Enclosing an area the size of almost 8 football pitches, Caerau Hillfort is by far the largest hillfort in south Glamorgan, yet it has never before been researched. In fact, we know very little about hillforts in this region at all – even basic questions like how the settlements were organised inside, how long they were occupied, or even who lived in them remain largely unknown.
This year’s excavations at Caerau are just beginning to provide some answers to those questions. Although there’s still much work to be done analysing artefacts, examining environmental samples and sorting out dating and phasing of the site, project Co-director Olly Davis’ gives his first thoughts about what we found…
With over 1,000 visitors to the site and 150 people directly involved in the excavations the last four weeks have been some of the most incredibly rewarding, but very tiring, of my life!
This year we managed to open up three trenches to explore the lives of the people who lived on Caerau hill more than 2,000 years ago. Our trenches were concentrated into the south-eastern area of the hillfort. Trenches 4 and 5 were designed to explore ditch defined enclosures identified from the geophysical survey completed by local school pupils and Time Team last year. Trench 3 was much larger and intended to examine the Iron Age occupation of the hillfort. In total we opened up an area of 760 square metres. It doesn’t sound much, but to put our excavations in perspective, you could fit all the other excavated areas that have ever taken place within hillforts in south Glamorgan into our Trench 3, with some room to spare!
The Time Team excavations last year had revealed a small part of a well-defined ring-gully of a roundhouse and a small pit outside of its eastern extent had produced a remarkable carinated bowl dated to around 500BC. This year we opened Trench 3 in order to fully expose, characterise and understand the construction, use and abandonment of this house. We also wanted to recover more dating material and environmental remains so we could begin to build a picture of the food that Iron Age people ate, the animals that they kept and the types of crafts and other activities that they undertook.
However, rather than just one house, we actually uncovered the remains of three houses, all overlapping, so they couldn’t all have been standing at the same time. This was clearly an area of intensive occupation probably over a considerable period of time.
Two of the houses were defined by ring-gullies, and were large, almost 15m in diameter, presumably the homes of a large extended family of perhaps 10 people. The final house was post-built, but only part of this was revealed as it was sealed beneath a layer containing Romano-British pottery (even so, that is important as the house therefore must pre-date the Roman period, and suggests there may be an intact floor of this house remaining in place).
We only found the foundations of these houses, but what might the houses have looked like 2,500 years ago? Well, you may have seen a reconstructed Iron Age roundhouse if you’ve ever been to St Fagans or Castell Henllys. They were constructed of a circular ring of posts, in-filled with wattle and daub, and with pitched rafters holding up a conical thatched roof. But, they were far from just simple dwellings to keep out the wind and rain – they were the focus around which much of prehistoric life must have centred. It was here that people prepared and cooked food, entertained guests, undertook their daily tasks and crafts, slept, and occasionally sheltered their livestock. They must have been sensual places, dark places, smelly places – not least from the smoke from the hearth as there were no windows or smokeholes – and filled with furniture, pottery, stored food and the chattering voices of the occupants.
More work needs to be done analysing the pottery and recovering radiocarbon samples to know when these roundhouses at Caerau were built, but it’s likely that they were constructed in the Early or Middle Iron Age (500-100BC).
Forthcoming analysis of animal bones recovered from the ring-gullies surrounding the houses combined with wet sieving soil samples (a process using water to wash soil through very fine-mesh sieves to recover small seeds and plant remains) will also tell us more about the diets and everyday lives of these prehistoric occupants of Caerau hillfort.
An interesting benefit of the position of Trench 3 is that it allowed us to also explore the inner earthwork boundary of the hillfort. By placing a cutting through it we could see that it was constructed in two phases, separated by thick layer of soil which had been trampled and squashed by people when the second phase of rampart had been constructed. Underneath the earthwork was an occupation layer – it still needs to be dated, but it suggests that the hillfort wasn’t built on an unoccupied hill – when the boundaries of the hillfort were constructed people had already been living on Caerau hill, probably for some time.
These earthwork boundaries and their associated ditches defining the hillfort were unlikely to have been built just for defence. People think of hillforts as defensive structures, but our attention is turning to whether the people who lived here were actually developing a community or collective identity for themselves through the construction of these boundaries.
An exciting new discovery in Trench 3 was a metalled surface – a stone-built path or road – running concentrically with the hillfort boundary, laid probably in the 1st century AD or later. Was this built just to prevent people getting muddy feet when walking between occupation areas or did it help define areas within the settlement much like a modern road in a village or town? In either case, it certainly suggests some considerable organisation and planning of the settlement within the interior of the hillfort.
Trench 4 was positioned over the entrance into a roughly oval ditch-defined enclosure. Time Team had excavated one of the ditch terminals and suggested that the enclosure was a Roman livestock corral. Our excavations this year have completely transformed that interpretation. By excavating the opposing ditch terminal we recovered lots of domestic remains, including pottery, particularly black, bead-rimmed pottery, typical of the Late Iron Age (1st century AD) in south Wales. Rather than a cattle corral then this enclosure was more likely a Late Iron Age farmstead, set up in the hillfort perhaps 100 years after it had been abandoned. Two postholes either side of the entrance gap must have represented the posts of a gate into this enclosed area.
Outside the enclosure was a pear-shaped pit full of burnt stone and charcoal. Closer inspection of the pit fills revealed very small carbonised grains of wheat and barley suggesting that this may have been a corn-drying kiln for drying grain after harvest.
A glimpse of the personality of the inhabitants was provided by the discovery of a glass bead of Iron Age date and an enamelled disc brooch of Roman date. Both these objects indicate that individuals were concerned with their appearance and that life in the past was a little more colourful that we sometimes think. The glass bead needs further analysis, but is similar to types manufactured at Meare in Somerset, so may suggest links with communities living the other side of the Bristol Channel.
Trench 5 examined the ditch of the oval enclosure further around its circuit. Again, we recovered lots of Late Iron Age pottery confirming our assumption that it dates to the 1st century AD. That is potentially very significant because it suggests that we have people, perhaps only a single family albeit possibly an important one, living up on Caerau hill around the time of the Roman invasion. This is an important time in south-east Wales – the subjugation of the Welsh tribes, particularly the Silures, took about thirty years in the face of stubborn resistance and is well documented by the Roman author Tacitus. This oval enclosure therefore allows for the examination of interesting questions about power relations, Roman control and native-Roman acculturation in this region during the first century AD.
One of the other exciting things about this trench though is that we also uncovered a Roman story. Dating to the 2nd or 3rd centuries AD, probably contemporary with the occupation of the Roman villa in Trelai Park, we excavated a deep circular pit full of iron working debris and Roman pottery. The pottery is fascinating – the assemblage is dominated by coarse ware bowls and jars while there were no fine tablewares or amphorae. Fine-ware pottery and amphorae were found in abundance at Ely Roman villa by Mortimer Wheeler in his excavations during the 1920s. This difference in material culture at both the sites suggests that at Caerau at this time we either have a relatively impoverished settlement or one that is actively rejecting some parts of Roman culture.
Despite the many finds, much more work is required to understand exactly when the hill was first occupied and the boundaries built. A cursory examination of the pottery assemblage suggests occupation starting perhaps around 500BC and continuing until at least the third century AD, well into the Roman period. However, there appear to be gaps – there is little to suggest occupation from about 100BC until the 1st century AD when the oval enclosure was probably constructed. There is another gap in the ceramics from the later first to early second centuries AD. Question of continuity, discontinuity and legitimacy of the communities that lived here are hard to address, but it could be that the communities founding the oval farmstead and later Roman settlement were claiming ancestral links with the builders of the hillfort. One thing I am sure of however is that Caerau Hillfort still has many archaeological secrets yet to be revealed!