishes in Kent held outlying pieces on the marsh and has concluded that .this reflects the pattern of Anglo-Saxon estates. A number of pre-Conquest charters refer to detached pasture lands on the marsh which belonged to even quite distant estates. Though the evidence is less good for Sussex, it seems likely that there was a similar arrangement there. A corrupt Anglo-Saxon charter describes the bounds of an estate at Icklesham. It is now impossible to locate the bounds, but the places mentioned indicate that the estate included a substantial area of marshland below the upland core (Sawyer 1968, no. 108). Part ofBroomhil1 seems to have been originally associated with Fairlight, which lies beyond the western side of the marsh. An area of Broomhill was considered to be parochially part of Fairlight parish, presumably that piece of land which lay within the fee of Fairlight. ' When Battle Abbey first acquired land in Broomhill in the early 12th century, and certainly before 1124, its attraction was said simply to have been that it was close to their estate at Dengemarsh (Searle 1980, 120-21). However, by the end of the 12th century both Battle and Robertsbridge Abbeys were beginning to extend their holdings and take a more active interest in the area, because of the potential of the region for large-scale reclamation. They had been preceded in their work by earlier enclosers who had probably inned the land which could be most easily drained. Little is known about this early period of inning. It seems to have been carried out by local entrepreneurs, wealthy tenants who obtained land by charter and then constructed banks against the sea and ditches to drain the land. One of those who was active in this work was called Doudeman. He must have been a considerable local figure for he not only held land in the south ofBroomhill, but his name is also connected with an area in the north- east of the parish1* and there is also a lost place-name, Dudmanswike in Hope All Saints, near New Romney (Wallenberg 1934, 48 1). By the early 13th century his heirs occupied a large part ofBroomhil1: just one piece of their land was 500 customary acres in extent which, if measured with a rod of 20 feet which was widely used in the marsh' 5, would have been equivalent to 600 statute acres (Curia Regis Rolls 1, 467; Homan 1937). A second person active in the Broomhill area was Ellis de Ria who had enclosed part of Grikes Marsh (Curia Regis Rolls 5, 203). He is also mentioned in a charter concerning the lease of mills near Rye suggesting that, like Doudeman, he was an entrepreneur with wider interests.16 The first phase of inning by a few enterprising individuals was succeeded by a period when seigneurial control became more signficant. The Abbeys of Robertsbridge and Battle became the major land reclaimers, but groups of men working with seigneurial agreement were also active at this time. This second phase began in the early years of the 13th century. In 1208, for example, an agreement was made to divide up between four groups of men, comprising at least eight individuals, the land which they had recently enclosed under the direction ofJohn de Gestling (Curia Regis Rolls 5, 20-24). A similar agreement was made at about the same time with another group ofnine men who, working with the co-operation of the Abbey ofRobertsbridge and the men of Broomhill, had enclosed further land (Anon. 1873, 81). The areas which were inned by this co- operative enterprise are not known, but they may have been considerable for some years later, in 1222, the Abbeys ofBattle and Robertsbridge made an agreement to work together and eventually enclosed some 1,200 customary acres between Capenesse and Chene.17 The early enclosure of land began in the areas of the marsh called Grikes and Spikespich (Mawer and Stenton 1929, 190 discuss this place-name). Ellis de Ria had enclosed land in Grikes and the heirs of Doudeman are subsequently found holding land both there1 and at Spikespich where the marsh had been divided between them through partible inheritance (Curia Regis Rolls 5, 202-3). Spikespich was otherwise known as the Marsh of Broomhill, and probably therefore adjoined the place called Broomhill itself (Historical Manuscripts Com- mission 1925,74, no. 446). The place-name suggests that this was a piece of land slightly raised above the surrounding marsh, sufficiently so to be called a 'hill' (Mawer and Stenton 1930, 529). Later reclamation works extended away from these areas, first of all to the Medieval Settlement and Society in the Broomhill Area marsh of Cumerse which lay between Spikespich and Winchelsea.lg Battle Abbey was particularly active in purchasing and enclosing land in Swanmere but its interest lay generally to the western side of Broomhill where it had been granted land in the fee of Guestling by Robert de Hastings and by members of the St. Leger family.20 Its land extended as far as the great fleet going towards Rye (Historical Manuscripts Commission 1925, 68, no. 97). Robertsbridge Abbey, while holding land at Swanmere, also had land to the east ofBroomhil1 behind Capenesse and in Pedling which lay in Kent (Anon. 1863, 217). The land around Jury's Gut was called the Terrae Perjuratae from whence the modern name derives, and this lay within the lordship of John de Gestling. The first half of the 13th century was a period of great activity. The three major manorial lords sought to expand their holdings both by purchasing land from existing tenants and by enclosing further land in the marsh. The large number ofcharters from this period are a record of the systematic accumulation of many small pieces of land, often of only a few acres in extent. At the same time, the two abbeys and John de Gestling were making agreements to take in additional lands and 115 would certainly have affected Broomhill, but the only reports of the troubles of this period are given by the 16th-century antiquarian William Camden, who records that, "the sea (driven forward by the violence of the winds) overflowed this tract, and ... threw down Prom-hill", whose inhabitants, he says, were forced to flee to Lydd (Camden 1722, 274). These losses were not rapidly made up. Only in 1309 did Robertsbridge obtain a licence to make up for their losses in the marshes of Winchelsea, Rye and Broomhill, and decayed rents are still recorded in the Battle Abbey accountsz3 in 1354 because the lands held by the Abbot of Robertsbridge were under the sea (&l. Pat. Rolls 1307-1 3, 152). A considerable change in the topography may have taken place at this period as land was newly re- enclosed: there are many references to places called New Innynge and New Land.24 A reference to the site of Ealde Promhull also suggests a shift in settlement. 25 The 14th-century recovery of land is less well recorded than the initial enclosure a hundred years before, but again large areas must have been taken in, for a document of 1374 refers to 1,564 acres adjoining Newelondefflet (Cal. Inq. Misc. 3, 349). maintain the common walls of the land enclosed. ' After the middle of the 13th century, however, there are no further references to newly enclosed land. This may be a result of a lacuna in the documentary record at this period, but equally it may show that the conditions which led to flooding in the marsh subsequently were already apparent. A note of caution is first given by an agreement made between the Abbots of Robertsbridge and Battle in 1243, for it allows that payment of rents and services for land enclosed at Cumerse would cease if the marsh were overwhelmed by the sea, and until the land was re-enclosed (Ancient Deeds 2, B2965). This clause, not found in earlier charters, may anticipate the increasing number of storms which affected the area in the second half of the 13th century. Though there is little specific evidence of events at Broomhill during this period, it is possible to surmise what was happening from the events elsewhere. The south-westerly winds mentioned by Matthew Paris were apparently driving the sea up into the Winchelsea estuary (Luard 1880, 272). By the 1250s sea water was being forced up this inlet so that the level was raised and salt water was running up as far as Appledore. A higher water level would have affected the Wainway Channel which was apparently tidal, for Swanmere adjoining it is described as a salt marsh,22 and would have impeded drainage from the marsh. Around Broomhill the water levels would have been raised and would have threatened the enclosed land. The progress of storm damage at Old Winchelsea would then directly relate to events in the Broomhill area and this can be followed in the Patent Rolls and in Matthew Paris's Chronicle. Violent storms recorded in 1250 and 1252 caused extensive flooding, and in February 1288 marsh walls were broken down and lands from the 'Great Wall of Appledore' southwards and westwards as far as Winchelsea were inundated. This Fourteenth-Century and Later Inning at Newland, Midley The subsequent changes in the geography and lack of early maps for the Broomhill area make it impossible to reconstruct the progress of marshland enclosure in the fourteenth century in Broomhill itself'. However, just to the north of Broomhill in the adjacent parish of Midley lies Newland, a well-documented estate belonging to All Souls College, Oxford from about 1438, and this illustrates the character and process in a neighbouring area (Fig. 10.2). An inquiry of 136526 discusses this area and suggests much of it had not yet been reclaimed, but was covered by sea water which flowed up the Wainway Channel and submerged land as far northwards as Sea Wall (Fig. 10.2). Seyntmaryeland is mentioned as land beyond the limits of the marsh overflowed by the sea, and this must have been protected by St. Mary Wall. That wall abuts on, and is therefore secondary to, Sea Wall. This implies that the land between Sea Wall and Midley Wall had already been inned at the time of the inquiry. In the late 14th or the first half of the 15th century, further land in the Wainway Channel was gradually enclosed by a series of walls built across the head of the creek, which cut off pieces of land which were then drained. The first of these enclosures, called Newe Innynges, was constructed jointly with Christ Church, Canterbury which held land in Agney to the north-west, and with the owner of land to the south, for a single continuous wall was built to separate all these lands from the water in the Wainway. After the enclosure had been made, the land inned must have been divided up between the parties concerned, because this is very evident from the boundaries created (Fig. 10.2). An Mark Gardiner Newland, Midley - 500 m Land af Christ Church, 'newlye enclosed ' Fig. 10.2 All Souls College artificially straight ditch runs across the middle of the former channel between the land of All Souls College and Christ Church, and it can only have been laid out when the land was apportioned. Similarly, on the south side of the Newe Innynges, a line was drawn from the junction of Sea Wall and St. Mary Wall in a westerly direction until it reached a creek, to define the boundary there. These straight boundaries contrast with the line of the wall built to protect the Newe Innynges on the south- west side. This meanders across the bed of the Wainway Channel in a very erratic manner. The direction of the wall appears to have been guided by a desire to skirt deeply incised creek channels and to' take in areas of raised land in the salt marsh. Some of the embayments created by constructing the wall around three sides of the creek channels were later enclosed by building further pieces of walling across the fourth side and in the process, straightening the line of the wall. All Souls added eight acres to the 277 acres of the Newe Innynges in this way. and the land gained is described in a rental as 'newlye enclosed' (Fig. 10.2). When a map was drawn up in 158g2' the piece of land furthest to the south-west in the estate was still salt marsh. This marsh stood close lands at Newland, Midley. Newe lnnynyes 4 Sea walls - F~eld boundar~es to the head of the Wainway Channel and had not yet been adequately drained. It brought the Newland estate southwards as far as the lands held by the College in Scotney and Okholte, where further land had been inned. A similar wall had been constructed there to take in further land at the edge of the Wain~ay.~~ Though this enclosure work was late medieval in date, the methods used in the 13th century for the initial enclosure of Broomhill marsh must have been very similar. Inning had proceeded in a similar way, gradually moving outwards from existing enclosures by cutting off areas of land with walls and then draining them. One agreement envisages a situation in which the walls used to enclose marshland might become unnecessary and be demolished if further walls were constructed in the future beyond them (Curia Regis Rolls 5, 203). The methods ofwall construction have been discussed by Smith (1939) using the records of Christ Church, Canterbury. The account rolls of the Dengemarsh manor, and Scotney and Okholte manor adjacent to Broomhill, suggest that once they had been built the cost of maintaining walls and scouring the ditches was a considerable and recurring expense.29 Smith noted that Medieval Settlement and Society in the Broomhill Area 7he Men of Broovnhill 1st Robert Doudeman I Eilw~n +l I 3rd Adam 4th Ellis , John , Roger Osbert Williarn de ; Iyscou;be Williarn , , Williarn=Ernma Thornas de 1 W~nchrlese , Robert 1 , Austin Walter = Helewise Adarn in; l Thornas Niger , Osbert inr Thornas Gilbert Vincent Iienry Nicholas Luke Nicholas Reginald Bartholomew Walter Robert Fig. 10.3 Four generations of the Men of Broomhill. while Henry of Eastry was Prior of Christ Church (1285-1331) an average of 14 per cent of the annual revenue of the manor of Ebony in the north-west of the marsh was spent on embanking and draining land, though there was no overall increase in the area under cultivation. Yet against this it should be borne in mind that this was a period of particularly bad storms. Common walls were not maintained by individual manors, but by all tenants with land in the vicinity, and they contributed according to the area ofland they held. Money was likewise collected for the scouring of the watergangs, or main ditches which drained the marshlands. Settlement and Society in Broomhill In the early 13th century it is evident that settlement was in scattered farmsteads situated throughout the reclaimed land. These farms are often mentioned as boundary marks, for in the flat marshland landscape there were few other distinguishing landmarks. It is not clear that there was ever a village at Broomhill, though William Camden (1722, 274) claimed that Broomhill was, at the time of the storms, a populous village. Another antiquarian affects to recall that Broonlhill was once a place so considerable that it included more than 50 inns and taverns (Cooper 1850, 5). More worthy of consideration is a report that in the 17th or 18th century foundations of houses were visible around the ch~rch,~ ' and recently Broomhill has appeared in a list of deserted medieval villages (Beresford and Hurst 197 1,204). Even so, it must remain an open question if there actually was a village at Broomhill. The impression given of the landscape by the charters is ofa flat, featureless expanse ofarable, pasture land and shingle, with the occasional meadow and isolated farm. Shingle is mentioned in a number of charters and there was evidently a much greater expanse of it than is now found, and in the Broomhill area at least, this is because it has been subsequently overlain by further sediments. Shingle was often granted as an appurtenance to other land and occasionally was even the subject itself of a grant (Historical Manuscripts Commission 1925, 77, no. 247). This suggests that the shingle may have had a use and value as rough grazing land. An area of sand is also referred to, and there were in the locality occasional beds of peat because turbaries, whence peat was cut for use as fuel, are also indi~ated.~ From the early 13th century it is also possible to reconstruct the society on the marsh. The charters give a remarkably detailed account of one particular family, the descendants of Doudeman. The members of this family were known to their neighbours as the Men of Broomhill, because they formed such a large and cohesive group of tenants (Fig. 10.3). Like other families on the marsh they practised partible inheritance, which was the common inheritance custom in Kent and some adjoining parts of Sussex. By the third generation their marshland had been divided up between the 11 grandsons of the original encloser, Doudeman. Both the Abbeys of Robertsbridge and Battle were very interested in obtaining land from members of the family, because they held such a large area in Broomhill. The purchase of land proceeded piecemeal and Robertsbridge, in particular, had to buy up rights of the cousins to assemble its holding. In the earliest records the members of the family were treated and behaved as a group rather than as individuals, as if they considered themselves to be joint shareholders in the land, not single tenants in their own right. They prosecuted court proceedings jointly, alienated and received land ~ollectively.~ When Adam, son of Eilwin wished to sell his portion of the marsh he first gave his brothers and nephews other land in exchange for their shares34 and only then was able to complete the sale of the land they had jointly held (Historical Manuscripts Commission 1925, 734, nos. 118 Mark 444,58). John Roger, a member of the fourth generation who took his surname from his father, described his holding in Spikespich as all his portion in that marsh which he held by right of inheritance "contra coparticipes meos", against his CO-par~eners.~~ This too suggests that he perceived his land as that held out of the common share. By the 1220s the original pieces of Doudeman's marshland had been subdivided to some degree. The marsh of Spikespich had been split up into three parts which corresponded to the portions of the three brothers of the second generation, and in the earliest surviving Battle Abbey rental for the area, the Men of Broomhill paid rent in three groups.36 The land may no longer have been thought of as the common property of the whole family, for it was possible for Roger, son of Robert to buy land from Thomas and Gilbert, his nephews3' Yet other documents, apparently of the same date, contradict this and suggest that the areas of marshland were still considered as a whole. Thomas and Gilbert de Setelescumbe, for example, gave a fourth part of a third share of Spikespich to Robertsbridge, and William, son of Eilwin likewise granted one fifteenth part of the land ofBroomhill (Historical Manuscripts Commission 1925, 74, no. 446) .38 Reference to parceners, allies, brothers and joint heirs are not infrequent among rentals of land in Kent where the partible inheritance custom of gavelkind applied, but explicit detail of how such arrangements worked in practice is less common (Baker 1965, 166). It has been doubted if indeed by the end of the 13th century such groups were really groups of tenants holding land jointly, or if they were simply described as such through conservatism or for clerical convenience. The Broomhill charters, however, make clear that in the early 13th century joint tenancy was the common practice and not only was land held in common, but the Men of Broomhill even had joint shares in the mill there (Historical Manuscripts Commission 1925, 75, no. 447). This type of inheritance and tenure is an interesting survival and seems to be derived from a pattern which is thought to have been common in the Anglo-Saxon period. It seems likely that at the time of the Norman Conquest the normal inheritance custom had been partible, but was changing in the upper levels of society to an impartible custom (Holt 1982; 1983). The custom ofimpartibility slipped down the social scale towards the peasantry and except in a few areas of England, which included Kent, unigeniture was the normal inheritance type from the time manorial records become available in the mid- 13th century (Hyams 1980, 77n). Though the attitudes to land and inheritance at Broomhill in the early 13th century were, no doubt, not exactly the same as in the Anglo-Saxon period, the pattern ofsettlement on the marsh is likely to have been similar to that which has been inferred for the earlier period. In the Anglo-Saxon period it has been argued that agnatic rule of inheritance and patrilocal character of settlement would have produced clusters offarms held by kinsmen (Charles-Edwards 1972, 29). This seems to have been the situation in Broomhill too. When Adam, son of Eilwin sold land to Robertsbridge Abbey, for example, he described it as lying in width between Midston on one side and the land of Austin his brother on the other; and in length from the house and land of Thomas Niger to the land of the sons of Robert, his brother.39 Though most is known about the Men of Broomhill, they were not unique. Partible inheritance was widely practised on the marsh and indeed beyond it, so much of the land would have been occupied, like this area of Broomhill, by groups of farms of close relations4 (cf. the Men of Misleham, Tatton-Brown, 1988). The erosion of the unity of the extended peasant family and of its tenure of land has been noted throughout England in the late 12th and early 13th centuries, and this has been connected with the increasing control lords were exerting over their tenants (Dyer 1980, 86; Harvey 1984, 354-6; Hilton 1965). At Broomhill the growing interest in the area by larger landowners, notably Robertsbridge and Battle Abbeys, certainly led to attempts to strengthen seigneurial control and restrict the freedom of the tenants. This may be followed in charters and disputes concerning the marshland. The earliest surviving grant made to the Broomhill family was from Robert de St. Leger who gave land in his fee to Doudeman to hold for rent and on the condition that if any wrecks, whales or porpoises should be beached there, then Robert and Doudeman should have halfeach. A subsequent charter made by Robert de Hastings to members of the third generation of the Men of Broomhill was on similar terms.41 Yet, as already mentioned, the Abbot of Battle later claimed all such finds as his own. At about the same time, the Abbot was also arguing that the Men of Broomhill were tenants on his demesne in the manor of Broomhill with the intention of reoccupying it. They disputed this, saying they held their land freely, paying only rent. In settlement of the case the members of the Broomhill family were forced to give up their claims and received in compensation a smaller area of land (Curia Regis Rolls 1,467; 2, 102, 3 12; Salzman 1902, 2 1, 22). The importance of this case for the Abbey was that it cleared the way for extensive reclamation work in Swanmere and Chene. The success of this case allowed greater control to be exercised over the tenants , and in due course a manorial court was instituted by the Abbot of Battle.42 More specific limitations on the tenants on the marsh were imposed in an agreement between John de Gestling and men who had just enclosed marshland. It required that they should attend his court, where they might be fined up to 21- amercement, and that they should pay Id. fine on alienation of the land and relief (Curia Regis Rolls 5, 202-4). In the early years of the 13th ceiitury there was a decisive shift of power away from free men who had held land directly from the lord by charter, in favour of the lords who were extending their manorial powers and restricting their tenants' rights. Coincidental with this was a change discussed above in the initiative for Medieval Settlement and Society in the Broomhill Area marshland enclosure, which passed from individual tenants to the lords. This was a boom period in demesne farming, and land which had once been granted out to tenants was again being taken in hand. The inflation of the years around 1200 made it more profitable for lords to extend the land under their direct cultivation. The assertion of power at this time reflects the growing interest of lords in their land and the wish to extend the demesnes by enclosing large areas of marsh, and extend their revenues through the purchase of existing land. It has been estimated that Robertsbridge alone spent £1 80 in buying land in Broomhill (Chapman 1977, 170). Both the Abbeys of Robertsbridge and Battle established granges in Broomhill to cultivate the land which they had enclosed and to oversee their interests in the marsh. In the land on which tenants were already established, though, they were content to continue letting the land which they had bought (Historical Manuscripts Commission 1925, 734, no. 58).43 119 is, however, not otherwise known and does not occur in any later documents referring to Broomhill which have been examined. As a possession of an alien priory, the rectory and advowson ofBroomhill were seised by Henry V and were granted to Wye College. When in 1524 a vicar ceased to be maintained at Broomhill, divine service had already ceased there. The parish by then may have been sparsely populated, as it was from at least the late 16th century onwards, and there was little subsequent need for a ~hurch.~ Excauation Broomhill Church Documentary evidence Even after the storms of the late 13th century, Broomhill continued to suffer from periodic flooding. An inquiry of 1365 mentions the danger of flooding to Lydd and one of 1474 reported that the area of Broomhill was daily subject to the danger of inundation, and recommended repair on the sea walls (Dugdale 1772,47). It may have been this that prompted Robertsbridge Abbey in 1478 to begin a series ofgrants by which they gave up to Richard Guldeford, who was active in enclosing and draining land in the area, their interests in Broomhill (Gal. Pat. Rolls 1476-85,228). By the mid-16th century changes in the coastline were again causing concern. An area of 1,500 acres had been enclosed in Broomhill Marsh, but a few years later in 1570 the sea breached the walls and part of the newly won land was lost. This state is clearly shown on a map in 1589 which depicts an inlet of the sea which had apparently flowed up the Jury's Gut sewer, having broken through the shingle barrier. It had covered the site of Broomhill church which is described as ruined, and this is confirmed a few years later when it is described as 'an old piece of church'.44 This inlet was apparently sealed, and maps of the late 16th century show a large enclosure in the Broomhill area, though the church remained outside this. The earlier history of the church has been collected by Salzman (1937, 149-50) and this account substantially follows his history. He suggests that Broomhill church should be identified with that ofAlschot which was given by Manasses, Count of Guisnes, to the French Abbey of St. Leonard of Guisnes in 1129 X 39 together with the churches of Newington and Celpham (Dunkin 1845, 170). The Abbey is later recorded as holding the churches ofBroomhil1, Newington and Brenzett. A 19th- century history also mentions that Broomhill was called Alcotch in old deeds, though no evidence is cited in favour of this (Cooper 1850, 5). The place-name Alschot The site of Broomhill church has been marked on many maps of the marsh and survives on the ground as a low mound ofstones in the middle ofa pasture field, so that it can be readily located. It is one of the few landmarks which can be identified in a landscape which has been transformed since the medieval period. It was therefore a useful place at which to begin archaeological work in the area, and two short seasons of excavation have now been completed. In 1985 the purpose of the excavation was to locate the extent of the church and determine its preservation. 'Two trenches (Fig. 10.4, A and B) were cut across the mound which was presumed to be the remains of a collapsed wall. It became clear, however, that the mound had been created by ploughing either side and throwing stones on to the bank in the middle. Resistivity survey was used to locate the other walls, and the results confirmed by cutting a small trench, C in the north-east corner of the nave (Fig. 10.5). A second season examined the structural history of the building by excavating trench D in the north-west corner of the nave. Although only a small part of the church has now been excavated and very limited areas removed down as far as pre-church ground level, it is possible to suggest a tentative account of its history. The church has been remarkably well preserved by a layer of sediment 0.60 m. thick which overlies its floor. The upper levels of this have been reworked into ploughsoil, and subsoiling has penetrated deeper in places. Bedded within the layer of sediment is a band of tiles from the collapse of the roof which must have taken place when some of the sediment had already been laid down. Within the mass of fallen tile it was possible to isolate the line of a roof timber, now rotted. It is therefore likely that the roof collapsed at a time when the church was flooded by the sea. Both before the sediments were deposited, and while they were being laid down, the windows of the church fell in, for stained glass with lead cames was found lying both directly on the floor of the church and within the overlying material (Fig. 10.6). Also on the well-preserved floor of the church was a thin layer of dirt, which would have accumulated after the building went out of use. The lower part of the walls of the church has been little disturbed and stands at least up to two courses high and, beneath the mound where they are protected, to a greater height. Four phases of church building could be Mark Gardiner Rroo~zlzill Church: final plzrlsc~ robbed wall l I .i _... I l l I I I * ,\ t 1- > C ' / \ ' ', A Fig. 10.4 Broomhill Church: a plan of thejnal phase from the excavated evidence and resistivity survey. identified in the area which has been excavated, and there was evidence for activity preceding the first construction phase. The natural soil underlying the church was a combination of shingle and sand on top of which a thin layer of sediment had been laid down, and which contained very occasional pieces of charcoal and sometimes shell. A ditch (1 1 1) located in one section here also belongs to this period (Fig. 10.7). The pottery from this layer is typically from reduced cooking pots in a hard, sandy fabric tempered with occasional pieces of shell, and a few pieces from glazed jugs were also found. Similar pottery has been noted at Eynsford Castle and that site, though a good distance from Broomhill, provides the best source of comparative material (Rigold 1971). Pottery of Rigold's SSa type was found at Broomhill with rim forms characteristic of Eynsford periods B to D, which are dated there from the mid- 13th to the beginning of the 14th centuries. The occasional flint-tempered sherd from Broomhill would also fit into this period. The pottery assemblage here contrasts with the medieval material from excavations in New Winchelsea in which Saintonge Ware, Rye Ware and 'Winchelsea' Black Ware were the major components, and again emphasises the earlier date of this level at Broomhill (Orton, forthcoming). Cut through this layer was a ditch (81) 0.50 m deep and tightly packed with gravel (82). This had been intended as a foundation for a large masonry building, presumably a church, but there was no evidence that any structure was ever erected on this base. There was no trace of mortar on the gravel, nor could any building debris be specifically associated with it. It appears to be the remains of an abortive attempt at the construction of a church building. More substantial remains survive from the second phase. This is represented by a wall built of roughly shaped sandstone laid on a foundation of alternate layers of gravel and clay. Traces of this (127) survived under a later floor of the church, and part was incorporated as a buttress (64) in the third phase building. From this third phase the whole of the north wall and part of the west wall survives (Fig. 10.8). The craftsmanship of this period of construction is crude, and the church must have had a heavy and inelegant appearance with its close buttressing. The north wall was almost one metre thick and was built with an outer facing (119) of unknapped flint pebbles with occasional water-rounded sandstone blocks. The core of the wall (1 18) was of Medieval Settlement and SocieQ in the Broomhill Area 121 Fig. 10.5 A dot density plan of Broomhill Church (from a resistivity survey by Andrew Woodcock and Mark Gardiner, June 1986). rubble construction and was made of similar materials. The corner (120) had evidently been built first and was laid more carefully, but on a sl&htly different alignment to the remainder of the wall. The foundations ofthe wall had been made by digging a trench and filling it with a concrete and flint pebble mixture. Similar foundations were found beneath a large base in the north-west of the church (36) and extended 0.8 m below the floor level. Clearly such massive foundations were intended to take a substantial load, and the most likely explanation is that this was a pier for a west tower standing over the nave. The tower may also have been supported by the west wall and a second base which adjoined it (125). Around the outside of the church, in trench D, was a considerable depth of material that had accumulated during construction. The building debris occurred in a series of thin layers, some no more than lenses of material, and consisted of very fine gravel, sand and quantities of roof tile and slate. Among this debris was found a number of small pieces of glazed Rye Ware, the majority apparently from jugs and frequently decorated by combing or horizontal grooving. The final phase of the church was an extension made on the south side by piercing the south wall, replacing it Fig. 10.6 Trench D. Fragments of a stainedglass window lying on thejoor of the church where they had fallen before marine jooding took place. Scale length: 30 cm. 122 Mark Gardiner Fig. 10.7 Broomhill Church: trench D. with an arcade, and adding a south aisle and doorway. The wall was finely made of knapped, coursed flint pebbles and survived up to one metre high, protected by the accumulated soil in the mound. The inner face of the wall had been faced with plaster, areas of which, remarkably, had survived in situ. Scratched into the plaster by the doorway was a graffito of a ship (Fig. 10.9). Drawings such as these are not uncommon in churches and in this case may be associated with a shrine to a saint connected with seafaring, such as Saint Nicholas I I Fl~nt e* 0 Sandstone I I Mortar (Dr. D. Jones-Baker, personal com- munication). The standard of building in this fourth phase was much higher than in the earlier periods. The doorways were decorated with well-carved mouldings, (Fig. 10.10), and the bases of the aisle piers were neatly made octagonal columns with a capping ofstone (Fig. 10.11). Though elegant, these columns were not adequate to support the arcade. They had been subsequently encircled by a very crude surround oflarge stone pebbles Medieval Settlement and Socze~ zn the Broomhill Area --W-' , ,.h. ' v , . A $ + = = . ; : *- 123 . :*-- .c "V , W%- - Fig. 10.8 Trench D. The north-west corner of the nave. Foreground: seclion lhrough thegrauel foundations of lhejrsl phase. The third phase has been robbed oul in lhe foreground but .rurvive.r elsewhere. Scale length: I m. Fig. 10.9 Ship grafito from Broomhill Church. 124 Mark Gardiner Fig. 10.10 Trench A. Stone moulding on the west side of the south doorway, surrounded by a dense scatter of tile fallen from /he roof Scale length: 20 cm. Fig. 10.11 Octagonal pillar base with a crude stone surround. Scale: I m. Medieval Settlemen1 and Socie~ in the Broomhill Area 125 Fig. 10.12 Trench B. An octagonal aisle pier with crude stone surround. The dense scatter of tile from [he roof can be seen on the right and in section. Scale lenglh: 2 m. to give additional strength (Fig. 10.12). Another late modification to the church was the construction of a low stone bench (132) found in trenches B and D around the inside of the north, west and south walls, to provide rudimentary seating. Discussion From the documentary and archaeological evidence it is possible to suggest an interpretation of the history of the church and the surrounding area. No evidence of structures or any remains have been found which could be dated before the mid-13th century. This may simply reflect the limited area which has so far been excavated, and future work may locate earlier church building. It may, on the other hand, suggest that the site of the church had been moved, and it is tempting to associate this with the storms and flooding in the later 13th century. The cooking pots in shelly-sandy fabric and other finds together with the ditch (1 1 l), which predate any church buildings so far located, may be of a domestic character and not related to ecclesiastical usage. The church may then have been moved to a site close to existing buildings. Further archaeological and historical work should enable us to confirm or rebut this suggestion. The first attempt at building work was apparently abortive and was given up after considerable effort had becn cxpcnded in digging deep foundations. When the second-phase wall was built, it did not use the existing foundations although it was constrlirtrd on the same alignment. Little can be said about the second-phase church until a greater area has been dug, but it may not have stood for long for it was replaced by the third-phase church. This event cannot yet be precisely dated, as the Rye pottery only suggests the broad period of the 14th or early 15th century. On the dating given to phases two and three will depend the period of building with which a carved stone head is associated. This was found, badly weathered, on the surface of the field (Fig. 10.13). It cannot therefore aid the chronology, though the carving itself may be compared with work in other churches which would suggest a date in the first half of the 14th century. The fourth phase, in which the church was enlarged by the construction of a south aisle, can be dated on the basis of the constructionaI type used and the stone mouldings. It belongs to the early 15th century, at which time the parish must have been populous and had sufficient resources to fund the expansion of the church. The confidence and wealth implied by tIlc investment in building at this time is in contrast to the situation just a hundred year later when church services had already ceased and a vicar no longer served there. Thr windows fell in, rubbish accumulated on the floor, and in due course the sea swept in, and in sheltered conditions 126 Fzg, 10.13 Stone head from Broomhill Church. (drawing by Lys Drewett) Mark Gardiner with surviving churches, it is evident that they are accurately drawn. The church shown at Broomhill is illustrated with a west tower, reinforcing the archaeolog- ical evidence for such a structure. Conclusion The excavation of the church is intended as a starting point for a wider study of the area. The deposits found at the church suggest that the surrounding land may be covered under a layer of sediment which is sufficiently deep to protect it from the attrition of ploughing. In this zone of deposition, an entire late medieval and early modern landscape may be buried and preserved underneath a blanket oflater deposits. Resistivity survey has already proved an effective method of locating archaeological remains, and a survey of the church area was able to detect not only substantial stone walls, but also a bank and ditch which encircles the north side of the building and may be the boundary to the churchyard (Fig. 10.5). When the survey is extended beyond this, it may be possible to establish if there was a village around the church and locate other elements of the marshland landscape such as ditches, marsh walls, metalled tracks and further settlements. The history of settlement in Broomhill is more complex than in many areas. In other places, changes in settlcrncnt can be interpreted as a result of the altrring rconomic and social situation. Broomhill, however, lay at the edge ofland which could, in favourable periods, be settled through enclosure from thc sea, but when conditions deteriorated lay beyond what could be deposited sediment on top of the iloor. With exposure to the weather, the roof timbers rotted and collapsed. Then, after the land was reclaimed from the sea, robbing began on the walls. The building material in the ploughsoil does not account adequately for the entire fabric of the church, much of which must have been taken for use elsewhere, and the west wall in particular has been partly removed to foundation level by a robber trench cut down through the sediment. Symondson's 1594 map of'the decayed harborough of Rye' may give an anachronistic picture of Broomhill church, for it shows a building still standing. The churches shown on his map cannot be merely dismissed as depicted by cartographic convention, because all those shown appear to be different and, by comparison References (Superscrzpt numbers in the lexl reJk to unpubhs/zed sources, listed below.) Published sources Ancient Deeds Calendar of Inquiritions Miscellaneous Calendar of Patent Rolls Curia Reps roll^ maintained as dry land. The settlement here was a product not only of the society and the economy, but also of the environment. A study of Broomhill of necessity must be multidisciplinary to take into account these factors, and the present paper is intended as an interim report in such a programme of research. Acknowledgements The excavation at Broomhill Church was made possible by agrant from East Sussex County Council through the kind offices of Dr. Andrew Woodcock, and with the co- operation of the farmer Frank Cooke. I am also grateful for advice or assistance on various points from Dr. John Blair, Richard Coates, Jill Eddison, Dr. Doris Baker- Jones, Mr. J. S. G. Simmons and Christophcr Whittick. Anon. 1863: Charters of Cumbwell Priory. Arch. Cant. 5, 194-222. Anon. 1873: Calendar of Charters and Documents relating to the Abbey of Robertsbridge. (London). Baker, A. R. H. 1965: Some fields and farms in A4edieval Kent. Arch. Cant. 80, 152-74. Beredvrd, M. and Hurs~, J. G. 1971: Deserled Medieval V~llages. (London, Lutterworth Press). Medieval Settlement and Soc :iety in the Broomhill Area Brooks, N. P. 1981: Romney Marsh in the Early Middle Ages. In Rowley, R. T. (editor) The Ez,olution of Marshland Landscapes 127 Anglo-Saxon Towns in Southern England (Chichester, Phillimore), 1-36. (Oxford. Dept. for External Studies) 74-94. Brooks, N. P. 1988: Romney Marsh in the Early Middle Ages. In this volume, chapter 8. Camden, W. 1722 edn.: Britannia. (London). Chapman, A. 1977: Grangesand Other Landholdings of Robertsbridge '4bbey. (Michigan (U.S.A.), Xerox University Microfilms). Charles-Edwards, T. M. 1972: Kinship, status and the origins of the hide. Past and Present 56, 1-33. Cooper, W. D. 1850: The History of IVinchelsea. (London, John Russcll Smith). Dugdale, W. 1772: The History of Zmbanking and Draining of Diners Fens and Marshes. (2nd edn.) (London). Dunkin, A. J. (editor) 1854: Proceedings of the British Archaeological Association Meeting held at Canterbury, 1844. (London, John Russell Smith). Dyer, C. C. 1980: Lords and Peasants in a Changing Society: the estates of the Bishopn'c of Worcester. (Cambridge). Eddison, J. 1983: The reclamation of Romney Marsh: some aspects reconsidered. Arch. Cant. 99, 47-58. Green, R. D. 1968: Soils of Romney Marsh. Soil Survey of Gt. Britain, Bull. 4. (Harpenden). Harvey, P. D. A. 1984: Conclusion. In Harvey, P. D. A. (editor), The Peasant Land Market in Medieval England (Oxford, Clarendon Press), 328-56. Hilton, R. H. 1965: Freedom and villeinage in England. Past and Present 31, 3-19. Historical Manuscripts Commission 1925: Report on the Manurcripls of Lord de L'lsle and Dud@ Preserved at Penshurst Place 1. (London, HMSO). Holt, J. C. 1982: Feudal society and the family in early Medieval England: I. The revolution of 1066. Trans. Roy. Hist. Soc. 5th series, 32, 193-212. Holt, J. C. 1983: Feudal society and the family in early Medieval England: 11. Notions of patrimony. Trans. Roy. Hist. Soc 5th series, 33, 193-220. Homan, W. M. 1937: The use ofa Flemish land measure in the marshes between Hythe and Pett in the Middle Ages. Sussex Notes and Queries 6, 228-30. Homan, W. M. 1938: The marshes between Hythe and Pett. Sussex Arch. Cf1. 79, 119-223. Hyams, P. R. 1980: Kings, Lords and Peasants in Medieval England: the common law of villeinage in the twelflh and thirteenth centuries. (Oxford, Clarendon Press). Luard, H. R. 1880: Matthaei Parisiensis Chronica Majora 5. (London, H.M.S.O.). Matthew, D. 1962: The Norman Monasteries and their English Possessions. (Oxford, Clarendon Press). Mawer, A. and Stenton, F. M. 1929: The Place-Names of Sussex 1. (Cambridge, English Place-Name Society 6). Mawer, A. and Stenton, F. M. 1930: The Place-Names of Sussex 2. (Cambridge, English Place-Name Society 7). Orton, C. forthcoming: The pottery. In Rudling, D. R., Excavations at Winchelsea. Sussex Arch. Coll.. Rigold, S. E. 1972: Eynsford Castle and its excavation. Arch. Cant. 86, 109-72. Salzman, L. F. 1902: Feet offines relating to the county ofSussex from 2 Richard I to 33 Henry 111. Sussex Record Soc. 2. Salzman, L. F. 1937: Victoria History of the County of Sussex 9. (Oxford). Sawyer, P. H. 1968: Anglo-Saxon Charters: an annotated list and bibliography. (London, Roy. Hist. Soc.). Searle, E. 1980: The Chronicle of Battle Abbq. (Oxford). Smith, R. A. L. 1939: Marsh embankment and sea defence in Medieval Kent. Econ. Hist. Rev. 1st series, 10, 29-37. Tatton-Brown, T. 1984: The towns of Kent. In Haslam, J. (editor), Tatton-Brown, T. 1988: The topography of the Walland Marsh area between the eleventh and thirteenth centuries. I11 this volume, chapter 9. Wallenberg, J. K. 1934: The Place-Names ofKent. (Uppsala). Unpublished Sources Abbreviations used: BL British Library ESRO East Sussex Record Office PRO Public Record Office KAO Kent Archives Office 1. PRO, E134135 and 36 Eliz. Mich. 7. 2. ESRO, X.4314, no. 4. 3. PRO, E134135 and 36 Eliz. hiich. 7. 4. ESRO, XA3/19, f. 31r. 5. Lincoln's Inn, Hale MS. 87, f. 49r. 6. BL, Cotton Galba E IV, f. 152r. 7. ESRO, X.4314, no. 4. 8. BL, Campbell Ch. XXVI, 7. 9. e.g. Lincoln's Inn, Hale MS. 87, f. 49v. 10. PRO, E134117 Jas. I Mich. 9. 11. BL, Campbell Ch. XXV, 2; Egerton Ch. 383. 12. e.g. BL, Campbell Ch. XXV, 5. 13. Lincoln's Inn, Hale MS. 87, f. 29r. 14. BL, Campbell Ch. XXV 7, 15. 15. PRO, SC2/180/57, m. 2; BL, Cotton Galba E IV, f. 152r. 16. Rouen Archives Departmentales Seine-Maritime, 7H57. I owe this reference to Christophcr Whittick. 17. BL, Campbell Ch. XXVII, 2; Egerton Ch. 375; Lincoln's Inn, Hale MS. 87, f. 50v. 18. Lincoln's Inn, Hale MS. 87, ff. 50v-51r. 19. BL, Campbell Ch. XXV, 8. 20. ESRO, XA3119, ff. 29v-31r. 21. Lincoln's Inn, Hale MS. 87, E. 49v-51v. 22. BL, Campbell Ch. XXV, 3-5. 23. PRO, SC6/889/15. 24. e.g. PRO, SC12/1/34. 25. BL, Campbell Ch. XXVI, 11, 12. 26. Bodleian Library, DD All Souls C184/1. 27. Ibid., C130/85; C130194; All Souls College, CTM 417al2 (Hovenden 111, 10). 28. All Souls College, CTM 226164 (Hovenden 111, 8). 29. PRO, SC6/889/13-15; Bodleian Library, DD All Souls C1831 51a, b. 30. PRO, SC121 1/34; Bodleian Library, DD All Souls C 183151 b, 52b, c. 31. BL, Add. MS. 6343, col. 470, quoting Miscellanea Relica. It has not been possible to trace this work. 32. BL, Campbell Ch. XXVI, 7; Egerton Ch. 383. 33. BL, Campbell Ch. XXV, 5; ESRO XA3119, f. 30v. 34. BL, Campbell Ch. XXVII, 5. 35. Ibid., XXVII, 26. 36. PRO, E315/18, f. 29v. 37. BL, Campbell Ch. XXVII, 4. 38. BL, Egerton Ch. 383. 39. BL, Campbell Ch. XXV, 15. 40. Ibid., XXVI, 19; XXVII, 12. 41. ESRO, XA3/19, E. 29v-31r. 42. ESRO, X.4314, nos. 1, 4. 43. Lincoln's Inn, Hale MS. 87, f. 52r. 44. BL, Add. MS. 6343, col. 747; PRO, E134135 and 36 Eliz. Mich. 7; All Souls College, CTM 226164 (Hovenden 111, 8). 45. BL, Add. MS. 26785, f. 128; KAO, U442 0212.