A villa unveiled

7 mins read

Uncovering luxury living and ‘ritual activity’ in Roman Oxfordshire

Archaeological work in rural Oxfordshire has uncovered the remains of a winged corridor villa that was occupied for much of the Roman period. Carly Hilts spoke to Louis Stafford to learn how the story of this long-lived, high-status residence is evolving as investigations continue.

Hypocaust tiles with archaeologist working in background.
The remains of a Roman villa are being excavated on the outskirts of Grove, near Wantage in Oxfordshire. Although most of its buildings appear to have been systematically demolished and their materials recycled elsewhere in the late Roman period, vivid clues to its former status can still be seen, including hypocaust tiles testifying to the fact that some of the complex’s rooms had underfloor heating. PHOTO: Red River Archaeology Group

On the outskirts of Grove, a village near Wantage in Oxfordshire, archaeologists exploring an area scheduled to become a new housing development for Barratt and David Wilson Homes have uncovered traces of a much earlier ‘des-res’: a previously unknown Roman villa. Red River Archaeology have been excavating the site for over a year, working from west to east and recording a diverse array of features, including extensive Roman field systems. When they reached the eastern extent of the planned estate, however, they found signs of something even more significant: the unmistakable outline of a modest villa complex, lying almost outside the area set for investigation.

Pieces of painted wall plaster in bright reds and yellows.
6 pieces of mosaic tiles (tesserae) in red and white
Scraps of colourful painted plaster and mosaic tiles (here in a modern placement) speak of the grand designs that once adorned the villa’s interior spaces. PHOTOS: Red River Archaeology Group

The remains of a Roman villa are being excavated on the outskirts of Grove, near Wantage in Oxfordshire. Although most of its buildings appear to have been systematically demolished and their materials recycled elsewhere in the late Roman period, vivid clues to its former status can still be seen, including hypocaust tiles testifying to the fact that some of the complex’s rooms had underfloor heating.

The Roman residence was a winged-corridor-type villa, a fairly common design in British archaeology, comprising a central block of rooms linked by a passageway, with flanking ranges of other buildings to the sides. In this case, the (larger) western and (more subsidiary) northern wings of the east-facing villa have been exposed; its eastern range, if it has survived in the heavily ploughed soil, lies outside the scope of the present excavation. Despite this partial picture, though, Red River Archaeology have been able to estimate that the villa complex would have measured around 50m by 45m: not a sprawling, palatial estate, but one that would nevertheless have been a dominant feature of the local landscape, and one that still bore many of the key markers of high-status occupation. Fragments of hypocaust tiles indicate that at least some of its rooms boasted underfloor heating, while quantities of colourful painted plaster – decorated with bright banding and floral motifs – as well as orange and white tesserae from mosaic floors, speak of an attractive interior design. The villa’s inhabitants were evidently influential enough to access far-reaching trade networks: analysis of pottery from the site is ongoing, but they were using Samian table-wares imported from Gaul.

Even more impressive than the villa, however, was the monumental aisled structure that had once stood next to it. Its well-made and expertly squared-off chalk foundations were much deeper and more substantial than those of the villa, easily sturdy enough to have supported a second storey, and within these walls four large, square chalk bases – possibly supports for columns or large wooden posts – hint at an internal colonnade.

Aisled buildings are known from many villa sites in Britain, though their precise purpose – and their place in the chronology of their respective complexes – remains debated. Some examples are thought to have been built first, before later being incorporated into a larger complex, while others appear to have replaced existing villas. In the case of the Grove aisled building, while its exact phasing is still being pinned down, the current interpretation from the excavation team is that it is probably a later element of the site – and, at 43m by 15m, it is one of the largest examples of its kind known in Britain.

Reconstructing the villa

However imposing the Grove complex’s buildings would have been in their heyday, today their remains are reduced to robbed-out footprints plundered so comprehensively that in some places even the foundations have been taken out. The villa appears to have been demolished at the end of its life, probably during the late Roman period, and much of its fabric recycled elsewhere. Senior Project Manager Louis Stafford described his ‘brilliant’ site team led by Francesca Giarelli as recovering ‘maybe a couple of wheelbarrows’ of ceramic building material (CBM) from within 100m of the villa, and very little in the way of roof or floor tile: finds that you would normally expect in abundance. By contrast, the team have identified an estimated 50kg of wall plaster, not in situ on collapsed walls, but mainly as tiny fragments found among the rubbish that had been used to back-fill ditches – the opposite way round to what you would expect, Louis said, suggesting that the CBM had been systematically removed from the site.

An aerial view of the villa site with overlay of coloured lines marking a provisional site plan.
This provisional site plan, overlaid on a drone photograph of the villa remains, reconstructs its possible appearance. It was a winged-corridor type, and the image below picks out the complex’s northern and western ranges, as well as the aisled building. IMAGE: Red River Archaeology Group

In direct contrast to these ghostly structures, however, artefacts lost by the people who lived and worked there have proven plentiful. Over 2,000 small finds have been documented to-date, including more than 400 Roman coins (mainly low denomination) spanning the 1st/2nd centuries through to the late 4th century. It is possible that the villa’s lifespan was even longer, however: one key piece of evidence to support this suggestion was a distinctive buckle decorated with two outward-facing horse heads. Classed as ‘Type 1B’ by Hawkes and Dunning’s 1961 typology, these objects are thought to have been worn between the mid-4th and mid-5th centuries, and are often associated with individuals linked to the late Roman army. They are known from locations across Roman Britain – search for DUR-455F88 on the Portable Antiquities Scheme’s database (https://finds.org.uk), for example, for a recent discovery from North Yorkshire, as well as links to other examples from locations as disparate as Northamptonshire, Somerset, Hampshire, and Warwickshire. The collections of Amgueddfa Cymru – Museum Wales include one from Caerwent, too. These brooches also appear to have been carefully curated as curiosities or family heirlooms, as they have been recovered from a number of early Anglo-Saxon graves as well. In the case of the Grove find, this particular buckle might suggest that its owner had connections to social elites linked with the army, or that they at least wanted to be associated with that kind of prestige.

As well as extensive Roman field systems, structural evidence of agricultural activity was found just to the north of the villa remains, including T-shaped dryers that would probably have been used for processing corn and/or hops.

Aerial photo of distinctive T-shaped outline in stone of a corn or hops dryer.
As well as extensive Roman field systems, structural evidence of agricultural activity was found just to the north of the villa remains, including T-shaped dryers that would probably have been used for processing corn and/or hops. PHOTO: Red River Archaeology Group
A complete bronze buckle decorated with two outward facing horse heads.
This buckle, dating to AD 350-450 and adorned with two horse heads, is one of the most telling clues hinting at the villa’s longevity. Such objects are often associated with the late Roman army, although they have also been found in a number of early Anglo-Saxon graves. PHOTO: Red River Archaeology Group

Other items of jewellery, including rings and brooches, most likely represent the personal possessions of the villa community, but some of the metalwork recovered by the team is more enigmatic. Chief among these are around a dozen small strips of lead, tightly wound to create little coils that are strikingly similar to artefacts commonly called defixiones or ‘curse tablets’. Around 300 such items are known from other locations in Roman Britain, concentrated (as Roger Tomlin describes in his 2021 article ‘The Latin curses from Uley and other sanctuaries in Britain’) in the non-military south, and associated with two key sites in particular: Bath, which has produced around 130, and Uley in Gloucestershire, where around 80 inscribed pieces of lead were excavated in the 1970s. Further finds (albeit in much smaller numbers) are known from elsewhere in south-west Britain, such as at Lydney, also in Gloucestershire; Brean Down and Pagans Hill, south-west of Weston-super-Mare and Bristol respectively; and Leintwardine in Herefordshire; as well as the legionary fortress at Caerleon, in Newport, Wales – but they are not exclusive to this region, and others are known from places including London, Farley Heath in Surrey, and just outside Venta Icenorum, the walled Roman town near Norwich.

Where their text can be deciphered, these items typically record the prayers of individuals who have been wronged in some way – usually through the theft of items ranging from clothing and jewellery to livestock – and who are seeking the intervention of their chosen deity to secure justice and the return of their property. Various gods and goddesses appear to have been entreated in this way – Uley’s tablets mainly call on Mercury, Venta Icenorum’s example names Neptune, Bath’s favour Sulis Minerva, and Lydney’s invoke another syncretic deity, Mars-Nodens. What these finds have in common, though, is their fairly formulaic phrasing, and the fact that many of them have been recovered from sites where the remains of temples or shrines have also been excavated. No such structure has been identified on the Grove villa estate, though it is possible that one of its rooms or buildings had a ceremonial role on a more informal or domestic level. Unfortunately, although the scrolls were certainly associated with the villa and its occupation, they were not tellingly concentrated in one place but scattered around the site.

Frontal and overhead view of 6 rolled lead curse tablets.
Unrolled lead curse tablet, blank.
Around a dozen lead scrolls have been found scattered around the villa site. Similar examples from other locations such as Bath and the temple at Uley bear texts that have been interpreted as curses, but the ones from Grove are blank. PHOTOS: Red River Archaeology Group

This is an extract of an article that appeared in CA 411. Read on in the magazine, or click here to read it online at The Past, where you can read all of the Current Archaeology articles in full as well as the content of our other magazines, Current World ArchaeologyAncient Egypt, and Military History Matters.

Leave a Reply

Your email address will not be published.