Compared to the number of surviving hillforts – there are almost as many on Toby Driver’s distribution map as there are settlements in modern Wales – only a tiny handful have been excavated. As a result, there is much that we do not know about the later prehistory of Wales, or, as Toby puts it, there are ‘many dark corners that require illumination, or the spark of imagination’ if we are to piece together the lives of our Iron Age ancestors.
Hillforts, along with promontory forts, are one of the three main types of later prehistoric settlement in Wales, the other two being small, defended enclosures and hut groups/field systems. The Historic Environment Records for Wales, accessed via the online portal Archwilio, currently list 764 hillforts, 1,191 defended enclosures, and 1,332 hut groups. In parts of Montgomeryshire and Pembrokeshire, there is a prehistoric fort or farmstead for every modern farm, hinting at the density of settlement, population size, and extent of land cultivation in the period from the earliest Iron Age (around 800 BC) to the early medieval period. Although Rome occupied parts of Wales from c.AD 70, the hillforts and farms of Wales changed little as a consequence, and some that had fallen out of use under Roman rule were reoccupied after AD 410, especially those on coastal sites.
Hillforts are categorised by the size and strength of their enclosing ramparts, and many have elaborate gateways, marked by outworks, hornworks, and annexes. Most occupy striking natural locations, often on prominent hilltops, ridges, or plateaus, though some were constructed on low-lying knolls, ridges, and hillocks. It is more than likely that these commanding sites were regarded as significant long before they were given the additional definition of hillfort ditches and banks. Finds of worked flints and stone axes at Breiddin and Ffridd Faldwyn in Montgomeryshire, at Pen Dinas in Ceredigion, and at many of the coastal promontory forts of Pembrokeshire indicate an importance for these locations at least two millennia before they became hillforts. Neolithic causewayed enclosures have been found underlying hillforts such as at Caerau, Cardiff, where one of the later gateways is aligned on the main eastern entrance of the Neolithic enclosure, which must therefore have been visible as an earthwork when Iron Age people began work there.
Bronze Age burial mounds and cairns were also frequently enclosed by later hillforts, to survive untouched among later roundhouses and ramparts, and thus clearly of continuing importance or reverence in Iron Age society. An upstanding barrow occupies a prominent position within Moel Fenlli, just south of the highest point in the Clwydian Range, Denbighshire, while mounds or stone burial cairns command the summits of Foel Trigarn in Pembrokeshire, Pen y Gaer at Llanbedr-y-Cennin in the Conwy Valley, Penycloddiau and Foel Fenlli in the Clwydian Range, Ysgyryd Fawr (the Skirrid hillfort) in Monmouthshire, and Caer y Twˆr on Holyhead Mountain.
Why did thr late Bronze Age and earliest Iron Age communities begin enclosing these places of ancient ritual and burial? The beginning of the 1st millennium BC was a time of change in Wales, and Britain in general, to a cooler and wetter climate. It is likely that the resulting competition for resources, especially in upland regions, created the need for defence. The development and rapid spread of iron tools and weapons to all levels of society must have had a profound effect on the ways in which elite members of any community had been used to establishing their status through the ownership and display of elaborately ornamented bronze objects. It is no coincidence that the hoarding and deposition of bronze objects suddenly ends in south Wales around 600 BC – bronze had ceased to be regarded as a sufficiently special gift for the chthonic deities or to be a precious material that needed to be hidden from view and secured against possible theft until needed.
Perhaps it is in this context, of changing status values and new pressures on territorial ownership, that we can find the reasons for the rise of a new and highly visible type of monument. The place of conspicuous metalwork was subsumed by a new form of power: the ‘power of place’, to use a phrase made popular by the publication by English Heritage of a report on the future of the historic environment in 2000.
For, in order to build on a prominent hill that probably had considerable spiritual significance for the people of the region, you must have been recognised by the community as having the right to make such an intervention. You must have been acknowledged to be the effective owner of the hill, with the spiritual authority to convince everyone that the heavens and/or ancestors will smile upon your appropriation of an ‘iconic’ location. And you must have had the unquestioned power to be able to command the necessary resources in terms of people, time, and tools.
Community consent was surely vital. Viewed like this, it is clear that hillforts were a symbol of regional power, whether that power was vested in a high-status individual, an elite family, a social class, the community as a whole, or a combination of all four. But as well as representing power, they gave power. From the highest summits, such as the striking Breiddin hillfort, which rises 300m from the floor of the Severn Valley to the north-east of Welshpool, in Powys, one could survey 100km of the surrounding countryside on a clear day, including the lands of the Cornovii to the east.
The builders of these high forts acquired considerable knowledge of the surrounding landscape that could be difficult to gain at ground level: a ‘map of the known world’, as Toby terms it. From these high places, they could observe other settlements and hillforts; watch over travellers, trade routes, and approaching visitors; keep a watchful eye on fields and livestock; and be warned about approaching weather systems. Hillforts overlooking rivers and seas could also observe waterborne traffic. Pen Dinas, on the coast at Aberystwyth, looks across Cardigan Bay to the Llyˆn Peninsula, 70km to the north-west, while several of the hillforts of south Wales take in views across the Severn Estuary to the lands of the Dumnonii in modern-day Devon and Cornwall.
Clues in the construction
Further clues as to the purpose of hillforts can be gleaned from imagining what was involved in their design and construction. Building a large hillfort was akin to the raising of a great medieval tithe barn or large church. In place of the daily agricultural cycle of waking, working, eating, and sleeping, taking part in hillfort construction was surely a novel and exciting activity, perhaps akin to the communal effort of the annual harvest, or kindling friendships, alliances, and romantic attachments, as well as enmities and rivalries.
Not all the work was carried out at the hillfort site. William Manning, in ‘The use of timber in Iron Age defences’ in Studia Celtica (1999), has calculated that a medium-sized 4ha hillfort required the builders to harvest 76ha of woodland and haul 4,680 trees all the way to the upland construction site. Large quantities of bronze and iron tools were required for converting timber to planks, posts, and stakes; the workers on site would have needed antler picks, buckets, baskets, ladders, and ropes. Stone for the revetment walls would also have to have been scoured from hill, field, or beach, or specially quarried. Specialist iron fittings were required for the hinges and closing mechanisms of the gate. Supplying adequate food and drink for all the workers was another important consideration.
Hillforts were scenes of constant activity, and without good organisation, chaos could reign on the construction site. There is good evidence from several excavations to show that gang-work was employed to organise the otherwise formidable challenge of breaking a big project into a series of smaller, manageable tasks. Chris Musson, in excavating the Breiddin hillfort in Montgomeryshire, detected clear subdivisions in the rampart walling where groups had worked in sections, and similar evidence can be seen at Caer Drewyn hillfort, Corwen, where different gangs used different sized blocks of stone. Harold Mytum, who excavated the hillfort at Castell Henllys in Pembrokeshire, suggests that various teams were involved in the digging of ditches, the scouring of hillsides for stone, spoil removal, and dragging sledges loaded with stone to the point where skilled wall-builders constructed each day’s stretch of rampart or wall.
Some of these walls and ramparts were so well-built that they have lasted two millennia. Considerable care was invested in the foundations to ensure that ramparts were structurally stable and internally robust. Sometimes the natural slope was sculpted into terraces before construction began, and trenches were dug and lined with clay or gravel to create a foundation layer. The builders incorporated natural features and rock outcrops to great effect, not only to take advantage of the strength and solidity of the natural bedrock, but also to give the impression of a wall that springs from the earth, where nature and artifice are hard to tell apart. Within the rampart, layers of clay, turf, gravel, or rock were used at intervals as levelling layers, and timber bracing was used within both the ramparts and walls to ensure structural stability.
Maximum care and attention were lavished on the gateway. This was the most important part of the defensive circuit – the focal point for anyone arriving and the culmination of everybody’s journey towards the hillfort. It is there that the most elaborate design features are always found, and this is the most frequently rebuilt and redesigned section of the hillfort’s defences. It was the critical point for conveying power, strength, and liminality – the sense of a transition was carefully choreographed, from the approach to the hillfort along a well-defined path to the narrow-walled corridor through which the interior private space was entered from the public realm.
In a typical Welsh hillfort, the gateway passage was dark, lined by massive stone slabs or drystone walls, and roofed in timber. At some hillforts, there were chambers to either side. Earlier interpretations of hillforts as military strongholds suggested that these were guard chambers, but new thinking about hillfort gateways proposes instead that these may have been ceremonial niches. Some entrances incorporated a dogleg, or sharp bend. At Tre’r Ceiri hillfort on Llyˆn, the passage is blocked by a stone pillar that forces visitors to step to the left. It seems that access into the hillfort was controlled and choreographed.
Like the castles of the medieval age, hillforts undoubtedly acted as controlling entities in the wider landscape, and as a garrison, perhaps, for those charged with exerting authority – hence the ‘fort’ element in the name. But recent commentators have sought to play down a purely military role on the grounds that many hillfort gateways and ramparts are, at the same time, over-designed for this purpose but also under-designed, with incomplete defences and unprotected ‘back doors’ that would make them vulnerable in the event of a sustained attack. Their purpose was, perhaps, to give the appearance of great defensive strength and, as A H A Hogg wrote in Hill-forts of Britain (1975), to look ‘so strong as to make any thought of attack appear hopeless’.
This is not to deny that conflict took place at some sites, nor that they served a defensive purpose from time to time, but that cannot be the whole story. Toby Driver’s book explores the evidence that hillforts were instrumental in creating community identity, binding people together in a common cause. This would account for the initial construction of a hillfort and the subsequent additions, modifications, and renewals, but also for its continuing relevance. Structures that exist to control and intimidate people tend not to last – on Crete, for example, the theory that Minoan culture collapsed because of volcanic eruptions just does not stand up: more likely is a palace revolt, motivated by resentment of the ruling elite’s control of olive oil and grain resources, and its habit of sacrificing children to the gods. For hillforts to be maintained and used for centuries suggests that they were widely valued, and Toby quotes Professor Robert Dodgshon’s paper in Celtic Chiefdom, Celtic State (1996) in which he compares hillfort communities to Highland clans. People drew meaning and relevance from being rooted in a particular area: without their land, they would become ‘broken clans’.
Thus the control of land and all its resources is key to understanding hillforts and Iron Age society. The more land the leaders and their community owned or controlled around the hillfort, the more secure the range of food they could access, from bread and meat from the lowland pastures and freshwater fish from river valleys to wildfowl and game from the high moorlands. Ownership might extend to the sources of metal ore and mining sites, for there is plenty of evidence of the use of hillforts and promontory forts for metal-processing in a safe environment, away from prying or acquisitive eyes. The point about security is well-illustrated by Akira Kurosawa’s 1954 epic Seven Samurai, in which a small rural community is threatened with starvation until they hire the eponymous samurai to protect them against the bandits who annually steal their rice harvest just as it is gathered in.
Toby’s distribution map of hillforts in Ceredigion reinforces this point by showing that hillforts are often located at sites that look like fringe locations, but that make sense when you realise that these are central points between various resources: open upland grazing, wooded valley slopes, good- or medium-quality arable land, an adequate water supply, rivers, intertidal zones (for salt, seaweed, and seafood), and sometimes mines, quarries, and places of coastal or riverine trade. A similar positioning of hillforts in relation to mixed resource zones has been noted in the Yorkshire Wolds, north-west Wales, and Pembrokeshire.
Question of purpose
Archaeologists have searched in vain for a catch-all explanation for what went on inside hillforts – what use people made of the spaces so elaborately enclosed and controlled. Like dwellings on a housing estate or a Victorian street terrace, superficial similarity disguises a range of different lifestyles. Some hillforts have evidence for dwellings so closely packed that the roof eaves must have touched, like the jettied timber buildings of a crowded medieval town. Initial excavation of Pen Dinas at Aberystwyth by the Dyfed Archaeological Trust in 2021 uncovered part of a crowded group of house platforms set just inside the south gate of the south fort, where each one was stepped up the steep, rocky slope so that the door of the house above would have opened to the roofline of the house below.
Another common feature of hillforts in Wales and elsewhere is the raised granary, set on four or six posts, which served the dual function of being a practical method of storing the community’s food surplus and of standing as a conspicuous display of wealth. But views vary as to whether this represents the entirety of the food supply, gathered there for distribution by a controlling elite; an emergency store as an insurance against times of scarcity or starvation; or a form of tax or tribute extracted from farmers of the surrounding landscape, like the tythes that medieval landowners paid to the church. All three are possible, and storage of grain in defended sites speaks of the constant threat of raiding, the risk of poor harvests, and the ways in which some Iron Age leaders ruled their territory and people in pre-Roman Wales.
There are simply no generalisations that work for every hillfort, and Toby notes that some of the most remote hilltop enclosures, such as Castell Rhyfel in the high Ceredigion moorlands above Tregaron, the Cefn Cil-Sanws scree enclosure high on a bare limestone ridge overlooking Merthyr Tydfil in southern Bannau Brycheiniog (Brecon Beacons), or the huge and well-built Garn Goch in Carmarthenshire, have empty interiors and are sited far away from good agricultural land. They must have been used for other purposes, such as seasonal herding or for gatherings at the junction of several remote mountain territories. Kate Waddington has theorised (The Settlements of Northwest Wales, 2013) that they acted as ‘focal points in the prehistoric landscape, where different communities met for festivals, feasting, markets, and to forge new alliances’, just like the causewayed enclosures during the Neolithic.
Toby’s book ends with suggestions for visiting some of the most impressive hillforts in Wales, from Caerau hillfort in the Ely suburb of Cardiff or Pen Dinas in Penparcau, south of Aberystwyth (both easily accessible urban sites that are the subject of Heritage Lottery-funded community excavations at the moment), to Pen y Crug hillfort, north-west of Brecon – which Toby calls a ‘masterpiece’, whose ‘steep ramparts and deep ditches present a baffling labyrinth to those who explore them’. Pen y Crug remains unexcavated, as are the majority of hillforts, and that is what makes the study of Iron Age Wales still so vibrant: every new archaeological find and every new excavation sheds fresh light on the intriguing subject of hillfort archaeology.
Further reading: Toby Driver, The Hillforts of Iron Age Wales (Logaston Press, £20, ISBN 978-1910839676).
All images: T Driver, unless otherwise stated