Roman Caerwent: antiquarian and archaeological research

During one of his many itineraries in the early 1540’s, the antiquary and poet John Leland visited the village of Caerwent in Monmouthshire. Occupied by 16 or 17 cottages, mostly built with stone robbed from Roman buildings, Leland described Caerwent as ‘sumtyme a fair and large cyte’.

Other scholars would follow in Leland’s footsteps and they began recording ancient monuments more methodologically than had been the case before. William Camden, for example, was the first to recognise that Caerwent was the site of the Roman city of Venta Silurum and in his great work Britannia, first published in 1586, he wrote: ‘Venta, a very ancient city … whose name neither the rage of men nor time has yet extinguished’.

Caerwent is called Venta Silurum in the surviving copies of Roman maps and itineraries, meaning ‘Market of the Silures’. Indeed, Caerwent’s name itself is a reference to the village’s ancient past: Caer is a Welsh word corresponding with the Latin castra, meaning ‘fortification’, ‘camp’ or ‘stronghold’, while Went is derived from Venta, or ‘market’.

The earliest recorded archaeological discovery at Caerwent comes from 1777 when a rich mosaic pavement was discovered to the south-east of the village (the landowner initially erected a cover-building over it, but later took this down exposing the pavement ‘to vandalism and weather’). The mosaic’s location would be explored by Octavius Morgan (MP for Monmouthshire and a notably antiquary in his own right), who in 1855 conducted the first recognisably modern excavation at Caerwent, revealing part of a small bath-house and an adjacent building.

It would be another 44 years until further investigative work was undertaken at Caerwent and the plan of Venta Silurum comes to us largely as a result of the remarkable excavations that took place at Caerwent between 1899 and 1913. These campaigns were organised and run by the Caerwent Exploration Fund (CEF), established in 1899 by the Clifton Antiquarian Club in Bristol to undertake ‘the systematic exploration of the site’. The CEF excavations were supported by Godfrey Morgan, 1st Viscount Tredegar (also the Fund’s President) and directed by Alfred Hudd and Thomas Ashby.

About two-thirds of the Roman city’s interior was explored over 14 excavation seasons (investigating 11 of the city’s 20 insulae), producing a wealth of new information about Venta that was published in the journal Archaeologia (the CEF also established a small museum in Caerwent where finds were put on public display). Although much of what we know today about the layout and history of Venta derives from the work of these Edwardian pioneers, their rapid excavations inevitably concentrated on recovering the plans and functions of the uppermost masonry buildings (any earlier structures, especially those constructed of timber, were left largely untouched or unrecognised).

Chronological detail is generally absent from Ashby and Hudd’s reports and, consequently, the plan we have of Venta probably shows mainly the late Roman city’s layout. Ashby’s involvement with Caerwent lessened after his appointment as Director of the British School at Rome in 1906 and, although Hudd continued to excavate at Caerwent until his death in 1920, the outbreak of the First World War led to the closure of Caerwent Museum in 1915 and the CEF was formally dissolved in 1917.

After the First World War, the recently founded National Museum of Wales (AC/NMW) began its long involvement at Caerwent when V.E. Nash-Williams investigated the public baths opposite the forum (1923), before examining the southern side of the curtain wall prior to its consolidation (1925). The city’s walls were often densely overgrown with trees and other vegetation and between 1930 and 1935 Nash-Williams also cleared the west wall, while W.F. Grimes investigated the north wall in 1930 in advance of the construction of a new bypass road (the current A48).

Excavations at Pound Lane in 1946-7 by Gerald Dunning (Inspector of Ancient Monuments at the Ministry of Works), were the first to take place in the centre of the village since the CEF campaigns. Dunning revealed a sequence of workshops fronting onto the main road, as well as a large courtyard house to the rear, and his excavations are significant because, for the first time at Caerwent, they identified the remains of timber buildings from Venta’s earliest occupation (unfortunately, these important excavations have never been fully published).

In 1981 the AC/NMW proposed a new series of excavations within the walls of Venta Silurum. This programme of research began with an excavation in Insula I (not excavated during the CEF campaigns), which uncovered a late 2nd-century building that was rebuilt at the end of the 3rd century, before being replaced again at the beginning of the 4th century by a substantial residential house with rooms arranged around two central courtyards. The late Roman house provided a high standard of living with painted wall-plaster, mosaic floors and hypocaust heating systems in two rooms (the excavation of the courtyard house has not been published).

The Romano-Celtic temple immediately to the east of the forum-basilica was re-excavated between 1984 and 1988. The investigation of the earliest levels on this site showed that the temple had been built on top of a sequence of roadside strip-buildings and yards. The first of these was constructed in the late-1st or early-2nd centuries and was a timber post-built structure, after which the plot was remodelled several times when the buildings were rebuilt in stone.

It is likely that these were workshops, with small rooms at the front serving as shops onto the city’s main road. The last of these industrial, commercial and domestic buildings was demolished around 330 and replaced with the temple within a sacred precinct that remained in use, albeit with alterations, throughout the 4th century (the excavation of the temple has not been published).

The AC/NMW’s proposal to re-excavate the forum-basilica highlighted a lack of knowledge regarding the chronology of Venta Silurum and it was anticipated that the project could provide a reliable date for the construction, use and disuse of this public building complex. The Edwardian excavators were unlikely to have identified any traces of pre-basilican occupation and the new excavations would look for remains that might aid the understanding of the settlement’s early history. The end of the forum-basilica was another primary objective of the excavation and it was thought that there was a good chance of locating ‘sub-Roman’ structures.

Although the excavation of the forum-basilica has not been published, an Open Access article summarising the architecture and history of the building appeared in the journal Britannia in 2022.

In 2001, Cadw initiated the Research Framework for the Archaeology of Wales. Caerwent was highlighted as a key site in southeast Wales, which is recognised as of ‘more than national significance’, where ‘large areas are available for study’.

Although geophysical techniques had been widely used in archaeology since the 1980s, the first surveys at Caerwent were undertaken only in 2008 and 2012. Cardiff University and GeoArch organised several geophysical surveys in the NE and SE corners of the Roman town, as well as significant swathes of the zone immediately outside the southern and western walls. These surveys provided significant new information about numerous buildings and other archaeological features buried beneath the modern ground surface at Caerwent, although the absence of cemeteries or evidence for extramural occupation outside the city walls is surprising.

Further resistivity and ground penetrating radar (GPR) surveys were carried out by Geophysical Surveys of Bradford in advance of Time Team excavations at Caerwent in 2008. The geophysical surveys located an unknown and extensive villa-like building in the northern part of Insula I, as well as the walls of buildings in the previously unexplored Insula XIV.

The Time Team excavation consisted of 7 evaluation trenches in Insulae I and XIV, as well as outside the east gate. The two trenches excavated in Insula I investigated the courtyard villa complex located during the GPR survey, revealing a large multiphase building surrounding a central courtyard with an apsidal-ended range at its eastern end (perhaps a bath-house).

This sophisticated building was provided with tessellated and opus signinum floors, painted wall plaster and roofed with Old Red Sandstone tiles. No evidence for the date of its construction was found, but coins and pottery suggest that it was occupied into the second half of the 4th century when the building possibly was destroyed by fire and subsequently robbed.

Wall footings identified in the Insula XIV trenches probably represent the remains of a narrow roadside strip building, altered and perhaps re-aligned on at least two occasions. There was little evidence for the outward appearance of this structure, but numerous glass fragments and part of an opus signinum floor, probably added during the late 3rd century at the earliest, indicate the presence of relatively comfortable living quarters within it. The only archaeological features located on the eastern approaches to the town consisted of a probable roadside ditch and a gravelled track or path leading towards the east gate. As with the GeoArch surveys to the south and west of the city, surprisingly little evidence for extramural cemeteries or occupation was discovered (the Time Team evaluation excavation has not been published).

Reconstruction of Venta Silurum by Alan Sorrell (Crown copyright (2020) Cadw)


Reconstruction of the forum-basilica at Venta Silurum by John Banbury (Crown copyright (2020) Cadw)