Excavating the CA archive – Saxon Shore, Part I

The Saxon Shore (litus Saxonicum) was a military command consisting of fortifications built on both sides of the English Channel, which were established in the late 3rd century AD to defend Britain against the incursion of Saxon raiders. Its nine English forts are mentioned in the Notitia Dignitatum, a document that details the administrative organisation of the late Roman Empire. Over this and next month’s column, I will visit each of these forts in turn, from the exposed north Norfolk coast to the shelter of Portsmouth harbour. Current Archaeology has covered all nine sites over the years, but as CA 182 (November 2002) notes, ‘archaeology provides an incomplete record, and in the case of the shore forts, the gaps in the evidence are chasms’. This is perhaps why there has never been a comprehensive survey of these sites in Current Archaeology, and why there is no straightforward starting point for this column. Nonetheless, I will flag three interesting articles to commence with: CA 275 (February 2013) on North Sea coastal communities, CA 332 (November 2017) on the Anglo-Saxon inhabitants of the Fens, and CA 336 (March 2018) on how the Roman military shaped industrial infrastructure in Britain, all of which refer to the workings of the Saxon Shore. What these articles reveal is more than just the history of each fort in isolation, but rather a palimpsest that includes the local farming communities that supplied them, the troops who garrisoned them, and what finds from these sites reveal about changes in late Roman society.

Ground-penetrating radar (GPR) surveys of Branodunum were conducted by Time Team in 2012, revealing the layout of the shore fort in remarkable detail, as was reported in CA 276.

BRANODUNUM, BRANCASTER, NORFOLK

Branodunum lies to the east of the modern village of Brancaster, halfway between Hunstanton and Burnham Market. The fort was built to guard the Wash approaches, and during the Roman period its northern wall lay directly on the seashore, which served as a harbour; since then, the shoreline has shifted, and the fort now lies inland of salt marshes. The site has been subject to periodic investigation, including by Time Team in 2012, who carried out ground-penetrating radar (GPR) surveys. Producing one of the great ‘wow’ moments of the television show, these revealed a wealth of detail about its layout (see CA 276, March 2013). What is especially interesting about the site is that it has been studied both inside and out, including explorations of the civilian settlement immediately beyond its east gate, as reported in CA 44 (May 1974), which showed the community’s complete layout down to individual house plots. Later evidence for settlement has also been found there, including during the Anglo-Saxon era, as was reported in CA 85 (December 1982), and – as will become a pattern across this column – this is usefully compared to other settlements both near and far away, such as the early medieval village at Dunston’s Clump, near Babworth, Nottinghamshire. Closer to home, CA 211 (September 2007) suggested links between the fort and the Roman villa at Gayton Thorpe near King’s Lynn (perhaps as a supplier of grain and other materials), 32km (20 miles) south along the Icknield Way. The larger context of these Roman Fenlands is examined in CA 295 (October 2014): was this in fact one great imperial estate, with the farmsteads and the forts inexorably interlinked and the roads, rivers, and canals shipping grain, salt, and other commodities to military distribution centres like Brancaster? The magazine has over the years flagged finds, too, that show the shifting cultures of these forts – notably, from Brancaster, a 4th-century ring bearing a Christian inscription which was reported in CA 284 (November 2013).

GARIANNONUM, BURGH CASTLE/CAISTER-ON-SEA, NORFOLK

The archaeological record of England’s Saxon Shore forts is incomplete, with some sites better studied than others. The example at Portchester, Hampshire, made the cover of CA 187 when the magazine examined the career of Stephen Johnson and particularly his work at the Heritage Lottery Fund (now the National Lottery Heritage Fund).

Gariannonum has been identified as either Burgh Castle on the River Waveney, located just before its confluence with the River Yare to the west of Great Yarmouth, or at Caister-on-Sea, 9km (6 miles) away at the mouth of the River Bure to the north of Great Yarmouth. During the Roman period, the sites lay on opposite shores of a large estuary, the modern remnant of which is Breydon Water. The significance of landscape- and seascape-change in this area since the Roman era is crucial to our understanding of both sites: as the map in CA 270 (September 2012, p.32) demonstrates, the ‘great estuary’ of the River Yare made water transport around this area the easiest option, including travel inland to the major local town of Venta Icenorum/Caistor St Edmund. This last site, just to the south of modern-day Norwich, was the capital of the Iceni people. Its interior is explored in CA 270 (September 2012); see also CA 344 (November 2018), CA 356 (November 2019), CA 380 (November 2021), CA 393 (December 2022), CA 406 (January 2024), and CA 419 (February 2025), as well as next month’s issue, for coverage of more recent work outside the settlement’s walls. Both Burgh Castle and Caister-on-Sea have been investigated since the 1950s, revealing a range of finds that indicate their high-status interconnections: for example, a collection of late Roman glass from Burgh Castle, which was reported in CA 186 (June 2003).

OTHONA, BRADWELL-ON-SEA, ESSEX

On p.32 of CA 270, a map depicts the ‘great estuary’ of the River Yare, and the sites of Burgh Castle and Caister-on-Sea on opposite shores.

Othona lies near the modern-day village of Bradwell-on-Sea, at the edge of the Dengie Peninsula, controlling entry to the estuaries of the rivers Blackwater and Colne – the latter waterway leading to Camulodunum/Colchester (see my column in CA 424, July 2025). Although some of the fort’s Roman building material was reused in the extant 7th-century chapel of St Peter-on-the-Wall, most of the former site has since been lost to coastal erosion. (St Peter is compared to St John the Baptist at Reedham, which has similar late Roman origins, in CA 420, March 2025.) Enough of the ramparts survived into the 17th century for it to be described by the local historian Philemon Holland as a ‘huge ruin’, but given the scale of land-loss, the site has never been subject to in-depth investigation. It remains the most elusive of the nine shore forts (and the only one not to have had any direct reporting in Current Archaeology). However, the Anglo-Saxon chapel of St Peter, standing lonely among the salt marshes, remains a wonderfully atmospheric place to visit.

REGULBIUM, RECULVER, KENT

Regulbium lies near the modern-day village of Reculver, on the north Kent coast to the east of Herne Bay. The fort stood on the mainland side of the northern entrance to the mile-wide Wantsum Channel, which separated the Isle of Thanet from the mainland. By the 7th century, Reculver had become a landed estate of the Anglo-Saxon kings of Kent, and the site of the Roman fort was given over for the establishment of a monastery, subsequently developing into a township, until coastal change forced its decline in the 17th century, causing it to be largely abandoned by the late 18th century. Due to the site’s multi-period significance, it has been investigated from the 1950s onwards (see CA 196, March 2005). A marker of the cosmopolitan nature of Roman Reculver was reported in CA 58 (September 1977), which featured a review of Roman mirrors found in Britain, including a rare example of a silvered glass mirror from the fort, comparable with similar finds from Ospringe, also in Kent, and York. As at Gariannonum, though, the most interesting reporting from Current Archaeology comes not from the forts themselves, but from the places the forts were connected to. CA 193 (August 2004) reported on Minster-in-Thanet, a Roman villa that is landlocked today but in the past lay close to the shore of the Wantsum Channel. This site was excavated between 1996 and 2001, revealing an Iron Age farmstead that evolved by the 1st century AD into a substantial villa complex, feeding into both the military networks at Reculver and the civilian ones at nearby Richborough. A connection from later in the site’s history then comes in CA 284 (November 2013) and CA 355 (October 2019), highlighting the architectural similarities between the 7th-century royal settlement (especially its church) at Lyminge, 40km (25 miles) south-west of the fort, hinting at the cultural connections that existed across Kent in the period.