Known as ‘Banna’ to the soldiers who garrisoned it almost 2,000 years ago, Birdoswald Roman fort has long been a key focus for Hadrian’s Wall research. In fact, it is one of the most-extensively excavated forts on the line of the Wall. The area immediately outside its fortifications has seen virtually no digging, however, although extensive geophysical surveys by Alan Biggins and David Taylor in the 1990s revealed an impressive extra-mural settlement that, at its peak, was two-to-three times the size of the fort itself.
Now a new project headed by Newcastle University and Historic England (facilitated by English Heritage, who care for the site) has set out to learn more about this external settlement, and how it related to the military community based inside the fort. Leading the project are Tony Wilmott, who has been directing excavations at Birdoswald for over 30 years, and Professor Ian Haynes – and, on a hot early August day this summer, they kindly welcomed CA to the Roman frontier and told us about what they had found so far.
The project is scheduled to run for five digging seasons spread over four years, and the initial 2021 phase saw three trenches opened outside the fort. Two of these were located to investigate different parts of the extra-mural area to the east, while a third was placed a short distance away to explore the northern side of Hadrian’s Wall. This latter trench aimed to shed light on the flat area between the foot of the Wall and its outer ditch, known as the berm. Far to the east, previous excavations at several sites in Tyneside have found the berm there to be peppered with pits that are thought to have held sharpened stakes creating an obstacle almost like barbed wire (see CA 195). Did this barrier feature more widely along the frontier fortifications, and might similar traces be found at Birdoswald?
As it happened, the Birdoswald trench revealed no trace of pits like those to the east – which raises further questions about how these defences may have functioned. The western stretch of the Wall was originally built in turf on a different line to its stone successor, so it could be that investigating this earlier incarnation might reveal the presence of ‘barbed-wire’ pits at Birdoswald. Alternatively, Ian Haynes said, the pits could be a local peculiarity specific to the Tyneside area that was not used further to the west – or, indeed, Birdoswald might be the anomaly in not having them.
Moving from ephemeral evidence to rather more substantial features, the project’s other two trenches revealed the unmissable remains of sturdily built structures. Trench A was designed to investigate an area where geophysical survey showed a complex array of substantial structures, and to include an area where Ian Richmond had briefly dug in the 1930s (one of the few previous excavations to explore the extra-mural area). There, Richmond had recorded the remains of a structure that he interpreted as a tower or turret, but the present project has revealed instead to be a monumental rectangular building.
Although robbed in places, its walls can still be clearly seen, with 13 courses of large, finely worked stone blocks surviving to an impressive extent. The topmost course terminates in a smooth, even surface, suggesting that the building had been carefully brought down in a methodical demolition at the end of its life, and Tony Wilmott suggests that the building may have been in part subterranean or semi-basement, with four steps leading down into it from the ground surface.
Trench A also uncovered the wall of a second large building a short distance from the first, together with two adjacent chambers, one of which had a distinctive curved (apsidal) end. Running through these remains were a series of water mains that had been added at a later date, distinguished by the line of their course as well as the metal collars that once joined their long-decayed wooden pipes. The date of these features is not yet known, but they raise intriguing questions about the longevity of the settlement, Ian Haynes said, pointing to a possible parallel at Verulamium (Roman St Albans) where Sheppard Frere found similar pipes that he argued were mid-5th century in date or later (CA 237). ‘It’s one thing to find people living in the shadow of Roman buildings after AD 410, but the idea that they had that kind of infrastructure in the post-Roman period is quite potent,’ Ian said. ‘The presence of the iron collars and their relatively late position in our stratigraphic sequence is what reminds us of the Verulamium finds, but we have no evidence that they are that late in date.’
Tony added: ‘What is clear is that the water supply system post-dated and cut through the apsidal structure, as did a further stone building with an impressive flagstone floor. The water main bifurcated into two lines, one on each side of this structure, with which it was certainly contemporary in use.’
Evidence for post-Roman occupation within the fort is much more persuasive: previous investigations found that the Roman granaries within its walls were later replaced by two successive large timber halls thought to date to the 5th and 6th century (CA 116 and 240). Might this new phase of fieldwork find traces of early medieval activity in the extra-mural settlement too? While the jury remains out for whether the water mains represent a post-Roman addition, the 2021 excavation has uncovered some possible hints of later visitors to the site. While investigating the rooms adjoining the second large building in Trench A, the team discovered a long, narrow pit cutting through the Roman floor surface, which Tony and Ian believe is a grave cut. Given that the remains of the demolished building’s walls are thought to have remained visible above the surface for some time after the structure went out of use, perhaps the monumental stonework caught the interest of people moving through the landscape some time later.
‘Thirty-four years of involvement with excavation at Birdoswald did not prepare me for the astonishing level of preservation,’ Tony said.
Ian agreed: ‘Looking at the buildings in Trench A, you really appreciate the quality of the preservation at Birdoswald. It is something that we are blown away by. These structures also highlight the array of forms that were in the extra-mural settlement beyond strip houses and farm buildings – these aren’t identikit Roman structures, and we are often wrestling with questions of interpretation. We think that these may have been some of the more important buildings associated with the extra-mural area: their masonry is very finely worked and we suspect that there may have been a bath suite close by – we know that there are hypocausts associated with some of the other structures in the area, and within our own trench we found the carved foot of a bath bench similar to those seen at Chesters.’
Signs of life
Tony and Ian also wanted to examine more ‘ordinary’ houses within the settlement, and Trench B has uncovered the footprint of what is thought to be a strip house associated with the north road leading out of the fort. Interestingly, while this structure originally fronted on to the road, it was later extended so that it truncated part of the route, even incorporating some of its flagstones into the new floor. Finds from the front of the house speak of hospitality, including gaming counters, fragments of fine tableware, and part of a black slipped ware (Moselkeramik) flagon with its clay stopper still in place – this container would have held wine from Trier, speaking of wide-ranging interests and commercial connections within the settlement. Further back from this public-facing area, though, there may have been workshops within the structure – Ian described the discovery of ‘globs of glass’ that might point to glass-working having taken place in this location.
Beside this structure was a structure that was harder to characterise, perhaps a second building or some kind of platform, and in the gulley between this and the strip house the team found two small portable altars made of local sandstone. One of them, around a foot tall, bears a faint inscription that the team has not yet been able to read, though attempts to decipher its meaning, aided by white light scanning, are ongoing. The other altar is almost pocket-sized, about as long as a human hand – ‘you can’t get much more portable than that,’ Ian said.
The vast majority of the small finds from this summer’s excavation have been 2nd- and 3rd-century in date. This provides a marked contrast to the much longer occupation of the fort, but does fit into a wider pattern of change in settlement evidence that is known to have taken place in the 3rd century. It has been noted during excavations at several other forts that their extra-mural settlements seem to have been abandoned at around this time, while life still went on within the walled military area. Tony and Ian hope that the project’s remaining digging seasons will add more to this picture, allowing the cumulative finds from all the planned extra-mural trenches to be compared to material from the fort, as well as to discoveries from their previous excavation of an extensive but eroding 3rd-century cremation cemetery just west of the fort in 2009 (CA 353). ‘Taking a holistic approach to the spaces, the structures, the timeline of the settlement, we will be in a better position to people some of these places,’ Ian said.
Tony Wilmott is Senior Archaeologist at Historic England. Ian Haynes is Professor of Archaeology at Newcastle University.
For more information about Birdoswald Roman fort and visiting the site (please note that this season’s excavations have now concluded), see www.english-heritage.org.uk/visit/places/birdoswald-roman-fort-hadrians-wall.
Ian Haynes and Tony Wilmott have also published another major research project, focusing on Maryport and covering five years of excavations at the Roman cult site. Watch out for a feature drawing on their discoveries in a future issue of CA; for more information on the book, see https://cumbriapast.com/cgi-bin/cwaas/cp_main.pl?action=cp_publications_list.
All photos: Historic England / Newcastle University (drone photos taken by Lesley Davidson).