The Roman occupation of Britain, which began with the invasion of Kent in 43 A.D., took over a generation to reach as far north as the Tyne. Then, under the governorship of Agricola, the Romans advanced into the foothills of the Scottish highlands in the early ‘eighties. This campaign brought the Tyne-Tweed region under Roman control, and the area remained on the frontier of the Roman empire for the next 300 years. The forward line around the Highlands, with forts at Inchtuthil and Fendoch, was only held until about 87 A.D., but the Romans continued to hold the Lowlands until the end of the century.
Under Agricola a whole network of roads, supply-bases and forts were built. Across the Tyne-Solway gap the Stanegate road was constructed, with an important base at Corbridge at its east end. Dere Street, the main supply road to Scotland, was built north from York, through West Durham to Corbridge, then up the North Tyne and Rede valleys and over the Cheviots to the major Scottish base at Newstead. Northumbrian forts were established at High Rochester (Bremenium), north of Otterburn, at Blakehope, slightly to the south, and at Chew Green, high up on the border line. This last fort, a bleak posting for any Roman soldier, was probably to house a few men and horses to help convoys over this difficult stretch of road.
Another road was built from north of Corbridge, travelling north¬east towards Berwick through Ryal, Hartburn, and Bridge of Aln. The Devil’s Causeway, as the road is known, was partly intended to supply a base at Tweedmouth that was never completed, but was also designed to police the native population. In the late 1930s parts of the road were surveyed and excavated by R. P. Wright, revealing a road about 20 feet wide, with kerbstones, a central rib and a cambered surfacing of sandstone over well-prepared foundations. Some stretches, such as that over Rimside Moor to Edlingham, can still be easily followed across the landscape. Bridges on the road were probably wooden, and in 1834 the local antiquary John Hodgson noted that at Hartburn there was ‘a double row of square holes still remaining in the bed of the Hart for fixing it in’. A lateral road running from High Rochester to Bridge of Aln through Holystone and Callaly joined Dere Street to the Causeway. Close to the junction a fort was built at Low Learchild. This may be the Alauna of the Ravenna Cosmography, a gazetteer of the Roman world, and there may be another fort, referred to as Coccuveda in the Cosmography, still to be found on the Coquetdale Road.
After the Scottish lowlands were evacuated, the forts along the Stanegate were heavily garrisoned. To what extent Northumberland continued to be held is uncertain; the very limited pottery evidence from Learchild fort suggests early 2nd-century occupation, and it may be the present Border rather than the Tyne-Solway line that was the frontier. After 120 A.D., however, the frontier was very definitely drawn across the Tyne-Solway isthmus with the construction of Hadrian’s Wall.
The Wall, designed by Hadrian, and built between 122 and 128 is the best-known Roman remain in Britain. Originally designed to start at Pons Aelius (Newcastle), it was to run westwards through the Tyne gap, taking advantage of the fine, north-facing scarps of the Whin Sill that give the Wall its most exciting stretches, to Bowness on the Solway. Milecastles were planned for every 1620 yards, with the main garrisons behind the Wall, but easily called up, advancing beyond the Wall through the milecastle gates. During the building of the Wall, however, the design was altered. Construction had almost reached the North Tyne from the east when it was decided to bring the main forts and garrisons up onto the Wall itself. Where possible these forts projected beyond the Wall, the main side gates as well as the north gate leading to the outland, emphasising that the strategy was not to fight from the Wall itself, but to make sorties beyond it. Along the length of the Wall a large ditch, now known as the Vallum, was dug behind it in the middle of a 100-foot cleared strip of ground. This was only crossable at controlled causeways, and enabled a frontier or Wall zone to be defined. The Wall was built by the IInd, VIth and XXth legions, though the operational Wall was manned by auxiliaries from various parts of the Empire. The VIth legion came by sea from Lower Germany in 122, and two stone inscriptions set up to Neptune and Oceanus on their safe arrival have been dredged out of the Tyne near Newcastle Swing Bridge. In the hinterland behind the Wall, forts like Corbridge became redundant, whilst none of the forts beyond the Wall in Northumberland were occupied.
Although a tactical success, Hadrian’s Wall created strategic problems. Potential enemies were free to combine north of the Wall, whilst friendly groups, like the Votadini of Northumberland and East Lothian, were left without protection. The result was a re-advance into Scotland in the reign of Antoninus Pius. Dere Street was re-occupied, new building put up at Corbridge, and then a turf wall, the Antonine Wall, was constructed on the Forth-Clyde line in 142. On the Northumbrian stretch of Dere Street, High Rochester was re-commissioned. An inscribed stone from the fort, now at Ainwick Castle, records this rebuilding by the first mounted cohort of Lingones, a tribe from the East-Gaulish district of Langres. A new fort at Risingham (Habitancum) was built between Corbridge and High Rochester to guard the crossing of the river Rede. Chew Green was re-occupied, and the road fitted with milestones. One survivor, marking the twelfth Roman mile north from Corbridge, can now be seen on Waterfalls Hill to the east of the road. On Hadrian’s Wall itself, the forts were largely de-garrisoned, the Vallum breached to give easy access, and in the forts and milecastles the gates were torn down. In some milecastles the broken pivot stones can be traced.
The Antonine advance once more brought the Northumberland area within Roman occupation, and it is from this date onwards that Roman influences on native settlement come. The field survey and excavations of G. Jobey have shown that during the Roman period the native settlements of the Votadini became less defensive. The ramparts of the hillforts are often overlain by scattered huts, frequently stone-built, sited particularly on the more sheltered eastern slopes. A good example occurs at Lordenshaws near Rothbury. Other non-defensive settlements sprang up away from the hillforts, around the North Tyne and Rede valleys and in south-east Northumberland. A major village of at least 30 huts exists at Greaves Ash in the Breamish Valley. Roman objects did not reach these native sites in any bulk, but continued in a trickle, in some cases, as at Huckhoe, a few miles to the south, right through to the fourth century.
The two main types of Roman pottery found in excavations are coarse pottery, often made in Britain and usually only dateable by typology and association, and Samian pottery, a brilliant red-glazed pottery imported from Gaul. Samian was usually either decorated or, if plain, stamped with the potter’s name, so it is usually closely dateable. Mortaria or kitchen mixing-bowls were also by custom stamped by the potter, as with the Halton Chesters bowl stamped Camulacus fecit, ‘Camulacus made it’, an early Hadrianic piece probably made in kilns near Bedford. This detailed ceramic evidence is often crucial in determining the occupation history of sites: lack of Central Gaulish Samian ware was used by Pryce and Birley in the 1930s to confirm the Scottish evacuation around 100 A.D. Similarly pottery has been central to recent re-writing of the Antonine period in the north.
The Romans abandoned the Antonine Wall in about 155, possibly because of a revolt south of Hadrian’s Wall. In 157-158 many forts on Hadrian’s Wall were being re-commissioned, but in 159-160 the Antonine Wall was re-occupied and Hadrian’s Wall again largely evacuated, though left in working order. Two years later Calpurnius Agricola was sent to Britain to meet a threatening situation. Until 1972 the general view was that occupation of the Antonine Wall continued until about 180, but B. R. Hartley has used the Samian pottery finds to suggest a re-interpretation. He showed that the stamps of potters working in 140-160 were (as expected) absent from most of Hadrian’s Wall, but present on the Antonine Wall, whereas stamps definitely dateable to after 165 (for example those found in a shipwreck of brand-new Samian pottery on the Kent coast at Pudding Pan Rock) were absent from the Antonine Wall. This Wall was abandoned by Calpurnius Agricola in 163 in favour of Hadrian’s Wall.
This new picture creates its own problems. Newstead was held as an outpost until about 180. Presumably Dere Street was also held, but the 1930s excavations along it were not geared to this question, and there is a need to re-investigate Risingham, Chew Green and High Rochester to gain more pottery evidence. Also a major destruction at Corbridge and Halton Chesters, previously dated to 197, must now be dated to about 180. Yet the rebuilding by Severus in the first decade of the third century when he campaigned in the north following further enemy invasions, was over 20 years later. Did parts of the frontier lie derelict for this time?
After 211 there was a long period of relative peace. The Severan reconstruction revised the frontier strategy, and a strong emphasis was placed on outposts beyond the Wall. Artillery defences were built at High Rochester and Risingham. In addition, long-range scouts or exploratores toured the Cheviots area, reporting on hostile move¬ments, and keeping contact with friendly groups. At Sunnyside, south of Tweedmouth, pottery from a late third-century mortarium has been found, probably left by scouts patrolling up the Devil’s Causeway. In this period local recruiting to the auxiliaries became common, linking the soldiers to the local population; they were also allowed to marry and lease lands. This was the main age of civilian settlements or vici in the frontier zone. The highly developed urban life of southern Roman Britain had no place in the frontier zone, nor did the villa (though there was a villa at Old Durham), but around the military camps civilian settlements grew up, as at Corbridge, Chesterhoim and Housesteads. At Chesterhoim (Vindolan*da), north of Bardon Mill, there have been major excavations in the last nine years by Robin Birley. This vicus was self-governing, with shops, workshops, and a large 15-room courtyard house, probably designed for travelling Roman officials.
In 296 Roman troops were withdrawn to fight for the usurping Allectus, and the northern tribes took advantage to destroy as far south as York and Chester. There is, however, little dateable evidence after 276 and much of the destruction may date from then. The new Emperor Constantius, who defeated Allectus in the south, reacted swiftly, and the frontier was soon recontrolled and new building took place. But north of the Wall, at Risingham, for example, the vici were no longer inhabited. In 343 the outposts at Risingham and High Rochester were burnt down in an invasion not of British tribes, but of Picts from north of the Tay. In the excavations at Risingham the red calcined threshold to the west gate provided clear evidence of this burning. After this High Rochester was abandoned, but Risingham kept because the local iron-ore deposits weie needed for the military forges at Corbridge. Twenty-five years later, in 367, a ‘barbarian conspiracy’ of hostile groups, aided by treachery of the scouts, again destroyed the frontier.
After this incursion the frontier was restored by Theodosius, but it was no longer based on Roman military might. The remaining outposts were abandoned, and the Wall and forts rather chaotically restored by what were now soldier-settlers, not military garrisons. The vici contracted, though Corbridge and Vindolanda remained into the fifth century. A key element in the new policy was the relationship with treaty-states of the Votadini and other tribes beyond the Wall. There was no dramatic leaving-ceremony on the Roman Wall: by the time of formal Roman withdrawal from Britain in 410, the local inhabitants were dominant on the frontier. There was no collapse of order, and the Britons showed a capacity for organised policy, evidenced by the migration of a group from Manau, a treaty-state around what is now Clackmannan, to North Wales to drive out Irish invaders. But the Roman elements, particularly the urban life of the vici, did not last long in this thinly-Romanised frontier zone.