Building the hillforts

Although archaeology can rarely tell us exactly why a hill-fort was built in a particular location, or even who built it, it can usually reveal something of the way the fort was built, and how it developed over time. In addition we can draw on other archaeological evidence to improve our understanding of the people

Ditch Celtic
At Caburn, Sussex there is evidence that a second rampart and a broad but shallow ditch were added to the earlier defensive works in response to the Roman threat.Traces of both phases of construction can be seen here. (Courtesy of Steve Danes)

The hill-fort at Caburn, Sussex had only one entrance, sited on its north-eastern face. Archaeologists believe the mid-lst-century AD gateway consisted of a four-post box housing a recessed gate. The rampart itself was lined by a substantial palisade consisting of wooden posts set a foot apart, laced with horizontal braces and filled with chalk rubble. (Courtesy of Steve Danes)

The well-sited hill-fort on Beacon Hill, Hampshire enclosed an area of approximately 3.6 hectares. It was protected by a bank, a ditch and a counterscarp bank - pictured here as a path that now encircles the site.The traces of 20 roundhouses have been detected on its plateaulike enclosure. (Courtesy of Marcus Cowper)

The well-sited hill-fort on Beacon Hill, Hampshire enclosed an area of approximately 3.6 hectares. It was protected by a bank, a ditch and a counterscarp bank - pictured here as a path that now encircles the site.The traces of 20 roundhouses have been detected on its plateaulike enclosure. (Courtesy of Marcus Cowper)

who inhabited Britain during the Late Bronze Age and the Iron Age, and to show what tools and techniques they could apply to a major building project such as this.

The site of the fort seemed to be crucial. In some places hill-forts were built within sight of each other, which has led to the theory that they constituted an interconnected network of defences - akin to an Iron Age Maginot Line. While this notion has largely been discredited, it does appear that forts were built to dominate the immediate hinterland, usually on a hill overlooking an arable and populous valley. Whatever the reason for construction, these Iron Age 'engineers' certainly knew how to take advantage of the terrain. Almost without exception British Celtic hill-forts were sited to make the best possible use of the contours of the hill on which they were built. In other words the steepness of the slope on one or more sides of the position, the proximity of rivers and streams, or the visibility of any approaching force were all factors used to decide where the fort should be built.

The following account of the building process is gleaned from the archaeological investigation of Ladle Hill fort, Hampshire, which was excavated by Professor Piggott in 1931. What was unusual about the site was that it was never completed, and so traces of several phases of building could be identified. About a third of the 3.3-hectare (8-acre) site was already delineated by an existing boundary ditch before the construction work began, and this feature was utilized by the builders to help them. The rest of the circumference was delineated by a series of ditches, where the excavated earth was piled into mounds on the inner side of the ditch. Piggott detected the hand of at least 12 digging teams at work, but the project ended before the various teams could link their ditches together. In effect, Ladle Hill is a perfect example of an Iron Age building site.

The first stage must have been to mark out the near-circular course of the ditch. We know enough about Celtic society to imagine that some form of druidic religious ceremony would have been involved. In at least one hill-fort the remains of what might have been a sacrificial victim have been found, suggesting that the British Celts took the idea of honouring a deity very seriously. However, all our direct accounts of druidic practices come from non-Celtic or post-Roman Celtic sources, and the exact nature of these ceremonies is very much a matter of speculation. After the ceremonies came the hard labour - digging the ditch. As already noted, several teams began work at the same time in different parts of the perimeter. At Ladle Hill this first ditch was never completed, but the sections that were dug were of a fairly uniform size, approximately 1.5m deep and 3m wide.

Rather than simply dumping the soil on the inner face of the ditch to form the rampart, the teams carried their soil a few yards inside the fort, creating a series of temporary soil dumps. This allowed the walls to be built with some care: a base and inner revetment of large chalk boulders was created, and the rest of the soil was used as infill. A timber framework would have been constructed - essentially a wooden cage within which the rubble and soil would have been poured. The slope on the outer side of the bank would have been more substantial than on the inner face, so a stone or timber revetment would have been essential to keep the soil in place. This inner revetment was substantial, as it was also designed to reinforce the bank and prevent it from collapsing. This was not an easy task, given the simple tools available, and the construction of a revetment of this kind represented a feat of engineering that belied the idea that the builders lacked engineering skills. Amazingly, in sites where evidence of this timber framework has survived, there is no sign that nails were used; the structure was either slotted together, or tied into place. It also involved a lot of work. Evidently the idea of carrying the same load of soil twice was clearly considered a necessary sacrifice in order that the finished wall could be built in a methodical manner. At Ladle Hill the eastern portion of the rampart was built to what must have been its full planned height of around four metres, with a substantial chalk revetment bracing it. The bank was approximately seven metres wide, tapering on its outside face to a narrower platform of approximately one metre. A similar smaller section of rampart was completed on the fort's western side, but the two sections remained unconnected. It is almost as if the number of skilled wall-builders was smaller than the number of men required to dig and transport the soil - which was almost certainly the case. This suggests the presence of experienced supervisors.

Ladle Hill is not the only fort where the work remained unfinished. In several other cases improvements to existing fortifications show signs of having never been completed, particularly in the south of England. As many of these features have been dated to around the mid-lst century AD, these may well have represented emergency construction programmes begun as a response to the Roman invasion of Britain. The fact that they were unfinished may be explained by either the speed of the Roman advance through the region, or by the diversion of the labour force into mobile armies, charged with contesting the invasion. It is interesting that in most of these cases the work also appears to have been undertaken by several teams working simultaneously. The time taken to build a fort in this manner would obviously depend upon the size and availability of the workforce, or the size of a perceived threat to the community that undertook the work. In the case of larger hill-forts such as Maiden Castle construction of the perimeter defences could well have taken several years, particularly as the workforce almost certainly had to fit the work in between the seasonal demands imposed by arable farming during the Iron Age.

Once the rampart and ditch had been completed, the fort-builders would begin work on the palisade, which would surmount the bank. In most cases this involved the sinking of a line of timber posts about one metre apart around the outer edge of the rampart, then linking these with cross-braces. The structure would then be completed by filling the gaps between the posts with large stakes, thereby creating a solid palisade. We shall look into the nature of these timber defences in more detail, but at Ladle Hill the wooden perimeter would have been completed by the building of one or more gateways. As the fort was never finished and its perimeter never completely dug, we cannot be sure how many gates were planned to be incorporated into the Ladle Hill defences. However, if we look at the evidence from other forts such as Danebury in Hampshire we can imagine how these would have looked. Stout posts would have been dug between two inwardly turning spurs of rampart, and the palisade would have run up to these gateposts. For easier opening the wooden gate would have been built in two parts; these were secured to the posts using iron fittings.

The completed structure would have involved a substantial ditch - no doubt deeper and possibly wider than the preliminary ditch that was dug before work stopped on the site. In forts of similar size the ditch could be up to three or four metres deep. From there a steep glacis would rise up seven metres towards a wooden palisade, supported and braced using stout timber posts and beams. A narrow wall-walk behind the palisade would have provided a fighting platform, and from there the bank would have fallen sharply away towards the interior of the fort. A stone revetment would have prevented this bank from being eroded through use. We cannot tell if a more elaborate structure was planned at Ladle

The hill-fort of St. Catherine's Hill, Hampshire was excavated in the late 1920s, the first scientific hill-fort excavation in Britain.The oval perimeter encloses a gently sloping hilltop area of nine hectares.This main gateway shows signs of being hastily re-fortified, the builders creating an elaborate defensive system incorporating a chalk-walled approach and guardhouses.The hill-fort shows signs of being stormed in the mid-lst century AD, presumably by the Romans. (Courtesy of Marcus Cowper)

The hill-fort of St. Catherine's Hill, Hampshire was excavated in the late 1920s, the first scientific hill-fort excavation in Britain.The oval perimeter encloses a gently sloping hilltop area of nine hectares.This main gateway shows signs of being hastily re-fortified, the builders creating an elaborate defensive system incorporating a chalk-walled approach and guardhouses.The hill-fort shows signs of being stormed in the mid-lst century AD, presumably by the Romans. (Courtesy of Marcus Cowper)

left The defences of a hill-fort: Old Oswestry, 50 BC

This impressive hill-fort encloses an area of 5.3 hectares (13 acres), and was the subject of an excavation during the late 1930s. It appears the fort was built in three phases from around 600 BC until the arrival of the Romans in the mid-1 st century AD. During each phase an additional bank and ditch was added to the structure, although the final (outermost) phase of construction undertaken around 150 BC was by far the most extensive. During this last phase the gatehouse was strengthened with the addition of a series of banked enclosures or annexes on either side of it. Most probably these served as miniature fortified positions from which the defenders could fire upon any attacker who attempted to reach the gateway via the walkway bridge. Like most hill-forts, the principal line of defence was the innermost bank or rampart - the outer banks and ditches would have served to slow down an attacker, or even to keep him beyond slingshot range of the inner defences.The weak point of any hill-fort was its wooden gateway, so this was where the defenders concentrated their efforts, developing 'killing grounds' which covered all approaches to the vulnerable gateway.

Hill, where defensive works outside the gateway or additional counterscarp ditches would have been added. What we can calculate is the time taken for the builders to complete the structure they began work on.

Had it been completed, the hill-fort at Ladle Hill would have involved the transportation of some 11,000m5 of soil, boulder and rubble dug from the ditch; in addition, if suitable stones could not be found these would have had to be moved to the site from elsewhere. Once the material from the ditch had been graded and sorted it would have had to be moved again, this time to the site of the rampart. If we apply the construction yardstick where a man can dig and transport one cubic metre of soil per day (given the simple tools available), and assuming that half as much time again would be needed for the subsequent re-transportation of the soil at the rampart site, this means the digging and soil moving alone would have taken around 16,500 man days of work. Included in this total is the time taken to build the revetment. Given the perimeter of the bank would have been about 680m long, we can assume the builders would have needed about 1,000 small trees to provide the posts, braces and stakes needed to build the palisade and gateway. We can add another 500 man days for this work, giving a rough total of 17,000 man days. Archaeologists have determined that for each hectare enclosed within a fort the defences could house and protect approximately 60 people. That gives us a workforce of 200, a total that would include women and children (who would have been used to transport the soil in baskets). If we reduce the workforce by a quarter to include the lesser contribution of the elderly and the children, we are left with a total project time of approximately 112 days, or roughly four months.

The hill-fort on Beacon Hill, Hampshire was served by only one entrance sited on its southwestern face, consisting of bank and rampart that turned back on itself to house a simple gate.This entrance was mirrored by a similar feature on the counterscarp bank, while the entrance was further protected by an additional smaller semi-circular bank and ditch beyond the counterscarp. (Courtesy of Marcus Cowper)

The Scottish Border hill-fort of Woden Law, Roxburgh originally consisted of a single stone bank and ditch, sited on the edge of a precipitous slope. During the 1st century AD an additional perimeter of two banks and corresponding ditches were added to the fortification. Archaeologists are still debating whether the Roman siegeworks located close to the fort were used in anger, or rather served as a training ground. (RCAHMS)

The same equation can be extended to include other similar hill-forts, including Maiden Castle. The only difference would have been that in several of these sites the defences were built in several phases, often a century or more apart. These earlier works were often incorporated into the new design, which reduced the amount of work involved. However, in its final phase Maiden Castle was protected by no fewer than three banks and a counterscarp bank, and three ditches. In addition the gatehouse defences were significantly grander than those that would have been built at Ladle Hill. The hill-fort historian A.H.A. Hogg estimated that in time of peace the defences of Maiden Castle could have been improved from one building phase to another in a period of approximately five to seven months. As this would have involved an unacceptable degree of disruption to the farming life of the community, he argues that the work would have been spread out over two or even three seasons. In many ways Ladle Castle was the ideal size - a smaller hill-fort would almost certainly have served a smaller population, which meant the work involved would have taken longer. Here again the work would have imposed on the agrarian calendar, and so would probably have been undertaken over at least two years.

0 0

Responses

  • ileana
    How to build a fortified fort?
    7 years ago
  • Tony
    How to build a celtic hill fort?
    7 years ago
  • Ross
    How did the celtic the celtic build the hill forts?
    7 years ago
  • mathias bumgarner
    How the celts built hill forts?
    7 years ago
  • maura
    How to build a hillfort?
    7 years ago

Post a comment