Excavation of an Iron Age and Saxon Site at South Willesborough Ashford

237 EXCAVATION OF AN IRON AGE AND SAXON SITE AT SOUTH WILLESBOROUGH, ASHFORD simon deeves with Barry Bishop, Nick Branch, Natasha Dodwell, Malcolm Lyne and Alys Vaughan-Williams During July and August of 2001 an excavation was carried out at South Willesborough, Ashford (TR 0200 4060), following a field evaluation undertaken in the February of that year (Fig. 1). At the time of the excavation, the site was derelict agricultural land covered with light scrub growth. A proposed residential development by Fairview New Homes Plc was granted planning permission, but with archaeological conditions attached. The investigations were prompted by the known existence of Bronze Age, Iron Age and Romano-British sites in the immediate vicinity. The site lies on the lower valley side of the East Stour River, which runs along its southern boundary. Its northern extent lies at 42m od, falling toward the river at a level of c.38m od. The underlying geology of the area consists of Atherfield Clays with Hythe Beds. Toward the south of the site, the natural clay was capped by alluvial deposits associated with the East Stour River. The presence of features, along with relatively undisturbed potsherds, on this alluvial sandy silty clay sub-soil suggested that it represented a palaeo-landsurface dating from the Middle to Late Iron Age into the Romano-British period. The excavation Fifty-five trenches were investigated to evaluate the site. These revealed Middle to Late Iron Age/Romano-British pits across the south-western part of the site close to the East Stour River. An open-area excavation followed which measured c.145m e-w x 90m n-s. Beneath the topsoil and cutting the underlying clay were more pits producing comparable Iron Age/Romano-British pottery. The pits had charcoal-rich fills and some produced large quantities of fired clay or daub, which may indicate their use as domestic ovens, or possibly furnaces with a more industrial SIMON DEEVES 238 function. Most of the pits were clustered near a buried Iron Age pot which contained the cremated remains of an infant. In addition, several hearths which produced no dating evidence were uncovered. Radiocarbon dating of these features indicated a sixth- or seventh-century date. Greater LondonAshfordTHE SITE050kmNrailway601500/140550ExcavationEast StourSouthWillsborough602450/140550A20700400mFigure Fig. 1 Site Location. EXCAVATION OF AN IRON AGE AND SAXON SITE AT SOUTH WILLESBOROUGH 239 Phase 1 – Early Prehistoric The early prehistoric periods were represented by struck flints recovered mainly from unstratified contexts, although some occurred as residual finds in Iron Age features. The flints showed a great deal of variation in condition, raw material and technological styles, representing a wide range of flintworking traditions dating from the Mesolithic or Early Neolithic to the middle Bronze Age and possibly later. A few pieces may actually be contemporary with the Iron Age activity recorded although the small size of the assemblage and its evident chronological mix means that this must remain in doubt. Fifteen retouched implements were identified, nearly half consisting of scrapers, including denticulated types and a long endscraper. The remainder included notches, piercers, edge trimmed flakes, knives and a crudely manufactured chopping-type tool. The presence of small chips, miss-hits, core fragments and other knapping waste would suggest some core reduction was occurring on site, although the diversity and size of the retouch component would suggest a variety of activities involving tool use were being conducted, rather than just tool production. Phase 2 – Iron Age/Romano-British This phase of activity was represented largely by pits characterised by their charcoal-rich fills (Fig. 2). The pits were generally sub-rectangular or ovoid in shape with steep sides and a slightly concave or flat base. They were all of a similar size, typically measuring no more than 1.0m across by 0.20m deep. The fills of these pits produced relatively little dating evidence, although some Mid-Late Iron Age pottery was found along with sherds of an indeterminate prehistoric date, burnt and struck flint, and daub fragments. Larger pottery assemblages were recovered as unstratified finds in the topsoil, doubtless residually deposited during past ploughing episodes. They consisted of Middle Iron Age wares, coarse ‘Belgic’ wares dating from the Late Iron Age to ad 100+, and Upchurch/Hoo wares dating from ad 43 to 120 which were the only sherds that definitely post-dated the Roman conquest. There was no ceramic evidence for occupation after c.ad 100. A complete inverted saucepan pot recovered from a small pit dug to accommodate the pot provides further dating evidence. It measured 0.1m in diameter and would have been at least 0.15m deep had its base not been absent, presumably as a result of plough damage. It was made in a pale siltstone grog-tempered fabric which was used for Middle Iron Age style saucepan pots as found here, as well as later ‘Belgic’ forms. The pot appears to have been intentionally buried as part of the mortuary ritual. Its fill contained a small amount of cremated bone (16g), including SIMON DEEVES 240 fragments of skull and six partially formed tooth crowns indicating the cremation was of an infant about 6 months old. The recovery of cremated tooth crowns is not that common since the enamel shatters as it expands rapidly in the heat of the pyre. Nevertheless, the buff/white bone colour is indicative of full oxidation and efficient cremation; the survival of tooth crowns, and indeed any recognisable skeletal elements, suggesting the pyre was less hot, or the process as not as prolonged, as would be needed to cremate an adult. The inclusion of such small bone fragments (generally less than 10mm) suggests that considerable care was taken in collecting bone from the pyre. Of the recognised middle and late Iron Age burial practices, this burial TrenchEvaluationTrenchKEYSaxon hearthIron Age/Romano-British pitCremationSubsoil overlying natural clayEvaluation Trench040m Fig. 2 The open-area excavation showing archaeological features. EXCAVATION OF AN IRON AGE AND SAXON SITE AT SOUTH WILLESBOROUGH 241 seems most comparable to the La Tène III cremation tradition (Whimster 1977, 323-5), or Aylesford/Swarling culture (Taylor, 2001, 68), which saw the reintroduction of cremation into south-eastern England. This tradition, being an aspect of the influence of Belgic culture, became common from about 50 bc and grew to be more widespread and highly developed in the early Roman period. The Middle Iron Age form of the pot, in a fabric that continued in use into the Late Iron Age suggests a possible date of c.150-50 bc indicating that this burial may date to the very early period of the cremation tradition. Indeed similar saucepan-pots in the same fabric were present in a large pottery assemblage of this date recovered from a double ring-ditch at Beechbrook Wood to the west of Ashford (Lyne, forthcoming). A concentration of 79 pot sherds weighing 692g from a coarse ‘Belgic’ grog-tempered jar, dating from Late Iron Age - ad 100+ was found less than 20m to the south-west of the cremation. Although no cremated bone was recovered, this pot could also represent a cremation that had been almost entirely destroyed by ploughing. This cremation activity, along with other contemporary cremation sites in the surrounding landscape – such as Sevington Bridges to the east, probably had connections to nearby settlement. Several Iron Age settlements have been recorded along the East Stour river valley within the vicinity. These include to the south-east Waterbrook Farm (Rady 1992; 1995); on the higher ground to the east of the site, near Old Boys Hall and Sevington (Bennett 1988 and Glass 1999); and Kingsnorth, Park Farm to the south (Hicks 1993). Further settlement evidence has also been indicated from initial investigations at Park Farm East to the south (Wragg 2002). These settlements tend to be located further up the valley and away from the river. There appears to have been quite a dense pattern of Iron Age settlement in the vicinity of the site and the river would presumably have been exploited over quite a wide range away from the areas of settlement. The burial, therefore, is not too far from the focus of the local Iron Age community, despite its isolated nature. Locating this burial next to the East Stour is almost certainly significant, with rivers being one of the obvious foci for ritual practices during the Iron Age (Cunliffe 1975, 297). Many of the contemporary pits, which were grouped near the cremation may have been associated with the mortuary ritual, although other nearby pits indicated another function. The primary fill of one pit contained a notable amount of charcoal and daub fragments, plus nearly 2kg of large burnt flint pebbles which formed a ring around the edge of the feature. Its secondary fill was mixed with a high proportion of fired clay fragments (226g were collected) which suggests that this feature was a sunken hearth or oven; the fired clay representing a collapsed ‘capping’ over the pit, whilst the ring of burnt flint may be the oven lining or wall. A second SIMON DEEVES 242 possible sunken hearth or oven was uncovered in Trench 3 to the east of the main excavation area during the evaluation (see Fig. 1). This feature also contained a charcoal-rich fill covered by fragments of fired clay that were interpreted as the remains of a collapsed clay structure or lining. Also recorded across the site were a number of shallow sub-circular pits, again with charcoal-rich fills. Recovered from the fill of one of these was a fragment of daub that showed evidence of structural detail. On its outer face were the impressions of three overlapping withies (wattle rods) that appeared to create a convex surface. The interior, concave face was covered in soot suggesting it was part of a domed oven roof or lining, or part of an enclosed hearth. Phase 3 – Saxon A distinct group of features, interpreted as hearths, was present as shallow circular and sub-circular cuts, between 0.75-1.45m in diameter, with fills composed primarily of medium to large charcoal fragments (Fig. 2). These hearths differed from the Iron Age/Romano-British features in that the charcoal in their fills appeared less fragmented and the natural clay exposed in their bases was scorched indicating in situ burning. Stakeholes recorded in the base of three of these features were filled with a similar burnt material indicating that they were associated with the burning episode and perhaps represented structures built over the fires. In the absence of datable finds, charcoal from three of the hearths was radiocarbon dated, indicating an early-mid Saxon date range from the fifth to the seventh century ad (probably sixth). During the Saxon period, this area is thought to have been open agricultural land although analysis of environmental samples from the hearth fills suggested a mix of woodland and grassland, possibly pasture, but also with an indication of damp wasteland vegetation. There was only limited evidence for agricultural land-use in the form of hedgerow species, which could indicate field-borders and a few weeds of cultivation. It seems most likely, therefore, that during the Saxon period this low-lying area by the river was not being farmed for crops although nearby higher ground may have been cultivated. The Pottery by Malcolm Lyne The excavations produced 273 sherds (2,292g) of pottery dating from the Early Iron Age to Early Roman times, the majority being Middle to Late Iron Age material. All of the overtly Roman pottery came from the 69 unstratified sherds. EXCAVATION OF AN IRON AGE AND SAXON SITE AT SOUTH WILLESBOROUGH 243 Middle Iron Age Fabrics MIA.1 Handmade black fabric with profuse ill-sorted up-to 2.00 mm. calcined flint filler. MIA. 2. Friable brown-black fabric with up-to 1.00 mm. grog and occasional up-to 0.50 mm angular white alluvial flint grit and larger chert inclusions. MIA.3 Friable black fabric with profuse silt-sized quartz and sparse larger up-to 0.30 mm. grains. Late Iron Age/Early Roman Fabrics (as per Canterbury Arch. Trust codings) B2 Coarse ‘Belgic’ grog-tempered ware. B2.1 Similar but with pale siltstone grog. This fabric was also in use during the Middle Iron Age for saucepan pots as well as later ‘Belgic’ forms. R16 Upchurch greyware. R17 Upchurch/Hoo oxidised ware. Middle Iron Age Assemblages (c.300-100 bc) The few assemblages of this date are for the most part tiny and scrappy and totally lacking in rims and other diagnostic sherds. The one exception is the inverted cremation pot; a small saucepan-pot in black Fabric B2.1 with siltstone-grog filler fired rough brown. All of the pot is present except for the base. Ext. rim diameter 100mm (Fig. 3.1). Similar saucepan-pots in the same ‘Belgic’ Late Iron Age type fabric are present in a large transitional Middle Iron Age/Late Iron Age (c.150-50 bc) pottery assemblage from a double ring-ditch at Beechbrook Wood on the west side of Ashford (Lyne, forthcoming). Fig. 3 Examples of Iron Age and Romano-British pottery. 010cm12Figure SIMON DEEVES 244 Late Iron Age to Early Roman Assemblages (c.100 bc-ad 100) The bulk of the pottery from the site belongs to this date-range but, here again, is largely made up of small assemblages which tend to be deficient in rims and other diagnostic sherds apart from furrowed jar fragments. All of the native ‘Belgic’ material is in grog-tempered fabrics B2 and B2.1 and includes fragments of a jar of Thompson Type B2-1 or 2-3 (1982) with slightly-everted rim and rippled shoulder, in patchy red/black Fabric B2.1. c.50 bc-ad 50 (Fig. 3.2). All of the sherds in Romanised wheel-turned fabrics were unstratified in the topsoil and comprise nine fragments from a ?Monaghan Class 2E beaker in grey Upchurch ware Fabric R16 (1987, c.ad 43/70-110) and a Hoo ?flagon of c. ad 50-120 date in Fabric R17. There is no ceramic evidence for occupation after c.ad 100. TableABLE 1. RESULTS OF THE RADIOCARBON DATING Context Sample Result (Uncal 14C years BP) Code δ13C % Cal BC/AD (95.4%) 185 35 1452±44 Wk10727 -25.1 530-670 AD 178 32 1405±48 Wk10726 -27.4 540-710 AD 142 28 1588±45 Wk10725 -27.0 380-600 AD discussion The excavation showed a human presence in this area on the banks of the East Stour during much of prehistory. The lithic material shows a range of technological styles that span the Mesolithic/Early Neolithic to the Middle Bronze Age or after, although the lack of features dating to these periods indicates temporary riverside activities rather than settlement of any great duration. The low-lying area along the East Stour River was probably a focus of hunting expeditions away from the main settlements during these periods. The river itself would have been a source for fishing and fowling, plus would have attracted larger game to its banks, which could then be exploited. The significant proportion of retouched implements among the lithic assemblage suggests that activities involving tool use were dominant. Indeed, the high number of scrapers provides an association with animal processing, suggesting that such activities may have taken place on site. Perhaps temporary hunting camps were established in the area during these prehistoric periods. Although the study site showed no evidence of settlement, it would have been a place of much greater activity by the middle of the Iron EXCAVATION OF AN IRON AGE AND SAXON SITE AT SOUTH WILLESBOROUGH 245 Age, when a denser pattern of settlement is evident in the surrounding area. The river would still have been exploited in similar ways to the earlier prehistoric periods, and there is surviving evidence of charcoal-rich pits dating to this period that appear to be associated with burning and possibly cooking. Indeed, two of the pits revealed what appeared to be a collapsed clay ‘capping’ of a domed roof suggesting they were sunken hearths or ovens. The presence of vitrified charcoal, indicating very high temperatures, does not necessarily reflect their use for industrial processes and the mix of wood types present would suggest they did not have a specialist function. Again, the idea of temporary hunting camps in the low lands by the river is conceivable for this period. Alternatively, these features may be associated with the herding of animals, which could have occurred over quite a large range away from the main settlements, creating the need for temporary camps. The lack of crop remains from the environmental samples may be indicative that this area was pasture as opposed to agricultural land or perhaps just damp wasteland, although it is noted that the lack of recovered material makes interpretation difficult. Nonetheless, an area such as this, open ground on the lowlands with a source of water, would have been a good location for the herding of animals. Such activity may well have occurred from the Middle Iron Age on into the early Roman period. The cremation also falls within this period, being most likely to date from the Mid-Late Iron Age transitional period (c.150-50 bc), and so relatively early as part of the ‘Belgic’ tradition. The lack of any associated grave goods with this cremation shows it to be a relatively simple grave, although its seemingly isolated location does make it somewhat distinct, as does its status as an infant burial. A concentration of pottery from a Late Iron Age/Romano-British jar may represent a much disturbed second cremation, leaving the possibility that other cremations once existed on the site. More formal cremation cemeteries are known in the vicinity of the site, such as at Westhawk Farm to the south-west (Booth and Lawrence 2000). However, these tend to date more from the very Late Iron Age and Romano-British periods, which saw the centralisation of both cemetery and settlement (Pearce 1997). During the preceding centuries, when the ‘Belgic’ influence first appeared in south-eastern England, rural communities were living in smaller group settings that were a part of more general tribal structures. Why is it, then, that this burial is so isolated, being some distance away from the nearest settlement? Perhaps its location is somehow linked to the use of the land in this area. Burial ceremonies are often seen as rites of passage that reinforce the order and structure of a society and, in rural society, the time scales of the living are partially based around the cycle of the agricultural year. The dead may have been harnessed to promote fertility, or were used to hold land rights or some other resource (Pearce 1997). It is conceivably in this SIMON DEEVES 246 capacity that the location of this burial was chosen. Possibly associated with this cremation were the two pit clusters to its south and south-west. The overlapping date ranges of the small pottery assemblages from these pits suggests they date from the Mid-Late Iron Age transitional period, perhaps the late second/early first century bc and therefore broadly contemporary with the cremation burial. Although the pit fills contained a notable amount of charcoal, the environmental samples did little to enhance understanding about their possible function, therefore making it difficult to assess whether they were associated with the mortuary ritual, or merely a close concentration of fire/cooking pits. The function of the Saxon hearths across the site is also difficult to ascertain. The environmental samples produced no food residues, with the exception of raspberry and fig seeds, and no residues indicating an industrial function. It seems most likely, therefore, that they merely represent campfires used for heat. Those with stakeholes recorded at their base may just have been campfires with associated stake-built structures, possibly for cooking. The environmental samples suggested this region was a mix of woodland and grassland at this time, with little evidence of agricultural activity. It seems likely, therefore, that these campfires were built at times when the riverside was being exploited in ways similar to those that occurred in earlier prehistoric periods. bibliography Bennett, P., 1988, ‘Archaeology and the Channel Tunnel’, Archaeologia Cantiana, cvi, 1-24. Booth, P. and Lawrence, S., 2000, ‘Westhawk Farm, Ashford’, Current Archaeology, 168. Cunliffe, B., 1975, Iron Age Communities in Britain, Book Club Associates, London. Deeves, S., 2002, ‘Assessment of an Archaeological Excavation at South Willesborough, Ashford, Kent’, Pre-Construct Archaeology (unpubl. report). Glass, H.J., 1999, ‘Archaeology of the Channel Tunnel Rail Link’, Archaeologia Cantiana, cxix, 189-220. Hicks, M., 1993, ‘Park Farm, Ashford’, in CAT Annual Report 1992-93. Lyne, M.A.B., forthcoming, The Middle Iron Age, Late Iron Age, Roman and Medieval pottery from south of Beechbrook Wood (provisional title). Macpherson-Grant, N., Savage, A., Cotter, J., Davey, M. and Riddler, I., 1995, Canterbury Ceramics 2: The Processing and Study of Excavated Pottery. Monaghan, J., 1987, Upchurch and Thameside Roman Pottery, BAR Brit. Ser. 173, Oxford. Pearce, J., 1997, ‘Death and time: the structure of late Iron Age mortuary ritual’, in A. Gwilt and C. Haselgrove (eds), Reconstructing Iron Age Societies, Oxbow, Oxford. Rady, J., 1992, ‘Waterbrook Farm, Ashford’, in CAT Annual Report 1991-92. Rady, J., 1995, ‘Waterbrook Farm, Ashford’, in CAT Annual Report 1994-95. EXCAVATION OF AN IRON AGE AND SAXON SITE AT SOUTH WILLESBOROUGH 247 Stace, C., 1997, New Flora of the British Isles (2nd ed.), Cambridge University Press, Cambridge. Taylor, A., 2001, Burial Practice in Early England, Tempus, Stroud, Glos. Thompson, I., 1982, Grog-tempered ‘Belgic’ Pottery of South-eastern England, BAR Brit. Ser. 108, Oxford. Whimster, R., 1977, ‘Iron Age Burial in Southern Britain’, Proceedings of the Prehistoric Soc., 43. Williams, A., 2002, ‘Environmental Assessment’, in Deeves 2002. Wragg, E., 2002, An Archaeological Evaluation at Park Farm East, Ashford, Kent, Pre-Construct Archaeology (unpubl. report).