Flag Fen

The jumble of timbers found on the floor of the Bronze Age hall being excavated at Flag Fen (Cambs). Shaped and jointed timbers can be seen in the photograph. (Francis Pryor)

Reconstruction of the late Bronze Age settlement at Springfield Lyons, Essex. (Tracey Croft)

On the edge of the Cambridgeshire Fens near Peterborough lies Flag Fen, an artificial island built about 1000 BC and made up of more than a million timbers. It is connected to the mainland by a ritual avenue of posts 820 m. long. It is likely that a number of buildings stood on the island, but at present only one large rectangular hall 6.5 m. (7.1 yd) wide and at least 20 m. (22 yd) long has been uncovered. It was constructed with three aisles, and posts supporting a thatched roof. Many of the timbers used in building it have survived, due to waterlogging, although they now lie in a jumbled confusion. Conspicuous, lying just off the shore in a lake of open water into which offerings were dropped, the site was one of prestige, perhaps a ceremonial centre, built by people who commanded respect and could control workmen.

There is however an alternative explanation for Flag Fen. Its position on an island might be for reasons of defence; further excavation is required to resolve this issue. We have noticed the great increase in military weapons, particularly swords in the south-east and elaborate spearheads elsewhere. There have also been hints that society was becoming stratified with peasants and farming groups, perhaps to be interpreted as workers and landowners with rich metal distributors somewhere in the picture. Weapons may have existed only as a prestigious deterrent, but it is most probable that they meant conflict between groups possibly disputing land ownership. The appearance of linear dykes to separate territories hints at land division between tribal groups.

At the same time, around 1000 BC, the climate was changing. As it became cooler and wetter areas of upland like Dartmoor and the North York Moors became impossible to farm due to the growth of blanket bog. The Dartmoor boundary reaves and Deverel Rimbury settlements seem to have been abandoned, so too was the marginal land, and there were new incursions into the chalk lands of the south and east. In northern Britain another natural catastrophe overwhelmed many impoverished settlements. Climatologists have recorded the volcanic activity of Mount Hekla in Iceland about 1159 BC. Prolonged clouds of volcanic dust blocked out the sun and caused low pressure and temperature, resulting in extremely high rainfall and cold weather all over Scotland and northern England. This led to an exodus of highland folk to the south. Settlements at Strath of Kildonan in Caithness were suddenly deserted, as were similar sites on the island of Noth Uist.

In Roxburghshire open settlements on Eildon Hill North were fortified with timber palisades for the first time, and at Broxmouth near Dunbar in East Lothian a large, circular wooden house was enclosed by a strong fence and guarded by dogs. Close by at Dryburn Bridge the same need for defence was observed, and this was repeated many times over in the Cheviot Hills of Northumbria. Hownam Rings in Roxburgh has become the type-site for this sudden enclosure. Over many years it was defended by a succession of two palisades, a stone wall and eventually multiple earthen ramparts. Inhabitants of these enclosures seem to have been pastoralists who found their new defences adequate protection for their animals.

Together, the effect of tribal grouping and environmental stress in the uplands due to climatic deterioration, aggravated by the effects of volcanic activity, seems to have been folk migration from the north which caused pressure on land in southern Scotland. Ripples spread out across England leading to an increase in population and heavy demand on resources. This in turn led to the enclosing of many settlements that were formerly open, and the first attempts at fortification and the protection of herds and flocks against human predators.

Major changes were taking place in the countryside. The old familiar rectangular fields of Wessex, sometimes called Celtic fields, were beginning to disappear. Sinuous linear dykes, with deep V-shaped ditches and often with banks on either side, perhaps planted with hedges, started to straddle the countryside, often running for several kilometres. Some included earlier fields within their bounds but often they cut across them, suggesting that the old arable plots were being replaced by sweeping ranches. The new boundaries usually respected the barrows that dotted the landscape, and often seem to have been aligned on them. Similar dykes were appearing elsewhere. Extensive systems are known on the Yorkshire Wolds, the North York Moors and in the Midlands. Shorter cross-ridge dykes found on the chalk of Wessex, Sussex, Berkshire and the Chilterns probably relate to this same period.

Ann Ellison has suggested that certain major palisaded settlement sites which appear at the beginning of the first millennium fulfilled the function of exchange centres for metalwork and fine pottery. It is not clear how they would have operated but an examination of materials, especially pottery, from the sites suggest that each was situated on a boundary between style zones and acted as a communal meeting-place or market centre. Suggested centres were Rams Hill in Oxfordshire, Norton Fitzwarren in Somerset, Highdown Hill in Sussex and Martin Down in Dorset. With more investigation others may be expected on Dartmoor, in Kent and the Thames Basin.

A group of rather specialized fortified settlements of the late Bronze Age, around 1000 BC are known. One of the best examples has been excavated at Springfield Lyons in Essex, where a circular area 65 m. (71 yd) in diameter was strongly defended by a ditch 1.5 m. (1.6 yd) deep, broken by six causeways, and a bank topped by a timber palisade. Inside were three houses, the main one with an elaborate entrance porch facing the enclosure entrance. Another hut may have been a workshop—the site produced an important find of clay moulds for bronze casting. Similar sites have been excavated nearby at Mucking North and South Rings, and at Thwing in Yorkshire.

At Thwing a chalk bank was piled up inside a massive outer ditch which was more than 100 m. (109 yd) in diameter and 3 m. (3.3 yd) deep. Inside the bank were the paired post holes of a box-type rampart. At the centre of the enclosure was an enormous building 28 m. (30.6 yd) in diameter, with an internal ring of posts 19 m. (21 yd) in diameter which would have supported a roof. At the centre a cremation was found in an urn—a dedication perhaps? The building also produced many domestic items including a saddle quern, loom weights and spindle whorls, personal jewellery and weapons, and lots of pottery. Animal bones included those of cattle, pig, sheep, horse and deer.

Hilltop settlements were in existence by 1000 BC that mark the advent of the hillforts and are found initially in the highland zone. Radiocarbon dates show that Dinorben (Clywd) 1032 bc, Mam Tor (Derby) 1274 bc and Grimthorpe (Yorks) 1058 bc were already established, often as open settlements soon to be fortified. Metalwork from Ivinghoe Beacon (Bucks) suggests a late ninth-century BC date for a lowland site.