When I arrived at the silt, I couldn’t believe my eyes. There weren’t any dark marks that might have revealed buried pits or wells, but instead I was able to glimpse something unique – and far more exciting. Running across the silt in a gently curved line were two absolutely parallel dark marks, each about two inches wide. About 20 metres (22 yd) away I could just make out two other parallel lines that seemed to cross and intersect with the main set. And I knew exactly what they were. I called over a couple of experienced excavators and together we outlined the edges of the marks with the tips of our trowels. As the marks started to fade, we found they could be brought back using sprinkler watering cans.
The marks were clearly the ruts made by a two-wheeled vehicle and the thickness of the wheels was about the same as the Flag Fen wheel.9 The two wheels were about 1.1 metres (3½ ft) apart, which would suggest a small cart or carriage, appropriate for a single person. At one point, the vehicle had got stuck in the mud and we could see how it had been rocked forwards and backwards to release it. There was no possibility that it had been a four-wheeled cart-like vehicle. The silts in question had probably been laid down by the River Welland flooding at various times between about 800 and 600 BC.
13.6 A plan of later Bronze Age (800–600 BC) wheel ruts at Welland Bank Quarry, near Market Deeping, Lincolnshire.
© Francis Pryor
When I was a boy in the 1950s, I can remember seeing similar-sized vehicles being towed by donkeys (always known as ‘asses’) in the Irish village in County Carlow where my mother’s family lived. They were used to take their owners to the village pubs and it was said that the asses knew the way home without being driven or steered. I think the Bronze Age vehicles were used in a similar way, mainly to transport light goods and people. The wheels could not have been cheap to make, so I suspect the people who owned the little carriages would have been quite well off and fairly senior in the local tribal structure. All we have to do now is find conclusive evidence for the pubs.
a Medium Density Fibreboard.
b As we saw in Scene 11.
c Sadly, this is a process that we can witness to this day in places like the Amazon Basin, although in Brazil it has recently been hastened by some disastrous political decisions.
d A pin through the end of an axle to keep the wheel in position.
e See Scenes 2 (page 25) and 9 (page 172).
Scene 14
Life in the Sky: Hillforts (1200–100 BC)
Maiden Castle – Danebury – Dorset Hillforts – Pen Dinas and Cardigan Bay
Exciting archaeology is about great sites and great archaeologists – and in some happy instances it can be about both. There are, however, dangers in linking certain people with particular places for too long. Ideas about the place that were expressed during their lifetime can become fossilized and fixed. Up-and-coming younger people tend to hesitate before offering alternative views and interpretations, which may often be based on good, solid evidence. I have seen this happen in my own area of expertise – indeed, I have sometimes been a bit of an old fossil myself – until the emerging facts forced me to change my ways. Old ideas can sometimes prove hard to dislodge, because they can be so comforting – and they are often linked to good friends and colleagues who are no longer with us. So simply to discard them would seem somehow disloyal. Indeed, the relationship between prehistorians and the sites they are trying to untangle can be as complex as the lives of the people being investigated. As time passes, I find this complexity both exhilarating and humbling. It’s what keeps me going.
Britain’s best-known Iron Age site is undoubtedly Maiden Castle, the imposing Dorset hillfort just over 3 kilometres (2 miles) outside the county town of Dorchester. The decades after the last war saw the rapid rise of television and the creation of a new cluster of celebrities. Quiz shows were a popular feature of early post-war television and one of the best-known and most widely watched was Animal, Vegetable, Mineral? It was broadcast from 1952 until 1959 and by far the most popular of the panellists was the strikingly handsome archaeologist and soldier Sir Mortimer Wheeler. Wheeler was undoubtedly a remarkable man with a very strong – some would say domineering – personality, which was matched by his considerable archaeological achievements, both in Britain and across the territories of what was then still the British Empire. He also had a distinguished military career, which probably explains his impressive moustache. He was elected BBC Television Personality of the Year in 1954 and it’s fair to say that his influence still lives on.1
Wheeler’s best-known British excavation took place between 1934 and 1937 at Maiden Castle and the report appeared in 1943. It had been assembled and largely written during the darkest years of the war, when the author admits his mind was largely elsewhere, for entirely understandable reasons. There is also a tragic backstory to the project: the field team was run by three archaeologists, including Wheeler and his wife Tessa. Two years into the project, Tessa Verney Wheeler died and the volume is dedicated to her memory; she is now widely regarded as a pioneering female archaeologist.2 Her death meant that she wasn’t able to take part in the writing-up and the volume’s interpretations are very much framed within Wheeler’s world view, which is perhaps best described as militaristic and Empire-based. Tessa’s influence would undoubtedly have softened it. Having said that, he did publish it, quite promptly (by the standards of the time) and under extremely difficult circumstances.3 The report is certainly well organized and clearly written. Most importantly, and in common with all good excavation reports, it can be reinterpreted and reworked. In other words, no data have been removed that might undermine the explanations proposed by the author. It’s fair and balanced.
The big problem with Wheeler’s view of Maiden Castle was that he placed huge emphasis on a story that made good reading, and which appealed to the soldier side of his personality and background, but which can now be proved to have no basis in reality. As Wheeler saw it, the massive build-up of defences that happened later in the Iron Age was a direct response to the rise of the Roman Empire and the threat it posed to Britain. I can completely sympathize with that view, given the fact that in the late 1930s, while he was carrying out much of the later work in the report’s preparation, the Nazi threat to the peace of Europe was growing as Hitler’s regime became increasingly militaristic. By 1940, when Wheeler must have been writing or editing the final chapters, attacks on Britain had begun in earnest. So I don’t find it even slightly surprising that he made the climax of his Maiden Castle story centre on the growing threat from Rome and its conquest of Britannia, culminating in a final battle and the creation of a war cemetery within the hillfort’s eastern entrance.
With their enclosing ramparts of massive banks and deep ditches, hillforts are undoubtedly the most spectacular of prehistoric monuments and they dominate the landscapes around them. Maiden Castle is one of the most impressive examples, but there are others in Dorset that can rival it. Some 32 kilometres (20 miles) to the south-east, the imposing ramparts of Hambledon Hill – like those of Maiden Castle – completely surround and seem to take over the hill itself, while the spectacular so-called promontory fort at Hengistbury Head, on the coast near the Dorset/Hampshire border, would have been visible from a great distance, both inland and across the sea. Given the prominence of their locations, it is easy to see why these monuments demanded that epic histories, involving identifiable tribes and individuals, attach to them. Anonymous terms like ‘the Bronze Age’ or ‘Iron Age hillfort’ may be fine for archaeologists or prehistorians, but they hardly fire the popular imagination. To do that, you need names.
The earliest recorded British names appear in the writings of Roman authors, such as Tacitus. Tacitus (c. AD 56–c.120) was writing in the late first and early second centuries AD and one of his books was the Agricola, a biography of his father-in-law, who was governor of the Roman province of Britannia from AD 78 to 84.4 Tacitus gives us the names of the late Iron Age tribal kingdoms, many of which (but not all) fought t
o repel the Roman conquest of AD 43. The most famous tribe of all were the Iceni of East Anglia, who rebelled against the injustices of Roman rule in AD 60–1, under perhaps the best-known pre-Roman monarcha of the day, Queen Boudicca, or Boadicea, as she was known in Victorian and Edwardian times. A famous statue of her in her chariot (complete with revolving knives) marks the entrance to Parliament Square from Waterloo Bridge.5 The most celebrated late Iron Age kingdom of southern England was that of the Durotriges, who have been very closely identified with the enlargement of the Dorset hillforts at Hod Hill and Maiden Castle. This was the explanation favoured by Wheeler in his report on Maiden Castle, and by the distinguished Romanist and archaeologist Sir Ian Richmond, excavator of Hod Hill in the 1950s. Richmond was Professor of the Archaeology of the Roman Empire at Oxford University.6
It used to baffle me that the two leading archaeological authorities of the time, Wheeler and Richmond, should be so completely won over by the idea that hillforts were a response to an impending threat from the growing Roman Empire across the Channel. That, however, was before the Covid-19 pandemic. Now, having lived for many weeks under lockdown, I am beginning to understand how the appearance of a new, external threat can influence one’s judgement. Wheeler and Richmond lived through a brutal era in which mass slaughter had become routine. And it didn’t end with Victory in Europe Day in 1945: Stalinist repression continued in the Soviet Union and eastern Europe well into the 1950s and beyond. Britain must have appeared to be an island of peace and calm at the centre of a raging storm. So I can see why, in a Second World War and early Cold War context, these men placed so much emphasis on such a ruggedly pro-Ancient British and anti-Roman interpretation of two iconic hillforts. In theory, we should protest at the ‘distortion’ of objective truth in this way, but on the other hand we can never remove all subjectivity from a discipline of the humanities such as archaeology. The skewed interpretation of the hillforts provides us with a clear example of the extent to which everyone, including intelligent, highly educated people like Wheeler and Richmond, can be affected by the events of their time. Maybe that simple observation provides us with a lesson that is just as significant as the importance of objective truth. Interpretations, like the archaeological sites themselves, must always be seen in context.
There has always been a tendency in the world of archaeology to isolate and categorize different types of sites and finds. Hillforts were an early example of this trend. Some eighty were excavated in the 1920s and 1930s during a surge of research known by some as ‘hillfort mania’.7 The approach was overwhelmingly focused on hillforts as defensive military constructions. This view remained fashionable in the 1940s and then, after the war, in the 1950s and early 1960s. The researches of people like Wheeler and Richmond belonged to this general tradition; there was a great interest in hillfort defences and the details of their construction, which sometimes involved quite complex internal timber reinforcements.
The complex layout of hillforts’ defended entranceways, with their many blocking banks and other diversions, led to a lively debate on how they might have developed through time. These debates continued into the 1960s. It was all very interesting (although I must confess that as a student in the mid-1960s, the intricacies of hillfort defences never floated my boat even slightly) and it did advance the story to a certain extent, but while it continued we were losing sight of the more general questions, such as: what were hillforts used for? And were they all the same? To find answers to these and other problems, hillforts had to be examined in their settings. It was time to return them to their contexts; to see them as places inhabited and built by human beings, rather than military artefacts: the prehistoric equivalents of tanks or guided missiles.
*
One or two younger archaeologists working in the 1960s were keen to approach hillforts from such a broader perspective. Principal among them was Barry Cunliffe, whose long-term (1969–88) researches at Danebury, a hillfort on the edges of the Test valley, near Stockbridge in Hampshire, have had a profound influence on our understanding of these extraordinary places. He was able to demonstrate that the interior of the hillfort was very carefully laid out, with well-defined roads, often lined with houses or other buildings.8 I visited Danebury while it was being excavated and it undoubtedly exerted a big influence on the way my own researches were heading. But when it came to hillforts, I have always found there was something special about those spectacular ramparts at Maiden Castle. Maybe, too, my interest was heightened by the memory of Sir Mortimer – and by the association with the war. I don’t know why, but the place has always fascinated me.
In the mid-1980s I was frantically busy running excavations at Etton, while simultaneously starting work at Flag Fen, so I couldn’t find the time to visit some new research that was being organized at Maiden Castle by English Heritage.9 I knew many of the archaeologists involved and I was getting word that the project was producing amazing results. But it was all coming through to me second-hand, which was very frustrating. The fieldwork took place over two seasons, in 1985 and 1986, and readers will have noticed that I refer to it as ‘research’ or ‘fieldwork’ rather than the more usual ‘excavations’. Yes, excavations did take place and they were extraordinarily productive, but the whole project was very carefully planned from the outset and the dug trenches were just part of a much larger programme of work. I think history will judge this Maiden Castle project as one of the best planned and executed pieces of archaeological research of modern times.
Detailed field surveys were carried out after ploughing and included the plotting and collecting of thousands of pieces of pottery, flints and other artefacts. These clearly showed that the landscape around the hillfort had been settled from Neolithic times; this coincided well with the construction and occupation of a double-ditched causewayed enclosure that is located in the eastern half of the later hillfort and can be dated to the early third millennium BC. A distinctive kink in the layout of the massive ramparts in both the north and south sides of the hillfort shows where the Iron Age defences were slightly diverted to respect the ditches and banks of the earlier site.
The trenches were carefully positioned to avoid those of Sir Mortimer Wheeler – the idea being to build on and enhance what he had already revealed. This also meant that they could test and verify many of his conclusions. The dig was also planned and laid out in ways that amplified the experiences and research carried out by Barry Cunliffe at Danebury. The two projects were intended to be closely comparable, which indeed they were. The excavations at Danebury and Maiden Castle both showed clear evidence for a slightly smaller early phase in the fifth and sixth centuries BC, followed by massive enlargements and developments in the Middle Iron Age, from about 300 to 100 BC.10 Thereafter, the use of both sites is at best casual, or non-existent. So there is absolutely no evidence whatsoever for them being built to repel Roman invasions. If we transpose the Roman conquest to modern times, so that AD 43 becomes 1943 (coincidentally, the date Wheeler published his Maiden Castle report), then most hillforts would have been abandoned by 1800: several years before the Napoleonic battles of Trafalgar and Waterloo – a long time previously.
These early dates also suggest that most hillforts could never have been built by the famous pre-Roman tribal kingdoms, people like the Durotriges or the Iceni. Instead, we must seek their origins in much earlier times, possibly even in the final centuries of the Bronze Age, from around 800 BC. So could these early hillforts have arisen as a result of certain tribes becoming richer or just being militarily successful? That would be the obvious inference but the trouble is, it makes the assumption that all hillforts were essentially the same: regional defended centres for larger or smaller populations. It’s a nice, simple explanation, but sadly it’s also probably wrong.11 Later projects that investigated hillforts followed the lead of Cunliffe at Danebury and the English Heritage team at Maiden Castle: they shifted their gaze away from the massive ramparts to the large area of land they enclosed and to
the landscape that surrounded them. With a little help from modern science-based technology, these broader approaches have opened our eyes and provided some remarkably exciting new discoveries, which have transformed our understanding of life in Iron Age Britain.
*
We know about the existence of some thirty-five hillforts in the county of Dorset. About a third of them were examined in one way or another during the pre-war ‘hillfort mania’, but only four have so far been subject to proper archaeological excavation and even at Maiden Castle the area of the interior examined has been very small. This partly reflects the expense of modern research excavation, but it also demonstrates the legal protection provided by the Ancient Monuments Acts in action: these sites are safeguarded for the future and they must not be unnecessarily damaged – even by well-meaning archaeologists. One way round these problems is to use techniques that do not disturb the ground. Fans of Channel 4’s Time Team will recall that we always relied heavily on John Gater and his small team of ‘geofizz’b professionals, whose futuristic instruments used electronics and radar to look at what lay hidden below the surface. The hundreds of detailed surveys John’s team produced have proved of enduring value to archaeologists and are still regularly consulted today.
Since 2009, archaeologists and geophysicists based at Bournemouth University have been carrying out a detailed survey of prehistoric and Roman sites in Dorset, including about two-thirds of its known hillforts.12 It had traditionally been assumed that most of the large areas enclosed by hillforts remained essentially unoccupied: open areas of grass, perhaps used as grazing. These were, after all, sites placed on the tops of high hills and the prehistoric inhabitants, just like the geophysicists carrying out the survey, would have been frequently battered by strong winds and driving rain. Even in sunny Dorset, high hilltops are not the kindest of environments. But the geophysical plots showed a very different picture. Their surveys have been remarkably complete, revealing the extent of occupation across the large areas enclosed by the ramparts. Hod Hill provides a particularly clear example.
Scenes from Prehistoric Life Page 27