Introduction

 

The Cotswold Water Park Project

The Cotswold Water Park Project is a landscape study centred on parts of the Upper Thames Valley in the southern and eastern hinterland of the Roman town of Cirencester (Fig. 1.1: Location of Project Area).
Over the past 40 years, much of this area has been subject to gravel extraction in order to cater for the boom in the construction industry. The resultant flooded gravel quarries saw the creation of the Cotswold Water Park, an area of nature reserves, country parks and recreational zones spreading over 40 square miles (Fig. 1.2: Location of key sites). The large scale quarrying along these gravel terraces has led to numerous extensive archaeological investigations which have highlighted dense areas of settlement, ranging from early Neolithic to the post-medieval period.
The current project has incorporated four key Iron Age and Roman archaeological sites within this region (Fig. 1.2).

  • Claydon Pike
  • Somerford Keynes
  • Whelford Bowmoor
  • Stubbs Farm

Most of the key sites were the subject of archaeological investigation in the late 1970s and 1980s as part of a co-ordinated program of research carried out by Oxford Archaeology (OA, then Oxford Archaeological Unit, OAU) in the Gloucestershire Upper Thames Valley.
The overall aim of the current project has been to present the data and interpretations of the four key sites and to examine the socio-political and economic development of the wider region from the middle Iron Age to the end of the Roman period.


Summary of the key sites

Fairford Claydon Pike

Excavations on the first gravel terrace at Claydon Pike between 1979 and 1983 revealed two areas of settlement, ranging from the middle Iron Age to the medieval period. Middle Iron Age activity at Warrens Field was recorded over three gravel islands, and probably represents the settlement of one or two families, that shifted eastwards over time. A maximum of four roundhouse structures were in use on an island at any one time, and the structures contained varying quantities of occupation refuse in their surrounding drip gullies and associated enclosure ditches. The inhabitants were pastoralists with a subsistence led mixed animal economy.
In the early 1st century AD a nucleated settlement was established about 120 m to the south at Longdoles Field, characterised by a series of large and intensively re-cut enclosures, gullies, pits and substantial boundary ditches. Within the site was identified a number of activity areas associated with domestic habitation, small scale metalworking, and stock management. It appears to have operated a largely subsistence economy associated with cattle ranching, and in this respect was very similar to the nearby site at Thornhill Farm. However, there were noticeable differences between these sites, with Claydon Pike having an increased emphasis upon the settlement boundary, along with larger numbers of imported goods associated with eating and, to a lesser extent, drinking.
The early 2nd century saw dramatic changes. The enclosures, gullies and ditches of the earlier phase were replaced by two large rectangular enclosures, a substantial aisled barn and an aisled house with a tiled roof and painted plaster interior. This was probably the residence of an extended family group, and there is evidence to suggest that they utilised elements of Roman style dress as well as new eating and drinking habits. The economic basis of the site incorporated the management of hay-meadows, probably on a commercial basis to sustain the needs of growing local population centres such as Cirencester. By the mid 2nd century AD the settlement expanded onto adjacent gavel platforms which seem to accommodate the lower status estate workers, as well as providing small stock enclosures and industrial areas. What may have been a religious precinct was also established at the heart of the settlement, overlooking a central open space. At some point during the early 4th century AD there appears to have been deliberate and widespread clearing of the site with much domestic and structural material deposited within pits and ditches. This was undoubtedly connected with the establishment of a modest masonry footed villa on the site comprising two separate structures, the southern of which incorporated a hypocaust room. The villa was substantially modified during the 4th century, and had two successive enclosures built around it, probably signalling a shift in emphasis towards a perception of an increased need for security.
A small inhumation cemetery plot lay 100 m to the west of the villa and a well built masonry footed shrine was constructed to the east. The surrounding gravel terrace and floodplain were no longer used for haymaking and had reverted to grassland used for grazing animals. It is possible however that some arable production may have occurred on certain gravel islands within the villa estate. Other economic activities may have included bee keeping, fishing, and salting and/or curing of meat and fish. The final abandonment of the villa at Claydon Pike is unclear, but there is some evidence to indicate activity of some kind until the start of the 5th century.
A small group of inhumation burials were cut through the villa building, three of which were assigned to the middle Saxon period by radiocarbon dating. Further intermittent activity took place on site in the medieval period.

Top of page

Somerford Keynes, Neigh Bridge

A series of salvage excavations between 1986 and 1988 at Somerford Keynes Neigh Bridge revealed part of a late Iron Age and Roman settlement. The earliest features were a series of curvilinear enclosures dating from the early/mid 1st century AD to the early 2nd century AD, although the evidence hints at some level of middle Iron Age occupation in the vicinity. These represent part of a farmstead, although a late Iron Age/early Roman religious focus is also suggested by an unusually large number of coins and brooches. A system of trackways and ditched enclosures and a large aisled building replaced the earlier features in the early 2nd century AD. The building was associated with a large quantity and variety of tile, indicating that it may have served at least in part as a tile depot, associated with wider changes in landscape organisation during this period. Sculptural fragments of the Capitoline triad point to an official religious presence. No features can be securely dated much beyond the end of the 2nd century AD, although a substantial number of late 3rd and 4th century coins and small finds suggests late Roman activity of some kind. An official or military presence is hinted at during this late phase.

Top of page

Whelford Bowmoor

Three archaeological investigations were undertaken at Whelford Bowmoor in 1983, 1985 and 1988. The earliest features revealed during excavations comprised a regular system of sub-rectangular enclosures, dating to the early/mid 2nd century AD. The enclosures were probably used for livestock management. Another group of smaller enclosures lay further to the west, which were probably a later development, but still possibly used in some aspect of livestock management. It seems likely that many of these enclosure ditches went out of use by the latter part of the 2nd century AD, and the only convincing evidence for domestic activity within the site dates to this period, taking the form of a rubble building platform and associated 'midden' deposits, dating from the later 2nd to early/mid 3rd century AD. The finds assemblage on the whole suggests that the later phase of the site was of higher status, with relatively high quantities of imported fine and specialist wares in addition to finger rings, bracelets and remains of hobnail shoes.
The site lies upon the immediate floodplain of the river Coln, and there is some reason to believe that incidences of flooding were slowly increasing throughout the 2nd and 3rd centuries AD. In response to this, the site may have been occupied only on a seasonal basis, and if such were the case, then the quantities of fine ceramics associated with eating and drinking, together with the much higher quantities of animal bone from the later phase, may be explained in terms of seasonal feasting associated with the re-occupation of the site. The general absence of later 3rd or 4th century AD material from the site suggests that settlement and structurally defined agricultural activity may have shifted from the area entirely, towards drier locations further up the gravel terrace, which were less prone to flooding and waterlogging.

Top of page

Kempsford Stubbs Farm

Archaeological evaluation and excavation took place at Stubbs Farm from 1991 to 1995, specifically targeting a multi-ditched circular and rectangular enclosure known from cropmark evidence. The two enclosures seem to relate to separate phases of activity, with the circular feature tentatively dated to the Iron Age/early Roman period. The scarcity of finds associated with the use of the feature suggests a non-domestic function, possibly the corralling of animals. The rectangular enclosure clearly belongs to a later phase of activity, in the 2nd and 3rd centuries AD. The overall character of this phase is indicative of a low status rural farmstead operating a largely pastoral economic regime amidst the grasslands of the lower gravel terrace and floodplain. Despite the paucity of what may be termed high status material culture, the rectangular enclosure itself represents a considerable investment of labour. The site clearly relates to a wider system of field boundaries, trackways and settlements in the local area, with the rectangular enclosure linked with the 2nd century settlement just to the west at the Multi-Agg quarry site. Both the Stubbs Farm and Multi-Agg settlements appear to have gone out of use by the second half of the 3rd century, with the latter probably lasting slightly longer.

Top of page


Location and geology

Most of the key Cotswold Water Park Project sites lie within the triangular area between the modern settlements of Lechlade, Fairford and Kempsford, centring on the confluence of the rivers Coln, Leach and Thames (Fig. 1.3: Geological map of project area ). The main group lay along an east-west line from Whelford Bowmoor, by the River Coln in Kempsford Parish, via Thornhill Farm and Claydon Pike, Longdoles Field in Fairford Parish to Claydon Pike, Warrens Field in Lechlade Parish. These sites all lay within a block defined by the river Coln to the west and south, and the river Leach to the east. The Kempsford Stubbs Farm site lies about 2 km south of Whelford Bowmoor, south-west of the river Coln and about 1 km north of the Thames. The 'Survey' sites were mainly concentrated to the east in Lechlade Parish, to south-east of the Leach and north of the Thames. Other survey sites lay between the Coln and Leach and one lay to the south of Thames in Buscot Parish, Oxfordshire. In addition to this main concentration of sites was Somerford Keynes, Neigh Bridge, which lies about 17 km WSW of Claydon Pike. It was located adjacent to the Thames, 6 km due south of Cirencester and just south of the village of Somerford Keynes.
Geologically, all of these sites lay on the 1st gravel terrace of the river Thames, with Whelford Bowmoor being sited upon the immediate floodplain of the river Coln (Fig. 1.3). To the south, past the main low-lying areas of alluvium on the Thames floodplain lie areas of Oxford clay. The ground then rises significantly towards the sands and clays of the Corrallian Ridge, which also includes significant outcrops of Coral Rag. To the north of the sites lie higher gravel terraces, then in some cases further areas of Oxford clay. Bands of Cornbrash, Forest Marble and Great Oolite limestone then mark the rise into the Cotswolds.

Top of page


The post-excavation process

The post-excavation programmes of Claydon Pike, Somerford Keynes and Whelford Bowmoor followed on from the fieldwork and continued until the late 1980s and early 1990s. During this phase of post-excavation work, substantial progress was made particularly in stratigraphic phasing and on the analysis of finds assemblages and environmental data for Claydon Pike. Matrices were produced and a phasing scheme for the site was developed. Digital archives for contexts and finds data were created and specialist reports were produced. Interim reports were also published for Claydon Pike (Miles and Palmer 1983, 1984, 1990).
Little substantive work was undertaken on the stratigraphic data or finds for Somerford Keynes, but the site at Whelford Bowmoor was phased and a draft publication text produced, including site description and finds reports. A draft report was also written in the late 1990s for Kempsford Stubbs Farm, while the finds from the survey sites were catalogued and a brief report made in the 1980s.
With the exception of Stubbs Farm, most of the post-excavation work for the Cotswold water Park sites was carried out in the 1980s, and an assessment of this data indicated that many significant revisions were needed. Therefore, all of the original finds and stratigraphic reports have been fully revised and updated, and much additional work has been carried out, leading to the publication of four complete site reports. The original emphasis on landscape interpretation has been maintained, so that the sites are viewed as components within their local and regional context. Thornhill Farm has been the subject of a separate post-excavation programme that has now been completed (Jennings et al. 2004), although the wider landscape aspects of this site still form a prominent part of the current print volume.

Top of page


Research aims and objectives

The following are the research aims and objectives for the current publication, based upon an assessment of the archive and the need to address contemporary research agendas in British Iron Age and Roman studies.
Aim 1: Settlement landscapes and people: the landscape as social expression

  • Settlement development and building forms: what light do the Cotswold Water Park sites throw on later Iron Age and Romano-British settlement development?
  • Power: What light can study of the late Iron Age and Romano-British artefact assemblages from the Cotswold Water Park sites throw on our understanding of social structure and power within settlements and between settlements?
  • Resources and their control: can we observe regional settlement hierarchies that can be related to the control and exploitation of resources?
  • Religion: how does the evidence from the Cotswold Water Park sites throw light on religion and beliefs in the later Iron Age and Romano-British periods?
    Identity: how does the data from the Cotswold Water Park sites provide direct evidence for expression of identity?


Aim 2: Regionality: Dobunni: the Cotswolds and the Thames Valley; Atrebates: Berkshire Downs; Catuvellauni: Oxfordshire and Hertfordshire.

  • What light do the Cotswold Water Park sites throw on the question of the regional variations in the material culture and agricultural practice of Iron Age and Roman Britain?
    Aim 3 Chronologies, and agencies and processes of change, including the dynamics of Romanization
  • How far can the Cotswold Water Park sites throw light on the processes of change - settlement development, agricultural intensification and increasing craft specialisation during the middle and late Iron Age?
  • What evidence is provided by the Cotswold Water Park sites for the transition from Iron Age to Roman and for the role of the Roman military?
  • How does the data from the Cotswold Water Park sites contribute to our understanding of local, regional and provincial patterns of development in the 2nd and 3rd centuries AD?What evidence is there for the changes in occupation and material culture in the late and sub-Roman period?

Top of page