Two thousand years of farming in the Severn Vale

Middle Bronze Age to Roman remains at Cleevelands, Bishop’s Cleeve

Bishop’s Cleeve, a popular village near Cheltenham, has expanded rapidly since the 1990s, providing opportunities for archaeologists to investigate the underlying remains. It occupies low-lying ground within the Severn Vale, with the Cotswold uplands nearby. Although much of the vale lies above clays, Bishop’s Cleeve sits on well-drained sandy gravels, and this probably attracted the Iron Age, Roman, Anglo-Saxon and later people whose traces have been recorded during previous investigations. When Persimmon South Midlands decided to develop land north-west of Bishop’s Cleeve for housing, Cotswold Archaeology were called in to investigate the archaeology.

The site, looking east towards Bishop’s Cleeve, with the Cotswold hills in the distance
The site, looking east towards Bishop’s Cleeve, with the Cotswold hills in the distance

The earliest remains were a few flint tools spanning the long duration (10,000–3000BC) of the Mesolithic to Early Neolithic and dropped by hunter-gatherers visiting the valley floor.

Log ladder
Log ladder

Longer term settlement was established during the Middle Bronze Age (1500–1100BC), by which time the site occupied an established agricultural landscape, with grazing, arable fields, hedgerows and woodland patches. Within the site, a single roundhouse and ancillary building were built. Some 85m away, a small cemetery contained the cremated remains of seven individuals, perhaps the former inhabitants of this family sized farm which would have been one of many within this well managed landscape. Several large pits were found 200m from the roundhouse, within what was then damp ground alongside woodland or a hedgerow. These deep pits had been lined with wood, but their functions are uncertain: they are unlikely to have been waterholes or quarries, and one possibility is that they were shafts used for ritual purposes, such as communing with underworld deities.

scutching knife
Scutching knife

One of the pits contained the remains of a log ladder, hewn from a single piece of oak, as well as a tankard-sized vessel made from stitched bark, and a wooden implement used for processing plant fibres. The ladder had been used to access the pit floor, but the bark vessel, wooden implement and other items (antlers, a gold strip, a flint arrowhead, pottery and animal bone) were probably cast into the pits as offerings.

Middle Bronze Age pit with log ladder in situ
Middle Bronze Age pit with log ladder in situ


The site was re-occupied during the Early to Middle Iron Age (700–100BC) when several roundhouses were built across what continued to be an agricultural landscape. Enclosures on drier sandy gravels in the southern part of the site were probably for penning cattle and sheep/goats that were grazed on open clay land alongside the Dean Brook. Several grain storage pits suggest that this farmstead also had an arable element.

This pattern of enclosures on the drier sands/gravels and open grazing to the north persisted beyond the Roman invasion. Isotope analysis of sheep and cattle bones from the site indicates that most of the farm’s animals were reared there; the exception may have been brought in from the Cotswold uplands, although whether this points to transhumance or the importation of breeding stock is unknown. A smithy within the farm had been used to repair and recycle tools and jewellery.

Arable production intensified from the 2nd/3rd centuries AD. An unusually high number of millstones suggests proximity to a watermill, whilst the remains of portable clay ovens reflect the provision of hot food to agricultural labourers during harvesting or sowing. Harvested crops would have been winnowed in the fields then dried within ovens to prevent spoilage when stored. One such oven contained charred cereals from a subsequent stage of processing, where the grains taken from storage are heated to harden them to facilitate milling.

The implications are that during the later Roman period, grain was milled at a central mill and labourers worked the fields; this suggests intensification and centralisation, perhaps indicating that the farm had become part of a villa estate. Indeed, Roman roof and flue tiles from the site do suggest proximity to a Romanised building. As part of the wider economy, the farm could have supplied nearby Glevum (Gloucester), accessible within a day’s travel.

A small number of these farmers were buried within the site. They were robust, having lived long enough to display diseases associated with ageing and to show signs of their farm labouring. They had adopted elements of Roman culture, seen for example in their choice of pottery for food storage, preparation and serving, and in their use of brooches to fasten clothing. They also retained Iron Age traditions, burying a neonate and cow in ditches, burying some adults in crouched positions and enjoying an Iron Age-style diet rich in meat, dairy and emmer wheat. They probably viewed themselves primarily as farmers, rather than ‘Roman’, although no doubt they interacted in different ways with different people depending on context. The range and quality of the finds is typical for a farmstead of this period, and these farmers’ wealth would have been invested largely in their livestock and crops. They were able to trade for jewellery, pottery and, presumably, perishables, but were frugal, recycling and repairing when possible, whilst depositing material within pits and ditches as propitiatory acts to ensure future productivity.

What effect the development of a villa nearby would have had on their lives and attitudes is unknown but there are hints that during the mid/later Roman period they were integrated into a villa estate, the owner of which may have had more Romanised aspirations. Burials from these centuries were placed in the extended body position typical of Romano-British traditions and the impression is that engagement with a Romanised market economy went hand in hand with greater acceptance of Roman cultural expressions. 

Early post-Roman remains were sparse, but settlements of this date can be difficult to detect and are under-represented in the archaeological record. The latest dated cremation grave at Cleevelands dates to the 5th/6th centuries AD, whilst Anglo-Saxon organic-tempered pottery from a pit dates to the 5th–8th centuries.

A full report can be downloaded from the Reports Online page of this website (report 18495).

Author: Jon Hart

Share this!

A Roman site at Llanwern, South Wales

Between September 2019 and January 2020, Cotswold Archaeology undertook an excavation at Llanwern in South Wales on behalf of RPS Planning and Development and Redrow, on a site occupying the east face of a hill overlooking the Gwent Levels.

The investigations revealed a complex including several stone-founded Roman buildings, terraced into the hillside. These included two stone-built circular buildings, a rectangular structure and an apsidal structure, with evidence for activity between the 2nd and 4th centuries AD. At least three other buildings were represented by partial ring-ditches, one containing a kiln or oven. Environmental samples taken from the oven may allow its function to be determined.

The rectangular structure, which measured 17m long and 7m wide, appears to have been the main building at the site. It had been subject to at least three phases of modification over time. The western side of the building had been cut into the terrace, with made ground to the east used to provide a level construction surface for the structure’s eastern extent. Two areas of mosaic flooring were identified, possibly representing a room and an associated corridor, which had potentially been separated by a wooden partition wall. The mosaics have been dated on stylistic grounds to the 4th century AD.

The main rectangular building, facing east
The main rectangular building, facing east
Remnants of mosaic from one of the rooms of the building
Remnants of mosaic from one of the rooms of the building

The function of the rectangular building is currently uncertain, although a complex series of stone-lined culverts drained water from a group of springheads at the west of the site down to the east, indicating that water-management played an important role in the use of the site. Many finds were recovered, including large quantities of Roman pottery, animal bone, ceramic building material (including tile), brooches and coins. The presence of unusual plate brooches amongst the assemblage raise questions as to whether the complex may have had a religious focus. One of the tile fragments bears the stamp of the Roman military unit Legio II Augusta, which was stationed at nearby Caerleon, suggesting a possible connection between the sites.

To the immediate north of the rectangular building, an impressive and well-constructed east-west aligned metalled trackway was formed from large unworked limestone blocks; the trackway measured at least 50m long and 4.8m wide, with its size and method of construction suggestive of a high volume of traffic.


The well-constructed trackway
The well-constructed trackway

At the west end of the site, an apsidal structure was terraced into the hillside; the internal area of the structure contained natural stone that had been exposed to create an area of hard standing, with a surface constructed on top, presumably for extra consolidation. The function of the structure is currently uncertain; interpretative possibilities being considered include a nymphaeum, a church, a shrine or possibly even a theatre, which utilised the slope as a spectator area.

The apsidal structure and adjacent rectangular building
The apsidal structure and adjacent rectangular building

This structure abutted a heavily truncated rectangular building, which contained a cremation burial and an inhumation burial. The latter burial had been placed in a prone (face-down) position. These burials raise the possibility that this building formed part of a mausoleum.

Other features identified at the site include an area of metalworking, water tanks and features interpreted as possible quarry pits.

It is hoped that analysis during the post-excavation stage will allow questions about some of the unusual aspects of the site to be answered.

Share this!

Weyhill Medieval Punishment Cemetery – Monograph Release

Following four years of hard work across all our teams, CA has just published a monograph on the medieval punishment cemetery discovered in Hampshire.

Weyhill cemetery monograph coverIn 2016, a watching brief in advance of construction at a site between Andover and Weyhill revealed disturbed human remains. This formerly isolated roadside location, which had been partly developed in the later 20th century, lies close to the historical boundary between Andover and Abbotts Ann Hundred. The discovery led to the excavation of a densely crowded group of partly disturbed and heavily intercut graves, and it soon became clear that this was not a normal community cemetery. The remains of some 124 individuals, mostly male and adult, were revealed, with much additional disarticulated bone. Many had clearly been buried with their hands tied behind their backs or had been interred face-down, and some had been decapitated.

As might be expected with a rural ‘execution’ (or punishment) cemetery of this type, very few artefacts were recovered. In this case only one find, a silver penny dateable to AD 979-985, allowed close dating, which suggested that the site was a distinctive type of Saxon cemetery known elsewhere in Hampshire. However, radiocarbon dates from articulated burials, selected for their spatial and stratigraphic positions, demonstrate that the cemetery was in use from at least the 10th century (and possibly as early as the 8th century), and continued in use beyond the Norman Conquest to at least the 13th and possibly the 14th century. Stable isotope analysis shows that the people buried here were mostly of local origin.

The significance of the site lies in this evidence for continuity, at a local level, of the organisation of judicial power from Saxon times well into the Middle Ages. It is the first site to provide clear evidence for such duration of a ‘punishment’ cemetery.

The book can be purchased from Oxbow for £19.95



  1. Introduction and Background
  2. Excavation Results
  3. Human Bone
  4. Artefacts and Faunal Remains
  5. Radiocarbon Dating and chronological modelling
  6. Isotope Analysis
  7. The Medieval background and Legal System
  8. Discussion

Appendix 1 – Analysis of Local Early Medieval Sources

Appendix 2 – After the Cemetery

Share this!

Cotswold Archaeology Wins Prestigious Book of the Year Award!

The 12th annual Current Archaeology Awards were held on Friday 28th February at Senate House in London, and we are very pleased to announce that our collaborative book, Life and Death in the Countryside of Roman Britain, won the award for Book of the Year!

Life and Death in the Countryside of Roman BritainThe book is the third and final volume in the ‘New Visions of the Countryside of Roman Britain’ series; it deals with the rural people of Roman Britain…how they looked, lived, worshiped, died, and how they were treated in death. Cotswold Archaeology was a major partner in this hugely successful Leverhulme Trust funded project, along with the University of Reading, the Archaeology Data Service and Historic England. CA’s Neil Holbrook acted as a series editor for the volumes, and the volume was co-authored by our very own Tom Brindle (Post-Excavation Manager). More information about the project is available on the ADS website, where you can download a free PDF of volume 1 in the series, and access the online database.

A review of the book was published in the January 2019 edition (no. 347) of Current Archaeology, and the volume can be purchased from Oxbow Books.

2020 is CA’s most successful year so far at the awards, with our first win to date and a record three nominations in total. CA’s two other nominations were also for collaborative projects, with Revealing Boxford’s Ancient Heritage nominated for Research Project of the Year, and Exercise Shallow Grave up for Rescue Project of the Year.

Congratulations to all of the award winners!

Share this!

Keynes, trains and automobiles

We were clearly trying too hard to come up with a witty headline but the facts are clear, when it comes to the big infrastructure projects we stand side-by-side with our clients, managing their complex archaeological requirements. Whether it is in the planning or construction stages of projects, we have the skills and experience to handle the scheme specific logistical tests that invariably accompany the more traditional archaeological challenges. The last few years has seen the Milton Keynes team busy working on significant energy, airport, road and rail infrastructure projects across central and eastern England.

Drone Image of archaeological works
Excavations in advance of construction of the A120 Little Hadham Bypass and Flood Alleviation Scheme, Hertfordshire

During the late summer of 2019, while working on a highway scheme in Hertfordshire, our team topped 50 archaeologists on site. While, by most measures, this has to be acknowledged as a large archaeological project, in this instance the significance was not the scale but the unique circumstance that we had colleagues from all of our five offices working alongside each other. One company; one high standard; one large talent pool to draw from.

large-scale trial trenching project satellite image
Trial trenching during the most recent phase of work at Rugby. Credit: Google Earth

But energy and transport projects don’t have the monopoly on scale; we have been working on some seriously large-scale strategic housing schemes too. Several different five-hundred plus trial trenching projects in Northants, Leicestershire, Rugby and Essex have kept our teams occupied. At Rugby, we are nearing our 10th anniversary working together with the client team; with more site work planned for later this Spring.

These longer running, sustainable urban extension projects present a different set of challenges such continuity of personnel, retaining invested-knowledge and keeping the momentum going on post-excavation work and publications; all matters that we are well-versed with. In the next few weeks we are looking forward to returning to Bidwell West, Central Beds; another large-scale strategic housing scheme that we have been working on for several years.

Share this!

A moated site near Tewkesbury, Gloucestershire

Last autumn, the CA field team excavated a moated site near Tewkesbury, Gloucestershire. The moat had largely been infilled during the late 20th century, while post-medieval domestic buildings within the interior were demolished following a fire in 2005. The moat is now surrounded by a large industrial estate and business park, although it originally occupied a rural location. Despite this, it was deeply stratified, requiring three phases of mechanical stripping.

The demolished remnants of the post-medieval domestic buildings within, and contemporary bridge across, the moat
The demolished remnants of the post-medieval domestic buildings within, and contemporary bridge across, the moat

At the west of the site, the first site strip identified the later moat fills as well as the 18th and 19th-century domestic structures, all of which correlated closely with the available 19th-century cartographic evidence. Disappointingly, the moat here was relatively shallow, typically no more than 2m deep, and only 18th to 20th-century artefacts were recovered from within the four hand-excavated interventions.

The island created by the moat contained a number of later medieval/early post-medieval ditches and a stone culvert, which were clearly cut by the moat’s western arm before extending beyond the limit of the site. This suggested that the western part of the surviving moat was not medieval in origin, but possibly part of a much later formal landscaped garden, contemporaneous with the post-medieval buildings.

Outside the moat, immediately to the north, re-stripping revealed the footprint of a large medieval aisled barn. This stone building was very close to a post-medieval brick bridge across the moat’s northern arm, indicating that the building and the bridge were not contemporaneous. The stone barn was at least 35m long, 8m wide, and had eight surviving central stone post-pads within its interior. It had been roofed with Cotswold stone slates, although only the smaller pegged tiles survived.

The post-medieval western arm of the moat
The post-medieval western arm of the moat

Hand-excavation within the moat’s northern and eastern arms revealed a strikingly different sequence, and here it reached a depth of 3.5m. Organic-rich fills lay beneath the post-medieval deposits and, importantly, the earliest fills produced exclusively 12th to 14th-century pottery.

A bridge, contemporaneous with the medieval moat fills, was identified crossing the eastern arm, comprising interior and exterior stone abutments of finely dressed and coursed limestone. Two large, squared timbers (each 5m long and 0.4m square) were incorporated into the stonework, founded on the exposed natural clays at the moat’s base. Several upright timbers had been incorporated into the base timbers with mortice and tenon joints. While the interior stone abutment survived in good condition, the exterior stonework had suffered a catastrophic failure, which presumably rendered this bridge unusable.

The medieval stone abutments and exposed timber structure
The medieval stone abutments and exposed timber structure
Evidence for the failure of the exterior stone abutment pushing against the upright timber
Evidence for the failure of the exterior stone abutment pushing against the upright timber

The findings bring into question the true extent of the medieval moated site. It is now speculated that the original medieval moat enclosed a much larger area, probably extending west into an area truncated by modern flood alleviation works. This postulated part of the moat may have been truncated (or completely abandoned) when an early post-medieval mill race was constructed along the western limit of the site.  The surviving western arm is certainly much later (possibly 18th century) and runs parallel with the mill race; its most likely interpretation is as a post-medieval garden feature, albeit one that perhaps re-instated the dynamics and grandeur of a moated site.

Cliff Bateman


Share this!