|ISSN 1357-4442||Editor: Simon Denison|
Frontier territory along the Thames
The Thames valley today is the epitome of a peaceful landscape. It was not always so, writes George Limbrick
Today we think of Oxford as lying at the heart of a homogeneous geographical region formed by the Thames valley, but for much of the last few thousand years this part of England was an area of intense political and economic rivalry and territorial dispute. In the Iron Age, and perhaps earlier, the River Thames and its northern tributary the Cherwell appear to have served as political, tribal boundaries. The story was to be repeated in Anglo-Saxon times when the Thames again divided kingdom from kingdom, Mercia to the north and Wessex to the south.
The political map of the Iron Age is complex, but distributions of coinage for this area suggest that the Catuvellauni occupied the area to the east of the Thames / Cherwell, the Atrebates to the south of the Thames, and the Dobunni to the west (with a possible `sub-Dobunni' group to the south-west). The section of the Thames extending from the Cherwell downstream to Wallingford represents the area where all three tribal powers met.
An overall sense of this historical geography has been understood for years; but the picture became much clearer recently when a massive, 33-hectare, late Iron Age defensive enclosure, or oppidum, was discovered underlying the town centre of Abingdon, south of Oxford, during excavations in 1990-91 by Tim Allen of the Oxford Archaeological Unit. The oppidum was enclosed by a substantial multiple-ditched defensive circuit, and by projecting the line of the defences it appeared they cut off the angle between the Thames and its tributary the Ock, which together acted as its other defensive sides. The oppidum was established on a site already occupied in the early to middle Iron Age, and after its main period of use it continued in occupation throughout the Roman period.
Most importantly, however, the oppidum joins a series of four or five major earthworks of about the same period - roughly 100BC to the time of the Roman conquest - in this short stretch of the Thames valley. Together they amount to one of the most significant concentrations of late Iron Age and early Roman defensive or territorial earthworks in Britain.
At the northern end of this political and economic frontier zone, near the modern towns of Woodstock and Charlbury, in what was probably the Iron Age territory of the Dobunni, there is the massive North Oxfordshire Grim's Ditch, an earthwork enclosing an area of initially about 13km2 and eventually 80km2, within which a notable cluster of early Roman villas later appeared. No-one knows exactly what this enclosure was for; but it may have been intended to demarcate a special-status area for settlement. To the south, a much smaller, possibly unfinished defensive enclosure of about ten hectares, known as Cassington Big Ring, lay on the banks of the Thames near Eynsham.
Next in the sequence going south is Abingdon in the territory of the sub-Dobunni or the Atrebates, and then the Dyke Hills oppidum at Dorchester, another very large, 40- hectare defensive enclosure, very similar in form to Abingdon, in what might have been Catuvellaunian territory. This had two substantial banks either side of a large ditch, probably an artificial channel linking the Thames to its tributary the Thame to form a defensive moat.
Finally, near Wallingford, and least well dated, was the South Oxfordshire Grim's Ditch, a substantial linear earthwork across the west end of the Chilterns, apparently cutting off a huge loop of the Thames between Wallingford and Henley.
We do not know enough about the ebb and flow of each tribe's political and economic fortunes to be certain of the tribal allegiance of each of these earthworks. However, what we do see is a string of major enclosures and linear earthwork systems that may have been used as trading entrepôts, bridgeheads or strongholds along the Thames.
The impression of a prehistoric political rivalry and economic interchange is further indicated by many of the non-coin finds from the Thames gravels. Salt, for example, was brought into the Upper Thames valley from Droitwich but appears not to have reached south of the river, while in the territory south of the river, Abingdon was receiving salt containers, querns, brooch styles and many other artefacts from Hampshire and Sussex. There are also possible differences in fineware pottery styles either side of the Thames. Apart from the briquetage, so far the differences revealed by finds distributions either side of the river are not mutually exclusive, but nevertheless are noticeable.
Another indication that the Thames may have been a political boundary in prehistoric times is the quantity of late Bronze Age and Iron Age weaponry deposited in the river. Usually this metalwork is thought to represent sacrificial or symbolic offerings of wealth; but it is interesting that swords were often ritually `killed' by being deliberately twisted into unusable forms before they were deposited. It is also clear that some of the weaponry had been used in earnest - a bronze shield from Dorchester, for example, had been punctured by a Bronze Age socketed spearhead, while a human pelvis was found from the same part of the river with a Bronze Age spearhead embedded in it.
If the Thames was already a long-standing tribal boundary by the Iron Age, the deliberate deposition of weaponry in the river - which continued through the Iron Age - may have had political as well as religious and social meaning. Whether deposited individually, possibly with bodies, or as groups, the weapons may have conveyed messages of peace, or perhaps the opposite - aggressive displays of power.
The pervasiveness of Roman culture and administration eventually suppressed any sign of the pre-existing tribal politics. But it is beginning to become clear that this was not sudden and may have varied across the region. It is clear that many Iron Age customs survived well into the 2nd century in some rural parts of the Thames valley at places like Stanton Harcourt. It is notable that the three late Iron Age defensive enclosures at Cassington, Abingdon and Dyke Hills had very different fortunes around the time of the conquest, possibly reflecting differences in the political standing of their tribal allegiances with the new Roman administration: Cassington never really took off at all, perhaps overshadowed by whatever was happening within the North Oxfordshire Grim's Ditch; Abingdon continued to flourish within its defensive circuit; while Dyke Hills appears to have been replaced by the Roman town of Dorchester.
After the demise of Roman rule, this same overall area was one of the earliest regions to be settled by Anglo-Saxons, forming a fairly coherent social and political tribal grouping called the Gewisse in the 5th and 6th centuries AD, whose influence later shifted south to become better known as Wessex. In the middle to late Saxon period the Upper Thames valley once again became an area of political and territorial rivalry as a result of Mercian expansion around AD630. Territorial control fluctuated somewhat during the 7th and 8th centuries, with the 7th century Mercian king, Offa, pushing south of the Thames to occupy much of the Vale of White Horse. In the early 9th century the Thames was once again firmly established as a political boundary as a result of the resurgence of Wessex. Oxford itself originated as a crossing point or bridgehead close to the minster shrine of St Frideswide, and may well have started as a Mercian settlement, but became one of a string of fortified Wessex towns along the Thames, with others at Cricklade and Wallingford.
It is intriguing how the probable extent of Mercian penetration south of the Thames mirrors the extent of the distribution of coins of Cunobelin, the Catuvellaunian king, who may have had ambitions in the same area some 750 years earlier. Another parallel worthy of note is the re-occurrence in this period of weaponry (swords and seaxes of late Saxon and Viking origin) found in the Thames between Oxford and Reading - again with several pieces ritually `killed' by bending.
The prehistoric and Saxon political boundary of the Thames survived for millennia, latterly as the old boundary between Berkshire and Oxfordshire. The county reorganisation of 1974, however, finally gave official recognition to the more recent perception of the Upper Thames valley as a homogeneous region.
George Lambrick is the Deputy Director of the Oxford Archaeological Unit
Return to the British Archaeology homepage
Return to the CBA homepage
© Council for British Archaeology, 1998