Chapter 11: Concluding comments

by Frank Meddens, Stuart Foreman, Martin Bates and Damian Goodburn

Physical limitations of the investigation deposits and the often ephemeral character of Managing impacts to archaeological deposits within archaeology in alluvial environments (such as a busy urban environment, alongside a major road prehistoric waterlogged wooden structures). construction project, presented numerous logistical A third limiting factor was the physical difficulty challenges, which were addressed by the road in accessing some of the deeper deposit sequences. engineers through ingenious and innovative ways. The most significant waterlogged archaeological For instance, the preservation of the important deposits were found along the margins of Neolithic site identified by evaluation trenching at palaeochannels, buried beneath variable depths of Woolwich Manor Way was achieved by using light - alluvium and made ground. The project lies on the weight polystyrene fill in the road embankment periphery of Greater , in an area which was construction, thus substantially reducing the predominantly undeveloped marshland and frequency of concrete piles required to support the agricultural land until the latter part of the 19th structure. The nature of the working environment century. Nevertheless deep modern made ground meant that investigations were for the most part was a significant obstacle in some areas (where limited to relatively small windows where there was covered by the existing A13 road embankment, for a demonstrable below ground impact from construc - example). The engineering efforts required to reach tion activity. This can be frustrating for archaeolo - the archaeology in these cases meant that the gists who naturally wish to explore the full extent of archaeological work was both costly and poten - a site or feature, but is a necessary constraint on tially hazardous. The deposit sequences were not developer-funded archaeology. In most cases this universally deep however; in areas where the should mean that the unexcavated portions of those terrace gravels lay very close to the present ground sites are preserved in situ and available for future surface, the width of the new road was stripped of investigation should the opportunity or need arise. topsoil to expose quite extensive sections of the The important Neolithic site at Woolwich Manor terrace surface, in ground conditions comparable in Way, for example, was briefly glimpsed in an evalu - most respects with investigations in rural dry land ation trench and subsequently preserved for environments. The least effective interventions posterity. On the A13 project some of the key excava - proved to be the Watching Brief work. This can be tion areas were defined by the extents of flyover explained both by the inevitable constraints on abutments, which were very small areas indeed, access and visibility, and the nature of the alluvial comparable in size with evaluation trenches. The deposits, in which it was often difficult to recognise most extensive deep investigation was the coffer features even when conditions were ideal. The dam excavation at Prince Regent Lane (Freemasons difficulties of creating a coherent archaeological Road Underpass) in which a substantial Bronze Age record under general Watching Brief conditions are piled structure was found. The extent of the well-known. cofferdam was largely dictated by the plan of the new underpass (although informed by evaluation trenching results). Contribution of the archaeological data to The scope of investigation was also limited to regional research areas where reasonable prospects existed for intact In spite of the practical limitations the particular archaeological deposits. Disturbance caused by topographical niche occupied by the A13 corridor, construction of the existing A13 and the extent of skirting the northern edge of the Thames flood - 20th century development on either side of the road plain, offered valuable insights from an archaeolog - meant that these areas often lay in narrow strips of ical and palaeoenvironmental point of view, which surviving intact deposits, flanked on either side by have fully repaid the time, money and effort truncated, disturbed or inaccessible deposits. The invested in the archaeological work. The extensive relatively undisturbed areas were identified initially presence of waterlogged sediment sequences on the basis of desk-based studies and geotechnical offered exceptional potential for the recovery of investigations, and then refined by excavating organic materials of various periods in stratified archaeological test pits and trenches during the alluvial and archaeological deposit sequences. The Phase I and II evaluation. Geophysical survey formal excavation areas and evaluation trenches, methods were not a viable option due the disturbed taken together, provide a rare series of compara - nature of the ground, the depth of overlying tively large scale investigations for the Lower

147 Landscape and Prehistory of the East London Wetlands

Thames in the London region, linked by a common Europe, which have yet to be satisfactorily resolved project design and geoarchaeological research (Brown 2008, 1-18). framework. The archaeological discoveries range from the Mesolithic through to the post-medieval period, Prehistoric archaeology in the Lower Thames although material from the 2nd millennium BC was floodplain most commonly encountered, in the form of timber Until the explosion of development-led archaeology structures and in particular trackways. This volume in the 1990s, knowledge of prehistoric human therefore marks an outstanding contribution to our activity in the floodplain of the Lower Thames and understanding of the prehistory of the Lower its tributaries in the London area was limited, Thames area, particularly the Bronze Age. Together comprising occasional tantalising glimpses, such as with other recent large scale investigations within the discovery of significant quantities of metalwork, the Thames Floodplain in London, such as High mostly weaponry from the river itself, and Speed 1 (formerly the Channel Tunnel Rail Link), occasional other finds such as the idol the Jubilee Line Extension and the Lea Valley unearthed in 1922 (Coles 1990, 326). In addition Mapping Project, the A13 investigations have there were references to wooden structures, inter - helped to clarify the sequence of inundation of the preted as crannogs or pile dwellings of uncertain Thames floodplain caused by sea-level rises during date which had been found in the later 19th century the Holocene, shedding particular light on the on the , during the excavations of the reactions of Bronze Age communities to these Maynard (Smith 1907), Warwick (Needham and landscape changes. The data is most immediately Longley 1980), Banbury and William Girling reser - relevant to the narrow but important topographical voirs. The extent and nature of the peat beds found zone occupied by the A13 corridor, at the interface in the floodplain alluvium had been remarked upon between the marshland occupying the valley floor at least as early as 1721 by John Perry in his account and the gravel terrace to the north. However it also of the stopping of the Dagenham Breach. In this contributes to discussions of contemporary cultural work Perry not only refers to the extensive presence landscape changes beyond the floodplain. As of brushwood in the peats, he also mentions observed throughout the region, there is a widespread finds of hazelnuts and yew trees and widespread change, between the Neolithic and the occasional finds of deer antlers in these deposits. He middle Bronze Age (broadly the late 3rd and early also describes the basic sequence of clays overlying 2nd millennium BC) from a landscape focused on peats, which in turn lay on top of blue clays over dispersed ceremonial monuments to an agricultural gravels and sand (Perry 1721). landscape, which in some areas is characterised by The A13 project has contributed some new data a formally delineated environment comprising co- to add to models of Mesolithic and Neolithic axial field-systems associated with stock manage - landscape evolution in the Lower Thames Valley, ment and pastoralism (Pryor 1998; Yates 2001). The but has produced relatively limited archaeological different types of archaeological data from the evidence from these periods. Other recent projects Thames floodplain provide a complementary view in the region, such as High Speed 1, have been more of this phase of settlement transformation. forthcoming in this respect (Bates and Stafford forthcoming). Nevertheless the identification of elements of a buried Neolithic landscape, including Regional sea-level and climate studies artefacts and components of carbonised wood in a This volume also presents multi-proxy palaeoenvi - varied range of woods, charred hazelnut shells and ronmental evidence from a series of radiocarbon the cereal crop remains at Woolwich Manor Way dated sediment sequences from the margins of the offer rare glimpses of Mesolithic and Neolithic Thames floodplain. Apart from the important infor - human activity within the floodplain. When set mation that these provide for the landscape context alongside other contemporary evidence in the of the associated prehistoric archaeological finds vicinity (summarised in Chapters 1 and 9) it is clear (discussed in detail in this volume) these represent that the A13 Neolithic finds contribute significantly valuable additions to the Holocene palaeoenviron - to regional discussions of this period. The Neolithic mental record for southern Britain, particularly evidence, such as it is, appears generally compatible given their close association with significant, radio - with patterns in south-east more generally, carbon dated, in situ archaeological structures. The in supporting suggestions of low density, moder - relationship between sea-level rise, climate change ately mobile, seasonally based activities, with a and cultural change are complex and multi-facto - preference for ecotonal locations along channel rial, and the subject of current debates which lie margins. The cereal crop remains from Woolwich largely beyond the scope of this study. Suffice it to Manor Way contribute evidence to current debates say, the changes in landscape during the Neolithic concerning the chronology, extent and context and Bronze Age, as noted in the A13 environmental of arable cultivation in the Neolithic. The jet belt sequences and other projects in the , slider from Movers Lane contributes to debates are likely to be related to more widely observed concerning long-distance movement of individuals fluctuations in climate across Britain and north-west and groups, and exchange mechanisms.

148 Chapter 11

As the edges of the Thames floodplain have been access was not evenly distributed across the flood - periodically pushed inland and upwards by rising plain. sea levels, areas of earlier prehistoric dry-land or Taken together the evidence suggests a phase of terrace edge activity have been inundated, sealing intensive but unevenly distributed exploitation of archaeological sites beneath layers of alluvium. The the Thames marshes throughout the Bronze Age. line followed by the A13 appears to correspond The main period of trackway construction appears broadly with the edge of the floodplain as it was in to begin in the late 3rd millennium BC and continue the Bronze Age. More extensive Mesolithic and to the late 2nd millennium. The dating of individual Neolithic remains may well be preserved in equiva - structures is often very broad, but it is possible to lent former terrace edge situations in the Thames discern a peak in trackway related radiocarbon floodplain to the south of the A13 corridor or along dates from the A13 project in the middle period of the margins of tributary channels, masked by thick the 2nd millennium. 1500 BC coincides broadly layers of later alluvium. Locating ephemeral with a period of cultural transformation across the Mesolithic and Neolithic sites by currently available region in the middle and late Bronze Age, seen as survey methods in such deep sequences is notori - the point of transition to a later prehistoric world of ously challenging. Physical access is logistically sedentary communities, field systems and hillforts, difficult and extremely costly due to the depth of large scale polities within complex social hierar - overlying deposits. In general only exceptionally chies and increasingly intensive agricultural and large construction projects offer opportunities for craft production (Bradley 2007, 178-202). The extent archaeological excavation in this environment. to which this transformation reflects a peak in Exceptionally dense concentrations of artefactual population levels is unclear. Part of the explanation materials are likely to be preferentially detected in for the appearance and distribution of the track - these situations (for example Tank Hill Road on ways may be increasing population pressure in High Speed 1, Bates and Stafford forthcoming). certain ecotonal localities during the first half of the Because of these difficulties, development-led 2nd millennium BC, leading to increased exploita - investigations have increasingly focussed on tion of the marshland environments in their modelling the alluvial deposit sequences by various immediate vicinity. geoarchaeological methods. Current approaches A number of key questions surrounding the emphasise modelling the inundation of the Thames trackways remain unanswered, as investigations floodplain in the early Holocene and attempting to have tended to be focused in the relatively shallow identify the shifting ecotonal locations that seem to alluvial sequences near the terrace edge. Why are have been favoured by prehistoric communities. more trackways found in some areas than others? The most likely locations for prehistoric occupation Proximity to a major settlement focus on the gravel sites in particular periods can be suggested by terrace is one possibility. There may be as yet identifying areas of what would have been higher unknown factors which made this part of the river drier ground in terrace edge situations in the valley particularly attractive to Bronze Age commu - Mesolithic and Neolithic. (Bates and Stafford forth - nities. We know that the trackways ran from the coming). Buried landscape features such as terrace edge deep into the marshes, but as yet we palaeochannels can also be identified and investi - have little notion of where they ended. Did they gated by such methods; such features may have terminate in the marsh, or continue as far as the attracted prehistoric activity and acted as sediment Bronze Age foreshore of the Thames? ‘sumps’ which provide optimum conditions for the Further work is also needed to relate the preservation of organic artefacts, structures and trackway evidence to contemporary settlement and palaeoenvironmental evidence. economic exploitation on the adjacent gravel The A13 investigations have produced a remark - terraces. Even in this period, which is the richest in able record of the efforts of communities of the 2nd terms of archaeological evidence, no certain millennium BC to adapt to landscape change as the evidence for settlement was found in the A13 sites. terrace edge migrated northwards and became However the density of remains at certain locations, wetter during the Bronze Age. The appearance of particularly at Prince Regent Lane (Freemasons trackways at numerous locations was presumably Road Underpass) is particularly striking when driven by a need for continued or increased access compared with the typically ephemeral and through the alder carr wetlands. The structure at dispersed nature of Bronze Age settlement remains Freemasons Road, comprising a double row of in dry-land contexts in south-east England. The substantial timber piles, undoubtedly constitutes range of features encountered, including wooden the most significant Bronze Age structure uncov - trackways and platforms, a piled timber structure, ered in the A13 work and represents one of the post and stake holes, pits and ditches, animal bone, largest structures of its kind in the London region. pottery, struck and burnt flint, suggests that these The distribution of Bronze Age trackway discov - sites are likely to have been located in close eries in the East London marshes, including a high proximity to settlement sites, probably located on concentration of wooden trackways and platforms the adjacent gravel terrace. The varied character of in the area and a limited number of other the wooden trackways, and the inclusion of burnt apparent focal points, suggests that the need for material in their construction at both Woolwich

149 Landscape and Prehistory of the East London Wetlands

Manor Way and Movers Lane, point to significant, later Bronze Age at the Golf Drivers Range has been possibly seasonal habitation in the immediate interpreted as evidence that stock management vicinity of these sites in the Bronze Age. The invest - became increasingly important at this time, as ment in substantial timber structures, such as that at flooding increased and trackways were constructed Freemasons Road, surely implies a need for to maintain or improve access to the marshland frequent passage through the terrace edge land - (Carew et al . 2010), although cereal-type pollen is scape and may suggest more permanent settlement present further up the profile, when conditions in the near vicinity. Unfortunately the impact of were at least as wet. Cereal-type pollen was, 19th century and later development along the line however, present in later Bronze Age samples from of the A13 is likely to have severely damaged or Woolwich Manor Way, although there are known removed much of the evidence for these postulated difficulties in identifying cereal type pollen as terrace edge sites. In particular the construction of opposed to wild grasses in coastal wetlands, as the Great Northern Outfall Sewer is likely to have discussed in Chapter 8. As is generally the case had a significant adverse effect on the survival of during this period in the British Isles, there is no archaeology along the inferred line of the Bronze evidence for the exploitation of wild wetland Age terrace edge, although bands of relatively resources such as fish and birds (Carew et al . 2010). undisturbed deposits are likely to survive, as The recovery of several tiny human long bone discovered in the A13 corridor. The areas of high fragments, from alluvial layers which contained potential indicated by the A13 project suggest Bronze Age material, is of interest, although the priority targets for any future development-led evidence is clearly very slight and reworking of investigations that may be required in the area. earlier material is a clear possibility. The evidence Some evidence has been recovered for economic for relocation and redistribution of skeletal remains exploitation of the marshes in the Bronze Age. The has been noted elsewhere and it has been suggested A13 data, especially the worked wooden structures, that veneration of the skeletal remains of ancestors provide evidence for woodland management (such may have played an important part in the lives as coppicing for firewood and building materials). of Bronze Age communities (Brück 1995; Halstead The pollen evidence indicates increasingly open et al . 2001; Owoc 2001; Parker Pearson 1999). Pre - conditions and a decline in alder carr woodland in historic human long bones and skulls, whole as well the late 3rd and early 2nd millennium, which could as fragmented, have been noted as selectively result either from increasingly wet conditions or occurring in placed deposits, practises which are impacts from livestock grazing (or a combination of also found in Iron Age and Roman contexts (Butler the two). There is a small but gradually accumu - 2006, 38-44; Harding 1985; Merrifield 1987) as well lating body of evidence for the presence of as in the case of skulls found in the livestock, particularly cattle, in the former Thames (Bradley and Gordon 1988; Edward, Weisskops and marshes from the Neolithic into the Iron Age (see Hamilton 2010). for example Carew et al. 2010; Jarrett 1996; Crockett Many of the research questions posed in advance et al . 2002). From the A13 sites, fragments of cattle of the project, particularly those related to the and sheep/goat bone of late Bronze Age date came historic periods, from Roman times on, have not from ditch fills at Prince Regent Lane. Butchered been addressed to any great extent, as significant cattle and sheep bones, part of a pig skull and a remains and datable palaeoenvironmental evidence piece of a cattle skull came from a flood deposit at from these periods were rarely encountered in the the same site. There was also a buried surface of late A13 sites and did not represent clear settlement or Bronze Age / early Iron Age date present here other focii. Rising sea-level in the Thames Estuary in which included possible poaching by cattle hooves, later prehistory is likely to have shifted the line of and a possible hoof print was identified at the the terrace edge in a northerly direction, beyond the Movers Lane site (this volume). The small Bronze limits of the A13 investigations. The traces of Age wattle fenced enclosure at the Bridge Road site Roman and later activity encountered along the A13 in Rainham is likely to have been for stock manage - are largely confined to elements of field systems ment (Meddens and Beasley 1990, 243). Many sites found on spurs projecting from the gravel terrace in the area have occasional evidence for the into the floodplain. The relative scarcity of Roman, presence of dung beetles from Bronze Age contexts, medieval or post-medieval archaeological sites but their numbers in each instance are so low that within the A13 sites is somewhat surprising as the they cannot be used as evidence for domesticated archaeological sites at Prince Regent Lane/ herd animals in the marshes. The Bronze Age Freemasons Road Underpass, Woolwich Manor ditches on the gravel terrace margin at the Prince Way and Movers Lane lie along local historic route - Regent Lane site may have had a purpose similar to ways and at natural river crossing points. the enclosed fields found at Fengate in Lincolnshire, close to the former dryland margin of the fen (Pryor 1998). These are interpreted as the enclosed winter Methodological issues pastures for cattle, which, during the summer The adoption of a geoarchaeological approach has months would have been fattened up on the benefited the project in two distinct areas: marshes. A gap in the record for cereal pollen in the

150 Chapter 11

1. Through the implementation of a seamless to write a report which is relevant to field archaeol - strategy of investigation from desktop assess - ogists who find themselves working in alluvial ment through field evaluation to excavation and environments, as well as to palaeoenvironmental analysis and geoarchaeological specialists. Some of the 2. As an aid to interpretation in the field during decisions made in presenting the data, such as the assessment and excavation and to contextu - use of interpretative graphics, and inclusion of alise the archaeology of the route corridor common species names in the main report text, are within the framework of the local and regional intended to make the data more accessible to non- environment. specialists.

There is little doubt that the adoption of a geoar - chaeological framework for the project aided Lessons learned progress by providing a common theme for all It is worth highlighting key areas where lessons can stages of the investigation. The identification of be learned for the benefit of future projects, partic - sediment bodies likely to contain archaeological ular those in alluvial environments: remains and the location of that archaeology was in With regard to the waterlogged wood, the usual part a function of following this approach. The problems of wetland archaeology had to be faced in complex interaction of different organisations at an urban setting where the zone of construction various stages of the project required a common impact on archaeological deposits was linear and thread to keep the investigations focussed on restricted. Many of the excavation areas were defined objectives, particularly at times when relatively small, which caused problems in inter - switching between excavators with differing levels preting some of the timber and roundwood struc - of expertise and familiarity with the archaeology of tures. Chief amongst these was the probable corner East London, which might otherwise have impacted of a substantial artificial timber platform at Wool- detrimentally on the project. A similar approach wich Manor Way (Str. 61). was adopted within the High Speed 1 project (Bates Although specialists attended on-site as required, and Stafford forthcoming). in certain respects unfamiliarity of many of the field Secondly the geoarchaeological framework staff in some of the special features of prehistoric provided a basis for training and informing field wetland archaeology, such as the distinguishing of staff during the excavation process. This was partic - wood chips from bark fragments, inevitably caused ularly important given the complex nature of the some waste of resources. At the initial post-excava - primary stratigraphy, which was largely controlled tion stage time was spent cleaning and checking by natural sedimentation processes. Such condi - many kilograms of bark fragments for worked tions are familiar to the minority of field archaeolo - material. This highlights the importance of regular gists who are accustomed to working routinely in site visits by specialists and targeted staff training. alluvial environments, but are unfamiliar territory The most successful method of wrapping lifted for many. In these circumstances it is often common woodwork for recording and sampling off-site was practice for field archaeologists to resort to semi- to place all but the smallest items in doubled zip- interpretative ‘descriptions’ of deposits or to simply lock bags, were first wetted and then tightly cease excavation when ‘natural’ is reached. Clearly wrapped in light, black rubbish bin liners of fine within the context of the floodplain such an plastic which clings to the wet wood. Then the approach is flawed and limits the excavators ability worked wood was wrapped in heavier duty rubble to interpret the sequence. sacks and double labelled. The inner membrane A third avenue of geoarcheological benefit was was thus protected by the tough outer covering. available to the project but, for a variety of reasons, Material wrapped in cling film was the least well- was not adopted. Although much important infor - protected and the extra handling required to mation about the sediments and sequences was unwrap this material both tended to damage the obtained during the field and laboratory investiga - material, and was very time-consuming time. It is tions, these investigations were not always strongly recommended that wrapping with cling conducted at a scale commensurate with the film is not used in future similar projects. original questions posed by the excavators and As with waterlogged wood, wherever possible, archaeological analysts. The reasons for this are in assessment and analysis of waterlogged bulk and part a function of cost rather than absence of monolith samples should proceed swiftly, since evidence, as well as difficulties in pinpointing those once removed from the ground delicate water - parts of the sequences that could address the logged material is prone to degradation and decay. questions. Experience from palaeoenvironmental analyses in An important characteristic of this project is the connection with this project and others with long extent to which it has proved possible to compare delays between collection and final analysis, has directly palaeoenvironmental evidence for demonstrated significant reduction in the range of landscape change with archaeological evidence at pollen and plant remains present after a delay of a both a site-specific and a landscape level. The inten - year or more. In some cases on the A13 project, tion of the authors in writing this volume has been alternative monolith sample sequences had to be

151 Landscape and Prehistory of the East London Wetlands selected for detailed analysis due to the deteriora - design and based on specialist advice. The respon - tion or loss of key samples between the assessment sibility for collecting, documenting, sub-sampling and analysis phases. Any sampling undertaken and curating sample material needs to be clearly set should be in accordance with a written site-specific out and managed at each stage. Samples should be sampling strategy, with clearly stated aims, formu - kept cool and dark, and stored following the latest lated within the context of the overarching project English Heritage guidelines.

152