Subscribe now for full access and no adverts
On the edge of the Great Estuary, into which flow the rivers Bure, Yare, and Waveney, St John the Baptist, Reedham, stands on a small promontory of sand and shingle. Although the existing structure was mostly built c.1300 (though its west tower dates to the mid-15th century), it stands out from Norfolk’s many other medieval churches for the quantity of reused Roman building material in its fabric. Initially, archaeological attention focused on the Roman brick and tile, which is widely used in the nave and choir, and is especially prominent in the north wall, where a section from foundation to eaves is entirely built in this way. More recently, though, following a disastrous fire in 1981, Edwin Rose observed that the grey quartzite stone that is also abundant in the church has similarly ancient origins. The same stone was used to build the Roman shore fort at Brancaster on the north Norfolk coast. Roman brick and stone therefore account for the great majority of building materials used in the medieval church – but what kind of structure had they been recycled from?
Tellingly, Professor John Allen traced the grey quartzite stone to the Leziate Beds of the Lower Cretaceous formation that overlooks the shores of the Wash in the Hunstanton–King’s Lynn–Downham Market area, in the north-west of the county. This is relatively close to Brancaster, but represents a journey of around 80 miles (130km) by sea to Reedham. Sourcing materials over such a distance suggests organised efforts, and it is likely that the Roman building for which they were destined had an official, possibly military, purpose. There has been considerable speculation over the years as to the nature of this structure and, given the position of the church in signalling distance of the Roman forts at Caister-on-Sea and Burgh Castle, a favoured interpretation has been that of a pharos (lighthouse) or signal station – or a building serving both functions – presumably located close to the site of the church. But where, exactly?
Over the years, several locations around the village have been suggested for where the Roman structure may have stood, but the obvious starting point is the church itself. In order to try to resolve this question – and, perhaps, establish the character of the Roman building itself – between 2013 and 2016 a small team from the University of Reading’s Department of Archaeology undertook geophysical surveys inside the church and within the churchyard, using ground-penetrating radar. A subsequent, more targeted survey was carried out too in 2017, focusing on an area close to the walls of the church. The results of these investigations were intriguing: within the church, strong anomalies could be seen along the north and south sides of the nave, indicating wall remains running roughly parallel with each other but not quite lining up with the footprint of the present structure (deviating from the church’s orientation by about 10°). Meanwhile, across the churchyard there were further hints of possible building remains to the west.
Unfortunately, it was not possible to follow up on the results of the survey within the church by excavation, but there were areas within the churchyard without headstones (and, potentially, without graves) where we could carry out very limited archaeological explorations, carefully targeted in order to avoid disturbing nearby burials. In total, we opened eight small trenches, each initially limited in size to 2m by 1m, though three were subsequently extended. Most were dug directly outside the church, positioned as close as possible to the interior locations that had produced the strongest radar signals, while two other sondages were excavated a short distance to the west, to explore geophysical anomalies within the churchyard. Neither of these latter trenches revealed any underlying archaeology – but three of the church-focused ones produced positive results.
Roman remains
The clearest evidence of a Roman building preserved beneath the church came from Trench 7, which we opened on the south side of the nave, towards its eastern end. This trench had been positioned in order to explore why the interior radar survey had shown such strong results on the other side of the wall, and it did not disappoint, revealing a 2.2m-long section of Roman masonry. Measuring over 1m wide, this gently curving wall had a core of grey quartzite rubble and broken tile, faced with three tile courses above two courses of grey quartzite stone. The wall was set into a surface of hard, creamy mortar and, with its impressive dimensions and curved line, might represent the base of a tower or bastion.
On the north side of the nave, diagonally opposite Trench 7, a second, smaller fragment of Roman masonry, aligned north–south, emerged from Trench 4. Although only the lowest course of construction had survived (set on foundations of rough, unworked blocks), it was built in a similar way to the Trench 7 remains, comprising a rubble core faced with grey quartzite blocks, surrounded by creamy mortar. Finally, although Trench 3 – dug against the west face of the church tower – did not reveal any trace of actual masonry, it contained a mortar-rich fill suggestive of a robber trench that might testify to the presence of another Roman wall long-since plundered of its materials.
These discoveries were encouraging, but the biggest surprise came from Trench 5, which had been placed against the north wall of the nave, either side of a blocked-up arch which was loosely dated to c.1200. There, at a depth of about 0.45m, we found a floor surface made up of fragments of Roman brick and tile in a crumbly pale cream mortar, into which a few sherds of 10th-century Saxon pottery (identified by Alice Lyons) had been trampled. Extending the trench a little further to the north produced a better-preserved surface of white plaster, as well as the remains of a wall aligned east–west and made up of mortared pieces of Roman brick and tile.
Our immediate interpretation was that this, too, represented the remains of a Roman structure, but two samples taken from birch and oak charcoal found immediately beneath the mortared surface, and very close to the section of nave wall that is constructed solely of Roman brick and tile, gave radiocarbon dates of between the mid-7th and mid-to-late 8th century. Not what we were expecting! In fact, we had chanced upon evidence of the early history of the church, including an area – perhaps a side chapel or porticus – which had been accessible from the nave via the arch of c.1200. Measuring at least 11.5m by 3.75m, it survived long enough to be incorporated into the new, larger church that was built around 1200, but it was demolished around a century later, when the archway was blocked up and a new window inserted into the resulting wall.
Interpreting the evidence
What can we conclude from these investigations? We have found unequivocal evidence of a Roman building beneath the church, whose scale and style of construction – including the evidence of a possible tower or bastion in Trench 7 – is consistent with that of a military structure. The distance that some of its materials had travelled to be used in the main body of the building (as opposed to in decorative elements like columns) would also be unusual in a domestic building.
If we assume that the excavated evidence corresponds to the possible wall remains indicated by the geophysical survey within the church, we might envisage a single, substantial structure measuring at least 12.5m north–south by at least 20m east–west. Indeed, if our interpretation of the remains in Trench 3 as a robber trench for another Roman wall is correct, we could extend the east–west dimension by a further 9m, making the building c.30m long. Meanwhile, the fragment of Roman masonry discovered in Trench 4, to the north of the church tower, is a little to the north of the east–west-aligned anomaly picked up by the radar within the nave, forming a right angle with this line. This might suggest that the Roman structure also extended further to the north, though it is possible, too, that this fragment might have belonged to a separate, second structure.
While our small trenches produced no dating evidence for the Roman remains, the shared use of the Leziate rock with the Saxon Shore fort at Brancaster might suggest contemporaneity, placing the Reedham building between the late 2nd and mid-3rd century. However, the build style of the curving length of wall in Trench 7 recalls that of nearby Burgh Castle, raising the possibility of a later date in the second half of the 3rd century. As for the nature of this structure, assuming a military role, the scale of the excavated remains is not at all consistent with what might be expected of a fort, but they could fit a fortlet (albeit one on the small side), or a defended pharos or signal station.
Tentatively, it is possible to offer an interpretation that gives us a tower a least 12.5m square, with a potential height of 20m, standing within an enclosing wall around 4.5m high and at least 20m in length in one dimension. This reconstruction is much more modest in scale than the known, possibly 2nd-century pharos at Dover, with the remains having more in common with the much later (late 4th-century) signal stations that have been identified along the North Yorkshire coast. No obvious parallel is known with the same date as that proposed for Reedham, but a lighthouse or signal station would fit with a period when other defensive measures were being taken along the east coast.
As for how such a structure might have functioned, a lighthouse at Reedham could have guided vessels into the Great Estuary and towards the Yare, presumably working in conjunction with a second light near Caister-on-Sea, where the estuary opened into the North Sea. As a signal station, it would have been intervisible with both Caister and Burgh Castle – but was it also designed to give warning of raiders to the local civitas capital, Venta Icenorum, which lay 15 miles up the Yare at what today is Caistor-St-Edmund? (See CA 270 for more about the walled town itself, and CA 419 for the latest update on the Caistor Roman Project’s excavations on its outskirts.) In this scenario, there would probably need to have been at least one intermediary station further up the river. And, if Reedham’s role had been as a signal station at the mouth of the Yare, would there have been similar provision close to the mouth of the Bure, meaning that (with Burgh Castle by the Waveney) there were three installations giving cover to all the rivers flowing into the estuary?
The wider picture
Just as intriguing as the Roman remains is the unexpected insight that we gained into an early phase of the church at Reedham, probably dating from between the mid-7th and the mid-to-late 8th century. While the use of Roman brick and tile in church construction is widely attested elsewhere in Norfolk (as was recently illuminated by Peter Wade-Martins; see CA 416), and in eastern England more generally, reused blocks of Roman limestone, rather than brick, are dominant at 8th-century St Peter-on-the-Wall at Bradwell in Essex. Reedham is, therefore, a very rare instance where a good case can be made for the first church having been built predominantly of brick and tile, not just its north wall. As for other examples in England thought to have been built initially entirely of these materials, parallels can be found in Canterbury: at St Martin’s, and at the churches associated with St Augustine’s Abbey.
Within the latter complex, St Peter and St Paul (the primary church, built c.597-619) and St Mary’s (c.620) survive only as fragmentary footprints – but the foundations of both structures are made entirely of Roman brick and tile, hinting that their whole fabric could have had a similar composition. Meanwhile, the 7th-century church of St Pancras is better preserved, with the north wall of its west porch still standing to a height of over 3m, and its construction is also comprised of recycled Roman brick and tile. In contrast to these ruins, St Martin’s (which is located a five-minute walk from the abbey and forms part of the same World Heritage Site, together with Canterbury Cathedral) stands intact and still operates as an active church. Although it was expanded and altered in both the medieval and Victorian periods, substantial above-ground elevations built solely or largely of reused Roman brick and tile still survive, laid in horizontal courses.
St Martin’s was possibly founded as early as the late Roman period (and is otherwise dated no later than St Augustine’s mission to Kent in 597), and this early incarnation offers an interesting parallel with the site that we have been investigating, as its dimensions are comparable with Reedham’s c.6m-long nave. Unlike St Martin’s and the other churches discussed above, however, Reedham represents the only example where the brick courses have been laid alternately horizontally and herringbone fashion. Might Reedham, though, have had similarly early origins? In the late 12th-century Liber Eliensis there is an intriguing reference to St Felix founding the first church at ‘Redham’ in the 620s or 630s. If ‘Redham’ is Reedham, perhaps our investigations may have identified the remains of St Felix’s church, incorporated into the later medieval structure of St John the Baptist.
The building of the first church at Reedham was a labour of love. Assuming a total or near-total build of reused brick and tile, we have estimated that around 10,000 pieces would have been required. Bearing in mind that, in the Roman structure(s), courses of tile were regularly interspersed with several courses of stone, we can easily envisage how the original building would have had to have been totally dismantled in order to obtain a sufficient number of tiles. The perfectly useable stone was initially deliberately set aside, though it was subsequently chosen to be employed along with the brick and tile in the later medieval church, as we see today. However, although this work was done with care, we can also see from the surviving work that few, if any tiles came away from the hard mortar intact. Moreover, the wastage rate was high: our small trenches alone yielded around 1,000 small fragments of brick and tile. Although a labour of love, this would have been an undeniably labour-intensive task – but one that has left a distinctive and intriguing legacy for us to explore today.
Further reading: This article summarises a longer report: M Fulford, A Clarke, E Durham, S Machin, N Pankhurst, R Fry, D Thornley, and R Williams-Burrell (2024) ‘The Roman fortlet or defended watch tower/pharos at St John the Baptist, Reedham, Norfolk and the early medieval origins of the church’, Archaeological Journal 181(2): 208-242; https://doi.org/10.1080/00665983.2024.2412412 (open access).
All images: Mike Fulford, unless stated otherwise
