Most of England’s monumental mounds are assumed to be Norman castle mottes built in the period immediately after the Conquest – but could some of them have much earlier origins? Jim Leary, Elaine Jamieson, and Phil Stastney report on a project that set out to investigate some of these mighty constructions.
Eleven years ago, a team of archaeologists, engineers, and miners opened a tunnel into the heart of the huge Neolithic round mound of Silbury Hill in Wiltshire. At 31m tall, Silbury is the largest prehistoric mound in Europe, and well deserves its classification as a ‘monumental mound’ – a monument type that is among the rarest and least well understood in Britain. Investigating and recording inside Silbury was a wonderfully complex project, and it took almost exactly a year to complete, between 2007 and 2008 (see CA 293).
As this exploration began to wind up, though, and with the post-excavation phase looming, the project team found themselves talking about other round mounds to which this site could be compared. Of course, nothing actually compares to the great hulking eminence of Silbury – in that respect it remains unique, ‘but’, we said, ‘there is another large round mound not far cut the spiral path terraced into the mound’s sides that leads to its summit. The origins of the Marlborough Mound, however, have been the subject of debate for decades. The celebrated antiquarian Richard Hoare was, in 1821, among the earliest to suggest it could be prehistoric in date, and you can see why – although much smaller than Silbury, Marlborough is of similar form, and in a comparable topographic location: low-lying and next to a river and springs. Without any dateable material from the mound itself though, whether Marlborough was indeed a prehistoric sister to Silbury had to remain an entirely theoretical question. Or did it?
We had no intention of starting a whole new project along the lines of the Silbury tunnel – far too destructive and costly (the former investigation was in direct response to conservation needs and not solely for research). But we needed a way of accessing material deep within the Marlborough Mound without any major intrusive excavations. The answer we came up with was to drill a small diameter core (only 10cm across) running from the top of the motte right the way down through the mound make-up until it had reached – and gone through – the underlying old ground surface.
These cores, which come out in metre lengths of plastic tubing, could then be taken back to the labs and analysed. It seemed a neat solution and, happily, when we suggested this method to the Marlborough Mound Trust, who were in the process of undertaking conservation work on the site, they immediately approved the plan and agreed to fund it. October half-term in 2010 saw a specialist geotechnical company drill two bore holes from the summit to the base of the mound, and the resulting cores were analysed at the Historic England (then English Heritage) labs at Fort Cumberland and samples sent off for radiocarbon dating.
The results were striking: they clearly showed that the main body of the mound was contemporary with Silbury, dating to the second half of the 3rd millennium BC. Our hunch had proved correct: the Marlborough Mound was another Neolithic monument hiding from us in plain sight. It was a prehistoric mound that had been reappropriated as a motte in the Norman period, just as it had been reused as a garden mount later on.
AN IMPROBABLE PROJECT
The obvious question that arose from this work was: how many other ‘Norman’ mottes up and down the country were also constructed in the prehistoric period? Their supposed medieval date is actually largely speculation – an idea put forward in the early 20th century by castle scholars like Ella Armitage – when, in fact, no one has ever attempted to date castle mounds archaeologically; they are too large to excavate and too expensive to tunnel into, before even considering the damage that work like that would cause to the sites themselves.
Seen in this light, could other large prehistoric monumental mounds exist, fossilised as later medieval mottes? Can we at least put rough dates on these huge and much-loved monuments? When we applied for a research grant, we put these questions to the Leverhulme Trust, who loved the idea. It was, as they say in management circles, a ‘no-brainer’ – whatever dates we got back, the project would be a success, as it would significantly advance our understanding of the mottes under investigation, prehistoric or otherwise – so, in early 2015, the Trust agreed to fund the Round Mounds Project. This would be a two-and-a-half-year initiative run by a team from the University of Reading with colleagues from the Scottish Universities Environmental Research Centre (SUERC) dating laboratory in East Kilbride. We had one simple, central aim: to identify mottes in England with prehistoric potential and, through a programme of targeted, minimally intrusive investigation, determine their date of construction, how they developed, and the environmental context in which they were built.
A MOTLEY CREW
The fieldwork is now complete, the dates have been processed, and the results are in. Radiocarbon dating revealed that 70% of our sample mounds were indeed constructed in the medieval period; more specifically, in the decades around the late 11th or early 12th centuries AD, placing their probable construction in the years following the Norman Conquest. For example, the motte at Wallingford Castle returned a minimum age (a terminus post quem – ‘limit after which’) at 95% confidence of AD 1026-1151. Northamptonshire’s Fotheringhay Castle, the birthplace of Richard III, returned a minimum date of AD 1033-1176, while Pilsbury Castle in Derbyshire yielded a minimum age of AD 1056-1215.
So far, the mounds we had studied appeared to be exactly what they had always been suspected to be: Norman castle mottes. Our coring work has produced a host of new information on the material make-up of such monuments, granting new insights into the siting and construction of motteand- bailey castles. Assessment of the cores from Clifford Hill castle motte (located towards the eastern fringe of the county town of Northampton), for example, revealed clear layering of deposits from the time that the motte was raised – but radiocarbon dates from three samples taken from different locations within the mound were statistically indistinguishable, suggesting the motte had in fact been constructed in a relatively short period of time. Over at Tickhill Castle in South Yorkshire, coring revealed that this motte was constructed almost entirely of sand, its medieval builders sculpting and enhancing a natural sandstone outcrop to create an impressive mound appearing to stand over 17m high.
Meanwhile, our earthwork surveys have also produced huge amounts of new information on not only the mounds but the castles themselves. Investigations at Fotheringhay, for example, revealed the layout of the buildings within the inner bailey, including the likely location of the castle’s Great Hall, where Mary Queen of Scots was executed in 1587. At Castle Hill, Bishopton, we were able to determine that the bailey had been accessed from the north by way of a raised causeway – a feature which was clearly secondary to the initial castle construction, and which gives us an insight into how these sites could change through time. Finally, we have identified echoes of how many of these sites were reused long after the castle itself was abandoned: at Bramber Castle in West Sussex, for example, where the site was turned into a visitor attraction in the early 20th century – with the castle’s curtain wall enclosing a tea garden and fairground – the remnants of this tearoom can still be seen as slight earthwork remains.